The Oxford Handbook Of Bioethics - PDF Free Download (2025)

the oxford handbook of

BIOETHICS

the oxford handbook of ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

BIOETHICS ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Edited by

B ONNIE STEINB O CK

1

1

Great Clarendon Street, Oxford OX2 6DP Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offices in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York © the several contributors 2007 The moral rights of the author have been asserted Database right Oxford University Press (maker) First published 2007 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this book in any other binding or cover and you must impose the same condition on any acquirer British Library Cataloguing in Publication Data Data available Library of Congress Cataloging in Publication Data Data available Typeset by Laserwords Private Limited, Chennai, India Printed in Great Britain on acid-free paper by Biddles Ltd., King’s Lynn, Norfolk ISBN 978–0–19–927335–5 1 3 5 7 9 10 8 6 4 2

To Lindy Linder, my first cousin and first best friend

Ackn ow l e d g men ts

••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

This book is the brainchild of Peter Momtchiloff, who asked me to edit it many years ago, and refused to take ‘no’ for an answer. While I remain uncertain as to whether I owe him ‘thanks’ for this, I must acknowledge that the book would not have come into being, but for his persistence. I am grateful to all my contributors, especially those who actually handed in their work on time, or at least within a year of when they said they would get it in. (The delays led to John Arras suggesting that the contributors could be paid, not in cash or Oxford books, but death benefits, or possibly, assisted-living expenses.) I could not have finished this project without the painstaking work of Juliette Stevens. Juliette, you’re wonderful. Thanks also to Darren Anderson for putting the manuscript on disk and putting together the index, saving me a lot of time and my sanity. Finally, I had the good fortune to have the best copy editor in the world: Laurien Berkeley. I am deeply grateful for her careful attention and unfailing good humor.

Contents

•••••••••••••••••••••••••••••••••••••••

Notes on the Contributors Introduction Bonnie Steinbock

xiii 1

PART I: THEORETICAL AND METHOD OLO GICAL ISSUES 1. Methods in Bioethics James F. Childress 2. The Way We Reason Now: Reflective Equilibrium in Bioethics John D. Arras

15

46

3. Autonomy Bruce Jennings

72

4. Mental Disorder, Moral Agency, and the Self Jeanette Kennett

90

5. ‘Reinventing’ the Rule of Double Effect Daniel P. Sulmasy

114

PART II: JUSTICE AND POLICY 6. Policy-Making in Pluralistic Societies Søren Holm 7. Tiers Without Tears: The Ethics of a Two-Tier Health Care System Benjamin J. Krohmal and Ezekiel J. Emanuel

153

175

x

contents

8. Justice and the Elderly Dennis McKerlie

190

PART III: B ODIES AND B ODILY PARTS 9. Organ Transplantation Ronald Munson 10. Biobanking Louise Irving and John Harris 11. For Dignity or Money: Feminists on the Commodification of Women’s Reproductive Labour Carolyn McLeod

211 240

258

PART IV: THE END OF LIFE 12. The Definition of Death Stuart J. Youngner

285

13. The Aging Society and the Expansion of Senility: Biotechnological and Treatment Goals Stephen G. Post

304

14. Death Is a Punch in the Jaw: Life-Extension and its Discontents Felicia Nimue Ackerman

324

15. Precedent Autonomy, Advance Directives, and End-of-Life Care John K. Davis

349

16. Physician-Assisted Death: The State of the Debate Gerald Dworkin

375

PART V: REPRODUCTION AND CLONING 17. Abortion Revisited Don Marquis

395

contents

18. Moral Status, Moral Value, and Human Embryos: Implications for Stem Cell Research Bonnie Steinbock 19. Therapeutic Cloning: Politics and Policy Andrea Bonnicksen

xi

416 441

PART VI: GENETICS AND ENHANCEMENT 20. Population Genetic Research and Screening: Conceptual and Ethical Issues Eric T. Juengst 21. Enhancement Thomas H. Murray 22. Genetic Interventions and the Ethics of Enhancement of Human Beings Julian Savulescu 23. Pharmacogenomics: Ethical and Regulatory Issues Matthew DeCamp and Allen Buchanan

471 491

516 536

PART VII: RESEARCH ETHICS 24. Clinical Equipoise: Foundational Requirement or Fundamental Error? Alex John London

571

25. Research on Cognitively Impaired Adults Jason Karlawish

597

26. Research in Developing Countries Florencia Luna

621

27. Animal Experimentation Alastair Norcross

648

PART VIII: PUBLIC AND GLOBAL HEALTH 28. The Implications of Public Health for Bioethics Jeffrey Kahn and Anna Mastroianni

671

xii

contents

29. Global Health Ruth Macklin

696

30. Bioethics and Bioterrorism Jonathan D. Moreno

721

Index

735

Notes on the Cont r ibu tors

•••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Felicia Nimue Ackerman is Professor of Philosophy at Brown University. Her essays on bioethics have appeared in the Hastings Center Report, Midwest Studies in Philosophy, Physician-Assisted Suicide: Expanding the Debate (Routledge, 1998), Ethical Issues in Modern Medicine (McGraw-Hill, 2002), and elsewhere. Her short stories on bioethical themes have appeared in Commentary, Mid-American Review, Prize Stories 1990: The O. Henry Awards (Doubleday, 1990), Clones and Clones: Facts and Fantasies About Human Cloning (Norton, 1998), and elsewhere. She is writing Bioethics Through Fiction, a book of essays and short stories, forthcoming in the Rowman & Littlefield series Explorations in Bioethics and the Medical Humanities. John D. Arras is Porterfield Professor of Biomedical Ethics and Professor of Philosophy at the University of Virginia, and a Fellow of the Hastings Center, Garrison, NY. His primary areas of interest currently include research ethics (with a focus on international trials), global justice and human rights, and methods of practical ethics. He consults regularly with the Centers for Disease Control and Prevention, Atlanta, Georgia, and the National Institutes of Health on ethical issues in research and public health practice. Andrea Bonnicksen teaches courses in biomedical and biotechnology policy in the Department of Political Science at Northern Illinois University, where she is Presidential Research Professor. She is the author of three books, including Crafting a Cloning Policy: From Dolly to Stem Cells (Georgetown University Press, 2002). She is co-chair of the Ethics Committee of the American Society for Reproductive Medicine. Allen Buchanan is James B. Duke Professor of Philosophy and Public Policy at Duke University. Buchanan’s work is mainly in Bioethics and in Political Philosophy. His most recent books are From Chance to Choice: Genetics and Justice (co-authored with Dan W. Brock, Norman Daniels, and Daniel Wikler; Cambridge University Press, 2000) and Justice, Legitimacy, and Self-Determination: Moral Foundations for International Law (Oxford University Press, 2003). James F. Childress is the John Allen Hollingsworth Professor of Ethics at the University of Virginia, where he teaches in the in the Department of Religious Studies and in the Schools of Law and Medicine, and directs the Institute for Practical Ethics and Public Life. He is the author of numerous articles and several

xiv

notes on the contributors

books in bioethics, including (with Tom L. Beauchamp), Principles of Biomedical Ethics, (5th edn., Oxford University Press, 2001). He has served on a number of governmental bodies addressing bioethics and public policy, including the National Bioethics Advisory Commission 1996–2001. John K. Davis is an Assistant Professor of Philosophy at the University of Tennessee, Knoxville. His research and teaching interests include bioethics, ethical theory, and philosophy of law. In addition to autonomy and end-of-life care, he is interested in the methodology and epistemology of applying ethics to cases. Matthew DeCamp entered the Duke University Medical Scientist Training Program (MD/Ph.D.) in 2000 and is currently pursuing graduate work in the Department of Philosophy. His current research focuses on the relationship between intellectual property rules and global distributive justice in global health. Gerald Dworkin is Professor in the Philosophy Department at the University of California at Davis. He is also an adjunct Professor in the School of Law. He is the author of many articles and books in moral, political, and legal philosophy. His most recent book (co-authored with Sissela Bok and R. G. Frey) is Euthanasia and Physician-Assisted Suicide: For and Against (Cambridge University Press, 1998). Ezekiel J. Emanuel is the Chair of the Department of Clinical Bioethics at the Clinical Center of the National Institutes of Health. He is a Fellow of the Institute of Medicine, the Association of American Physicians, as well as the Hastings Center, Garrison, NY. He is author of The Ends of Human Life: Medical Ethics in a Liberal Polity (Harvard University Press, 1991) and No Margin, No Mission: Health-Care Organizations and the Quest for Ethical Excellence (Oxford University Press, 2003), and co-editor (with Robert A. Crouch, John D. Arras, Jonathan D. Moreno, and Christine Grady) of Ethical and Regulatory Aspects of Clinical Research: Readings and Commentary (Johns Hopkins University Press, 2003). John Harris is Sir David Alliance Professor of Bioethics at the Institute of Medicine, Law, and Bioethics, University of Manchester. In 2001 he was the first philosopher to have been elected a Fellow of the Academy of Medical Sciences. He has been a member of the Human Genetics Commission since its foundation in 1999. The author or editor of fourteen books and over 150 papers, his recent books include Bioethics (Oxford University Press, 2001), A Companion to Genetics: Philosophy and the Genetic Revolution, co-edited with Justine Burley (Blackwell, 2002), and On Cloning (Routledge, 2004). Søren Holm is Professorial Fellow in Bioethics at Cardiff Law School and Director of the Cardiff Centre for Ethics, Law, and Society. He is also adjunct Professor of Medical Ethics in the Section for Medical Ethics, University of Oslo. He has written on many subjects in bioethics and the philosophy of medicine, and his most recent publications have been on biobanking and on stem-cell research.

notes on the contributors

xv

Louise Irving was formerly at the Institute of Medicine, Law, and Bioethics at the University of Manchester, where she was a research fellow on a three-year project, funded by the European Commission, which seeks to develop a legal and ethical framework for stem-cell research. She has written on the commodification of the human body, the relationship between analytic moral philosophy and bioethics, and issues of public health. Her main research interest is in the nature of freedom. Bruce Jennings is Director of the Center for Humans and Nature, a private operating foundation that focuses on ethical issues in environmental and health policy, and Senior Consultant at the Hastings Center, Garrison, NY. He also teaches at the Yale University School of Public Health. He is author of numerous books and articles on ethical and social issues in health-care and public policy. He is co-author (with Willard Gaylin) of The Perversion of Autonomy: The Uses of Coercion and Constraint in a Liberal Society (2nd edn., Georgetown University Press, 2003). He is currently at work on a book on dementia, quality of life, and the ethics of long-term care. Eric T. Juengst is Professor of Bioethics at the Case Western Reserve University School of Medicine, where he directs the Center for Genetic Research Ethics and Law. He has written widely on conceptual and ethical issues in genetics, and has served on the DNA Advisory Board of the FBI, the NIH National Advisory Council for Human Genome Research, the NIH Recombinant DNA Advisory Committee, the Ethics Committee of the American Society for Gene Therapy, and the National Academy of Sciences Committee on Human Genetic Diversity. He is an elected Fellow of the Hastings Center, Garrison, NY, and the Center for Genetics and Society, University of California, Los Angeles, and in 2006 was awarded the Golden Eurydice Prize for pioneering work integrating science and ethics from the European Commission’s International Forum on Biophilosophy. Jeffrey Kahn is the Maas Family Endowed Chair in Bioethics and Director of the Center for Bioethics at the University of Minnesota. He holds additional faculty appointments in the university’s Medical School, School of Public Health, and Department of Philosophy. He has published over seventy-five articles in both the bioethics and medical literature, serves on numerous state and federal advisory panels, and speaks nationally and internationally on a range of bioethics topics. From 1998 to 2002 he also wrote the bi-weekly column ‘Ethics Matters’ on CNN.com. Jason Karlawish is Associate Professor of Medicine, Fellow of the Center for Clinical Epidemiology and Biostatistics, and Senior Fellow of the Center for Bioethics and the Leonard Davis Institute of Health Economics at the University of Pennsylvania. He is the Associate Director of the Penn Memory Center and the Director of the Alzheimer’s Disease Center’s Education and Information Transfer Core, University

xvi

notes on the contributors

of Pennsylvania. His research focuses on ethical issues in human subjects research and the care of persons with dementia. Jeanette Kennett is Principal Research Fellow in the Centre for Applied Philosophy and Public Ethics, Australian National University, and School of Philosophy and Bioethics, Monash University. She is the author of Agency and Responsibility (Clarendon Press, 2001) and has published widely on philosophical and ethical issues related to moral responsibility, the self, and mental disorder. Benjamin J. Krohmal is a fellow in the Department of Clinical Bioethics at the Clinical Center of the National Institutes of Health. His current focus is on research ethics and the ethics of health-care reform. Alex John London is Associate Professor of Philosophy and an executive member of the Center for the Advancement of Applied Ethics and Political Philosophy at Carnegie Mellon University. The recipient of a New Directions Fellowship from the Andrew W. Mellon Foundation in 2005, he is also co-editor (with Bonnie Steinbock and John D. Arras) of Ethical Issues in Modern Medicine (6th edn., McGraw-Hill, 2003). Florencia Luna is an Adjunct Researcher at CONICET (the National Scientific and Technological Research Council), Argentina. She is the Director of the Program of Bioethics at FLACSO (the Latin American University of Social Sciences) and co-director with Ruth Macklin of a research training grant of the National Institutes of Health in the United States. She was also the President of the International Association of Bioethics (2003–5). She is the author of Bioethics and Vulnerability: A Latin American View (Rodopi, 2006), and of Ensayos de bio´etica: Reflexiones desde el Sur (Fontamara, 2001), and co-editor (with Arleen L. F. Salles) of Decisiones de vida y muerte (Sudamericana, Buenos Aires, 1995) and (with Arleen L. F. Salles) Bioetica (Sudamericana, 1998). Ruth Macklin is Professor of Bioethics at Albert Einstein College of Medicine in the Bronx, New York. She is the author or editor of twelve books and has published more than 200 articles in scholarly and professional journals. She is an elected member of the Institute of Medicine of the National Academies of Science, chairs the External Ethics Committee of the Centers for Disease Control and Prevention, Atlanta, Georgia, and serves as an adviser to the World Health Organization and the Joint United Nations Programme on HIV/AIDS. Her latest book is Double Standards in Medical Research in Developing Countries (Cambridge University Press, 2004). Don Marquis is Professor of Philosophy at the University of Kansas. He teaches applied ethics and ethical theory. Anna Mastroianni is Associate Professor at the School of Law and Institute for Public Health Genetics at the University of Washington. She holds additional

notes on the contributors

xvii

faculty appointments in the University’s School of Public Health and Community Medicine and Medical School. She has held a number of legal and federal policy positions in the United States and has served on several national advisory panels concerning issues of biomedicine and health policy. She has published numerous articles in the legal and bioethics literature, with specific interests in the legal, ethical, and policy issues related to human subjects research, public health, and the use of genetic technologies. Dennis McKerlie is Associate Professor of Philosophy at the University of Calgary. He has published articles on egalitarianism, justice between age groups, and ancient philosophy. Carolyn McLeod is Associate Professor of Philosophy at the University of Western Ontario. From 2004 to 2006 she was also a New Faculty Fellow in the Comparative Program on Health and Society at the Munk Centre for International Studies at the University of Toronto. She has published in bioethics and moral theory, on topics such as trust, autonomy, integrity, conscience, objectification, and commodification. Her book Self-Trust and Reproductive Autonomy (MIT Press, 2002), is part of the Basic Bioethics series of the MIT Press. Jonathan D. Moreno is the Emily Davie and Joseph S. Kornfeld Professor and Director of the Center for Biomedical Ethics at the University of Virginia. He is an elected member of the Institute of Medicine and has served as senior staff to two presidential advisory committees and as a consultant to numerous public and private organizations. Among his books are Undue Risk: Secret State Experiments on Humans (Routledge, 2001) and In the Wake of Terror: Medicine and Morals in a Time of Crisis (MIT Press, 2003). His latest book is Is There an Ethicist in the House? On the Cutting Edge of Bioethics (Indiana University Press, 2005). Ronald Munson is Professor of the Philosophy of Science and Medicine at the University of Missouri—St Louis. He has served as a bioethicist for the National Cancer Institute, the National Eye Institute, and the Veterans Administration, and is a member of the Human Studies Committee at Washington University School of Medicine. His books include Reasoning in Medicine (Johns Hopkins University Press, 1994), Outcome Uncertain: Cases and Contexts in Bioethics (Wadsworth, 2003), and Raising the Dead: Organ Transplants, Ethics, and Society (Oxford University Press, 2002). He is also the author of the novels Nothing Human, Fan Mail, and Night Vision. Thomas H. Murray is President of the Hastings Center, Garrison, NY. He has written or edited many books including The Worth of a Child (University of California Press, 1996), Cultures of Caregiving (with C. Levine; Johns Hopkins University Press, 2004), Genetic Ties and the Family (with M. Rothstein, G. Kaebnick, and M. A. Majumder; Johns Hopkins University Press, 2005), The

xviii

notes on the contributors

Encyclopedia of Ethical, Legal, and Policy Issues in Biotechnology (with M. Mehlman; John Wiley, 2000), and Feeling Good and Doing Better: Ethics and Nontherapeutic Drug Use (Humana Press, 1984). He is Chair of the World Anti-Doping Agency’s Ethical Issues Review Panel. Alastair Norcross is Associate Professor of Philosophy at Rice University. His articles on ethical theory and applied ethics have appeared in such journals as Philosophical Review, Journal of Philosophy, Philosophy and Public Affairs, and Social Theory and Practice. He is co-editor, with Bonnie Steinbock, of Killing and Letting Die (2nd edn., Fordham, 1994). Stephen G. Post is Professor in the Department of Bioethics, Case Western Reserve University School of Medicine, and was a Senior Research Scholar in the Becket Institute at St Hugh’s College, University of Oxford. He is Editor-in-Chief of The Encyclopedia of Bioethics (3rd edn., Macmillan Reference, 2004). He is an elected member of the Medical and Scientific Advisory Panel of Alzheimer’s Disease International, the recipient of a ‘distinguished service’ award from the Association’s National Board (1998), and the author of The Moral Challenge of Alzheimer Disease: Ethical Issues from Diagnosis to Dying (2nd edn., Johns Hopkins University Press, 2000). Julian Savulescu is the Uehiro Professor of Practical Ethics and Director of the Oxford Uehiro Centre for Practical Ethics at the University of Oxford, and Director of the Program on Ethics and the New Biosciences in the 21st Century School, University of Oxford. Qualified in medicine, bioethics, and analytic philosophy, he has published over 100 articles and is co-author (with R. A. Hope and Judith Hendrick) of the book Medical Ethics and Law (Churchill Livingstone, 2003). Bonnie Steinbock is Professor of Philosophy at the University at Albany/SUNY. She is the author of Life Before Birth: The Moral and Legal Status of Embryos and Fetuses (Oxford University Press, 1992) and the editor of several books, including (with John Arras and Alex London) Ethical Issues in Modern Medicine (6th edn., McGraw-Hill, 2003). Daniel P. Sulmasy, OFM is Sisters of Charity Chair in Ethics at St Vincent’s Hospital Manhattan, and Professor of Medicine and Director of the Bioethics Institute of New York Medical College. He is Editor-in-Chief of the journal Theoretical Medicine and Bioethics. His latest books are The Rebirth of the Clinic and A Balm for Gilead: Mediations on Spirituality and the Healing Arts (both Georgetown University Press, 2006). Stuart J. Youngner is the Susan E. Watson Professor of Bioethics and Chair of the Department of Bioethics at Case Western Reserve University School of Medicine. He has written and spoken extensively about definitions of death, organ and tissue transplantation, end-of-life decisions, and clinical ethics consultation.

•••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

INTRODUCTION •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

bonnie steinbock The aim of this Handbook is to provide an up-to-date picture of the state of the art in bioethics. It makes no attempt to cover every issue in the field—there are excellent encyclopedias that serve this purpose (Post 2004; Murray and Mehlman 2000). Instead, it is a selective representation of some of the most important issues in contemporary bioethics. The chapters are surveys in that they inform the reader of what has been happening recently in each area, through a discussion of the relevant literature. At the same time, they are not neutral, encyclopedia-like articles, but original essays that reflect the particular ‘take’ of their authors. They are not intended for undergraduates or general readers, but rather for those with some knowledge of the field (scholars and graduate students) who want an authoritative and stimulating account of bioethics today. My task as editor was to get the best and most interesting individuals in the field as contributors. This might be interpreted as meaning the best-known or most prominent bioethicists, and certainly this Handbook includes some of the most respected names in bioethics, people who have shaped, and continue to shape, the field. However, I also wanted to include some who are not as well known because they are in the early stages of their careers. These ‘up-and-comers’ represent the next generation of bioethicists. In addition, the aim was to reflect the interdisciplinarity that is fundamental to bioethics. Therefore, the volume includes individuals not only from my own discipline (philosophy), but also from theology, medicine, law, political science, social science, and public health. Until recently, bioethics has been dominated by Americans. However, in recognition that bioethics is becoming increasingly international, a third of the contributors come from outside the United States. Equally importantly, the authors represent a diversity of opinions and

2

bonnie steinbock

viewpoints, sharpening and deepening the debate on a number of controversial issues.

A Short History of Bioethics ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Bioethics was preceded by medical ethics, which focused primarily on issues arising out of the physician–patient relationship. The ancient Hippocratic literature (which includes but is not limited to the Hippocratic Oath) enjoins doctors to use their knowledge and powers to benefit the sick, to heal and not to harm, to preserve life, and to keep in the strictest confidence information that ought not to be spread about (though precisely what must be kept confidential is not detailed). These basic values and principles remain an essential part of contemporary bioethics. However, after the Second World War it became clear that the old medical ethics was not sufficient to meet contemporary challenges. Unprecedented advances in medicine, including the use of penicillin and immunizations against childhood diseases, have saved literally millions of lives. So have open heart surgery and cardiac catheterization, chronic hemodialysis, and organ donation. At the same time, many of the tools of modern medicine are very expensive, and thus out of the reach of many who might benefit from them. Medicine’s success thus led to a debate about how to pay for health care. In most industrialized countries, the provision of health care is viewed as the responsibility of government, comparable to the obligation to provide public education. By contrast, in the United States many still regard payment for health care as an individual responsibility, or at least something that employers, not the state, should provide. Among those who agree that some kind of national health insurance is both fair and fiscally sound, a debate continues between egalitarians, who insist that no care should be provided unless it is available to all who need it, and those who favor a tiered health care system that allows some medical services to be distributed by the market. (This issue is thoroughly explored in Part II, Chapter 7, by Benjamin J. Krohmal and Ezekiel J. Emanuel.) Medicine’s successes in the post-war years raised another issue: the value of preserving life. Respirators were originally invented for people who were expected to recover and be able to breathe on their own. Within a short period of time they began to be used on people in persistent vegetative states, forcing medical professionals to ask whether this was an appropriate use of technology. Should people who are permanently and irreversibly unconscious be kept alive indefinitely? A similar issue resulted from the development of neonatal intensive care units (NICUs), which have saved the lives of many premature babies who would have died in earlier decades. Many of these babies go on to have normal, healthy lives, but many face a lifetime of severe disabilities and serious health complications.

introduction

3

Thus, NICUs raise the question: Ought life to be preserved regardless of the nature or quality of that life? And if there are times when life should not be preserved, who should be authorized to make these decisions? During the 1960s these questions began to be debated at academic conferences and in scholarly journals, giving birth to the field of bioethics. In 1969 the Hastings Center in Garrison, NY, an independent, nonpartisan, and nonprofit bioethics research institute, was founded by Dan Callahan and Willard Gaylin to explore fundamental and emerging questions in health care, biotechnology, and the environment. Its journal, the Hastings Center Report, first appeared in June 1971. In July of that year the Kennedy Institute of Ethics at Georgetown University opened, with two research scholars: LeRoy Walters, who soon became its Director, and Warren Reich, who was the editor of the first edition of the Encyclopedia of Bioethics, published in 1978. The term ‘bioethics’ was coined in the early 1970s by biologists who brought to the public’s attention two pressing issues: the need to maintain the planet’s ecology, on which all life depends, and the implications of advances in the life sciences toward manipulating human nature. In his book, Bioethics: Bridge to the Future, published in 1971, Van Rensselaer Potter focused on the growing human ability to change nature, including human nature, and the implications of this for our global future. This issue has been revisited in recent years in a growing literature on enhancement, genetic and otherwise, and is addressed from very different perspectives in Part VI Chapters 21 and 22, by Thomas Murray and Julian Savulescu respectively. Although the term ‘bioethics’ has referred almost exclusively to problems in biomedicine, in recent years the field has returned to ‘the wider context provided by the life scientists of the early 1970s, including their environmental and public health concerns’ (Post 2004, p. xi). (The relevance of public health for bioethics is covered in Part VIII, Chapter 28, by Jeffrey Kahn and Anna Mastroianni.) While bioethics has been interdisciplinary since its inception, theology played a foundational role in its creation. It continues to have a profound influence today, as reflected in the careful analysis and defense of the rule of double effect by Daniel Sulmasy in Part I, Chapter 5. Three theologians in particular were instrumental in the birth of bioethics: Joseph Fletcher, an Episcopal minister; Paul Ramsey, a Methodist minister; and Richard McCormick, a Jesuit moral theologian. The theologians were soon joined by philosophers who rejected the emphasis in contemporary analytic ethics on meta-ethics, to the exclusion of normative ethics. Events in the 1960s—opposition to the war in Vietnam, the civil rights movement, and other social movements it spawned, such as the women’s movement, the disability rights movement, and the gay and lesbian rights movement—played a role in the revitalization of normative ethics, and philosophical interest in applied ethics. Students began to demand that their courses were ‘relevant’, and professional philosophers also became interested in writing on the issues of the day. Philosophers now specialize in bioethics, but even philosophers not usually thought

4

bonnie steinbock

of as ‘bioethicists’, including Jonathan Bennett, Ronald Dworkin, Joel Feinberg, Jonathan Glover, Thomas Nagel, Onora O’Neill, Judith Thomson, and Bernard Williams, have made important contributions to the bioethics literature. A new journal, Philosophy and Public Affairs, appeared in 1971, with lead articles on war and abortion. John Rawls’s A Theory of Justice appeared the same year. The book had a huge impact on Anglo-American political philosophy, and indirectly on bioethics. As Jonsen (1998: 74) puts it, ‘Rawls’s thesis excited many moral philosophers, restored faith in a rational approach to ethics, provided a carefully articulated version of contractarianism, and bequeathed to some future bioethicists the basis for a theory of the allocation of medical resources.’ The influence of Rawls can be seen in many of the chapters in this volume, especially in Parts I and II.

Central Issues in Bioethics ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

At its inception, the central issues in bioethics were research with human subjects, genetics, organ transplantation, death and dying, and reproduction. As a glance at the table of contents reveals, these continue to be important issues today. Methodology has been a central theoretical issue since the very beginning of bioethics. While some hoped that bioethics would generate a single correct normative theory, that idea has been given up by most practitioners. Indeed, bioethics reflects a wide range of theoretical approaches in normative ethics, including utilitarianism, deontology, natural law, contractarianism, virtue ethics, communitarianism, pragmatism, and feminist ethics (Steinbock et al. 2003). All of these approaches, and variations within these, are carefully enunciated by James Childress in Part I, Chapter 1. As Childress perhaps ruefully concludes, ‘It is probably too much to expect consensus about the best possible method(s), all things considered—at most, we can take advantage of the strengths of each method and compensate for its special deficiencies.’ A related issue concerns the relationship between abstract principles and concrete particulars. An approach that has come to be known as ‘principlism’, famously advocated by Beauchamp and Childress in their classic text Principles of Bioethics (1979–2001), attempts to derive answers to bioethical dilemmas from the basic principles: autonomy, beneficence, and justice. Principlism has been criticized, as John Arras points out in Chapter 2, by casuists, feminists, partisans of narrative ethics, and pragmatists as ‘too abstract, deductive, and ‘‘top down’’ ’; as being ‘insufficiently attentive to particulars, relationships, storytelling, and process’. In response, principlists have attempted to incorporate these insights into a principlebased approach, using Rawls’s method of ‘reflective equilibrium’. However, while this approach has considerable intuitive appeal, it is far from clear exactly how reflective equilibrium is to be interpreted or applied. Arras provides a clear and

introduction

5

thorough analysis of the difficulties, ending on a moderately skeptical note about the value of method in moral inquiry. No method, he says, is a guarantor of truth, or even intersubjective agreement, but careful attention to the various approaches can improve practical reasoning and ‘facilitate our quest for moral justification’. In addition to methodological issues, theoretical issues in bioethics include certain concepts and principles. Of these, perhaps the most central is autonomy. The prominence of autonomy in biomedical ethics can be traced back to the Belmont Report (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research 1979), so called because it had its origins in a meeting held in February 1976 at Belmont House, a conference center of the Smithsonian Institution. Congress had instructed the Commission to identify the basic ethical principles that should underlie the conduct of biomedical and behavioral research involving human subjects. The Commission articulated three basic principles: respect for persons, beneficence, and justice. The articulation of these principles had a major impact on the development of bioethics, and indeed on virtually every chapter in this volume. These concepts found their way into the general literature of the field, and evolved from the principles underlying the conduct of research into the basic principles of bioethics. ‘Respect for persons’ became understood as ‘respect for autonomy’ or simply ‘autonomy’. Yet while autonomy is unquestionably a central principle in bioethics, both its interpretation and its moral weight remain controversial, as Bruce Jennings illustrates in Chapter 3. The issue of autonomy is considered from the perspective of moral psychology in Chapter 4, by Jeanette Kennett. While mental illness clearly can rob a person of the capacity for autonomous choice and action, it is also true that this has been exaggerated in the past by prejudice and stigmatization. What, then, is the correct attitude to take toward those suffering from mental illness? Kennett reminds us that ‘agency comes in degrees, that autonomy is an achievement, and that respect for autonomy may require our active support for the agency of those in adverse circumstances’, including mental illness. Acknowledging the difficulty of adopting ‘the participant stance’ toward those who are seriously mentally ill, Kennett argues that nevertheless we are morally required to adopt it, so far as possible. Part II turns to issues of justice and policy. Like Arras in the preceding section, Søren Holm, a Danish philosopher and physician, starts with Rawls, although Holm is less interested in reflective equilibrium as a methodology, and more interested in the requirements of deliberative democracy. He notes that while policy making shares some of the values of ethical decision making (such as reasonableness, reciprocity, consistency, and integrity), it also incorporates features that may be anathema to philosophers, including political compromise, moratoria (waiting periods while scientific uncertainty is resolved), and the accommodation of minority positions. These political tools, Holm argues, are as important as philosophical reasoning for achieving ‘peaceful public decision making in a context of fundamental moral disagreement’.

6

bonnie steinbock

The next chapter, by Benjamin J. Krohmal and Ezekiel J. Emanuel, considers whether a just health care system must be strictly egalitarian (as in Canada and Norway), or whether a better approach allows for a tiered health care system, in which some people can pay for greater access to services than others with the same needs. After examining the arguments pro and con, the authors conclude that an egalitarian approach is unjust. Justice calls for a two-tier health care system with universal public coverage (as in Britain and Israel). Another question of distributive justice arises because people have different needs for health care. For example, as people age, they are likely to need more and more expensive health care, including prescriptions, tests, and hospital stays. In Chapter 8 Dennis McKerlie, a Canadian philosopher, addresses this issue, asking, ‘Are the elderly receiving less, or more, than their fair share of health resources and economic wealth?’ An important part of the answer depends on how principles of justice are applied. McKerlie contrasts a ‘complete lives’ view, which focuses on people’s lifetime expectations of receiving primary goods, with the prudential lifespan account, proposed by Norman Daniels and Ronald Dworkin, in which principles of justice have a temporal scope. He rejects both in favor of what he calls the ‘life-stage view’, which is more generous to the elderly than either the complete lives view or the prudential lifespan account, in large measure because they need help the most. The chapter ends with a thoughtful reflection on the effects of aging on individuals’ abilities and values, and the implications this has both for respect for autonomy and for a just distribution of resources. Part III addresses a set of issues turning around human bodies and parts of bodies. In Chapter 9 Ronald Munson outlines the history, economics, and ethics of organ transplantation, an issue that has been prominent in medical ethics since the first organ transplant in 1954. Most transplants are taken from dead donors, but many more lives could be saved if the number of living donors of kidneys and livers was increased. The risks to donors, however, are considerable, and the chance of coercion or undue influence is thought by some to undermine voluntariness and autonomy. However, Munson argues that paternalistically denying someone the opportunity to be a living donor is a greater threat to autonomy, especially since measures can be taken to protect voluntariness. The next chapter, by Louise Irving and John Harris, on biobanking discusses the methods and purposes of tissue storage and collection. A major theme is the great importance of tissue collection to medical research, which at the same time raises issues of confidentiality, consent, stigmatization, and risk. While these need to be taken seriously, Irving and Harris argue that the ethical problems raised by biobanking are neither unique nor unresolvable. Carolyn McLeod, in Chapter 11, addresses the issues of oocyte vending and commercial contract pregnancy from a feminist perspective. The commodification of women’s reproductive capacity presents feminists with a classic double bind. Should a market in reproductive labor be viewed as degrading and rife with the

introduction

7

potential for harm and exploitation of women? Or should the ability to sell one’s labor (reproductive or otherwise) be viewed as empowering, and the denial of this right inconsistent with women’s status as autonomous persons? McLeod does not resolve this issue, but suggests several ways in which the debate could be refined and clarified. Part IV opens with an update on the problems with the definition and determination of death. Stuart J. Youngner illustrates the ways in which this is not only a scientific issue, but a matter of philosophy and policy as well. His chapter dovetails nicely with Munson’s, on organ transplantation, since the possibility of transplanting organs from deceased donors has influenced, if not driven, the conceptual debates about death. One of the challenges of the twenty-first century is the increase in lifespan, and the consequent increase in numbers of the very old, issues addressed by Stephen Post and Felicia Nimue Ackerman in Chapters 13 and 14. While both Post and Ackerman disagree with the natural law approach taken by, for example, Leon Kass, they disagree with each other about the value of extending life when this entails illness and deterioration. John K. Davis returns, in Chapter 15, to the issue of autonomy in his discussion of advance directives (ADs). A number of writers have argued that ADs pose an insoluble problem. They express the values and preferences of the competent individuals who write them, but they go into effect only when the individuals are no longer competent. Once incompetent, they are likely to have very different interests than the interests they had when competent. They may no longer care about, or even be cognizant of, issues such as dignity and independence. Indeed, some would argue that the presently incompetent individual is so radically changed that he or she is not the same person as the individual who wrote the AD, and that therefore the writer of the AD does not have the moral authority to decide that lifesaving treatment should be withheld from the present incompetent person. Drawing a distinction between loss of capacity and loss of concern, Davis defends ADs against this particular criticism. Gerald Dworkin ends this part with a careful analysis of the arguments for and against physicianassisted death, and a defense of physician-assisted suicide, though not, for strategic reasons, euthanasia. Part V concerns issues at the beginning of life, specifically the morality of abortion and embryonic stem cell research. In 1989 Don Marquis published a now classic article, ‘Why Abortion Is Immoral’, in which he argued that both pro-life and pro-choice approaches were seriously flawed. In their place, he offered a new approach, the future of value argument. It maintains that the best explanation for the wrongness of killing is that killing deprives us of our valuable futures. Because we were once fetuses, fetuses have the same futures of value that every one of us has, and therefore abortion is ‘seriously presumptively wrong’. In this new chapter Marquis elaborates on the notion of having a future of value, and argues that this

8

bonnie steinbock

argument is superior to recent accounts of moral status provided by Mary Anne Warren and David Boonin. My own contribution to the book concerns the moral status of extracorporeal embryos, and the implications this has for embryonic stem cell research (ESCR). The chapter builds on the interest view, which I elaborated in Life Before Birth (1992). I argue that very early embryos (blastocysts) have no moral status, because they lack interests of their own. This means that we cannot act in their interest or on their behalf; they cannot be harmed or benefited. Nevertheless, as developing forms of human life, they have moral value, and are entitled to respect. I distinguish between the respect owed to human embryos and Kantian respect for persons, and argue that the respect owed to embryos does not rule out using them merely as means to our ends. Instead, it requires that the ends served be morally important ones. Potentially valuable scientific research qualifies as morally important, and is therefore consistent with respect for human life. In the last chapter in Part V Andrea Bonnicksen returns to an issue addressed in Part I by Søren Holm: policy making in pluralistic societies. In the United States ESCR is politically extremely divisive. There is no federal support for it, and hence no federal oversight. Yet several states (California, New Jersey, Massachusetts, Illinois) have passed legislation to allow and/or fund ESCR, including the cloning of human embryos. Moreover, several private laboratories are attempting to clone human embryos and derive stem cell lines, without, however, the kind of oversight and ethical reflection characteristic of the Human Fertilization and Embryology Authority in the United Kingdom. Bonnicksen suggests that a ‘policy community’, consisting of academics, policy analysts, interest group members, and others with shared interests, might be able to overcome the current stalemate, both within the United States and internationally, and address some of the pressing ethical concerns raised by therapeutic cloning. In addition to the issue of enhancement previously mentioned, Part VI addresses issues in genomic medicine. In Chapter 20 Eric Juengst notes that while the interest in mapping the human genome was initially genealogical, interest soon developed in the possibility of tailoring diagnostic protocols, therapeutic interventions, and preventive measures to each patient’s genetic profile. This would have profound implications not only for medicine, but also for public health (this is addressed in Part VIII, Chapter 28, by Jeffrey Kahn and Anna Mastroianni). At the same time, it has raised old questions about racism and discrimination, as well as new questions about the collective interests of groups being studied. In Chapter 23 Matthew DeCamp and Allen Buchanan focus specifically on the role pharmaceuticals play in genomic medicine. In addition to problems resulting from the reifying of race, pharmacogenomic technologies raise serious issues of distributive justice: who will get these new drugs, on what basis will decisions to produce them be made, and who will pay for them?

introduction

9

Part VII concerns the ethics of research on human subjects, a subject that has been at the heart of bioethics since its inception. For most of its history, medicine was primarily therapeutic. Experimentation was limited to trying new curative or palliative approaches on individual patients. As medicine became more of a science, physicians wanted to know if their treatments, standard and innovative, actually worked. To do this, one needed to perform experiments on people, not primarily to benefit them, but to learn things that might benefit others. This put physician–researchers into a double bind. In so far as they were acting as the patient’s doctor, their aim was to benefit the patient. In so far as they were acting as scientific researchers, their aim was to perform a controlled experiment and get accurate results. Typically, this means assigning subjects randomly to the various arms of a clinical trial without regard to whether the standard treatment, the new treatment, or placebo would be in the patient–subject’s best interest. It seems that physicians cannot satisfy both roles simultaneously. As Alex London points out in the first chapter in Part VII, Charles Fried attempted to solve this dilemma through the notion of ‘equipoise’. Physicians were justified in enrolling their patients in clinical trials only if they were uncertain about the relative therapeutic merits of various arms of the trial. However, as London notes, there is considerable uncertainty about how precisely to formulate the requirement of equipoise. After clarifying different versions of the equipoise requirement, and demonstrating their respective strengths and weaknesses, London proposes a novel approach to research ethics that better avoids the exploitation of research participants. The problem of exploitation is central to the next three chapters as well. In Chapter 25 the physician Jason Karlawish examines the problem of research on individuals who, due to cognitive impairment, cannot give informed consent. Is it possible both to accomplish valuable research and also to protect vulnerable individuals? This issue is also addressed by Florencia Luna (and again in Part VIII in the chapter by Jeffrey Kahn and Anna Mastroianni) from the perspective of research on vulnerable populations, namely, individuals in developing countries. The problems of such research were brought into sharp relief in the late 1990s, when clinical trials were done in sub-Saharan Africa, Asia, and Latin America to find a more economical and effective treatment for preventing mother-to-baby transmission of HIV/AIDS. A proven treatment for preventing transmission existed, and was standard in the United States and other developed countries. However, it was not only expensive (far beyond the health care budgets of developing countries), but also burdensome to administer—perhaps even impossible in the conditions in most developing countries. To see if a less expensive, less burdensome regimen was effective, half the subjects in the trial were given the proposed regimen and half were given placebos. This was extremely controversial because normally a new treatment is tested against the proven treatment to determine if it is as effective.

10

bonnie steinbock

This was not done in the so-called ‘AIDS-African trials’ because the proven regimen was beyond the economic reach of developing countries. The issue for them was not whether the new regimen was as effective as the proven regimen, but whether it was better than nothing (i.e. placebo). If it was not, there would be no point in using scarce health care dollars to provide it to HIV-infected pregnant women. This seemed reasonable, and yet such a protocol would never be approved in a developed country. Critics complained of introducing a double standard into international research—‘a standard for industrialized countries that can provide the best existing therapies and another for poor countries with limited funding and a deficient healthcare system’. Exploitation of vulnerable populations is a real problem, but not one that can be solved with slogans, Luna argues. Obligations to research subjects during and after clinical trials must be considered, as well as the benefit to of individuals and communities who participate in research. Part VII ends with a discussion of research on animals. Ever since Peter Singer’s Animal Liberation appeared in 1975, there has been a lively debate about the moral status and treatment of animals. Although the treatment of human beings in research has at times been scandalous (with the experiments on Jews and others in Nazi concentration camps being the most heinous example), such treatment of non-human animals is pervasive. Animals are routinely killed, maimed, shocked, burned, and caused terrible pain, in the name of scientific and medical progress. In Chapter 27 Alastair Norcross examines various justifications for inflicting pain on non-human animals, and concludes that they are unsuccessful, and that this renders much animal experimentation morally unacceptable. Part VIII explores some relatively new directions for bioethics: the turn to public and global health. On the one hand, public health seems at odds with bioethics because the focus in public health is the health of the population or community, whereas the focus in medicine is the health of the individual. Moreover, even as bioethics diverged from traditional medical ethics, it retained the principle of respect for individual autonomy, and correlative principles of confidentiality and privacy. On the other hand, there has been a kind of self-correction in the field of bioethics, away from an excessive individualism and toward engagement in questions of public health and global justice. In the first chapter in Part VIII Jeffrey Kahn and Anna Mastroianni discuss these issues, using as examples HIV/AIDS, resource allocation, and public health genomics. In the next chapter Ruth Macklin explores the notion of ‘globalization’, and explains why in today’s globalized world public health is global health. In part, this is because intercontinental travel makes the spread of disease so rapid, increasing the possibility of global pandemics. In addition, actions and policies carried out in one part of the world are likely to have significant, and often harmful, effects on people very far away. Thus, it becomes increasingly impossible to reject responsibility for those effects. Finally, the world’s problems can only be solved by international cooperation.

introduction

11

In the last chapter Jonathan Moreno discusses the ethical problems raised by bioterrorism, including the tension between respecting civil liberties and protecting the public’s health, triage in a bioterror attack, and the responsibilities of emergency health care professionals. Moreno’s chapter is an excellent example of the way in which bioethics can draw on historical parallels even as it adapts in response to new developments and challenges. As science and medicine continue to advance, new issues are constantly being raised, for example, in neuroscience, bioengineering, and nanotechnology. Some of these emerging sciences and technologies may pose genuinely new questions for bioethics; others can be characterized as ‘old wine in new bottles’. Controversy over this question (and others) is inevitable, as bioethicists identify and analyze new issues, thus framing the debates of the twenty-first century.

References Beauchamp, T., and Childress, J. (1979–2001), Principles of Bioethics, (1st–5th edns. New York: Oxford University Press). Jonsen, A. R. (1998), The Birth of Bioethics (New York: Oxford University Press). Marquis, D. (1989), ‘Why Abortion Is Immoral’, Journal of Philosophy, 89: 183–202. Murray, T. A., and Mehlman, M. (eds.) (2000), The Encyclopedia of Ethical, Legal and Policy Issues in Biotechnology (New York: John Wiley). National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research (1979), The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects of Research (Washington, DC: Government Printing Office). Post, S. G. (ed.) (2004), The Encyclopedia of Bioethics, (3rd edn. New York: Macmillan). Potter, V. R. (1971), Bioethics: Bridge to the Future (Englewood Cliffs, NJ: Prentice-Hall). Singer, P. (1975), Animal Liberation (New York: Random House). Steinbock, B. (1992), Life Before Birth: The Moral and Legal Status of Embryos and Fetuses (New York: Oxford University Press). Arras, J. D., and London, A. J. (2003), Ethical Issues in Modern Medicine, (6th edn. Boston: McGraw-Hill).

part i •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

THEORETICAL AND M E T H O D O LO G I C A L ISSUES •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

chapter 1 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

METHODS IN BIOETHICS •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

james f. childress

Introduction: Questions About Method ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

In the preface to the first edition of his classic work The Methods of Ethics, Henry Sidgwick noted its focus on ‘the different methods of obtaining reasoned convictions as to what ought to be done’ (Sidgwick 1962, p. v). He defined a ‘method of ethics’ as ‘any rational procedure by which we determine what individual human beings ‘‘ought’’—or what it is ‘‘right’’ for them—to do, or to seek to realize by voluntary action’ (Sidgwick 1962: 1). One major reason for the widespread interest in methods of doing bioethics is to determine how best to guide human action. Hence, assessments of different methods consider, in part, how well a bioethical theory, framework, or perspective guides action—other criteria include clarity, consistency, coherence, completeness, and comprehensiveness, as well as congruence with moral experience (Beauchamp and Childress 2001, ch. 8). But, even on Sidgwick’s definition, there can be strong and weak conceptions of method. For instance, a method may illuminate an agent’s choices without fully prescribing or determining what he or she should do. Indeed, a method may provide a complex rather than a tidy answer to the agent who asks, ‘What should I (or we) do?’ Early bioethics, in the 1970s, was often viewed as a species of ‘applied ethics’: bioethics denoted the reflective activity of applying an ethical theory or ethical principles to the domains of the biological sciences, medicine, and health care.

16

james f. childress

The language of ‘applied ethics’ implies more action guidance from theories or principles than is usually available, and it has now been largely discarded in favor of the language of ‘practical ethics’, a phrase that Sidgwick (1998) also used. Not all methods for addressing practical moral problems entail applying, in a deductivist fashion, ethical theories, frameworks, or perspectives. The philosopher R. M. Hare (1996), in examining ‘the methods of bioethics’ (the title of one of his articles), observes that the designation ‘ethical or moral theories’ covers different kinds of activities, as captured, for instance, in their broad and narrow meanings. On the narrow conception of ethical theory, the task is to examine the logic of moral reasoning, and representative theories, often called metaethics, including naturalism, intuitionism, subjectivism, emotivism, and prescriptivism. Hare (1996) believes that we have to start with these narrow theories, because they can make the greatest contributions to bioethics, but he focuses instead on normative theories, such as utilitarianism, virtue ethics, and ethics of care. I too will concentrate on normative ethical methods and theories rather than on metaethics, while recognizing that the line between them is not always clear. Other broad uses of the term ‘method’ encompass a variety of descriptive approaches. Observing the wide range of methods used in scholarly inquiry about physician-assisted suicide, Daniel Sulmasy concludes that scholars who have employed such methods as history, law, theology, philosophy, quantitative methods, ethnographic methods, and so forth, are all ‘properly called ‘‘medical ethicists’’, and their research is properly called ‘‘medical ethics’’ ’ (Sulmasy 2001: 259). However, this usage is too broad to be helpful. Although numerous methods of scholarly inquiry can and do make important contributions to medical ethics or bioethics, not all those contributions actually involve doing medical ethics or bioethics in the normative sense, and I will limit my attention to methods in normative bioethics. This restriction in no way denigrates other methods and their contributions—indeed, they are frequently illuminating, and often indispensable. Nevertheless, a scholarly inquiry into bioethics, or into some topic within bioethics, does not necessary translate into ‘doing bioethics’, however important it may be for ‘doing bioethics’. This chapter then will stay largely within boundaries of normative bioethics. It will examine major types of principle-based methods (consequentialist, deontological, and pluralistic principlist methods), case-based methods, virtue ethics, ethics of care, and communitarian perspectives, along with some critical points from feminist perspectives and from rule-based theories. One cautionary note is in order: most of these types of method, theory, or perspective encompass a number of approaches that involve some degree of family resemblance. Since it will be impossible to examine all of these approaches in detail, I will highlight some major themes and criticisms, discuss a few representative positions in more detail, and sketch a few of their implications for practical decision-making about physician-assisted

methods in bioethics

17

suicide and active euthanasia, topics addressed by proponents of all these methods, theories, and perspectives.

Principle-Based Methods ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

A principle-based method, some claim, dominates bioethics. This claim may be accurate when the whole range of principle-based positions is surveyed. However, principle-based methods encompass a wide and rich variety of more specific methods, all of which stress principles without necessarily sharing much more. Minimally, a bioethical method that is principle-based must hold that general moral action guides are central to moral reasoning in bioethics. It need not, however, reduce all moral reasoning to explicit principle-based reasoning. It may concede, for instance, that appeals to principles most often occur when there is uncertainty or conflict about the appropriate course of action. Action guides are frequently distinguished into principles and rules. The most general action guides—for instance, utility or respect for autonomy—are usually labeled ‘principles’ while the more specific action guides—for instance, respect for confidentiality—are usually labeled ‘rules’. Nevertheless, the terms ‘principles’ and ‘rules’ are frequently used interchangeably, and the lines between them are often unclear because they reflect different degrees of generality and specificity. I will later consider different interpretations of the relationship between principles and rules; for now I merely note that rules are often regarded as derivative from broader principles (Solomon 1978). No single approach can be called the principles approach; hence, criticisms directed against one principle-based method may not apply to other such methods. For instance, criticisms aimed at a principle-based deontological theory may not apply to a principle-based consequentialist theory. Sometimes the language of ‘principles’ is mistakenly restricted to deontological theories, that is, theories holding that some inherent or intrinsic features of acts, such as truthfulness or lying, make them right or wrong. This restriction is misleading because consequentialist theories, which focus on the probable effects of actions, may also be principle-based. For example, utilitarianism, the most prominent contemporary consequentialist theory, appeals to the principle of utility in assessing acts or rules.

Consequentialist Principles Most consequentialists focus on both the intended ends and anticipated effects of actions but consider those intended ends only in relation to the action’s probable overall effects. Many, perhaps most, contemporary consequentialists are also utilitarians: the principle of utility provides the fundamental point of reference for their assessment of actions. This principle—in its simplest form, the requirement

18

james f. childress

to produce the greatest good for the greatest number—presupposes some values for the evaluation of states of affairs that might result from different actions. Prominent values include pleasure, happiness, and individual preferences, but a variety of other values can be employed. Whatever the locus and range of values, consequentialists take different tracks, depending on whether they focus on the effects of particular actions (act-consequentialist methods) or on the effects of types of action and hence rules governing those types of action (rule-consequentialist methods).

Act-Consequentialist Methods Contemporary act-consequentialists analyze the probable consequences of different courses of action and assess those actions according to their probable balance of good effects over bad effects. Many act-consequentialists are utilitarians. Act-utilitarians doing bioethics, such as the late Joseph Fletcher (1966) or Peter Singer (1993), apply the principle of utility directly to different possible acts in a situation to determine which would probably produce the greatest good for the greatest number; that act is then right and even obligatory. Act-utilitarians may face uncertainties about which course of action would satisfy the principle of utility, but they never face moral dilemmas created by conflicting principles. No moral dilemmas arise because act-utilitarians recognize only one principle (utility) as binding and view other principles and rules as mere maxims or rules of thumb that may usefully summarize agents’ experiences in the application of the principle of utility. Such principles or rules can help agents see the tendencies of different acts to produce good or bad consequences, but they lack prescriptive power. In short, the principle of utility binds, while other principles and rules, as generalizations based on past experience, only illuminate decisions. From the standpoint of the act-utilitarian, both rule-utilitarians and rule-deontologists are more alike than different: both make too much of principles and rules (other than utility) and too little of the consequences of particular acts. Such principles and rules create victims: people suffer bad consequences as a result of others’ adherence to principles and rules (other than utility). It is not surprising then that act-utilitarians often support changes in laws, policies, and practices to allow agents to assist in suicide or to engage in voluntary active euthanasia because such actions can in some circumstances relieve patients’ pain and suffering (Singer 1993). Critics contend, among their other charges, that act-consequentialists, including act-utilitarians, fail to attend to the necessity of principles and rules to solve or at least to reduce problems of coordination, cooperation, and trust in human interactions. For example, G. J. Warnock (1971) considers the expectations that would be appropriate in a clinical encounter between an act-utilitarian physician and a patient. He notes that the patient could only expect the physician to attempt to cure him of his afflictions ‘unless his [the physician’s] assessment of the ‘‘general happiness’’ leads him to do otherwise’ (Warnock 1971: 33). Asking the act-utilitarian physician

methods in bioethics

19

to declare his intentions truthfully or to promise to consider only the patient’s welfare, in accord with the Hippocratic tradition, would not be helpful because all the physician’s declarations and promises would themselves be subject to utilitarian calculation. Trusting that the act-utilitarian physician would act only or even primarily in the patient’s best interest would be unwarranted because such a physician must always, in every situation, attend to the greatest good of the greatest number. Such problems of trust, coordination, and cooperation lead many utilitarians and other consequentialists to focus on rules rather than only on particular acts.

Rule-Consequentialist Methods Many rule-consequentialists, as already suggested, are utilitarians, but some have a different and larger set of considerations than happiness, pleasure, or individual preferences for evaluating the consequences of various actual and proposed moral rules. Brad Hooker (2002) provides a major recent example of a well-developed ruleconsequentialist approach. He contends that the ethical assessment of acts should be based on rules that can be justified impartially and that impartial justification occurs if and only if the reasonably expected overall value of the general internalization of those rules by the overwhelming majority is greater than the reasonably expected overall value of any alternative rules. In short, we have impartially justified rules if those public rules internalized by ‘the overwhelming majority in each new generation’ would have the greatest expected value and their implementation would be cost-effective (including the costs of internalization) (Hooker 2002). This method employs rules that are justified by their anticipated overall consequences if implemented. Clearly this method presupposes a different moral psychology than the one we find in act-consequentialism. Hooker contends that his version of rule-consequentialism does not collapse into act-consequentialism because the moral agent aims at justifiable rules rather than at overall maximization of the good—the rules are designed to achieve that end. Furthermore, his moral psychology recognizes that human agents are susceptible to both cognitive errors and affective distortions, including impure motivations, and these points enter into his view of different rules (Hooker 2002: 187). Several questions arise. Questions about the relevant values plague consequentialist theories. Despite the apparent simplicity of consequentialism, the debates about values are as serious as the debates about principles in deontological approaches. For Hooker (2002), the relevant value is objective well-being, and he adds distributional patterns that assign some priority to the worst off. In addition, questions arise about specification: What counts as an ‘overwhelming majority’? How do we choose between actual rules that function fairly well and possible rules whose effects we may not be able to predict with great certainty? And what evidence is required for determining the probable balance of expected value and disvalue of possible rules?

20

james f. childress

One of Hooker’s main examples concerns a rule that would permit euthanasia under certain conditions. Such a rule might appear to be prima facie warranted because it would respect autonomous choices and reduce suffering—both important consequences. Whether, however, it would be ethically justified overall depends on a reasonable prediction and assessment of its total impact if it were internalized by an ‘overwhelming majority’. Major concerns focus on whether allowing euthanasia would be accompanied by an unacceptably high level of intentional abuse and would erode communal inhibitions on killing the unconsenting innocent. Hooker recognizes that any answers to these questions must be at least ‘partly speculative’ (Hooker 2002: 186). The best judgment must be based not only on the prediction and assessment of probable outcomes of a permissive rule but also on an appreciation of the costs—in suffering and disrespect for autonomous choices—of current prohibitive rules (Hooker 2002: 187). In view of the facts of human psychology—susceptibility to both cognitive errors and affective distortions, including impure motivations—Hooker finds good reasons for imposing and enforcing ‘tight restrictions on the use of euthanasia’ (2002: 187). Then he concludes: ‘With rigorously enforced restrictions, a rule allowing euthanasia, even active euthanasia, has (I believe) greater expected value than a complete ban’ (2002: 187). In response to those who might suppose that empirical evidence from the experiment with euthanasia in the Netherlands would be helpful and perhaps even decisive, Hooker stresses the divergent interpretations and evaluations of that experiment, most of which reflect different ethical viewpoints (Hooker 2002: 187). (For a rule-utilitarian argument for not treating and even for ‘active termination by anesthetic’ of certain ‘defective [sic] newborns’, see Brandt 1992.)

Deontological Principles Deontological approaches are usually contrasted with consequentialist ones. At a minimum, deontologists hold (1) that some features of actions other than or in addition to their consequences make those actions right or wrong, obligatory, or optional, and (2) that deontological considerations always, generally, or sometimes trump consequentialist considerations. Even if a bioethical theory recognized (1) but held that consequentialist considerations always triumph in a conflict, that theory would not be considered deontological. While the label ‘Kantian’ is now often used rather than ‘deontologist’ for such positions, that common usage is more likely to distort than to illuminate. To be sure, Immanuel Kant and later interpreters of his ethical theory have greatly influenced modern bioethics, but the sources of deontological theories are broader, and few contemporary bioethical approaches are Kantian in a strict sense even if they draw

methods in bioethics

21

on some Kantian language and themes, such as human dignity, autonomy, and respect for persons. Onora O’Neill (2002) is a distinguished exception. Indeed, deontologists in bioethics are as likely to draw on religious sources, texts, and traditions as on Kant and the Kantian tradition. Some contemporary deontological approaches to bioethics are secular in nature, even if partially inspired by religious perspectives. For instance, Robert M. Veatch (1981, 1995), whose position I will discuss later as a pluralistic principle-based approach with a lexical or rank order, identifies several deontological principles (veracity, fidelity to promises, avoidance of killing, autonomy, and justice), which collectively take priority over consequentialist principles. Hence, he satisfies both conditions for a deontological theory. In addition, Kevin Wildes (2000) affirms deontological principles, based on natural law and reason, but he also notes the absence of the epistemological conditions for securing agreement on those principles in our pluralistic society. In contemporary debates, two major deontological principles are often viewed as competitors: sanctity of life and respect for autonomy. They may point in different directions in debates about assisted suicide and active euthanasia. Defenders of a principle of sanctity of life usually oppose those acts and any moral, professional, or legal rules that would allow those acts, while defenders of respect for personal autonomy often (but by no means always) recognize the moral rights of individuals to choose suicide or euthanasia, and of others, including health care professionals, to assist them in committing suicide or to carry out the request for euthanasia. Hence, two prominent examples of deontological principles in contemporary bioethics appear on the side of libertarians—for example, the earlier Engelhardt (1986)—and on the side of religious thinkers with a commitment to the sanctity of life—for example, Pope John Paul II (1995) and Paul Ramsey (1970, 1978). In the next section, on pluralistic principlism, I will examine the various moves that deontologists make. There I will explicate principle-based approaches that recognize at least one deontological principle and at least one consequentialist principle and then have to connect those principles to concrete cases through such maneuvers as application and deduction, specification, or balancing.

Principlist Approaches Pluralistic Principlism The term ‘principlism’ was coined by critics (Clouser and Gert 1990) to designate and disparage a particular principle-based approach to bioethics, especially the one associated with Principles of Biomedical Ethics by Tom L. Beauchamp and James F. Childress (1979–2001). Even though Clouser and Gert would oppose any

22

james f. childress

principle-based framework, for reasons that will be clear later, their main target, designated ‘principlism’, is an ethical framework that employs several general and unranked principles that are all prima facie binding. A broader interpretation of principlism would encompass a wide variety of positions that appeal to moral principles, rules, and other guidelines, including some positions that rank principles and rules. Many principle-based approaches incorporate consequentialist principles along with non-consequentialist ones without deriving one set from or reducing it to the other; for instance, a principle of utility may be included on equal footing with principles of respect for autonomy and justice. Such approaches can be called pluralistic. In this chapter, then, ‘principlism’ will refer to pluralistic principlebased positions that recognize both deontological and consequentialist principles and hence have to develop ways to address actual and potential conflicts within and between these two kinds of principles, particularly in moving from principles to cases in order to guide action. Principlists differ greatly in the principles they affirm, how they justify those principles, and how they connect them to concrete cases. In Principles of Biomedical Ethics (5th edn., 2001, as well as earlier editions), Beauchamp and Childress identify four primary principles—respect for autonomy, non-maleficence, beneficence (including utility), and justice—and several derivative rules—veracity, fidelity, privacy, and confidentiality—along with various other rules, such as informed consent. Others have called this framework the ‘four principles approach’ (Gillon 1994; Gillon and Lloyd 1994). These principles and derivative rules are all prima facie binding; that is, they are binding other things being equal, but each can be outweighed in a particular context by another principle or rule. However, the principles’ different weights cannot be assigned in advance; they can only be determined in particular contexts in addressing cases or policies. In a similar vein, the influential National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research (1978) identified three unranked, basic principles that should govern research involving human subjects: beneficence (which includes non-maleficence), respect for persons (which includes respect for autonomous choices and protection of persons who are non-autonomous), and justice—principles that still guide federally funded research in the United States (see also Childress et al. 2005). As previously noted, Robert Veatch, in A Theory of Medical Ethics (1981) and subsequent works (e.g. 1995), offers a different list, but one that has substantial overlap with the above approaches: beneficence, contract-keeping, autonomy, honesty, avoiding killing, and justice. He also recognizes several moral rules, such as informed consent. The major difference between Veatch’s principle-based method and the two already mentioned in this section is that he offers a rank order of his principles. (I will later examine Veatch’s method of lexical ranking for resolving conflicts among these principles and rules.)

methods in bioethics

23

Rule-Based Critique One major type of criticism of principlism holds that broad, general moral principles—of the sort associated with Beauchamp and Childress’s work—are unnecessary because specific moral rules, such as non-deception, and moral ideals, such as going out of the way to help others, cover the whole moral terrain; that principles are too general and vague to guide action; and that, in the absence of a strong theory, it is not possible to resolve conflicts among principles (Clouser and Gert 1990). This critique, which can be described as a rule-based (or strong theory-based) approach, attacks any principle-based approach. Clouser and Gert write: ‘for all practical and theoretical purposes, there are no moral principles’, and ‘Our quarrel is not so much with the content of the various ‘‘principles’’ as it is with the use of ‘‘principles’’ at all’ (Clouser and Gert 1990: 235, 220). Nevertheless, they mainly target a pluralistic principlist approach that does not rank principles or provide a clear-cut decision procedure for resolving actual or potential conflicts. Regarding the charge that moral principles are unnecessary and moral rules are sufficient, principlists could respond that they too recognize moral rules, often derived from their principles; for instance, rules of voluntary, informed consent can be derived from the principle of respect for persons or respect for autonomy. However, they doubt that all that is important in moral principles can be fully captured in specific rules, such as the ones proposed by Gert and colleagues: don’t kill; don’t cause pain; don’t disable; don’t deprive of freedom; don’t deprive of pleasure; don’t deceive; keep your promise; don’t cheat; obey the law; and do your duty (Gert et al. 1997). These rules largely specify the harms that are to be avoided under the general requirement—what others might call a principle—of nonmaleficence, but they do not adequately express what many principlists construe as obligations of respect for autonomy, beneficence, and justice. Since space does not permit a full discussion, a single example must suffice. Elsewhere I have argued (Childress 1982) that neglect of a principle of respect for autonomy leads Gert and his colleagues to bizarre interpretations of particular cases, especially in their discussions of paternalism, as is evident in their reasoning about a particular case: Following a serious accident, a patient while still competent refuses a blood transfusion on religious grounds but then he falls unconscious and his physicians believe he will die unless he receives a transfusion. Gert and Charles Culver (1979) argue that the physician’s provision of a blood transfusion under these circumstances would be paternalistic because, after the patient regains consciousness, it would lead to a violation of either the moral rule against deception or the moral rule against causing pain. If the physicians fail to tell the patient, they would violate the moral rule against deception; if they tell him, they would violate the moral rule against causing pain. Gert and Culver are forced to take such a circuitous and problematic path of moral analysis precisely because they neglect the principle of respect for autonomy.

24

james f. childress

They fail to see that the reason the transfusion is paternalistic, and prima facie wrong, is that it violates the patient’s autonomously expressed wishes and choices in order to provide a medical benefit (Childress 1982). As a result, the Gert–Culver analysis implies that if the patient dies without regaining consciousness, no paternalistic or wrongful act has been committed, because no moral rules have been violated. The content of the moral rules proposed by Gert and colleagues does not adequately express the principle of respect for persons and their autonomy; such moral rules as ‘do not deprive of freedom’ will not cover all the critical cases involving self-determination. Nor will their subsequent expansion of this moral rule adequately address this problem; the later expansion appears in the interpretation added in parentheses after ‘Do not deprive of freedom (includes freedom to act and from being acted on)’ (Gert et al. 1996: 44). Although this expansion comes closer to addressing the problem created by the narrow rule, it does so only by building in enough to bring the rule closer to a principle, and clarity would be served by explicit recognition of a principle of respect for autonomy. Regarding the charge that principles are too vague to guide action, principlists can respond that they too seek to specify their principles in rules that move closer to concrete action guidance. And they attempt to resolve conflicts in various ways, for instance, through lexical ordering, specification, and constrained balancing—all methods that will be examined in the next section on ‘From Principles to Cases’. Even so, many principlists, especially those who seek ‘reflective equilibrium’ in a coherentist strategy, will not satisfy the Gert et al. requirement of a strong and unified theory: ‘the value of using a single unified moral theory to deal with the ethical issues that arise in medicine and all other fields, is that it provides a single clear, coherent, and comprehensive decision procedure for arriving at answers’ (Clouser and Gert 1990: 233; for an examination of reflective equilibrium, see Chapter 2, by John Arras, in this volume). For many principlists, including this one, that goal is unrealistic and efforts to realize it are misguided and ultimately distort morality and moral decision-making.

From Principles to Cases Moral principles require interpretation because they are indeterminate. It is often unclear whether a case falls under a principle or rule; a single principle or rule may point in two different directions in the same situation; and there may be apparent or real conflicts among principles and rules—perhaps even dilemmas as well as interpersonal conflicts. Henry Richardson (1990) has identified three models of connection between principles and cases: (1) application, which involves the deductive application of principles and rules, (2) balancing, which depends on intuitive weighing, and (3) specification, which proceeds by ‘qualitatively tailoring our norms to cases’. I will use these three models for analytic purposes (without

methods in bioethics

25

considering and assessing Richardson’s constructive argument). I will indicate how these models become even more illuminating when we consider two dimensions of moral principles (and rules): (a) their meaning, range, and scope of application (e.g. broad or narrow), and (b) their weight, strength, and stringency (e.g. absolute, prima facie, or suggestive). These two dimensions are closely related rather than totally independent. For instance, if we interpret ‘lying’ as ‘any intentionally false or deceptive statement’, it is utterly implausible to defend an absolute rule against lying. However, if we understand ‘lying’ as ‘deliberately deceiving someone who has a right to the truth’—another definition commonly offered for lying—it is not implausible to hold that a rule against lying, with this meaning, range, and scope, is absolute. But then all the important moral analysis in conflict cases would center on who has a right to the truth in particular circumstances.

Applying Principles As I noted at the outset, bioethicists now generally eschew or at least criticize the language of application. This point holds for most principlists as well as for their critics, even though act-utilitarians may reasonably be viewed as applying the principle of utility. While there may be some genuine applications of principles to concrete cases, the language of application does not cover all or even the most significant connections between principles and particular judgments about cases. Not all such connections involve rational deduction of particular case judgments from general moral principles. On the one hand—as will be developed further in conjunction with casuistry—most principlists concede that particular case judgments are often made in relative independence of general moral principles. We often know what we ought to do without explicit reference to general moral principles even if we could articulate such principles if we were challenged to do so. On the other hand, principlists can and should recognize that judgments about particular cases can and sometimes should also lead us to modify our general moral principles. The application framework can function effectively only where we can assume that (a) the principle’s scope and range of applicability can be firmly established, (b) the principle’s weight or strength can be established a priori, and (c) the principle will never come into conflict with other equally significant principles. In concrete cases, conflicts between moral principles—e.g. between benefiting patients and respecting their autonomous choices—generate moral perplexities that lead to adjustments in (a) or (b). In such situations, we often proceed by specifying or balancing the principles in conflict. Specifying principles is a way to try to reduce or eliminate the conflict; balancing principles is an effort to resolve the conflict through determining which principle outweighs the other in the circumstances.

26

james f. childress

Some bioethicists seek to resolve conflicts not by specifying or balancing moral principles but by arranging them in some a priori order. This represents yet another version of the application–deduction framework for connecting principles to concrete cases. For instance, a bioethical framework might hold that a rule against killing always outweighs a principle of beneficence or—in an example that combines specification with a rank order—that the principle of respect for autonomy always overrides the principle of beneficence when the only beneficiary is the competent person whose wishes, choices, or actions are at stake. Much of Robert Veatch’s work, including A Theory of Medical Ethics (1981), attempts to find a defensible lexical ordering of principles, which can provide a framework within which some balancing can occur. Focusing on what he considers to be the relatively neglected but critical task of resolving conflict among competing principles, he proposes a ‘mixed strategy’ (Veatch 1995). This ‘mixed strategy’ presupposes a sharp distinction and separation between consequentialist principles (beneficence and non-maleficence, which are consequence-maximizing) and non-consequentialist principles (veracity, fidelity to promises, avoidance of killing, autonomy, and justice). While moral agents can balance various consequentialist principles and various non-consequentialist principles, they cannot balance consequentialist principles against non-consequentialist ones. Instead, there is a lexical rank order: the balanced non-consequentialist principles are lexically ranked over the balanced consequentialist principles (Veatch 1995). Such an application–deduction framework is subject to the same challenges that confront other application–deduction frameworks: plausible counterexamples.

Specifying Principles Specifying general moral principles is another way to connect them to particular cases. This process of specification presupposes a distinction between general and specific and between degrees of generality and specificity. However general our moral principles—for example, respect for personal autonomy or justice—we interpret them in part by formulating them more specifically or by delineating the types of cases that we believe fall under them. This process is inevitable because, as R. M. Hare observes, ‘any attempt to give content to a principle involves specifying the cases that are to fall under it . . . . Any principle, then, which has content goes some way down the path of specificity’ (Hare 1989: 59). Indeed, drawing on the distinction between principles and rules, it is plausible to view many rules in bioethics as specifications of broad principles so they can function as concrete action guides. For instance, rules of voluntary, informed consent specify the requirements of the principle of respect for persons and their autonomous choices. Specifying ethical principles is often indispensable and valuable for action guidance, especially but not only in conflict situations. In face of apparent moral dilemmas, specification always merits a trial to see if the conflict can be avoided,

methods in bioethics

27

eliminated, or reduced. How far the method of specification can be expected to succeed will depend in part on fundamental beliefs about the moral universe, including harmony and conflict within that universe. It is important to note that specification proceeds by restricting the range or scope of a principle’s applicability, rather than by adjusting its weight or strength. However, one effect of narrowing a principle’s range or scope of applicability may be to increase its weight or strength within that narrower range or scope. A good example of specification appears in the way Roman Catholic moral theology has over time interpreted the Decalogue’s prohibition of killing. This use of specification is complex because it is combined, as specification often is, with application and deduction and with absolutism: ‘from the beginning, faced with the many and often tragic cases which occur in the life of individuals and society, Christian reflection has sought a fuller and deeper understanding of what God’s commandment prohibits and prescribes’ (John Paul II 1995, para. 55). Specification was undertaken in light of what the Church deemed to be fundamental and unchanging beliefs and values—human beings as created in God’s image and God’s sovereignty over human life. In this light, the Church determined the precept’s meaning by restricting its range and scope of application in at least two ways: first, to innocent human life; second, to direct action. Hence, given the first restriction, killing in self-defense, warfare, and capital punishment could be justified. Given the second restriction, a few cases of indirect fetal death (e.g. in the case of ectopic pregnancy) and some cases of letting patients die could be justified. Specification has been praised as a way to reduce the role of intuition in concrete decisions (Richardson 1990), but there is debate about how far it can actually succeed in doing so. Some critics suggest that balancing occurs in the very process of specification or at least that specification falls prey to the same problems that many see in balancing. As John Arras asks: ‘what motivates and guides the modification and specification of abstract principles, what compels one to lard them with qualifying clauses, if not precisely the sort of countervailing values and principles encountered by the principlist [engaged in balancing]?’ (Arras 1994: 997). Specification thus may be as arbitrary as intuitive balancing in cases of conflict, especially in the absence of controls over the interpretation of key moral categories, such as ‘lying’ or ‘killing’. Some specification, as noted earlier, is unavoidable as part of the process of giving concrete content to broad, abstract principles, and it enables principlists to avoid or rebut some of the charges leveled by the strong rule critics. Indeed, specification generates rules. In addition, in conflict situations, specification, as developed by Richardson (1990), and extended by David DeGrazia (1992) in ‘specified principlism’, is promising. Beauchamp and I have also featured this strategy in later editions of Principles of Biomedical Ethics (1994, 2001; for a critique, see Strong 2000, with a response by Beauchamp 2000). However, as suggested above, specification’s overall value in connecting principles and particular judgments in

28

james f. childress

situations of apparent conflict will depend in part on more fundamental beliefs about whether moral conflict, not only between people but also within the moral universe, is inevitable, since the goal of specification is to reduce if not eliminate those conflicts.

Balancing Principles While specification addresses the meaning, range, and scope of principles, a second question also arises: How much weight, or what degree of stringency, do different principles have in relation to each other if they come into conflict? If moral principles are more than merely suggestive, illuminative, or advisory, then it is important to determine just how binding they are in order to address any conflicts that emerge. Absolutism presents one extreme possibility: absolutists maintain that some moral principles and rules are absolutely binding whatever circumstances arise. But they face irresolvable moral dilemmas if they recognize more than one absolute principle and those principles come into conflict. As a result, absolutists often carefully specify the meaning, range, and scope of their principles in order to avoid such conflicts (see Ramsey 1968, 1970). Another possibility—discussed above under application—is to try to establish a lexical or rank order that would itself be absolute. However, this approach too seems implausible in the face of counterexamples that appear to constitute legitimate exceptions: it is hard to establish an absolute priority for all cases. A single-principle approach could be absolutist, as act-utilitarianism is, but then it will be forced to view all other principles (and rules) as mere maxims or rules of thumb with illuminative but no prescriptive power. Yet another approach—one often associated with principlism in its narrow sense—views moral principles as prima facie or presumptively binding, rather than as absolutely binding or lexically ordered (Beauchamp and Childress 2001). It thus balances various principles when they come into conflict in particular cases, if the process of specifying the principles does not eliminate the conflict. An act is morally right or obligatory in so far as it has the features that, according to the relevant principles, establish moral rightness or obligatoriness. For example, an act is right in so far as it is truthful, wrong in so far as it is a lie. However, a particular act, in particular circumstances, may have features that express some principles while contravening others; for instance, a truthful act may also be unjust or cause harm. In such a case, the agent must determine whether one principle or the other is weightier or stronger, a judgment that cannot be made on the basis of prior, abstract formulations. In part to reduce (but not eliminate) the role of intuition, Beauchamp and Childress (2001) offer several conditions for restricting or constraining judgments about balancing conflicting prima facie moral principles. These conditions are:

methods in bioethics

29

1 Better reasons can be offered to act on the overriding norm than on the infringed norm . . . 2 The moral objective justifying the infringement must have a realistic prospect of achievement. 3 The infringement is necessary in that no morally preferable alternative actions can be substituted. 4 The infringement selected must be the least possible infringement commensurate with achieving the primary goal of the action. 5 The agent must seek to minimize any negative effects of the infringement. 6 The agent must act impartially in regard to all affected parties . . . (Beauchamp and Childress 2001: 19–20)

These conditions or constraints express the logic of prima facie principles. Prima facie principles are weightier or stronger than mere maxims or rules of thumb, but lighter or weaker than absolute principles. They are morally binding other things being equal. Hence, in a case of conflict, infringing or overriding one prima facie principle in order to protect another prima facie principle requires attention to various circumstances, including the often overlooked or neglected conditions of necessity, or last resort, and least infringement. Nevertheless, according to its critics, this form of principlism still relies excessively on intuition, whatever procedures it introduces to reduce the role of intuition in the process of balancing its various principles and rules. Instead of viewing application, balancing, and specification as three mutually exclusive models, it is better, I believe, to recognize that all three are important in parts of morality and for different situations or aspects of situations, as well as often intertwined and overlapping. Sometimes principles (and rules) can be applied, and sometimes they need to be specified, but at times conflicts may emerge that can only be resolved by (constrained) balancing. It is a substantive, and not merely a formal, moral debate as to which method works where and when.

Case-Based Methods ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

The Gert–Clouser critique, as we have seen, faults principlism for failing to develop a rationalistic, unified theory with concrete directives for action (in the form of rules). According to this critique, principlism is insufficiently ‘top-down’ and deductivist. By contrast, proponents of case-based methods charge that principlism is too ‘top-heavy’ and ‘top-down’ because it fails to attend sufficiently to particular judgments about cases. Case-based methods involve a ‘bottom-up’, inductive approach to ethical justification. Casuistry, as one case-based approach, can be defined as ‘a method for arriving at justifiable decisions about what to do in specific cases’ (Strong 2000). There are several important case-based and casuistical methods (Brody 1988, 2003; Strong 2000; Kuczewski 1997), but I will concentrate

30

james f. childress

on the casuistical approach offered by Albert Jonsen and Stephen Toulmin in their influential book The Abuse of Casuistry (1988), and in their various individually authored articles. The Abuse of Casuistry claims, among other points, (1) that ‘casuistry is unavoidable’, (2a) that ‘moral knowledge is essentially particular, [(2b)] so that sound resolutions of moral problems must always be rooted in a concrete understanding of specific cases and circumstances’, and (3) that moral reasoning proceeds by paradigm cases (moral precedents) and analogical reasoning (Jonsen and Toulmin 1988: 330). The first point (1) is relatively uncontroversial. While the essential particularity of moral knowledge (2a) can be interpreted in several different ways, major qualifications are necessary on any interpretation; however, (2b) is defensible. Finally, while identifying an important feature of moral reasoning, the last claim (3) tends to exaggerate the role of this mode of reasoning to the neglect of principle-based reasoning. In explicating and justifying claim (2a) about the essential particularity of moral knowledge, The Abuse of Casuistry vigorously resists what Toulmin had earlier called ‘the tyranny of principles’ (Toulmin 1981). This rhetoric is exaggerated, because, in fact, Jonsen and Toulmin have as their target not the inevitable tyranny of any and all principles but only the tyranny of certain conceptions of principles, that is, ‘eternal, invariable principles, the practical implications of which can be free of exceptions or qualifications’ (Jonsen and Toulmin 1988: 2). They contend that such principles lead to deadlocks and fruitless standoffs, among other problems. In short, this casuistry-based attack on tyrannical principle-based methods focuses on absolutist versions, rather than on versions that view principles and rules as prima facie binding and require specification, balancing, and other modes of interpretation in situations of decision. This rhetoric is also overblown because Jonsen and Toulmin themselves do, in fact, recognize principles. At one point, they state that their aim is to argue for ‘good casuistry’, that is, casuistry ‘which applies general rules to particular cases with discernment [in contrast to] bad casuistry, which does the same thing sloppily’ (Jonsen and Toulmin 1988: 16). And elsewhere Toulmin (1981) concedes that principles have special relevance and importance in relations between strangers rather than intimates. If so, then one important question is how we can best characterize relations in research and also in medicine and health care—they are often relations between strangers rather than relations between intimates. In view of their nod to principles, how should we understand and assess Jonsen and Toulmin’s claims about the primacy or priority of particular judgments? On the one hand, such a claim might mean that particular judgments came first chronologically and gave rise to general judgments (principles). Even if that claim could be established—a difficult and perhaps impossible task—its implications may be quite limited for the ways we are now acculturated and reason ethically. We all participate in communities of moral discourse that embody and extend traditions of

methods in bioethics

31

moral reflection. These traditions incorporate both general principles and particular case judgments, including paradigm cases, and these communities build both into their formal and informal moral instruction. Indeed, moral education is virtually unimaginable without general principles and rules as well as paradigm or precedent cases that help specify their content. On the other hand, claims about the primacy of particular judgments could refer to logical or normative priority of case judgments over principles. However, the relation between particular judgments and general judgments (principles) is better viewed as dialectical, with neither fully and completely derived from the other and each potentially modifying the other. Where there are conflicts, adjustments may need to be made in either the particular or the general judgments (principles), but there is no justification for insisting that either one should always take priority. Indeed, what is sought is a kind of ‘reflective equilibrium’ within a coherentist strategy (see Arras 1991; Beauchamp and Childress 2001). Suppose that, upon reflection, we make a judgment that an action in a particular set of circumstances is wrong, without regard to general moral principles. What are the implications of this particular judgment, if we take seriously both ethical consistency and universalizability, which entail treating similar cases in a similar way? If we have made a judgment about a case—that X is wrong or right—then, R. M. Hare notes, ‘we have acquired a precept or principle which has application in all similar cases. We have, in some sense of that word, learnt something’ (Hare 1989: 55–6). And if we learn something useful from reflection on a particular case, the principle we gain, Hare contends, must be somewhat general rather than having unlimited specificity. Since no two real cases are exactly alike, the results of reflection can be useful in the future only if we ‘have isolated certain broad features of the cases we were thinking about—features which may recur in other cases’ (Hare 1989: 56; italics added). To isolate these ‘broad features’ is to identify a principle or rule (for the principlist) and a paradigm or precedent case (for the casuist). Casuists also make another important claim ( (3) above) about their inductive method of moral reasoning and justification: Practical reasoning proceeds by analogy from paradigm cases, i.e. settled cases or precedents, to new or unsettled ones. Such reasoning is widespread in bioethics, among principlists as well as casuists. Hence, it is not surprising, for example, that much of the ethical debate about a controversial case of research involving human subjects will reason analogically in relation to the negative paradigm case of the Tuskegee syphilis study in which close to 400 African American men were left untreated for syphilis for decades so that researchers could study the history of untreated syphilis in the African American male. If a current case is relevantly similar to the Tuskegee case, then, on grounds of ethical consistency and universalizability, we are committed to judging the current case as wrong. Such an analogical mode of moral reasoning can and often will be illuminating.

32

james f. childress

The principlist could agree but could also argue that cases such as Tuskegee are paradigms or precedents for moral reasoning precisely because of the principles they embody—for instance, the Tuskegee syphilis study became a negative paradigm case because it violated fundamental principles of justice, respect for autonomy, beneficence, and non-maleficence. The differences between principles, on the one hand, and paradigms and precedents, on the other, appears to be minimal, especially because each identifies ‘broad features’ of types of cases. They usually differ in degree of generality or specificity—the paradigm or precedent case may have more details and greater specificity. However, to use a legal analogy, even with the rich, specific details of the case, it may still be necessary to identify the ‘holding’ in the case, and this will bring us closer to a principle or principles. Furthermore, some of the maxims that casuists see embedded in cases appear to function as mid-level principles or rules. Other important—but often neglected—questions for casuists (and for others) concern cases and their descriptions: How are cases identified and labeled? How does the casuist determine what kinds of cases they are? These are necessary first steps in the process of taxonomic analysis. However, identifying and labeling cases often seems to be more implicit than explicit, more intuitive than reasoned, with insufficient attention to the process of ‘evaluative description’, which is perhaps the most adequate way to conceive case presentations. Narrative analysts who have turned their spotlight on bioethical cases, as mini-narratives, have directed our attention to the evaluative and other assumptions that often structure cases and lead to both classifications and conclusions that may not be adequately examined or warranted (Chambers 1999). Consider two evaluative descriptions of cases, the first relatively uncontroversial and the second quite controversial. First, several years ago the Journal of the American Medical Association (JAMA 1988) reported a case under the title ‘It’s Over Debbie’. In this case, the authenticity of which has been questioned, a medical resident injects a terminally ill woman with enough morphine to end her life in response to her request, uttered in their first encounter, ‘Let’s get this over with’. The casuist Jonsen classifies this case as one of killing, hence bringing it under a taxonomy of cases of killing, governed by various maxims, and then reasons analogically in relation to paradigm cases in this taxonomy (Jonsen 1991). Jonsen’s case description, classification, and analysis appear to be quite straightforward in this case, but conflicts can emerge about the evaluative descriptions of cases according to their type and classification, as is evident in the next example. A few years ago a clinical case involving the disconnection of a ventilator maintaining the life of a patient with amyotrophic lateral sclerosis (Lou Gehrig’s disease) was presented as an end-of-life case, in which the ‘patient’ decided to discontinue the ventilator. However, members of the audience, many of whom had themselves experienced long-term ventilator use, disputed this classification, viewing the case instead as a disability case in which the clinicians should have provided better care,

methods in bioethics

33

fuller information, and more options to the ‘consumer’, particularly after the recent death of his spouse (Kaufert and Koch 2003). Interpreters contend that this conflict illustrates the importance of narrative analysis, which examines the assumptions and perspectives that operate in the description of cases: ‘What to the clinicians was a textbook case of ‘‘end-of-life’’ decision making was, for their audience, a story in which a life was ended as a result of failures of information and assistance by the presenters themselves’ (Kaufer and Koch 2003: 462). At the very least, casuists (and others) need to pay more explicit attention to ‘moral diagnosis’ (Arras 1991) and hence to how they describe and frame cases. Of particular importance is the recognition and reduction of bias in ‘describing, framing, selecting and comparing of cases and paradigms’ (Kopelman 1994: 21). Bias reduction strategies, at a minimum, should include richer, fuller descriptions of cases, and the incorporation of a wide range of possible descriptions. Case-based and casuistical methods lack a clear way to identify relevant features of cases, in part because they lack content. According to some critics, casuistry is ‘a method without content. It is a tool of thought that displays the fundamental importance of case-comparison and analogy in moral thinking, but it lacks initial moral premises’ (Beauchamp and Childress 2001: 395). Or, stated differently, it is ‘more a method than a doctrine, more an engine of thought than a moral compass’ (Arras 1998: 112). Hence, the engine’s direction depends on the values—individual, professional, communal, etc.—casuists bring to bear on the case. If, of course, casuists were to identify the maxims in an array of cases, e.g. on killing, they would perhaps then have a framework of mid-level principles or rules that could be critically examined and employed in describing cases and recommending actions—but then it would be even harder to distinguish casuists from principlists. General principles may also provide ways to criticize practices that are not available in case-to-case analysis. This point may hold even if the principles themselves are discerned in practices, rather than established by an ethical theory. John Arras (1991) worries that, in the taxonomic approach of moving from actual case to actual case, the casuist may be limited to what practitioners and others present for ethical analysis and assessment because of felt problems or dilemmas. However, general principles may help us identify other cases that should be on the moral agenda because of some ‘broad features’, and they may direct our attention to real problems and dilemmas that have not yet been experienced as such. In addition, principles, such as justice, may help us identify and correct the distortions that our social structures, policies, and practices create in our perception of cases and our analogical reasoning. The contrasts between principle-based and case-based methods or between generalist and particularist approaches may be less significant and less illuminating if these methods gravitate more and more toward the middle as they mature (Sumner and Boyle 1996). Indeed, most debates about these methods and positions already tend to feature caricatures rather than real opponents. One question is whether

34

james f. childress

these methods become virtually identical or whether they are complementary. Even though casuists and principlists—in their best moments—attend to both principles and cases and seek rigorous and imaginative ways to relate them—for example, dialectically—the methods are better understood as complementary than, as Mark Kuczewski (1998: 521) claims, ‘largely the same method’.

Virtue Ethics ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

A modern alternative to principle-based (and rule-based) approaches appears in virtue ethics, which draws on ancient philosophical resources, especially Aristotle. In a sense, as Michael Slote (1997: 233) observes, virtue ethics has awakened from ‘a long slumber’, but, in view of its extended dormancy, it is, practically speaking, ‘the new kid on the block’ compared to consequentialist and deontological approaches. It (re)emerged as a critique of the inadequacies of various principle-based approaches, but has now morphed into a distinctive, constructive alternative. (Re)emerging as a critique, virtue ethics has routinely been characterized as a corrective. According to proponents of virtue ethics, principle-based as well as casebased methods—which from the standpoint of virtue ethics are more similar than different—tend to neglect moral character, discernment, motivation, emotions, etc. By contrast, virtue ethics attends to the agent rather than the act; to character rather than conduct; to what sort of persons we should become rather than what sorts of actions we should perform; and so forth. While partially illuminating the different positions, such characterizations and contrasts also inevitably oversimplify and distort them. Principle-based approaches—as well as casuistical approaches—usually recognize the importance of the virtues. While the list of specific virtues will vary from theory to theory and method to method, they all presuppose certain conceptions and kinds of virtues, as established traits of character, including motivation to act according to certain principles and rules or to act on casuistical judgments. Nevertheless, it is fair to say that neither principle-based nor case-based approaches have adequately attended to the virtues they presuppose for their own successful operation. Principle-based approaches tend to identify (at least some of) the relevant virtues by their correspondence to different principles—e.g. the virtue of benevolence corresponds to the principle of beneficence—or by their value for morality as a whole—e.g. courage (see Beauchamp and Childress 2001). From this standpoint, the virtues, as motivational structures and dispositions, are important because they enable agents to adhere to moral principles. As William Frankena put it, ‘principles without virtues are impotent’ (Frankena 1973: 36). There is thus little dispute that the virtues are important and even indispensable in ethics and in bioethics.

methods in bioethics

35

However, for principled and casuistical approaches, the virtues are important and indispensable in a secondary, derivative way—to provide motivation to realize what is already known independently through principle-based or case-based judgments. For principle-based and case-based methods, then, the virtues have no independent cognitive or normative significance for human action. Virtues do not guide action but, instead, motivate action. Indeed, emphasizing that virtues do not direct action, Frankena holds that ‘virtues without principles are blind’ (Frankena 1973: 36). In a similar vein, R. M. Hare claims: ‘It looks as if any ethics of virtue would have to borrow extensively from an ethics of principle in order even to tell us what virtue consists in’ (Hare 1996: 22). Given these points, it might be unclear just what virtue ethics can contribute to bioethics other than calling attention to neglected virtues as indispensable motivational resources for principle-based and case-based methods. One problem is that conceptions of the virtues as dispositions to act, in a oneto-one correspondence with independent principles and rules, reduce each of the virtues to what Hursthouse (2003) calls a ‘single track disposition’. However, such a reductionist conception distorts the virtues. For instance, the virtue of honesty cannot be reduced to a ‘single track disposition to do honest actions, or even honest actions for certain reasons’—instead, it is ‘multi-track’ because it denotes a ‘complex mindset’, consisting, in addition to honest actions, of other actions, such as disapproving of others’ dishonesty, along with various associated emotions, attitudes, perceptions, sensibilities, and the like (Hursthouse 2003). One important implication is that not even a wide range of agent X’s acts of honesty can warrant the judgment that X is honest, in part because X’s reasons (motives) are crucial to the judgment but also because honesty is a ‘multi-track disposition’ that may be embodied and expressed in degrees. While one set of action-guiding virtues—variously called practical wisdom, prudence, or discernment—functions in part through the interpretation and application of principles, it too is more complex. As Alisa Carse rightly argues, ‘recognizing that a general principle or rule is relevant to the situation at hand, and knowing how it is fittingly to be acted upon requires a capacity for discernment that is distinct from, and presupposed by, the application of principles themselves’ (Carse 1991: 11). She further observes that discerning responses are not always principle-driven or the result of principled deliberation; rather they may involve sensitivity to other people through a sympathetic attunement to their needs and concerns. Principle-based approaches need not and should not deny these points: principles do not exhaust the moral life and its decisions, and even when principles are relevant they must be discerningly interpreted and employed in the situation. Proponents of virtue ethics contend that virtues provide more action guidance than other theorists often recognize. A major misconception that pervades much of the contemporary debate about ethical theory and method, according to Rosalind Hursthouse, a virtue ethicist herself, is that ‘virtue ethics does not, and cannot,

36

james f. childress

provide action guidance, the way utilitarianism and deontology do’ (Hursthouse 2003). Against this misconception, she argues that virtue ethics provides a specification of ‘right actions’—as ‘what a virtuous agent would, characteristically, do in the circumstances’—and such a specification can be regarded as generating a number of moral rules or principles (contrary to the usual claim that virtue ethics does not come up with rules or principles). Each virtue generates an instruction—‘Do what is honest’, ‘Do what is charitable’; and each vice a prohibition—‘Do not . . . do what is dishonest, uncharitable’. (Hursthouse 2001: 17)

Despite such points about virtue ethics’ capacity for action guidance, critics still often view an examination of the virtues as a subject better suited for the novelist than for the ethicist, because they are so numerous, unspecifiable, and potentially in conflict without some way to arrange and order them. Hence, beyond claims that virtues lack independent action-guiding content, questions arise about whether the virtues can guide action because of their limited specifiability. As Justin Oakley observes, ‘a virtue ethics criterion of rightness is less precisely specifiable and less easily applicable’ than some other approaches (Oakley 1998: 94). The very nature of the virtues, according to Robert Louden, means that we can reasonably expect ‘a very limited amount of advice on moral quandaries . . . from the virtue-oriented approach. We ought, of course, to do what the virtuous person would do, but it is not always easy to fathom what the hypothetical moral exemplar would do were he in our shoes’ (Louden 1984: 229). In response to concerns about whether virtue ethics can guide decisions in conflict situations, Hursthouse argues that the goal should not be ‘to provide a decision procedure which any reasonably clever adolescent could apply [without moral wisdom or discernment]’ (Hursthouse 2001: 18). This is part of her—and virtue ethics’—rejection of the idea that ethics is ‘codifiable’. She also distinguishes ‘resolvable dilemmas’, or hard cases, from ‘irresolvable and tragic dilemmas’, noting that the virtue ethicist may try to specify the virtues and to distinguish merely apparent from real dilemmas—a task that presupposes moral wisdom and discernment (Hursthouse 2001). (As Hursthouse recognizes, some forms of principlism also distinguish the apparent from the real by further specifying principles to avoid or reduce the apparent conflict.) Sometimes, however, irresolvable and tragic dilemmas remain. Even though action guidance is impossible in such situations, virtue ethics can depict how the agent should respond, that is, with appropriate attitudes and emotions, such as distress and guilt. One critique of virtue ethics, leveled by a variety of principlists, is that good or virtuous people can act wrongly and do bad things (Hare 1996; Oakley 1998). Neither a settled motivational structure nor guidance from the virtues themselves, including prudence, may be sufficient to preclude wrong actions. Indeed, even if an agent’s motivation is virtuous, certain dominant virtues may lead him or her to act wrongly. For example, a caregiver’s virtue of benevolence may lead her to find

methods in bioethics

37

active euthanasia acceptable for a particular patient, even though that patient is not able to speak competently for himself. Virtue ethicists could respond that the virtues need to be considered as an integrated whole, as a unity, and that a virtuous person has discernment to draw appropriate lines. Furthermore, unless virtue ethics reduces all general standards of action to virtue standards—a move that is unnecessary and indefensible—then, as previously noted, virtue ethics can incorporate other principles of action guidance. Even though many theories have grounded the moral virtues in conceptions of human flourishing or fulfillment, those conceptions have floundered in the wake of the collapse of teleological notions of science and the prevalence of diverse conceptions of the good life. As a result, some virtue ethicists have resorted to traditions of particular communities—such as religious communities—which may embody a unified conception of human flourishing that can warrant a set of virtues. A promising alternative in bioethics grounds the virtues in particular professional traditions, such as medicine and nursing. In contrast to general ethics, Edmund Pellegrino argues that ‘professional ethics offers the possibility of some agreement on a telos—i.e. an end and a good’, which, in the case of the relationship between the health care professional and the patient, would be the good of the patient (Pellegrino 1995: 266). From the clinical relationship with its telos of healing, Pellegrino develops a list of virtues: fidelity to trust and promise, benevolence, effacement of self-interest, compassion and caring, intellectual honesty, justice, and prudence. According to Pellegrino, this list, while not exhaustive, identifies the virtues ‘most essential to the healing purposes of the clinical encounters’ (Pellegrino 1995: 268). In short, ‘they are ‘‘entailed’’ by the end of the healing relationship; that is, they are required if the end is to be attained’ (Pellegrino 1995: 268). These different virtues are mutually reinforcing—compromising one would compromise the others—such that no priority order is possible. Questions still arise about whether this list of virtues is sufficiently comprehensive. Critics note the absence of a virtue that would correspond to respect for patients’ autonomous choices and wonder whether Pellegrino’s virtue-based framework thus allows and supports excessive professional paternalism. In addition, the professional relationships and traditions, on which Pellegrino builds, have suffered, in recent decades, from widespread consumerism in market-driven interactions. Hence, it is unclear whether they can engender and sustain the virtues Pellegrino affirms. Virtue-based methods would approach questions about euthanasia in part by drawing out the implications of character and specific virtues in light of the telos that the particular method recognizes. For example, focusing on the telos of human flourishing, Philippa Foot examines the moral possibility of various kinds of euthanasia (passive or active and voluntary or non-voluntary) from the standpoint of the virtues, contending that charity as the virtue of attachment to the good of others can support euthanasia while justice sometimes opposes and sometimes

38

james f. childress

allows it (e.g. justice opposes non-voluntary active euthanasia, may accept nonvoluntary passive euthanasia, and supports active voluntary euthanasia) (Foot 1977: 106). However, in considering changes in legal and professional rules, Foot’s argument differs little from many rule-consequentialist arguments, in stressing the importance of keeping a ‘psychological barrier’ and the dangers of abuse as well as the alteration of social expectations about care of the sick and elderly (Foot 1977). By contrast, Pellegrino (1998) totally rejects active euthanasia, at least by physicians, because it is opposed to specific professional virtues in clinical relationships aimed at healing.

Ethics of Care ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Another approach that also seeks to provide an alternative to principle-based methods is an ethic of care or caring. It is similar in many respects to virtue ethics, with care or caring viewed as a specific virtue. After decades of research into human moral development, some psychologists noticed that principle-based approaches tend to echo male voices about male experiences, while neglecting women’s voices and experiences. They further observed that theories of moral development were based exclusively on male models. In her pioneering research along these lines, Carol Gilligan (1982) reported that women tend to concentrate on narratives, contexts, and relationships of care. By contrast, males tend to emphasize what Gilligan calls an ethic of justice, involving tiers of general moral principles and employing a logic of hierarchical justification. (These tendencies do not mark gender exclusivity—men may take a care perspective and women a justice perspective.) From a care perspective, the moral agent is relational and interdependent with others rather than an independent decider who applies abstract, general principles—even autonomy is often construed as ‘relational autonomy’ (Mackenzie and Stoljar 2000). Beyond studies of moral development, philosophers have highlighted a variety of features of a care perspective. Viewing an ethic of care as ‘a way of understanding one’s moral role, of looking at moral issues and coming to an accommodation in moral situations’, the philosopher Rita Manning finds in it five central ideas: moral attention, sympathetic understanding, relationship awareness, accommodation, and response (Manning 1998: 98). Another central theme is ‘appropriate trust’, which Annette Baier (1985, 1994) emphasizes. Given the focus on care as a moral sentiment and response in particular relationships, and its similarities to virtue ethics, it is not surprising that ‘care as a standard does not prescribe specific actions in the way an ethic of rules or decision procedures strives or claims to do’ (Blum 2001: 186). Methodologically, as Alisa Carse observes, ‘ ‘‘care’’ reasoning is concrete and contextual rather than abstract; it

methods in bioethics

39

is sometimes principle-guided, rather than always principle-driven, and it involves sympathy and compassion rather than dispassion’ (Carse 1991: 17). In so far as the ethics of care is contextual, it overlaps with casuistry. However, turning from principles to context, including relationships, still leaves open difficult questions about setting priorities when responsibilities to concrete others in particular relationships come into conflict with each other or with responsibilities to stakeholders outside those particular relationships. In many cases, moral agents have to determine, at least in part through principles and rules, how much weight to give to different relationships. For example, a physician may have to determine whether to breach confidentiality in order to warn a stranger of a patient’s threatened violence. An ethic of care appears to be best suited for intimate relations, including some in health care. However, it must also address the stranger, the person outside specific relationships. Indeed, a major challenge to an ethic of care focuses on its particularity and its partiality. The question is whether care, in many of its formulations, is too partial and parochial, without pressures generated by a more principled approach. For instance, care in particular relationships may need to be limited in strength or weight even as care is expanded in scope—i.e. rendered more general and universal—perhaps under a principle of beneficence (or virtue of benevolence) or a principle or virtue of justice. As Gilligan (1982) recognizes, care and principles (which she characterizes in terms of ‘justice’) are generally complementary rather than opposed. Even though much of the interest in and early work on an ethic of care emerged from feminist contexts, many feminists challenge such an ethic, especially if taken by itself without further attention to principles. Certainly principles of justice, particularly in the form of equality and impartiality, as well as respect for personal autonomy, strongly support often neglected women’s rights. In addition, some feminists distinguish a feminist approach to ethics from a feminine approach, associating an ethic of care with the latter. For example, some feminists, who take seriously the oppression of women, are suspicious of a feminine ethic of care that originates and operates under oppression (and thus lacks independent standing) and that may actually perpetuate oppression. Such an ethic may foster further oppression through the liberal division of private and public spheres, with care being deemed appropriate in the private sphere but not in the public sphere, and with women being viewed as the primary bearers of caring—for instance, in the care of elderly family members (Sherwin 1992). For others, the care perspective is potentially transformative, since the personal and the interpersonal are also publicly and politically significant. One possible formulation is ‘just care’ (see Manning 1998: 103–4 on care and justice). Just as a virtue ethic, an ethic of care can lead in different directions in debates about physician-assisted suicide or active euthanasia. At the very least, it would give priority to long-standing relationships in which care has been evident—for

40

james f. childress

instance, it would tend to find Dr Timothy Quill’s assistance to his patient Diane in her suicide much more acceptable than Dr Jack Kevorkian’s assistance in the suicide of persons whom he barely knew and had not cared for (Quill 1991). Still the appropriate, fitting, responsible response in such situations will depend greatly on what is built into the notion of care and its limits. In the final analysis, the care perspective offers an important corrective to some principle-based approaches by attending to context, narrative, relationships, emotion, compassion, and the like. At a minimum, principle-based approaches must attend to—and, if necessary, be reformulated in light of—the whole range of human moral experience, including women’s experiences of caring as well as of oppression. Fidelity to moral experience is one important criterion for any acceptable ethical theory, perspective, or method.

Communitarian Perspectives ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Communitarianism, another family of approaches to ethics—a somewhat loosely related family because of the variety of meanings of community—has often criticized principle-based approaches to bioethics, whether deontological, consequentialist, or pluralist. Some communitarians charge that principlism is a foundationalist approach that neglects the role of community and tradition in moral reflection. However, principlists are not necessarily foundationalists in the theoretical sense. Some principle- and rule-based approaches appeal to at least a thin sense of community in notions of ‘common morality’ (Gert 1989; Gert et al. 1997; Beauchamp and Childress 2001). And some principlists argue further that communities and traditions regularly embody and convey moral principles, just as they transmit settled judgments about cases and seek to engender certain traits of character. Some communitarians have stressed shared, deep, thick communal values that can engender, support, and sustain both casuistical practices as well as virtuous persons (Kuczewski 1997, 2001). One difficulty in thinking from a communitarian perspective, as these points suggest, is determining the relevant community, e.g. whether a whole society or a particular community within that society. Communitarianism may be primarily important as a perspective on ethical problems rather than as a way to resolve them. Most communitarians would agree with Daniel Callahan that ‘the first set of questions to be raised about any ethical problem should focus on its social meaning, implications, and context, even in those cases which seem to affect individuals only’ (Callahan 2003: 287). With its ecological bent, the point of communitarianism in Callahan’s sense is to offer ‘a way of thinking about ethical problems, not to provide any formulas or rigid criteria for dealing with them’ (Callahan 2003: 288). It presupposes both

methods in bioethics

41

analytical skills (rationality, imagination, insight) and personal virtues. It often seeks deliberative methods for the community, however defined, to develop and express its values (Emanuel 1992). In addition, communitarians often criticize principle-based approaches for their tendency to ‘block substantive ethical inquiry’ (Callahan 2003: 288). The debate about principles has been caught in the larger and tangled web of debate about individualism versus communitarianism. Communitarians charge that individualistic interpretations of bioethical principles have dominated moral discourse and now need to be corrected. Even though some defenders of principlebased approaches are individualists—for example, deontologists who concentrate on respect for autonomy and liberty or utilitarians who simply sum up the effects of actions on individuals’ interests—many principlists also recognize that moral agents are social and that communal values are very important. Furthermore, communitarians need not reject principles—though they may ground, interpret, and weight them differently than individualists and may propose additional principles, for instance, what some bioethicists call a principle of community or respect for community. At the first meeting of the National Bioethics Advisory Commission in October 1996, Ezekiel Emanuel, then a member, contended that the three Belmont principles (respect for persons, beneficence, and justice), developed by a predecessor national body (the National Commission for the Protection of Human Subjects), and related guidelines for research involving human subjects do not adequately address community. Such a challenge could mean, among other possibilities, that we should add community as a fourth principle to the Belmont list—the approach that Emanuel recommends—or that we should interpret all these principles through the lens of relationships and community. According to Emanuel and Weijer (2005), an independent principle is needed in order adequately to recognize that communities have moral status, values, and interests (such as avoidance of communal stigmatization) that merit protection—beyond the sum of individual values and interests—and to describe and address the conflict between individual and communal interests. However, communitarians are rarely clear about how such conflicts are to be resolved, particularly in view of the different kinds of community involved. An alternative way to correct the putative individualism of principlism is through a richer interpretation, or reinterpretation, of ethical principles through the lens of relationships and community. After all, the question for principles is not only their content, or their weight, but also their scope, as is indicated by the rubric Raanan Gillon (1994) has used for Beauchamp and Childress’s principles—‘four principles plus scope’. Among other things, scope can include such matters as moral status—for instance, of communities as well as embryos and the environment. Reinterpreted through the lens of relationships and community, appeals to the principle of respect for persons, or respect for autonomy, would consider persons

42

james f. childress

not merely as isolated individuals, who consent or refuse to consent, but also as members of communities. Individuals are embedded, to varying degrees, in their communities with various traditions, beliefs, values, and practices, and respect for cultural values is important. However, it is not possible or justifiable to determine an individual’s wishes and choices by reading them off community traditions, beliefs, values, and practices, and it is not ethically acceptable to subordinate the individual’s autonomy to the community’s will without satisfying stringent justificatory conditions. Similar points can be made about other basic bioethical principles: each one can be interpreted through the lens of relationships and community. For instance, researchers should recognize that some communities can be harmed in genetics research—for example, through stigmatization—in violation of the principles of beneficence and non-maleficence, and that the principle of justice may require involving such communities to participate in the design of genetics research involving their members.

Conclusions ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Significant conflicts remain about which bioethical method, if any, provides the best guidance for decisions and actions in clinical and policy settings. From the standpoint of action guidance, as well as other criteria, each method has significant problems, and yet each method is also helpful in some respects, particularly in identifying relevant features of agents, actions, ends, consequences, contexts, and relationships that merit attention in deciding and acting. Furthermore, several methods overlap with each other, or converge, or supplement each other. It is probably too much to expect consensus about the best possible method(s), all things considered—at most, we can take advantage of the strengths of each method and compensate for its special deficiencies.

References Arras, J. D. (1991), ‘Getting Down to Cases: The Revival of Casuistry in Bioethics’, Journal of Medicine and Philosophy, 16: 29–51. (1994), ‘Principles and Particularity: The Role of Cases in Bioethics’, Indiana Law Journal, 69: 983–1014. (1998), ‘A Case Approach’, in H. Kuhse and P. Singer (eds.), A Companion to Bioethics (Oxford: Blackwell), 106–14. Baier, A. C. (1985), Postures of the Mind: Essays on Mind and Morals (Minneapolis: University of Minnesota Press). (1994), Moral Prejudices: Essays on Ethics (Cambridge, Mass.: Harvard University Press).

methods in bioethics

43

Beauchamp, T. L. (2000), ‘Response to Strong on Principlism and Casuistry’, Journal of Medicine and Philosophy, 25: 342–7. and Childress, J. F. (1979–2001), Principles of Biomedical Ethics, (1st–5th edns. New York: Oxford University Press). Blum, L. (2001), ‘Care’, in L. C. Becker and C. B. Becker (eds.), Encyclopedia of Ethics, (2nd edn. New York: Routledge), 185–7. Brandt, R. B. (1992), ‘Public Policy and Life and Death Decisions Regarding Defective Newborns’, in Brandt, Morality, Utilitarianism, and Rights (Cambridge: Cambridge University Press), 354–69. Brody, B. (1988), Life and Death Decision Making (New York: Oxford University Press). (2003), Taking Issue: Pluralism and Casuistry in Bioethics (Washington, DC: Georgetown University Press). Callahan, D. (2003), ‘Principlism and Communitarianism’, Journal of Medical Ethics, 29: 287–91. Carse, A. L. (1991), ‘The ‘‘Voice of Care’’: Implications for Bioethical Education’, Journal of Medicine and Philosophy, 16: 5–28. Chambers, T. (1999), The Fiction of Bioethics: Cases as Literary Texts (New York: Routledge). Childress, J. F. (1982), Who Should Decide? Paternalism in Health Care (New York: Oxford University Press). Meslin, E. M., and Shapiro, H. T. (eds.) (2005), Belmont Revisited: Ethical Principles for Research with Human Subjects (Washington, DC: Georgetown University Press). Clouser, K. D., and Gert, B. (1990), ‘A Critique of Principlism’, Journal of Medicine and Philosophy, 15: 219–36. DeGrazia, D. (1992), ‘Moving Forward in Bioethical Theory: Theories, Cases, and Specified Principlism’, Journal of Medicine and Philosophy, 17: 511–39. DuBose, E. R., Hamel, R., and O’Connell, L. J. (eds.) (1994), A Matter of Principles? Ferment in U.S. Bioethics (Valley Forge, Pa.: Trinity Press International). Emanuel, E. J. (1992), The Ends of Human Life: Medical Ethics in a Liberal Polity (Cambridge, Mass.: Harvard University Press). and Weijer, C. (2005), ‘Protecting Communities in Research: From a New Principle to Rational Protections’, in Childress et al. (2005: 165–83). Engelhardt, H. T. (1986), Foundations of Bioethics (New York: Oxford University Press). Fletcher, J. (1966), Situation Ethics: The New Morality (Philadelphia: Westminster Press). Foot, P. (1977), ‘Euthanasia’, Philosophy and Public Affairs, 6: 85–112. Frankena, W. K. (1973), Ethics, (2nd edn. Englewood Cliffs, NJ: Prentice-Hall). Gert, B. (1989), Morality: A New Justification of the Moral Rules (New York: Oxford University Press). and Culver, C. (1979), ‘The Justification of Paternalism’, Ethics, 89: 199–210. et al. (1996), Morality and the New Genetics (Sudbury, Mass.: Jones and Bartlett). Clouser, K. D., and Culver, C. (1997), Bioethics: A Return to Fundamentals (New York: Oxford University Press). Gilligan, C. (1982), In a Different Voice: Psychological Theory and Women’s Development (Cambridge, Mass.: Harvard University Press). Gillon, R. (1994), ‘Medical Ethics: Four Principles Plus Attention to Scope’, British Medical Journal, 309: 184–8. (ed.) and Lloyd, A. (asst. ed.) (1994), Principles of Health Care Ethics (Chichester: John Wiley).

44

james f. childress

Hare, R. M. (1989), ‘Principles’, in Hare, Essays in Ethical Theory (Oxford: Clarendon Press), 49–65. (1996), ‘The Methods of Bioethics: Some Defective Proposals’, in L. W. Sumner and J. Boyle (eds.), Philosophical Perspectives on Bioethics ( Toronto: University of Toronto Press), 18–36. Hooker, B. (2002), Ideal Code, Real World: A Rule-Consequentialist Theory of Morality (Oxford: Oxford University Press). Hursthouse, R. (2001), On Virtue Ethics (Oxford: Oxford University Press). (2003), ‘Virtue Ethics’, in E. N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Fall 2003),

, accessed 8 Oct. 2005. Jansen, L. A. (2003), ‘The Virtues in Their Place: Virtue Ethics in Medicine’, Theoretical Medicine, 21: 261–76. John Paul II, Pope (1995), Evangelium vitae ( The Vatican). Jonsen, A. R. (1991), ‘Casuistry as Methodology in Clinical Ethics’, Theoretical Medicine, 12/4: 295–307. and Toulmin, S. (1988), The Abuse of Casuistry (Berkeley: University of California Press). JAMA (1988), ‘It’s Over, Debbie’, 259/2: 272. Kaufert, J. K., and Koch, T. (2003), ‘Disability or End-of-Life: Competing Narratives in Bioethics’, Theoretical Medicine, 24: 459–69. Kopelman, L. M. (1994), ‘Case Method and Casuistry: The Problem of Bias’, Theoretical Medicine, 15: 21–37. Kuczewski, M. G. (1997), Fragmentation and Consensus: Communitarian and Casuist Bioethics (Washington: Georgetown University Press). (1998), ‘Casuistry and Principlism: The Convergence of Method in Biomedical Ethics’, Theoretical Medicine and Bioethics, 19: 504–24. (2001), ‘The Epistemology of Communitarian Bioethics: Traditions in the Public Debates’, Theoretical Medicine, 22: 135–50. Louden, R. (1984), ‘On Some Vices of Virtue Ethics’, American Philosophical Quarterly, 21: 227–36. MacKenzie, C., and Stoljar, N. (eds.) (2000), Relational Autonomy: Feminist Perspectives on Autonomy, Agency, and the Social Self (New York: Oxford University Press). Manning, R. C. (1998), ‘A Care Approach’, in H. Kuhse and P. Singer (eds.), A Companion to Bioethics (Oxford: Blackwell), 98–105. National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research (1978), The Belmont Report: Ethical Guidelines for the Protection of Human Subjects of Research, DHEW publication no. (OS) 78–00 (Washington, DC: Department of Health Education and Welfare). Oakley, J. (1998), ‘A Virtue Ethics Approach’, in H. Kuhse and P. Singer (eds.), A Companion to Bioethics (Oxford: Blackwell), 86–97. O’Neill, O. (2002), Autonomy and Trust in Bioethics (Cambridge: Cambridge University Press). Pellegrino, E. D. (1995), ‘Toward a Virtue-Based Normative Ethics for the Health Professions’, Kennedy Institute of Ethics Journal, 5/3: 253–77.

methods in bioethics

45

(1998), ‘The False Promise of Beneficent Killing in Regulating How We Die’, in L. Emanuel (ed.), Regulating How We Die (Cambridge, Mass.: Harvard University Press), 71–91. Quill, T. (1991), ‘Death and Dignity: A Case of Individualized Decision Making’, New England Journal of Medicine, 324: 691–4. Ramsey, P. (1968), ‘The Case of the Curious Exception’, in G. Outka and P. Ramsey (eds.), Norm and Context in Christian Ethics (New York: Charles Scribner’s). (1970), The Patient as Person (New Haven: Yale University Press). (1978), Ethic at the Edges of Life (New Haven: Yale University Press). Richardson, H. S. (1990), ‘Specifying Norms as a Way to Resolve Concrete Ethical Problems’, Philosophy and Public Affairs, 19: 279–20. (2000), ‘Specifying, Balancing, and Interpreting Bioethical Principles’, Journal of Medicine and Philosophy, 25: 285–307. Sherwin, S. (1992), No Longer Patient: Feminist Ethics and Health Care (Philadelphia: Temple University Press). Sidgwick, H. (1962), The Methods of Ethics (London: Macmillan). (1998), Practical Ethics: A Collection of Addresses and Essays (New York: Oxford University Press). Singer, P. (1993), Practical Ethics, (2nd edn. Cambridge: Cambridge University Press). Slote, M. (1997), ‘Virtue Ethics’, in M. W. Baron, P. Pettit, and M. Slote, Three Methods of Ethics (Oxford: Blackwell), 175–238. Solomon, W. D. (1978), ‘Rules and Principles’, in W. T. Reich (ed.), Encyclopedia of Bioethics, i (New York: Free Press), 407–13. Strong, C. (2000), ‘Specified Principlism: What Is It, and Does It Really Resolve Cases Better Than Casuistry?’, Journal of Medicine and Philosophy, 25: 324–41. Sulmasy, D. P. (2001), ‘Research in Medical Ethics: Physician-Assisted Suicide and Euthanasia’, in J. Sugarman and D. P. Sulmasy OFM (eds.), Methods in Medical Ethics (Washington, DC: Georgetown University Press), 247–66. Sumner, L. W., and Boyle, J. (1996), ‘Introduction’, in Sumner and Boyle (eds.), Philosophical Perspectives on Bioethics ( Toronto: University of Toronto Press), 18–36. Toulmin, S. (1981), ‘The Tyranny of Principles’, Hastings Center Report, 11/6: 31–9. Veatch, R. M. (1981), A Theory of Medical Ethics (New York: Basic Books). (1995), ‘Resolving Conflicts Among Principles: Ranking, Balancing, and Specifying’, Kennedy Institute of Ethics Journal, 5/3: 199–218. (2005), ‘Ranking, Balancing, or Simultaneity’, in Childress et al. (2005: 184–204). Warnock, G. J. (1971), The Object of Morality (London: Methuen). Wildes, K. W. (2000), Moral Acquaintances: Methodology in Bioethics (Notre Dame, Ind.: University of Notre Dame Press).

chapter 2 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

THE WAY WE REASON NOW: REFLECTIVE EQUILIBRIUM IN BIOETHICS •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

john d. arras In the world of bioethics, the air is abuzz with reflective equilibrium. Not too long ago, this same air echoed the din of clashing moral methodologies. Casuists,1 feminists,2 narrativists,3 and pragmatists4 had been collectively engaged in a tag This chapter is dedicated to the memory of John Fletcher. Many thanks to Alex London and Bonnie Steinbock for helpful discussion along the way. 1 Casuistry in bioethics has been championed by Albert Jonsen and Stephen Toulmin as an alternative methodology to principlism, one that supposedly works from the ‘bottom up’ by means of analogical case analyses, much like the common law, rather than ‘top down’ by means of moral theories and theoretically derived principles. (See Jonsen and Toulmin 1988; Arras 1991.) 2 Feminist criticism in bioethics has focused primarily on gender-related power imbalances in the health care system. As contributors to the debate over method in bioethics, feminist critics of principlism have lamented the latter’s alleged overemphasis on abstract principle and neglect of personal relations, the emotions, and power in the analysis of moral problems. Many feminists are also moral particularists, and thus have much in common with casuists and narrativists. (See e.g. Wolf 1995.) 3 Narrative ethics, like casuistry, gives pride of place to the particularities of persons and social contexts as these are articulated within personal narratives. The emphasis here is on the trajectory of the ‘patient’s story’, rather than on the abstractions of theory and principles. (See Arras 1997.) 4 Bioethical pragmatists embrace the rejection of epistemological and ethical foundationalism that they find in traditional pragmatists like John Dewey and William James. They view ethical principles as

the way we reason now

47

team debunking exercise at the expense of the embattled defenders of principlism, the heretofore dominant method in the field of practical ethics (Beauchamp and Childress 1979–2001). These methodological malcontents shared a common core of grievances: Principlism was too abstract, deductive, and ‘top down’; it was insufficiently attentive to particulars, relationships, storytelling, and process. Each claimant to the throne adamantly pressed its case for methodological supremacy in bioethics. In response, the avatars of principlism, Tom Beauchamp and James Childress, politely thanked all their critics for making so many truly helpful suggestions, assuring them that they would each be respectfully subsumed into principlism’s evolving grand synthesis. Each hostile critique would, however, first have to be shorn of its excesses before being assigned its very own section in the next edition of Principles of Biomedical Ethics. Thus, to the casuists, Beauchamp and Childress granted that principles not only ruled over case judgments but were actually derived from particular intuitions in a dialectic between ‘top down’ and ‘bottom up’. To the feminists, they conceded the importance of relationships and the need to address inequality and domination in the relations between the sexes. To the partisans of narrative ethics, they affirmed the importance of narrative detail for the successful deployment and specification of principles. And, finally, to the pragmatists, they admitted the flexible, tool-like nature of principles, and the indispensability of good judgment in their application (Arras 2002). Central to this project of principlist assimilation was the notion of reflective equilibrium (henceforth, RE), a method of doing moral and political philosophy originally developed by the great political philosopher John Rawls (1971, 1975). According to Rawls, the project of justifying ethical beliefs ideally involves the attempt to bring our most confident ethical judgments, our ethical principles, and our background social, psychological, and philosophical theories into a state of harmony or equilibrium. Our most confident moral judgments or intuitions (e.g. ‘Slavery is wrong’) provide a touchstone for the adequacy of our principles; any moral principle that justified slavery would be either reformulated or rejected. Meanwhile, principles invested with a great deal of confidence could be used to reject some conflicting intuitions while extending our ability to judge confidently in less familiar moral settings. We thus zip back and forth, nipping an intuitive judgment here, tucking a principle there, building up or reformulating a theory in the background, until all the disparate elements of our moral assessments are brought into a more or less steady state of harmonious equilibrium. According to this view, moral justification must be sought, not in secure, incorrigible foundations outside of our processes of reflection, but rather in the coherence of all the flotsam and jetsam of our moral life. flexible tools that evolve over time in the service of social problem solving, rather than as absolute dos and don’ts; and they tend to emphasize the importance of democratic process as well as the substance of ethical judgments. (See McGee 2003.) It should go without saying that all of these brief vignettes in nn. 1–4 are hopelessly oversimplified.

48

john d. arras

Although Rawls limited the deployment of RE to the theoretical construction of his social contract theory, applied ethicists in many fields have more recently taken up this method as a vehicle for solving practical moral problems in their respective disciplines. (See e.g. Benjamin 2002; Nussbaum 1992; Daniels 1996; van der Burg and van Willigenburg 1998.) Although this more practical deployment of RE might raise eyebrows among those philosophers who view ethical theory and practical ethics as existing on two entirely different planes, it has made perfect sense for the growing number of practical ethicists who regard their work as existing on a continuum with that of theorists in normative ethics. It has, in fact, become something of a commonplace for philosophers straddling the theoretical and practical domains to remark that their modes of thinking and justification are pretty much identical in both areas, even if the role of abstraction is obviously greater in the domain of theory construction (Beauchamp 1984; Brock 1995). Although there is no doubt much to be said on behalf of this claimed continuity between practical and theoretical ethics, it remains to be seen whether RE, at least in its most expansive contemporary manifestations, can serve the theorist and practical ethicist equally well. As I explain below, I have my doubts on this score. One of the many attractions of RE as a method in practical ethics has been its ability to appeal to just about every faction in the method wars. It is agreeably flexible, non-dogmatic, and non-foundationalist in claiming that there are no incorrigible elements of morality on which everything else must be grounded and from which all justification flows. It enforces an appealing egalitarianism with regard to all the various elements of our belief systems, including our beliefs about particular cases, moral principles, and background theories. Within the method of RE, there are no privileged beliefs. Every belief is fair game for pruning in the service of more strongly held beliefs of the same or other kinds. Thus, casuists are happy to hear that intuitive case judgments are crucially important in moral justification; principlists are pleased with the robust role of moral principles; and high-flying philosophers and social theorists are relieved to hear that there’s even a place for background theorizing about the nature of persons and society. It was thus no great surprise, then, when Beauchamp and Childress, confronting that unruly mob of rival methodologists gesticulating from the other side of the moat, hoisted the unifying flag of RE, declaring it to be henceforth the method of principlism in bioethics (4th edn., 1994). All our methodological differences would henceforth merely be matters of emphasis. We would all just be fellow bozos on Neurath’s boat—out at sea, unable to reach dry dock where foundational work could be done, we patch, mend, and stitch our moral bark with the disparate materials at hand.5 As Mark Kuczewski observed at the time, ‘Who could ask for anything more?’ (1997). 5 ‘We are like sailors who on the open sea must reconstruct their ship but are never able to start afresh from the bottom. Where a beam is taken away a new one must at once be put there, and for this the rest of the ship is used as support. In this way, by using the old beams and driftwood the ship can be shaped entirely anew, but only by gradual reconstruction’ (Neurath 1966).

the way we reason now

49

Indeed, who could ask for more than this? As a philosopher, however, my job is to make life more difficult for people, so I will proceed to ask some hard questions about the method of RE in practical ethics. I do this not simply to make trouble, but rather because this method raises some very difficult and troubling questions which I shall explore below. I do so with some trepidation, however, because I have previously recommended a modest version of this very method (Arras 2003), and I am hard-pressed to identify a better way of justifying our judgments of right and wrong in practical ethics. I begin, then, with some preliminary remarks about the general features and basic varieties of RE in moral reflection. I shall then consider a couple of preliminary doubts about this method. One claims that the most plausible interpretation of RE is so comprehensive that it risks paralyzing our thinking, while the other claims that this same version of RE is insufficiently determinate in practical contexts and will thus fail to be sufficiently action-guiding. I then proceed to explicate the sense in which RE qualifies as a coherence theory of justification, and I consider several objections to RE that flow from its reliance on the putative connection between coherence and moral justification. I will then bring this chapter to a close with some reflections on the very idea of method in ethics—that is, on what we might realistically expect of methodology as a vehicle to advance justification and truth in the domain of practical ethics.

Standard Features of Reflective Equilibrium ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Considered Judgments and Principles in RE In his first paper devoted to a ‘decision procedure’ in ethics, Rawls elaborated an interesting picture of the relationship between moral principles and the intuitive or ‘considered’ judgments out of which they develop (1951). We begin, says Rawls, with a notion of ‘competent moral judges’—i.e. people who are intelligent, impartial, reasonable, well informed, imaginative, empathetic, and so on. Now, let us assume, first, that these judges are capable of filtering out their less plausible moral judgments. They are on guard against operating under conditions that usually yield bad or untrustworthy decisions—i.e. they are not judging in haste, under a cloud of intense emotions, driven by their own self-interest, and so on. Let us then suppose that these competent judges confront a wide spectrum of moral situations or cases and deliver judgments based not upon some sort of sophisticated theory or set of principles, but rather, assuming there to be such, simply upon their direct, unmediated, most reliable intuitions of right and wrong. Later on, in A Theory of Justice, Rawls calls these responses ‘considered moral judgments’—i.e. those moral judgments in which we have the most confidence (1971: 47/1999: 42). Although

50

john d. arras

Rawls, as always, is loath to cite concrete examples here, the paradigm cases might be examples of free speech, religious liberty, and racial equality. Putting it mildly, a competent moral judge would look unfavorably upon both the goals and methods of the Spanish Inquisition. Forcing all people to abandon their own visions of the good and the nature of the universe at the altar of Catholic orthodoxy—and to do so by threatening the rack, thumbscrew, or burning pyre—would definitely strike our competent moral judges as being morally out of bounds. So the initial ‘data’ of moral reflection are these intuitive judgments of competent judges directed at various cases. A good bioethical example of this would be our intuitive negative responses to the infamous Tuskegee syphilis study, in which US government researchers charted the doleful effects of untreated syphilis in black sharecroppers over a period of forty years. The next step in the method is to develop moral principles that ‘match’, ‘explicate’, ‘accord with’, ‘fit’, or ‘account for’ the body of intuitions amassed by competent moral judges. Thus, a principle of religious freedom might explicate the fact that various competent moral judges would intuitively condemn the torture and burning of heretics and similar behaviors. Rawls likens this process to the inductive scientific method, whereby inquirers assemble a set of observation statements or data points, and then attempt to formulate a principle or mathematical function that best makes sense of them. According to this interpretation of RE, moral principles are hypotheses advanced to make sense of the set of considered moral judgments of competent judges. Principles ‘explain’ moral judgments if we could deduce exactly the same judgments just from the principles and relevant facts alone, without the benefit of any moral intuitions or sentiments. In short, moral principles are supposed to yield conclusions in particular cases that would match our considered judgments. If our suggested principles mesh perfectly with our considered judgments, then we are in equilibrium; if they don’t fit, then we have to amend either our particular moral judgments or our principles, depending upon which element of our moral system merits the most confidence. In addition to their explanatory function sketched above, moral principles also have normative functions within the framework of RE. First, principles that have been forged over time in the crucible of RE can help us recognize and reject mistaken moral judgments; second, Rawls asserts that principles should help us resolve moral perplexities posed by conflicting intuitions in difficult cases. Deploying a set of firmly held principles can assist us in extending the reach of our convictions, even in those situations where we initially lack confidence in our judgments. A number of questions arise with regard to this initial Rawlsian sketch of RE. First, it is reasonable to ask whether our considered judgments about cases really present themselves in isolation from more ‘theoretical’ considerations in the way that Rawls initially suggested they might (1951). I suspect, with Brian Barry, that most such intuitive judgments about cases already harbor the germ of some sort of larger, quasi-theoretical reflection. For example, our repulsion at the burning of heretics

the way we reason now

51

or at the withholding of proven treatment from syphilitic black sharecroppers is not some sort of brute ‘datum’, but is rather most likely a reaction already suffused with the ‘quasi-theoretical’ judgment that the Inquisition and Tuskegee violated, inter alia, important moral norms mandating equality among human beings (Barry 1989: 264; Arras 1991). Conversely, it is equally problematic to assert that moral principles can be relied upon to yield, exclusively on the basis of the relevant facts, the very same conclusions reached by confidently uttered case judgments. Just as case judgments are infused with quasi-theoretical elements, so principles, in order to actually reach practical conclusions in concrete circumstances, must be supplemented with all sorts of judgments, intuitions, and analogous comparisons with other cases. Although we might be able to ‘derive’ or ‘deduce’ correct practical conclusions from principles in clear-cut cases of moral evil like the Tuskegee experiment, we cannot do so in hard cases involving conflicting principles or difficult problems of interpretation—that is, in precisely those cases that provide the grist for most bioethical reasoning. A second problem with Rawls’s initial formulation of RE is his claim that moral principles should be understood as hypotheses developed to explain or match whatever deeply felt intuitions we happen to have (Barry 1989: 263). At the least, proponents of RE need to explain how moral intuitions can play the same justificatory role as observation statements in the physical sciences, each one providing its own particular kind of basic ‘datum’ for further theorizing. (I shall say more about this issue later on.) Even some of the most zealous defenders of RE have abandoned this suggested analogy between observation statements in science and considered moral judgments (Daniels 1979). More plausible recent expositions of RE simply acknowledge the existence of most of our commonsensical moral principles—e.g. keep promises, respect autonomous choices—and then attempt to show what is valuable and important about the norms they articulate and to state the best reasons why it is wrong to violate them (Scanlon 1992). In other words, instead of deploying RE to discover new principles, these theorists harness RE for the more modest but still crucially important task of becoming more reflective about the meaning, functions, and justifications of whatever moral principles we happen to endorse. Sometimes the process of RE will reveal that we have misunderstood the values protected by a principle, and this will have implications for what the principle is now taken to sanction. More rarely, this process may reveal that we have simply given up entirely on the reasons behind a principle (‘Shield your patients from troubling diagnoses’), and at that point we jettison the principle. Finally, and most importantly, it is unclear how RE, as explicated so far, has any serious normative bite. At best, such a bouncing-back-and-forth between considered judgments and derivative principles will yield a fully rounded inventory of our collected moral intuitions, but the justification of those very intuitions remains unsettled. It is a commonplace that people often feel supremely confident in their most basic moral judgments, but it is unsettling, to say the least, to acknowledge

52

john d. arras

that some of those very judgments have in the past affirmed the naturalness of slavery, the necessity of burning heretics at the stake, and the unsuitability of atheists and women for public office. Construed narrowly to encompass only particular judgments and the principles that explain them, RE offers an easy target to critics, who claim that it is hopelessly parochial and conservative (Hare 1975; Singer 1974; Kagan 1998). This suspicion of moral intuitions is nicely captured by James Griffin, who observes that ‘It is especially in ethics that intuitions have risen so far above their epistemological station’ (Griffin 1996: 5). In order to meet this challenge, the partisans of RE reach for a distinction between ‘narrow’ and ‘wide’ versions of the method.

Narrow versus Wide RE The version of RE that we have considered so far is called ‘narrow reflective equilibrium’ (henceforth NRE); it can help us identify our most confidently held moral judgments and the principles that best explain them, but it apparently lacks the resources to actually justify those intuitions and matching principles. To do that, RE must expand its inventory of moral considerations and widen its scope. Towards this end, Rawls and Norman Daniels have developed an alternative account of RE that they call ‘wide reflective equilibrium’ (henceforth WRE; see Rawls 1975; Daniels 1979: 258). In addition to considered judgments and principles, then, WRE encompasses a wide variety of theoretical considerations, including (1) alternative ‘moral conceptions’ (e.g. utilitarianism, perfectionism, Kantian ethics) and their respective philosophical warrants; (2) theories of moral personhood; (3) theories of procedural justice; (4) theories of moral development; and (5) empirical theories bearing on the nature of society and social relations. In spite of the evident shortcomings of our considered moral judgments with regard to the problem of justification, the partisans of WRE contend that we have no other choice but to embrace them. Crucially, however, they contend that these judgments are only ‘provisional fixed points’ that must be scrutinized from every possible angle. The above theoretical considerations are imported into RE in order to provide just that kind of critical scrutiny. Importantly, Daniels insists that these background theories are not dependent upon our moral intuitions in the same way that our principles are, so they can, he argues, be counted upon to provide independent justification for the deliverances of RE (1979: 260). WRE thus provides us with a highly complex and multi-layered approach to moral justification. Considered judgments, principles, and background theoretical considerations incorporate as many moral and empirical beliefs as possible, and allow us to test each of these elements or strata against all of the others. Crucially, as mentioned before, no single element or stratum of this dynamic mix of beliefs is considered to be foundational or immune to criticism. We shuttle back and

the way we reason now

53

forth from judgments, to principles, to theories, and back again—always adjusting, pruning, and seeking coherence among the widest possible set of relevant beliefs. WRE is thus both a coherentist and non-foundationalist approach to moral justification.

WRE and Methods of Bioethics It’s time to pause in our exposition and draw out some important initial implications of WRE for our methodological debates in bioethics. Remember, Beauchamp and Childress have explicitly embraced WRE as the official methodology of principlism. (In their fourth edition they write, ‘we have agreed with Rawls that justification is ‘‘a matter of the mutual support of many considerations, of everything fitting together into one coherent whole’’ ’ (1994: 23). One important question is whether this move is consistent with their long-standing advocacy of principlism as a distinct methodology. As I understand it, principlism has traditionally emphasized the centrality of moral principles in bioethical reflection. As the very title of their oft-revised book implies, for Beauchamp and Childress bioethical reasoning is ultimately about the identification, justification, specification, weighing, and balancing of moral principles against one another in the context of specific cases. As Beauchamp once remarked, principles provide the ‘spine of ethical analysis’ (personal communication). Now, it may well have been a commendable move for these distinguished partisans of principlism to endorse WRE and thereby usher in the peaceable methodological kingdom, but it is difficult to comprehend how they could do so while still giving pride of place to principles in moral analysis. We must recall in this connection that WRE doesn’t play favorites with regard to the various kinds of beliefs, whether they are about cases, principles, or background theories. No single stratum or cluster of moral considerations is privileged. All that matters as we go about our business of adjusting, pruning, and rendering our beliefs coherent is the strength of particular beliefs, our degree of commitment to them, rather than the objects of belief or the level of their concreteness (Scanlon 1992: 14; DePaul 1993: 157). Thus, our beliefs about principles are always subject to revision at the bidding of more strongly held beliefs about particular case judgments or background theories of the person, due process, or society. If this is the case, then principlism seems to have effectively placated and silenced its critics at the cost of its own methodological distinctiveness. Were brevity and aesthetics not factors to consider, the title of their next edition should read: Considered Case Judgments, Principles, and Background Theories in Bioethics: How They Can All Be Brought into Coherence Within the Ambit of Reflective Equilibrium. Principles no longer deserve top, let alone exclusive, billing. What’s true for principlism is also true for each of the other rival methodologies in bioethics. Just to take one additional example, casuistry can also be taken up into

54

john d. arras

the larger synthesis of WRE, but only at the cost of sacrificing its distinctiveness. According to Albert Jonsen and Stephen Toulmin (1988), the prime movers of casuistry as a rival method to principlism, the primary locus of moral justification and certitude is the paradigm case. Modeling their vision of ethics on the common law, these authors assert that moral knowledge results from the slow accretion of cases and our efforts to distill principles out of them. According to Toulmin (1981), moral principles are just so many afterthoughts trailing behind our intuitive responses to paradigm cases. If we are looking for normativity, we will find it, asserts Toulmin, in the paradigm case—not in principles distilled post hoc, and certainly not in abstract ethical theories. Thus, just as principlism gave pride of place to moral principles (the ‘spine’ of ethical inquiry), casuistry locates the ‘real action’ in ethical reflection at the level of the paradigm case. In so far as both of these rival methodologies favor one level of moral belief over others, they both must be shorn of such favoritism before being allowed to play their respective roles in our search for wide reflective equilibrium. The take-home message here is that if WRE is taken to be the preferred method in moral philosophy and bioethics, then both principlism and casuistry as traditionally understood must be rejected as partial and incomplete moments in a grander, allencompassing methodological synthesis. As Hegel would put it, both methods are aufgehoben—that is, they are shorn of whatever is partial, fragmentary, or one-sided about them, while their remaining valuable features are preserved and elevated within a more comprehensive synthesis. Again, who could ask for anything more? The beauty of WRE is that it makes room for all beliefs that might potentially contribute to a richer synthesis. As Michael DePaul (1993: 107 ff.) notes, WRE is the only fully rational method of moral inquiry. Whatever its faults or limitations might be, WRE is uniquely capable of leading the moral inquirer to accept a rational system of beliefs through a set of rational steps. Alternatives to WRE cannot make this claim. Take, for instance, R. M. Hare’s project of grounding ethics on a foundation of meta-ethical propositions bearing on the meaning of moral terms like ‘good’ and ‘right’ (1963). According to Hare, moral intuitions and considered moral judgments are far too untrustworthy to help guide moral reflection in a process like RE. (‘Garbage in, garbage out.’) He therefore urges us to submit all of our moral beliefs, including beliefs about moral intuitions and moral principles, before the tribunal of epistemic principles and background theories. DePaul asserts that Hare’s proposed method is irrational in so far as it would require us to subordinate or jettison our strongly held moral beliefs about cases and principles even if we think, on due reflection, that they are more likely to be true than any theory of moral language that Hare or others might concoct. This, DePaul concludes, cannot be a rational move (1993: 124). The same thing could be said about a method like casuistry that submits background theories and moral principles to the tribunal of paradigm cases. Even if someone were suddenly to come to believe that a moral or political theory (e.g.

the way we reason now

55

Marxism) was more likely to be true than any other item in their inventory of beliefs, the casuist would require her to give up her theory if it conflicted with her pre-existing moral intuitions about paradigm cases, even if the born-again Marxist now views those intuitions (e.g. those relating to private property) to be the products of false consciousness. The intuitive attractiveness of WRE, then, rests upon its inclusiveness. If you don’t like the way the process of RE is going, if you think that it currently overlooks some crucial pieces of the moral picture—such as a different moral outlook or a background theory of social stability—then WRE simply asks you to toss it into the mix alongside all our other beliefs. Who could ask for a more accommodating method of moral reflection?

Preliminary Doubts About Wide Reflective Equilibrium ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Notwithstanding the inclusiveness and intuitive attractiveness of WRE, there are difficulties and objections that must be squarely confronted. Although I think that the most philosophically interesting and important challenges to WRE relate to its embrace of coherence as the engine of moral justification, I shall begin with worries focused on the scope and action-guiding potential of this method.

WRE is too Comprehensive Although the inclusiveness of WRE initially strikes us as a major advantage over foundationalist theories, it is also a source of pragmatic concerns about the method’s practicability. Consider the length and breadth of WRE’s welcome mat for the ingredients of moral reflection. Judgments about cases, moral principles, competing moral outlooks (i.e. moral theories), the accompanying philosophical arguments for those rival outlooks, theories of the role of morality in society, theories of moral personhood, notions of procedural justice, general social theories, and theories of moral development all have a role to play in the working out of RE. If we wish to know whether any particular proposition is morally justified, we have to subject it to the competing pushes and pulls of this entire network of beliefs. The daunting nature of such a mission comes to light when we consider just one ingredient of this overall process of justification, namely, those competing moral conceptions and their accompanying philosophical justifications. Suppose we wish to ascertain whether our views on cloning are morally justified. The method of WRE would demand, inter alia, that we at some point consider the rival claims of all the live options in moral theory—e.g. utilitarianism, Kantian theories of

56

john d. arras

autonomy, Thomistic perfectionism, feminism, Aristotelian virtue theory—assess their competing arguments, and embrace the one moral conception that best coheres with the rest of our moral intuitions and background theories. This, of course, could well constitute the life’s work of a professor of moral philosophy, but it is not all that would be required. We would then have to perform similar work on competing theories of the person, of social organization, and of the role of morality in society. Once all this work (and more) was done, we would need to see how all of our disparate beliefs and theories fit together, making sure to nip and tuck those that conflicted with those beliefs and theories in which we had the greatest confidence. In contrast with casuistry, which views justification as a relatively simple and straightforward matter of bringing our judgments within the gravitational pull of paradigm cases, this picture of justification in WRE is daunting in the extreme. It truly seems like a job not for ordinary mortals, but rather for Ronald Dworkin’s legendary but fictional judge Hercules.6 This problem of over-inclusiveness need not necessarily be a huge problem for the moral theorist. Rawls (1971: 49–50/1999: 43), for example, concedes that ‘it is doubtful whether one can ever reach this state’. He contents himself, however, with the thought that the theorist can at least canvass some of the most salient options in moral theory, if not all such options, and still regard WRE as a regulative ideal towards which the theorist should strive. It is less clear, however, whether the practical ethicist has this same luxury. Once we transcend narrow RE to encompass standard moral theories and their philosophical justifications—not to mention all those other background moral, political, and empirical theories—the ordinary working stiff bioethicist is likely to find WRE to be a hopelessly cumbersome method of moral justification. If she is to make any progress at all, she will no doubt have to bracket many beliefs and theories that would normally play integral roles in an ideal process of RE. Indeed, it would not be surprising if Reflective Equilibrium for Working Bioethicists, Version 1.0 were to bear a striking resemblance to narrow RE in the majority of cases.

WRE is too Indeterminate A closely related problem concerns a disconcerting lack of precise guidance in coming to terms with all those competing moral outlooks and their corresponding supportive arguments. Exactly how is the process of analysis and comparison 6 Hercules is portrayed by Dworkin as an idealized omniscient judge who decides hard cases by forging legal principles that best ‘fit’ all the legal precedents, best mesh with our legal history and institutions, and are most compellingly justified by our best moral and political theories of justice and equality. Dworkin realizes that fallible human judges cannot conceivably work their way through all these steps, but he asserts that this would be the ideal process of legal justification (Dworkin 1977: 197).

the way we reason now

57

supposed to proceed? Rawls and Daniels say precious little about this. Even Rawls’s attempt in A Theory of Justice to vindicate his famous two principles of justice against the claims of various versions of utilitarianism has run into a barrage of cogent criticisms. The same worry obviously haunts the project of choosing among rival social theories or theories of the person. What criteria should we use to select one such theory over others? How should the reasoning go? Here again, WRE appears to be more a rather massive effort of hand waving than a precise road map to moral justification (Raz 1982: 309). Given the inherent vagueness in the charge to review all these various objects of belief, to vindicate some but not others, and to bring all of the remaining beliefs into the broadest possible circle of coherence, it appears highly unlikely that WRE will eventually yield definite action-guiding results. The combined effect of these two related criticisms of WRE—i.e. that this method is both too comprehensive and too indeterminate—is to suggest a distinction between an ideal method of justification and a more rough-and-ready decision procedure that might helpfully guide our thinking in practical contexts.7 WRE might well serve as a regulative ideal of moral justification for theorists, especially if we were given more information about how the various moral conceptions and background theories should be evaluated and compared; but it is hard to imagine a more cumbersome or less action-guiding program for practical moral decision making. The omniscient but unfortunately fictional judge Hercules might be able to manage all those justificatory hoops, but ordinary mortals will have to settle for much less.

WRE and the Limits of Coherence As we have seen, WRE rejects foundationalism in ethics, the view that a certain favored set of beliefs—e.g. meta-ethical propositions, human nature, paradigm cases, moral theory, certain key intuitions, or the Bible—constitutes the incorrigible bedrock of moral reflection from which all other beliefs must flow. Within foundationalist moral systems, beliefs are justified by being ‘derived from’ or ‘based upon’ those more fundamental or foundational beliefs. Within WRE, by contrast, beliefs are justified—i.e. they acquire the greatest measure or warrant or support—by being brought into coherence with the widest possible set of other beliefs we hold. As Rawls (1971: 21/1999: 19) puts it, the justification of moral principles ‘is a matter of the mutual support of many considerations, of everything 7 Alex London has cogently developed this theme in a series of articles on method in practical ethics. Rather than viewing the various proposed methods of bioethics as routes to justified true beliefs, he views them as helpful procedures for ensuring that non-ideal agents deliberating under real-world circumstances have the best chance to decide an issue on the merits (London 2000).

58

john d. arras

fitting together into one coherent view’. RE is, thus, a particular version of the ‘coherence theory’ of moral justification, which has applications not just in ethics but also in the theory of knowledge generally (Pollock 1986). While much of the attraction of WRE derives from its repudiation of foundationalism, its embrace of coherence theory is viewed by some as deeply problematic. In this section I shall try to sketch, albeit all too briefly, the most salient of these doubts and worries and to assess their significance for the viability of WRE as a bioethical method.

What Kind of Coherence Justifies? If beliefs are justified by being brought into coherence with other beliefs, we first have to determine exactly what we mean by ‘coherence’. There are several different interpretations of coherence, and it matters which one we select because they have very different implications for the project of moral justification. We might, for example, interpret coherence to mean the consistency of each element in the overall belief system, including both moral and empirical beliefs, vis-`a-vis all the other elements. Although this gloss on coherence would no doubt be somewhat helpful, allowing us to ferret out contradictions among our various beliefs, it’s hard to see how mere consistency with other beliefs can serve to justify any particular belief. A more robust notion of coherence can be found in commentaries on the natural sciences, where the justificatory power of coherence is bolstered by two special features. First, coherence in the natural sciences is buttressed by observation statements that provide the data for theory building. True, even in the natural sciences some observations might be ignored if they happen to conflict with an especially powerful theory, so science, like RE, takes a holistic approach to justification. Still, most of us think that the observation statements in physics are on a much firmer epistemological footing than the considered moral judgments of various people, especially when we note that these sources of considered judgments may reflect major and irreconcilable cultural, religious, and class differences. Second, the kind of coherence available in science features not just mutually consistent beliefs, but also mutually supporting beliefs or ‘credibility transfers’ that can reliably raise the level of the whole set of beliefs (Griffin 1996: 16). Assuming that the same natural world is the subject matter of all the sciences, beliefs developed in one zone of inquiry will have the effect of supporting similar beliefs developed in others. The upshot of this cross-cutting system of mutual support among scientific beliefs is that coherence in science exhibits a certain ‘bootstrapping’ effect that appears to be lacking in the moral domain. We appear, then, to confront a dilemma: If we construe coherence weakly to mean consistency with our other moral beliefs and known facts, then just about any live option in moral philosophy will pass the test of coherence, and we won’t have a test that will allow us to choose between such live options. Alternatively, if we construe coherence to entail the sort

the way we reason now

59

of bootstrapping and credibility transfers that we encounter in scientific webs of belief, then we would indeed have a conception of coherence with legs, one that offered real justificatory power. Unfortunately, however, it is highly doubtful that the sort of relationships connecting our considered judgments, moral principles, and background theories can rise to this level. Certainly, various relationships can be discerned among these disparate elements of RE; but it is highly uncertain, to say the least, whether they can bestow the sort of heightened credibility on display in mutually supportive scientific beliefs.

Can Coherence Teach Us What and How to Prune? According to Daniels (1979: 258), the method of WRE is an attempt to generate coherence in an ordered triple of sets of beliefs encompassing (a) considered moral judgments, (b) a set of moral principles, and (c) the set of relevant background theories. We thus attempt to bring our beliefs at one distinct level (e.g. considered judgments) into harmony with our beliefs at the other levels (e.g. moral principles or background theories). If we encounter a disparity between any two of these disparate elements of our moral system, then RE calls for a movement to and fro, pruning a bit here, tucking a bit there, until the discrepancies are ironed out and harmony among our beliefs is restored. But now a question arises: Assuming the appearance of a conflict between two different elements of our moral system—e.g. between a considered judgment and a given moral principle—which one should be pruned? As D. W. Haslett (1987) observes, coherence considerations by themselves are not enough to enable us to decide between any two conflicting elements within the ambit of RE. While coherence can indeed direct us to prune either the judgment or the principle, it cannot tell us which one should be sacrificed in its name. Given any two sets of considered judgments and matching principles, there could, then, be innumerable different wide reflective equilibria corresponding to different (arbitrary) choices for nipping and tucking. Haslett’s objection gives us an additional reason to demand more from coherence than mere consistency. If that is all coherence means, then Haslett has indeed delivered a crushing blow to the claims of RE. The defenders of RE can respond, however, that coherence encompasses not just the bare-bones notion of consistency, but also such important relations as providing the ‘best fit’ or ‘strongest mutual support’. While this gloss on coherence could be helpful in pointing our pruning shears in the right direction, it generates problems of its own. As Bo Petersson (1998) points out, the larger the network of intersecting elements in RE, the more disparate their contents and varieties of interdependence, the more difficult it will be to assign a definite meaning to such notions as ‘maximal coherence’ and ‘strongest mutual support’. In WRE we have a vast network of considered judgments, moral principles, and both moral and empirical background theories. The way in which

60

john d. arras

empirical background theories may support various moral theories, for example, may differ significantly from the way in which considered judgments support moral principles. How are we to determine the relevance and strength of these different kinds of support in coming to an all-things-considered judgment about ‘maximal coherence’? For this purpose should the number of supportive relations count for more than their ‘strength’? Since WRE contains such a huge mix of disparate elements and differing kinds of supporting relationships—e.g. exhibiting logical entailment, inductive support, or mere consistency—the notion of degrees of coherence implied by this alternative approach turns out to be yet another exercise in hand waving. If we cannot clearly ascertain which particular arrangement of all these disparate elements is the ‘most coherent’, then the method of WRE will fail to provide a useful guide to either theory construction or practical ethics.

Will the Coherence Approach Yield Convergence? It’s hard to think of a better method than RE to help each of us organize, systematize, and smooth out inconsistencies within our respective inventories of moral beliefs. Indeed, most discussions of RE stipulate that it is a method to be used by each individual inquirer to achieve moral justification of his or her own beliefs. What happens, however, when we abandon this first-person perspective and consider the reflective equilibria of other people in society? Will the widespread use of WRE tend to foster convergence of belief with regard to individual cases, moral principles, and background theories? This is an important question because it goes to the heart of the whole rationale for employing a method of moral reasoning in the first place. In most fields of inquiry, the promise of method, as opposed to mere ad hoc, episodic ruminations, is that it will reliably guide us to discover correct results. At its most rigorous, the notion of method works as a kind of procedural guarantee that anyone who submits to its discipline, going through all the prescribed steps, will come out at the other end with an intersubjectively verifiable truth. A more realistic expectation for method in bioethics is that following its dictates will at least increase our chances of converging upon the truth. I will come back later in this chapter to the very notion of method in ethics, but for now I want to focus on the claims of WRE to facilitate convergence around proposed solutions to bioethical problems. One indicator of intersubjective reliability is the ability of a method to produce a confluence of opinion or belief among those who use it. Although mere consensus is obviously no guarantee of moral truth—witness the Third Reich, which was remarkably successful in achieving consensus—it’s hard to imagine a method of thought being reliable in the required sense unless it can produce convergence of belief among those who deploy it. It is thus an important question whether WRE can be counted upon to effect this sort of convergence with regard to moral beliefs.

the way we reason now

61

Interestingly, Rawls (1971: 50/1999: 44) never committed himself on this question, and apparently for good reason. A moment’s reflection on the matter should make it rather painfully obvious that in a liberal, pluralistic democratic society, different groups of people will enter WRE with very different considered moral judgments. Although such judgments are not sacrosanct and incorrigible within WRE, they do exert a significant influence upon the moral principles we eventually embrace and the sorts of background theories we are willing to accept. Thus, we should expect fundamentalist Christians from the Midwestern United States to start with dramatically different considered judgments from those of Jewish leftists living on the upper west side of Manhattan; and we should, moreover, expect both of these types to differ fundamentally with the considered moral judgments of Shi’ite clerics in Baghdad or Shining Path Marxist revolutionaries in Peru. It is reasonable to expect, then, that the members of all four of these demographics will end up in very different moral places once they have gone through the required motions of WRE in their own heads. Clearly, their considered judgments will not all be mutually exclusive; they will share a good deal with regard to nearly universal values bearing on truthfulness, respect for property (except those Marxists and anarchists who equate property with theft!), and the avoidance of gratuitous cruelty. They will, nevertheless, tend to differ on such key bioethical issues as the fair distribution of health care, the permissibility of artificial reproductive technologies, and the enhancement of human nature through biomedical technology. In short, then, the method of WRE can ideally yield justification to those individuals who engage in the process, but it cannot be counted upon to yield convergence with regard to typical front-burner issues in bioethics; and if it can’t yield convergence among different sorts of people in a pluralist society, then it cannot be considered to be a reliable method of arriving at moral truth. There may well be no other method of moral inquiry that could possibly achieve this kind of convergence and reliability; the moral world may simply be too fragmented for that. Even Rawls’s staunchest defenders have admitted that WRE cannot achieve intersubjective consensus within the context of pluralistic societies, at least with regard to those controversies that implicate divergent visions of the good beyond the ‘overlapping consensus’ on basic principles of justice. As Daniels (1996: 144–75) now puts it, it was a ‘philosopher’s dream’ to imagine that the kind of philosophical reflection driving WRE could bring about convergence within pluralistic societies, and he has therefore turned to other methods of justifying practical moral solutions within such societies, such as the political procedures of ‘deliberative democracy’ (Daniels and Sabin 2002). Along with Rawls, Daniels still believes that RE can generate an overlapping consensus on the core principles governing the basic structure of a liberal democratic society, but he now contends that we must seek the resolution of controversies arising beyond the basic structure, where most bioethical problems lurk, in a political process bounded by the norms of basic justice.

62

john d. arras

This is in many ways a salutary move. One problem with the method of RE as developed so far is that it tends to focus on the solitary theorist and her struggle to achieve coherence among all of her disparate beliefs. Notwithstanding the importance of this project, it tends to neglect the social dimension of reasoning in an area like bioethics. More often than not, we reason together in bioethics in the context of hospital ethics committees, governmental ethics commissions, and collaborative group projects. In contexts like these, it is important not simply to align our respective beliefs in reflective equilibrium, but also to justify them to other free and equal citizens within a process of inquiry governed by democratic norms.

Local versus Global Coherence? According to the partisans of WRE, we should seek coherence among the widest possible set of moral and non-moral beliefs. But how far and wide can we expect this set to extend into the entire domain of morality? What, in other words, is possible? On one view, morality can be properly regarded as a unified system of beliefs, principles, and theories. Any inconsistencies between disparate regions of the ‘moral world’ should, from this angle, be ultimately remediable. On this reading, the sort of coherence we seek in moral reflection is ultimately global. This is a very ambitious thesis. Just how ambitious it is can be gauged by reflecting for a moment on the sort of phenomenon morality is. For those of us who don’t believe that morality is given to us by God, helpfully inscribed on extremely durable stone tablets, or by nature in such a way that the solutions to our disputes in practical ethics can be ‘read off’ the nature of humans and society, morality will present itself to us as a historically and culturally conditioned achievement. Different considered judgments and moral principles developed in different historical settings exist side by side in contemporary cosmopolitan societies, and these differences may well prevent morality from being viewed as a coherent whole. If the domain of law can be aptly characterized by one of our most distinguished theorists as a ‘higgledypiggledy assemblage of the remains of contradictory past political ambitions and beliefs’ (Raz 1994: 296), then, a fortiori, the same could be said of morality (Nagel 1991; Scanlon 1992; Brock 1995). This does not mean that achieving coherence isn’t a desirable thing, for it clearly is; it does mean, however, that the sort of coherence that we can realistically seek among our moral beliefs is most likely local rather than global. As Raz (1994: 317) nicely puts it, we should expect to find ‘pockets of coherence’ rather than vast unified expanses of it. Thus, the principles developed to govern the physician–patient relationship may be inadequate or counterproductive in other domains of the moral life, such as public health, environmental ethics, or assessing our obligations to the distant needy in other lands (Scheffler 2001).

the way we reason now

63

What’s So Great About Coherence? I wrap up this excursion into coherence theory and RE with some brief remarks about the epistemological value of coherence. This topic runs very deep, so I can only scratch the surface here. This is the problem: Once we abandon foundationalist approaches to justification in ethics, we are apparently left with coherence as the only remaining source of justifiability. Yet coherence, on the face of it, seems an unlikely candidate for this role. The first thing to notice is that coherence per se is not necessarily a compelling virtue of moral outlooks. It’s quite possible for someone to inhabit a seamlessly unified moral world-view and yet fail to be justified in his moral judgments.8 Consider the case of Rush, a hypothetical teenager of middling intellectual gifts, sub-acute moral perception, and an embarrassingly bad complexion, who immerses himself in the world of right-wing politics in order to provide himself with ‘an identity’. All day long, he tunes into the rants of right-wing ideologues on talk radio, reads and rereads their screeds in pamphlets and books, and avidly participates in their website chat rooms. Rush emerges from this ideological bath feeling much better about himself as a committed Republican, free-marketeer, political libertarian, and a sworn enemy of the welfare state, which he decries as a haven for losers and a drain on the energies of virtuous, wealthy entrepreneurs. What shall we say about Rush’s world-view and his dismissal of the poor as a bunch of pathetic losers? Clearly, Rush’s moral universe is exceedingly coherent. His moral, political, economic, and even artistic views all hang together quite nicely now, and his judgments about events and policy flow spontaneously from that coherent world-view. Still, many of us, even many conservatives, would probably say that Rush’s moral and policy judgments are anything but morally justified. They would perhaps point to Rush’s lack of intelligence, his lack of experience in the real world, and the urgency of his need for self-validation as factors vitiating the trustworthiness of his confident denunciations of the poor. Even if he has reached, within his own mind, a state of reflective equilibrium, most of us would deny that this equilibrium justifies anything about Rush’s judgments. We would, for starters, recommend that he broaden his experience of the world, perhaps by joining Habitat for Humanity or the Peace Corps, and enlarge his reading list to include (at least) Charles Dickens, Victor Hugo, John Steinbeck, Frantz Fanon, and Martin Luther King. Once he has exposed himself to such potentially transformative texts and experiences, he might well abandon his former beliefs; but even if he continues to embrace some version of political conservatism, his former beliefs will have been tested and transformed in the crucible of conflicting evidence. At least compared to his earlier beliefs, Rush’s new beliefs will be more justified for having emerged from that crucible, even if his old web of belief was entirely coherent. This is where 8

The same reservations about coherence surface with regard to particular moral theories. One of the most coherent moral outlooks, by far, is the theory of act utilitarianism, yet this theory is regularly assailed by its many critics as overly simplistic and tolerant of immoral results.

64

john d. arras

Daniels’s exhortation for us to achieve coherence among the ‘widest possible’ set of beliefs becomes absolutely crucial. Here’s another case. Consider Sophia, a hypothetical young woman—a smart, well-educated, widely read, well-traveled, progressive, and deeply thoughtful early twentieth-century amateur eugenicist who advocates sterilization of the ‘unfit’. Like Rush, Sophia inhabits an extremely coherent moral viewpoint. All the disparate elements of her inventory of belief—including the Bible, as she has learned to interpret it; ‘common sense’; the then-ascendant social Darwinist theory of political economy; and, of course, then-current theories of genetics—tell her the same thing: namely, that the white race is superior, that it is under siege, and that sterilizing the ‘unfit’ represents the quintessence of social responsibility. In contrast to our response to Rush, it’s harder to say that Sophia is intellectually dim, sheltered, badly educated, or psychologically deformed. She is, in the opinion of her contemporaries, a very serious, thoughtful, and politically progressive person.9 Still, most of us would say that, in retrospect, her beliefs about sterilization were wrong and destructive, primarily because they were based upon background scientific and social theories that we now know to be empirically false. Were we to use time travel miraculously to inject our contemporary knowledge of genetics into her constellation of beliefs, Sophia would have to prune many of her considered judgments and background theories right down to their stumps. Here we need to make an important distinction between justifying moral theories or positions on controversial issues, on the one hand, and, on the other, asking whether a given agent is justified in believing some proposition. With regard to this latter question, we can ask, as we did in Rush’s case, whether a given agent is ‘justified’ in making the claims that he does. Perhaps he really isn’t epistemologically ‘entitled’ to make various claims because he is lazy or uninformed. Although important in its own right, this latter question is not within the original purview of the method of RE, which is concerned with the marks or criteria that are most likely to indicate the presence of a true belief. More specifically, the method of WRE claims that any given belief is most likely to be true if (and only if) that belief can be shown to belong to a maximally coherent set of other beliefs.10 The case of Sophia shows us, then, that even when we abstract from the question of whether a particular agent is epistemologically entitled to believe a proposition, we can still raise questions about the truth of beliefs that happen to meet the coherence criterion of justification. Thus, even supposing that Sophia had subjected her belief in eugenic sterilization to the test of maximal coherence, and even if that belief had passed the test in 1920, we would now say that it was wrong, both because it was based upon what we now know to be junk science, but also because we now place a 9 I don’t mean to suggest here that enlightened opinion at that time was unanimously in favor of eugenic sterilization. There were eloquent dissenters, such as G. K. Chesterton, but they were unfortunately in the minority (Chesterton 1927). 10 I thank Alex London for helpful discussion on this point.

the way we reason now

65

much higher value upon individual autonomy and bodily integrity than people did back then. Our reflections on these two hypothetical individuals yield the following preliminary conclusions. First, any given agent’s state of reflective equilibrium can fail to justify various elements of his moral outlook and practice if that agent is sufficiently obtuse, inexperienced, unimaginative, callous, or otherwise psychologically warped. So when Daniels defines WRE as the search for the ‘widest possible’ network of coherent beliefs, we have to ask whether we should develop an objective or subjective gloss on ‘widest possible’. A subjective gloss, as might be rendered by Rush at the nadir of his immersion in right-wing ideology, would appear to be a wholly inadequate basis of moral justification. Second, someone can possess all the requisite ‘epistemic virtues’—i.e. intelligence, open-mindedness, zeal in the pursuit of truth, etc.—and yet still be wrong about some very important moral issues, such as eugenic sterilization. So even if we grant that, in cases like Sophia’s, WRE can yield maximal coherence, we have to acknowledge that this is not the same thing as actually yielding moral truth. So even if we insist on a more objective gloss on the notion of the ‘widest possible’ network of beliefs in Daniels’s definition of WRE—a gloss that would include reference to all those beliefs that a reasonable person would have to consider in her quest for WRE—this would mark a big improvement, but would still tend to underscore the important difference between the weak sense of moral justification provided by RE and moral truth. Another way to put this is to say that while any true moral belief is likely to pass the test of WRE, merely passing this test does not guarantee that any particular belief will, in fact, be true. The second major problem with coherence theories of moral justification is summed up in what I’ll call the ‘circularity objection’. According to this objection, RE yields results as reliable (or unreliable) as the considered moral judgments that get the ball rolling. As we have already seen, if we limit ourselves to the relatively narrow ambit of NRE, the problem is obvious: considered moral judgments give rise to moral principles that ‘explicate’ or ‘fit’ them, yet why should we credit the moral bona fides of those considered judgments? If we cannot give a satisfactory answer to this question, we will be highly motivated to adopt WRE as our method, since it promises to advance both moral and empirical background theories to test the adequacy of our considered judgments and our principles derived from them. Now, it won’t do us much good at this juncture if the background theories deployed in WRE turn out to owe their credibility to the same considered moral judgments that gave rise to our moral principles. That would obviously amount to question begging. So defenders of WRE must contend that the background theories they endorse, both moral and empirical, do not owe their existence or credibility to the same set of considered moral judgments that animate NRE. They must, that is, demonstrate that the considered moral judgments that ultimately give rise to various morally saturated background theories (e.g. theories of the person,

66

john d. arras

of the role of morality in society, etc.) are somehow independent of those that shaped the moral principles in NRE. Daniels calls this the ‘independence constraint’ (1979: 260). This is a plausible defense of WRE, and Daniels carefully works it out with his usual blend of scholarly precision and philosophical sophistication. There is, however, one remaining problem. Supposing that the background theories we use to discipline both our considered moral judgments and our principles all satisfy the independence constraint; suppose, in other words, that whatever moral judgments go into the development of those theories are drawn from a different set than those that go into the development of our moral principles. This would solve the circularity problem, but it remains something of a mystery how it is that the stock of our considered or ‘pretheoretical’ moral judgments could be compartmentalized in this way (Haslett 1987). One would naturally think that all of our considered moral judgments would be pretty much cut from the same cloth of moral sensibility. If they are, then the circularity objection re-emerges. But if they aren’t, if some of our considered moral judgments can morph into background theories capable of criticizing the considered judgments that animate NRE, then we would have to explain how we are capable of generating two conflicting sets of considered moral judgments within the structure of the same moral personality, only one of which falls under the purview of NRE. It’s unclear, to me at least, how this can be done. It’s also unclear why this second set of considered judgments, lurking in the background until the stage of theory formation in RE, would have any greater degree of epistemic warrant than the first set that went into the formation of NRE. The fact that, as incorporated into morally informed background theories, this latter set of judgments might be used to criticize the first set does not establish that they actually provide firmer moral footing. The fact that they are different does not necessarily make them more trustworthy.

Conclusions ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

At the end of this long rumination on method in bioethics, we can draw together our conclusions. First, we can say that RE deserves its status as a regulative ideal for achieving ethical justification, but only on certain conditions. We are justified in our actions and beliefs to the extent that we have maximal confidence in them, and there appears to be no better way of achieving such confidence than by testing those beliefs against the widest possible set of other beliefs, including those that conflict with the belief in question. In this connection, RE is clearly a powerful engine of rationality and consistency, which afford it a good measure of critical edge. On the other hand, we have seen that mere confidence in our beliefs is not enough, and that this confidence must be sufficiently warranted. It’s not enough, in other

the way we reason now

67

words, merely to achieve coherence among the beliefs that we happen to have on hand, no matter what our degree of intelligence, our level of empathy, and moral perspicacity, and no matter what efforts we have made to enhance the quality and credibility of those beliefs. An insensitive, provincial, and intellectually lazy dullard, like our hypothetical friend Rush, may have a coherent set of beliefs in which he is supremely confident, but most of us would (correctly) see this not as a justification of those beliefs, but rather as evidence of his intellectual and moral inadequacy. We can thus conclude that RE may be a necessary condition of moral justification, but coherence alone is not sufficient. Second, we have seen that even warranted confidence in a coherent system of beliefs, such as that possessed by our other hypothetical friend Sophia—i.e. confidence earned through the diligent application of intelligence and moral perception to the thorough testing of all the various strata in our system of belief—we have seen that even this kind of moral justification is not enough to guarantee moral truth or credibility. Sophia was well brought up, smart, and progressive, and conscientiously tested her beliefs against the best that contemporary religion, genetics, and social science had to offer. Although many of us might perhaps still fault her for a lack of empathy for the victims of involuntary sterilization, she could make a strong case that her belief system was justified (because maximally coherent) even though its scientific bases were later shown to be manifestly false. So even though the process of RE at its very best can promise ethical justification as a regulative ideal, we should not mistake this for a promise of ethical truth. Our third conclusion has been that wide RE is an unattainable ideal, especially in practical contexts, and that both its requirements (e.g. the critical evaluation of all live options in moral and social theory) and its putatively global scope will have to be significantly scaled back before this method can achieve traction in an area like bioethics. If we take seriously the suggestion developed above that the most we are likely to get is local, rather than global, coherence, then WRE, at least in its most ambitious global form, is not even a necessary condition for moral justification. Finally, our fourth conclusion is that even if WRE, properly hedged with the conditions stipulated in conclusion 1, can serve as a regulative ideal of ethical justification, we should not expect it to deliver intersubjective convergence around particular moral judgments bearing on actions and policies. As we have seen, the fact of ethical, cultural, and religious pluralism within contemporary cosmopolitan societies will insure a broad multiplicity of reflective equilibria bearing on bioethical questions. Each of these differing sets of belief in equilibrium will be shaped by differing respective visions of the good belonging to the agents in question. So even though we can reasonably expect WRE to have some salutary potential for rendering our beliefs more justifiable, and even though WRE practiced on the societal level might occasionally be helpful in bringing the disparate members of society together on some issues, we should not expect this method to deliver intersubjective agreement on most bioethical controversies.

68

john d. arras

Some Final Thoughts on Method ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Similar deflationary thoughts loom over the very notion of method in moral inquiry. As we have seen, the unspoken assumption behind appeals to method in various disciplines and practices is that following all the prescribed steps will at least make it more likely that we will reach correct results. At its most daring, some might say presumptuous, this assumption amounts to the claim that just about anybody can reach the correct result, just so long as he or she adheres to the proper method. In the domain of morality this would mean that no matter who employs it, the method of RE can be counted upon to lead the enquirer to moral justification and truth. The focus here is on purifying, as far as possible, our initial considered judgments, and then systematizing all the disparate elements of our moral and empirical inventories. As DePaul (1993: 181) observes, this project fits within a very intellectualist or even scientistic picture of moral philosophy. Now, the main point that I want to make in closing is that if the project of RE actually harbors the above assumption about the power of method, then we should abandon that assumption. It is highly unlikely that method can live up to this billing anywhere else than in mathematics and the physical sciences, if even there. If we continue to speak of method in the area of moral philosophy and practical ethics, we will have to lower our expectations considerably. In this domain, while it is certainly important to try to get the various sectors of our moral world to fit together into as coherent a whole as possible, it is equally if not more important, as Nussbaum (1992) urges, to try to improve our capacity for making sensitive and discriminating moral judgments. As DePaul (1993: 174) puts it, tweaking what he takes to be the intellectualist–scientistic pretensions of RE, we need to spend equal if not more amounts of time attempting to refine the measuring instrument itself—i.e. our own sensibility as moral agents. Thus, he suggests that we need to focus more on developing the inquirer’s capacity for making discriminating moral judgments, rather than merely tidying up and systematizing whatever judgments this capacity happens to crank out. If DePaul is right about this, then our efforts to achieve RE would need to be supplemented, as I have suggested above in conclusion 1, by traditional Aristotelian concerns about character formation and training in virtue. And within the story of RE’s development, we would have to dust off and rehabilitate Rawls’s initial emphasis, dropped from later versions of the method, on the important role of ‘competent moral judges’ in this scheme (Rawls 1951). Were we to go this route, then the cultivation of good judgment would share center stage with the cultivation of consistency and coherence; and formative works of art, such as novels and films, would assume much greater importance both in moral education and in our quest to improve the output and reliability of RE. What, then, can or should we expect of method in the domain of practical ethics? While this is surely the subject of another paper, a few modest propositions can be

the way we reason now

69

advanced here. First, if method is to play a role in our work, it cannot be the role that Descartes (1637/1967) envisioned for it in ‘first philosophy’, or metaphysics. That is, we should not conceive of method as something we should first have to learn before we go off and start arguing about cloning, justice in health care, and randomized clinical trials in Africa. It is highly unlikely that there are any sure-fire steps that, once mastered, will lead us to correct results. Second, we need to distinguish between an idealized method of justification like WRE and helpful suggestions for ‘going forward’ in ethical analysis and argument (London 2000). As we have seen, in order to head off charges of circularity and insularity, NRE had to morph into WRE; but WRE turned out to be hopelessly clunky and complex as a recipe for moral justification. Not only do we have to make sure that our considered judgments and moral principles are in equilibrium, but we also have to consider all the available ‘moral outlooks’ and their respective philosophical defences, as well as all those background theories—a morning’s work before breakfast for Hercules, perhaps, but a crippling justificatory burden for us mortals. So if we are going to extend the function of RE from contractualist theory building a` la Rawls to practical ethics, we will have to direct those pruning shears to WRE itself. In the vast majority of cases (but not in all, to be sure), it will probably turn out that NRE, for all its limitations, will be a tolerably helpful methodological tool. Such an approach would, I suspect, be more or less extensionally equivalent to the more sophisticated varieties of casuistry that envision a dialectical relationship between case judgments, moral principles, and background institutions (Jonsen 1995; Kuczewski 1997). Finally, just as we should not view method in bioethics as a foolproof guarantor of justification—no matter how callow or unperceptive the moral agent might be—neither should we view it as a complete waste of time in our search for answers. Anyone who has seriously studied the methodological reflections of principlists, casuists, feminists, narrativists, and pragmatists realizes that each of these disparate approaches to thinking about morality points us in helpful directions. It is important to know, for example, about the processes of specifying, weighing, and balancing competing principles; about reasoning by analogy from paradigm cases; about the effects of power and domination on moral relationships; about the importance of narrative for comprehending the rich texture of practical situations; and, finally, about the flexible, tool-like character of moral principles and the importance of deliberative processes for bioethics. These are all important lessons that will surely improve our practical reasoning and facilitate our quest for moral justification, even if they won’t guarantee truth.

References Arras, J. D. (1991), ‘Getting Down to Cases: The Revival of Casuistry in Bioethics’, Journal of Medicine and Philosophy, 16/1: 29–51.

70

john d. arras

Arras, J. D. (1997), ‘Nice Story, But So What? Narrative and Justification in Ethics’, in H. Nelson (ed.), Stories and Their Limits: Narrative Approaches to Bioethics (New York: Routledge), 65–88. (2002), ‘Pragmatism and Bioethics: Been There, Done That’, Journal of Social Philosophy and Policy, 19/2: 29–58. (2003), ‘The Owl and the Caduceus: Does Bioethics Need Philosophy?’, in F. G. Miller, J. M. Humber, and J. C. Fletcher (eds.), The Nature and Prospect of Bioethics ( Totowa, NJ: Humana Press), 1–42. Barry, B. (1989), Theories of Justice (Berkeley: University of California Press). Beauchamp, T. L. (1984), ‘On Eliminating the Distinction Between Applied Ethics and Ethical Theory’, Monist, 67/4: 514–31. and Childress, J. (1979–2001), Principles of Biomedical Ethics (New York: Oxford University Press; 1st edn., 1979; 2nd and 3rd edns., 1983; 4th edn., 1994; 5th edn., 2001). Benjamin, M. (2002), Philosophy and This Actual World: An Introduction to Practical Philosophical Inquiry ( Totowa, NJ: Rowman & Littlefield). Brock, D. (1995), ‘Public Moral Discourse’, in R. E. Bulger, E. M. Bobby, and Harvey V. Fineberg (eds.), Society’s Choices: Social and Ethical Decision Making in Biomedicine (Washington, DC: National Academy Press), 215–40. Chesterton, G. K. (1927), Eugenics and Other Evils (New York: Dodd, Mead). Daniels, N. (1979), ‘Wide Reflective Equilibrium and Theory Acceptance in Ethics’, Journal of Philosophy, 76: 256–82. (1996), Justice and Justification: Reflective Equilibrium in Theory and Practice (Cambridge: Cambridge University Press). and Sabin, J. E. (2002), Setting Limits Fairly: Can We Learn to Share Medical Resources? (New York: Oxford University Press). DePaul, M. (1993), Balance and Refinement: Beyond Coherence Methods of Moral Inquiry (London: Routledge). Descartes, R. (1637/1967), ‘Discourse on Method’, in The Philosophical Works of Descartes, i, ed. E. S. Haldane and G. R. T. Ross (Cambridge: Cambridge University Press), 79–130. Dworkin, R. (1977), ‘Hard Cases’, in Dworkin, Taking Rights Seriously (Cambridge, Mass.: Harvard University Press), 81–130. Griffin, J. (1996), Value Judgement: Improving Our Ethical Beliefs (Oxford: Clarendon Press). Hare, R. M. (1963), Freedom and Reason (Oxford: Oxford University Press). (1975), ‘Rawls’ Theory of Justice’, in N. Daniels (ed.), Reading Rawls (New York: Basic Books), 82–107. Haslett, D. W. (1987), ‘What Is Wrong With Reflective Equilibria?’, Philosophical Quarterly, 37/148: 305–11. Jonsen, A. R. (1995), ‘Casuistry: An Alternative or Complement to Principles?’, Kennedy Institute of Ethics Journal, 5: 237–51. and Toulmin, S. (1988), The Abuse of Casuistry: A History of Moral Reasoning (Berkeley: University of California Press). Kagan, S. (1998), Normative Ethics (Boulder, Col.: Westview Press), 11–16. Kuczewski, M. (1997), ‘Bioethics’ Consensus on Method: Who Could Ask for Anything More?’, in H. L. Nelson (ed.), Stories and Their Limits: Narrative Approaches to Bioethics (New York: Routledge), 134–48.

the way we reason now

71

London, A. J. (2000), ‘Amenable to Reason: Aristotle’s Rhetoric and the Moral Psychology of Practical Ethics’, Kennedy Institute of Ethics Journal, 10/4: 287–305. McGee, G. (ed.) (2003), Pragmatic Bioethics, (2nd edn. Cambridge, Mass.: MIT Press). Nagel, T. (1991), ‘The Fragmentation of Value’, in Nagel, Mortal Questions (Cambridge: Cambridge University Press), 128–41. Neurath, O. (1966), ‘Protocol Sentences’, trans. G. Schick, in A. J. Ayer (ed.), Logical Positivism (New York: Free Press), 199–208. Nussbaum, M. (1992), ‘Perceptive Equilibrium: Literary Theory and Ethical Theory’, in Nussbaum, Love’s Knowledge: Essays in Philosophy and Literature (New York: Oxford University Press), 168–94. Petersson, B. (1998), ‘Wide Reflective Equilibrium and the Justification of Moral Theory’, in van der Burg and van Willigenburg (1998: 130–3). Pollock, J. L. (1986), Contemporary Theories of Knowledge ( Totowa, NJ: Rowman & Littlefield). Rawls, J. (1951), ‘Outline of a Decision Procedure for Ethics’, repr. in S. Freeman (ed.), John Rawls: Collected Papers (Cambridge, Mass.: Harvard University Press, 1999), 1–19. (1971/1999), A Theory of Justice (Cambridge, Mass.: Harvard University Press; rev. edn., 1999). (1975), ‘The Independence of Moral Theory’, in John Rawls: Collected Papers, ed. S. Freeman (Cambridge, Mass.: Harvard University Press, 1999), 286–302. Raz, J. (1982), ‘The Claims of Reflective Equilibrium’, Inquiry, 25/3: 307–30. (1994), ‘The Relevance of Coherence’, Ethics in the Public Domain: Essays in the Morality of Law and Politics (Oxford: Clarendon Press), 277–325. Scanlon, T. M. (1992), ‘The Aims and Authority of Moral Theory’, Oxford Journal of Legal Studies, 12/1: 1–23. Scheffler, S. (2001), ‘Individual Responsibility in a Global Age’, in Scheffler, Boundaries and Allegiances: Problems of Justice and Responsibility in Liberal Thought (New York: Oxford University Press), 32–47. Singer, P. (1974), ‘Sidgwick and Reflective Equilibrium’, Monist, 58/3: 490–517. Toulmin, S. (1981), ‘The Tyranny of Principles’, Hastings Center Report, 11/6: 31–9. van der Burg, W., and van Willigenburg, T. (eds.) (1998), Reflective Equilibrium: Essays in Honor of Robert Heeger (Dordrecht: Kluwer Academic Publishers). Wolf, S. (ed.) (1995), Feminism and Bioethics: Beyond Reproduction (New York: Oxford University Press).

chapter 3 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

AUTONOMY •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

bruce jennings

Introduction ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

No single concept has been more important in the contemporary development of bioethics, and the revival of medical ethics, than the concept of autonomy, and none better reflects both the philosophical and the political currents shaping the field. In this chapter I propose to consider autonomy in three of its facets and functions: first, as a concept in ethical theory; second, as a concept in applied ethics; and finally, as what might be called an ideological concept—that is, one that both draws from and reinforces non-philosophical interests at work in the profession of medicine, biomedical science and technology, and the broader liberal individualistic culture of Anglophone countries, particularly the United States, where a bioethical discourse centred on autonomy has flourished. Two theses stand behind this discussion. First, there is a significant difference between the meaning and history of the concept of autonomy in moral philosophy (and political theory) and its appropriation in the normative and applied work of bioethics. This difference is often overlooked, and the deployment of autonomy in bioethics has usually been presented as the straightforward ‘application’ of a philosophically grounded concept and principle to particular cases or decisionmaking situations. In fact, the relationship between the meaning of autonomy in bioethics and its status in deontological or rights-based theory—particularly its most natural homes, contractarian and libertarian liberalism—is highly selective and tenuous. This is best illustrated, I believe, by contrasting autonomy as Kant

autonomy

73

understood it with liberty as it has been understood by John Stuart Mill and more recently Isaiah Berlin. Indeed, to talk about ‘autonomy’ in bioethics is something of a misnomer and a category mistake; the concept that is actually much more influential in bioethics is a version of Millian ‘liberty’, the version that Berlin dubbed ‘negative liberty’. Bioethicists by and large have been suspicious of ‘positive liberty’, as Berlin himself was, and which is closely akin to Kantian or rationalistic autonomy. Modern bioethics has largely rejected the theoretical tradition and legacy of autonomy not so much by explicit critique as by rendering it invisible. This is done when the term ‘autonomy’ is used to mean and to do the ethical work of negative liberty. My second thesis is that this conceptual displacement has occurred in bioethics partly owing to the sociology of the field and partly owing to the conditions that have been required of it to gain intellectual legitimacy (and thereby some measure of power and influence) in professional and policy domains.

Autonomy’s Influence in Bioethics ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

I Beginning with the Belmont Report of the National Commission on Ethical Issues in Human Subject Research (or perhaps even earlier, with Telford Taylor’s resounding statement at the Nuremberg Doctors’ Trial), the concept of autonomy (sometimes referred to as ‘self-determination’ or ‘respect for persons’) has played a central role in the modern field of bioethics. This centrality was reinforced by the tremendous influence of the seminal work of Tom Beauchamp and James Childress The Principles of Biomedical Ethics, first published in 1979, which has since become the standard work in the field and has gone through four significant revisions and expansions. There the principle of autonomy is presented as a middle-range principle, not dependent upon any single more general or abstract ethical theory and compatible with many. It stands alongside the other principles in the Beauchamp and Childress edifice—beneficence, non-maleficence, and justice—as a primus inter pares. In soon-to-follow agenda-setting documents for the field, such as the influential reports of the President’s Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research in the early 1980s, the principle of selfdetermination again took a leading role in the analysis and recommendations made about everything from whistle blowing to genetic testing to health care access to decisions to forgo life-sustaining treatment. Before long autonomy-centred work in bioethics was beginning to appear in other countries beside the United States, in the Canadian Law Reform Commission, in the Waller Commission in Australia, in the work of the noted Australian philosopher Max Charlesworth (1993), and in the United Kingdom with the work of Raanan Gillon (1986), to name only a few.

74

bruce jennings

This is not the proper place to rehearse all the countless ways and occasions in which the concept of autonomy has been pressed into service in bioethics. By the late 1960s and early 1970s it was an idea whose time had come. Let four areas of debate in bioethics provide brief examples to indicate the power of autonomy (or, more precisely, various conceptions of autonomy, for there are several) in bioethics. These are abortion, medically assisted reproduction, genetic counselling, and decisions to forgo life-sustaining medical treatment.

II One thinks first of the evolution of the abortion debate during these years, particularly polarizing in the United States, and culminating in the landmark decision of the US Supreme Court Roe v. Wade, in which the woman’s autonomous control of her own body and her own pregnancy became the central tenet. Autonomy was clearly central for those who did not ascribe moral significance to the early embryo and fetus. But even more interesting and indicative of the moral force of the concept of autonomy was an argument put forward by Judith Jarvis Thomson (1971), who was willing to grant even personhood status to the fetus and yet still argued that the pregnant woman’s autonomy overrode any moral claims of fetal persons. As the pro-choice or pro-abortion line of argument developed, the question of the morally right choice came to be subsumed under the question of who had the moral right to choose. Some, like Daniel Callahan (1970), would argue that the law should not interfere with a woman’s autonomy by prohibiting abortion, but that from a moral point of view there should be good reasons, some account beyond appeal to the woman’s subjective will alone, for terminating a pregnancy; good reasons, that is, for cashing in one’s autonomy. In general, I think it is fair to say that Thomson’s position prevailed over Callahan’s position, but in so far as they both appealed to moral reasons (other than the autonomous right to choose) to justify a particular choice, neither of these positions gave the concept of autonomy the strength that it acquired in the more popular, less philosophical debate. The abortion controversy became so polarized and so bitter that asking for moral reasons for autonomous choice or action was rejected by the pro-choice side of the debate as a covert means of taking autonomy and power away from women. Being autonomous came to mean not being second-guessed by anybody or anything about one’s choices or the use that one makes of one’s freedom. Closely related was the larger issue of medically assisted reproduction and reproductive autonomy. New technologies and drugs to overcome infertility, such as in vitro fertilization (IVf) grew in efficacy and popularity in the 1980s. Autonomous control of one’s gametes—for donation, stimulation, implantation, and cryo-preservation—was the dominant orientation of bioethics, as various

autonomy

75

early restrictions were set aside that would make IVF and related technology accessible only to married heterosexual couples, or only to those couples who could demonstrate good health and good character and potential to be good parents and to provide well for the child. Concerns about the unknown effects on the child, psychological effects, social stigma, even possible genetic damage caused by the procedure itself were all set aside as so much idle speculation, and, in any case, were not considered sufficiently weighty to override the moral importance of protecting autonomy and choice for infertile men and women (Robertson 1994). In the new field of genetic counselling, dealing as it did with a sordid history of racism, discrimination, and eugenics, the concept of respecting autonomy was the foundation upon which the doctrine of non-directive counselling was built. That same orientation and issue quickly moved over into the field of reproductive counselling as well. The ideal of ‘non-directive’ behaviour by the counsellor raises a number of interesting problems in relation to the concept of autonomy. In essence, this doctrine limits the counsellor’s role to providing clear, understandable factual information about test results, treatment options, and the like. Counselling should never discuss which course of action may be more reasonable, wise, prudent, or ethically appropriate. Even non-binding advice on such matters is regarded, not as helpful additional information, but rather as intrusive, even coercive, influence by a professional authority figure. As such it interferes with, instead of assisting or promoting, the client’s exercise of autonomy. This makes sense when autonomy is conceived as being the independent exercise of individual will or choice, but it seems odd and overly restrictive if autonomy requires some degree of reason and is subject to some norms of cogency and correctness (perhaps norms built into the very concept of reason itself). Here many of the difficult issues related to the concept of autonomy become apparent. If we stress the notion of reason and rational choice, how tightly do the norms of reason constrain action? Is there one right and reasonable decision to be made in each situation, and are only those who embrace the ‘right’ choice truly autonomous? On the other hand, if we stress the notion of liberty and independent choice, does the value of autonomy then preclude all other ethical judgements concerning actions (including perhaps the agent’s own reflexive judgements)? Does it set up an internal conceptual conflict between autonomy and all other ethical values? Kant, for example, held that to be autonomous was to be rational in the sense of adopting principles that met certain conceptual tests, such as universality. John Stuart Mill held that diversity and experiments in ways of life were important and that what is the right way to live for one person may not be the right way for another. Finally, a century later Isaiah Berlin developed this side of Mill’s thought even further and eloquently articulated a defence of liberty and ethical pluralism that was quite far removed from Kant. Often, in bioethics, as is illustrated by the

76

bruce jennings

doctrine of non-directive counselling, the term ‘autonomy’ more closely resembles Berlin’s liberty than Kant’s reason. Finally, in the so-called right to die area, the concept of autonomy plays a central role in the notion that persons have the right to forgo life-extending treatment, even when they have lost decision-making capacity and another person must exercise that autonomy right on their behalf. Until the 1990s the boundaries of autonomy in end-of-life medical care had been drawn by the active–passive or killing–letting die distinction. The national grassroots efforts to legalize assisted suicide (in Oregon and briefly in the Northern Territory of Australia) and even active euthanasia by lethal injection (in the Netherlands) pushed the domain of autonomy further than the court cases and laws on the right to refuse treatment. Few official documents pertaining to bioethics are more influenced by the ethical concept of autonomy than the majority ruling of the US Ninth Federal District Court (covering the states of California, Oregon, and Washington) striking down the state laws criminalizing physician-assisted suicide in those jurisdictions (Compassion in Dying v. Washington 1996).

III Further testimony to the influence of autonomy in bioethics is the fact that autonomy is facing a backlash. Patients do not always want to be the partners of their doctors in making health care decisions, argues Carl Schneider in The Practice of Autonomy (1998). Feminist bioethicists and others have tried to blunt the radical individualism that use of the concept of autonomy so often seems to imply. Such arguments can be found in work by George Agich, Autonomy and Long Term Care (1993), in the important collection Relational Autonomy, edited by Catriona Mackenzie and Natalie Stoljar (2000), and in my own work with Willard Gaylin, The Perversion of Autonomy (Gaylin and Jennings 2003). Another notable work in this vein is Onora O’Neill’s book Autonomy and Trust in Bioethics (2002: 75), where she argues: The claims of individual autonomy, in particular of patient autonomy and reproductive autonomy, have been endlessly rehearsed in bioethics in recent decades. By themselves . . . conceptions of individual autonomy cannot provide a sufficient and convincing starting point for bioethics, or even for medical ethics. They may encourage ethically questionable forms of individualism and self-expression and may heighten rather than reduce public mistrust in medicine, science and biotechnology. At most individual autonomy, understood merely as an inflated term for informed consent requirements, can play a minor part within a wider account of ethical standards.

O’Neill holds that the concept of autonomy that bioethics has appropriated bears little resemblance to Kant’s understanding and use of this concept, and she offers a counter-reading of Kant, stressing the role of practical reason in his ethics and

autonomy

77

the fundamental importance, in the ethical life, of the social fabric of mutual and institutional trust.

Autonomy: A Complex Concept ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

I Let us turn now to matters of definition and the place of autonomy in ethical theory outside bioethics. The word ‘autonomy’ comes from the Greek autos (meaning ‘self’) and nomos (meaning ‘rule’, ‘governance’, or ‘law’). Literally, therefore, ‘autonomy’ means ‘the state of being self-governed or self-sovereign’; living autonomously means living by a law that you impose on yourself. In other words, autonomy is the right to live your own life in your own way. Beauchamp and Childress (1994: 121) state that autonomy is the ‘personal rule of the self that is free from both controlling interference by others and from personal limitations that prevent meaningful choice. . . . The autonomous individual freely acts in accordance with a self-chosen plan.’ Numerous social and political philosophers express the same general idea. ‘To regard himself as autonomous’, says Thomas Scanlon (1972: 215), ‘a person must see himself as sovereign in deciding what to believe and in weighing competing reasons for action.’ For his part, Robert Paul Wolff (1970: 14) places autonomy in the context of a struggle among human wills for control: As Kant argued, moral autonomy is a submission to laws that one has made for oneself. The autonomous man, insofar as he is autonomous, is not subject to the will of another. The autonomous . . . man may do what another tells him, but not because he has been told to do it . . . . By accepting as final the commands of the others, he forfeits his autonomy.

Other philosophers and psychologists, such as R. S. Peters and Lawrence Kohlberg, say that autonomy requires a deliberate self-consciousness about obedience to rules. They place it at the pinnacle of moral development. ‘Children finally pass to the level of autonomy’, Peters writes (1972: 130), ‘when they appreciate that rules are alterable, that they can be criticized and should be accepted or rejected on a basis of reciprocity and fairness. The emergence of rational reflection about rules . . . central to the Kantian conception of autonomy, is the main feature of the final level of moral development.’ Kohlberg (1981) also pictures the best moral agent as standing judgementally above the existing rules, laws, traditions, habits, and norms of his society, and choosing with rational detachment which rules to follow and which to disregard. Autonomy is a catalyst for moral growth and social enlightenment. It is a sign of moral adulthood in the individual; only autonomous persons are truly grown-up, regardless of their age. Kant (1963) likewise saw autonomy as a sign of the enlightenment and maturity of mankind.

78

bruce jennings

Lawrence Haworth, author of one of the few book-length studies of the concept of autonomy, defines autonomy as having what he calls ‘critical competence’. He writes: ‘Having critical competence a person is first of all active and his activity succeeds in giving effect to his intentions. Having critical competence, the active person is sensitive to the results of his own deliberation; his activity is guided by purposes he has thought through and found reasons of his own for pursuing’ (1986: 46). These philosophical views of autonomy are widely shared in intellectual circles. Perhaps they are summed up most succinctly, if not most grammatically, by the philosopher Joel Feinberg: ‘I am autonomous if I rule me, and no one else rules I’ (1972: 161).

II I rule me, myself. It is important to note that autonomy is not a single idea but a cluster of closely related, overlapping ideas. And it is not only a philosophical or theoretical construct. The idea, if not the actual term, is part and parcel of contemporary culture and everyday self-consciousness in the West. One of the things that has distinguished bioethics from academic philosophy is that each has a different audience. If one wishes to address a lay audience or a professional audience that is not widely read in philosophy, then one both must use the concept of autonomy, if one is to be appealing and understood, and must only use it in certain ways. This is why O’Neill’s attempt to reconstruct and make more robust the concept of autonomy, whatever its appeal to other professional philosophers, is quite difficult within the context of bioethics, at least within American bioethics. Similarly, her attempt to concentrate attention elsewhere in our ethical understanding of medicine, to focus on the crucial concept of trust, for example, will also be difficult to achieve. The fact is that in the broader audience to which bioethics must appeal, the need to mean much more than ‘informed consent requirements’ when one appeals to autonomy is not apparent. Nor is this audience likely to agree on which forms of individualism and self-expression are ethically questionable. Again, at least in the American context, individualism, self-expression, and autonomy are closely conjoined and enormously appealing.

Autonomy and the Cultural Context of Bioethics ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

I The idea of autonomy in contemporary culture and society has been the focus of important studies by leading social theorists and social scientists. How far back one

autonomy

79

should trace this is debatable (Tocqueville?), but the groundwork for modern work on autonomy in American sociology was already established in the influential work of David Riesman et al., The Lonely Crowd (1950).1 A generation later, an implied critique of autonomy can be found in what the historian Christopher Lasch (1978) called the ‘culture of narcissism’. It has affinities as well with what the philosopher Charles Taylor (1991) has referred to as the ‘ethics of authenticity’. Robert Bellah and his colleagues (1985) called it ‘expressive individualism’. Finally, at the heart of what I am concerned with under the heading of autonomy is a notion that the sociologist Alan Wolfe (2001) has called ‘moral freedom’. He defines his subject in the following way: The old adage that America is a free country has, at last, come true, for Americans have come to accept the relevance of individual freedom, not only in their economic and political life, but in their moral life as well. The defining characteristic of the moral philosophy of the Americans can therefore be described as the principle of moral freedom. Moral freedom means that individuals should determine for themselves what it means to lead a good and virtuous life. Contemporary Americans find answers to the perennial questions asked by theologians and moral philosophers, not by conforming to strictures handed down by God or nature, but by considering who they are, what others require, and what consequences follow from acting one way rather than another. Some of our respondents adopt moral freedom as a creative challenge . . . they find themselves quite comfortable with the idea that a good society is one that allows each individual maximum scope for making his or her own moral choices. (2001: 195)

Wolfe goes on to argue that the widespread (but by no means universal) embrace of moral freedom amounts to a cultural revolution in America: never have so many people been so free of moral constraint as contemporary Americans. Most people, throughout most of the world, have lived under conditions in which their morality was defined for them. Now, for the first time in human history, significant numbers of individuals believe that people should play a role in defining their own morality as they contemplate their proper relationship to God, to one another, and to themselves. (2001: 199)

Many social commentators have been critical of the autonomy orientation of contemporary culture. These writers usually direct their criticism of the culture of autonomy along one of two lines. One such line, developed by the theologian Stanley Hauerwas (1974), is that autonomy is morally flabby because it is little more than an ethic of selfishness and self-indulgence. In a similar expression of the same kind of criticism the sociologist Daniel Bell (1976) decried what he called the dominant ‘hedonism’ behind the cultural transformations begun in the late 1960s. The second line of criticism of the general cultural emphasis on individual autonomy is that the widespread deference accorded to autonomy today in the 1 Riesman employs the term ‘autonomous self ’ to designate an alternative character type that avoids the respective weaknesses of the ‘inner-directed’ and the ‘other-directed’ personalities. A useful sampling of social commentary on individualism and what we today would probably refer to as autonomy can be found in Bellah et al. (1987).

80

bruce jennings

law and in secular morality threatens to undermine the moral order itself. A spate of books published in the last few years has developed variations on this theme, including Robert Hughes (1993) and Jean Bethke Elshtain (1995). These books have in common a sense of disorder, albeit a disorder still just short of normlessness or anomie. Claims to rights and demands for dignity and respect, no matter how well deserved, are overloading the system. This sense of disorder and the need to respond aggressively to it are captured in the very titles or subtitles of many of these works: the ‘fraying of America’ (Hughes’s subtitle), ‘on trial’, ‘defence’, ‘death’, and ‘suffocating America’. In other works, America is being ‘de-moralized’ (Himmelfarb 1995) or ‘disunited’ (Schlesinger 1992).

II Many of these recent criticisms of the effects of the concept of autonomy in our moral discourse and in our social lives cast an overly broad net. Rather than subjecting autonomy to a more careful conceptual analysis, they tend to use the term (or the concept) as a lightning rod for many different social complaints and criticisms. To be sure, autonomy has been very influential in the broader moral culture as it has been in the field of bioethics. What is most needed now, however, is not a broad-brush critique but a more careful sifting of meanings and inconsistencies in the way the term is used. I have already mentioned the distinction between autonomy understood as the exercise of reason and autonomy understood as the exercise of liberty, and I will return to that distinction below. First, though, let us consider the broad and general meaning of autonomy further. Autonomy is not an ethic of selfishness, nor is self-indulgence the same as rational self-fulfilment. Autonomy means freedom from outside restraint and the freedom to live one’s own life in one’s own way. To be autonomous is to live by your own law, or only by laws that you have embraced and accepted as your own. It is to be self-sovereign. It is to be the author of your life, your self, and your actions. Nothing in these notions necessitates selfishness or egoism. Self-determining conduct need not be exclusively self-serving. Being the author of your own life says nothing per se about the moral content of that life. As an ethic, autonomy means living according to your own values and principles, as these are refined in the light of informed, rational deliberation and settled conviction. This ethic is a far cry from an anything-goes, do-your-own-thing morality. There is no obvious reason why an autonomous person cannot identify with principles of justice and moral virtue that are founded on reason. Not only is autonomy compatible with just and altruistic ideals, it seems to go naturally with them. The free and secure person is a more open and charitable person. It is insecurity about one’s worth that often breeds defensiveness and hostility.

autonomy

81

While autonomy is not an ideal of selfishness or disorderly self-indulgence, it is individualistic. It can be—according to most philosophers it has to be—based on rational will rather than licence; but it is the self’s reason, not that of a community or a higher authority. Autonomy’s mood is always possessive. It speaks in the first-person singular rather than the first-person plural. It is a vision of free Is, not free Wes. In order to illuminate the special role of autonomy in modern culture, Joel Feinberg (1980), a leading proponent of autonomy and civil liberties, has asked what would be missing if we lived in a society that did not understand the concepts of autonomy and rights. He imagined a prosperous and benevolent society, not a tyranny or a despotism. He imagined a society in which individuals possessed all the other virtues and lived by the Ten Commandments and the Golden Rule. Even in such a good and happy society, Feinberg said, something essential to morality and humanity would be lacking. Individuals would not be able to make claims on each other for dignity and respect. Autonomy allows one person to demand respect from another as a matter of right. Not to ask meekly for respect as a matter of goodwill, friendship, or charity, but to demand it. Proudly. And not necessarily on the basis of some accomplishment, but just because one is a person, an adult human being, a first-class member of the moral community. It is through the concepts of autonomy and rights that we gain our sense of identity as moral persons and free agents worthy of the respect of others. Self-sovereignty in the moral realm; the right to live your own life in your own way so long as you do others no harm; being true to yourself above all—these notions are at the core of individualism, authenticity, moral freedom, and autonomy, if not as all professional philosophers would define them, then at least as many ordinary people define them in their own social and self-consciousness.

III In sum, there are several key tenets that give autonomy its extraordinary moral power and appeal. They can be stated as follows. One tenet is moral individualism, the belief that the human individual is the centre of the moral universe, the subject of ultimate worth to whose well-being all other things that may be said to have moral value must finally contribute. Another is what might be called moral constructivism, the notion that the basic features of society are ultimately products of human choice and artifice, and as such the social world is a human world, a world fashioned not by God or nature but by the amazingly plastic, adaptive human will. The final root idea underlying autonomy is the doctrine of moral voluntarism or consent. The moral life of the individual is made up of relationships, commitments, and obligations that the individual has freely chosen, not those into which she was

82

bruce jennings

born or that were imposed upon her. Rather, autonomy means living in accordance with rules that one gives to oneself. If an autonomous person does her duty, it is because she has freely and rationally chosen to do so. Autonomy gives an inward turn to moral duty, obligation, or responsibility, grounding them not in nature or history but in the domain of will and rational choice.

Objective Autonomy (Reason) and Subjective Autonomy (Liberty) in Bioethics ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

I There are, of course, a number of ways in which the concept of autonomy can be used in ethical theory. It can be a good or goal to be obtained in some type of consequentialist theory, but such traits or aspects of human flourishing are usually differentiating concepts, something that only a few can obtain by special effort or circumstance, while autonomy is more commonly taken to be a universal potentiality, a common property or constitutive feature of our humanness as such. If a given individual does not possess autonomy it is rarely taken to be the result of failed effort and more often taken to be the result of some extrinsic condition, oppression, coercion, ignorance, or mental illness, for example. Thus autonomy is often used in deontological or contractarian type theories to designate a condition of mind and agency that adheres to individuals as a matter of right. The right to be allowed to be autonomous is an individual right or moral claim that can be made against others who might interfere with that condition and who thereby have a correlative obligation of non-interference with the autonomy of the other. Whether this is a negative obligation of forbearance only, or also a positive obligation of assistance, is a nice question. For our purposes here the most important set of theoretical questions about the concept of autonomy concerns the issue of how and on what basis autonomy imposes any moral restrictions on the will of the autonomous person. This has to do with the difference between what might be called an objective understanding of autonomy versus a subjective one. Alternatively, as I would prefer to formulate it, the question has to do with the relative emphasis one gives to reason or to liberty in one’s understanding of autonomy. This difference marks out the two principal pathways within the liberal tradition that the concept of autonomy has followed. The first may be seen in Locke and Rousseau, but certainly finds its most sophisticated philosophical expression in the ethical writings of Kant. The second can be seen as a course charted by John Stuart Mill in On Liberty and pushed even further a century later by Isaiah Berlin in his famous essay ‘Two Concepts of Liberty’.

autonomy

83

II Kant was more concerned with duty than with freedom.2 Autonomy is obedience to self-imposed law, but not just any law or indeed any aspect of the self. Autonomy is obedience to the moral law or categorical imperative as it is discerned by the self exercising reason. Autonomy is objective because reason and the moral law it discerns are objective. Moreover, Kant was primarily concerned with universalism and the formal preconditions for something to be a universal and rational moral law or principle. Mill was a naturalist and was more concerned with individual choice and action in a context of political and social constraint. Reason is not a formal and universal kind of knowing for Mill and it does not establish the formal preconditions for either autonomy or right. Reason is a faculty for determining conduct and for the choice of means to protect and preserve interests and desires, particularly those compatible with cooperative and mutually beneficial life with others similarly motivated and inclined. Hence, with Mill one finds a version of subjectivity introduced into the notion of autonomy (a term that Mill rarely used) because each person is the most reasonable custodian and definer of his or her own interests and objectives. If the power to determine those interests is exercised over one by others, especially by officials of the state, one is deprived of liberty (autonomy) and one is hampered in the development of intelligence, skill, and self-reliance that Mill considered to be some of the hallmarks of human flourishing. On this account of autonomy there is no independent standard of moral knowledge or reason to determine if one use of freedom is inherently superior to another; the individual should decide as a matter of right, and if individuals are permitted by social and political arrangements to have this liberty, the society as a whole will prosper and the arrangement will be justified from a utilitarian point of view. In this line of thinking, reason as a universal standard of right gives way to liberty as the exercise of judgement and choice by each individual person. Moral law and duty as bridles on natural desires and preferences give way to freedom as a claim-right to be asserted by the individual against others. Autonomy moves from being the basis of obligation to being the object of obligation. One obvious question arises at this point. If there is no universally rational or objective basis for distinguishing qualitatively among the choices individuals make or the way each uses liberty, on what basis can any limits to the exercise of individual liberty be set? Mill’s own attempt to answer this question is sketched in the following famous passage from On Liberty (1956: 13): the sole end for which mankind are warranted, individually or collectively, in interfering with the liberty of action of any of their number is self-protection . . . . the only purpose for which power can be rightfully exercised over any member of a civilized community, 2

In the following discussion I have relied heavily on Schneewind (1998) and Wolff (1970).

84

bruce jennings

against his will, is to prevent harm to others. His own good, either physical or moral, is not sufficient warrant. He cannot rightfully be compelled to do or forbear because it will be better for him to do so, because it will make him happier, because, in the opinions of others, to do so would be wise or even right. These are good reasons for remonstrating with him, or reasoning with, or persuading him, or entreating him, but not for compelling him or visiting him with any evil in case he do otherwise. To justify that, the conduct from which it is desired to deter him must be calculated to produce evil to someone else. The only part of the conduct of anyone for which he is amenable to society is that which concerns others. In the part which merely concerns himself, his independence is, of right, absolute. Over himself, over his own body and mind, the individual is sovereign.

This shifts the question from the definition of reasonable, rightful choice to the definition of harm and to discriminations among various kinds and degrees of harm. It is not obvious that this latter is an easier concept with which to deal, but harm does have a certain solidity, a naturalistic and tangible quality that the concept of reason seems to lack. Nonetheless, problems abound. Define harm too expansively, and autonomy shrinks. How serious does harm have to be? Is it only serious physical harm or injury, or do psychological damage, pain, and suffering count as well? What is the gradation between harm, offence, annoyance, and inconvenience? Where should we draw the line? Generally speaking, the species of harm that has most preoccupied bioethics is precisely the denial of autonomy or negative liberty of the other. Closely related, and of special importance in medicine, is the right to privacy and bodily integrity, in a word, informed consent. The autonomy or liberty limiting harms that are most prominent in bioethics are those things that encroach on my private space, those things that I don’t invite in, whether it be an experimental drug, a life-saving surgery, or the implantation of an artificial nutrition and hydration tube that serves to prolong my dying from an underlying incurable disease.

III There are some limitations on voluntary transactions imposed on grounds of moral standards, to be sure. But they are coming fully into the literature of bioethics only rather slowly, and they tend to come in the area of health policy and questions of the allocation of scarce resources. For example, the insurance system now confounds John Stuart Mill’s distinction between other-regarding and purely self-regarding behaviour. Corporations will soon begin to take a greater interest in the healthrelated behaviours and characteristics of their employees—diet, exercise, smoking, genetic make-up—in order to hold down their insurance costs. This interest poses a substantial threat to the privacy and autonomy of large numbers of individuals in the coming years. But for those bioethicists inclined to justify it, it will be justified, as autonomy as negative liberty holds that all abridgements of freedom must be, on the basis of protecting the rights of healthy, well-behaved, and genetically

autonomy

85

well-endowed employees against the costly profligacy and self-abuse of those who get sick and drive up health care costs. Not because the choices of the profligate are inherently objectionable, but because their social effects cause harm to others. Nonetheless, it is difficult, when discussing morality, especially within the domain of will, autonomy, and freedom, not to hold out some kind of interpersonal ideal, some notion of what counts as a higher kind of life and a better kind of choice. The tension between objective autonomy and subjective autonomy lingers even today in bioethics, although I believe that it is fair to say that the subjective conception of autonomy, which stresses liberty over reason, holds sway, subject pretty much only to those side constraints that Mill and his liberal successors have placed on it, such as the harm principle, justice, equity, and fairness. This is why it is helpful to consider Berlin as a kind of missing step as we move from Mill to modern bioethics. One rarely acknowledged source that anticipates, if it did not directly influence, an important intellectual trend in bioethics is the key distinction between negative and positive liberty that was coined and first suggested by Berlin in his ‘Two Concepts of Liberty’. Negative liberty has to do with establishing a zone of privacy and non-interference around each person, a zone within which the person can exercise his own faculties and pursue his own life in his own way. Berlin (1969: 127) explains the concept this way: The defence of liberty consists in the ‘negative’ goal of warding off interference. To threaten a man with persecution unless he submits to a life in which he exercises no choices of his goals; to block before him every door but one, no matter how noble the prospect upon which it opens, or how benevolent the motives of those who arrange this, is to sin against the truth that he is a man, a being with a life of his own to live.

Autonomy understood as negative liberty appeals to metaphors of space. It wants elbow room, a place of one’s own. It is the single-family dwelling of ethics. Negative liberty requires fences and boundaries for protection against outside intruders. It rests on a conflict-ridden and antagonistic picture of social existence, in which each individual struggles with everyone else to control his own patch of ground. Negative liberty does not appeal to Napoleons or Don Juans. It appeals to ordinary folks who simply wish not to be dominated, who wish to be left alone. Autonomy is also linked to liberty in another sense, through what Berlin and others have called positive liberty. Positive liberty is very close to self-mastery and detached judgement. Berlin (1969: 131) explicates it this way: The ‘positive’ sense of the word ‘liberty’ derives from the wish on the part of the individual to be his own master. I wish my life and decisions to depend on myself, not on external forces of whatever kind. I wish to be the instrument of my own, not of other men’s, acts of will. I wish to be a subject, not an object; to be moved by reasons, by conscious purposes, which are my own, not by causes which affect me, as it were, from outside. I wish to be somebody, not nobody; a doer—deciding, not being decided for, self-directed and not acted upon by external nature or by other men as if I were a thing, or an animal, or a slave

86

bruce jennings

incapable of playing a human role, that is, of conceiving goals and policies of my own and realizing them.

One is free in the positive sense when one’s reason, one’s higher self, is in charge of one’s conduct. Negative liberty is the absence of control by others; positive liberty is more like self-control. Berlin is suspicious of positive liberty, which at least once he calls ‘autonomy’, as a manifestation of a dangerous objectivist, universalist conception of the fully human person and the fully human life. In the name of attaining this ideal, individuals have been asked, or required, to subordinate their ordinary freedoms and interests to Causes, with a capital C. Totalitarian ideologies of the mid-twentieth century talk of reforming and improving ‘human nature’. This is the dark side that Berlin sees not only in various romantic and irrationalist bodies of thought, but in the legacy of Kantian rationalism itself. Autonomy and authenticity are dangerous elements in the Enlightenment. Better to stay on the more solid, mundane ground of ordinary interests and negative liberty.

IV While acknowledging this danger in autonomy as positive liberty, I would still maintain that some vital elements of both moral philosophy and bioethics are missing without it, and that autonomy understood as negative liberty alone is too thin. The missing element in negative liberty has mainly to do with moral duties beyond rights, with relationality, a reaching out to establish modes of solidarity, mutual assistance, and care. These values, I argue, are constitutive of an autonomous form of life and human flourishing not simply instrumental to some self-directed project that contingently one happens not to be able to pursue alone. The world of autonomy as negative liberty is a world of absences and omissions. As such, it is a clean, well-lighted place. One of autonomy’s best emblems is the revolutionary-era flag in America that shows a coiled snake and the motto ‘Don’t Tread on Me’. The world of positive liberty is a messier space, filled with shadows. It is a much more human space. In it people do things to and with one another. They cannot get by simply by steering clear. ‘Don’t tread on me’ gives way to ‘Help me up’.

Autonomy and the Success of Bioethics ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

One of the primary things to be observed by a review of the concept of autonomy in bioethics is a kind of conceptual displacement. This, at I noted at the outset, has occurred in bioethics partly owing to (a) the sociology of the field and partly

autonomy

87

to (b) the conditions that have been required of it to gain intellectual legitimacy, and thereby some measure of power and influence, in professional and policy domains. The sociological factor to which I refer is the growing distance in the field between those who practice moral philosophy and political theory in academic settings and the rather insular literature in bioethics that purports to address its theoretical, conceptual, and philosophical underpinnings. William F. May (1980) once quipped that the relationship of the applied ethicist to the moral philosopher was as one who draws and carries water from wells they have not dug. Today that requires amendment. Applied ethicists are digging their own wells and carrying ideas from them to share with professional practitioners in medicine, science, and other fields. However, the new wells are not so deep nor their water always pure. That is perhaps unfair and too harsh. Another way of saying it is that the conditions under which theorizing takes place in bioethics have changed and are different from what they are in philosophy and cognate disciplines. To be in a bioethics centre in a medical school is not the same as being in a philosophy department in a school of arts and sciences. Thirty-five years ago bioethics was made up mainly of individuals who were at home in both places; today few move easily between them. The problem of legitimacy in bioethics arises from the necessity for the field to make a transition from its origins in movements that arose in opposition to mainstream medicine, health care, and biomedical science to an established consulting discipline that has gained access to both clinical and policy settings. These movements of the late 1960s and early 1970s voiced concerns about obstetric practices and women’s health generally, consumer rights, health care access for the elderly, children, and the poor, and advocacy for persons living with disabilities, and for those mistreated in certain types of medical research, such as the syphilis study conducted around Tuskegee, Alabama, by the United States Public Health Service. An important part of gaining this type of respectability and acceptance, while at the same time avoiding a total and abject co-optation, was to build an ethical stance around a concept such as autonomy that effectively straddled the powerful forces of libertarian individualism in the broader political and moral culture while redirecting the use of the expertise and authority of medicine without fundamentally challenging or undermining that authority. The concept of autonomy, understood as it has been in bioethics as negative liberty, has been well suited to this task. The practice of medicine in the service of respect for the rights, dignity, and personhood of the patient replaces medicine in the service of beneficent paternalism. But in this shift medicine retains its power based on scientific expertise, cultural influence, control over access to medical technology and treatment modalities, and considerable (albeit somewhat attenuated of late by the fiscal crises of universal insurance systems or by the commercial interests of managed care systems in the United States) leverage over the allocation of resources, conditions of practice, lobbying power, and the like.

88

bruce jennings

Conclusion ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

The component ideas of autonomy—individual rights, freedom of choice, privacy, independence, freedom from outside interference—are vested with great moral legitimacy. They express deeply felt aspirations and worthy ideals. At the same time, other principles that remind us of our social natures and our interdependence as human beings—community, citizenship, authority, obligation, responsibility, reciprocity, tradition, rules, limits—are also ideas of potent resonance. They are the words of civilization and moral order; they are the elements of an essential vision of the human good. In fact, we perceive them as indispensable, since they are firmly rooted in our psychological natures. These ideas derive from the fact that the human identity is shaped by—and the human good is lived within and through—social roles (such as being a parent), relationships (such as providing care), and institutional structures (such as the family and civic and religious groups). No moral discourse is truly adequate to the richness of human moral experience if it lacks either vocabulary—the vocabulary of autonomy or that of relationality and mutuality. It is critical that we retain what is of enduring value in the language of liberty and autonomy, but it is no less critical that we sustain the language of responsibilities and relationships. If there is a danger inherent in bioethics today—and I am convinced that there is—it comes from an excessive emphasis on autonomy and too little appreciation of human interdependence and mutual responsibility.

References Agich, G. J. (1993), Autonomy and Long Term Care (New York: Oxford University Press). Beauchamp, T., and Childress, J. (1994), Principles of Biomedical Ethics (4th edn. New York: Oxford University Press). Bell, D. (1976), The Cultural Contradictions of Capitalism (New York: Basic Books). Bellah, R. N., Madsen, R., Sullivan, W. M., Swidler, A., and Tipton, S. M. (1985), Habits of the Heart: Individualism and Commitment in American Life (Berkeley: University of California Press). (eds.) (1987), Individualism and Commitment in American Life: Readings on the Themes of Habits of the Heart (New York: Harper and Row). Berlin, I. (1969), ‘Two Concepts of Liberty’, in Berlin, Four Essays on Liberty (New York: Oxford University Press), 118–72. Callahan, D. (1970), Abortion: Law, Choice, and Morality (New York: Macmillan). Charlesworth, M. (1993), Bioethics in a Liberal Society (Cambridge: Cambridge University Press). Compassion in Dying v. Washington (1996), 79 F. 3d 790 (9th Cir.) Elshtain, J. B. (1995), Democracy on Trial (New York: Basic Books). Feinberg, J. (1972), ‘The Idea of a Free Man’, in R. F. Dearden, Education and the Development of Reason (London: Routledge & Kegan Paul).

autonomy

89

(1980), ‘The Nature and Value of Rights’, in J. Feinberg, Rights, Justice, and the Bounds of Liberty (Princeton: Princeton University Press), 143–58. Gaylin, W., and Jennings, B. (2003), The Perversion of Autonomy: Coercion and Constraint in a Liberal Society (2nd edn. Washington, DC: Georgetown University Press). Gillon, R. (1986), Philosophical Medical Ethics (New York: John Wiley). Hauerwas, S. (1974), Vision and Virtue (Notre Dame, Ind.: Fides). Haworth, L. (1986), Autonomy (New Haven: Yale University Press). Himmelfarb, G. (1995), The De-Moralization of Society: From Victorian Virtues to Modern Values (New York: Alfred A. Knopf ). Hughes, R. (1993), Culture of Complaint: The Fraying of America (New York: Oxford University Press). Kant, I. (1963), ‘What Is Enlightenment?’, in L. W. Beck (ed.), On History (Indianapolis: Bobbs-Merrill), 3–10. Kohlberg, L. (1981), Essays on Moral Development, i: The Philosophy of Moral Development (New York: Harper and Row). Lasch, C. (1978), The Culture of Narcissism: American Life in an Age of Diminishing Expectations (New York: W. W. Norton). Mackenzie, C., and Stoljar, N. (eds.) (2000), Relational Autonomy: Feminist Perspectives on Autonomy, Agency, and the Social Self (New York: Oxford University Press). May, W. F. (1980), ‘Professional Ethics: Setting, Terrain, and Teacher’, in D. Callahan and S. Bok (eds.), Ethics Teaching in Higher Education (New York: Plenum Press), 205–41. Mill, J. S. (1956), On Liberty (Indianapolis: Bobbs-Merrill). Olmstead v. United States (1928), U.S. 438, 478. O’Neill, O. (2002), Autonomy and Trust in Bioethics (Cambridge: Cambridge University Press). Peters, R. S. (1972), ‘Freedom and the Development of the Free Man’, in R. F. Dearden (ed.), Education and the Development of Reason (London: Routledge & Kegan Paul). Riesman, D., with Glazer, N., and Denny, R. (1950), The Lonely Crowd: A Study of the Changing American Character (New Haven: Yale University Press). Robertson, J. A. (1994), Children of Choice (Princeton: Princeton University Press). Scanlon, T. M. (1972), ‘A Theory of Freedom of Expression’, Philosophy and Public Affairs, 1: 204–26. Schlesinger, A. M., Jr. (1992), The Disuniting of America (New York: W. W. Norton). Schneewind, J. B. (1998), The Invention of Autonomy: A History of Modern Moral Philosophy (Cambridge: Cambridge University Press). Schneider, C. (1998), The Practice of Autonomy: Patients, Doctors, and Medical Decisions (New York: Oxford University Press). Taylor, C. (1991), The Ethics of Authenticity (Cambridge, Mass.: Harvard University Press). Thomson, J. J. (1971), ‘A Defence of Abortion’, Philosophy and Public Affairs, 1/1 (Fall), 47–66; repr. in M. Cohen, T. Nagel, and T. Scanlon (eds.), The Rights and Wrongs of Abortion (Princeton: Princeton University Press, 1974), 3–22. Wolfe, A. (2001), Moral Freedom: The Search for Virtue in a World of Choice (New York: W. W. Norton). Wolff, R. P. (1970), In Defense of Anarchism (New York: Harper and Row).

chapter 4 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

MENTAL DISORDER, MORAL AGENCY, AND THE SELF •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

jeanette kennett

Introduction ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Let me begin by quoting from Ann Deveson’s account of life with her son Jonathan, who suffered from chronic schizophrenia: At the end of the afternoon I was driving through North Adelaide when I spotted [Jonathan] walking along the middle of the road. He waved at me to stop. He asked me for a milkshake so we went into a cafe, and talked for a few minutes before he began glowering at me, and muttering. The tension of the past few weeks . . . was bound to erupt sometime. It erupted over me. One minute Jonathan was blowing into his chocolate malted. The next he had thrown the milkshake into my face, followed by the pepper and salt, upturned the table, and chucked a chair at me. People gasped, the waiter came running and Jonathan shot off, out the door and up the street. I would like to thank Dean Cocking, Steve Matthews, Franc¸ois Schroeter, Janna Thompson, and audiences at the universities of Melbourne, Edinburgh, and Glasgow, the Australian Catholic University, and the Australian National University for helpful comments on earlier versions of this chapter.

mental disorder, moral agency, the self

91

I shook the milk off me and tried to rub the pepper out of my eyes, which made it worse. The waiter hovered, hoping for an explanation. I couldn’t think of one that wouldn’t take half an hour so I paid the bill and left. Brenda and Margaret [workers] thought I should charge Jonathan with assault. They said I had to set limits. The idea appalled me. But I did feel angry: angry with Jonathan for hurting me, angry with the system for not helping him, angry with the illness. The hardest anger to deal with was the anger with Jonathan, because of its paradox. Can you be angry with someone if it is their illness that makes them so destructive? But I was angry, so angry that I felt like thumping anyone and everyone, so angry that I had to belt my rage out on some cushions, and even then I could not assuage it because I felt so powerless. (Deveson 1998: 82)

Here and at other points in Deveson’s narrative we see a wide variety of responses to Jonathan and the manifestations of his illness, from blaming, excusing, pity, anger, and resentment, to a detached managerial approach, to withdrawal altogether from interaction. Two connected questions are raised here. What happens to the person who is the subject of such attitudes, and what should our moral responses be towards those who suffer serious mental disorders? Or, to put it in terms of a central debate in moral philosophy, how does mental disorder affect the agency, responsibility, and the moral standing of the affected person? In the philosophical debate, the more general question, as I first put it, of what happens to the person whose responsibility is of such theoretical interest is left almost untouched. Yet, a full understanding of the impact of mental illness on agency and responsibility must also examine the impact of mental illness on the selfconception1 of people with mental illness, and on their interpersonal relations, for these are intimately connected with the exercise of their agency and their capacities for autonomy. A person suffering a mental illness or disorder may differ dramatically from his or her previous well self. Family and close friends who knew the person before the onset of illness tend to regard the illness as obscuring their loved one’s true self and see the goal of treatment as the restoration of that self. ‘He is not really like this,’ they will say with increasing desperation. Treatment teams and others, who have no acquaintance with the person when well, respond to what they see in front of them and do sometimes make harmful judgments of character based on the person’s presentation when ill. ‘He knows exactly what he is doing’, ‘He’s just being manipulative’, ‘There’s no excuse for that kind of language’, and so forth. One mother I know took to carrying around a scrapbook filled with photos, letters, school reports, and testimonials recording her son’s academic and sporting prowess, his popularity, and his deep moral concerns prior to the onset of schizophrenia. She kept this record to try to show treating teams what they could not or would not see for themselves, 1

When I talk about self-conception I mean it in the most everyday sense: that is, how we see ourselves, our history, character, personality, relations with others, and so forth, and how we project ourselves into the future. It encompasses both self-esteem and personal narrative.

92

jeanette kennett

to try to ensure that her son was not seen either as a collection of symptoms or as straightforwardly difficult and bad: uncooperative, rude, aggressive, and dirty.2 The identity of those who suffer from mental illness, their character, and their capacities, can thus become contested by those surrounding them. But what happens to the person at the centre of this contest? I think that at least in a very common subset of cases—those where the mental illness is chronic or cyclical with onset during late adolescence—the central project we all have of constructing and maintaining our identity is profoundly undermined. This is in part because of the disunifying effect of such illnesses on one’s agency; it is in part because certain identity-constituting relationships, such as our friendships and other close voluntary reciprocal relationships, may be largely unavailable when in the grip of the illness; and it is in part because those family and patient–client relationships left over provide an impoverished and perhaps even a stultifying environment for self-understanding and self-creation. The person’s self-conception and agency is, over time, affected by his illness, both directly and indirectly, to the point where we may well conclude that there is no univocal answer to the question. ‘Is it him or is it his illness?’ When we ask this question we are partly interested in issues surrounding the person’s capacities for autonomy and responsibility. For example, we might want to know whether it would be appropriate to charge Jonathan with assault and punish him, or conversely to subject him to involuntary treatment. But we are also, and in everyday circumstances, perhaps primarily interested in establishing what kinds of interpersonal dealings and relations are possible with the person. Is he always like this? When can we rely on what Jonathan tells us? Do we have special obligations towards him? Would we be able to take him to the theatre or the football or a family gathering? Or should we keep him at arm’s length? Some philosophers have held that the capacities necessary for responsible agency and the capacities required for decent reciprocal relationships are two sides of the same coin. In what follows, I try to show how they are related but argue that there is a strong moral case for drawing a distinction between our assessment of someone’s present capacity for autonomous choice and action and our preparedness to engage in reciprocal relations with him or her.

Autonomy and Responsibility ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

I would prefer jail or halfway house to nuttery. I will always tell the truth because I have metal in my brain. ( Jonathan, quoted in Deveson 1998: 157) 2 Here and at other places in the text I am grateful to members of the Bridging the Gaps Forum, a parent support and advocacy group in Melbourne, for sharing their experiences and insights with me.

mental disorder, moral agency, the self

93

[Jonathan] said jail was heavy, but hospital was humiliating. (Deveson 1998: 164)

As Jennifer Radden (2002) notes, questions of autonomy have been held to be especially important in psychiatric ethics as well as more broadly in moral philosophy. The assessments we make of each other’s capacities for autonomous choice and action have significant moral and practical implications. For those suffering mental illnesses and disorders it may, at the limit, mark the difference between forced treatment on the one hand, and imprisonment on the other. But well before we reach the criminal justice system we find a polarization in community and professional attitudes to sufferers, each of which risks injustice. Radden (2002: 399) puts the problem this way: This inquiry is a delicate one. Stigmatizing attitudes towards mental disorder have undeniably exaggerated the extent to which the mentally ill are imperfectly rational and autonomous agents, and modern day efforts to extend the presumption of autonomy . . . have gone some way towards undoing the ill-effects of such prejudicial attitudes. Nonetheless, it must be possible to acknowledge and avoid the wrongs associated with the cruel and discriminatory attitudes of the past without resorting to a misapplication of the autonomy model. Unwarrantedly attributing capabilities to the psychiatric patient when such capabilities are at least temporarily compromised involves its own inhumanity and injustice.

The task is indeed a difficult and delicate one, and part of the project of this chapter is to draw out why this is so. In this part of the chapter, I outline the capacities necessary to autonomy and their relation to responsibility as it has been traditionally conceived. I then consider a more fundamental condition of effective agency, unity of agency, which is disrupted in many cases of mental disorder but which is largely ignored in standard accounts of responsibility. I then examine the ways in which the exercise of our agency can be affected in and through our interactions and relations with others, and the disruptions caused by mental disorder to these relations, before returning to questions of responsibility and moral standing.

Autonomous Agency and Reflective Self-Control What capacities do agents need in order to act autonomously and be held responsible for their actions? To what extent are people suffering mental illness impaired in these capacities? At one extreme of the spectrum Thomas Szasz (in The Myth of Mental Illness, 1961/1974, and in many places since) famously urges the view that the odd behaviour of those conventionally regarded as mentally ill is chosen as a way of dealing with ‘problems in living’ and that any form of compulsory treatment is a brutal measure of social control. If such persons commit crimes, they should be punished according to law. Anything else is dishonest and disrespectful.3 He argues 3 Szasz says in a recent piece, ‘I do not deny that there are in our society, poor, homeless, lonely people who talk to themselves. However I do disagree . . . that this phenomenon is evidence that these

94

jeanette kennett

that ‘insane behaviour, no less than sane, is goal directed and motivated’ (Szasz: 113, cited in Moore 1975/1982). Szasz apparently believes that the mere capacity for intentional action is decisive evidence of the autonomy of the actor. This is a very thin account of the conditions of autonomous agency and one that is not shared by many moral philosophers. For, on such an account, young children and even dogs would count as autonomous and so as morally and legally responsible for their actions.4 R. Jay Wallace, in adopting a broadly Kantian account of agency, notes that to be an accountable agent—an agent who is subject to moral requirements, and who may be appropriately held responsible for failing to comply with such requirements—it is not enough merely to be subject to desires and capable of acting to promote the ends set by such desires. Moral agency requires, in addition, the capacity to step back from one’s given desires and to assess the ends they incline one to pursue in the light of moral principles. (Wallace 1994: 13)

Wallace articulates a widely shared view in insisting that free autonomous agents must possess the powers of ‘reflective self-control’.5 What are these? First, the person must possess moral concepts and have the capacity to recognize and apply moral reasons appropriately in their practical deliberation. Second, they must have the capacity to control their behaviour in accordance with these reasons. These capacities are not all or nothing. They come in degrees, and so responsibility, too, comes in degrees. While most persons with psychiatric illnesses or disorders will have acquired ordinary moral concepts (psychopaths are arguably the exception), disordered thinking, delusions, hallucinations, clinically depressed or elevated mood, and so forth may reduce or remove their capacity to recognize, or to weigh appropriately, relevant moral considerations and to judge accordingly.6 Further, persons suffer from a mental-brain disease . . . and that ostensibly compassionate–therapeutic attitude towards such individuals justifies our depriving them of their liberty (when they are innocent of crime), and of excusing them of responsibility for crime (when they are guilty of it)’ (Szasz 1998: 205). I take it that by ‘guilt’ here he means only that they performed the action proscribed by the law. 4 Unfortunately the incidence of mental illness and other psychiatric disorders in our prison populations suggests that the view expressed by Szasz—that mentally disordered persons are rightly held responsible for their offences—is not uncommon. A recent New York Times report puts the figure at about 20 per cent of the prison population (‘Study Finds Hundreds of Thousands of Inmates Mentally Ill’, 22 Oct. 2003). This is a time when it is increasingly difficult to mount a successful insanity defense in the United States even for defendants who are clearly severely impaired. Among philosophers, though, there is fairly consistent support for the idea that people with mental illnesses are not, or not fully, responsible for those actions symptomatic of their illness. Discussion here has tended to focus on the case of schizophrenia. See Christian Perring (2002). 5 The idea that autonomous agency centrally involves the capacity to stand back from and evaluate one’s first-order desires is widely shared by philosophers of both a Humean and a Kantian cast. See e.g. Bratman (2000); Kennett (2001); Korsgaard (1996); Watson (1982); Frankfurt (1971); Velleman (1997). 6 For an argument that mentally ill defendants usually do know both what they are doing and that it is wrong, see Schopp (1991). Schopp argues that, nevertheless, such individuals may be substantially

mental disorder, moral agency, the self

95

these conditions may impair the ill person’s capacity to exercise control over their actions in accordance with the moral considerations they do accept. Deveson writes movingly of finding desperate notes written by Jonathan to himself, perhaps as reminders or injunctions: ‘Don’t hurt Anne’. But such attempts to control his future actions were always vulnerable to defeat by his symptoms. On standard accounts Jonathan cannot count as an autonomous, responsible agent. He, at least, ought not be blamed for many of his actions. But even where we can make such a clear determination on responsibility for particular harmful actions, it does not determine in full what our attitude should be towards Jonathan or how he should be treated. For that we need to acquire a more substantial understanding of the losses the illness imposes on Jonathan himself.

Unified Agency and Self-Authorship David Velleman has remarked that the philosophy of autonomy and the philosophy of personal identity seem to intersect at something called ‘the self’.7 My discussion here is located at this intersection: at the interrelation between effective agency and self-conception. Christine Korsgaard (1988: 101), in providing a Kantian account of personal identity, argues that a person is ‘both active and passive, both an agent and a subject of experiences’. She argues that effective agency requires unity both at a time, in order that we may eliminate conflict among our motives and do one thing rather than another, and over time, since many of the things we do form part of longer-term projects and make sense in the light of these projects and life plans. And she says that in pursuing these long-term projects and commitments ‘we both presuppose and construct a continuity of identity and of agency’ (1988: 113; my emphasis). What does this mean? We do not simply act, as Michael Bratman (2000: 40) points out, from moment to moment. We conceive of ourselves as agents who persist over time, and so we commit ourselves to future directed plans, which we intend should structure and coordinate our more particular decisions and activities. But we can only do so if we can be confident that our future directed decisions will be effective, that we are able to ‘determine today what gets done by us tomorrow’ (Velleman 1997: 45). If we don’t have this ability we lack autonomy here and now. For, as Korsgaard says: ‘to the extent that you regulate your choices by identifying yourself as the one who is implementing something like a particular plan of life, you need to identify with your future in order be what you are even now’ (1988: 113–14). Velleman argues that we achieve diachronic autonomy because the plans we adopt now provide reasons for us into the future, reasons we will then buy into impaired in their capacities of rational choice. They may have, for example, delusional beliefs of such force that they cannot reason properly about their actions. For further discussion of issues surrounding criminal responsibility and mental impairment, see Kennett and Matthews (2004). 7 Formerly on his website.

96

jeanette kennett

but might not otherwise have had. We identify these plans and commitments as our own and implementing them thus usually makes the best sense of our life stories (Velleman 1997: 45–8; 2000). Hence it would seem that autonomous agency depends upon significantly unified agency. Unity is fundamental to the capacity to make, shape, and revise plans and to carry them out: to engagement in those long-term projects and commitments, such as an education, a career, a marriage, or a friendship, which are the sources of our deepest satisfactions. Unity is an essential condition of both the capacity for competent deliberation and the capacity for self-control over time that standard accounts of moral agency and responsibility insist upon. In short, it underwrites our authorship of our lives. While not all aspects of the self, or even, I would argue, all valuable aspects of the self, are subject to agential control, agency is surely central to the self and is a condition of access to a wide variety of important normative goods. Of course unity of one’s agency does not, on its own, ensure a good life. Successful self-authorship relies on the possession of a variety of fairly sophisticated skills that fall under the rubric of either competent deliberation or self-control. There is the rational requirement of getting one’s projects to cohere, there are means–end managerial skills; there are value judgments about the quality of different projects; there are prioritization skills; and importantly, predictive imagination—‘If I take that job, then I’ll turn into that kind of person, can I imagine myself as being like that, is that the way I want to be . . . ?’, and so on. A view of the self as author allows for rational revision of plans in the light of changed circumstances or new information, and so allows for personal development and change over time. Indeed we may have projects that quite explicitly aim at changing our very characters, and so at changing the way we may come to view ourselves, and our projects, in the future. One of my projects may be to write a cookbook, another may be to learn to service my car. Yet another may be to develop a greater sensitivity to the suffering of others and a less judgmental attitude to their misfortunes. If I am successful in this latter character-building project, I may come to see new reasons for action in the future; these lead me to take on new projects, and so by incremental steps I become very different from how I am now. Nevertheless, I retain authorship of the overall process and identify with each stage of it. But now consider what happens when we undergo dramatic changes—not in our external circumstances, but in ourselves—which we have not chosen. What if we find we can no longer rely, in the normal way, on our future selves being appropriately affected by our present plans and decisions? What if our projects, large and small, were at constant risk of derailment? What if even those traits and dispositions which we might think of as fundamental to our characters—openness, generosity, efficiency, and so forth—were in danger of being masked or reversed? Then it would seem that we would lack authorial control over our lives, and our identity, our self-conception, as well as our autonomy would be gravely

mental disorder, moral agency, the self

97

undermined. In cases of serious, recurrent, or chronic mental illness I suggest that this is what happens. The onset of illness is like a grenade or a series of grenades exploding in one’s path. The impact of the repeated failure of one’s plans and one’s hopes for the future must damage one’s self-conception and autonomy in ways that extend beyond the more obvious and direct disabling of the capacities for rational reflection and self-control. As Korsgaard says, we need to see ourselves as agents of our future, we need to identify with our future, to be what we are even now. If mental illness changes a person’s view of their future, and of their ability to shape their future, it changes them now. Agnetti and Young (1993: 69) argue that, for patients (and their families) with a long-standing psychiatric disorder: life seems like it has always been dominated by the illness and always will be. The past is seen as the same as the present and the future offers nothing new. . . . If the temporal dimension is lost in one’s view of the world, then the possibility of hope, growth and change is also lost . . . events stand still and individual action is paralysed.8

Repeated experiences of illness, of hospital, of (perhaps coercive) treatment, of failed plans, of other ill people, and of the reactions of others to one’s illness, must all over time profoundly affect the way the agent sees themselves and influence the ways in which they can project themselves into the future. This is vividly illustrated in Jay Watts and Stephen Priebe’s (2002: 447) study of users of assertive community treatment services: Most of the participants described the psychiatric system and the labelling of diagnosis as an attack on their identity. For example: ‘I get sick, I get cracked up, and this is my life. I can’t keep fighting against it and trying to be someone else . . . I had people that could depend on me. I liked independence . . . To lead, and I am not in a position to lead no more.’

Though people with chronic mental illnesses still wish for the same goods as everyone else, and may hold much the same values as they did prior to illness, a loss of confidence in their power to see through valued plans and projects may mean that they are given up, and the person lives an impoverished life dominated by short-term considerations. They may come to see themselves quite generally as subjects, rather than agents, of their experiences. There is much more to be said, and much more precisely, than there is room for here, about how the various symptoms of the various illnesses might affect or impair one’s authorial capacities as well as about the possibilities of regaining authorship of one’s life in a way that integrates and organizes one’s experience of illness. I leave further exploration of that to the reader.9 Here I want to focus on one 8

There is also evidence that treating teams are vulnerable to this paralyzing sense of timelessness when dealing with patients or clients with chronic psychiatric conditions. For discussion of this and of ways in which treating teams can ‘restart the clock’ for their clients, see Young (1994). 9 But see Wells (2003); Kennett and Matthews (2003); Phillips (2003); Radden (2003); Velleman (2005).

98

jeanette kennett

particularly central way in which one’s sense of self, and opportunities for effective agency, may be affected when one suffers a mental illness, which has especially strong moral implications for those with social, professional, and institutional contact with persons suffering mental disorders or illnesses.

Reciprocal Interaction and the Visibility of the Self ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Successful negotiation of the social environment is a condition of success for almost all of our activities and projects, from going shopping to holding down a job. Further, our most central and long-term commitments may be to loving reciprocal relationships. Talking in terms of an individual person’s projects and capacities, as I have until now, may suggest that to the extent to which persons have authorship at all, we are sole authors of our identity and life story. But of course, as social beings, as agents whose actions impact on others and who regularly need the cooperation of others in order to complete action, we are not. Other people contribute very significantly to how we see ourselves, and so to the selves that we become. How do these social interactions and relationships proceed? How do they make their contribution to the self-building and maintenance project? And how does this enhance or impair our agency and autonomy? Nathaniel Branden (1993: 67), in an essay entitled ‘Love and Psychological Visibility’, argues that it is to a large degree through our interactions with others that we become visible to ourselves. He begins by analyzing the pleasure he gets from playing with his dog: The key to understanding my pleasurable reaction . . . was in the self-awareness that came from the nature of the feedback she was providing. From the moment I began to ‘box’ she responded in a playful manner . . . Were I to push or jab at an inanimate object, it would react in a purely mechanical way; it would not be responding to me; there would be no possibility of its grasping the meaning of my actions, of apprehending my intentions, and of guiding its behaviour accordingly . . . The effect of Mutnik’s behaviour was to make me feel seen, to make me feel psychologically visible . . . And as part of the same process, I was experiencing a greater degree of visibility to myself ; I was making contact with a playfulness . . . which . . . I generally kept severely contained, so the interaction also contained elements of self-discovery . . .

Branden argues that this experience of visibility to oneself and others is a product of all positive human interaction that reaches its pinnacle in romantic love. He further claims that to do the job of providing visibility, the interaction must be both positive and appropriately responsive to the other and to what is conveyed by the other. Let’s look at some of those interactions.

mental disorder, moral agency, the self

99

Reciprocity, Interpretation, and the Reactive Attitudes ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

As social beings we must, necessarily, engage in a wide variety of transactions and interactions with others. Social life is shot through with interpretive practices. Grasping each other’s meaning, recognizing each other’s intentions, guiding our actions accordingly to coordinate with each other, these all involve a degree of interpretation of the other. The interpretation is largely implicit. For the most part we don’t think about the particular messages we give to others: we are skilled at mind reading, at picking up behavioural cues and responding to them appropriately, and many social situations and transactions are governed by rules, roles, or conventions that we have internalized. All social transactions, from getting a haircut, to asking for directions, to having dinner with friends, require a degree of mutual responsiveness and interpretation of the other for success. In these transactions with other our default stance, our natural stance, is what Peter Strawson (1982: 62) has called the participant stance. Within this stance we are exposed to the reactive attitudes of, for example, ‘gratitude, resentment, forgiveness, love, and hurt feelings’. My attitude to you will depend crucially on my picture of your motivations towards me. As Strawson (1982: 63) remarks: ‘If someone treads on my hand accidentally, while trying to help me, the pain may be no less acute than if he treads on it in contemptuous disregard of my existence . . . but I shall generally feel in the second case a kind and degree of resentment that I shall not feel in the first.’ Strawson claims that just in being prone to these kinds of reactive attitudes we manifest respect for each other. In taking up the participant stance, he says, we signal that we regard those with whom we engage as fit for ordinary adult human relationships. So successful social interactions should provide, at least, a basic visibility and affirmation of the self as social participant with rightful expectations of mutual goodwill. Often enough they do more. We might receive more specific feedback on how we’re doing and on how we’re seen, for example a compliment on our good service, thanks for some basic assistance like helping someone pick up the contents of a spilled trolley, or we might share a moment’s camaraderie with a stranger at a football match. Such interactions involve positive reactive attitudes and provide positive reinforcement to the self. They allow us to proceed with confidence. But now consider those of the negative variety. Whether we stand to the other as friend, lover, colleague, customer, or patient, we expect and indeed demand that their actions where we are concerned reflect good rather than ill will. That we make this demand, and that the attitudes of others matter to us, is evident in our reactions when our ordinary expectations are disappointed. The rudeness of a shop assistant, the brusque dismissiveness of a doctor, each occasions hurt feelings and a degree of resentment that may bear little relation to any actual disadvantage we suffer. The shop assistant rolls her

100

jeanette kennett

eyes at my clumsiness with my purse, ignores my attempts at pleasantries, and displays obvious impatience in answering my queries, but she wraps my purchases efficiently and gives the correct change. The injury here is in the attitude not the action. Attitudes of contempt or indifference are in themselves injurious; attitudes of goodwill and esteem, in themselves, confer a benefit on the recipient. But what is the nature of the injury? I suggest that it is, at least in part, an injury to our conception and presentation of ourselves. When we lose control of how we appear to others, when our self-presentation misfires, we will typically experience some anxiety about who we are. I am trying to be pleasant and cooperative but am coming across as fussy and incompetent. Maybe I am fussy and incompetent. Or maybe she is just rude. In the case of one-off interactions such as with the shop assistant where there is a failure of reciprocity and interpretive charity there is no chance to negotiate or revise the interpretation as there may be in longer-term relationships. I am stuck with the dissonance between the way I thought I was and the picture I gained in seeing myself through her eyes. (Of course I may come to believe that her interpretation of me is correct and I will then modify my view of myself and maybe make an effort to change in order to avoid the negative feedback.) In cases of the worst kind I cannot see myself at all through the other’s eyes. I am simply invisible to them. This is frequently the experience of members of minority or oppressed groups in interactions with members of dominant groups. In these cases the default assumption of mutual goodwill and respect doesn’t even get off the ground. The effect of such invisibility and exclusion from social interaction on the agent’s sense of efficacy, and on their capacity to act, is particularly severe.

Failed Reciprocity and Fragile Identity ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

I started hearing voices, I started talking to myself, she couldn’t take it any more, the relationship broke up . . . and I ended up on my own. (quoted in Watts and Priebe 2002: 445)

In cases where the agent suffers a mental illness, the kinds of social misfirings I’ve described above happen more often. Though most people with a mental illness manage well enough with most public social interactions, going to the corner shop, catching public transport, etc., the features of their illnesses can often create the conditions for things to go wrong. Depressed or anxious persons may seek to make themselves invisible, others go the opposite way; their lack of judgment or social disinhibition leads them to press too much personal information on strangers, resulting in withdrawal. Other oddities in behaviour and presentation, for example, talking too loudly or standing too close, can also lead to avoidance. Sufferers may

mental disorder, moral agency, the self

101

have difficulty seeing themselves through others’ eyes and so may not make the appropriate modifications to their behaviour: alternatively they may be too ready to adopt a negative self-evaluation on the basis of others’ reactions to them. Thus these experiences will fail to provide visibility of the right kind and may feed back in a negative way to the agents’ conception of themselves. Of course these kinds of public social interactions are not of a deep, extended, or intimate nature and so they would not normally provide, contribute to, or challenge a deep view of the self. Our close and intimate relationships do, on the other hand, contribute very significantly to the way we see ourselves and to the selves that we become. The more stable, positive, and sensitive interpretations offered in the context of our intimate relationships can counter and resolve dissonant experiences in the public social sphere. The reassurance and affirmation we receive there strengthen and protect the self. Unfortunately these relations too are often weakened or unavailable to the agent suffering a mental disorder, since many such people are impaired precisely in the responsiveness and attentiveness to others needed to form and maintain them. Since Aristotle it has been accepted that friendship is a central human good and an important source of self-knowledge. In addition it provides a means to self-growth and development. Let’s examine how. In close friendships we engage in the broad interpretive practices outlined above of grasping each other’s meaning, recognizing intentions, coordinating actions. But, ideally at least, these will be more finely tuned and attentive in friendship. One’s close friend will be better able than acquaintances or strangers to read one’s mood, and friends, as I’ve claimed elsewhere, are especially receptive to responding to each other’s wishes and coordinating their activities in accordance with each other’s suggestions.10 So I might notice, where others would not, a slight stiffening in my friend’s demeanour when she is cornered by someone she dislikes but doesn’t wish to be rude to at a social gathering, and will move to rescue her from the situation. If she then says ‘Let’s get out of here’, I will take that as sufficient reason to leave. My responses to her are highly particularized and thus provide a high degree of visibility and affirmation of self. Beyond this particularized implicit interpretation of her I also, as her close friend, offer explicit interpretations that contribute in a deeper way to her understanding of who she is. I may, for example, point out, after rescuing her, that she quite generally exhibits discomfort with men of a certain status and that perhaps this is colouring her view of the individual who cornered her. I may say that he’s not so bad, that she may have misread his intentions, or I may affirm her own assessment of the situation. Assuming that what I say is, as Branden insists, appropriately responsive, she is likely to take it on board and this will help shape and develop 10 The account of friendship given here draws upon that developed in Dean Cocking and Jeanette Kennett (1998, 2000). See also Dean Cocking and Steve Matthews (2000). The example below originated in a talk given by Dean Cocking, ‘Intimacy and Privacy Online’, at the Australian National University in 2002.

102

jeanette kennett

her view, not just of the situation, but also of herself. Mutual deep interpretation is a central and defining activity of close friendship without which much of its intimacy would be lost. Within friendship we receive and accept input and feedback that importantly contributes to our self-building project, and affirmation, which is vital to its maintenance. This, what we might call, drawing of each other, further enhances the relationship between us: there is a positive and dynamic feedback loop. Such interpretations may not be taken on board at all, and certainly not in the same way, if made in the context of a different relationship, say by one’s mother or even by one’s therapist. The point is that in close friendship we are especially receptive to such interpretations and so friendship provides an important arena for self-development and the acquisition of self-knowledge that may not be replaceable. Aristotle (MM 1213a 10–26) says: ‘If, then, it is pleasant to know oneself, and it is not possible to know this without having someone else for a friend, the self-sufficing man will require friendship in order to know himself.’ There is here, as John Cooper (1980) says, recognition ‘of the social bases of a secure self-concept and of the role intimacy plays in providing the means to it’. This supposed necessary base of a secure self-concept is placed at risk when one party develops a chronic mental illness. The nature of the difficulty that, for example, delusional individuals must face in establishing and maintaining close friendships is obvious when we look at the role reciprocal interpretation plays in friendship. Delusional individuals often have beliefs about themselves and others that are bizarre and resistant to revision. A lack of response or a highly inappropriate response to the interpretations offered by another will constitute a barrier to friendship. Conversely the ill person is often not well placed to offer the kinds of interpretations of others that the other can integrate into their self-conception. So an important mutual activity of friendship that contributes to the intimacy between friends, as well as to self-knowledge and self-development, may be less available to those suffering delusions. More generally the mere presence and expression of the delusion can disturb the friendship since this may lead to wariness on the part of the friend with respect even to non-delusional communication. Other psychiatric conditions will also affect one’s fitness here. For example, deep depression, paranoia, or wild mood swings may so bias interpretations that they are unacceptable to the recipient. Even the apparently less demanding broad interpretive activities of grasping the other’s meaning and anticipating their intentions, forming joint intentions, and coordinating activities can be more difficult for people with serious mental illnesses. The presence of delusions or obsessions, for example, may dictate activities or routines that cannot be altered and which others cannot understand or share; or the illness may affect the agent’s motivation and planning capacities such that they cannot reliably take part in joint activities, even in the absence of positive symptoms. Poor or absent motivation is well recognized as one of the negative

mental disorder, moral agency, the self

103

symptoms of schizophrenia. Further, the loss of confidence in one’s future that comes from repeated interruption of one’s projects does, as we have seen, affect how the agent thinks of themselves in the present and so affects the way they proceed in relationships as in other matters. The loss of friendships and other close relationships leads to social isolation and contributes to the social invisibility of the ill person, and it is well known that such isolation has a significant detrimental effect. As Watts and Priebe (2002: 446) point out, the literature on chronic illnesses shows that ‘the overwhelming implications of illness on identity is mitigated by intervention from significant others’. But this mitigation was largely unavailable to the psychiatric patients in their study, many of whom had lost significant others ‘early in their psychiatric career’. Without the stabilizing influences that good relations with others provide, psychological function deteriorates further.11 A vicious loop can develop as the feedback from others becomes increasingly impoverished, negative, and defensive, or somewhat artificial and paternalistic, as in many professional encounters.12 It is hard to construct a rich positive sense of self as we become increasingly invisible to others, as they refuse eye contact, turn away at our approach, end conversations abruptly, and show less and less interest in what we say and do. The argument to this point suggests that the consequences of impaired social interaction and the loss of deep interaction on self-concept and the development and exercise of one’s agential capacities can be severe. What it does not support is the idea of a neat division between the person and the symptoms of their illness that we can use in allocations of responsibility, since the illness can affect the person via its effects on their plans, projects, expectations, and relationships even during periods of quiescence. The person they would have become if they had not become ill is an increasingly distant fiction, and the self they are now is constructed out of experiences of illness and of remission, and of their own responses to these experiences and others’ responses to them. It is not identical either to a set of symptoms or to the self prior to illness. If the illness now forms part of the person and has shaped their identity, we cannot dismiss it or glide over it in our moral evaluation of the person. Our responses to the person must also be responses to their illness. However, if we accept the claim that in our relations with others we can become, invited, welcomed, or not, part authors of that person’s self-conception and life story, we need carefully to consider the responsibility that is thus placed on us, particularly in our dealings 11 Indeed impaired adolescent friendship patterns are a significant predictor of a sustained delusional outcome in persons with schizoaffective disorder (Coryell, Keller, et al. 1990). 12 Watts and Priebe outline the therapeutic experiences of their subjects as follows: ‘the approach focused on medication . . . Relationships were seen as impersonal and paternalistic because of short appointment times and frequent turnover of staff ’ (2002: 448). They quote one subject, Mr B: ‘When you are on the ward, you get staff that think it’s a prison or something. And that you’ve done something and they try and boss you around. Control you, control you’ (2002: 446).

104

jeanette kennett

with those whose identity and agency is already fragile. Given the undermining effects of social failure and withdrawal by others on the self, it may seem obvious that in both professional and personal contexts we should not withdraw, as people often do, from those suffering a mental illness. Simple beneficence dictates that we should actively seek to maintain engagement with the other even where their capacity to reciprocate appropriately is impaired. But there is a problem with this simple conclusion. For mental illness is widely (and properly) seen as exculpatory, and in several influential writings it is argued that reciprocal relations and the presumption of responsibility are not separable.

Failed Reciprocity and the Objective Stance ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

The consequences of the mentally ill agent’s impaired capacity for interpersonal relations and social interaction have deep and troubling implications not only for the development and maintenance of the self and the exercise of agency, but also for the ill person’s moral standing in the community. As we have seen, to the extent that we find someone’s actions odd or unintelligible and unmodified by our reactions to them, to the extent to which they seem incapable of understanding or taking an interest in our interests, or of responding to us as we are, we may decide that our interaction with them is insufficiently rewarding and withdraw from it. Peter Strawson (1982: 66) takes a position at the opposite end of the spectrum to Szasz in suggesting that in the case of the severely mentally ill, the participant reactive attitudes are altogether out of place. When faced with such individuals we move, and should move, to the objective stance. Here is his description of this stance:13 To adopt the objective attitude to another human being is to see him, perhaps, as an object of social policy; as a subject for what, in a wide range of sense, might be called treatment; as something certainly to be taken into account, perhaps precautionary account of; to be managed, handled, cured or trained; perhaps simply to be avoided. . . . The objective attitude may be emotionally toned in many ways, but not in all ways: it may include repulsion or fear, it may include pity or love, though not all kinds of love. But it cannot include the range of reactive feelings and attitudes which belong to involvement or participation with others in inter-personal human relationships; it cannot include resentment, gratitude, forgiveness, anger, or the sort of love which adults can sometimes be said to feel reciprocally for each other. If your attitude towards someone is wholly objective, then though you may fight with 13 The following passage may present Strawson’s view as rather more black and white than it is. Strawson does acknowledge that we may move between the two stances particularly in the case of children. I argue below that there is a distinction between a fitting objectivity of judgment that may inform aspects of our relations with others, including professional relations, and the objective stance as described by Strawson.

mental disorder, moral agency, the self

105

him, you cannot quarrel with him, and though you may talk to him, even negotiate with him, you cannot reason with him. You can at most pretend to quarrel, or to reason, with him. Seeing some one then, as warped or deranged or compulsive in behaviour . . . tends, at least to some extent, to set him apart from normal participant reactive attitudes on the part of one who sees him, tends to promote, at least in the civilized, objective attitudes.

Though the objective attitude, so described, sounds anything but civilized, we may find ourselves moving towards it, in both personal and professional contexts, for the reasons I have given. Disappointed reciprocity and failures of interpretation can lead to feelings of frustration and impotence. Then the objective stance may indeed act as a refuge from the strains of involvement. But further argument is needed to support Strawson’s implied suggestion that we should, normatively speaking, adopt the objective stance when dealing with people suffering severe mental illness. If reciprocal relations are, as Strawson suggests, the ordinary basis even for respect and goodwill, then we would expect to find, what we largely do find, that mentally ill persons are treated with less respect and less goodwill than other adult members of the community including by some members of the helping professions. It does not seem right or desirable that one’s moral entitlements should depend so heavily upon one’s social competence. So what can be said in favour of Strawson’s position? For Strawson, resentment is arguably the most central of the reactive attitudes, so I will focus on that. Let’s consider the kinds of consideration that allowably modify or remove resentment and the disposition to blame in circumstances where one would normally expect it to be operative; that is, in circumstances where one has suffered some injury or offence at the hands of the other. Strawson suggests that the considerations fall into two kinds, which, following Wallace (1994), I shall call excuses and exemptions. Excuses do not call into question those capacities of the agent that seem essential for engagement with others and for making and responding to the demand for goodwill and respect. They do not invite us ‘to view the agent as one in respect of whom these attitudes are in any way inappropriate’ (Strawson 1982: 64). Excuses draw our attention to the fact that, for example, the action was performed inadvertently or on the basis of misinformation, or that it was coerced or accidental. ‘I couldn’t help it’, ‘I didn’t know’, ‘I was pushed’, ‘The accelerator jammed’. Resentment is inappropriate here, since the agent’s actions do not signal any lack of goodwill, and the suspension of interpersonal attitudes is also inappropriate, since there is no indication that the agent is unfit for social life. The agent is seen as in general accountable for her actions, but the presence of an excusing condition indicates that the agent did not, on this occasion, do anything deserving of blame. Exemptions, on the other hand, do invite the suspension of some or all of the reactive attitudes. It is admitted that the agent did do something wrong, but it is suggested that this agent, in particular, is not accountable for it on this occasion. Strawson believes that there are two subgroups to the exemptions. The first points

106

jeanette kennett

to a temporary lack of accountability—e.g. ‘He’s been under a very great strain lately.’ The second suggests a more thoroughgoing lack of accountability. Strawson gives the following examples: ‘He’s only a child’, ‘He’s a hopeless schizophrenic’, ‘His mind has been systematically perverted’, ‘That’s purely compulsive behaviour on his part’. These pleas, Strawson claims, do invite the suspension of our ordinary reactive attitudes. The second subgroup, in particular, invites us to see the agent as ‘psychologically abnormal—or as morally underdeveloped’, as ‘incapacitated for normal human relationships’. He argues that when we see someone in such a light, ‘all our reactive attitudes tend to be profoundly modified’ (1982: 66). It is apparent that Strawson places mental illness in the class of type-2 exemptions rather than excuses and sees the presence of serious mental illness as a reason to suspend or ‘profoundly modify’ our reactive attitudes, rather than to see them as undeserved in just this instance. He argues that ‘the participant attitude, and the personal reactive attitudes in general, tend to give place and, it is judged by the civilized, should give place, to objective attitudes, just in so far as the agent is seen as excluded from ordinary human relationships by deep-rooted psychological abnormality’ (1982: 67). Strawson believes that holding people responsible for what they do, praising or blaming them, or even excusing them on some occasion for their actions, is something we can only do in so far as we, actually or vicariously, occupy the participant stance with regard to them. Gary Watson (1987: 257) points out that, for Strawson, the reactive attitudes ‘are constitutive of moral responsibility; to regard oneself or another as responsible just is the proneness to react to them in these kinds of ways under certain conditions’. We find a similar position put forward by Christine Korsgaard (1992: 305) in her paper ‘Creating the Kingdom of Ends: Reciprocity and Responsibility in Personal Relations’. There she argues that to hold someone responsible ‘is to adopt an attitude towards him rather than to have a belief about him or about the conditions under which he acts’. It is, effectively, to enter a certain kind of relationship with the other: a relationship of reciprocity. Reciprocity, she claims, involves acting in concert with another, sharing ends and sharing reasons with each other. You will only enter into such a relationship on the expectation that the other will deal with reasons rationally as you would. ‘In this sense’ she argues, ‘reciprocity requires that you hold the other responsible’ (1992: 311). She says: If my friend fails me in a serious way, and I do not blame her, shrugging it off as I would the misdemeanours of a child or a pet, then I was not holding her responsible after all and probably I was holding myself back. . . . Blame is important . . . as an expression of the tenacity of disappointed respect. At its best it declares to its object a greater faith than she has in herself. Yet it is still not central. The willingness to take a chance on some form of reciprocity is the essence of holding someone responsible. . . . In everyday personal interaction, we cannot get on without the concept of responsibility. (1992: 311–12)

mental disorder, moral agency, the self

107

Both Strawson and Korsgaard see fitness for reciprocal relations, rather than the possession of the capacities mentioned earlier, for deliberative judgment and self-control, as primary to an account of responsibility. It appears that for them there is a conceptual connection between the reactive attitudes, to which reciprocal relations expose us, and holding people responsible for what they do. And they do not detach this from according the other moral standing and respect. So there is a two-step argument in defense of Strawson’s seemingly unpalatable suggestion. The first step is to point out the person’s impaired capacity for reciprocal relations, and to say that in such cases the participant stance is, in truth, unavailable. We delude ourselves if we think it is; the reactive attitudes simply misfire. Deveson’s anger, understandable though it is, lacks a proper object when directed at ill Jonathan. She might as well be angry with her car for failing to start. She might as well try to set limits for her car as for Jonathan, for all the good it would do in getting either of them to shape up. Second, if we agree with Strawson and Korsgaard that it is actual or vicarious engagement in or fitness for reciprocal relationships that grounds attributions of responsibility, the reactive attitudes of resentment and so forth to which engagement exposes us are simply unfair to the mentally ill. To the extent that they cannot share our reasons and ends and coordinate their activities with ours, they don’t deserve to be the subject of such harmful attitudes as resentment. Such attitudes may also get in the way of our helping them and managing their behaviour. We can do this much better if we can remain calm and detached in the face of odd or ‘challenging’ behaviour. However, if we move all the way to the objective stance, if we treat persons as things ‘to be managed, handled, cured or trained’, we make them socially invisible and so, as I have argued, to a significant extent, invisible to themselves. The person is visible merely as a threat, a nuisance, an embarrassment. Their already fragile agency is further frustrated and undermined by the lack of recognition, support, and positive feedback. It looks as though whichever stance we adopt will be damaging. People with psychiatric conditions too often walk a thin line between being the recipients of undeserved blame and resentment on the one hand, and exclusion from the participant stance and so from social life on the other.

Reciprocity and Responsibility ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

We have arrived back at the dilemma outlined by Radden at the beginning of this chapter. It is a very real one and desperately hard to resolve in practice, as Deveson’s account of life with Jonathan makes clear. However, I think Strawson’s

108

jeanette kennett

solution rests on a mistake. First, it is not clear that there is the strong conceptual connection claimed by Strawson and Korsgaard between occupying the participant stance with respect to another and attributions of responsible agency. So it is not clear that adopting the participant stance will be unfair to the mentally ill. Second, the participant stance is elastic; it encompasses a wide variety of reciprocal interactions that are commonly adapted to take account of the particular capacities and characteristics of those with whom we engage. It is unlikely that in any but the most extreme cases the participant stance would be wholly unavailable to us. It seems clear in both Strawson’s and Korsgaard’s accounts that judgments of responsibility made from within the participant stance must presuppose the existence and importance of just those capacities for deliberation, planning, and self-control over time that I earlier described as belonging to autonomous agency. Indeed it is precisely the presence, impairment, or absence of these capacities that Strawson implicitly, and sometimes explicitly, draws upon in distinguishing the varieties of excusing conditions from each other and from those that exempt one from blame and modify or remove resentment. Resentment serves, often, to alert us to the violation of an other-regarding reason. But we may only fairly hold people responsible for such violations if they have both the capacity to recognize the reason and the capacity to conform their conduct to it. Though those capacities are for the most part developed and supported through our relations with others, they are nonetheless conceptually distinct from these relationships. Further, we can make judgments about those capacities and about the degree of responsibility to be borne by the agent without entering, actually or vicariously, into a relationship with them and exposing ourselves to the participant reactive attitudes. According to what Korsgaard calls the theoretical conception of responsibility, which is the conception familiar in philosophical and legal discussions, a person is responsible for her actions just in case she possesses certain characteristics or capacities essential for responsible agency and is in a condition to exercise those capacities in the circumstances (or to determine the condition she is in). It will be a matter of fact, though perhaps difficult to determine, whether or not the person has those capacities. Korsgaard (1992: 313) believes that this is not the model we adopt in our practical interactions with others. On the practical construal that she favours, she argues that holding someone responsible ‘is adopting an attitude towards her, or, much better, placing yourself in a relationship with her’. On this conception, facts about the agent’s condition at the time of action guide decisions about responsibility but do not determine them. She says: I do not believe there is a stable relationship between the voluntariness of an action or attitude and the appropriateness of holding someone responsible for it. If a bad action is found to have been involuntary in some straightforward sense we will withdraw blame; we may also do this if a person is under severe emotional stress. But there is neither need nor reason to reduce the second kind of excusing condition to the first and say that people under

mental disorder, moral agency, the self

109

severe emotional stress cannot control themselves. We do not need to understand a form of debilitation as a form of impossibility in order to make allowances for it; we only need to know what it is like. (1992: 313)

Korsgaard is right, I think, to point out that judgments of an agent’s responsibility made within relationships—from within the participant stance—are not, and should not be, strictly forensic. But though the facts about the agent’s capacities may need to be put in their place, they still have a central place. Given that even judgments of responsibility made from within the participant stance are highly (and properly) dependent on our assessment of the person’s capacities for reflective self-control, it seems that perhaps Korsgaard, at least, is claiming, not that responsible agency and reciprocal relations are conceptually inseparable, but that the capacities that constitute us as responsible agents are prerequisites for decent reciprocal relations. She says: ‘in everyday personal interaction we cannot get on without the concept of responsibility’. Is this right? Korsgaard (and Strawson too, to some extent) presents a model for relations between persons, one requiring responsive and responsible agency. The model is one of perfect reciprocity, described in Kantian terms as a sharing of reasons and ends.14 Now, the relationship Korsgaard uses to illustrate this model is close friendship. But close friendship is not an appropriate model to govern all our relations with others, requiring, as we have seen, a very high degree of equality and responsiveness across domains. This sets the bar for reciprocity too high. While one might need to qualify as an autonomous agent to be a fit subject for the reactive attitude of resentment, it is not at all obvious that full autonomy is needed to fit one for the many other participant reactive attitudes, or must be present to ground the moral demand for respect and goodwill. (It is not even clear that it is necessary to every case where there is a sharing of ends—for example, the end of recovery or a normal life.) This puts the cart before the horse. We are born into the participant stance. We engage in reciprocal exchanges long before we become autonomous agents; indeed it is entirely plausible that such exchanges are essential to the development and maintenance of our agency. Admission to the participant stance cannot and should not depend on the equal capacity of the parties. Annette Baier (1995: 28–9) points out: we begin as helpless children, at almost every point in our lives we deal with both the more and the less helpless . . . equality of power and interdependency, between two persons or groups, is rare . . . The recognition of the importance for all parties of relations between those who are and cannot but be unequal, and of their effect on personality formation and so on other relationships, goes along with a recognition of the fact that not all morally important relationships are freely chosen.

14

The description also recalls Aristotle’s description of virtuous friendship.

110

jeanette kennett

Imperfect Reciprocity and the Grounds of Respect ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Let’s go back, briefly, to Mutnik. Here is an example of an (admittedly primitive) reciprocal social interaction that is not subject to the demand that the other deal with reasons as I would—and yet does provide the psychological visibility and affirmation that is so important to the development or maintenance of a positive sense of self. And there are many other examples of exchanges of affection, shared amusement, sadness, and so forth where this demand is similarly absent, highly domain-specific, or beside the point of the interaction. A father and his young son may share their enjoyment of the slapstick comedy of the Three Stooges without there being any requirement on the son that he be able to appreciate any more sophisticated humor and without the fact that he cannot leading to any diminution of the respect and goodwill owed to him quite generally by his father. Similarly, the person whose spouse develops Alzheimers may come to accept and find value in a changing relationship worked around and respecting those aspects of the self—the values, memories, habits, and emotions—that remain.15 These and the earlier examples of interactions between friends serve to highlight the fact that what is valuable in our relations with each other, and our moral standing within those relations, is not given wholly by the features that make us accountable agents. I can value your clumsiness, your ill-fated attempts to hide your fear of spiders, your warmth, your laugh, the way you look at me. These sorts of valuings are part and parcel of relationships and help structure the kinds of communication that go on in them and the conditions under which the particular reactive attitudes arise. Impairments in agency and social responsiveness such as occur in many mental disorders also structure what is possible in interpersonal interactions; they may place restrictions on its scope and depth, they may require the modification of some, though not all, of the reactive attitudes. It is commonly the case, though, in relationships that we adjust our responses and expectations in the light of facts about the other, their sensitivities, temperament, etc., and we do this without abandoning the participant stance. A survivor of the Changi prisoner of war camp may be utterly rigid in his views of all things Japanese and perhaps understandably so. If I say ‘Don’t talk to Granddad about Japanese art. You’ll only get him going’, I am taking an objective view of this feature of Granddad and of the situation, but I need not thereby depart from the participant stance. I may in fact be showing a highly particularized concern and respect for him. The ‘sort of love’ that adults have with their unimpaired close friends or partners may be out of reach for some people suffering mental illness or disorder, and this, I have argued, can be a great misfortune, but it would be a mistake to think of such 15

See Agnieszka Jaworska’s perceptive discussion in Jaworska (1999).

mental disorder, moral agency, the self

111

high-level, wide-ranging mutual responsiveness as providing a description or a norm for interpersonal relations in general. We are neither forced nor normatively required to move to the objective stance in our dealings with the mentally ill, since not all social interactions must presuppose full autonomous agency and since we can establish or maintain areas of reciprocity with persons whose agency and social responsiveness is underdeveloped, impaired, or intermittent. However, given that it can be difficult to remain in the participant stance without falling prey to misplaced resentment or other harmful reactive attitudes, we might wonder whether anything obliges us to remain, or to try to remain, there, and to pursue engagement when our ordinary expectations of reciprocity and reason sharing are disappointed. Here I believe that the best argument (though perhaps not the whole argument) is to be found within a Kantian perspective. I’ve argued that a wholesale shift to the objective stance denies the mentally ill person psychological visibility and so further depletes or undermines their picture of themselves and their efficacy as agents. It fails to acknowledge or to support their stake in constructing a unified autonomous self and in regaining authorship of their actions and lives. In moving to the objective stance we deny or ignore the agency of the other. That is what it is to treat someone as ‘an object of social policy . . . to be understood and controlled in the most desirable fashion’ (Strawson 1982: 66). Now, as beings with the goal of autonomy—as beings constitutively concerned with discovering and acting for reasons—Kant believed that persons had intrinsic worth and dignity and as such were never to be used merely as means. To treat a person as an object is clearly to mistreat them. The demand for respect and goodwill in our dealings with each other is, in the first place, the demand that we approach each other from within the participant stance. It is minimally the demand that we do not ignore or undermine each other’s agency. For many latter-day Kantians it is more. They recognize that agency comes in degrees, that autonomy is an achievement, and that respect for autonomy may require our active support for the agency of those in adverse circumstances. Mental illness is such an adverse circumstance. In especially severe cases we may not have any great reason for optimism about the prospects of the ill person regaining control of their lives and achieving the goods conditional upon unified agency; even in less severe cases we may not always be sure to what extent the person’s actions are symptoms of her illness rather than manifestations of her (albeit impaired) agency, and so we may not be sure that our responses are appropriate. But our uncertainty does not counter the moral reason to adopt as far as possible the participant stance and to adapt it in response to the capacities of the ill person. The demand for goodwill from each other is a powerful moral demand and one that mentally ill people are as entitled to make as mentally well people. In the professional context and the broader social context as well as in more personal contexts it is a demand for patience and charity of interpretation of the other’s words and actions particularly as they reflect on the person’s character. It is the demand that we remain receptive to, and promote, opportunities for engagement;

112

jeanette kennett

for shared activities and shared goals. Where mental illness plays a causal role in cases where an agent causes injury or offence this should inhibit blame, though maybe not always, but it need not inhibit the participant reactive attitudes more generally. Given the importance of social relations in building, supporting, and strengthening agency, and so in securing access to central human goods, the default stance in both personal and professional dealings with those suffering a mental illness or disorder, as with anyone else, must be the participant stance.

References Agnetti, G., and Young, J. (1993), ‘Chronicity and the Experience of Timelessness: An Intervention Model’, Family Systems Medicine, 11/1: 67–81. Aristotle, (1984), Magna Moralia, ed. J. Barnes, in The Complete Works of Aristotle: The Revised Oxford Translation, ii, Bollingen Series (Princeton: Princeton University Press), 1868–1921. Baier, A. (1994), ‘The Need for More Than Justice’, in Baier, Moral Prejudices: Essays on Ethics (Cambridge, Mass.: Harvard University Press, 1995), 18–32; first pub. in Canadian Journal of Philosophy, suppl. vol. 13: 197–220. Branden, N. (1993), ‘Love and Psychological Visibility’, in N. K. Badwhar (ed.), Friendship: A Philosophical Reader (Ithaca, NY: Cornell University Press), 65–73. Bratman, M. (2000), ‘Reflection, Planning, and Temporally Extended Agency’, Philosophical Review, 109: 35–61. Cocking, D., and Kennett, J. (1998), ‘Friendship and the Self ’, Ethics, 108/3: 502–27. (2000), ‘Friendship and Moral Danger’, Journal of Philosophy, 97/5: 278–96. and Matthews, S. (2000), ‘Unreal Friends’, Ethics and Information Technology, 2: 223–31. Cooper, J. (1980), ‘Aristotle on Friendship’, in A. Rorty (ed.), Essays on Aristotle’s Ethics (Berkeley: University of California Press), 301–40. Coryell, W., Keller, M., Lavori, P., and Endicott, J. (1990), ‘Affective Syndromes, Psychotic Features and Prognosis. II: Mania’, Archives of General Psychiatry, 47/7: 658–62. Deveson, A. (1998), Tell Me I’m Here (Ringwood, Vic; Harmondsworth: Penguin). Frankfurt, H. G. (1971), ‘Freedom of the Will and the Concept of a Person’, Journal of Philosophy, 68: 5–20. Garety, P. A., and Freeman, D. (1999), ‘Cognitive Approaches to Delusions: A Critical Review of Theories and Evidence’, British Journal of Clinical Psychology, 38: 113–54. Jaworska, A. (1999), ‘Respecting the Margins of Agency: Alzheimer’s Patients and the Capacity to Value’, Philosophy and Public Affairs, 28/2: 105–38. Kennett, J. (2001), Agency and Responsibility (Oxford: Clarendon Press). and Matthews, S. (2003), ‘The Unity and Disunity of Agency’, Philosophy, Psychiatry, Psychology, 10/4: 305–12. (eds.) (2004), International Journal of Law and Psychiatry, 27/5, special issue: Responsibility and Mental Impairment. Korsgaard, C. (1988), ‘Personal Identity and the Unity of Agency: A Kantian Response to Parfit’, Philosophy and Public Affairs, 18: 101–32.

mental disorder, moral agency, the self

113

(1992), ‘Creating the Kingdom of Ends: Reciprocity and Responsibility in Personal Relations’, in J. E. Tomberlin (ed.), Philosophical Perspectives, vi: Ethics (Atascadero, Calif.: Ridgeview), 305–32. (1996), The Sources of Normativity (Cambridge: Cambridge University Press). Moore, M. S. (1975/1982), ‘Some Myths About Mental Illness’, Archives of General Psychiatry, 32: 1483–97; repr. in R. B. Edwards (ed.), Insanity, Rational Autonomy and Mental Health Care (New York: Prometheus Books, 1982), 33–49. Perring, C. (2002), ‘Mental Illness’, in Edward N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Summer 2002 Edition),

. Phillips, P. (2003), ‘Psychopathology and the Narrative Self ’, Philosophy, Psychiatry, Psychology, 10/4: 313–28. Radden, J. (2002), ‘Psychiatric Ethics’, Bioethics, 16/5: 397–411. (2003), ‘Learning from Disunity’, Philosophy, Psychiatry, Psychology, 10/4: 357–60. Schopp, R. F (1991), Automatism, Insanity, and the Psychology of Criminal Responsibility (Cambridge: Cambridge University Press). Strawson, P. F. (1982), ‘Freedom and Resentment’, in G. Watson (ed.), Free Will (Oxford: Oxford University Press, 1982), 59–80; first pub. in Proceedings of the British Academy, 48 (1962), 1–25. Sweeney, P. D., Anderson, K., and Bailey, S. (1986), ‘Attributional Style in Depression: A Meta-Analytic Review’, Journal of Personality and Social Psychology, 50/5: 974–91. Szasz, T. (1961/1974), The Myth of Mental Illness (New York: Harper and Row). (1998), ‘Commentary on ‘‘Aristotle’s Function Argument and the Concept of Mental Illness’’ ’, Philosophy, Psychiatry & Psychology, 5/3: 203–7. Velleman, J. D. (1997), ‘Deciding How to Decide’, in G. Cullity and B. Gaut (eds.), Ethics and Practical Reason (Oxford: Clarendon Press), 29–52. (2000), ‘Well-Being and Time’, in Velleman, The Possibility of Practical Reason (Oxford: Clarendon Press), 56–84. (2005), ‘The Self as Narrator’, in J. Christman and J. Anderson (eds.), Autonomy and the Challenges to Liberalism (Cambridge: Cambridge University Press), 56–76. Wallace, R. J. (1994), Responsibility and the Moral Sentiments (Cambridge, Mass.: Harvard University Press). Watson, G. (1982), ‘Free Agency’, in Watson (ed.), Free Will (Oxford: Oxford University Press), 96–110. (1987), ‘Responsibility and the Limits of Evil: Variations on a Strawsonian Theme’, in F. Schoeman (ed.), Responsibility, Character, and the Emotions: New Essays in Moral Psychology (Cambridge: Cambridge University Press), 256–86. Watts, J., and Priebe, S. (2002), ‘A Phenomenological Account of User’s Experiences of Assertive Community Treatment’, Bioethics, 16/5: 439–54. Wells, L. A. (2003), ‘Discontinuity in Personal Narrative: Some Perspectives of Patients’, Philosophy, Psychiatry and Psychology, 10: 297–303. Young, J. (1994), ‘The Loss of Time in Chronic Systems: An Intervention Model for Working with Longer Term Conditions’, Australian and New Zealand Journal of Family Therapy, 15/2: 73–80.

chapter 5 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

‘REINVENTING’ THE RULE OF D OUBLE EFFECT •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

daniel p. sulmasy The Rule of Double Effect has played an important role in bioethics, especially during the last fifty years. Its major application in bioethics has been in providing physicians who are opposed to euthanasia with a moral justification for using opioid analgesics in treating the pain of patients whose death might thereby be hastened. It has also prominently been applied to certain obstetric cases. The scope of application of double effect is actually much broader than medical ethics, extending to cover such topics as strategic bombing in warfare. This chapter, while general in theory, will concentrate on its applications in medical ethics. The Rule of Double Effect has been heatedly debated in the literatures of philosophy, theology, and bioethics. Some have tried to derive all of morality from it (Knauer 1979). Some have pronounced it a dead idea, riddled with confusions and inconsistencies (Davis 1984; Quill et al. 1997). Some have felt that it contains some kernel of truth and have either tried to revise it radically in order to preserve that truth (Quinn 1989) or have suggested that whatever kernel of truth it contains must be defended in some other form of moral analysis (McIntyre 2001; Foot 1978). Still others have defended it in a more or less traditional formulation (Boyle 1980). Even the name given to the notion of double effect has proven controversial. Some have called it a Dogma; some a Doctrine. Some have called it a Principle; others a

‘reinventing’ the rule of double effect

115

Rule. Following Beauchamp and Childress,1 I will call it the Rule of Double Effect. It seems to me that it is best to label it a rule since it is not sufficiently fundamental to any ethical theory to be called a principle. Rather, it seems to be a meta-rule used in deciding about certain types of case in which more fundamental rules appear to clash. Calling it a dogma or a doctrine has the potential to obscure debate by suggesting theological and/or legal overtones, or by suggesting a prominence that may be unwarranted. While many philosophers have dismissed it as an anachronism, the persistence of the debate about double effect suggests that the idea has some staying power. In this chapter, I will propose a substantial reformulation of the requirements of the Rule of Double Effect (RDE), making use of recent developments in the philosophy of mind and taking seriously the critiques of those who quite simply have found many suggested applications of the RDE to be incredible. One should understand that by ‘reformulation’ I will not simply be defending the traditional form of the RDE. But one should also understand that I will not be completely revising it. My intention is to show how the rule itself is fundamentally sound, both morally and conceptually, but that the conditions governing the rational application of the RDE, particularly with respect to the intention–foresight distinction, must be substantially tightened if that soundness is to be preserved.

The Traditional Formulation of the RDE ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Most commentators turn to Mangan (1949) for an authoritative historical account of the RDE and its traditional formulation. While I agree with Mangan that the historical discussion and evolution of the RDE finds its beginning in Aquinas’ discussion of killing in self-defense, Aquinas himself never articulated the RDE or anything that looked much like the rule as it is known today. Properly speaking, the RDE emerged after centuries of reflection on Aquinas’ discussion of self-defense and its application to an ever widening circle of cases.2 As will become clear from what I argue below, I do not think the RDE actually works as a justification for using lethal force in cases of self-defense.3 Aquinas’ justification of killing in self-defense 1 Beginning with the fourth edition of their very widely used textbook, Tom L. Beauchamp and James F. Childress (1994: 206) changed from calling it a ‘principle’ to a ‘rule’. 2 The first clear formulation of the RDE that approximates its present form does not come until Jean-Pierre Gury in the 19th century (see Kaczor 1998). 3 Thomas A. Cavanaugh (1997) thinks that Aquinas would not permit the use of force that one knew would be lethal, while Joseph M. Boyle (1978) thinks Aquinas would permit this. Cavanaugh seems correct about Aquinas’ logic, while Boyle seems correct about what Aquinas thought he was proving. That such confusions have never been resolved suggests the need for a fresh approach.

116

daniel p. sulmasy

would fall outside the narrow constraints of the intention–foresight distinction that I articulate below. Nonetheless, part of the genius of Aquinas was to articulate enough ideas that an RDE eventually emerged. And that RDE, I believe, can be rendered in a manner that is conceptually and morally sound. The traditional formulation, according to Mangan, is as follows. A person may licitly perform an action that he foresees will produce a good and a bad effect provided that four conditions are verified at one and the same time: 1. 2. 3. 4.

that the action in itself from its very object be good or at least indifferent; that the good effect and not the evil effect be intended; that the good effect not be produced by the evil effect; that there be a proportionately grave reason for permitting the evil effect.

The classical case in medical ethics to which the RDE has been applied has been the use of morphine to treat the pain of dying patients. Applying the RDE to morphine injections, one notes that (1) they are at least morally indifferent actions; (2) one may presume that the physician only intends to relieve the patient’s pain and not to hasten the patient’s death; (3) the cessation of respiration and subsequent death of the patient are not the means by which morphine relieves pain; and (4) given serious pain and imminent death, there seems to be a proportionately grave reason for permitting the risk of hastening death. Therefore, even a scrupulously conscientious physician who is morally opposed to euthanasia may licitly perform the action of injecting the morphine according to the RDE. Nonetheless, the RDE has raised serious controversies in the literatures of medical ethics and philosophy. I will argue that much of this discussion has been hampered by two significant errors that have made it difficult to evaluate this discussion.

Two Significant Errors ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Two significant errors have plagued the last fifty years of debate about the RDE. These errors have resulted in many fruitless efforts to defend and to refute the rule. I intend to discuss these errors briefly. Following this discussion, I will present a ‘reinvented’ version of the RDE—a version that I will argue is philosophically sound and morally credible.

Lack of Appreciation of Developments in the Philosophy of Mind The first significant error that has plagued recent debates about the RDE is one that has afflicted both its opponents and its proponents, although the impact of this error has been greater for opponents. The error is this: contemporary discussions

‘reinventing’ the rule of double effect

117

of the RDE have largely ignored important developments in action theory and have been written as if moral philosophy and philosophy of mind were parallel universes that never communicate or influence each other. An important example in this regard is the treatment of the RDE by Beauchamp and Childress. They write that ‘the distinction between what is intended and merely foreseen in a planned action is not viable’, and cite John Searle in support of this position.4 However, in citing Searle, the authors appear to misinterpret him.5 In the passage they cite, Searle is actually making a distinction between different kinds of unintended consequences of actions. He is not discussing the distinction between the intended and the foreseen. Searle actually fully and explicitly defends the distinction between the intended and the foreseen in his later book Intentionality, stating, ‘A common confusion is to suppose that if someone knows that something will be a consequence of his action then he intends that consequence’ (1983: 103). That bioethicists simply insist (without philosophical argument) that there is no distinction between the intended and the foreseen, and pay so little attention to recent developments in philosophical action theory, seems a serious lapse in scholarly rigor. In this chapter I will offer an explanation and defense of the distinction between the intended and the foreseen based on the work of John Searle (1983), Michael Bratman (1987), and Alan Donagan (1987). This will lead me to a reformulation of the requirements for the rational application of the RDE that will address some of the worries expressed by opponents of the RDE. I fully realize that this approach will never satisfy those who simply hold that intentions never matter in the moral evaluation of human acts, or who insist that contemporary theories of human action are all mistaken. But the burden of proof would seem to be upon those who raise such objections to show how contemporary scholars in philosophy of mind are all wrong about the distinction between the intended and the foreseen, or admit that the distinction is valid but supply moral arguments about why it is nonetheless morally irrelevant.

Historical Misapplications of the RDE Now Function as Paradigm Cases A second significant error has been committed by proponents of the RDE. In the early twentieth century they misapplied the rule in a way that has subsequently stretched the distinction between the intended and the foreseen beyond the point 4

See Beauchamp and Childress (2001: 131 and n. 44). See Searle (1980: 65). Searle’s concern in this passage is with the unintended aspects of actions that seem to count as relevant aspects of the action (such as Oedipus’ marrying his mother) and the unintended aspects of the action that seem so far removed from the core of what the agent is doing and thinking that they cannot be considered parts of the action at all (such as Oedipus’ causing some neurons in his brain to fire and say, ‘I do’). 5

118

daniel p. sulmasy

of credibility.6 This may have solved certain moral problems from the perspective of Catholic theology in the short run, but in the long run it has proven disastrous to the defense of the RDE in the wider world.7 I would argue that the mistake was the application of the RDE to the cases of tubal ectopic pregnancy and the cancerous gravid uterus. Both proponents and opponents seem to accept the traditional view that these are typical applications of the RDE (even paradigmatic instances of its use), and have conducted their debates accordingly. But if these cases represent a misapplication of the RDE, then the moral and conceptual soundness of this rule does not rise and fall with discussions of these cases. In hindsight, it seems that the RDE was applied to these cases by Catholic medical moralists in an attempt to justify their intuitive belief that life-saving operations upon women who carried tubal ectopic pregnancies or were pregnant while suffering from cancer of the uterus were morally justifiable and distinct from the ethics of abortion. Since Catholic teaching in opposition to abortion has been absolute, and the theological climate of the day was not very open to innovation, I suspect that the only principle to which these Catholic moralists felt they could appeal in order to justify this procedure was the RDE. Unfortunately, however, this application of the RDE has rendered the whole notion (which I will contend is basically sound) incredible to many observers both inside and outside the Catholic faith. In extending the RDE to cover these cases, they have jeopardized the intellectual standing of this important moral rule with respect to cases in which it does readily apply. Once one gives up on the idea that the RDE applies to these cases, however, the task of defending it becomes much easier. In ‘reinventing’ the RDE, I will formulate the conditions for the rational application of this rule in such a way that the cases of ectopic pregnancy and the cancerous gravid uterus are excluded. I believe that surgical interventions for these conditions can be justified, morally and conceptually, on grounds other than the RDE that do not require an acceptance of the moral permissibility of abortion. Explaining fully how this is so is beyond the scope of this chapter. For the duration of this chapter I will proceed as follows: First, I will present my ‘reinvented’ set of requirements for the correct application of the RDE. Second, I will present the underlying action theory that corrects the first significant error and supports and explains my reformulation of the distinction between the intended and the foreseen in the RDE. Third, I will apply this reinvented RDE to the cases of morphine and ectopic pregnancy to show how this corrects the second significant 6 This complex and fascinating history is detailed by David F. Kelly (1979: 274–309). The morality of hysterectomy in the cancerous gravid uterus case was justified under double effect at the very beginning of the 20th century. After decades of debate, the doctoral dissertation of an American Jesuit named Bouscaren, by applying the rule of double effect, finally won acceptance within the Church for surgery for ectopic pregnancies in 1933. 7 See, for instance, relatively early attacks by Williams (1957: 200–5) and Glover (1977: 86–91).

‘reinventing’ the rule of double effect

119

error. Fourth, I will consider some major objections and suggest how I think the arguments that I have presented can move the debate about the RDE forward constructively.

A Reformulation of the Rule of Double Effect ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

If one encounters a conflict between a specific duty to do good and one’s general duty to avoid evil, and 1. this conflict arises from one intentional act with at least two foreseeable effects; 2. the act itself is either morally good or morally neutral; 3. the conflict of duties arises because intentionally bringing about one effect is morally good while intentionally bringing about another effect is morally bad; 4. all other reasonable means of achieving the good effect with less risk of causing the bad effect have been exhausted;8 5. the good and bad effects are not foreseen as coming about by way of intervening agents;9 6. one’s prior intention is to act in such a way that one’s intention in acting is to bring about the good effect while the foreseen bad effect is not part of one’s intention in acting; 7. one is sincere and rational in one’s report of one’s complex prior intention, such that, at the very least: (a) the intended act itself is not an alternative definite description of the bad effect; (b) the good effect is not an alternative definite description of the bad effect; (c) the bad effect is not wholly spatiotemporally contained within the intended act itself; (d) the bad effect is not wholly spatiotemporally contained within the good effect; (e) the intended act does not entail the bad effect logically; (f) the good effect does not entail the bad effect logically; (g) the bad effect is not one’s further intention in so acting; (h) the bad effect is not an empirically necessary causal condition for the good effect; (i) one does not intend the good effect by way of the bad effect; 8 Space precludes me from defending each of the amendments I am making to the classical formulation of the RDE. For a discussion of Requirement 4, see Sulmasy (2000). 9 If there are intervening agents, the proper moral category for analysis is cooperation or complicity, not double effect. See Sulmasy and Pellegrino (1999).

120

daniel p. sulmasy

8. and the act is undertaken with due proportionality, which is to say: (a) the good effect is proportionate to the bad effect; (b) the means under consideration are proportionate to the expected effects,10 then one is morally responsible for having undertaken the act with due diligence, in accordance with this rule, and in this sense one is morally responsible for all of the good effects and bad effects of the act one has undertaken intentionally. But one is not morally culpable for having brought about the bad effect of the act.11

The Action Theory Underlying This Reformulation of the RDE ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

The first point to establish is that intentions matter in the moral evaluation of human acts. This is, of course, denied by utilitarians and other consequentialists, who discount the moral significance of agency and look only to consequences. The importance of intention is sometimes discounted by those who worry about the public verifiability of intentions and prefer to limit the moral evaluation of human acts to the observable aspects. Some will accept the moral relevance of intention, but reduce it to a character trait: something that is good in so far as it produces morally good actions or morally good results. But it seems plain to the common person that intentions matter in the moral evaluation of acts. One learns from childhood that one is far more deserving of blame for what one has intentionally brought about than for what one has not intentionally brought about. Judges and juries expend great energy in distinguishing between manslaughter, homicide, and the degrees of murder based upon complex evaluations of the intentions of the defendant. Intentions are the backbone of the agency that is in moral acts. Moral acts do not merely happen. They happen because agents make them happen. And agents make their acts happen through intentions. To discount the importance of intentions in the moral evaluation of human acts is 10

This requirement is infrequently noted. Not only must the outcomes be proportionately balanced, the means used must ‘fit’ the end intended. For example, normally one would not be justified in using an expensive, toxic, intravenously administered drug such as amphotericin B to treat athlete’s foot. See Kaczor (1998: 310–12). 11 This is another very important point often lost on critics of the RDE, but perhaps because it is rarely made explicit by proponents. The RDE does not abolish moral responsibility. There is a distinction between what one does intentionally and what one intentionally does. The former simply refers to an act that one freely undertakes. The latter refers to what one aims at accomplishing in so acting. A physician who administers morphine to a dying patient does so intentionally, and is morally responsible for the act and all of its outcomes. This means that she should not do so without engaging in some form of double effect reasoning, so that she is acting responsibly. But this does not mean that she is intentionally aiming at making the patient dead. If she did, she would be morally culpable. See Bratman (1987: 124–5).

‘reinventing’ the rule of double effect

121

to discount the central feature that makes these happenings moral happenings in the first place. This is not to say that it is easy to explain what an intention is, or easy to know what an agent’s intentions actually are in any given act. But for moral philosophers to give up on ever doing so because it is too hard would be analogous to medical students giving up on molecular biology because it is too hard. Nearly half a century ago moral philosophers were challenged by Elizabeth Anscombe (1958) to take up the questions of philosophical moral psychology. Few have done so. Even fewer bioethicists have done so. Bioethicists have sometimes dismissed the importance of intention in the moral evaluation of clinical acts because they claim that intention is an unclear or philosophically underdeveloped notion.12 And yet, exciting work in the philosophy of mind has emerged over the past few decades, full of rich insights that can contribute importantly to ongoing work in moral philosophy and in bioethics. This work undergirds the reinvention of the RDE that I am proposing, especially with respect to the task of making the distinction between the intended and the foreseen clearer and more rigorous.

Definitions The reformulation of the RDE that I have offered uses a number of technical terms from the philosophy of mind that may not be familiar to readers of the bioethics literature. I will therefore provide some useful definitions: Definition 1: Acts. Donagan holds that the voluntary doings of agents constitute the class of things that are known as acts, and that acts are ‘a species of event’ (1987: 19). Acts are voluntary human doings that can be explained by their doers’ propositional attitudes (Donagan 1987: 23). Definition 2: Propositional attitudes. A propositional attitude is a cognitive relationship between a person and a proposition (Wagner 1995). Examples include Joan’s believing, hoping, desiring, being disappointed, or being surprised that her mother has been cured of cancer. Definition 3: Events. An event is a spatiotemporal occurrence. Events are changes or persistences in states of continuing objects. Ontologically, ‘events are genuine individuals’, not states of affairs (Donagan 1987: 38–9). They are represented linguistically, as other individuals are, by individual names or definite descriptions, not by sentences (Donagan 1987: 19). Definition 4: Intentions. An intention is an irreducibly simple attitude, ascribable to an agent, that takes as its object a proposition about an event that the agent himself chooses to bring about under a definite description. If the event occurs in the way the agent has chosen, this attitude will explain the event. 12

See e.g. Kuhse (1987: 24) and Brock (1993: 172–5).

122

daniel p. sulmasy

Definition 5: Conditions of fulfillment. Searle (1983: 14) notes that as propositions have truth conditions, so intentions have conditions of fulfillment. If an event comes about in the way the agent intends it, the agent’s intention has been fulfilled. Definition 6: Prior intentions. To have a prior intention means that an agent has consciously chosen to bring about an event (i.e. an end) in advance of an act. When fully mature, this prior intention also includes a choice of a means of bringing about the intended end. When complex, this prior intention includes all those (generally multiple) cause and effect relationships that are part of the agent’s plan. Definition 7: Intention in acting. To have an intention in acting means that an agent has acted or is acting, and that one can ascribe to the agent’s act the choice of a complete act—both an end and a means of achieving that end. Definition 8: Further intention. To have a further intention means that the agent intends to act upon a complex prior intention that includes not only the immediately intended event, but also some further event that the agent has chosen as part of her plan.

Intentional Causation Causation is itself a vexatious philosophical topic. For the purposes of this chapter, however, all that is required is an intuitive grasp of what is meant by causation. The theory of intention that I present should be compatible with almost any theory of causation one would choose to believe. Agents act in the world. Acts are, as defined above, those events that take place in the world that are explained by the intentions of agents. Agents cause events in the world by way of their intentions. Precisely how this comes about is beyond the scope of this chapter, but it is a great marvel that the events we call acts are causally explained by way of the intentions of agents. Yet one can be more precise. Acts are causally explained by the propositional content of intentional choices that must refer to some agent’s intention in acting (Searle 1983: 85 ff.; Donagan 1987: 88). Acts are events that are explained when the conditions of fulfillment of the intention of an agent are realized by way of the agent himself or herself carrying out that intention. There is a necessary self-referentiality to one’s intentional causation. Thus, not everything that happens after an agent acts can be explained as having been intentionally caused. While an agent may have caused events other than the condition of fulfillment of her intention, properly speaking, these other events are not part of the agent’s act. Ethically speaking, events that are not properly part of the agent’s act are evaluated differently from those that are. For example, suppose that Tim should intend to write Diane a prescription for 2 grams of secobarbital so that she could commit suicide, and were to have told her that this was his intention. Suppose that he were to have written a prescription

‘reinventing’ the rule of double effect

123

for Debbie for 2 grams of secobarbital, but had not yet written the prescription for Diane. Now, suppose further that Diane were to have shown up at Tim’s office, seen the prescription for secobarbital written for Debbie, thought it was for her, had it filled, consumed all the pills, and died. It would be true that Tim’s intention caused Diane to die, but it would not be true that Diane’s death could be explained as having been intentionally caused by Tim. Her death did not come about by way of Tim’s acting upon the full propositional content of his actual intention. If the district attorney were to investigate him as the agent of Diane’s death, he would need only to show that the prescription that Diane filled was written for Debbie, and the charges would be dropped. Agents need no theory of causation in order to act, but agents grasp the causal potential of their intentions. The plans that agents make often involve complex causal paths. As I shall explain in greater detail below, if an agent has formed a complex prior intention that includes further intentions as part of a plan, the conditions of fulfillment of that complex prior intention include all of the events that are foreseen as causally required in order to fulfill the furthest intention in that plan. As I will explain in greater detail later in this chapter, this is why conditions (g), (h), and (i) of Requirement 7 of the Reformulated RDE are so important. They address the agent’s intended causal chain, not merely the agent’s intended results. For example, if Jack intends that Janet should die by way of her pushing a button that causes a programmed series of injections of drugs that will first cause her to be sedated, then cause her to be paralyzed, then cause her heart to stop beating, and thus cause her death, the condition of fulfillment of Jack’s intention is that Janet’s death should come about by way of his having that complex prior intention—a complex prior intention that includes all of the causally necessary conditions for her dying in just this way.

The Irreducibility of Intention Knowing something about what one means by intentional causation still does not ensure that one has adequately understood what an intention is. Understanding intention as a propositional attitude raises questions about the relationship between intention and other propositional attitudes. It became common in the twentieth century, under the influence of positivism and behaviorism, for many philosophers to hold that intentions were reducible to beliefs, desires, or some combination of these other propositional attitudes. Against this, Aristotle (1985: 1111b 12) wrote that, ‘Those who say decision is appetite or emotion or wish or some sort of belief would seem to be wrong.’ Contemporary philosophy of action now holds that Aristotle was correct. Donagan takes an Aristotelian position and argues that the propositional attitude of intention cannot be reduced to any other propositional attitude. And he and

124

daniel p. sulmasy

others have amassed a significant array of arguments to make it clear that while certain beliefs and desires might be relevant to an intention, no combination of belief and desire is sufficient to account for an intention. While intentional acts require a whole nexus of background beliefs, intention cannot be reduced to belief alone. There are several reasons for this. First, beliefs alone, as Brand (1984: 94–7) points out, cannot initiate acts the way intentions can. For example, Tim’s belief that he will write a prescription for a lethal dose of secobarbital for Diane is not sufficient for him to act on that belief, nor would it be sufficient to explain his so acting if he were to write such a prescription. No matter how strong his belief, the act will not take place unless Tim is moved to do so, and a belief is not a being-moved-to-act. Second, Searle (1983: 3) simply points out that the ‘direction of fit’ differs in intention and belief. Belief has a ‘mind to world direction of fit’. That is, the truth conditions for beliefs involve the conforming of the believer’s mind to the world. On the other hand, intention has a ‘world to mind direction of fit’. The conditions of satisfaction of intentions involve the world coming into conformity with the agent’s mind. Given this fundamental asymmetry between intentions and beliefs, it is hard to see how one could be reduced to the other (Searle 1983: 9–10). Finally, it does not seem necessary for an agent to believe that the intended outcome of an intentional act will be accomplished by the act. It would seem that one can very strongly intend what one believes one will not accomplish, and there is nothing illogical about this. One can intend what one does not believe will happen. For example, one could intend to withstand torture even if one were to believe that the overwhelming likelihood were that one would eventually crack (Gustafsen 1986: 68). Similarly, a doctor could intend to save a patient’s life and act accordingly even if she were to believe that this would be, to a reasonable degree of medical certitude, impossible. One can intend to do either what one believes one will accomplish or what one believes one will not accomplish.

Intentions Are Not Desires Neither can intentions be reduced to desires (Anscombe 1969: 5–6). It is certainly true that desires have the same direction of fit as intentions, directed towards the conforming of the world to the mind of the agent, and this may have led certain thinkers to try to account for intention solely on the basis of desire. But having the same direction of fit does not mean that desire and intention are the same. Desire is neither necessary nor sufficient for intention (Brand 1984: 122–3). First, equating intention with desire makes accounting for the problem of wayward causal chains an insoluble problem (Gustafsen 1986: 44). The fact that something an agent desires actually comes to pass in no way implies that its occurrence was intentional. Suppose that Tim desired to give Diane the knowledge

‘reinventing’ the rule of double effect

125

and the means to commit suicide. Suppose that he had discussed this with his postdoctoral fellow, a young and enthusiastic general internist. Suppose that, while Tim was off giving a lecture about assisted suicide, Tim’s postdoctoral fellow were then to write a prescription for an overdose of secobarbital and were then to give it to Diane along with instructions on how to use it. If Diane were then to take the overdose, would one say that her death was intentional on the part of Tim simply because he desired it? Hardly. And if that is the case, how could one say that what happens intentionally is what one desires to happen and actually comes about? Second, it seems clear that one can easily intend what one does not desire, and that one can desire what one does not intend (Brand 1984: 122–3). My desire to eat ice cream does not mean that I intend to eat ice cream. At the moment, I have a strong desire to eat ice cream. I enjoy it very much. But I do not intend to eat ice cream. In the short run it would ruin my appetite for dinner. In the long run it would contribute to the ruin of my arteries by clogging them with cholesterol-rich atherosclerotic plaques. And because I think it would be noble to set a good example for my patients (whom I instruct to refrain from eating ice cream), I overrule my desire for ice cream and do not intend to eat it. Similarly, I often intentionally do what I do not desire to do. I frequently do not desire to stay at the office and see patients who show up late and happen to be very sick and in need of immediate medical attention. This is particularly true when my colleague who is supposed to be on call for such emergencies can’t be reached and I have had to stay late and have missed supper the two previous nights in a row and the patient who has shown up late but very sick has not followed my instructions and has a habit of showing up late and I have a tennis court reserved and love to play tennis. Despite my desires to the contrary, I act intentionally in staying with the patient and providing the necessary care. Intention cannot be reduced to desire alone. A frequent response to such arguments is to say that there is no such thing as an intention contrary to desire. On this theory, one really always does what one desires the most in any given situation. In the example given above, it can be argued, the desire to eat ice cream is simply not as strong as the desire to enjoy dinner or to maintain cardiovascular health or not to appear hypocritical to my patients. In the other example, it might be said (charitably) that my desire to help patients or (less charitably) my fear of a malpractice suit are my predominant desires and that I simply desire these more than playing tennis. But this argument can be dispensed with on several grounds. First, one must note that this begs the question in a serious way. If one has presumed that there is no such thing as an undesired intentional act, one simply reinterprets every other kind of explanation of an act as a disguised desire. Second, as Bratman (1987: 19) points out, even the postulation of such a predominant desire ‘does not guarantee that my desire will control my conduct’ the way that it seems an intention would. For example, Tim may begin an office

126

daniel p. sulmasy

visit with Diane feeling as if he can offer her no more help except by giving her the knowledge and the means to commit suicide and may have a predominant desire to do just that: to help her commit suicide. But this does not mean that he has settled upon giving her the means and the knowledge, nor has he settled on a plan for doing so (like a prescription for secobarbital). He still might not be sure that he will do so. He might also have a desire to make a breakthrough in the quality of medical care by launching a national campaign for assisted suicide and might see this case as a possible test case for his idea. But he might see his acts as constrained by certain principles. Tim might think that he would not assist Diane in suicide if she had relapsed in either her alcoholism or her depression, and so might not have formed an intention to write the lethal prescription and give the lethal instructions before talking to her, even though he might be quite desirous of doing so. This is not to say, of course, that intentions could not be formed on a provisional basis and later reconsidered. Bratman (1987: 5) is only arguing that intentions can constrain behavior even in the face of a dominant desire. Third, most of the examples of a conflict between desire and intention involve the concepts of duty and external coercion. As Brand (1984: 123) points out, in these conflict situations it is simply not the case that doing one’s duty or giving in to external pressures is equivalent to desiring what one does not desire. Brand gives the example of Richard Nixon resigning as President. Nixon could not truly be said to have resigned because he desired not to be impeached more than he desired to remain in office. Nixon’s predominant desire, one supposes, was to remain in office more than anything else. Yet his resignation was an intentional act. Or suppose that I desire that my patient should die. Suppose she has widespread metastatic cancer of the breast, is suffering great pain that is mostly, but not completely, relieved by medication, but that this incomplete pain relief had to come at the price of being sleepy most of the time. I might very well desire that she should die quickly, thinking that this would be for the best. I might even pray for her quick death. I might come to consider her death my predominant desire. But it seems clear that, even though I might strongly desire that she die, this does not mean that I intend that she should die. This desire would make my decision morally difficult, to be sure. If I were to believe that I had a duty not to practice euthanasia or assisted suicide, I would need to act against my desires. But I can do that. Intentions can override desires. And one can take this as evidence that desires are not intentions. To see why the reduction of intention to desire is so conceptually problematic, it might help to examine the problem of the ‘package deal’. Most choices made in cases of conflict can be regarded as choices of ‘package deals’. On the theory that intention can be reduced to a predominant desire, one would be forced to say that the choice of such a package deal implies that, on the whole, this ‘package’ is what the agent most desires. Consider again the case above regarding the physician who gets stuck late in the office caring for an acutely needy patient. On the theory that intentions can be reduced to predominant desires, one would say that the agent’s

‘reinventing’ the rule of double effect

127

intention represents a desire that the event (staying late to care for the sick and not playing tennis) should occur. But this would not seem to be the case, as can be seen when one examines the implications of this hypothesis carefully. There are three different ways by which desires for individual events can be related to the desire for the conjunction of these individual events: conjunctions of desires can be aggregative, complementary, or antagonistic.13 Desires for things that are valued for themselves individually (such as a desire for chocolate and a desire for a Nobel Prize) are aggregative. Those that are complementary (such as gin and tonic) are desired in conjunction, but not necessarily separately. Those that are antagonistic (such as a Hawaiian shirt and a tuxedo) may be desired individually, but not in conjunction. When desires are aggregative, these desires distribute, logically, over the conjunction, so that if one desires (A and B), one would desire A and one would desire B. Take an example of a tragic choice like Sophie’s (Styron 1979). Sophie chooses that (child A live and child B die)—tragically. But her desires cannot account for this choice. On the presumption that Sophie desires equally that each child individually should live, the most that the predominant desire theory could say would be that she desires ((that child A live and child B die) or (that child B live and child A die)) more than she desires that (child A die and child B die). But the disjunction is not a choice. If her desire for each child is truly equal, no amount of desire can result in a choice of which child should live and which should die. The predominant desire theory thus reduces to a Buridan case.14 Her predominant desire does not specify a choice. Further, her desire is that both children should live, and this desire is clearly aggregative: each child is valued for its own sake. But if that is the case, and one believes that her choice (that child A live and child B die) is simply explicable as her dominant desire, without any need to invoke intention as a distinct propositional attitude over and above her desires, it would follow (by the logical rule of the distribution of aggregative desire over conjunctions) that Sophie would, other things being equal, desire that child A otherwise should live and also desire that child B otherwise should die. And that is certainly not the case if we presume that she loves both equally. Nor could one suggest that Sophie desires the death of child B to the extent that this helps to fulfill her predominant desire that child A should live. She desires nothing of the sort. Nor does her choice (her fully determinate intention) prove that she desires that child A live more than she desires that child B live. Such accounts conflict violently with our deepest intuitions. 13 The first term is borrowed from economic theory. ‘A complement is a good which tends to be purchased when another good is purchased since it ‘‘complements’’ the first’ (Pearse 1992: 73). The rest of the analysis is based on permutations of this principle. 14 In a famous case, attributed to Jean Buridan but not found in his writings as they are known to us today, an ass starves to death because it is unable to choose between two equally attractive and equidistant bales of hay.

128

daniel p. sulmasy

Choice requires something that desire alone cannot supply—namely, intention. The nature of tragic choice and the moral life requires that intentions not be reduced to desires. Intentions, by contrast, do not logically distribute over conjunctions of events. To intend (A and B) does not imply that one otherwise intends A and otherwise intends B. Sophie’s choice of this conjunction does not imply, logically, that Sophie, other things being equal, otherwise intends that child A should live and otherwise intends that child B should die, any more than my choice to stay in the office and not play tennis implies that, other things being equal, I otherwise would intend to stay in the office and otherwise would not intend to play tennis. Intention cannot be reduced to desire because intention involves a commitment of some sort—a choice. And choices are often hard. They sometimes involve doing what one does not desire. A choice is, as Donagan (1987: 51) would put it, a fully determinate intention. In addition, it seems simply implausible that doing one’s duty should always be doing what one most desires, even for a saint. If a duty is a real duty, it would seem that it should be capable of constraining desire. And this would cast further doubt on the notion that intention can be reduced to desire. Further, as both Searle (1983: 103) and Brand (1984: 124) point out, desires can be contradictory while intentions cannot. For instance, Jack can desire that he should euthanize Thomas and also desire that he not transgress the law by doing so. But presuming that he is fully informed about the law and understands that euthanizing Thomas would break the law of the State of Michigan as it stood in September 1998, then he cannot intend to euthanize Thomas in the State of Michigan in September 1998 without intending to break the law. Brand (1984: 124) also points out that desires admit of degrees, while intentions do not. For instance, one can say of Tim that he greatly desires to help Diane commit suicide, but it makes no sense to say that he greatly intends to help her commit suicide. Intention requires, as Davidson (1980: 101) once put it, an all-out judgment. Intentions do not admit of degrees the way desires can. Intentions are not reducible to desires.

The Belief–Desire Model is Inadequate The most sophisticated model proposed by those who are not convinced that intentions actually exist is the belief–desire model. Such a model suggests that a combination of belief and desire accounts for intention. A good example of this approach is given by Faden and Beauchamp (1986: 242–8). By their account, a surgeon who desires to operate on a patient, believing that it is possible to operate and believing that it is possible that he will scar the patient, intends to operate on the patient and intends to scar the patient. While I suspect most surgeons would

‘reinventing’ the rule of double effect

129

find it perplexing to be told that they intend to scar their patients, for Faden and Beauchamp, nothing more need be said. But no such combination of beliefs and desires will accomplish what neither belief nor desire alone could do—account completely for intention. There are several reasons for this. First, Searle argues that while intentional states may involve belief and desire, intentionality is not reducible to belief and desire. For example, to be annoyed that p, to be remorseful that p, and to be sad that p all imply the belief that p and the desire that not-p. But annoyance, remorse, and sadness are not the same. Likewise, the fact that intentions might involve belief or desire does not imply that intention can be reduced to belief and desire (Searle 1983: 35–6). Second, the arguments made above point out that desire is not merely insufficient, it is in fact unnecessary to account for intention. If belief is insufficient and desire is unnecessary, then no combination of belief and desire will be both necessary and sufficient to account for intention. Third, as Aquinas pointed out long ago (Aquinas 1964: I-II q. 13 a. 6 ad 3), human agents are not paralyzed by equal choices. When one is faced with choices that are equally desirable or undesirable and one believes that both choices are available and can be achieved with equivalent effort and equivalent satisfaction of one’s desire, one is still able to choose between the two courses of action. Intentional choice is something over and above belief and desire. Bratman (1987: 11 ff.) makes the same point by discussing a set of problems known as Buridan cases (see note 14). If one sits before two equally desirable boxes of cornflakes believing that eating either will satisfy one’s desires equally, and believing that one will satisfy one’s desire by eating cornflakes and believing that it is equally likely that either box could be obtained with equal convenience and equal success, has one thereby formed an intention? Does one thereby intend to eat either box of cornflakes? Both boxes of cornflakes? It hardly seems so. Belief and desire do not add up to intention. It is important to note that the foregoing does not mean that one cannot simultaneously desire and intend that the same event should come about, or that one cannot simultaneously believe and intend the same event, or that one cannot simultaneously desire, believe, and intend the same event. Nor should the foregoing be taken as an argument that only intentions matter in morality, and that beliefs and desires do not matter or that consequences do not matter. On the contrary, moral agents can have multiple propositional attitudes towards the events they bring about intentionally, and a robust philosophical moral psychology will have something to say about all of these propositional attitudes. Intention is important, but morality is not limited to intentions. All I have tried to establish is that intentions cannot be reduced to desires and beliefs and that intentions have a special importance in the evaluation of the acts human beings undertake.

130

daniel p. sulmasy

Intentions Are Not Motives Neither should intentions to be confused with motives. A criminal investigation of homicide provides a good example of the difference between intention and motive. When the police look for a motive for a fatal knifing, or for a suspect with a possible motive, they have often already assumed that the act was intentional. They generally presume, that is, that the stabbing was the ascribable condition of fulfillment for the suspect’s intention in acting, and they often believe that the death of the victim by way of the suspect himself stabbing her was the condition of fulfillment of the suspect’s mature complex prior intention. When searching for a motive, however, they want to know what could have initiated or led someone to perform such an intentional act. As Davidson (1980: 105) puts it, a motive is a reason for intending. Gustafsen (1986: 181–2) agrees, pointing out that motives start intentions. However, ordinary language sometimes confuses motives with what I have called further intentions. Anscombe (1969: 20–1) distinguishes backward-looking motives from forward-looking motives. She means by a backward-looking motive that the main ground of the motive is something that happened in the past, or is happening now. By contrast, the main ground of a forward-looking motive is something that has yet to take place. It is forward-looking motives that are most easily confused with intentions. As an example of a backward-looking motive, one might be seeking to understand the motive for a person’s suicide. If the person’s motive were to ‘get even’ with those who had failed to show adequate concern for her suffering—to cause them suffering in return by making them feel guilty—it would be a backward-looking motive. Revenge is something that gets started by something in the past. One can look forward to the day of revenge. And the revenge would be present in the suicide, but it would have started because of past events. The memory of past hurts to the person committing suicide and a desire to respond by returning the hurt would be the complex state of affairs that constitutes the motive. By contrast, a person might commit suicide to prevent some undesired future state of affairs. This would be a forward-looking motive. For example, Janet might fear that as a carrier of the -4 allele of Apolipoprotein E, she is at such high risk for Alzheimer’s disease that at the first signs of memory loss she would rather commit suicide than face further dementia. If this were her motive for committing suicide, it would be forward-looking. While ordinary language can be sloppy in this regard (especially regarding forward-looking motives) a motive is not some further desired event. Death is what Janet intends, but death is not her motive. Her motive might be fear, or embarrassment, or some other psychological state regarding her future. Anscombe also names admiration, spite, and friendship as forward-looking motives, but these attitudes or states of affairs are not events that could function as the objects of genuine intentions. A motive is a psychological state of affairs that helps to explain

‘reinventing’ the rule of double effect

131

how an agent comes to have an intention. Properly speaking, the propositional content of an intention must refer to an event. Intentions and motives are not the same.

Event Identity Thus far, I have explained what intentions are and how they differ from beliefs, desires, and motives. However, merely establishing that intentions are not reducible to beliefs and desires does not suffice for establishing a plausible distinction between the foreseen and the intended. Certain beliefs must necessarily be ascribable to the agents that undertake acts. Those beliefs must be coherent and plausible if the description of the agent’s intentional state is to be coherent and plausible. Intentions, by definition, explain the events brought about by agency. So, while intention cannot be reduced to beliefs, incoherence or implausibility in an agent’s beliefs about the events that she brings about can render her explanation of those events as having come about under a particular intentional description incoherent or implausible. Critical to any proposed account of the distinction between the intended and the foreseen are beliefs regarding descriptions of events. Especially crucial is the question of whether two or more particular descriptions describe the same event or different events. With this in mind, I will propose a definition of event identity. This definition is modified somewhat from Myles Brand’s work. I will suggest that, formally, Definition 9: Event identity. If ‘x’ and ‘y’ are descriptions that pick out events, then if x and y occupy the exact same spatiotemporal coordinates, ‘x’ and ‘y’ pick out the same event. This means, for instance, that if x is an event (described, say, as ‘Oedipus marries Jocasta’) and y is an event (described, say, as ‘Oedipus marries his mother’) and both of these events occupy the same spatiotemporal coordinates, then they are the same event. ‘x’ and ‘y’ may be two true definite descriptions, but both descriptions pick out the same event.

Conditions of Plausibility and Coherence: Distinguishing the Intended from the Foreseen Armed with the philosophy of action I have just outlined, I can now specify a set of conditions that will permit one to test whether any particular report of a distinction between the intended and the foreseen is plausible and coherent. This, I believe, has always been the purpose of the causation requirement of the classical formulation of the RDE (Requirement 3 in Mangan’s formulation). In fact, if one is rational

132

daniel p. sulmasy

and sincere, the requirement that the unintended bad effect not cause the intended good effect adds nothing to the classical requirement that one must foresee but not intend the bad effect. Thus, I suspect that the purpose of the causation requirement in the classical formulation of the RDE has been to render plausible and coherent reports about what is foreseen but not intended, so that one would not deceive oneself or others in reporting on one’s beliefs and intentions. However, one may question whether the causal requirement is sufficient to quell doubts about the rationality and sincerity of reports about what someone claims to foresee but not to intend. I believe that exclusive reliance upon this ‘test’ has led to significant errors in the application of the RDE. There are quite a number of other ways in which one can, in reporting on what one believes and what one intends, be insincere or irrational. The purpose of Requirement 7 of my ‘reinvented’ RDE is to expand the set of tests for the coherence and plausibility of reports about what is foreseen but not intended. The correct application of this requirement of the Reformulated RDE must treat the object of an intention as a proposition regarding an event. Events, as I have discussed, are best understood as particulars. As particulars, events will have logical implications and alternative true definite descriptions, and will have the potential to cause other events. Double effect cases are typically quite complex. Often, when an agent forms a prior intention to bring about some event, there are many different implications, alternative true definite descriptions, and cause and effect relationships connected to that event. Accordingly, one must consider, at the very least, the following three types of scenario, each specifying a different set of relationships between intention and causation: linear causal chains, causal forks, and alternative causal routes: Scenario I: Linear Causal Chains (p causes r) & (r causes q) p

r

q

e.g. I intend to drink alcohol (p) in order to relax (r), but I do not intend that this experience should make me an alcoholic (q). Scenario II: Causal Forks (p causes q) & (p causes r) q

p

r

‘reinventing’ the rule of double effect

133

e.g. I intend to prescribe penicillin (p), in order to cure strep throat (r), but I do not intend to cause an anaphylactic reaction (q). Scenario III: Alternative Causal Routes (p causes q)& (p causes r) & (q causes r) q

p

r

e.g. I intend to perform a bone marrow transplant (p) in order to treat leukemia (r), but not intending to cause high-grade graft v. host disease (q), even though high grade graft v. host disease also has a profound anti-leukemic effect (r).15 Having described these causal scenarios, I can now set forth the list of conditions I propose for plausibly and coherently maintaining that one foresees but does not intend a particular upshot of one’s intentional act. I will set these forth without argument, since it seems to me that once one understands what is meant by intention and event identity, these conditions are self-evidently true.

Conditions for Rationally Distinguishing Between Intended and Foreseen Events ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

If an Agent (S) believes that the causal situation is such that it can be illustrated by any of the Scenarios (I–III), it is plausible and coherent for S to maintain that he or she foresees two results (q and r) from a chosen means of acting (p), but only intends r, given the following conditions: (C1) (C2) (C3) (C4) (C5)

p and q are not alternative true definite descriptions of the same event; q and r are not alternative true definite descriptions of the same event; q is not wholly spatiotemporally contained within p; q is not wholly spatiotemporally contained within r; p does not logically entail q;

15 Currently, clinicians are actually aiming at (intending) mild graft v. host disease (GVHD) because of its anti-leukemic effect, but no one aims at high grade GVHD.

134 (C6) (C7) (C8) (C9)

daniel p. sulmasy r does not logically entail q; q is not S’s further intention; q is not an empirically necessary causal condition for r; S does not intend r by way of q.

If conditions C1 through C9 are met, it would be coherent and plausible to claim to intend r, but not q, by means of p, in any of the three scenarios, as the examples amply demonstrate. This is not to say that any such claim would be a true and sincere report of the agent’s intentions and plans. Such a claim could be rebutted. But that rebuttal would require further inquiry and investigation of the facts and reported beliefs and intentions of the agent. Further, this analysis does not consider whether the agent actually has the requisite knowledge (or the duty to obtain such knowledge) in order to have true and correct beliefs about these events and their casual relations and logical implications. Some ignorance is culpable, and some excuses. For example, it matters whether one thinks that Oedipus had a responsibility to know that Jocasta was his mother. In addition, it must be noted emphatically that having a plausible and coherent report of what one foresees but does not intend is not sufficient to justify the act from a moral point of view. It is necessary that the agent also satisfy the other requirements of the Reformulated RDE. Finally, while these conditions are proposed as necessary, I am making no claim that they exhaust all the conditions for rendering a coherent and plausible account of a complex prior intention in which it is claimed that at least one effect of an act is foreseen but not intended. This analysis only presents a first defense of a tightened set of conditions for the coherence and plausibility of distinguishing between the foreseen and the intended. This analysis requires some further explication. First, one should note that each of these conditions corresponds to a matching condition under Requirement 7 of the Reformulated Rule of Double Effect, as it was laid out above. This more formal rendering of these conditions makes use of the technical vocabulary in philosophy of mind that I have presented. Second, this set of conditions can be viewed as a significant amplification of the third requirement of the classical RDE. I believe that the major purpose of this classical ‘causal’ requirement (that the purportedly unintended bad effect cannot be the cause of the intended good effect) was to provide a check on the sincerity of the agent—to provide at least one condition for a report on beliefs and intentions that was coherent and plausible. However, this was simply insufficient. There are other ways to be incoherent and implausible in one’s reports on what one foresees but does not intend. Thus, the purpose of these conditions is to ‘tighten the rules’ by setting up further tests for the coherence and plausibility of such reports.

‘reinventing’ the rule of double effect

135

It will doubtless prove helpful to the reader if I describe these conditions in words and give examples. In words, p is the Agent’s (S’s) intended means—the act itself. q is the purportedly foreseen but unintended causal result of p, and r is the intended causal result of p. Condition 1 states that the purportedly unintended result (q) cannot be an alternative true definite description of the intended means (p). Condition 2 states that the purportedly unintended result (q) cannot be an alternative true definite description of the intended result (r). Condition 3 states that the purportedly unintended result (q) cannot be an event that is completely subsumed within a wider event that constitutes the intended means (p). Condition 4 states that the purportedly unintended result (q) cannot be wholly spatiotemporally contained within a wider event that constitutes the intended result (r). Condition 5 states that the purportedly unintended result (q) cannot be logically implied by the intended means (p). Condition 6 states that the purportedly unintended result (q) cannot be logically implied by the intended outcome (r). Condition 7 states that the purportedly unintended result (q) cannot be, in fact, the agent’s actual further intention. Condition 8 states that the purportedly unintended result (q) cannot be an empirically necessary causal condition for the intended result (r). And Condition 9 states that it cannot be the case that the agent actually intends the intended result (r) by way of the purportedly unintended result (q).

Intention and Foresight: Electroconvulsive Therapy As an example, consider a physician who is contemplating the use of electroconvulsive therapy (ECT) for a patient suffering from a severe manic state, in the setting of bipolar affective disorder, refractory to medical treatment. According to C1, it would not be coherent and plausible for the physician to claim to use ECT as his means (p) of treating mania (r), and also claim that he did not intend to subject the patient to electro-shock psychotherapy (q), since these are two alternative definite descriptions of the same procedure. According to C2, it would not be coherent and plausible for the physician to claim to use ECT as his means of treating mania (p), intending as a result of ECT to produce a seizure (r), but not intending to produce an epileptic fit (q), since ‘seizure’ and ‘eplileptic fit’ pick out the same event. According to C3, it would not be coherent and plausible for the physician to claim that if, in the course of ECT, he were to draw the patient’s blood (p) as a means of monitoring the patient’s drug levels, he only intended thereby to measure lithium levels (r), but did not intend to remove any red blood cells (q), because the removal of red blood cells (q) is an event wholly spatiotemporally contained within the event he intends as his means—namely, the drawing of the patient’s blood (p). According to C4, it would not be coherent and plausible for the physician to claim to use electrical current as his means (p) of achieving his intended aim of

136

daniel p. sulmasy

producing a seizure sufficient to cause motor movement of a non-paralyzed limb (r), while not intending to cause synchronous electrical discharges in the patient’s brain (q), since the event of synchronous electrical discharges in the brain (q) is an event wholly contained within the wider event that he intends as his aim—namely, a seizure (p). According to C5, it would not be coherent and plausible for the physician to claim to use a ten-treatment course of ECT as his means (p) of treating mania, but claim that he did not intend to give a multi-treatment course (q), since a ten-treatment course (p) logically implies a multi-treatment course (q). According to C6, it would not be coherent and plausible for the physician to claim to be using ECT as his means (p) of treating mania, intending the complete remission of the manic episode (r), but not intending to end the patient’s decreased sleep requirement (q), because a decreased sleep requirement (q) is a defining characteristic of mania, and the complete remission of mania (r) logically entails ending the decreased sleep requirement. According to C7, it would not be coherent and plausible for the physician to claim to use ECT as his means (p), intending only to produce seizures (r), but not intending to cause memory loss (q), if it were the case that (for some nefarious motive) his actual further intention were to produce permanent memory loss in the patient (q) by means of ECT. According to C8, it would not be coherent and plausible for the physician to claim to use ECT as a means (p) of bringing the patient’s mania into remission (r), but not to cause any ongoing electrochemical changes in the brain of the patient (q), when the production of ongoing electrochemical changes in the brain of the patient (q) is an empirically necessary causal condition for the remission of the patient’s mania (r). According to C9, it would not be coherent and plausible for the physician to claim to use ECT as a means (p) of bringing about a remission of mania (r) but not to intend to cause seizures (q), when the physician’s plan was to produce ten seizures of approximately one minute’s duration each over the course of two weeks (q) as a way of bringing about the relief of the patient’s mania (r). However, it would be coherent and plausible for the physician to claim to intend to perform ECT (p) as a means of relieving mania (r), while foreseeing but not intending permanent memory loss (q), under Scenario II (Causal Forks). It would also be plausible and coherent to make this claim under Scenario III (Alternative Causal Routes).16 The Alternative Causal Routes Scenario is especially interesting to consider in discussing ECT because it was once thought (although the view is now largely discredited) that the relief of psychiatric symptoms through ECT was at least in part caused by memory loss. Even were this to be true, it would still not be 16 Since the relief of mania (r) does not cause memory loss (q), Casual Scenario I (Causal Chains) would not seem to apply here.

‘reinventing’ the rule of double effect

137

implausible or incoherent to claim to intend to relieve the mania without causing memory loss (unless memory loss were known to be an empirically necessary causal condition for the relief of psychiatric symptoms by ECT).

The RDE and the Concept of a Side-Effect The distinction between the intended and the foreseen is of critical importance to any account of the practice of medicine. In fact, it is the structural basis of the concept of a side-effect. Side-effects (such as memory loss) are not merely unwanted. They are unintended, even if foreseen. To see why this is so, consider the fact that, occasionally, side-effects are foreseen, unintended, and yet desirable. For example, consider the case of a depressed and anxious patient who smokes, has difficulty sleeping, and has a chronic seizure disorder. Suppose the patient’s anxiety and depression have not responded to fluoxetine (Prozac). The clinician might then face a choice between bupropion and nortriptyline as the next drug to try. She quickly realizes that bupropion has a side-effect (unintended and very undesirable) of lowering the seizure threshold, and so chooses nortriptyline. In this case she intends to treat the anxiety and depression and to avoid seizures. The clinician might note that it would have been nice if she would have been able to use bupropion since it would also have helped the patient to quit smoking. Yet she is also aware that notriptyline is a second-line drug for smoking cessation (Hughes et al. 2003). In this case smoking cessation is foreseen as a potential beneficial and desirable side-effect, but it is not intended, since if it were the intention of the treatment, bupropion would have been the first choice. If, two weeks later, the patient returns and has quit smoking in addition to having improved anxiety and depression, the clinician and the patient would both have achieved a desirable, but unintended, side-effect. It would be coherent and plausible for a physician to say, ‘I did not choose to prescribe bupropion with the intention in so acting of causing the patient to quit smoking, even though I was able to foresee that this unintended effect was possible, and I am glad that it has happened’.

Application to Cases ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

The foregoing discussion of the RDE, as I have reformulated it, is significant for the debate about the place of the RDE in the philosophy of medicine and in medical ethics. The additional conditions I have proposed for judging the plausibility and coherence of a claim that a particular upshot of a proposed act is foreseen but not intended results in a reclassification of controversial cases that should affect this debate in important ways. Cases that have been used routinely by opponents

138

daniel p. sulmasy

of the RDE to show that the distinction between the foreseen and the intended makes no sense could no longer be used to discredit the RDE because these are not examples of its correct application. However, if the RDE were no longer a legitimate justification for the acts described in these cases, proponents of the RDE would either be required to consider these cases immoral or be forced to find some other means of showing that these acts are justifiable. While space limitations will not permit a full discussion of the implications of this reformulation of the RDE for many contested cases, an examination of its application to the tubal ectopic pregnancy case and the morphine case will readily illustrate the clinical and ethical implications of the Reformulated RDE.

The Tubal Ectopic Pregnancy Case Those who are morally opposed to abortion, but would defend the permissibility of surgery for tubal ectopic pregnancy as an application of the RDE, must assume that the intention of the operator is not to abort the fetus but to remove the diseased Fallopian tube. They suggest that one can hold that direct abortion is morally wrong in all circumstances yet claim that in operating under such conditions, foreseeing but not intending the death of the fetus, the abortion is indirect and therefore morally permissible (Griese 1987: 268–71). Is this a coherent claim according to the above analysis? Initially it would not seem that the case could be made that the removal of the diseased tube and the abortion of the fetus are two distinct events. Many would suggest that since the removal of the fetus and the removal of the tube require the same motor movements and occupy the same space and occur at exactly the same time, these two descriptions pick out the same event and are therefore nothing more than alternative definite descriptions of that event. If so, it would be incoherent, according to my analysis, to claim that one foresaw but did not intend the abortion of the fetus. However, the proponent of the applicability of the RDE to this case could respond that the fetus is contained within the Fallopian tube and is therefore, in fact, spatially distinct from the tube. Such a person might suggest that this justifies the intention–foresight distinction in this case. But one must question whether this claim is plausible. Suppose the proponent were to argue that this case involved a Causal Forks Scenario (Scenario II). The event of the operation (p) causes one good effect (r, the intended removal of the diseased tube), and also causes one bad effect (q, the unintended removal of the fetus). However, it would seem that this claim would be implausible because Condition 2 of Requirement 7 would still be violated. It would still seem that ‘q’ and ‘r’ pick out the same event. The diseased tube is diseased because it contains a fetus. The fact that it contains a fetus is what is wrong with it. Granted, the fetus and the tube are spatially distinct things. But

‘reinventing’ the rule of double effect

139

the only good event one could suggest that one hoped to achieve in such a case would be the removal of a diseased tube. In the case of tubal ectopic pregnancy, ‘the removal of the fetus’ and ‘the removal of the diseased tube’ seem to pick out the same event. So, even if one were to claim to foresee but not intend the removal of the fetus while intending the removal of the tube, it does not seem plausible to suggest that one intends the removal of a diseased tube but not the removal of the fetus. In other words, this is not a Causal Forks scenario. Further, even if one were to grant the (implausible) suggestion that ‘the removal of the diseased tube’ and ‘the removal of the fetus’ pick out different events, one would still be left to contend with the question of whether this procedure violates Condition 4. It seems that the event of removing the fetus is wholly spatiotemporally contained within the event of removing the diseased tube. No part of the event of removing the fetus would be spatiotemporally outside the event of removing the diseased tube unless one were to perform a salpingotomy (i.e. an incision in the tube) and were simply to remove the fetus from the tube. But then it would clearly be a directly intended abortion and the RDE could not apply. So, even if one were to grant that this construal of the act would survive scrutiny according to Condition 2, it would not survive scrutiny according to Condition 4. In fact, with methotrexate treatment, one may now remove the disorder from the tube without damaging the tube (Lipscomb et al. 1999). This may be safer for the mother since it is a medical, not a surgical, technique. Therefore, those who hold that the RDE justifies the surgery for tubal ectopic pregnancy are now in the awkward position of either (1) recommending the possibly more dangerous surgical procedure to the mother, or (2) claiming that the aim of intervention with methotrexate is to remove the diseased trophoblastic tissue that connects the fetus to the wall of the Fallopian tube, not to remove the fetus (Clark 2000). It should be clear that the latter construal would not stand up to scrutiny according to the conditions I have delineated for distinguishing the foreseen from the intended any more than for the surgical case. Many traditional defenders of the application of the RDE to the tubal ectopic pregnancy case have concluded that it is implausible and incoherent to claim that one intends only to remove the trophoblastic tissue, not to cause an abortion. Yet they fail to see that it is equally implausible and incoherent to say that one is merely removing a diseased tube, not the fetus inside it. This has led some to recommend the more invasive, expensive, and dangerous surgical approach, which is an awkward position to defend. It therefore seems better to give up on the idea that it is the RDE that justifies treatment in cases of tubal ectopic pregnancy, no matter what treatment modality one uses. An alternative move for the proponent of the application of the RDE might be to suggest that the tubal ectopic pregnancy case falls under the broad rubric of a type III scenario, not a type II scenario. It might be argued that what is intended (p) is better described more narrowly as the removal of the tube, rather than broadly as the performing of an operation for tubal ectopic pregnancy. The good one intends

140

daniel p. sulmasy

by this means is the health of the mother (r). The bad one foresees but does not intend is the removal of the fetus (q). One might also acknowledge fully that the event of the removal of the fetus (q) is an alternative causal route to the event of the health of the mother (r), but not the route one intends. But is this a coherent and plausible claim? The answer would again be no because it violates Condition 8 of Requirement 7 of the Reformulated RDE. The event of the removal of the fetus is a necessary causal condition for the event of the health of the mother. With current surgical techniques, the fetus cannot be left inside the mother to develop normally.17 And given current surgical techniques, if one did leave the fetus floating freely inside the abdominal cavity, or even partially attached to the abdominal wall, it would still constitute a grave threat to the life of the mother and the likely result would be two deaths instead of one. So, under a Scenario III interpretation, a plan to keep the fetus inside the mother would be inconsistent with one’s stated good intentions—to save the life of the mother. And this would seem quite irrational. Therefore, an attempt to argue that one foresees but does not intend the death of the fetus in operations for ectopic pregnancies is implausible and incoherent. A proponent might at this point concede that the removal of the fetus is within the scope of the intention, and offer one final move. The proponent might try to redescribe the act as the removal of the tube and the fetus, with the specific intention in acting of saving the life of the mother, while foreseeing but not intending the death of the fetus. However, this leads to a even more serious conundrum, for if this argument were acceptable for the ectopic pregnancy case, it would be acceptable for many, many other cases. Every act of abortion could thereby be justified by the same sort of act redescription: intending only to remove the fetus, not intending to kill it. One must bear in mind that the proponent of the use of the RDE to justify surgery for tubal ectopic pregnancies was initially led to use this argument because of opposition to abortion. It would be devastating to concede that the argument justifying the application of the RDE to the tubal ectopic pregnancy case could also justify abortion on demand. Operations for tubal ectopic pregnancies would therefore appear to be cases of killing the fetus.18 If such operations are justifiable, as almost everyone would agree that they are, and one is otherwise morally opposed to abortion, such operations must be justified some another way than by invoking the RDE. Space limitations (in an already very long chapter) do not allow a full discussion of the alternative.19 17 This is true at the present level of technology. However, if progress ever allowed physicians to do so safely, then one would presumably be morally impelled to reimplant the conceptus elsewhere (such as in the uterus) and there would be no point to saying that one did not intend the death of the fetus because it would be an avoidable event, and would violate Requirement 4 of the Reformulated RDE. 18 For a more complete definition of killing, see Sulmasy (1998). 19 For a more extensive discussion, see Sulmasy (1995).

‘reinventing’ the rule of double effect

141

But briefly stated, rather than its traditional formulation, the moral rule prohibiting killing, as I see it, ought best to be formulated as follows.

Prohibited Killing Acting on individual authority, one may never make the death of a human being the condition of fulfillment of one’s intention in acting, except in forced choice situations in which one cannot act intentionally without making the death of a human being the condition of fulfillment of one’s intention in acting. Classically, these are cases of self-defense or rescue, where self-defense and rescue are narrowly defined to include only direct, serious threats of death (and no lesser harms). In such cases, the threatened person (or the rescuer of the threatened person) is permitted to use lethal force against the attacker provided all other means have been exhausted, the threat is truly imminent, and it is truly direct (i.e. the attacker is the agent of the threat, not some other party such as a hired assassin or a subordinate ordered to kill). In these cases, one is not culpable for acting directly against human life, since one cannot act without taking human life. One is not acting against the value of life in such forced choice situations, because no choice is available except one that involves taking life. One either kills or is killed. In these cases, one may employ other considerations (such as number of lives involved, role responsibilities, or the like) in choosing one life or the other. In removing a tubal ectopic pregnancy, one is killing, but since one is doing so to rescue the mother from a serious imminent threat of death, caused by the fetus, despite its innocence, with no intervening agents, this is an allowable exception. When either one or the other must die (or, more specifically in such cases, either one dies or both die), one is allowed to choose, by action or omission, who will survive and who will die. Thus, vigorous defenders of the value of human life can justify surgery or methotrexate in cases such as tubal ectopic pregnancy without diminishing their allegiance to a principle of equal respect for the value of each human life, prenatal or postnatal. But they also ought not invoke the RDE in such an incoherent way that the significance of this important moral rule is jeopardized.

The Morphine Case In contrast to the way in which my Reformulated RDE excludes the tubal ectopic pregnancy cases from its ambit, the Reformulated RDE continues to justify its application in most traditional cases. To illustrate this, I will ‘walk through’ its application to the morphine case. In the morphine case, proponents of the RDE suggest that in giving powerful opioid analgesics such as morphine to dying patients, one intends only pain relief,

142

daniel p. sulmasy

not the hastening of death that one knows can occur through respiratory depression. According to my Reformulated RDE, such a combination of intentions is coherent, plausible, and morally justifiable. Although in such a case it might not seem immediately obvious, there are, in fact, two separable events, distinct in time and space: pain relief (intended) and respiratory depression (unintended). To see why these really are two distinct events, making the application of the RDE plausible, it is perhaps best to think about this case on a molecular level. The analgesic and respiratory depressant effects of morphine occur by the binding of morphine molecules at different subtypes of morphine receptors, populating different locations in the nervous system. The chemistry for each effect has a different time course (kinetics). Morphine achieves pain relief via µ1 receptors and respiratory depression via µ2 receptors. These molecular differences are manifested in the response of the patient to the drug. Pain relief occurs at lower doses and more rapidly than respiratory depression. Thus, while the effects are scattered throughout the body, conceptually this is still a Causal Forks Scenario (type II). So the claim that one intended pain relief and not respiratory depression is plausible and coherent. However, since the event of death is also a potential cause of the event of pain relief, the skeptic might suggest that this could be construed as a type III scenario—an Alternative Causal Routes Scenario. Accordingly, one must ask whether the morphine case, so construed, would meet the conditions of Requirement 7 of the Reformulated RDE. I argue that even under this construal, the distinction between the foreseen and the intended would be plausible and coherent. ‘Death’ and ‘administering morphine’ do not pick out the same event. And neither do ‘analgesia’ and ‘death’ pick out the same event. The intention to administer morphine does not logically imply intending the death of the patient. Intending analgesia does not logically imply intending the death of the patient. The event of the death of the patient is certainly not a necessary causal condition for the event of pain relief. If it is not the doctor’s further intention that the patient should die, and it is not the case that he intends analgesia by way of the patient’s death, then the claim is at least plausible and coherent, and meets all of the conditions of Requirement 7 of the Reformulated RDE. For the morphine case to be permitted, however, the remaining requirements of the Reformulated RDE must be met. This analysis turns out to be quite straightforward. A physician morally opposed to euthanasia has a conflict between a duty to relieve suffering and a duty not to kill (i.e. not to act with the specific intention in acting that a human being should die by way of that act except in cases of self-defense or rescue). In such cases, she may administer the morphine with the following provisos: (1) the administration of morphine is one intentional act with the two foreseeable effects of pain relief and respiratory depression;

‘reinventing’ the rule of double effect

143

(2) injecting morphine is at least morally neutral; (3) relieving pain is morally good and killing is morally bad; (4) other means of pain relief such as non-steroidal anti-inflammatory drugs are no longer effective; (5) no intervening agents are involved; (6) she has the intention in acting of relieving pain and not depressing respirations and thereby hastening death; (7) it is plausible and coherent for her to make this claim (as demonstrated by the above analysis), so one cannot rule out that she is sincere and rational in her report on her intentions; (8) there is due proportionality, such as when the patient’s death is inevitable and the pain is great, so that fears of hastening the patient’s death seem overwhelmingly small when proportionately compared to the benefits of pain relief. The act is then morally justified according to the RDE. While space considerations prohibit analyzing more cases, the upshot is that under the Reformulated RDE, the tubal ectopic pregnancy, the cancerous gravid uterus, and the case of using lethal force in self-defense, all fall out of double effect analysis. Cases such as morphine for the dying, and almost any other clinical case that could be described as a treatment with a side-effect, fall within the scope of the Reformulated RDE.

Consideration of Possible Objections ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

This is a considerable reworking of the RDE, and I can anticipate that objections will be raised. Here I consider seven objections, and will leave discussion of other potential objections to the thoughtful analysis of colleagues. First, some might wonder if the RDE has been rendered superfluous. If a case such as the tubal ectopic pregnancy case must be justified by some other moral rule, perhaps the moral rule about killing that I offered is sufficient, and the RDE unnecessary. Those who have suspected that the RDE is merely correlated with other moral intuitions that actually do all the necessary moral work might be tempted to think this way. However, this is not the case. A rule justifying killing only in self-defense or rescue would not cover the morphine case. The morphine case involves only one life, not a forced choice between at least two. So a rule that justifies direct killing only in cases of self-defense or rescue would not apply. Nor would it do to recast the morphine case as the ‘rescue’ of the person who is dying. While clever, such a move would implicate one in an untenable dualism, claiming that the person was being threatened by some ‘other person’ (perhaps the

144

daniel p. sulmasy

person’s own body). And, clearly, the threat would not be a threat of death, but a threat of pain. Similarly, the new rule prohibiting killing would not apply to a whole range of cases in which death is a side-effect, such as bone marrow transplantation for leukemia, in which the foreseeable risk of death is high, but not intended, but there is no ‘other party’ threatening death. Even cases such as the strategic bomber case would also still require the RDE, since the threat of harm posed to the bomber, her comrades, and her nation by the armaments factory (the intended target) would presumably not be imminent, and the schoolchildren next door (the unintended target) are not themselves the source of any threat. So, the RDE would be far from superfluous under my proposed reformulation. Second, some might object that this reformulation is too complicated. Surely, with so many conditions and qualifications, one might suggest, there must be something suspicious about this theory. My argument against this is to remind readers that the purpose of ethical theory is to explain and justify morality. This is different from being a moral person or, more specifically, a moral clinician, and different from acting in a morally upright manner. If I am correct that the RDE is the theory that explains the concept of a side-effect, then clinicians act according to the RDE almost reflexively every day. Explaining what one means by a side-effect in a rigorous philosophical manner is a far more complex matter than having a common-sense working notion of a side-effect in clinical practice. One resorts to theoretical analysis if someone challenges the meaning of a common-sense notion such as the notion of a side-effect, or if someone objects to the application or plausibility of a moral rule such as the RDE with respect to a particularly difficult case. Similar considerations apply to all sciences. Human bodies work according to the laws of biochemistry. One need not know all the theories of biochemistry in order to be a good doctor. But in exceptional cases one resorts to biochemistry to meet a particular clinical challenge or one explores biochemistry as a basic science in order to explain better what is common. No medical student would object to being taught about topoisomerases or enzyme kinetics ‘because it is too hard’. It is puzzling that anyone seriously interested in medical ethics would feel entitled to complain that medical ethics ‘is too hard’. This brings to mind Bernard Williams’s (Williams and Smart 1973: 149) critique of utilitarianism as having, in its reduction of all of morality to the simplicity of consequences, ‘too few thoughts and feelings to match the world as it really is’. If explaining biochemistry is complex, it seems reasonable to expect that the moral life of the human beings who arise out of these biochemical processes would be at least as complex. A third anticipated objection is that this is all a religious theory that has no place in secular philosophical analysis. This objection is often repeated, or at least intimated, in the bioethics literature. However, it is also a puzzling objection. The RDE is a substantial topic of philosophical discourse. Nothing in my argument

‘reinventing’ the rule of double effect

145

depends upon acceptance of any scripture, the teaching of any church, pope, or bishop, or the dogma of any religion. That it has its origin with the writings of a saint is no argument that it is an inherently religious notion, any more than one can argue legitimately that all theoretical defenses of the immorality of adultery are inherently religious simply because one of the Ten Commandments prohibits adultery. It is a form of the ad hominem fallacy to attempt to detract from an argument merely by citing what one considers its (arguably!) suspect source. In fact, a fourth anticipated objection is actually an additional counterargument to the third. Adherents to the traditional Roman Catholic formulation of the RDE might well object that the reformulation I am proposing is too radical a break with the Catholic tradition on double effect. I have suggested that the intention–foresight distinction, as it originated in the writings of St Thomas Aquinas, was misapplied from the beginning. While this might worry some Catholics, it also ought to suggest that the origins of the idea (whether Catholic or not) are irrelevant to my analysis and therefore no objection to it. I do think that the idea of the intention–foresight distinction is of critical importance, and that we can credit Aquinas with first presenting it, even if I do not agree that it applies to the case to which he first applied it. I think this distinction remains crucial to Catholic moral thinking, and that contemporary philosophy of mind helps to clarify and support it. But if the RDE has been misapplied in Catholic thinking, that mistake needs to be corrected within Catholicism as well as in philosophy. And I hope that Catholic proponents of the RDE will be able to see how my analysis strengthens their own defense of this important moral rule. A fifth anticipated objection is a variation on the fourth. Proponents of the RDE might accuse me of having so radically reformulated it that, in a parody of a phrase associated with the Vietnam War (and the phrase itself is a dark parody of sorts), ‘I had to destroy the RDE in order to save it’. There is no question that I have significantly restricted the use of the RDE by tightening the conditions for its rational application. But rendering a moral rule more rational can hardly be considered its destruction. Other attempts to reformulate this rule have changed its structure and moral tenor considerably more than mine. My reformulation simply makes the traditional formulation more stringent. If, as I believe is true, most of the objections to the RDE are raised by those who have found untenable the applications of the RDE that I also find untenable, then the proper metaphor is not that I have engaged in scorched earth warfare but that I have instead been pruning the vine. By pruning away branches of the RDE that are already dead, it is my hope that I will actually have breathed new life into this important moral rule and that, in its proper application, it will blossom. A sixth anticipated objection is that the RDE, with its emphasis on intention, is not morally relevant. It is true, as I stated at the beginning of this chapter, that if one does not believe that intentions matter in the moral evaluation of human acts, the RDE will make no sense. Many, like Bennett, are skeptical about whether

146

daniel p. sulmasy

intentions actually exist. But these critics often, like Bennett, beg the question in a very serious way. Bennett (1995: 194) states, ‘What a person intends in φ ing is defined, therefore, by which of her consequential beliefs explain her φ ing and by which of her desires do so.’ Notice that he says ‘consequential beliefs’ and not ‘consequences’. If he defines intentions in terms of beliefs and/or desires, it is simply vacuous for him to argue that intentions can be reduced to beliefs and/or desires. In yet another curious example of one of the two major mistakes I have tried to correct in this chapter, namely that of paying insufficient attention to developments in philosophy of mind, Bennett cites Searle, but then proceeds to ignore everything Searle has to say about why intentions cannot be reduced to beliefs or desires. Bennett offers no reply to any of the arguments to the contrary noted above, such as Buridan cases, wayward causal chains, or the differences in the logical properties of beliefs, desires, and intentions. He also makes no mention of self-referential causation. Therefore, nothing he says should cast any doubt on the theory of intention I have presented. Still, to accept the RDE, even if one concedes that intentions exist, one must be convinced that intentions mark a moral difference that cannot be accounted for purely by consequences. Space does not permit a full discussion here, but intentions make acts human and give acts a moral structure that must be evaluated if acts are to be evaluated as human and moral. At least two simple, initial lines of argument are available against the consequentialist alternative. First, one can point out that anyone who would dismiss the moral importance of how medical events come about, basing moral judgments solely upon the consequences, faces a monumental task. She must show that a formidable history of collective moral judgment has been ill-considered. Physicians, the US Supreme Court, and most plain persons continue to hold a belief that there is a moral difference between injecting 150 mEq of potassium chloride into the right ventricle of a dying patient and giving that same patient an injection of a dose of morphine sufficient to relieve the patient’s pain, and acting with that intention (Washington v. Glucksberg 1997; Vacco v. Quill 1997). The consequences of these two acts may be the same. I have argued elsewhere that the basis of the judgment that these acts are different depends upon intentions (Sulmasy 1998). It will not suffice, in the face of all that I have outlined about contemporary work in philosophy of action, simply to insist that intentions do not exist or to ‘prove’, via question-begging arguments, that everyone who believes these cases are morally different is misguided. One can also offer counterexamples. I will offer a variation on one first proposed by Philippa Foot (1978). Suppose a physician were to have a limited supply of a life-saving drug and were suddenly to face five patients who needed it. Four of these patients would be highly likely to survive with a dose equivalent to 1/5 of the supply of the drug, while the fifth would be likely to survive only if given the total dose. If one were to give each patient a dose equivalent to 1/5 of the supply, foreseeing but not intending the death of the patient who needed a higher dose, one would

‘reinventing’ the rule of double effect

147

fulfill all of the conditions of the Reformulated RDE and the act would be morally permissible. However, suppose one were to face a situation in which there were no drug supply, but the drug could be made by sacrificing one of the five, grinding him up, and making a life-saving serum out of him—enough to save the other four. This plan would not meet the conditions of the Reformulated RDE, since it would violate Requirements 7(h) and 7(i). One could not plausibly and coherently claim that one did not intend the death of the individual sacrificed to make the serum. Only the crassest of utilitarians would see these cases as morally equivalent—always making the best of a bad lot. What marks the moral difference that people of reason and goodwill sense here is not the net outcome (which is exactly the same in both cases). The difference can be ascribed to the fact that in the latter case the death was intended as part of the doctor’s plan, while the death in the former case was foreseen but not intended. Space limitations will not permit me to take up this debate in greater detail than I already have. Suffice it to say that the RDE matters in morality to the extent that intentions matter in morality. Seventh, some objections might be based upon particular moral judgments about the cases that I discussed. For example, some might hold that euthanasia is not immoral, and therefore hold that killing a dying patient is not bad, and therefore conclude that the RDE does not apply to the morphine case. However, this sort of objection is not an argument against the RDE itself. Discussion would need to shift to arguments for and against the morality of acting with the specific intention in acting of making a human being dead by way of one’s act in order to relieve that person’s suffering. Even so, such a critic might not object to the role of intentions in morality in general and thus be persuaded by my reformulation of the RDE with respect to its application to cases in which it could be granted that the unintended side-effect actually would be bad, such as the foreseeable and uncontroversial badness of the hair loss that often accompanies cancer chemotherapy.

Conclusion ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

I have presented a reformulation of the RDE that addresses what I consider two fundamental mistakes: the failure of ethicists (particularly bioethicists) to consider developments in the philosophy of mind in relation to the RDE and the misapplication of the RDE to a class of cases that has proven contentious. Critical to this reformulation of the RDE has been the introduction of various notions from philosophy of mind in order to tighten the conditions under which one can be considered plausible and coherent in a claim to foresee a certain bad outcome while not intending it. This reformulation precludes the cases that have seemed most incredible to critics, thereby lessening their objections. This reformulation

148

daniel p. sulmasy

also forces proponents of the RDE to reconsider the basis by which they can justify the acts that they mistakenly have thought to be justified by the RDE. Elizabeth Anscombe (1970) once observed that the denial of the RDE ‘has been the corruption of non-Catholic thought, and its abuse the corruption of Catholic thought’. It is my hope that this chapter has helped to expurgate a little bit of that corruption.

References Anscombe, G. E. M. (1958), ‘Modern Moral Philosophy’, Philosophy, 33: 1–19. (1969), Intention (2nd edn. Ithaca, NY: Cornell University Press). (1970), ‘War and Murder’, in R. A. Wasserstrom (ed.), War and Morality (Belmont, Calif.: Wadsworth), 42–53. Aquinas, T. (1964), Summa Theologiae, ed. T. Gilby (New York: McGraw-Hill). Aristotle (1985), Nicomachean Ethics, trans. T. Irwin (Indianapolis: Hackett). Beauchamp, T. L., and Childress, J. F. (1994), Principles of Biomedical Ethics (4th edn. New York: Oxford University Press). (2001), Principles of Biomedical Ethics, (5th edn. New York: Oxford University Press). Bennett, J. (1995), The Act Itself (New York: Oxford University Press). Boyle, J. M. (1978), ‘Praeter intentionem in Aquinas’, The Thomist, 42: 649–65. (1980), ‘Toward Understanding the Principle of Double Effect’, Ethics, 90: 527–38. Brand, M. (1984), Intending and Acting (Cambridge, Mass.: MIT Press). Bratman, M. (1987), Intention, Plans, and Practical Reason (Cambridge, Mass.: Harvard University Press). Brock, D. W. (1993), Life and Death: Philosophical Essays in Biomedical Ethics (New York: Cambridge University Press). Cavanaugh, T. A. (1997), ‘Aquinas’ Account of Double Effect’, The Thomist, 6: 107–21. Clark, P. A., SJ (2000), ‘Methotrexate and Tubal Pregnancies: Direct or Indirect Abortion?’, Linacre Quarterly, 67 (Feb.), 7–24. Davidson, D. (1980), Essays on Actions and Events (Oxford: Clarendon Press). Davis, N. (1984), ‘The Doctrine of Double Effect: Problems of Interpretation’, Philosophical Quarterly, 65: 107–23. Donagan, A. (1987), Choice: The Essential Element in Human Action (London: Routledge & Kegan Paul). Faden, R., and Beauchamp, T. (1986), A History and Theory of Informed Consent (New York: Oxford University Press). Foot, P. (1978), ‘The Problem of Abortion and the Doctrine of Double Effect’, in Foot, Virtues and Vices and Other Essays in Moral Philosophy (Berkeley: University of California Press), 19–32. Glover, J. (1977), Causing Death and Saving Lives (New York: Penguin). Griese, O. N. (1987), Catholic Identity in Health Care: Principles and Practice (Braintree, Mass.: The Pope John Center). Gustafsen, D. (1986), Intention and Agency, Philosophical Studies Series in Philosophy, 33 (Dordrecht: D. Reidel). Hughes, J. R., Stead, L. F., and Lancaster, T. (2003), ‘Antidepressants for Smoking Cessation’, Cochrane Database of Systematic Reviews, 2, CD000031.

‘reinventing’ the rule of double effect

149

Kaczor, C. (1998), ‘Double-Effect Reasoning from Jean Pierre Gury to Peter Knauer’, Theological Studies, 59: 297–316. Kelly, D. F. (1979), The Emergence of Roman Catholic Medical Ethics in North America (New York: Edwin Mellen). Knauer, P. (1979), ‘The Hermeneutical Function of the Principle of Double Effect’, in C. E. Curran and R. A. McCormick (eds.), Readings in Moral Theology, i: Moral Norms and the Catholic Tradition (New York: Paulist Press), 1–39. Kuhse, H. (1987), The Sanctity of Life Doctrine in Medicine: A Critique (Oxford: Clarendon Press). Lipscomb, G. H., McCord, M. L., Stovall, T. G., Huff, G., Portera, S. G., and Ling, F. W. (1999), ‘Predictors of Success of Methotrexate Treatment in Women with Tubal Ectopic Pregnancies’, New England Journal of Medicine, 341: 1974–8. McIntyre, A. (2001), ‘Doing Away With Double Effect’, Ethics, 111: 219–55. Mangan, J. T., SJ (1949), ‘An Historical Analysis of the Principle of Double Effect’, Theological Studies, 10: 41–61. Pearse, D. W. (ed.) (1992), The MIT Dictionary of Modern Economics (4th edn. Cambridge, Mass.: MIT Press). Quill, T. E., Dresser, R., and Brock, D. W. (1997), ‘The Rule of Double Effect: A Critique of Its Use in End-of-Life Decision Making’, New England Journal of Medicine, 337: 1768–71. Quinn, W. (1989), ‘Actions, Intentions, and Consequences: The Doctrine of Double Effect’, Philosophy and Public Affairs, 18: 334–51. Searle, J. (1980), ‘The Intentionality of Intention and Action’, Cognitive Science, 4: 47–70. (1983), Intentionality: An Essay in the Philosophy of Mind (New York: Cambridge University Press). Styron, W. (1979), Sophie’s Choice (New York: Random House). Sulmasy, D. P., OFM (1995), ‘Killing and Allowing to Die’, Ph.D. thesis, Georgetown University. (1998), ‘Killing and Allowing to Die: Another Look’, Journal of Law, Medicine, and Ethics, 26: 55–64. (2000), ‘Double Effect: Intention Is the Solution, Not the Problem’, Journal of Law, Medicine, and Ethics, 28: 26–9. and Pellegrino, E. D. (1999), ‘The Rule of Double Effect: Clearing Up the Double Talk’, Archives of Internal Medicine, 159: 545–50. Vacco v. Quill, 117 S. Ct. 2293 (1997). Wagner, S. J. (1995), ‘Proposition’, in R. Audi (ed.), The Cambridge Dictionary of Philosophy (Cambridge: Cambridge University Press), 658–9. Washington v. Glucksberg, 117 S. Ct. 2258 (1997). Williams, B., and Smart, J. J. C. (1973), Utilitarianism: For and Against (Cambridge: Cambridge University Press). Williams, G. (1957), The Sanctity of Life and the Criminal Law (New York: Knopf).

part ii •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

JUSTICE AND POLICY •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

chapter 6 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

POLICY-MAKING IN PLURALISTIC SO CIETIES •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

søren holm Now, the serious problem is this. A modern democratic society is characterized not simply by a pluralism of comprehensive religious, philosophical, and moral doctrines but by a pluralism of incompatible yet reasonable comprehensive doctrines. No one of these doctrines is affirmed by citizens generally. Nor should one expect that in the foreseeable future one of them, or some other reasonable doctrine, will ever be affirmed by all, or nearly all, citizens. Political liberalism assumes that, for political purposes, a plurality of reasonable, yet incompatible, comprehensive doctrines is the normal result of the exercise of human reason within the framework of the free institutions of a constitutional democratic regime (Rawls 1996, p. xviii; my emphasis). Bioethics is not only concerned with analysing the actions of individual moral agents; it also analyses policy decisions and thereby has an interface with political philosophy. Bioethicists often give unsolicited policy advice, but many also have more official roles on various kinds of ethics committees advising political decision-makers.1 The views put forward in this chapter have been developed through many years of friendly, but profound disagreement and discussion with John Harris, a good friend and practitioner of the virtues of mutual respect. The chapter relies on arguments I have previously discussed in Holm (2001, 2002, 2003a, b, 2004), and in Harris and Holm (2003). 1 I was myself for six years a member of the Danish Council of Ethics, which provides advice on bioethical issues to the Danish government and parliament.

154

søren holm

The purpose of this chapter is to discuss the issues that arise when a state or other lower public authority has to make policy decisions in areas where there is at the same time moral and scientific uncertainty. Perhaps the most recent example of this kind of problem is the still ongoing controversy with regard to public policy in the area of reproductive cloning and the therapeutic use of stem cells (see, for instance, the thematic issues of the journals Bioethics, the American Journal of Bioethics, and the Kennedy Institute of Ethics Journal, all in 2002), but there are many other examples in areas like assisted reproductive technologies, abortion, euthanasia, capital punishment, funding of the health care system, and research ethics.2 It is probably no coincidence that many of these questions touch upon moral issues where not only there are marked disagreements about moral issues, but where these disagreements are seen as fundamental by the persons involved and the corresponding views strongly held. In these cases we often end up in a situation where each side has arguments that it sees as compelling, but which the other side rejects utterly. In the following I will assume that the following two core features characterize the kind of societies we are discussing: (1) they are morally pluralistic, i.e. there is a variety of views concerning the moral issue at hand, and (2) they are liberal democracies, i.e. they are not dictatorships, autocracies, oligarchies, plutocracies, or populist democracies. That they are liberal democracies does not make these societies identical at the level of political systems or ideals. Most industrialized and many industrializing countries fall within the cluster of societies picked out by these two characteristics, but each has its own specific version of a morally pluralistic, liberal democracy. This is, for instance, evident in how the generally agreed principle that religion should not directly determine public policy is implemented. Some liberal democracies have an established church (e.g. England,3 Denmark, Norway), some have a weak separation between church and state (e.g. Sweden, Germany, Spain), and some have a strong separation between church and state (e.g. France, the United States). It is also evident at the structural level, where Montesquieu’s classic separation between the legislative, the judicial, and the executive powers is implemented in quite diverse ways, and where the power balance between these three branches of government differs widely between countries. As David Held points out, these differences concerning the meaning of the term ‘liberal democracy’ not only affect institutional design, they also permeate political philosophy: In the first instance, the ‘liberal’ component of liberal democracy cannot be treated simply as a unity. There are distinctive liberal traditions which embody quite different conceptions 2

Many of the examples in this chapter will be concerned with the stem cell debate, but I believe that the points made are generally valid. 3 But not Scotland or Wales.

policy-making in pluralistic societies

155

from each other of the individual agent, of autonomy, of the rights and duties of subjects, and of the proper nature and form of community. In addition, the ‘celebratory’ view of liberal democracy neglects to explore whether there are any tensions, or even perhaps contradictions, between the ‘liberal’ and ‘democratic’ components of liberal democracy; for example, between the liberal preoccupation with individual rights or ‘frontiers of freedom’ which ‘nobody should be permitted to cross’, and the democratic concern for the regulation of individual and collective action, that is for public accountability. Those who have written at length on this question have frequently resolved it in quite different directions. (Held 1992: 10; notes removed)

It may be the case that political philosophy will eventually reach universal agreement on the exact form of the ideal liberal democracy, but until that happens we will be unable to base our analysis on a canonical account of the ideal society. This multiplicity of forms of morally pluralistic, liberal democracies entails that the analysis in this chapter will have to be relatively general, and that it will therefore risk being slightly off the mark when applied to a particular jurisdiction.

Double Trouble ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

In morally pluralistic societies there are, by definition, many areas of moral disagreement. The scope of the current analysis is, however, only concerned with bioethical disagreements and only with those disagreements where a public policy decision has to be made. There are clearly a number of bioethical contentious issues that fall wholly within the private sphere and do not call for any public policy decisions, although it is important to remember that the extent of the private sphere is itself contentious. If the private is political, it is for instance not the case that activities wholly confined within the walls of the home necessarily fall within a protected private sphere. The class of issues analysed here will be further restricted to encompass only those issues where there is not only moral disagreement, but also significant scientific uncertainty of relevance to the ethical analysis. This immediately points to one potential problem for a general analysis of this type of public decision-making, since there is no general agreement on what scientific facts are of relevance to ethical analysis. Each ethical framework picks its own set of relevant facts. In the following I will, nevertheless, assume that facts about the effectiveness and usefulness of a therapy, its likely impact on major diseases, or its risks and side-effects are relevant facts of interest to most ethical analyses, as are facts about the consequences of implementing one type of policy and not another.4 4 Is it, for instance, the case that introducing capital punishment for a given class of crimes will reduce the number of these crimes?

156

søren holm

The chapter will initially be concerned with general issues of policy-making in liberal democracies in these circumstances, before moving on to an examination of some more specific issues where bioethical analysis and the policy-making process often clash.

Let Freedom Reign ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

One way of solving these policy-making problems that is sometimes explicitly or implicitly invoked in bioethics writing is the idea that we should allow people to act in the ways they want to act if it does not do any harm to others. A significant part of the literature on ‘reproductive freedom’ or ‘reproductive liberty’ does, for instance, adopt this approach, as does the whole libertarian segment of moral philosophy. If we do proceed in this way, most of the policy problems engendered by ethical and scientific uncertainty can be made to go away, since we only need to estimate whether the activity we are considering is likely to cause harm, and do not have to consider its eventual benefits. The benefit estimation can be left to individual actors. This idea is often traced to Mill’s dictum in On Liberty: ‘that the sole end for which mankind are warranted, individually or collectively, in interfering with the liberty of action of any of their number, is self-protection. That the only purpose for which power can be rightly exercised over any member of a civilised community, against his will, is to prevent harm to others’ (Mill 1987: 78). Philosophically this is a rather tempting idea. It is liberal, it is simple, and it seems to solve a whole range of policy-making problems in a way that happens to be in accordance with the views of many liberal bioethicists. But it is not an idea without problems. First, it relies on us being able to define rather precisely what constitutes harms to others. When the Millian principle is used in bioethics, it is often combined with a rather narrow concept of harm, partly because this is what Mill himself does in On Liberty, partly because a combination with a wider or thicker concept of harm could lead to rather illiberal conclusions. But it is not self-evident that the politically (or philosophically) relevant conception of harm can be restricted in a way that does, for instance, leave out harm to the reasonable sensibilities of others, or harm to the environment, or harm to social networks and cohesiveness. The exact demarcation of what harms should count as those kinds of ‘harms to others’ that can justify interference with liberty is itself a contentious issue. Second, the call for freedom often paradoxically has to be combined with a call for state action. Many of the goals that are aimed at through an effective exercise of individual freedom can only be achieved if the legal order is arranged in such a way that the actions create the legal effects people want. Stem cell researchers want to

policy-making in pluralistic societies

157

be able to patent stem cell lines, women pursuing pregnancy through insemination with their dead husbands’ sperm want the dead men named as fathers on the birth certificates, couples pursuing surrogacy arrangements want to be immediately recognized as the child’s parents, etc. Whether or not legal regulations should be changed to enable people to accomplish these goals cannot be decided by a reference to their liberty interests. A right to patent your invention is, for instance, not a natural right, but something that is granted by society in order to further the societal goals of economic prosperity and promotion of useful inventions, by creating positive incentives to invent. It is thus impossible to discuss whether stem cell lines should be patentable, without discussing whether society should promote the derivation of stem cell lines, since granting patentability would have the express aim of promoting the derivation of cell lines through the creation of larger and/or more certain economic incentives than exist without patentability. Third, unless we adopt a libertarian approach where the state has no role in trying to promote even a thin concept of the common good, we still have to decide how the state should allocate its resources between different claims, and that question is not answered by appeal to individual liberty. What is desired is often not only the freedom to do X, but state support in the pursuit of X. Whether all taxation is theft is a large issue in itself (for an interesting new approach, see Murphy and Nagel 2002), but there is clearly a distinct subsidiary issue about state funding of contentious activities out of tax revenues. The fact that the state allows me to pursue a certain activity (e.g. bondage and domination in the privacy of my home) does not necessarily entail that it should support this activity economically in a direct way, or that it should, for instance, support education about how best to carry out the activity. On the other hand, it is not clear either that the state should force everyone to contribute to contentious activities that the state pursues. It is now, for instance, generally recognized that there should be a right to conscientious objection to military service (for the best extant analysis, see the collected papers in Bedau 1969), as well as conscientious objection for health care professionals in the context of abortion provision (Holm 2004). Are there any good reasons for completely ruling out conscientious objection in the context of funding of these activities, apart from reasons of practicality and expediency? A direct appeal to liberty or freedom will, for the reasons outlined above, often fail when considering the doubly contentious issues we are discussing here. Either the activity or development in question will fall within a contested area of freedoms and the liberty argument will therefore not be (politically) decisive or compelling, or what is being sought is not only the liberty to pursue the development, but ‘liberty plus’, liberty plus other legal change, or liberty plus public support.

158

søren holm

Deliberative Democracy and Legitimate Policy-Making ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

If we cannot solve our policy-making problems by a hole-in-one approach, like a direct appeal to liberty rights as trumps, what are our options? Are we left with a choice between populist, majority rule where the winner takes all, or rule by philosophers (or perhaps theologians) as envisaged in The Republic, book 6? If this was our choice, we would be in trouble since both alternatives are rather unattractive. Pure majority rule often leads to highly problematic outcomes involving discrimination of minorities, and rule by philosophers is indistinguishable from other forms of, seemingly meritocratic, oligarchy. But, fortunately the dichotomy between populism and philosophocracy is a false dichotomy. Democracy has many forms and populist democracy is only one of them (Gutmann 1993). Democracy is not only a mechanism for the aggregation of preferences, as in pure populist democracy, but also a mechanism for an examination of and societal discourse about the preferences before they are aggregated. The legitimacy of a given policy is based not only on whether it was enacted according to the current legal rules, but also on the quality of deliberation preceding the policy, on the decisive reasons for the policy, and on its congruence with basic freedoms. Bioethicists therefore have at least to acknowledge that in liberal democracies the question of getting the right policies implemented cannot be divorced from the question of what constitutes legitimate political procedures and discourse. As the large literature on deliberative democracy shows, specifying the conditions for political legitimacy is in itself a contentious exercise and fraught with difficulty. There are many theories of deliberative democracy, each with its own normative foundation. In the following, I will mainly draw on the Rawlsian version as proposed by Rawls in ‘Political Liberalism’ (Rawls 1996) and on the considerations in Gutmann and Thompson’s ‘Moral Conflict and Political Consensus’ (Gutmann and Thompson 1990).5 This choice is made for two reasons, first because the practical implications of this version of deliberative democracy are relatively clear, second because its theoretical framework is congruent with the mainstream analytic tradition that also encompasses much of Anglo-American bioethics. I do, however, think that very similar conclusions would be reached if we applied the other main branch of thinking on deliberative democracy developed by J¨urgen Habermas (1992). There are distinct theoretical differences between Habermas and Rawls in the justification of deliberative democracy and its procedures (Habermas 1995; Rawls 1995), but very few differences at the pragmatic level of actual exercises of deliberative democracy. 5 Rawls himself only argues that his approach should be applied to the setting up and modifying of the basic structures of a given society, i.e. at the constitutional level.

policy-making in pluralistic societies

159

The basic premisses of deliberative democracy are (1) that there is an in practice irreducible plurality of reasonable comprehensive world-views in modern societies6 because of what Rawls calls the ‘Burdens of Judgment’;7 (2) that policy-making about fundamental issues should rely on public reason instead of the use of force; (3) that the aim of the use of public reason is to reach a policy that is acceptable to all as citizens because it offers them fair terms of cooperation; and (4) that 1, 2, and 3 combined justify certain restrictions on public political debate about fundamental issues, especially that it has to proceed within the confines of public reason. Within the context of deliberative democracy it is thus important to distinguish between different kinds of debate that may all be part of the wider policy-making process. Within a comprehensive world-view, in a philosophical debate, or when thinking through the issues for oneself, the adherence to public reason is not a requirement, it is only a requirement in the public debate, between persons holding different comprehensive world-views. What exactly are the requirements for participants in the public debate in a deliberative democracy—the features that distinguish public reason? The main requirement is a requirement for reciprocity in public discourse. Reciprocity has two components; it entails that the principles and standards that are proposed have to be principles and standards that are viewed as reasonable for everyone to accept as fair terms of cooperation, and that there is a willingness to discuss the fair terms that others propose. As Gutmann and Thompson point out, true reciprocity can only occur if there is mutual respect, and if the participants in the debate evince two particular civic virtues: the virtue of integrity and the virtue of magnanimity8 (Gutmann and Thompson 1990). Integrity in this context requires consistency of speech in different situations, performative consistency, and the willingness to recognize all the broader implications of the position one puts 6

See the quotation at the beginning of this chapter. The burdens of judgement are those factors that lead reasonable people to hold different views on the same issues. Rawls enumerates the following six: ‘(a) The evidence—empirical and scientific—bearing on the case is conflicting and complex, and thus hard to assess and evaluate. (b) Even where we agree fully about the kinds of considerations that are relevant, we may disagree about their weight, and so arrive at different judgments. (c) To some extent all our concepts, and not only moral and political concepts, are vague and subject to hard cases; and this indeterminacy means that we must rely on judgment and interpretation . . . within some range . . . where reasonable persons may differ. (d) To some extent . . . the way we assess evidence and weigh moral and political values is shaped by our total experience, our whole course of life up to now; and our total experiences must always differ . . . (e) Often there are different kinds of normative considerations of different force on both sides of an issue and it is difficult to make an overall assessment. (f ) Finally . . . any system of social institutions is limited in the values it can admit so that some selection must be made from the full range of moral and political values that might be realised . . . . Many hard decisions may seem to have no clear answer’ (Rawls 1996: 56–7; notes removed). 8 We will here only be discussing the civic versions of these two virtues. There are also personal virtues of integrity and magnanimity that affect our personal and intimate relations to others, but an exploration of them lie outside the scope of this chapter. 7

160

søren holm

forward. Magnanimity requires the recognition of the moral status of the opposing position (i.e. treating it as a moral view and not, for instance, as mainly a politically expedient view), openness to the possibility that I might come to change my own position, and a commitment to ‘search for significant points of convergence between our own understandings and those of citizens whose positions, taken in their more comprehensive form, we must reject’ (Gutmann and Thompson 1990: 82). Rawls himself summarizes the requirement of public reason in this way: What public reason asks is that citizens be able to explain their vote to one another in terms of a reasonable balance of public political values, it being understood by everyone that of course the plurality of reasonable comprehensive doctrines held by citizens is thought by them to provide further and often transcendent backing for those values. (Rawls 1996: 243)

When we enter the realm of public discourse concerning contentious issues, we have minimally to accept that there are other reasonable views than those we hold, and that the purpose of the public discourse is not to ‘win’, but to reach a solution that is mutually acceptable and respectful. In this way public discourse in a deliberative democracy differs substantially from standard philosophical discourse (see below). Here it might be important to note that the account of reciprocity and the virtues of integrity and magnanimity is not inextricably connected to Rawls’s further claim about the goal of public policy-making being an overlapping consensus. Reciprocity in public discourse is necessary even if we believe that an overlapping consensus is not realistically achievable or not the right goal at all. If we, for instance, agree with Sunstein that what we should aim for in most circumstances is an ‘Incompletely Theorized Agreement’ (Sunstein 1995), we would still have to discuss in a context of reciprocity to reach such an agreement. One attempt to specify further the requirements of deliberative democracy in an important, but contested, domain of public decision-making is the framework of accountability for reasonableness in the process of policy-making on priorities in health care developed by Norman Daniels and Jim Sabin (Daniels and Sabin 1997, 2002).9 It has four distinct components: publicity, relevance, appeals, and enforcement. Publicity entails a call for explicitness and requires that the rationale for priority decisions be made accessible to the wider public and open for scrutiny. Relevance is a requirement to screen out irrationality in priority-setting. Priority decisions must be made only in accordance with reasons that reasonable people will agree are relevant and adequate. The appeals component is an institutional mechanism 9 I owe significant parts of my account of accountability for reasonableness to a collaboration with Andreas Hasman.

policy-making in pluralistic societies

161

that provides persons with an opportunity to dispute and challenge decisions that have gone against them, as well as providing organizations with an option to revise decisions in light of further arguments. Enforcement simply entails public or voluntary regulation of the decision process, to ensure that the three other components are maintained (Daniels and Sabin 1998). A concrete decision process, which incorporates the components of accountability for reasonableness, will therefore have to meet requirements for both structure and content. An appropriate institutional setting must be put in place in the form of a group of reasonable people, and a procedure for appropriate deliberation amongst them established. An organizational structure that assures the dissemination of the group’s reasons and decisions to the public must also be implemented. Moreover, similar but separate decision processes will have to be set up to deal with appeals and enforcement of decisions internally. These are all requirements for the structure of the process. A different set of requirements relates to the content of the process of decision-making and the specification of appropriate deliberation. Daniels and Sabin think of the relevance component as a constraint on deliberation and contend that any rationale for a decision must be reasonable. A rationale is reasonable if it appeals to reasons that are accepted by people who are disposed to finding mutually justifiable terms of cooperation, as relevant to how the varied health needs of a defined population are met under reasonable resource constraints (Daniels and Sabin 1997). This approach has been tried in a number of health care institutions and has been shown to produce agreement around results that are seen as legitimate by the various stakeholders (Martin et al. 2003).

Is the Standard Model of Bioethics Analysis Incompatible with Deliberative Democracy? ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

The discussion so far raises an interesting question. Is the standard model of bioethics analysis incompatible with public discussion in a deliberative democracy? By the standard model of bioethics analysis I mean the following process (this is clearly a simplification, but not excessively so). (1) A putative moral problem is identified; (2) it is analysed and sharpened in various ways, often to the point where it is stated that ‘We can now see that the problem is essentially . . . ’; (3) a number of solutions to this refined problem are canvassed, and counter-arguments claimed to be decisive are marshalled against all but one of them (or in some cases against all of them); (4) positive arguments in support of the remaining solution (or a

162

søren holm

new solution brought in at this stage) are presented;10 (5) a definitive conclusion is reached. This model is an internalized variant of what James Sterba calls the ‘warmaking model of doing philosophy’ and describes in the following way: I was once asked by a well-known philosopher why I talked to libertarians. At the time, I was dumbstruck by the question, but now I believe that it reflects the dominant way that philosophy is being done these days, and maybe even the dominant way that philosophy has always been done. It sees philosophers as belonging to different groups within which there can be a significant degree of sympathetic understanding but between which there can only be hostile relations, a virtual state of war. If you believe this is the case, then there really is a question about whether you should talk to your philosophical enemies. You may perchance say something that indicates certain problems with your own philosophical view, which may in turn be used against you, and, as a result, you may lose an important philosophical battle and your reputation may decline accordingly. (Sterba 1998: 2–3)

Sterba’s description is rather stark, but it is not too far off the mark. Implicit in the standard model are a number of assumptions about the aims of philosophical analysis and the relationship between the view that is argued for, and the competing views—most importantly in the present context the assumption that most opposing views are not only wrong in some sense, but that this can also be shown decisively to be the case. The views are not only wrong; they are also irrational, or at least unreasonable. When the bioethicist enters the public debate, it is therefore with the firm conviction that most of the views being put forward by others have already been shown to be unreasonable. There is no reason to take account of them or to let them influence decision-making. Discussing them is only worthwhile as a tactical manoeuvre to get your own arguments on the table, or in order to show them to be unreasonable. The standard model, and to an even greater extent the rhetoric of bioethics, imply that there is only one right solution to each policy-making decision, that it is possible to find this solution, and to know that it is the right decision. ‘Knowing that you are right’ is in general a bar to a positive engagement with the views of others,11 but ‘knowing that you are right after proper philosophical reflection’ can be even more problematic. In contrast, theorizing about deliberative democracy proceeds from the assumption that, whereas there might be a right solution, it is often impossible to find this solution or to know that it is the right solution in any absolute sense. The best we can do is to outline the area of acceptable policies, and then choose a policy within this area through a deliberative, democratic process. 10 It is a feature of much bioethics writing (and of philosophical argument in general) that the negative arguments against other positions are often more convincing than the positive arguments for the author’s own position. 11 As is evidenced in many domestic rows.

policy-making in pluralistic societies

163

If philosophers participate in the public discourse about contentious issues in the same way as they engage in philosophical discourse, they are very likely to breach many of the requirements for an engagement in public reason. But they will also have engaged in a social category mistake. Attempting to reach a legitimate public policy in a morally pluralistic liberal democracy is not a social activity of the same kind as trying to win a philosophical argument. If all participants in a public debate adopted a war-making approach to the debate, we would eventually only have recourse to simple, non-deliberative voting for resolving our differences. Let us now move on to some more specific problems.

Hope and Hype, Doom and Gloom ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

A specific problem arises in those cases where our efforts of public decision-making, including public debate and deliberation, take place at such an early state of the development of a given technology that the scientific uncertainty is very large. In such cases there will typically be two scenarios painted by the proponents and the opponents of the new development, a ‘Hope and Hype’ scenario promising untold benefits if the technology is developed and a ‘Doom and Gloom’ scenario threatening all kinds of risks from the new technology. In retrospect, it often becomes evident that both of the two scenarios were wrong, and often for the very same reason—they both overestimated the new technology and its powers. In the case of human genetics it does, for instance, appear likely that Gregory Stock’s Redesigning Humans will eventually be shown to fall within the ‘Hope and Hype’ category, and Francis Fukuyama’s Our Posthuman Future in the ‘Doom and Gloom’ category (Stock 2002; Fukuyama 2002). Historically it is interesting that a significant segment of the early bioethicist community in the late 1960s and early 1970s were very sceptical towards the development of new medical technologies, seeing them as potentially dangerous, either because the technologies would increase the power of the medical profession, or because they would lay our lives open to more commercial influences. This technology-sceptical stream in bioethics seems to have been somewhat marginalized in recent times, at least in Anglo-American bioethics, where it has been replaced by a marked enthusiasm for new technologies that are seen as liberating. In contrast to this development in bioethics, there is still a significant sceptical attitude towards the blessings of new technologies in environmental ethics. Generally our ethical analysis and our policy-making are likely to go astray if we do not get the scientific facts right, a point often made in debates concerning the moral status of the embryo, but they are equally likely to go wrong if we do not get the scientific predictions right. Both too enthusiastic and too pessimistic

164

søren holm

evaluations of the science and its practical implications can seriously bias ethical argument. John Harris has pointed out many times that it must be the ends that justify the means, but this means that if we get our prediction of the actually achievable ends wrong, we may end up allowing, or prohibiting, the use of means that should not have been allowed or prohibited. In a society where technology is generally seen as a good, the tendency to overestimate the good effects of new technology is strong. As Khushf and Best point out, following Horkheimer and Adorno: Enlightened science promises the means for entering a realm of light and hope, but this means slides over into its opposite, resulting in an empty and dehumanizing rhetoric of justification. Minimally in the light of the accomplishments of science, we are forced to reassess what has been promised. One way of reading this history is to say that we should be sceptical of all utopian accounts that see the promised land emerging in this worldly terms. (Khushf and Best 2002: 37–8)

But underestimation may also occur. One example of this occurred in the ethical analysis of genetic testing. Before the development of gene-chips allowing the simultaneous detection of many gene variants, a lot of papers on the ethics of genetic testing and screening were written that explicitly relied on the empirical premiss that it was only possible to test for one or a few gene variants at the same time, and that testing and screening would therefore always be directed in the sense that one would always be looking for something specific. Worries about screening creating a genetic underclass were dismissed, because this could only happen if many genes could be screened at the same time. But we now know that the empirical premiss used in the argument was false. Overestimation of the final impact of a given development may be a conscious rhetorical ploy, in which case it is clearly ethically problematic,12 but it may also be an effect of a tendency to analyse new technologies in isolation from other concurrent developments. At a given point in time there will often be many different approaches to the solution of a particular problem, e.g. many different therapies being developed for the same condition. If we look at each of these in isolation we are likely to overestimate its importance, 12

A subset of this issue is the ethical problems raised by far too optimistic predictions of timescale between initial scientific breakthrough and routine application. In the case of stem cells, the public was promised real benefits within five to ten years. Many years have now elapsed of the five to ten years and the promised therapies are still not anywhere close to routine clinical use. There are similarities to the initial enthusiastic presentation of gene therapy in the late 1980s and the later problems encountered, and some reason to fear that stem cell therapies will have an equally long trajectory between theoretical possibility and clinical practice. It is clearly ethically problematic to raise false expectations in seriously ill people, and even more problematic if this is partly done from self-interest (e.g. to promote one’s own research in the media).

policy-making in pluralistic societies

165

because we bracket the consideration that one of the other concurrent developments may be developed first and corner the market, may turn out to be better overall or better for a subclass of people with the condition. Even good technologies are in a competitive race, and if we forget then we will often hugely overestimate their future positive and negative impact (an error made by many phone companies in assessing the value of third-generation cell phone licences). All of these problems could be solved for the bioethicist and the policy-maker if it was possible to get accurate, reliable, and unbiased estimates of the likely effectiveness and long-term impact of new scientific developments. Getting such estimates is, however, exceedingly difficult. It may be the case, as in the current stem cell debate, that the best experts all have an interest to defend and that their estimates of the success of the line of research they pursue is therefore likely to be consciously or unconsciously biased. But even where this is not the case, and where it is possible to find unbiased experts, predicting long-term impact is very difficult, partly for the reasons mentioned above concerning competing technologies, partly because long-term impact will depend on and interact with the social embedding of the technology. To give an example, it was predictable that the development of really small cell phones would change our modes of communication in profound ways, and perhaps predictable that they would become fashion accessories; but no one predicted (or could have predicted?) the immense success of the cell phone, not as a phone (i.e. a device for oral communication) but as a sender of small text messages, or the rapidly growing market in ring tones. It may even be impossible for the researchers involved in a given research programme to provide the necessary facts for a proper policy analysis, because the science is moving forward very quickly. Believing some aspect of a developing technology to be a stable feature of that technology is therefore always dangerous. One of the members of the Geron Ethicist Advisory Board describes her experiences with the information given to the Board: The Geron board was introduced formally to the notion of the hunt for hES [human embryonic stem] cells with a careful scientific explanation of the mechanism of the research and the motivation for the science itself. . . . A second focus of concern that arose early was the use of embryos. Our understanding was that embryos were graded (1–4) with grades 1 and 2 considered useable for implantation, and grades 3 and 4 considered too physically imperfect to be used. . . . We understood at that time that no embryos would be created for the purposes of research, that embryos would be donated under the most stringent system of informed consent, and that these embryos would have been discarded in any case. . . . In the next few months the science rapidly moved forward, and hence our elaborate rationale, although certainly interesting, became archaic. Since the norm for implantation was now two instead of five, many more embryos were available but would not be used by a particular

166

søren holm

couple, giving rise to a new ethical problem of ‘spare’ embryos and their use and storage.13 (Zoloth 2002: 4–5; notes removed)

Even if unbiased information can be gained, the problems are not over, because such information has to be interpreted and its significance assessed. At approximately the same time the American National Bioethics Advisory Commission and a British government expert group reviewed the evidence concerning the need for the creation of embryos specifically for stem cell research and came to two rather different conclusions. The National Bioethics Advisory Commission concluded that Currently, we believe that cadaveric fetal tissue and embryos remaining after infertility treatments provide an adequate supply of research resources for federal research projects involving human embryos. Therefore, embryos created specifically for research purposes are not needed at the current time in order to conduct important research in this area . . . We conclude that at this time, because other sources are likely to provide the cells needed for the preliminary stages of research, federal funding should not be provided to derive ES [embryonic stem] cells from SCNT [somatic cell nuclear transfer]. Nevertheless, the medical utility and scientific progress of this line of research should be monitored closely. (National Bioethics Advisory Commission 1999: 71–2)

Whereas the British Chief Medical Officer’s Expert Group concluded that For some people, particularly those suffering from the diseases likely to benefit from the treatments that could be developed, the fact that research to create embryos by cell nuclear replacement is a necessary step to understanding how to reprogramme adult cells to produce compatible tissue provides sufficient ethical justification for allowing the research to proceed. (Department of Health 2000: 40)

What was a fact for one group of experts was clearly not a fact for the other. What is at play here is a different evaluation of the available scientific evidence, but possibly also a different approach to the decision of whether a line of research should be deemed ‘necessary’. Is a particular line of research only necessary if it is the only way to get the knowledge we need for stem cell therapies, or is it necessary if scientific progress will otherwise be slowed down and will be much more costly, but will eventually lead to stem cell therapies anyway even if this particular line of research is not pursued (Holm 2001)? There is no principled way to solve the problems around scientific prediction and its interpretation. We can do better if we base our ethical and policy analysis on a range of possible developments of a technology, instead of relying on only one prediction, but we can never be sure that any of our scenarios will actually match the future. Policy decisions at the early stages of technology development should therefore almost always be seen as tentative, and not definitive. 13 It would be interesting to know how this could be classified as a new problem, since it has been discussed for more than ten years; see e.g. Holm (1993).

policy-making in pluralistic societies

167

Path Dependence and the ex ante Position ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Philosophical discussion of ethical issues and ethical policies sometimes takes place as if we are drawing policy on a blank slate, or deducing the right policy from first principles. Reference to already existing regulation is only made in two situations, either to criticize it for being wrong, inconsistent, and incoherent, or conversely in consistency arguments allegedly showing that already existing policies imply certain views concerning ethics, for instance concerning the ethical status of the embryo. But does it make sense to conceive of public policy-making as a series of unconnected acts where each can be considered on its own, without any reference to past decisions? Does where we are now, in terms of laws and policies, not at least to some extent determine where we can go and what means we should choose to get there? There are reasons to be sceptical about the strength of consistency arguments from existing policy. In order to get the consistency argument off the ground, we almost always have to reconstruct the underlying justification for the policy, because such a justification is very rarely given explicitly, except if the decision is a judicial decision. There are usually many possible reconstructions, and committing the policy-maker to only one of these may be impossible. Consistency arguments may, furthermore, fail completely if the policy is a result of a compromise, either because the policy-makers have openly engaged in a compromise with some other group, or because they have taken the views of some other group into account already when drafting the regulation (for instance, in order to ensure a smooth passage through the political process). If a regulation is based on a political compromise, it may well be the case that no one wants to defend or justify the position reached in the regulation, except as a legitimate result of a legitimate political process. This means that there is no argument to reconstruct, and no one saying A whom we can also commit to say B. An example could be the abortion legislation in countries allowing abortion on demand until a certain gestational age. It is unlikely that many people would actually claim to have an argument that can justify the exact limit legislated in a specific country, but many more might be willing to accept it as a legitimate political compromise between those wanting a higher limit and those wanting a lower limit. Therefore, such laws do not necessarily imply that something radical is thought to happen to the moral status of the fetus or the moral rights of the mother when the gestational age in question is reached. In implying further consequences of official policy we also need to take into account that there are limits to the preciseness with which one can draft regulation,

168

søren holm

and that ethically extraneous factors play a role. As Beauchamp and Childress write: Public policy is often formulated in contexts that are marked by profound social disagreements, uncertainties, and different interpretations of history. No body of abstract moral principles and rules can determine policy in such circumstances, because it cannot contain enough specific information or provide direct and discerning guidance. The specification and implementation of moral principles and rules must take account of problems of feasibility, efficiency, cultural pluralism, political procedures, uncertainty about risk, noncompliance by patients, and the like. (Beauchamp and Childress 2001: 8–9)

In evaluating official policy it is also very important not to forget that the ideal policy has to be both locally and globally coherent and consistent, and that maximal coherence in a body of regulations may be achieved by allowing some local inconsistencies. In many legal systems it is, for instance, accepted that a major change in legal status occurs at birth. This principle permeates many areas of law, including the laws outside those strictly dealing with reproductive matters, and it is at least arguable that global consistency is better served by upholding it in all cases, even if there are a few cases where it might seem to be locally inconsistent. Let us look at a specific example. In the context of stem cell research, one particular consistency argument is often put forward claiming that because destructive research using embryos for certain purposes is already allowed in a given jurisdiction, consistency requires that we allow the destructive use of embryos for stem cell research (Harris 2002). Many countries that presently allow experiments on embryos restrict these to experiments related to improving reproductive technologies and increasing our understanding of the biology of human reproduction. Many types of stem cell research fall outside this restriction and are therefore not permitted even though the jurisdictions in question allow other forms of research on embryos. Can a restriction of embryo research to reproductive matters be justified? The decision to allow embryo research for a restricted range of research questions could possibly be reconstructed as an attempt to achieve consistency in a situation where embryo research is believed to be (somewhat) ethically problematic.14 Most legal regulation of embryo research occurs within the broader context of regulation of assisted reproductive technologies, and it is evident that these technologies could only have been developed, and can only be improved, if embryo research is permitted. Any legislation that allows the use of assisted reproductive technologies and prohibits embryo research could therefore be charged with a form of performative inconsistency by prohibiting a necessary step in the development and improvement of a permitted technique. 14 Many public policies seem to indicate that people are ‘closet gradualists’ with regard to the moral status of embryos and fetuses.

policy-making in pluralistic societies

169

Would it then be inconsistent not to allow research using embryos with no connection to reproduction? Michael Walzer has famously argued that society contains several separate ‘spheres of justice’ and that the application of principles of justice from one sphere to another is not necessarily warranted (Walzer 1983). Maybe there are also separate ‘spheres of consistency’ when we discuss the consistency of public regulations. This is not as strange an idea as it might perhaps initially seem. Walzer’s argument for different spheres of justice is that different social goods are different, and that unless we understand and take account of this difference in our analysis of justice we will go wrong. He writes: No account of the meaning of a social good, or of the sphere within which it legitimately operates, will be uncontroversial. Nor is there any neat procedure for generating or testing different accounts. At best, the arguments will be rough, reflecting the diverse and conflict-ridden character of the social life that we seek simultaneously to understand and to regulate—but not to regulate until we understand. I shall set aside, then, all claims made on behalf of any single distributive criterion, for no such criterion can possibly match the diversity of social goods. (Walzer 1983: 21)

In the same way as different social goods are different, the different social practices aimed at producing these social goods are different. The practices involved in reproduction and in securing the goods secured by reproduction are different from the practices involved in research. There is thus no prima facie reason to believe that arguments and conclusions valid in one of these areas can be transferred without modification to the other area. In the area of reproductive ethics, the idea that there is a strong right to reproductive freedom or reproductive liberty has gained currency in recent years. If a given permissive legal regulation has been passed because of an appreciation of this right, then it is not immediately obvious that consistency requires the same kind of permissive regulation outside the sphere of reproduction. A right to reproductive liberty cannot, for instance, in itself support a permission to use embryos for non-reproduction-related purposes, like stem cell research! In looking at consistency arguments in the public policy sphere we might also want to question whether ‘as morally acceptable’ is a transitive relation in this context. If p is as morally acceptable as q and q is as morally acceptable as r, have we therefore committed ourselves to the judgement that p is as morally acceptable as r? Moral acceptability comes in at least two different forms, and the difference between them can be brought out by considering Tranøy’s analysis of moral disagreement and consensus. If two people A and B discuss the morality of p there are, according to Tranøy, three different possible outcomes: (1) Person A and B can agree that p is unacceptable, i.e. both mean positively that p should not be accepted.

170

søren holm

(2) A and B can agree that p is acceptable, i.e. both mean positively that p should be accepted. (3) A and B can agree that p is acceptable, but A or B might abstain from actively taking a stand on p. ( Tranøy 1998: 155)15

Tranøy calls the last option ‘open consensus’.16 Let us call the kind of moral acceptability where both decision-makers positively affirm the moral acceptability of p ‘moral acceptability’ and the kind of moral acceptability where one of the decision-makers abstains from taking a stand on p ‘moral acceptability*’. Let us assume that moral acceptability is transitive; does this entail that moral acceptability* is also transitive? Clearly not. The judgement of moral acceptability* may occur for a number of different reasons. The case p may for B be in a grey area between the clearly acceptable or the clearly unacceptable, or B may defer to A’s judgement because the question at issue is much more important for A. Moral acceptability* is thus only transitive in those cases where B’s reason for abstaining from actively taking a stand on p is shared by q and r, and since these reasons may be non-moral, transitivity of moral acceptability* only occurs if we can claim parity of reasoning for both the moral features and the non-moral features of p, q, and r. It should be relatively obvious that the kind of moral acceptability that lies behind public regulation of controversial areas in bioethics is most often not moral acceptability but moral acceptability*. This means that purely ethical consistency arguments are often misapplied, because they assume a transitivity of moral acceptability that is often not there.

Compromise, Moratoriums, and Accommodation ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

A further reason why philosophical analysis and public policy may part company is that a number of standard political tools are not very prima facie appealing to the philosopher, since they may lead to decisions that are formally inconsistent, eclectically justified, or only legitimized procedurally through a proper democratic process. These tools include compromise, moratoriums, and various means of accommodating minority positions. Prominent examples of successful and stable compromises are the abortion laws in many European countries that allow abortion on demand until a certain stage 15

I owe the translation from Norwegian to Jan Helge Solbakk. Open consensus can also occur around the judgement that p is unacceptable if either A or B abstains from actively taking a stand on p. 16

policy-making in pluralistic societies

171

in pregnancy and after that only allow abortion if certain criteria are fulfilled. The gestational age that is the legal cut-off point for abortion on demand varies between countries, as one would expect of a result of local political compromise. No one tries to defend their particular cut-off point as the uniquely right one, but the policies are remarkably stable within each society. They have given women significant reproductive freedom, they have removed the harm previously caused by illegal abortions, and they have to some degree accommodated the rather common belief that fetuses are morally significant and that this significance increases through pregnancy. This stability is clearly a political or societal and not a philosophical stability. There are strong arguments on one side to allow abortion on demand throughout pregnancy (this follows from some personhood views of moral status and from some analyses of reproductive liberty as a human right) and there are on the other side strong arguments for never allowing abortion on demand (this follows from some views giving significant moral status to all human individuals). It is, however, rather obvious that a policy implementing one of these philosophical positions would be politically highly unstable. Instituting a moratorium can also be a very useful and legitimate political technique. If there is genuine scientific uncertainty, a well-crafted moratorium blocking certain possibly problematic uses of a technology may allow time to resolve the scientific uncertainty before a final public policy decision is made. It could even be argued that in a deliberative democracy there should always be an initial moratorium, or at least a period of cautious development, while the necessary societal deliberation about the implications of a new technology takes place. Referring the issue at hand to a national ethical council or similar body often combines two functions: it creates a moratorium-like situation while the body deliberates, and it promotes wider societal discussion. Like all other human activities, policy-making carries opportunity costs. There are a limited number of policy decisions that can be made (e.g. because of limits on parliamentary time or on the number of issues in which the executive can simultaneously invest political capital), and this entails that it is important to get the policy right the first time and/or to have a policy that is sufficiently flexible to accommodate new technological and social developments. This is often better achieved if there is no rush to comprehensive regulation, but more time to assess the real implications of a given technology. Accommodating minority positions17 in different ways is not only politically expedient; it is also a way of showing proper mutual respect. If a society eventually decides to allow and support embryonic stem cell research, one way of accommodating those opposing this research could, for instance, be to provide generous 17 In a deliberative democracy it may be the case that the policy that is implemented is not the majority position but a position that is actually only held by a minority, if for instance the policy is passed owing to moral–political agreement in the third of Tranøy’s senses (see above). In that case it may be relevant to accommodate the majority position!

172

søren holm

funding for research into adult stem cells so that there would be real competition between the two alternative sources of stem cells, or to take measures to ensure that as few embryos as possible were used in research. Both accommodations would show that the reasonable concern for embryos that was suppressed in the final public policy was nevertheless considered with respect.

Conclusion ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Being a bioethicist myself with fairly clear and well-justified (at least in my own estimation) views on a range of contentious issues, I would have liked to be able to conclude that society ought to implement the results of good ethical analysis as public policy straight away. However, like everyone else, I have to recognize that, despite the fact that I think I have good (and to me decisive) arguments against those who disagree with me, their disagreement is often not unreasonable, given that we are unable to agree on some of the basic premisses even after all the arguments have been put on the table. They and I have to live together in the same society, and just as I would not like to have their views imposed as public policy, I should not try to impose my view, either directly or indirectly through the use of questionable rhetorical strategies in the public debate. It is only if we all adhere to the civic virtues of integrity and magnanimity that there is any chance of our finding those areas of practical convergence that allow peaceful public decision-making in a context of fundamental moral disagreement. This may require philosophers to shed some of their bad habits, but if they feel hard done by, they should think how much more difficult the task of developing these virtues will be for many politicians!

References American Journal of Bioethics (2002), 2/1. Beauchamp, T. L., and Childress, J. F. (2001), Principles of Biomedical Ethics, 5th edn. (Oxford: Oxford University Press). Bedau, H. A. (ed.) (1969), Civil Disobedience (New York: Pegasus). Bioethics (2002), 16/6. Daniels, N., and Sabin, J. (1997), ‘Limits to Health Care: Fair Procedures, Democratic Deliberation, and the Legitimacy Problem for Insurers’, Philosophy and Public Affairs, 4: 303–50. (1998), ‘Last Chance Therapies and Managed Care: Pluralism, Fair Procedures, and Legitimacy’, Hastings Center Report, 28/2: 27–41. (2002), Setting Limits Fairly: Can We Learn to Share Medical Resources? (Oxford: Oxford University Press).

policy-making in pluralistic societies

173

Department of Health (2000), Stem Cell Research: Medical Progress with Responsibility: A Report from the Chief Medical Officer’s Expert Group Reviewing the Potential of Developments in Stem Cell Research and Cell Nuclear Replacement to Benefit Human Health (London: Department of Health). Fukuyama, F. (2002), Our Posthuman Future (New York: Profile). Gutmann, A. (1993), ‘Democracy’, in R. E. Goodin and P. Pettit (eds.), A Companion to Contemporary Political Philosophy (Oxford: Blackwell), 411–21. and Thompson, D. (1990), ‘Moral Conflict and Political Consensus’, Ethics, 101/1: 64–88. Habermas, J. (1992), Faktizit¨at und Geltung (Frankfurt am Main: Suhrkamp). (1995), ‘Reconciliation Through the Public Use of Reasons: Remarks on John Rawls’s Political Liberalism’, Journal of Philosophy, 92/3: 109–31. Harris, J. (2002), ‘The Ethical Use of Human Embryonic Stem Cells in Research and Therapy’, in J. Burley and J. Harris (eds.), A Companion to Genethics (Oxford: Blackwell), 158–74. and Holm, S. (2003), ‘Abortion’, in H. La Follette (ed.), The Oxford Handbook of Practical Ethics (Oxford: Oxford University Press), 112–35. Held, D. (1992), ‘Democracy: From City-States to a Cosmopolitan Order’, Political Studies, 40: 10–39. Holm, S. (1993), ‘The Spare Embryo—A Red Herring in the Embryo Experimentation Debate’, Health Care Analysis, 1/1: 63–6. (2001), ‘European and American Ethical Debates About Stem Cells: Common Underlying Themes and Some Significant Differences’, in Nordic Committee on Bioethics, The Ethical Issues in Stem Cell Research (Copenhagen: Nordic Council of Ministers), 35–45. (2002), ‘Going to the Roots of the Stem Cell Controversy’, Bioethics, 16/6: 493–507. (2003a), ‘The Ethical Case Against Stem Cell Research’, Cambridge Quarterly of Health Care Ethics, 12/4: 372–83. (2003b), ‘ ‘‘Parity of Reasoning’’ Arguments in Bioethics: Some Methodological Considerations’, in M. H¨ayry and T. Takala (eds.), Scratching the Surface of Bioethics (Amsterdam: Rodopi), 47–56. (2004), ‘Conscientious Objection and Civil Disobedience in the Context of Assisted Reproductive Technologies’, T¨urkiye Klinikleri Journal of Medical Ethics, 11/4: 215–20. Kennedy Institute of Ethics Journal (2002), 12/2. Khushf, G., and Best, R. G. (2002), ‘Stem Cells and the Man on the Moon: Should We Go There From Here?’, American Journal of Bioethics, 2/1: 37–9. Martin, D., Shulman, K., Santiago-Sorrell, P., and Singer, P. (2003), ‘Priority-Setting and Hospital Strategic Planning: A Qualitative Case-Study’, Journal of Health Services Research and Policy, 8/4: 197–201. Mill, J. S. (1987), Utilitarianism, On Liberty, and Considerations on Representative Government (London: Dent). Murphy, L. M., and Nagel, T. (2002), The Myth of Ownership: Taxes and Justice (New York: Oxford University Press). National Bioethics Advisory Commission (1999), Ethical Issues in Human Stem Cell Research (Rockville, Md.: NBAC). Rawls, J. (1995), ‘Reply to Habermas’, Journal of Philosophy, 92/3: 132–80.

174

søren holm

Rawls, J. (1996), Political Liberalism (With a New Introduction and the ‘Reply to Habermas’) (New York: Columbia University Press). Sterba, J. P. (1998), Justice for Here and Now (Cambridge: Cambridge University Press). Stock, G. (2002), Redesigning Humans (New York: Profile). Sunstein, C. R. (1995), ‘Incompletely Theorized Agreements’, Harvard Law Review, 108: 1733–72. Tranøy, K. E. (1998), Det a˚ pne sinn: Moral og etikk mot et nytt a˚ rtusen (Oslo: Universitetsforlaget). Walzer, M. (1983), Spheres of Justice: A Defence of Pluralism and Equality (Oxford: Blackwell). Zoloth, L. (2002), ‘Jordan’s Banks: A View from the First Years of Human Embryonic Stem Cell Research’, American Journal of Bioethics, 2/1: 3–11.

chapter 7 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

TIERS WITHOUT TEARS: THE ETHICS OF A TWO-TIER HEALTH CARE SYSTEM •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

benjamin j. krohmal and ezekiel j. emanuel

The American health care system is a mess. The cost of the system has increased to nearly $2 trillion, over 16 per cent of the GDP (Levit et al. 2004; Smith et al. 2006). At least 45 million Americans are uninsured, while over the last few years more than 5 million jobs have lost health coverage (Holahan and Cook 2005). The demands of Medicaid are forcing the states to cut other services, especially education, forcing tuition hikes at public universities, and prompting states to consider ending Medicaid entirely (Coughlin and Zuckerman 2005; O’Dell and Goodwin 2005). The Medicare Trust Fund will be empty by 2020, and, according to the Medicare Trustees, enormous change is needed immediately to establish fiscal balance in the next seventy-five years: cutting benefits by 48 per cent or more than doubling the Medicare tax (Snow et al. 2005). While it may not be inevitable or imminent, escalating insolvency makes health care reform seem more probable. Indeed, leading conservatives including former Senate majority leader Bill Frist, Representative Bill Thomas, former chair of the House Ways and Means Committee,

176

benjamin j. krohmal and ezekiel j. emanuel

and Bill McGuire, chief executive of United Health Group have publicly called for comprehensive health system reform to repair inefficiency and under-coverage in the current health care system (McGuire 2004; Thomas 2004; Frist 2005). American health care reformers face a number of ethical issues, including familiar debates over the merits of a single-payer system and publicly provided universal health insurance. No matter how these debates are resolved, a further ethical question must be addressed. Both universal coverage and a single-payer system are compatible with permitting some patients to pay more for faster, better, or more health care choices. Should the United States continue to have a two-tier health care system in which wealth grants some patients access to medical services that others with the same needs cannot obtain? Critical evaluation of both principled objections to inequalities and practical objections to anticipated social and medical consequences of a two-tier health care system are needed.

Divergent Visions ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

In the United States public health coverage is currently limited. Most Americans rely upon whatever insurance they receive from their employer or can afford to purchase on their own. Many American insurance products are specifically designed with different tiers of service. For more money, people can access more services, have limited prior approvals and gate-keeping, and have greater choice of physicians and hospitals. Israel, the Netherlands, and the United Kingdom also have two-tier systems. Each provides an extensive public tier of health care that can be supplemented with market tier coverage. Patients may pay extra for private hospital rooms or longer stays, more treatment options, and faster access by jumping queues for procedures with long waiting times, like hip replacement or cataract surgery. In Israel as much as 70 per cent of the population purchases additional coverage (Boaz Lev, personal communication, 2006). Norway and Canada strive for a single tier of health care, though, short of draconian restrictions, completely eliminating the influence of wealth in health care may be impossible. For instance, wealth allows the affluent to obtain medical services out of country that are unavailable to their compatriots. Nevertheless, Norway and Canada each have explicitly egalitarian systems that prohibit a private market for any services covered by their universal national health care plans. However, in the 2005 decision Chaoulli v. Quebec, Canada’s Supreme Court ruled some of Canada’s egalitarian restrictions on health care distribution to be unconstitutional, requiring the expansion of private markets for health care. Ethical controversy in Canada pits those who support ‘equal access for equal needs’ against those who defend a

tiers without tears

177

tiered health care system that allows some medical services to be distributed by the market (Flood and Sullivan 2005). Reformers in America must consider whether it just for a system to allow a market tier of health care services, a tier in which a set of services or amenities are available only to those who can afford them. Do Canada and Norway’s prohibitions on payment for greater access to medical care provide an ethical model that the United States should follow? Does the two-tier British and Israeli model offer a more just alternative? What are the ethics of a two-tier health care system? The remainder of this chapter reviews the main ethical arguments against a twotier health care system, and then presents a rebuttal of those arguments (Table 1). The chapter concludes that justice not only permits but requires America to retain two tiers of health care, though significant changes are required in order to meet five criteria for an ethical two-tier system.

Objections to a Two-Tier System ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Opponents of a market tier usually begin their objections by noting that health care is special. Health is ‘a necessary condition for pursuing nearly all the goals around which we organize our lives’ (Brock and Daniels 1994). Health care is a universally recognized need and is central to the opportunities involved in leading a normal human life (Scanlon 1975; Walzer 1983; Daniels 1995). Without good health it is difficult to achieve and enjoy valuable ends: rich family life, meaningful work, the arts, nature, athletics, hobbies, and leisure. This makes access to health care more important than access to other resources. As Descartes claimed more than 400 years ago, ‘the preservation of health is undoubtedly the first good and the foundation of all the other goods in this life’ (Descartes 1988: 143). While market influences on many goods may lead to inegalitarian outcomes that are nevertheless ethical, such inequality is unjust when it comes to matters as fundamentally important as health. This conclusion is often left to stand on its own; however, it might be motivated by several considerations:

Only Need Matters Bernard Williams argues that wealth is morally irrelevant when it comes to distributing health care (1962). Justice, notes Williams, requires treating relevantly similar cases alike, and relevance, he claims, is determined by the goal of the activity in question. In hiring a lawyer, the goal is finding qualified representation, so experience is a relevant difference but race is not. In grading students, the goal is to assess academic achievement, so demonstrated knowledge is the relevant difference

178

benjamin j. krohmal and ezekiel j. emanuel

TABLE 1 Objections to a two-tier health care system and replies Objection

Description

Response

Only need is relevant

Need should be the basis for distributing health care. Wealth is not a relevant basis for the distribution of important services

Justice requires providing a certain level of medical services to all, but the liberties of health care workers cannot be ignored. Since providers have valid claims to compensation, patients’ need cannot be the only morally relevant consideration

Level of opportunity

Health must be restored for people to have a fair level of opportunity. This requires ensuring universal access to all medical services, or exhausting all public and private funds in the attempt

Health is not the only important good, and beyond a certain point further taxation and the resulting economic inefficiency limits more opportunity than it restores

Equality of opportunity

Only an equal share of opportunity is fair, and medical benefits increase opportunity. Therefore, there must be equal access to medical services

The priority of the health care system should be improving health, not equalizing it by lowering everyone to the level society can afford

Difference principle

Inequalities are only just when they are to the benefit of the worst off. A market tier of health allows inequality that does not benefit the worst off

The difference principle permits people to earn and spend unequal shares of wealth on desirable goods because it provides incentive for economic efficiency that benefits the worst off

Public good

Public funds extensively subsidize the health care system, so medical services must be equally available to all members of the public

Public health care subsidies are partial, and even full public ownership does not require that services be free or equally available

Undermining

A market tier will undermine access to health care for the disadvantaged by reducing physicians’ incentive to treat the poor and by limiting the incentive of the upper class to support adequately funding public health coverage

Undermining is avoided by meeting the five criteria outlined for an ethical health care system, including public funding for universal non-means tested coverage with an adequate core benefits package

but sex is not. In treating a patient, the goal is preserving and restoring health, so medical need is the relevant criterion for distribution, but wealth, race, and sex are not. A market tier is ‘straightforwardly the situation of those whose needs are the same not receiving the same treatment, though the needs are the ground of the

tiers without tears

179

treatment’ (Williams 1962: 122). Like racial or sex discrimination, a market tier is unjust because it allows different treatment for relevantly similar patients.

Fair Opportunity A strong entitlement to health care is most often defended by appealing to the principle of fair or equal opportunity (Daniels 1995). Rawls famously defends one version of this principle, which states that before any market produced social or economic inequalities can be just, they must be ‘attached to offices and positions open to all under conditions of fair equality of opportunity’ (1999: 266). Rawls argues further that society must take positive steps to ameliorate ‘social contingencies and natural fortune’ that arbitrarily stand in the way of opportunity (1999: 63). Ill health seems to be one component of natural fortune that arbitrarily limits opportunity. Someone who is sick will not have the same opportunity to attain advantaged offices or positions as someone who is healthy. Fair equality of opportunity may be construed with emphasis on the level of opportunity or emphasis on the equality of opportunity. Daniels takes the former approach to health care, and Rawls has endorsed his general method (Rawls 1996: 184). Daniels holds that there is an absolute level of health—‘normal species functioning’—that allows for fair opportunity when met or exceeded. Some might argue that if species’ typical functioning is required for fair equality of opportunity, economic inequalities cannot be just until society does everything possible to eliminate deficits in normal species functioning; in effect to eliminate ill health. After meeting this requirement, there will either be no medical advantages over the public tier that a market tier could provide, or there will be no disposable wealth left with which to pay for them. Alternatively, some hold that what matters for fairness is not the level of opportunity but rather that everyone have an equal share of opportunity. To the extent that health care improves health and increases opportunity, allowing a market tier of health care is unjust because it enables those who are already privileged to achieve an even more unequal share of opportunity.

The Difference Principle Rawls argues that once fair equality of opportunity has been achieved to the extent possible, inequalities are regulated by what he calls the ‘difference principle’. The difference principle states that economic and social inequalities must be ‘arranged so that they are . . . to the greatest expected benefit of the least advantaged’ (Rawls 1999: 1972). Unequal distributions are unjust unless the worst off are better off with the inequality than without it. Allowing the affluent to pay for greater access to a good as important as health care is unjust because it produces significant inequality without benefiting the poor. As one commentator explains Rawls’s position, a

180

benjamin j. krohmal and ezekiel j. emanuel

market tier ‘effectively severs the moral and economic bonds linking the ‘‘have nots’’ with the ‘‘haves’’. . . Whereas Rawls insists that any inequalities enjoyed by the rich should also benefit the poor’ (Arras 1981: 32). Non-mutually beneficial inequalities of this kind are made worse by undermining the self respect of the less affluent (Gutmann 1981; Dickman 1983; Daniels 1998). Social cooperation and solidarity fosters a sense of self worth, and flourishes when the advantages of the most fortunate serve also to bolster the standing of the least. On the other hand, a market tier in health services distances the fate of the ill and wealthy from the fate of the ill and poor, diminishing social solidarity. A means tested public tier of health care imposes further indignity by requiring proof of poverty before granting access to services.

A Public Good Two further objections are raised against a market tier of health care. Some have argued that health care workers owe their products and services to all, regardless of patients’ ability to pay, because of significant public subsidies of the health care system. In the United States tax revenue and deductions account for nearly half of all spending on health care (Smith et al. 2006). The American health care system is subsidized at every level—through funding research, construction of medical facilities, supplementing medical education, public health measures, and provision of medical services. Critics of a market tier object to making health care a public good in its creation and funding but a private, market good when distributed. It is claimed that ‘so long as communal funds are spent, as they currently are . . . the services that these expenditures underwrite must be equally available to all citizens’ (Walzer 1983: 90). And ‘it is unethical [for a physician] to take publicly-provided skills and use them on behalf of only those who can pay him whatever he charges, as though his education and skills were entirely his own to do with as he likes’ (Marcia Angell, personal communication, 2004). Those involved in the health care system are indebted to the public, and obligated to ensure that all medical services are equally available to all patients.

Undermining the Health Care System Finally, many opponents of a market tier emphasize a practical argument: A market tier inevitably makes the level of health care services provided to the poor inadequate. It is an old adage in American politics that programs for the poor become poor programs because when the relatively affluent are allowed to purchase more, they have no reason to support funding an adequate public tier. At the same time as a market tier undermines funding for an adequate public tier, it produces greater demand for the best medical practitioners to treat the richest patients,

tiers without tears

181

‘decreasing the quality of medical care received by the majority of citizens confined to the publicly funded sector’ (Gutmann 1981: 552). Rejecting a market tier is necessary to ensure the poor and others in the public tier get adequate coverage.

Defense of a Two-Tier System ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Defenders of a market tier usually begin their arguments by noting that public funds are scarce. Though the United States is a wealthy nation and can afford to spend massive amounts on health care, the fiscal instability of Medicare and Medicaid serves as a reminder that health care is expensive, and resources are inherently finite. Even wealthy nations must make tradeoffs—overtly or otherwise—between funding for health care, economic stability, and funding for other vitally important goods that government has obligations to provide: infrastructure, food, housing assistance, the environment, security, law enforcement, and education. America can no more afford to fund every possible medically beneficial service than it can afford to fund every possible highway project, educational opportunity, or security measure. This is not to say that health care costs might not be lowered by eliminating waste or fraud, or that the current level of taxation and distribution of funds is ideal. But even a rational economy would have to make tradeoffs in which needs are weighed against needs (Daniels 1995). Were health care available in unlimited abundance, a national health care system could provide comprehensive coverage: all medically beneficial services for everyone at any time. Unfortunately, inescapable scarcity prevents everyone from getting everything they may need or want (Eddy 1994)—which, as Rawls argues—is precisely why justice is needed to determine those goods to which people are entitled (Rawls 1999). In a nation as wealthy as the United States, there is considerable consensus that people are entitled access to a broad spectrum of health care that is both fundamentally important and compatible with providing other vital goods. But while health is clearly important, this ‘does not mean that all health care is of equal importance or that every beneficial service must be provided no matter how small or unimportant the benefit and how high the cost’ (Brock and Daniels 1994).

Efficiency and Incentive Scarcity and the need for tradeoffs lead Daniels to reject interpretations of his opportunity view as requiring comprehensive health care services (Daniels 1998). In making tradeoffs between public services, absolute priority should not be given to health care. This is especially true when the same amount of resources that could restore a marginal amount of species’ typical functioning could protect more

182

benjamin j. krohmal and ezekiel j. emanuel

opportunity for more people if spent on housing or education or some other good. After providing some level of services, society will reach a point at which further taxation and expenditure on services to promote opportunity becomes self defeating—a point at which economic inefficiency limits opportunity more than further services would restore and enhance it. A commitment to increasing opportunity requires not taxing every available dollar for spending on health care and other public services. This leaves medical services that are not socially guaranteed to be included in a market tier, and allows the retention of private wealth needed to pay for them. A similar appeal to economic efficiency is at work in applying the difference principle. Rawls recognizes that allowing some inequalities is to the benefit of the disadvantaged because the prospect of greater wealth provides an incentive for greater productivity and efficiency (1999: 68). But do the disadvantaged benefit in any way if the affluent are allowed to spend their legitimately earned wealth on health care? On the face of it, there may be no direct or obvious way that the poor benefit when the rich are allowed to spend their money on most goods. This, however, is no truer of orphan drugs than of riding lawnmowers, and the difference principle permits the wealthy to spend on either for the same reason: allowing the expenditure of wealth on desirable goods is an incentive for productivity that is to the benefit of the worst off. Provided that spending money on a market tier of health care is no worse for the poor than stuffing it under a mattress, a market tier is consistent with the difference principle. The difference principle may even provide reasons to prefer that the affluent spend on market tier health care than on luxury items: additional health benefits might make the wealthy even more productive, permitting a market tier is likely to stimulate medical innovation, and the expenditure of private funds instead of tax revenue for some medical services saves the public tier money that it can devote to expanded coverage for the poor (Fried 1976; Hoel and Saether 2003).

Greater Health, Not Equal Health A two-tier system is not compatible with a fair opportunity principle that requires equality of opportunity, but such a principle should be rejected, especially when applied to health care. In almost all cases, access to medical care is limited by fiscal scarcity, not by scarcity of material goods, like scarce vaccines or transplantable organs. When patients pay for brand name drugs or an MRI scan, they do not in any meaningful way exhaust the amount remaining; ‘as much and as good’ is left for others. Further, when fiscal scarcity limits access to a medical service, access to that service cannot simply be redistributed from those who can pay to those who cannot. In such cases, why should we prefer that the affluent spend on another luxury car than on an extra MRI? Should we also prohibit payment for other

tiers without tears

183

meaningful goods that not all can afford—like side impact airbags or Ivy League educations? It may be reasonable to prefer that, when possible, access to scarce resources be granted to the poor. But when this is not possible, as when cost is the limiting factor for access, it is perverse to regard the receipt of good things by the affluent as an evil to be avoided. To do so runs counter to the presumptive goal of the health care system to increase the health of the population, not to make everyone equally healthy, especially when this means limiting medical benefits by prohibiting access to care. The idea that we should increase equality by making some worse off and none better—the notion that if everyone cannot have something then no one can—is the basis of the powerful ‘leveling down’ objection to strict egalitarianism (Parfit 2000). Increasing equality by bringing everyone down to the level that society can afford deprives the rich of liberties and benefits without regard to helping the poor. Many share the goal of increasing equality by raising the level of health care available to the disadvantaged. However, there is at least the appearance of envy and resentment in preferring that some medical services benefit nobody than that they benefit those who are willing and able to pay for them. A just health care system focuses on whether all have enough, not on preventing the rich from getting more.

Costs are Relevant Even if medical services need not be distributed equally, however, it is still of concern that medical services be distributed on an ethically relevant basis. Patients’ ability to pay is one morally relevant consideration for determining the distribution of care. The argument that the only relevant ground for distributing services is the potential recipient’s need leads to the notorious conclusion that ‘the only proper criterion for the distribution of barbering services is barbering need’ (Nozick 1974: 234). The contention that only need is relevant fails to recognize that service providers, and not just recipients, have legitimate claims and interests. Saint-like selflessness may be a greater virtue in medicine than in barbering, but professional medical ethics do not demand, nor would it be just for the state to require, that physicians never consider compensation in accepting patients for treatment. While there is room for disagreement over what constitutes fair compensation for medical services, whether fair compensation is provided is clearly a relevant ethical concern.

No Unrestricted Access to Partial Public Goods Health care providers should be expected to provide some measure of public service in recognition of their reliance upon public support of the health care system (Astor

184

benjamin j. krohmal and ezekiel j. emanuel

and Sreenivasan 2003). However, public funds have been invested in the American health care system without any agreement from or expectation that recipients of funds ensure equal access to all subsidized products and services. Further, it is not reasonable to think that participation in a system 45 per cent funded with public dollars requires health care workers to provide 100 per cent of their products and services regardless of whether patients can pay. Even if the public fully owned the health care system, Americans are reminded each time they pay the postal service for shipping that public ownership of a service provider does not require free or equal availability of the service.

Freedom of Choice In a free society, people have the liberty to spend more of their money on enormous houses, designer clothing, travel, personal computers, and the arts. An argument claiming that private hospital rooms, shorter waiting times for cataract surgery, or Acyclovir for oral herpes are too important to be distributed by the market would, seemingly, also rule out the free exchange by market of cell phones, gourmet food, sporting goods, fancy cars, and innumerable other non-medical goods and services. At the margin, health care is simply not more important than many commodities that people should be free to buy and sell. Among more important medical services, some require too great a sacrifice of other necessary goods to justify public funding. Education and security are fundamentally important, but this does not mean that everyone is entitled to publicly funded postdoctoral education or a personal police escort. Similarly, justice does not require socially guaranteed access to full mouth reconstruction instead of dentures, Macugen for macular degeneration, lung volume reduction surgery for emphysema, or implantable cardio defibrillators. In distributing scarce public funds, justice requires consideration of the cost and value of medical services relative to other important goods. Respect for individual liberties and autonomy leads to an inherently different standard for private spending (Fuchs 2004). While government must avoid spending disproportionately for the value it receives in order to make prudent tradeoffs, individuals are free to spend in accordance with their own priorities. Indeed, principles of justice are devised precisely to determine what people should be socially guaranteed and what should be left to them to purchase with their own money based on their own values. Some people who are risk averse may prefer to buy coverage for every available health care service; others will prefer cooking lessons or skiing trips to a wider range of doctors or more vision coverage. So long as the objects of public and private spending are held to different standards—the state required and entrusted to make prudent policy level tradeoffs while the individual is free to pursue personal projects—there will inevitably be some medical services, including some with a substantial impact

tiers without tears

185

on health, that can and should be available only if patients pay for them. As a matter of individual liberty, patients should be allowed to pay for these medical services, and physicians who have worked and studied hard should be free to meet the demand for additional medical goods (Gutmann 1981; Ames 2003).

Designing an Ethical Tiered System ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

If the current US health care system is unjust, it is not because it has a market tier, but because it fails to ensure access to adequate health care for all Americans. The proper standard for evaluating the justice of the health care system is not whether everyone has access to the exact same services, or whether the rich can buy more services, but whether everyone is guaranteed, in a sustainable manner, an adequate set of benefits—a core benefits package. In a nation with America’s resources, universal coverage with a core benefits package is affordable, morally imperative, and compatible with allowing a market tier of health care (Emanuel and Fuchs 2005). A market tier would not simply be an add-on operating in parallel to an adequate public system. There are complex interactions between the market tier and the public tier that have unfortunately been subject to little empirical analysis (Hurley 2001). Largely without the benefit of data, critics have nonetheless reasonably speculated that a market tier could undermine public coverage for the poor. On the other hand, there is good reason to think that steps can be taken that would prevent a market tier from weakening the public system, and potentially even strengthen it by increasing medical innovation and reducing the public burden of paying for the health care of the affluent (Fried 1976; Hoel and Saether 2003). Problems arise in America’s two-tier system because the public tier is viewed as a safety net for the unlucky few who slip through the cracks in what is regarded as normal market tier health coverage. Consequently, few have a stake in ensuring an adequate minimum tier, and doctors have an incentive not to treat public tier patients. The solution is for the public tier to provide a solid floor that serves as the norm for health coverage, with the market tier serving as a ladder upward for those who want and can afford more. A sustainable and ethical tiered health care system should meet five criteria: 1. The core benefits package should cover an adequate level of health care. 2. The core benefits package should be guaranteed to all Americans, without means testing. 3. The core benefits package should be designed to attract a sizable majority of the population to use it without supplementation to the market tier. 4. Purchases of market tier services and coverage should be made with after tax dollars and should not provide exemption from tax obligations to support the core benefits package.

186

benjamin j. krohmal and ezekiel j. emanuel

5. Easy adjustment of the core benefits package should be possible in response to changes in technology, data about efficacy, and demand for market tier services. A tiered system that meets these criteria would dramatically decrease the incentive for doctors to serve only rich patients. With everyone covered by adequate health insurance, treating patients from the ‘lowest’ tier would no longer be a financial burden, and limiting a practice to the market tier would rarely be fiscally sound; only a small minority of physicians could be financially successful catering to the ‘rich and famous’. Political difficulties would also be avoided by making the system non-means tested, ensuring that a substantial majority of citizens have a stake in socially guaranteed service. A universal core benefits package would offer benefits to everyone, and even the minority who choose to pay for market tier services will pay less for supplemental health insurance than they would for a full health insurance plan. While means tested welfare programs face an uphill battle for popular and political support, non-means tested Medicare and Social Security enjoy broad popularity, despite the ability of recipients to upgrade with Medigap insurance and private retirement savings. Allowing a market tier would be a political strength rather than a liability (Pauly 2004). Any concerns about the impact of tiering on social solidarity and self respect should also be allayed by a system in which the majority utilizes the public tier for their healthcare needs. Indeed, Americans prefer and regard two tiers with an adequate minimum as the most just structure for the health care system (Frohlich and Oppenheimer 1992). Eliminating means testing circumvents the need for patients to prove that they are poor in order to receive benefits, and stigma associated with the public tier would cease when the public tier is the norm. The criteria offered are general and require elaboration. What is an adequate core benefits package? It should begin with services that, given a fair share of society’s wealth, many people would pay out of pocket to cover for themselves (Eddy 1991; Nord 1996; Dworkin 2000: 307–50; Danis et al. 2002). Such a core benefits package is likely to ensure that a small proportion of Americans buy market tier services. Importantly, determining the threshold level of participation for a sustainable universal public tier will be a challenge for policy makers who have speculated but produced little data on the topic (Rashi Fein, personal communication, 2005). While encouraging widespread use of a publicly provided core benefits package should not be difficult, the requisite threshold level of participation can be sustained by providing incentives to rely upon the core package and disincentives to buy into the market tier: eliminate the current tax exemptions for purchasing more health coverage, ensure diverse choices in the core benefits package, keep co-pays low in the core benefits package, or potentially tax higher tier medical care. A single-tier

tiers without tears

187

system is not required to ensure adequate health care for the poor. Indeed, of the top ten rated health care systems in the United Nations’ equality oriented 2000 World Health Report, all ten allow a significant market tier (World Health Organization 2000).

Conclusion ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

What sort of replacement should Americans want for their troubled health care system? Many who are rightly concerned with the ethical foundations of the American health care system look to Canada and Norway’s prohibition of a market tier of health care as an ethical model for the United States to emulate. This egalitarian approach, while well intentioned, is unjust. People should not be concerned that some of their neighbors might have ‘too much’ access to health care, but rather that many of their neighbors might have too little. A better example is provided by Britain, the Netherlands, and Israel’s two-tier systems with universal public coverage. Limited resources prevent the public from paying for access to all health care for all people at all times. Allowing a market tier of medical services above an adequate but necessarily finite universal public tier preserves freedom of choice and achieves more equitable health care by improving access for the disadvantaged, not by limiting access for the affluent. Many questions remain for American health care reformers, but on the question of tiering, justice calls for a two-tier health care system.

References Ames, J. T. (2003), ‘Proposals for US National Health Insurance’, JAMA 290/21: 2798. Arras, J. D. (1981), ‘Health Care Vouchers and the Rhetoric of Equity’, Hastings Center Reports, 11/4: 29–39. Astor, A., and Sreenivasan, G. (2003), ‘Providing Free Care to the Uninsured: How Much Should Physicians Give?’, Annals of Internal Medicine, 139/9: W78. Brock, D. W., and Daniels, N. (1994), ‘Ethical Foundations of the Clinton Administration’s Proposed Health Care System’, JAMA 271/15: 1189–96. Coughlin, T. A., and Zuckerman, S. (2005), ‘Three Years of State Fiscal Struggles: How Did Medicaid and SCHIP Fare?’, Health Affairs (Millwood), W5385-98,

, accessed 27 June 2006. Daniels, N. (1995), Just Health Care (New York: Cambridge University Press). (1998), ‘Rationing Medical Care: A Philosopher’s Perspective on Outcomes and Process’, Economics and Philosophy, 14: 27–50. Danis, M., Biddle, A. K., and Dorr Goold, S. (2002), ‘Insurance Benefit Preferences of the Low-Income Uninsured’, Journal of General Internal Medicine, 17/2: 125–33.

188

benjamin j. krohmal and ezekiel j. emanuel

Descartes, R. (1988), The Philosophical Writings of Descartes, ed. J. Cottingham, R. Stoothoff, and D. Murdoch, i (Cambridge: Cambridge University Press). Dickman, R. (1983), ‘Operationalizing Respect for Persons: A Qualitative Aspect of the Right to Health Care’, in R. Bayer, A. Caplan, and N. Daniels (eds.), In Search of Equity: Health Needs and the Health Care System (New York: Plenum), 161–82. Dworkin, R. (2000), Sovereign Virtue: The Theory and Practice of Equality (Cambridge, Mass.: Harvard University Press). Eddy, D. M. (1991), ‘What Care Is ‘‘Essential’’? What Services Are ‘‘Basic’’?’, JAMA 265/6: 782, 786–8. (1994), ‘Clinical Decision Making: From Theory to Practice. Principles for Making Difficult Decisions in Difficult Times’, JAMA 271/22: 1792–8. Emanuel, E. J., and Fuchs, V. R. (2005), ‘Health Care Vouchers: A Proposal for Universal Coverage’, New England Journal of Medicine, 352/12: 1255–60. Flood, C. M., and Sullivan, T. (2005), ‘Supreme Disagreement: The Highest Court Affirms an Empty Right’, Canadian Medical Association Journal, 173/2: 142–3. Fried, C. (1976), ‘Equality and Rights in Medical Care’, Hastings Center Reports, 6/1: 29–34. Frist, W. H. (2005), ‘Shattuck Lecture: Health Care in the 21st Century’, New England Journal of Medicine, 352/3: 267–72. Frohlich, N., and Oppenheimer, J. A. (1992), Choosing Justice: An Experimental Approach to Ethical Theory (Berkeley: University of California Press). Fuchs, V. R. (2004), ‘More Variation in Use of Care, More Flat-of-the-Curve Medicine’, Health Affairs (Millwood), suppl., VAR-104–7,

, accessed 27 June 2006. Gutmann, A. (1981), ‘For and Against Equal Access to Health Care’, Milbank Memorial Fund Quarterly Health and Society, 59/4: 542–60. Hoel, M., and Saether, E. M. (2003), ‘Public Health Care with Waiting Time: The Role of Supplementary Private Health Care’, Journal of Health Economics, 22/4: 599–616. Holahan, J., and Cook, A. (2005), ‘Changes in Economic Conditions and Health Insurance Coverage, 2000–2004’, Health Affairs (Millwood), W5-498-508, , accessed 27 June 2006. Hurley, J. (2001), ‘Ethics, Economics, and Public Financing of Health Care’, Journal of Medical Ethics, 27/4: 234–9. Levit, K., Smith, C., et al. (2004), ‘Health Spending Rebound Continues in 2002’, Health Affairs (Millwood), 23/1: 147–59. McGuire, W. W. (2004), ‘Business Opportunities in Transforming Health Care: An Interview with William W. McGuire’, Health Affairs (Millwood), 23/6: 114–21. Nord, E. (1996), ‘Health Status Index Models for Use in Resource Allocation Decisions: A Critical Review in the Light of Observed Preferences for Social Choice’, International Journal of Technology Assessment in Health Care, 12/1: 31–44. Nozick, R. (1974), Anarchy, State, and Utopia (New York: Basic Books). O’Dell, K., and Goodwin, J. (2005), ‘Work to Welfare Jarring Change’, Springfield News-Leader (Springfield, Mo.), 28 Aug., 1. Parfit, D. (2000), ‘Equality or Priority?’, in M. Clayton and A. Williams (eds.), The Ideal of Equality (New York: St Martin’s Press), 81–125. Pauly, M. V. (2004), ‘Conflict and Compromise Over Tradeoffs in Universal Health Insurance Plans’, Journal of Law, Medicine and Ethics, 32/3: 465–73. Rawls, J. (1996), Political Liberalism (New York: Columbia University Press).

tiers without tears

189

(1999), A Theory of Justice, rev. edn. (Cambridge, Mass.: Belknap Press). Scanlon, T. M. (1975), ‘Preference and Urgency’, Journal of Philosophy, 72/19: 655–69. Smith, C., Cowan, C., et al. (2006), ‘National Health Spending in 2004: Recent Slowdown Led by Prescription Drug Spending’, Health Affairs (Millwood), 25/1: 186–96. Snow, J., Chao, E., et al. (2005), Status of the Social Security and Medicare Programs: A Summary of the 2005 Annual Reports, Social Security and Medicare Boards of Trustees (Washington, DC); 2006 version available at

. Thomas, B. (2004), Vision for Health Care (Washington, DC: National Center for Policy Analysis). Walzer, M. (1983), Spheres of Justice: A Defense of Pluralism and Equality (New York: Basic Books). Williams, B. (1962), ‘The Idea of Equality’, in W. G. R. Peter Laslett (ed.), Philosophy, Politics and Society (Oxford: Blackwell), 110–31. World Health Organization (2000), The World Health Report. Health Systems: Improving Performance (Geneva: WHO), p. v.

chapter 8 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

JUSTICE AND THE ELDERLY •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

dennis mckerlie Ethical concerns about the elderly are usually treated as a matter of justice. This is partly explained by the special status of the elderly in almost all economically developed societies. Public pension systems provide them with financial support, and publicly funded health care programs help to meet their health care needs. Those who are now old will have contributed to these institutions when they were younger, but it remains true that the elderly receive a degree of support from public institutions that the members of other age-groups typically do not receive. This institutional background naturally raises questions about distributive justice. Are the elderly receiving less, or more, than their fair share of health resources and economic wealth? Moral philosophers have proposed many different theories of justice. However, it is difficult to apply their theories to the elderly. The elderly constitute a particular age-group. They are identified as people in a certain temporal stage of their lives, old age. And philosophical theories of justice are typically concerned with people’s complete lives or lifetimes, not with temporal parts of lives. For example, John Rawls’s influential account of justice is concerned with achieving fairness between different people’s lifetime expectations of receiving primary goods (Rawls 1971). Thomas Nagel comments, ‘Remember that the subject of an egalitarian principle is not the distribution of particular rewards to individuals at some time, but the prospective quality of their lives as a whole, from birth to death’ (Nagel 1991: 69). These theories compare the complete lifetimes of different people to decide whether they have been treated fairly. Because of their broad temporal scope, these theories cannot directly answer the question of whether a person

justice and the elderly

191

receives a fair share of income or health care during a particular temporal part of her life. One response is just to accept that the appropriate temporal scope for a principle of justice is a complete life. I call this the ‘complete lives view’. If we make this choice we cannot immediately object to treating the elderly differently from the members of other age-groups, so long as the treatment is maintained consistently over time so that everyone is treated in that way at the appropriate temporal stage of their life. For example, a policy of mandatory retirement could not be automatically criticized as unfair. If everyone at least potentially benefits from the policy when they are young, and everyone faces the same termination date of employment when they are old, the policy will not create inequalities between people’s complete lives. The same would be true, for the same reasons, of a policy restricting the use of kidney dialysis to those below a certain age. However, this does not mean that the complete lives view has nothing to say about fairness and the elderly. Certain ways of treating the elderly might be condemned as unfair because they are likely to lead to people being treated unfairly in terms of their complete lives. For example, many public pension systems—including American social security—have a degree of progressivity built into them. Poorer people contribute less to the institution than better off people, but they receive comparatively greater (although not absolutely greater) benefits in return for their contribution. A possible justification for the progressivity is that it reduces the inequality that would otherwise exist between people’s complete lives. So a non-progressive pension system might be regarded as unfair. Also, someone who holds the complete lives view can give other reasons—reasons that are not a matter of justice—for a particular form of income support or health care for those who are old. Someone who thinks that justice focuses exclusively on complete lives would probably suggest that once the requirements of justice for complete lives are satisfied we should divide resources between age-groups in whichever way would benefit people the most. For example, making a particular form of expensive health care available to the very old will be justified if this use of limited resources improves people’s lives more on balance than the alternative policy of concentrating the resources on younger people. According to this view, denying that form of care to the elderly would not be an injustice, but it would be the wrong choice to make. However, many writers believe that there are distinctive principles of justice for age-groups. For this to be the case, the principles must have a temporal scope other than complete lives. They must make claims about distribution between the temporal parts or temporal stages of lives. The most influential view of this sort has been developed by Norman Daniels, although many other writers share the same idea and apply it to a variety of questions about fairness to the elderly (see Daniels 1988, 1996; Dworkin 2000, ch. 8; Brock 1993, chs. 11, 12). Daniels calls the theory the ‘prudential lifespan account’.

192

dennis mckerlie

The theory says that the fair social distribution between age-groups should match how a hypothetical prudentially rational individual would choose to distribute the same resources over the different temporal stages of her own life. The elderly deserve, as a matter of justice, the share of resources that prudence would award to the temporal stage of old age in considering a single lifetime. The theory claims that the distribution established by prudence expresses genuine claims of justice concerned with temporal stages of lives. It differs from the complete lives view in holding that there can be injustice to an age-group even if the principles of justice for complete lives are fully satisfied (for example, even if people’s complete lifetimes are perfectly equal). However, the theory does not completely abandon the complete lives perspective. When prudence chooses a distribution over the different temporal stages of a life, its goal is to maximize the quality of the life as a temporal whole. So the distributional constraints that apply to temporal parts of lives are derived from a consideration of a complete lifetime and a concern with maximizing the overall quality of that life. Also, Daniels supposes that the prudential lifespan account operates within the constraint of a theory of justice that is concerned with lifetimes. First we must treat people fairly in terms of their complete lives. When that goal has been realized we turn to the special kind of justice that deals with distribution between the temporal stages of different lives. So the principles of justice focused on lifetimes outweigh the principles concerned with temporal parts of lives if there is a conflict. Finally, since prudence aims at maximizing the quality of a life as a whole, it seems that the distributional conclusions this theory generates will closely resemble the conclusions drawn by the complete lives view if the complete lives view does decide to distribute resources between age-groups in the maximally beneficial way. Those who appeal to prudence as the test of distributive fairness between agegroups characterize the prudential choice in different ways. Some (Dworkin 2000, ch. 8) allow the chooser to know her own age when the choice is made. Daniels (1988, ch. 3) suggests that the choice should be made behind a ‘veil of ignorance’. The chooser should not know her age, her specific needs in terms of health care and other resources, or the goals and values that she will endorse at different stages of her life. She only has generic information about the course that human lives typically take and the health problems that people are likely to face at different ages. The restrictions are designed to rule out bias and to permit a choice made by one hypothetical individual to be used to determine the just distribution over many different actual people who will have different goals and different medical problems. In principle the prudential lifespan account can provide specific answers to detailed questions about distribution. It is motivated by the recognition that we will not be able fully to meet all of the needs of all people of all ages. Even after we have done as much as we can to ensure that people are treated fairly when we compare their lifetimes, scarcity of resources might force us to choose between,

justice and the elderly

193

for example, providing one kind of health care to younger people or providing a different kind of health care to people who are older. The prudential lifespan account sees this choice as involving a distinctive kind of fairness that is concerned with distribution between temporal stages in lives. It tells us that the fair or just choice is the one that prudence would see as the most beneficial to the individual concerned if a parallel choice needed to be made inside one life. The distribution that the prudential lifespan account chooses will have two features that everyone would count as merits. First, it will not create unfairness at the level of people’s complete lives because the theory operates within the constraints of principles of justice for lifetimes. Second, the distribution will use the resources being distributed in the most efficient way to produce the greatest amount of benefits taking into account all of the people concerned. The distribution will have that second feature because the theory chooses the distribution between different people that would maximize lifetime well-being in the case of a single person. The distinctive and (I think) controversial claim of the prudential lifespan account as a theory of justice between age-groups is that fairness is simply identical to efficiency in the use of resources in the special case of justice between age-groups or justice between the temporal stages of lives. The person who would benefit most from the resources has a claim of fairness to receive them. When we consider the specific conclusions that writers have appealed to the prudential lifespan account to justify, some seem friendly to the claims of the elderly while others strictly limit the share of resources that they would receive. Daniels suggests that the theory will stipulate that people’s incomes during old age should be roughly equal to their incomes at earlier stages in their lives (Daniels 1988, ch. 7). On the other hand, in the case of health care resources, Daniels and others have argued that the prudential lifespan account would recommend (at least under certain conditions) rationing certain kinds of life-extending medical care on the basis of age (Daniels 1988, ch. 5). Some believe the prudential chooser would in general favour ensuring the availability of medical treatment during youth and middle age as opposed to old age (Dworkin 2000, ch. 8), on the grounds that treatment earlier in life would typically have a greater impact on the overall quality of a life as a temporal whole than treatment during old age. Some draw the strong conclusion that the prudential choice would prefer treatment in middle age that was not life-extending to life-extending treatment during extreme old age (Dworkin 2000, ch. 8). And some conclude that prudence would decide not to provide life-extending treatment in old age if the person were suffering from a serious form of dementia (Brock 1993, ch. 12; Dworkin 1993, ch. 8). Yet other writers are dissatisfied with both the complete lives view and the appeal to prudence (McKerlie 2002; Temkin 1993, ch. 8). Their case against the prudential lifespan account is partly based on the suspicion that its conclusions will sometimes be intuitively objectionable. They think that the theory, if rigorously applied, will tell us to give the elderly—especially those who are very old and very ill—much less

194

dennis mckerlie

than we feel they are entitled to receive. The severity of the problems faced by such people is precisely what makes it extremely difficult and expensive to promote their well-being. Prudence will recognize this in the case of a single life, and its response will be to concentrate resources in the earlier stages of the life where they can be used more effectively and efficiently to increase well-being. Also, in budgeting resources over a single lifetime prudence will realize that resources set aside or saved for extreme old age have a low probability of actually being used, since the life in question is unlikely to last that long. If this consideration is allowed to influence the prudential choice, it is another reason for favouring the earlier stages in a life. And the critics suggest that we will reject some of the strong conclusions that advocates of the prudential lifespan account themselves draw about health care and the elderly. The critics agree with the prudential lifespan account that there are constraints of justice applying to temporal parts of lives, as well as to complete lives. However, the prudential lifespan account derives these constraints from a prudential judgement about a lifetime made with a view to maximizing the overall quality of that complete life. The critics think that if people do indeed have moral claims based on the quality of their lives during particular temporal stages in their lives, there is no reason to expect that these claims will be revealed by such a prudential judgement. Because it is a judgement made about one person’s life, the prudential judgement aims simply at maximizing well-being. If we use it as the model for a just distribution across different lives, the view of justice that results will also aim at simply maximizing well-being across those different lives. However, the critics will contend that interpersonal justice—even the kind of interpersonal justice that is concerned with temporal stages of lives rather than lifetimes—does not aim at maximizing in the sense of producing the largest total amount of well-being taking everyone into account. It is more likely that the constraints of justice that apply to old age and the other temporal stages in our lives will be understood by applying some value that is more directly concerned with fairness in distribution—perhaps the value of having a quality of life that at least meets some minimum threshold, or the value of being given priority if one’s level of welfare is very low, or the value of equality—to a temporal stage of a life rather than to a lifetime. So these writers propose a theory of justice for age-groups that takes one of these values and changes its temporal scope from a complete life to a temporal part of a life. I will call this kind of theory the ‘life-stage view’. Defenders of the life-stage view should admit that considerations of justice concerned with parts of lives need to be weighed against reasons of justice concerned with lifetimes. Often the lifetime concerns will be stronger. Nevertheless, this view still differs from the prudential lifespan account. Unlike the prudential lifespan account, the life-stage view is not compelled to hold that the lifetime considerations will always be stronger than the reasons concerned with temporal parts of lives.

justice and the elderly

195

If we agree with the life-stage view about the temporal scope of principles of justice for age-groups, the next step is deciding which value to incorporate in the theory. The writers who support the life-stage view tend to choose a value that is, in a broad sense, egalitarian. No doubt this partly reflects the widespread sympathy for egalitarianism in contemporary moral philosophy, but the concern that many people feel for the elderly does seem to be egalitarian in its nature. Special institutions for helping the elderly exist because we realize how badly off they would be without them, and we are anxious to spare them from living in that degree of misery. If it does include a value like equality or priority, the life-stage view can be more generous to the elderly than either the complete lives view or the prudential lifespan account. Since it is focused on temporal parts of lives, it can tell us to help the badly off among the elderly because they are suffering now, during their old age. It does not limit their claim to what prudence would have provided for old age in the case of a single life, where prudence is governed by the aim of making that life viewed as a temporal whole as good as it possibly can be. In considering that single life, the prudential choice might see happiness during youth and extreme misery during old age and then decide to assign more resources to the temporal stage of youth because that would maximize well-being over the life as a whole. When this intrapersonal judgement is transferred to the interpersonal case of comparing the current lives of a well-off younger person and a badly off older person, the judgement would recommend helping the younger person to become even better off because she is the one who can be helped the most. A view that applies equality or priority to the temporal stages of different lives would resist that conclusion. It would compare the current state of the elderly person’s life and the current state of the younger person’s life and think that because the older person is worse off there is a reason for making them more equal now, or a reason for giving priority to helping the older person. We might decide in the end that this consideration is outweighed by considerations of fairness dealing with lifetimes, or simply by the greater size of the benefits that the younger person would receive, but the reason would still exist. The value of priority seems to be especially relevant to the elderly. The general notion of priority holds that benefiting a badly off person should be given a certain amount of priority over benefiting a person who is better off. Consequently, a smaller benefit for someone badly off can be more important than, or have more value than, a larger benefit for someone better off (the general notion of priority is explained in Parfit 1995). In the case of the elderly we think that the case for helping them is strongest when they are very badly off in the absolute sense, as well as being worse off than other people. We sometimes feel that we should attempt to do what we can to improve their lives, even if the resources that we devote to them could be used instead to make a greater improvement in the lives of other people. When we apply the notion of priority in the context of the life-stage view, we think that

196

dennis mckerlie

their needs should be given priority because they are worse off than other people now, considering the current state of their lives. They might not be the people who would count as the worst off if instead we compared people’s complete lives. A particular kind of priority has been regarded as a relevant consideration in the distribution of health care, apart from the debate between the three general approaches to justice between age-groups that I have described. When health resources are scarce, some people think that there is at least an apparent case for giving the health care to the person who needs it most. This person receives priority not because the quality of her life is in a general way worse than that of others, but because her medical condition is the most severe and the most urgent. There is disagreement about whether this reason can be strong enough to resist, at least sometimes, the consideration of efficiency, which tells us to distribute the medical resources to the people who will be helped the most by them so that we will produce the best outcome. Many writers have explained the criterion of efficiency, and applied it to specific choices that arise in medical practice. Less attention has been given to criteria that might compete with efficiency, including that of medical priority. However, Frances Kamm has defended the importance of the criterion of priority (which she calls ‘urgency’) in an especially detailed and thoughtful way (Kamm 1993, sect. iii). This issue arises independently of concern for the elderly. Nevertheless, the elderly will frequently be among those the criterion of priority or urgency would favour. Their health care problems are frequently more serious, cause more suffering, and are more debilitating. Often they would count as the worst off in the particular respect of illness. And it will often be the case that the criterion of efficiency would support offering the health care to someone else. Because of their age, and the nature of their problems, we would not achieve the best possible medical outcome by choosing to help them. In these choices there is a conflict between producing the best outcome and the criterion of priority. Writers about biomedical ethics seem typically to favour the criterion of efficiency. Perhaps this is due in part to the influence of a basically utilitarian moral outlook. The prudential lifespan approach also supports this answer in the end. However, an appeal to priority would at least provide a countervailing reason. There is also a case for helping the individual who needs help the most, even if this person can be helped less. If we agree with Kamm, we will see the choice as one where we must decide between competing ethical reasons that both are important and represent different kinds of moral concern. Apart from being worse off in a general way in terms of overall well-being, and being worse off more specifically in terms of their health problems, there is another respect in which the elderly are typically worse off than other people. They stand nearer to death than others. Old age is the final temporal stage in a life, and this fact might be thought to have ethical implications.

justice and the elderly

197

One implication again concerns the distribution of health care resources. Institutions providing health care are compelled to choose (in distributing existing resources, and at a higher level in terms of the resources that it chooses to create) between providing life-extending treatment to the elderly and important but not life-extending health care to younger people. The latter treatment does not prevent or postpone death, but it does make a person’s life better than it would otherwise have been for a certain temporal period of the life. The former treatment does prevent death, but when the treatment is provided to someone very old, death may only be briefly postponed and the quality of life during the remainder of the person’s life may not be high. Which choice should we make? To answer this question some would only consider the gain in terms of welfare that would result from each choice. In some cases, the younger person would experience a larger gain in well-being from her disease being cured than the elderly person would receive as a result of her life being extended. Then the right answer would be to treat the younger person. A theory that uses a prudential choice about a single life to answer this question would arrive at the same result. Others would find this conclusion disturbing. They feel that it underestimates the significance of death to weigh its importance simply in terms of the amount of well-being that the person in question loses by dying. They might claim that it can be more important to delay one person’s death than to improve someone else’s life, even if the second option would produce a greater gain in terms of well-being. For this view to be defensible, it must be subject to certain conditions: the postponement of death would have to be for a significant period of time, and the life the person led during this period would have to be of an acceptable level of quality. Deciding between the two views requires considering the difficult question of whether death counts as a harm or evil, and if it does, what kind of a harm it is. We need to answer the second question if we are to weigh death against other harms that people can suffer, like the harm caused by a debilitating but not life-threatening illness. There is a substantial body of philosophical literature focused on these questions (see Nagel 1979; Kamm 1993, sect. i; McMahan 2002, sect. 2). Most writers support some version of the so-called ‘deprivation view’, which maintains that the badness of death, and so its gravity as a harm, is simply a matter of the good that it deprives the person of (this is true of Nagel and McMahan). However, some think that the special nature of death as a harm is not completely captured by regarding it as a loss of welfare (Kamm). If this view is correct, then the idea of priority might apply in another way to the elderly. We might think that preventing the harm of death should be given a certain amount of priority over preventing the harm involved in illness, because death belongs in a different category of harm. Another fact about ageing raises both theoretical and practical problems in deciding what we owe to the elderly. People’s goals and values change during their

198

dennis mckerlie

lives. In the case of most of us what we want and strive to achieve during our middle years will differ significantly from our most important concerns when we are old. There are some characteristic differences between the aims of middle age and the aims of old age, although we should not expect to find those differences in every life. The differences in goals and values will be more striking when extreme old age is accompanied by failing mental powers or some degree of dementia. Changes in goals need to be taken into account when individuals make decisions that will affect their futures when their values may well be different. These changes also must be taken into account when society makes distributive choices about age-groups. This is especially so if we accept a view like the prudential lifespan account which uses a prudential decision about a lifetime as the criterion of fairness to the elderly. However, even if we endorse a different account of justice between age-groups, we might have to choose between satisfying the plans that younger people have made for their old age and helping them to achieve the goals they will actually have when they are old. This choice will be particularly difficult if we think that the goals they will aim at when they are old will be less valuable than their earlier objectives for their final years. It might seem unfair not to provide them with the kind of old age that they would relish while they lived it, but it might also seem unfair not to be guided by their own best judgement about how their lives should end. The problem of changing goals is easier to consider in the case of a choice by an individual. What should a rational person do when a choice he makes now will significantly affect his old age, and he knows that the goals that he holds now are ones that he will reject when he is old, and that the projects he will endorse when he is old are ones that he cannot acknowledge now as having value? One radical view is that changes in goals and values, and the other psychological changes that occur during our lives, are so fundamental that we should (although common sense does not) question whether it is really true that I am one and the same person in my youth and my old age (Posner 1995, chs. 4, 10–12). If we were to accept this suggestion, it would fundamentally change our understanding of our own lives and fairness to the elderly. Distribution across the different temporal stages of one life would then be appropriately viewed as morally equivalent to distribution between different people, and our views about justice between age-groups would have to be revised accordingly. I will assume that this view is too extreme, except perhaps in the case of people suffering from extreme dementia when we might literally challenge the application of the notion of personal identity. A more moderate view claims that identity as such is not what matters for rationality and morality. The application of moral values and concepts of rationality depends instead on relations of psychological continuity and connectedness between the different temporal parts of a life. This continuity will vary in strength in an ordinary life, and we should take this into account in deciding what attitude it is rational for a person to take towards his old

justice and the elderly

199

age and in making moral judgements about temporal stages in lives (such a view is defended in Parfit 1984, pt. 3, and it is applied to moral issues about the elderly in McMahan 2002, sect. 5). However, I think myself that the variations in continuity over time in an ordinary life are not important enough to be a significant factor in understanding our moral judgements and judgements about rationality. And I believe that judgements that we do make that might seem to support this view can be explained as responses to other considerations. Even if changes in character and values do not involve a breach of personal identity, or a serious weakening of the psychological relations that supposedly underlie identity, they do seem to be important for their own sake. How should we respond to them? Should I make provision for my pursuit of the goals I will hold when I am old, despite my current rejection of them? Or if there is something I could do now that would lock me into the kind of life in old age that I now think desirable, would it reasonable for me to make this choice despite my knowledge that my mind will change? This issue has been discussed in a theoretical way by moral philosophers without reference to its application to ageing (see Nagel 1970, ch. 5; Parfit 1984, pt. 2; Bykvist 2003). Three responses to it have been defended. The first claims that a reasonable person should make the present decision on the strength of his present goals and values, since those are the values that he is now committed to and believes to be valid. According to the second view, a reasonable person should decide on the strength of his future goals and values, since they are the ones he will accept when the effects of his choice on his own life will eventually be realized. The third view claims that a reasonable person should make the decision by taking both sets of values into account and giving them equal weight. He should be neutral between the conflicting values. The version of the prudential lifespan account that permits the prudential choice to be made by a young person who is aware of his present goals seems to choose the first answer. The other version of the theory appeals to a choice made by a hypothetical individual who does not know either his present or his future goals. This seems to me tantamount to requiring neutrality between present and future goals, and so amounts to a variation of the third answer. All three answers suppose that some attitude towards the importance of time itself is distinctively rational. The first view thinks that because of the special importance of the present, it is present values that matter for a decision that will be made now. The second view thinks that in assessing a change in the future, it is future values that matter because they are the values that you will hold when the change happens. This view also gives special significance to the present, but it identifies a different present as being relevant to the decision—not the time that is the present when the decision is made, but the time that is the present when the consequences of the decision will occur. The third view holds that rationality

200

dennis mckerlie

requires us to adopt temporal neutrality, to give equal weight to all times, and it supposes that this means we should also give equal weight to the different values that we hold at different times. I think that we should have reservations about accepting any of these answers. They all conflict with the belief that some goals and values can be objectively more important than others. If we agree with that belief, we cannot be satisfied with the claim that we should only use our present values whatever they are, or only our future values whatever they are, or that we must automatically give equal weight to both sets of values. If I am right, this would be another consideration that counts against the prudential lifespan account, which is committed to the first or the third answer. Arguably the best answer to the problem of changing goals is reached by thinking in terms of an agent’s well-being. If the agent decides on the strength of his present values then he will not approve of the change in his life when it happens, because he will then hold different values. And if he will not positively respond to the change, this will affect his well-being in the future or the quality of his life in the future. This gives him a reason to take his future goals into account now, even if they are objectively less valuable than the goals he now holds for his future. This suggestion appeals to what might be called the ‘positive response condition’ on well-being, an idea that has been adopted in one version or another by several recent writers on the nature of well-being (for examples, see Dworkin 2000, ch. 6; Darwall 2002, ch. 4). The point is not just that the person will feel less contentment if his life has been designed according to standards he used to endorse but has subsequently abandoned. That is the aspect of well-being that depends on hedonic goods. Some believe that the positive response condition can also matter in a different, and perhaps more important, way. They think that endorsing an activity that you perform enhances the value of the activity so that it makes a greater contribution to the quality of your life. A significant artistic achievement might increase the well-being of its creator even if she herself does not value what she has produced or her own actions in producing it. But it will do more to enhance the quality of her life if she herself appreciates what she has done. On this view, there are two factors to consider in deciding how much the activity will contribute to her well-being: the objective value of the activity, and whether she will endorse the activity when she performs it. The condition applies to choices that we must make about our future selves. A creative writer might correctly think that the most valuable goal she could pursue in old age would be to preserve her legacy as a writer. When she can no longer produce new fiction she can at least lecture on university campuses, prepare new editions of her works, and mentor young writers. Still, suppose she knows that when she is old she will instead think—wrongly—that relaxing with television, games, and socializing is more important. If her interest is in making her old age go as well as possible, she has a reason for planning her old age to fit the values that she will hold

justice and the elderly

201

when she is old. The objective value of golfing might be less than the objective value of helping young writers, but if she will positively respond to the former activity, but not the latter, playing golf might contribute more to her well-being. It may seem that the explanation in terms of well-being is identical to what I originally called the third solution to the problem of changing goals. The goals that matter are the goals the person holds at the time of the event or change in his life that we are evaluating. But the two views are not the same. The well-being view does not claim that because of the special importance of the present we are rationally required to evaluate the change in terms of the goals that the person holds when the change occurs. It explains the relevance of those goals in terms of the positive response condition on well-being, not in terms of the rationality of a particular attitude towards time itself. And the well-being explanation is compatible with the thought that if the person’s present goals are less valuable than his past goals, it might be better for his present life to match his past goals, despite the positive response condition. It has been suggested that younger people tend to disvalue or disparage the characteristic goals of old age. If so, the view that I have explained gives them a reason to change their attitude. However, it is an advantage of the view that it can lead to respect for the goals of the elderly even if we are willing to admit (as I believe many elderly people themselves would) that these projects are not as objectively important for the value of a life as the characteristic goals of middle age. There is another way—a sadder way—in which changes of goals are relevant to the ethical treatment of the elderly. The very old are sometimes afflicted by severe mental disabilities. The problems can take many different forms and arise in very different degrees of seriousness. The most tragic cases involve dementia, more specifically Alzheimer’s disease in its moderate and extreme forms. The frequency with which this disease attacks the elderly, and the increasing number of people who will live long enough so that it will pose a serious threat to them, mean that coping with the illness is now and will continue to be a major issue of public health. The illness affects characters and minds in complex ways. Impairment of shortterm memory, or rather the loss of the capacity to convert short-term memories into long-term memories, is the most familiar symptom, but radical changes in personality and deep cognitive impairments of other kinds are also experienced in more severe cases. One aspect of the problem is whether we should believe that the most problematic cases involve a change in a person’s goals or values at all. Some believe that in the worst cases the sufferer has become incapable of having goals, of having interests that take the form of valued projects rather than unreflective or automatic desires. And some believe that the afflicted person can no longer be regarded as an autonomous agent with the ability to control her life, and so cannot be regarded as possessing a moral claim to be left free to make her own choices about her own life.

202

dennis mckerlie

If this picture is correct, it obviously has profound consequences for how those who suffer from these diseases should be treated. It might lead us to give broad decision-making powers over their lives to relatives or medical personnel. If the patients had settled goals before their illness and had then expressed their wishes about questions that might arise later in their lives, it might lead us to give authority to their past decisions, even if there is some apparent evidence in their current behaviour that they are now of a different mind. And some would argue that if they lack autonomy and the capacity to hold values, this limits the value that would be gained (meaning value for those suffering from the disease themselves, not value for others) by a prolongation of their lives in a state of dementia. Here some conclude that we should not save their lives if they acquire a life-threatening but curable disease, especially if there is an advance directive from their earlier self requesting this lack of treatment. They would stand by this conclusion if the person seems placid and contented in their present state, and is not experiencing any pain or emotional distress. Sorting out these issues requires both settling difficult factual questions about precisely what capacities the people in question have and do not have, and answering broader ethical questions about the nature of autonomy and the source of its value, and about what if anything is special about endorsing something as a value as opposed to merely having a desire. This makes worthwhile and sensitive work on the problem very difficult indeed, but a number of writers have made valuable—though conflicting—contributions. Ronald Dworkin defends the strong thesis that a person afflicted with moderate or extreme Alzheimer’s disease lacks an autonomous will (Dworkin 1993, ch. 8). If before the onset of the disease she wished that her life would end without a final stage of dementia, that choice must be regarded as remaining in place as her standing will and it should determine what medical treatment she receives. Dworkin grants that the person can enjoy the circumstances of her limited life with the disease, and that if she experiences contentment, this counts as a good for her. Dworkin characterizes this as satisfying her experiential interests. However, he believes that the disease has rendered her incapable of holding values and applying them to her own life. She cannot form evaluative beliefs about her life as a whole—beliefs like her former view that her life would be better without the final state of dementia. So Dworkin considers her incapable of having what he calls ‘critical’ interests. Or rather he believes that her past evaluative beliefs, including her view about how her life should end, are still in force and constitute her present critical interests. Since Dworkin thinks that critical interests are more important than experiential interests, he thinks that allowing her life to end is in her best overall interest. According to Dworkin, the two considerations of implementing the patient’s will and doing what is best for her both support the conclusion that we should not provide her with life-extending medical treatment.

justice and the elderly

203

Dworkin’s critics contest his view of the capacities of such a patient (Jaworska 1999; Shiffrin 2004). They think that in the Alzheimer’s sufferer who is undeniably contented, we can discern something that deserves to be called a desire to go on living, even if she is no longer capable of thinking about her life as a temporal whole or capable of articulating her desire to others (at least they believe this about many sufferers from moderate Alzheimer’s disease and perhaps some patients with extreme forms of Alzheimer’s—they might concede that Dworkin’s view does fit the most extreme cases). And some also believe that when this desire is present it should be respected as representing her will. So they think that her will has changed, has taken a new direction, as a result of the disease. Respect for her autonomy should motivate us to implement what is in fact her current will, not what her will used to be. They also think she is capable of more than being pleased by her current experiential states. Her present interests are partly determined by values that she does hold now. In other words she possesses critical interests as well as experiential interests, and the critical interests that now belong to her might differ significantly from those she held before the disease. If she finds value in the conditions of her new life, these critical interests will support the continuation of her life. In effect the Alzheimer’s patient has undergone a change in her critical interests or values. Those who see the patient in this light will be tempted to think that what is in her best interest is determined by her new critical interests as well as her experiential interests, but not by the critical interests that she held in the past. So they would think that providing her with the medical treatment that will prolong her life is really what is best for her. It is not easy to decide between the two views. Most of us will be disturbed at setting aside the careful provisions that the person made, when her mind was clear, for just this eventuality. But we will also be disturbed by not offering life-extending treatment to a person who is now contented with her life and wants it to continue, even if she had previously wished not to live a life of that sort. Perhaps the issue about autonomy can be answered, assuming that we agree that the patient should be credited with an autonomous will that has as one of its objects the continuation of her life. She now wishes to live, and it seems to me that her wish should be honoured. Since her new will takes the direction it does partly because of the damage the disease has done to her, we might think that her new will is less reasonable—objectively less reasonable—than her previous will. Still, the value of autonomy counts in favour of respecting a person’s wishes—meaning by this the person’s current wishes—even if they are less reasonable than her past objectives and even if respecting them would be against her best interest. As a value, autonomy seems to operate in this ‘present tense’ way. Perhaps we would not respect a person’s current wishes if their will was grossly irrational, or disastrously opposed to their best interests. But this does not seem to be true of the Alzheimer’s patient. Given her condition, it is not grossly irrational of her to want the life she

204

dennis mckerlie

enjoys to continue, even if we think that it is against her interest for reasons that she can no longer understand. If we accept these ideas, we will feel obligated to provide the life-extending medical treatment on grounds of respecting autonomy, regardless of how we may answer other questions about her condition and interests. However, we might not find as ready an answer when we turn to the question of what would be in the patient’s best interest. Our judgement about the patient’s good might differ from our judgement about her autonomy, because autonomy and considerations about a person’s good are different moral reasons that might work in different ways. In considering the Alzheimer’s patient’s best interests, the crucial thing is the status of the critical interest represented by the patient’s past conviction that her life would be better if it ended without the postscript of a period of dementia. Does this conviction still apply to her life after she has been changed by the disease in precisely the way that she feared? On the one hand, it is clear that she no longer explicitly adheres to the conviction. We can suppose that she could not even begin to understand this belief if it were expressed to her. And I would agree with Dworkin’s critics that she may now have other interests, no doubt of a much simpler sort, that nevertheless count as critical interests because they involve a kind of valuing—for example, her desire for the company of individuals who love her, even if she might be uncertain of their identities or their role in her past life. Some would conclude that her critical interests have simply changed. The critical interest that Dworkin appeals to has been replaced, and it should not be applied to an assessment of her life with the disease. It is not relevant to a judgement of the value or quality of that life, or in deciding how much well-being she can enjoy. If we think of the change in the patient in this way we will suppose that her life will be better if it continues, so long as she is contented and the limited but important interests that she still possesses are satisfied. This is how we react to a self-conscious change of values during an ordinary life. We appeal to the person’s new values, not to the supplanted values, to determine what is in the person’s best interest from now on. We do this even if we think that the new values are less reasonable than the old ones, or that the person changed his mind for bad reasons. However, this comparison should lead us to challenge the view. The patient did not change her mind, her mind was changed by the disease. She did not reconsider the question of the shape she wanted her life to have and then decide (perhaps for inadequate reasons) that the years of dementia would not after all make her life worse. Rather, the disease took away from her any capacity to even think in those terms. The cause of the change in her—Alzheimer’s disease—is a disease or impairment, a condition that does not merely initiate changes in a non-rational way but destroys valuable abilities and damages its victim. In fact the person did not change her mind at all. In virtue of that, it is not clear that the old critical interest

justice and the elderly

205

fails to apply to her current life. Arguably it remains one of her critical interests, although she is now unable to understand or express it. The point is easier to see in an example that does not involve deciding whether to save a life. We can suppose that the Alzheimer’s sufferer had lived as a devout Roman Catholic until the disease destroyed her ability even to think in those terms. Perhaps in her final years she took pleasure in the cheerful non-denominational religious services in the nursing home. Her appreciation of them might amount in its own way to a critical interest, but I think we would feel that when her life ends she should be buried according to the rites of the Roman Catholic Church, not with a simple service in the nursing home. In this example, her previous critical interest has not been supplanted by the new one, and it continues to matter in determining what would be in her best interest. If we agree, then this interest must be taken into account in deciding what would be best for her. It is true that it must compete against the ‘new’ critical interests and experiential interests that might support enabling her life to continue. But it should be remembered that her own view, before the disease took hold, was that this critical interest was more important than whatever good she might be able to enjoy in the state of dementia. Her new critical interests do not include a revised view about the question of respective importance. So we might be led back towards Dworkin’s conclusion that a peaceful death is what is really in her best interest, although we are not basing the conclusion on Dworkin’s reasons. However, this tentative conclusion should not lead us to withhold the treatment if we also believe that the patient’s will is that her life should continue. As I have suggested, thinking about autonomy is different from thinking about what is in her best interest. I believe that her new will does supplant and replace her old will, even if her will changed because of the cognitive damage the disease did to her. By contrast we should conclude that her original critical interest was not simply replaced by her new critical interests, when we consider why her interests changed. The old critical interest still applies to her, with at least some degree of force, in her altered condition. Consequently those who find it unconscionable to withhold treatment in such a case should appeal to autonomy. This means being convinced of two things: that the Alzheimer’s patient does currently possess a will capable of some degree of autonomy, and that her will is directed at the continuation of her life. Some might grant the first but question the second. They might argue that because of her cognitive impairment she cannot be said to have as an object of her will something as complex as the prolongation of her life, even if it is true that she is living happily in her current state. According to this view we cannot be accused of frustrating her will if we withhold the medical treatment. If we agree then, as I understand the applicable reasons, we might after all be forced to the conclusion that the treatment should not be provided, assuming that we have decided that it would be against her best interest.

206

dennis mckerlie

I have discussed a series of issues that I believe involve a special kind of justice holding between the young and the old. Even if we agree that there is a distinctive problem of justice concerned with age-groups, we should also acknowledge that it differs in a fundamental way from justice between different races or different genders. At any given time, the age-groups of the young and the elderly will contain different people, but those who are now old were once young, and those who are young now will eventually be old. So the young should not understand a view about the claims of the elderly as simply a proposal about how they should treat someone else. In time that view will determine their own claims. The issues I have discussed—issues about distributing health care and other resources fairly and issues about deciding which goals and values to respect when they differ in significant ways at different times in a person’s life—are ones that we will experience from both sides. The difficult question for moral philosophy to answer is whether there is a way of giving the proper weight to this fact without retreating to the complete lives view and concluding that justice is only concerned with complete lifetimes. The prudential lifespan account responds to this fact in a distinctive way. When a young person views an elderly person, he sees someone who used to be young and someone who is now in the stage of life that he himself will eventually occupy. It is as though he is seeing himself in old age. According to the prudential lifespan account, this makes it fair for him to decide what if anything he now owes to that elderly person by thinking about a single life containing both youth and old age and asking how prudential rationality would divide resources between those two temporal stages. The single life can stand in for both his own complete lifetime and the elderly person’s complete lifetime. The result of the theory will be the distribution of resources that would maximize the well-being of the single life considered as a temporal whole, and I have tried to describe the apparent advantages and the apparent disadvantages of this solution to justice between the young and the old. However, that might not be the only way of responding to the fact that we all have lives that over time will contain the different temporal stages. When the young person sees that the elderly person is experiencing misery now, he may think that there is a reason to help, even if the resources that he contributes would create a larger amount of well-being were he to use them himself. This reason—if we are willing to suppose that it might exist—would naturally be expressed by a value like equality or priority applied to temporal stages in lives. Critics will object that we will only see this reason—or think that we see it—if we mistakenly regard this choice as a simple choice between the welfare of two different people in exactly the same way that a choice between two members of different races would be a choice between the welfare of two different people. I am not persuaded by their claim. I think the choice is concerned with interpersonal justice, even if it is not exactly the same as a choice between the members of different

justice and the elderly

207

races. When the young person views the elderly person, he sees someone else living in a distinctive temporal stage of a human life. Moreover, I also believe that if we see the reason for helping when we compare the current life of the young person and the current life of the elderly person, it might lead us to change our mind about the prudential choice made about a single life. A concern like that of priority might apply inside lives as well as across lives. If I recognize that my own old age will contain the same kind of misery that the elderly person feels now, I might decide that it would be better to alleviate some of my future suffering even if some other choice would maximize the total amount of well-being in my lifetime. Applying the value of priority to my own life does not require thinking of my elderly self as being virtually a different person. I am not suggesting that we should appeal to a view about personal identity or a view about the extent of the psychological unity of a human life over time in order to justify this way of using priority. The basic thought is that a benefit is more important when it is received by someone who is badly off, as it is in the case when we apply priority to a choice between benefits for different people. However, if this proposal is reasonable, it reverses the order of explanation of the prudential lifespan account. Instead of learning what fairness requires between the temporal stages of different lives by considering a prudential assessment of the diachronic temporal stages in one life, we might begin by considering how resources should be divided between the simultaneous temporal stages of different lives and discover something about the best way to allocate resources inside our own lives.

References Brock, D. B. (1993), Life and Death: Philosophical Essays in Biomedical Ethics (Cambridge: Cambridge University Press). Bykvist, K. (2003), ‘The Moral Relevance of Past Preferences’, in H. Dyke (ed.), Time and Ethics: Essays at the Intersection (Deventer: Kluwer Academic Publishers), 115–36. Daniels, N. (1988), Am I My Parents’ Keeper? (New York: Oxford University Press). (1996), ‘The Prudential Lifespan Account of Justice Across Generations’, in Daniels, Justice and Justification (Cambridge: Cambridge University Press), 257–83. Darwall, S. (2002), Welfare and Rational Care (Princeton: Princeton University Press). Dworkin, R. (1993), Life’s Dominion (New York: Knopf ). (2000), Sovereign Virtue (Cambridge, Mass.: Harvard University Press). Jaworska, A. (1999), ‘Respecting the Margins of Agency: Alzheimer’s Patients and the Capacity to Value’, Philosophy and Public Affairs, 28/2: 105–38. Kamm, F. M. (1993), Morality, Mortality, i (New York: Oxford University Press). McKerlie, D. (2002), ‘Justice Between the Young and the Old’, Philosophy and Public Affairs, 30/2: 152–77. McMahan, J. (2002), The Ethics of Killing (New York: Oxford University Press). Nagel, T. (1970), The Possibility of Altruism (Oxford: Clarendon Press).

208

dennis mckerlie

Nagel, T. (1979), ‘Death’, in Nagel, Mortal Questions (Cambridge: Cambridge University Press), 1–10. (1991), Equality and Partiality (New York: Oxford University Press). Parfit, D. (1984), Reasons and Persons (Oxford: Clarendon Press). (1995), ‘Equality or Priority?’, Lindley Lecture (Lawrence: University of Kansas Press). Posner, R. A. (1995), Aging and Old Age (Chicago: University of Chicago Press). Rawls, J. (1971), A Theory of Justice (Cambridge, Mass.: Harvard University Press). Shiffrin, S. (2004), ‘Autonomy, Beneficence and the Permanently Demented’, in J. Burley (ed.), Dworkin and His Critics (Oxford: Blackwell), 195–217. Temkin, L. S. (1993), Inequality (New York: Oxford University Press).

part iii •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

B O D IE S A N D B O D ILY PA RTS •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

chapter 9 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

ORGAN TRANSPLANTATION •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

ronald munson In 1954 the first successful kidney transplant was performed by Joseph Murray at the Harvard Medical School’s Peter Bent Brigham Hospital (Merrill et al. 1956). At that moment, we entered a new age. We had acquired, after decades of work by a small number of researchers, the power to snatch someone out of the grasp of death by replacing a vital organ. Since 1954 researchers have consolidated and extended that power. Improvements in surgical techniques, crossmatching tissues, experience in medical management, and, above all, the advent of Cyclosporin and other powerful immunosuppressive drugs have elevated transplantation to the category of standard therapy. Kidney transplants offered a preferable alternative to dialysis by the late 1970s, and the list of organs that can be transplanted with significant success has now expanded to include the heart, liver, lungs, intestines, and pancreas. Corneas, bone, bone marrow, blood, cells, blood vessels, heart valves, and skin are also transplanted, but because they are not functional wholes, they are not considered organs. Discussions of organ transplants are thus typically restricted to what are known as solid, or vital, organs.

A Success — Within Limits ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Each year about 50,000 Americans have their lives extended by receiving new organs. (Statistics are from United Network for Organ Sharing 2004 unless otherwise cited;

212

ronald munson

transplant and waiting-list numbers are estimates from 2003 data.) This is a number equal to the combined enrollments, graduate and undergraduate, of Columbia, Harvard, and Princeton universities. The number is particularly striking because three decades ago virtually all those now saved by transplants would have died. No matter how healthy the rest of a person’s body, without a functioning kidney, liver, or heart, death is the outcome. Yet transplants are not perfect fixes. Completely successful transplants would give people replacement organs without turning them into patients who must be treated with powerful immunosuppressive drugs for the rest of their lives. The drugs have disagreeable side effects, make recipients prone to infections, and increase their risk of cancer and other diseases. Chronic rejection remains a constant threat, and an organ that has functioned well for five or six years may, suddenly and unaccountably, be attacked by the recipient’s immune system and damaged so severely it has to be removed. A perfect transplant would restore a patient to health, be a one-time, long-term fix, and as free of negative consequences for the recipient as changing batteries is for a flashlight. Stem-cell technology may make this possible by engineering organs to be genetically identical with the ones they replace (Munson 2002, ch. 10). Yet while we wait for this marvelous future, transplants, though far from perfect, save lives right now.

A Shortage ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Every year nearly 10,000 people on the United States’ United Network for Organ Sharing (UNOS) national waiting list die without getting the organ they need to survive. They depart quietly, with little public notice. Yet the total of their deaths is roughly equivalent to three times the number of people who died in the 11 September 2001 terrorist attack on the World Trade Center. Almost 100,000 people are on the waiting list at any given time. Some are not as sick as others and, with medical help, are able to wait for months or even years. Those who are lucky may get a needed organ within weeks or a few days. But waiting is not always rewarded, and not everyone who needs an organ, no matter how desperately, receives one. The waiting list is growing at a rapid rate. A new name was added every eighteen minutes in 1998, every sixteen minutes in 1999, every fourteen minutes in 2001, and by 2005 it may be every ten minutes. Given our aging population, the list will grow longer at an increasing rate, and even now more names are put on the list than are removed from it.

organ transplantation

213

The need for organs is constant, pressing, and escalating. It may already be greater than can be met, even assuming complete efficiency in recovering organs from those recently dead (Evans 1989: 15). No more than 15,000 brain-dead potential donors are available annually, and even if the current average of 3.6 organs were recovered from each, this would amount to only 54,000 organs—less than half the number needed now. What makes the situation more desperate is that nothing like this number is actually recovered; hardly more than 50 per cent of those asked to donate the organs of a deceased family member agree to do so. Every organ transplanted may translate into a life extended. Thus, the failure of the present system of altruistic donations to secure enough organs to meet even current needs has produced recommendations for making the system more efficient. It has also led to more radical proposals to recover organs by ‘presuming’ consent and to supplement voluntarism with some sort of market scheme to reward donors or donor families. Then, too, animal organs, organs grown from stem cells, or artificial devices might ultimately eliminate or severely reduce the need for donor organs (Munson 2002, chs. 9–11). Such prospects are at best long-term, however. The ethical and social issues raised by transplants are so interrelated that the thread of any problem eventually leads to the whole tangled ball. I will, however, limit discussion to topics involving living donors in the United States. This restriction is not dictated solely by space constraints. Rather, the rise in the number of living donors gives a particular urgency to questions about informed consent, donor protection, and recipient needs. How we resolve conflicts of interest, address issues of consent, and define the scope of autonomy will shape the policies and practices that determine whether donors are protected and whether lives are lost or saved.

Liv ing-Donor Transplants ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

The most effective measure to reduce the shortage of the organs in greatest demand, kidneys and livers, is to increase the number of living donors (Spital 1989, 2001). Although several attempts were made during the 1940s and 1950s to transplant a kidney taken from a patient’s mother or father, all efforts failed until 1954 when Joseph Murray took a kidney from Ronald Merrick and transplanted it into his identical-twin brother Richard (Munson 2002: 125–9). A series of sixty successful twin transplants followed (Tilney 1986), but it was not until the advent of effective immunosuppressive drugs and crossmatching tissues that using kidneys from unrelated donors became feasible. Kidneys are no longer the only vital organ that—at least in part—can be donated by a living person. People can donate a liver lobe, lung lobe, or pancreas segment.

214

ronald munson

I will additionally restrict this chapter by focusing on issues associated with the living donors of kidneys and livers. Not only do these organs jointly constitute 80 per cent of all transplants; the ethical issues concerning donors are basically the same for other organs. Sixty per cent of all transplants are of kidneys. More than 14,000 of the 23,000 organs transplanted in 2001 (the latest year with complete figures) were kidneys. With 50,000 people waiting for a kidney transplant, the kidney is the organ with the highest demand and shortest supply. Liver transplants number about 5,000 a year, some 20 per cent of all transplants. Nearly 20,000 people are on the waiting list for a liver, making it the organ with the second highest demand and the second lowest supply. Only about 10 per cent of liver transplants use lobes contributed by living donors, but this number will likely increase as the surgical techniques involved become standardized and spread to more transplant centers. Pressure to increase living-donor liver transplants comes from the fact that there is no effective way to replace the liver’s function (unlike that of the kidney and the heart) for even a few days or weeks. The number of living donors increased by a factor of 2.5 during the period 1992–2000. Living donors constitute 52 per cent of all kidney donors, but they contribute only 40 per cent (6,000) of transplanted kidneys, because they can donate only one kidney. Most strikingly, the number of unrelated donors has reached 1,600, ten times the 1966 figure. The importance of living donors can be appreciated by the fact that if only one of every 3,000 people became a kidney donor, the kidney shortage would be solved. Easing the organ shortage is not the only reason for valuing living donors. Transplant surgery can be planned; organs are disconnected from their blood supply for a shorter time and thus remain in good condition; recipients may spend little or no time on the waiting list or undergoing dialysis, so their health does not deteriorate; organs from a living donor will be healthy and undamaged; and good immunological compatibility between donor and recipient can often be arranged. Also, when cancerous liver nodules prompt a transplant, the patient needs a new liver before the cancer metastasizes. A living donor can save the patient from a long wait for a deceased-donor liver and thus perhaps from developing metastatic disease. Kidney recipients benefit significantly from a living-donor organ. The one-year survival with a deceased-donor kidney is 94 per cent, but with a living-donor kidney, survival rises to 98 per cent. Five-year survival increases from 80 to 90 per cent. Liver recipients do not gain as much. Those getting a deceased-donor liver do slightly better (86 v. 85 per cent) during the first year. Yet by the fifth year the situation is reversed, with living-donor recipients significantly surviving longer (86 v. 73 per cent). These figures may change as living-donor transplants become routine and more frequently performed.

organ transplantation

215

Benefits to Recipients, Risks to Donors ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Living donors reduce the organ shortage and directly benefit transplant recipients, but what are the consequences for the donor? Donors risk death, as well as temporary and permanent injuries. They undergo abdominal surgery, and, in addition to the accompanying pain, they risk infection, blood clots, and a damaging or fatal reaction to the anesthesia. Removing a donor kidney via laparoscopy, as is now becoming more common, may reduce pain and shorten recovery, but risks remain. A UNOS survey of transplant centers and a twenty-plus-year follow-up study of living kidney donors show that the risk of dying from a kidney donation is 0.03 per cent (Najarian et al. 1992). (This is about 3 out of 10,000 donors or one donor death every four years.) Also, 56 kidney donors (as of 2004) have later required a kidney transplant themselves. Life-threatening or permanent complications occur in about a quarter of one per cent (0.23 per cent) of donors. No long-term difference between the longevity of donors and non-donors has been determined (Najarian et al. 1992; Park et al. 1996). Donors must also be prepared to alter their behavior (e.g. giving up contact sports) to reduce the chance of damaging their remaining kidney. Living-donor liver transplants are relatively recent. The first was performed in 1987, when surgeons at Brazil’s S˜ao Paulo Medical College transplanted the left lobe of a mother’s liver into her 4-year-old child (Crouch and Elliot 1999: 276). The procedure was restricted to children for the next few years, then several centers began transplanting the right lobe of an adult donor into an adult recipient. About 1,000 living-donor liver transplants are now performed every year. Because the number of cases is comparatively small and the procedure relatively recent, risks to liver donors are not yet well understood (Miller et al. 2001). The donor has 25–60 per cent of the liver removed (the left lobe for children and the larger right lobe for adults). The liver begins to grow back, but during the first several weeks the donor may develop liver failure so severe as to require a transplant. The gall bladder is removed when the lobe is removed, and bile leaks occur in 2–5 per cent of donors and may require additional surgery. Problems, major or minor, occur in 15–30 per cent of all donors. The mortality rate is estimated to be 0.2 per cent or 2 deaths per 1,000 donors. So far only two people are known to have died as a result of being liver-lobe donors. The following case called the public’s attention to the risks of becoming a donor and raised the question of what policies transplant centers ought to adopt with respect to living liver donors.

216

ronald munson Case 1. Mike Hurewitz, a 57-year-old Albany, New York, journalist, died at Mount Sinai Hospital in New York City on 13 January 2002 following an operation to remove a lobe of his liver. Hurewitz was in good health when he volunteered to donate part of his liver to his younger brother Alan. Alan made a good recovery from the surgery and continues to do well. Vickie Hurewitz, Mike’s widow, sued the hospital and six physicians, alleging negligence and malpractice. She also claimed that her husband had not been properly informed about the risks of becoming a liver donor. She recommended that transplant centers declare a national moratorium on the procedure until its value and safety could be established.

Neither federal nor state laws specify who is eligible to become a living donor, how informed consent should be obtained from a donor, nor how a donor’s interests should be protected. These are all matters determined by policies at transplant centers. Legislation, state or federal, will likely replace local rules soon, and this makes framing ethically justifiable procedures particularly pressing. We need to be sure that any new regulatory laws are grounded on sound moral reasoning.

Should Liv ing Donors Be Allowed? ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Thomas Starzl, who pioneered both kidney and liver transplants, argues against the use of living donors on the grounds that, in his experience, the weakest or least valued member of a family is targeted as the donor. Others in the family then manipulate the person into volunteering. Starzl’s view can be generalized into the claim that the risk that the decisions of candidate donors will not be voluntary is too great to permit the practice of using living donors. Surveys of donors provide no evidence that they believe their decisions were manipulated or coerced. The studies show that donors are motivated by a desire to help, take satisfaction in their role in benefiting another person, and experience an increase in self-esteem. They are pleased with their decision and would make the same one again. (See Riether and Mahler 1995: 338; Rhodes 1994: 78; Spital 1996: 376.) Yet surveys necessarily reflect how donors feel after the fact. Surveys cannot prove that donor decisions were not compromised. After all, an unconsenting person shoved off the platform at a bungee jump may later report a good experience. Also, some donors may have been so subtly manipulated that they were never aware of the factors influencing their decisions. Like Starzl, we tend to think of family members or the recipient doing the controlling. Siblings Sue and Tom look to their younger sister Beth to volunteer a lobe of her liver for their mother. Beth is unmarried, has no children, and works only part-time. She is not regarded by her siblings as worth as much as they

organ transplantation

217

are, because they are married, employed, and have young children. Beth, in her siblings’ and maybe her mother’s view, seems to owe it to the family to become a donor. Many similar scenarios are possible (Dwyer and Vig 1995), and, as Fox and Swazey (1978) point out in their classic study, where living donors and families are concerned, the potential always exists for moral blackmail. Less appreciated is that physicians, nurses, or others at transplant centers may unintentionally evoke guilt and so maneuver candidates into becoming donors. Candidates may be ‘actively encouraged’ (Spital 1996: 374) to become donors. This may involve something as simple as a nurse pointing out that ‘If you donated a liver lobe, your mom could beat her cancer.’ Or the encouragement may be a conversation with the patient’s physician, who urges the necessity for quick action to save a loved one.

Autonomy as the Basis ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Living-donor transplants can be morally legitimate, if the donor’s autonomy can be guaranteed. We act autonomously when our actions are the result of our own decisions, when they are self-determined. Autonomy is thus infringed when our behavior is coerced or manipulated. Our society is committed to recognizing the determining power of the individual in making self-regarding decisions. We thus let people decide how to live their lives, including deciding which risks to take. Some people, knowing the facts about becoming a donor, may be willing to undergo suffering and risk their lives to help a sister, mother, friend, or even a complete stranger. Others, for a variety of reasons or no reason at all, may decide otherwise. Autonomy, if it is to be exercised, must be protected, and that is the purpose of informed consent. Informed consent is a way to minimize the chance that, when it comes to decision making, people will be deceived, exploited, tricked, misled, duped, manipulated, or pressured so that their autonomy is violated. If a significant violation of autonomy occurs, the resulting decision is not, in a real sense, the individual’s. Informed consent is thus a means of making sure that the agent of an action is also its true author. For informed consent to be legitimate (valid, genuine, etc.), we require that adults be competent to make decisions—that their powers to understand not be compromised by drugs, mental confusion, disabilities, injury, or depression. We then demand that these competent people be provided with information relevant to the decision at hand and that the information be understandable and sufficient to allow them to weigh the character and consequences of the actions open to them. We require, finally, that people be protected from coercive forces, deception, situational pressures, or other factors that infringe on their autonomy and thus

218

ronald munson

take away some of their decision-making power. To emphasize the importance of this requirement, we often speak redundantly of consent that is ‘free and informed’. Requiring informed consent is a way of making sure that people understand what they might be getting into if they become a living donor, a participant in a clinical trial, or merely a patient considering surgery. The function of informed consent is not to protect people from the consequences of their actions. Rather, it is to make sure that they can know (so far as anyone does) the nature and results (the potential risks and benefits) of each course of action open to them before they make their decision. The generalized Starzl objection that people always risk being manipulated into becoming donors does not entail that using living donors is inherently wrong. Rather, it is a condemnation of any process of securing consent from a candidate that fails to guarantee the protection of the candidate’s autonomy. To be morally legitimate, any transplant program that permits living donors must meet the practical challenge of securing informed consent in a way that protects donor candidates from family and situational pressures and permits them to refuse consent (or withdraw it later) without suffering adverse personal consequences. (See ‘Summary: Rules Regulating Living Donors’ below for measures to protect autonomy.)

Inherent Coercion ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Some bioethicists suggest that where the life of a patient is a stake, someone strongly emotionally attached to the patient is not free to decide to become a donor. The attachment plus the patient’s grave condition (it is suggested) make the situation inherently coercive for the potential donor. Caplan, in commenting on liver-lobe transplants when a parent is confronting the potential death of a child, asks, ‘Does anyone really think parents can say ‘‘No’’ when the option is certain death for their own son or daughter?’ Annas echoes this opinion: ‘The parents basically can’t say no’ (both quoted in Crouch and Elliott 1999: 276). If ‘The parents can’t say no’ is construed to mean something like ‘The parents dare not say no, because they will be berated as bad parents’, the argument has merit. Our society expects parents to make sacrifices for their children, and this includes enduring suffering and, if required, running the risk of injury and death. We do not admire a father who refuses to be a liver-lobe donor and so fails to give his child the chance to live. Yet we also acknowledge that it would be wrong for us to violate his autonomy by forcing him to become a donor. Given our commitment to respecting autonomy, the most we can do is present the father with the option of becoming a donor and spell out its benefits and

organ transplantation

219

risks. Indeed, given the role of informed consent in preserving autonomy, we must make it possible for the father to refuse to be a donor without suffering rebuke. Even if we feel disapproval, we must avoid pressuring the parent into making a positive decision. (Whether the parent may later regret his decision is not our concern.) Expressing autonomy is not necessarily doing what others think is right, but exercising control over one’s actions. Crouch and Elliott (1999: 277) suggest, alternatively, that the ‘can’t say no’ of Caplan and Annas may be construed as invoking a certain notion of moral agency according to which agents are completely free only when they have no ties to others and so every decision considers only the agent’s interests. The authors rightly reject this notion, pointing out that the only people who fit such a description are sociopaths. Moral and emotional commitments, they write, ‘are not constraints on freedom, but are rather part of ordinary human life’ (Crouch and Elliott 1999: 278). I take this to mean that acting for the sake of someone we love does not make us less free. Deciding to act out of love is not a constraint on autonomy but an expression of it, and the Caplan–Annas claim rests on an error. ‘Can’t say no’, in a third interpretation, may be taken to mean ‘For anyone emotionally attached to the patient, no option other than becoming a donor is worth considering.’ Under this construal, when a transplant offers the best chance of saving the patient’s life, the volunteer donor is not interested in other possibilities. He knows at once what he wants to do and is ready to make an immediate decision. ‘Can’t say no’ means, in effect, ‘Doesn’t want to say no’, and if this is what Caplan and Annas are claiming, I find it unobjectionable. (The volunteer must be informed of risks and options and given a chance to consider them, of course, for otherwise his consent is not informed.) It is a mistake to believe, though, as Caplan and Annas may be asserting, that when a decision must be made in a stressful situation in which we care very much about the outcome, the decision is necessarily compromised. This confuses the external pressure that illegitimately influences a decision (family expectations, for example) with the pressure inherent in a situation that requires making a decision. Calling both ‘coercive’ obscures the crucial difference. Being forced to decide is not the same as being forced to decide a certain way. Buying a house is stressful for most people, and while we use laws to shield customers from being coerced into buying the house they are considering, we have no way to shield them from the need to make the decision to buy that house, another house, or no house at all. Similarly, while we can use the consent process to protect parents or others from being pressured into becoming donors, we cannot protect them from the need to make the decision in an inherently stressful situation. They are coerced (by the situation) into having to make a decision, but the decision whether to become a donor is not necessarily (and should not be) coerced.

220

ronald munson

Heroic Donors ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Because autonomy is the justifying foundation for using living donors, discussions of the practice usually focus on ways of making sure that candidates are free to decide if they want to take the risks. (See e.g. Gutmann and Land 1999: 516.) But what are we to say when volunteers insist on becoming donors against medical advice? Is it acceptable for a physician to reject a candidate who demands to become a donor because the risks would be greater than usual for donors? The case below illustrates how the question arises. Case 2: Donald Astrid’s wife died from a pulmonary embolism in childbirth, and his newborn daughter was diagnosed with biliary atresia. Surgery to bypass the child’s bile ducts by attaching a loop of intestine to her liver failed, and without an immediate liver transplant, she would die. Astrid is assessed as a donor at Bayshore Transplant Center and found medically unacceptable. He has a heart arrhythmia, diagnosed and treated with drugs for three years, that puts him at a higher than usual risk of suffering a stroke or dying during the surgery. Astrid insists on becoming a donor, despite the outcome of the assessment and against the recommendation of his physician. Spital refers to people like Astrid as ‘heroic volunteers’ and argues that if a physician had to accept a volunteer ‘against his best judgment’, this would mean the volunteer ‘has an absolute right to donate’, because the volunteer’s wishes ‘would be determinative’. The physician would have to do the transplant, ‘even though he considers donation to be dangerous and ill-advised’ (Spital 2001: 193). Spital’s talk about an ‘absolute right’ is difficult to interpret in the absence of a definition. What Spital appears to mean is: If A has an absolute right to do D, we must permit A to do D whatever the circumstances or consequences. But it is unreasonable to believe that a heroic volunteer is asserting anything so strong. A volunteer who insists on donating against the ‘best judgment’ of his physician is merely rejecting the advice of his physician. He is not claiming that no consideration will alter his decision and that he must be allowed to do D just because that is what he wants to do. He would, presumably, drop his demand if he learned that his blood type is incompatible with his child’s so that if his child received a lobe of his liver, the child would die. It is, furthermore, doubtful that anyone holds that we have any absolute rights in the sense Spital seems to mean. Spital’s argument seems beside the point, in any case. The real issue arises when we ask who is entitled to make the final decision about whether a volunteer can become a donor. Spital sees the autonomy of the heroic volunteer as conflicting with the autonomy of the physician who must carry out the volunteer’s wishes. Thus, ‘the physician must agree with the volunteer that the potential benefits of

organ transplantation

221

the procedure are worth the risks, as is true for any medical procedure’ (Spital 2001: 193). Potential benefits have to be worth the risks, Spital says. But whose benefits and whose risks is Spital talking about? The potential medical benefits of being a donor are never worth the risks. Indeed, the donor suffers medical harms. By contrast, the potential medical benefits of being the recipient of a healthy organ are always worth the risks, because without it the recipient will either die (in cases like Astrid’s daughter) or experience additional suffering. The benefit the donor achieves has to be something other than medical. The physician, on Spital’s view, must then have to balance the risk and nonmedical benefit of the donor against the benefit of the recipient. But this means that the physician is put in the position of deciding whether she is willing for the volunteer to take the risks. The result is that, in making it her decision, she is denying the volunteer the opportunity to decide how much risk he is willing to take for the benefit he seeks. Rather than the physician’s exercising her autonomy, as Spital suggests, she is exercising paternalistic power over the volunteer. It would be wrong for a physician to agree to kill even a willing volunteer for the sake of obtaining an organ for a recipient, even if the transplanted organ would save the recipient’s life. (Removing vital organs and thereby causing death would violate the dead-donor rule, which is the moral and social cornerstone of the practice of organ transplantation. While it is possible to challenge the rule on the ground that perhaps six lives might be saved by sacrificing one willing subject, it is not clear that rejecting the rule and permitting this would allow more lives to be saved. Indeed, the entire enterprise might collapse. I assume here, without argument, the legitimacy and utility of the rule.) Short of this extreme, however, the donation decision ought to be the volunteer’s. The physician should, of course, advise the volunteer of the risks to his health and life. It would even be appropriate for the physician to warn the heroic volunteer against becoming a donor on the grounds that he will be taking a greater than usual risk. Ultimately, though, the physician must let the volunteer decide whether he wants to put his life on the line. The physician may, as is the case with all medical treatment, exercise her autonomy by refusing to accept the heroic volunteer as a patient, but it is not clear, given her role as a physician, how she could justify her refusal. Spital offers what can be taken as an attempt to address this issue. ‘Physicians are responsible for the welfare of their patients, and should act in their best interests . . . ’, he writes. But the problem with this view, as with the initial one, is that because transplant surgery always causes harm to the donor, it can never be in the best medical interest of anyone to become a donor. This can be the case only if ‘best medical interest’ is understood to include a commitment to the welfare of the recipient. While this was the position taken by Justice Counihan (see below) in the

222

ronald munson

first of the twin-transplant cases, this construal of ‘medical’ seems arbitrary and no more than a dodge to permit causing medical harm. ‘Best interest’, as understood by the heroic volunteer, may include a commitment to the welfare of the recipient, despite the fact that the volunteer’s medical best interest will not be served. Thus, Astrid’s concept of his best interest will include doing whatever he can to save the life of his daughter. He will not find it acceptable, then, for a physician to refuse to let him become a donor because the physician does not consider it in his best interest. Why should the physician’s concept of the volunteer’s best interest always trump the volunteer’s concept? Why should the volunteer’s decision about the risk he is willing to take be supplanted by the physician’s decision? Spital offers no answer to these questions. Spital accepts the notion that a heroic volunteer rejected as a donor by one physician may legitimately find another who will accept him. This concession is enough to permit, in principle, heroic donors to take whatever risks they consider appropriate to further their concept of their best interest. But ‘in principle’ does not necessarily translate into ‘in fact’. What if a heroic volunteer cannot find any physician willing to accept him as a donor? We are then back in the position that Spital characterized as a conflict between the autonomy of the would-be donor and the autonomy of the physician. The donor cannot act to promote his understanding of his best interest (e.g. saving the life of his child) without the participation of the physician, while the physician believes she should not act, because it would not serve what she considers the best interest of the patient. Once again, I hold that the patient’s concept of his best interest should trump the physician’s. The physician has a duty to inform the volunteer that he would be taking a greater than usual risk, but the decision about whether to take it should ultimately be the donor’s, not the physician’s. This would be no more than a case of acting against medical advice, something long acknowledged to fall within the scope of patient autonomy. While the physician may regret the patient’s decision, refusing to abide by it would amount to a violation of the patient’s autonomy for paternalistic reasons. I also hold that, where living donors are concerned, an appropriate understanding of what it means for the physician to be committed to promoting the interest of the volunteer–patient makes clear that there is no conflict between the autonomy of the volunteer–patient and the physician. Ordinarily, the physician’s commitment to promoting the best interest of the patient is understood as limited to the diagnosis and management of disease in the medical context. But the practice of using living donors requires extending the notion of ‘best interest’ beyond that context. The volunteer’s best interest must include his non-medical wants and values. Thus, the physician’s commitment to promote the volunteer–patient’s best interest is a commitment to promote the broadened notion, the one that includes

organ transplantation

223

the volunteer–patient’s decisions about what is important and what risks he is willing to run to secure ends he values. A physician qua physician must act for the sake of the patient’s best interest, and, given the practice of using living donors, this means accepting the broadened notion of best interest as determined by the volunteer–patient. The physician, therefore, has an obligation to accept a heroic volunteer as a donor, even though the volunteer acts against medical advice. Because the physician qua physician is acting in the best interest of the volunteer, the physician’s autonomy is not in conflict with the volunteer’s. It would be in conflict only if ‘best interest’ is understood by the physician as limited to the medical best interest. Even if this argument is correct, finding a physician willing to operate on a willing heroic volunteer may remain a practical problem. This is not, however, a circumstance unique to transplant ethics. During the early years of the AIDS epidemic, some physicians refused to treat HIV-positive patients. While physicians lack moral grounds for refusing to accept heroic volunteers, given the circumstances in which transplants take place, the volunteers may not be permitted realize their intention to become donors.

Strangers as Donors ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

The initial basis for accepting living donors was a broadly construed concept of ‘medical interest’ (the sort of construal Spital needs to make his argument work). The purpose of medical treatment is to benefit the patient, but when surgeons remove an organ from a living donor, only the recipient appears to benefit. How then can physicians justify causing harm to a healthy person for the sake of someone else? Surgeons at Boston’s Brigham Hospital grappled with this question in 1957 at the dawn of kidney transplantation. Leonard Marsden, a 17-year-old, eagerly consented to donate a kidney to his identical twin, Leon. The surgeons then hesitated, questioning whether by subjecting Leonard to surgery they would be providing him with any benefit. Hoping to clear the way, the twins’ parents petitioned the Massachusetts Supreme Judicial Court to rule on the question. Justice E. A. Counihan, after hearing testimony about the brothers, decided that if the transplant were not done and Leon died, Leonard would suffer an emotional disturbance that would adversely affect his health and well-being (Curran 1959: 893). The surgery would thus confer a ‘medical benefit’ on both brothers. Consequently, the surgeons would not be harming Leonard just to benefit Leon. The judge’s insight was that benefit should not be understood too narrowly, even in the medical context. That the benefit for Leonard should be viewed as ‘medical’ was never persuasive. Without much discussion, as the frequency

224

ronald munson

of kidney transplants increased, centers began to construe Counihan’s ‘medical benefit’ as equivalent to the donor’s having an ‘emotional relation’ to the recipient. Thus, donors were limited by most centers to parents, spouses, siblings, or other blood relatives. In the 1980s and 1990s, however, as transplants became safer and deceased-donor organs scarcer, centers expanded the notion of ‘emotional relation’ and started accepting friends of recipients as donors. While some centers still adhere to such a policy, others have decided to accept also ‘altruistic strangers’ or ‘Good Samaritan donors’ (see Case 3 below). The transplant community, even when dealing with family donors, has always stressed altruism as a reason for becoming a donor, and some recent writers have claimed altruism as the sole basis for organ donation. This has occasioned a debate about how the altruism of family and friends (‘intimates’) compares with that of strangers. At stake is thought to be how much risk each group ought to be permitted to take. Ross and co-authors (Ross et al. 2002: 426) argue that intimates cannot be purely altruistic, because, given their sense of duties and obligations, they are both other-regarding and self-serving. Thus, intimates should be permitted to run higher risks than strangers, because intimates both act out of duty and do not identify their interest as being completely different from that of a recipient. Daar (2002) shows, however, that this conclusion depends on accepting the authors’ definition of an altruistic act as ‘one motivated primarily or solely by respect and concern for the preferences and needs of others, freely chosen rather than done out of a duty or obligation’ (Ross et al. 2002: 426). Daar argues persuasively that, even if we accept the definition, ‘altruism does not negate every element of self-interest’ and even strangers may have their own sense of intimacy and duty (Daar 2002: 424). Each potential donor, Daar holds, should be treated as an individual person and his or her acceptability based on such considerations as the level of risk, not on intimacy or degree of altruism. While Daar’s rejection of the Ross position is warranted, the debate itself rests on a misconception. Altruism may be a motive for organ donation and serve as an explanation for why people become donors, but it is not the moral basis for allowing living donors. Altruism is a value, but it is neither a duty nor an ethical principle, and it is a mistake to look to it to justify donation policies. Rather, it is our commitment to the autonomy of the individual, protected by the process of informed consent, that makes the use of living donors morally legitimate. Volunteers are given the opportunity to become informed and protected from pressures. They are given the chance to deliberate, with the opportunity to change their minds later, and then allowed to decide whether they wish to become donors. Perhaps some, maybe even the vast majority, will be moved by the wish to benefit others. But a misanthrope who wants to become a donor is as acceptable as an philanthropic superstar, so long as the process of informed consent is followed. While we may be curious about people’s motives (and may want to appeal to them to increase the number of donors), it is not motives that determine whether

organ transplantation

225

the selection of volunteers is legitimate. When the conditions of informed consent are satisfied, living donors, whether strangers or intimates, can reasonably be viewed as promoting their own interest. They are exercising their autonomy in deciding what is important to them and what they are willing to risk to secure it. They are deciding how they want to shape their lives. The assertion that altruism, rather than autonomy, is the basis for organ donation is a way of blocking what some see as an unacceptable consequence of allowing unrelated donors. The major objection to using such donors is that the practice may encourage the commercialization of transplantation (Kaplan and Polise 2000: 520). A mother is not likely to sell a kidney to her son, but a stranger might sell hers to the same person. Commercialization, which entails self-interest, is inherently incompatible with altruism. Thus, if altruism were required to legitimize donation, the very possibility of commercialization would be ruled out. Yet even if compelling reasons could be given against commercializing organ procurement, merely asserting that altruism must always be the basis for donation is not persuasive. We need an argument to show that there is something about selling organs that is morally different from selling cars or blood plasma.

Paying Liv ing Donors ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

The idea of paying donors or selling organs has been denounced by the transplant community, politicians, and religious leaders since the early 1980s (Munson 2002: 98–110), coincidental with the time that transplants were becoming successful. The United States National Organ Transplantation Act of 1984 makes buying and selling organs, whether from living or deceased donors, illegal, as do the laws of Great Britain, all European countries, China, India, Russia, Mexico, and South Africa. The World Health Organization condemns paying for organs under any circumstance (World Health Organization 1991), and, although trade in transplant organs takes place in parts of Asia, the Middle East, and South America, it is illicit (Cameron and Hoffenberg 1999: 727). The world ban on organ sales has been defended over the decades by a number of ethicists, lawyers, and transplant professionals who have generated a laundry list of objections to paying donors. (I shall limit discussion to living donors of kidneys; selling organs from deceased donors raises different issues.) Prominent and recurrent objections include: a paid donor loses the psychological benefits that reward an altruistic donor; paid donation reduces altruism in society; the quality of donated kidneys will decline; the donor may suffer harm and become a burden to society; paying donors may reduce the number of donations from deceased donors; organ selling puts the human body in the same moral category as slavery; organ selling involves putting a price on the priceless; paying for organs exploits the poor;

226

ronald munson

organ selling treats the human body as a commodity and thus violates our respect for persons. (See Phadke and Anandh 2002; Radcliffe-Richards et al. 1998; Russo and Brown 2003, for a review of objections.) Most of these complaints are about institutionalizing the buying and selling of transplant organs—that is, making organs goods in the market economy. Such objections are, for the most, consequentialist, and while the numerous issues raised are important, they go beyond the scope of this chapter. The question logically prior to the market and consequentialist issues is whether there is something about paying kidney donors that makes it inherently wrong. I claim there is not. If the autonomy of the individual is the basis for recognizing that, when the conditions of informed consent are met, donating a kidney to someone is a morally legitimate act, it must also be morally legitimate for the individual to be paid for donating the kidney. Either act follows as a result of a decision made freely by the person. In the first case, the individual decides to be altruistic, in the second case, she decides she wants money. Individuals may be said to own (or, at a minimum, have legitimate control over) their bodies in substantially the same sense in which they own their diamonds. Thus, in the way that individuals are free either to sell or give away one of their diamonds, they are free either to sell or give away parts of their bodies. Altruism might move someone to donate a diamond to a charity or to donate a kidney to a stranger; or he might decide to sell both. (I consider someone’s selling his own organ and being paid to be an organ donor as equivalent.) While differences between diamonds and body parts are numerous, I suggest that none is morally relevant with respect to the matter of getting paid to become an organ donor. Once we have agreed that autonomy is the ground for legitimizing an individual’s decision to donate a kidney, we must also acknowledge it as legitimizing his decision to sell a kidney. But what if someone wants to sell both his kidneys? Or his heart, liver, or lungs? While we may agree that, as their owner, he may dispose of his organs in any way that he sees fit, this does not mean that we are free to remove them or to buy them. We are constrained by the fact that by taking both his kidneys or his vital organs, we would be killing him. We would (to put the point another way) be violating the dead-donor rule, which requires us to establish that a donor is dead before any organ needed to sustain his life is removed. It is prima facie wrong to kill someone, even if he wants us to, and even if we could use his organs to save several lives. The most common defense of the claim that selling a kidney is morally wrong in itself is based on the Kantian view that it does not show respect for one’s humanity. Selling a kidney expresses disrespect for oneself and, as a consequence, disrespect for what it means to be human (Morelli 1999: 320). Gill and Sade (2002: 26) reject this complaint, justifiably, on the ground that it is not persuasive to consider one’s humanity as dependent on one of one’s kidneys. One’s humanity may be viewed, more reasonably, as dependent on one’s rationality and one’s capacity to follow

organ transplantation

227

self-given laws (autonomy) dependent on it. Selling a kidney thus has no destructive effects on one’s humanity. Hence, it cannot express disrespect for all humanity. A second and similar Kantian-type objection is that it is wrong to sell kidneys because human beings are not property, and, as Cohen says, to sell them ‘and those bits and pieces integral to them is to violate that which is essential to them’ (Cohen 2002: 28). While Cohen is right that to sell human beings violates their inherent worth, she wrongly assumes that the ‘bits and pieces’ of their bodies are likewise of inherent worth. If ‘integral’ means ‘essential’ or ‘indispensable’, a kidney fails to meet the description. Unlike selling oneself into slavery, selling one’s kidney will have no consequences on one’s capacity for self-governance. Indeed, Cohen’s argument appears to be a case of the fallacy of division. Gill and Sade (2002: 25) point out that even if the Kantian argument that selling one’s kidney violates the categorical imperative, because it involves treating oneself as a means only, were correct, it would not follow that paying a donor should be against the law. We do not base our laws on the Kantian duty to respect humanity by respecting oneself. The laws we make aim, rather, at protecting the (nonKantian) autonomy of individuals. We protect their freedom to make personal decisions about self-regarding acts, and, if the decision they make is to follow their understanding of a rational moral law (Kantian autonomy), they are free to do that as well. No one need sell a kidney. The transplant community is now in the process of rethinking its long-time condemnation of paying donors (Joralemon 2001; Cameron and Hoffenberg 1999: 724–5). The initial impetus for disapproval, in my opinion, was the fear of alienating the public by associating transplants with money and the unseemly business of trading in body parts. The community was afraid that a loss of public approval, due to bad associations, would result in a decline in the number of people donating their organs at death. Without donated organs, the entire transplant enterprise would then collapse. To sustain the system, the community has always stressed altruism and downplayed the commercial aspects of transplantation. That hospitals, surgeons, coordinators, laboratories, transport services, and organ procurement organizations make money from transplants is not a shameful truth, yet it is rarely mentioned in public. Inspiring stories of transplant miracles are the preferred sort of publicity. Yet now that the organ shortage is desperate and the public is more familiar with transplants (and perhaps more tolerant of commerce), some are saying that we need to reconsider the issue of paying donors. Extolling altruism has not produced nearly enough donors, so thousands of people are dying who might be saved. (The situation is especially critical in countries that cannot afford to buy and maintain the dialysis equipment required to sustain the lives of thousands in kidney failure.) Many do not find the arguments against paying donors compelling and believe that we could devise mechanisms to protect consent and prevent the exploitation of the

228

ronald munson

poor and disadvantaged. Because so many lives are at stake, the resolution of this issue is of more than academic concern.

Protecting the Donor: Promises Unkept ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Protecting a living donor must be understood as involving more than securing informed consent and guaranteeing the volunteer’s autonomy at the time of decision making. The following case points to a problem that needs solving. Case 3: Arielle Dove was so moved by the selfless acts displayed in the aftermath of the 9/11 terrorist attacks that she decided to donate a kidney to a stranger (Meckler: 2003). She located a living-donor web site and arranged for one of hers to be removed and transplanted into someone she had never met. After the surgery her life took a turn for the worse. More than a year later she still had episodes of vomiting and felt dizzy and listless. She was also very angry. The man who received her kidney assured her he would pay for her expenses not covered by his insurance. But he didn’t keep his promise. ‘I volunteered to put my life on the line, and I guess I’ve given up my good health for this, and nobody seems to care,’ she said. ‘It’s really hard not to cry.’ Living donors may develop long-term medical problems, may not be able to work for weeks or months, may require a liver or kidney transplant themselves, may run up medical bills not covered by their insurance or a recipient’s. Who is going to pay for the donor’s post-transplant expenses? Will it be the recipient? The transplant center? Or will the donor herself have to find some way? These are among the questions that need to be settled before a potential donor becomes an actual one. Yet often the questions are neither asked nor answered. No one, as matters stand, is committed to looking out for the longer-term interest of the donor. Some donors complain that once they have had a kidney or liver lobe removed and are out of the hospital, transplant centers no longer take any interest in their welfare. Promises that the donors thought were made are not kept, and their future health problems are not recognized as possibly related to the surgery or the loss of an organ. To make sure that the practice of using living donors functions in a morally legitimate way requires that we introduce into general practice three measures to protect the welfare of donors and future donors. 1. Living-donor advocate: medical. Some transplant centers provide donors with a medical advocate, and this should be required of all centers that accept living donors. (This is also a recommendation of United States Department of Health and Human Services Advisory Committee on Organ Transplantation 2002: 3.) A medical advocate should be a physician with expertise in transplantation who is not

organ transplantation

229

involved in the care of a potential recipient. An advocate should assist a candidate donor in understanding the process, risks, and benefits of becoming a donor and help the candidate frame appropriate questions and gather information relevant to making a decision. The advocate should, in general, take the measures needed to guarantee that consent is free and informed. If a volunteer becomes a donor, the advocate should then be responsible for making sure that the medical interest of the donor is served. This means not only seeing to it that the donor–patient receives appropriate hospital care, but making sure that she receives whatever follow-up care she needs. The advocate should serve as the donor’s medical advisor and champion, though not as her physician. 2. Living-donor advocate: legal. The Dove case illustrates, taking Dove at her word, what can happen to a donor when commitments made to her are only an informal understanding. A properly written consent document spells out the potential risks and benefits to the donor, alternatives to donation, and the opportunity to withdraw consent, but its usefulness is limited. It serves the dual purposes of informing the volunteer and offering partial proof that a donor’s choice was appropriately informed and risks were voluntarily undertaken. This (among other things) helps protect centers and physicians from lawsuits and professional censure, but the document, other than informing, provides little help to the donor. The center or the recipient’s insurer may agree to provide the donor with medical care that is immediately associated with the surgery, but if the donor loses income due to the hospitalization, will the money be reimbursed? And if the donor develops medical problems six months or a year later, will she be provided with free care? If, as Dove alleges happened to her, the recipient agrees to pay for costs not covered by insurance, then fails or refuses to do so, what remedy does the donor have? What the donor requires to protect her interest is a legally enforceable agreement—a contract—with the transplant center and with the recipient. The donor needs a legal advocate, as well as a medical one. The advocate should be an attorney whose fees are paid by the center, the recipient, or the recipient’s insurer, but whose client is the donor. The legal advocate, with the medical advocate, should consult with the potential donor as part of the consent process. The advocates should go over the consent document with the candidate, and the legal advocate should be available to offer advice before the volunteer makes the consent decision. (A potential donor may refuse legal advice or act against it.) Either as part of the consent document or in an additional document, commitments made to the donor with respect to such matters as financial compensation for time lost while hospitalized, the assumption of responsibility for health-care costs of the donor for donation-related problems, how a dispute about whether a complaint is donation-related should be resolved, and the limits of assumed responsibility should be addressed.

230

ronald munson

The legal advisor should consult with the donor after the donation is completed and the donor hospitalized. If either advocate questions whether the kind or quality of care promised is being delivered, then the attorney should advise the donor about the availability of appropriate legal remedies. Physicians and centers, fearing unwarranted litigation, are not likely to welcome legal advocates into the donation process. Yet donors put themselves at risk, and they deserve to be assured that guarantees made to them have the status of a legally enforceable contract. The addition of a legal advocate will add to the cost of a transplant. Quite apart from protecting the interests of donors, however, the knowledge that a legal advocate will be assigned to each donor may make becoming a living donor an acceptable option for many more people. This opens the possibility of saving more lives than can be saved at present. 3. Living-donor registry. Data about living donors are mostly from kidney donors. Even here, the data are for the most part confined to statistics about operative mortality and survival (Park et al. 1996). Liver-lobe donation is sufficiently untried that even the mortality rate associated with it is uncertain. Data for lung-lobe and pancreas-segment donors are similarly sparse. The long-term effects of becoming a living donor of any organ or organ-part have been little studied (Najarian et al. 1992). Thus, the information needed by donor candidates is not as good as it should be. Perhaps better information would do little to change the decisions parents make to donate to their children, but it might have a significant impact on others, particularly on those who want to deliberate before making a decision about donating an organ to a stranger. Because the liver regenerates, data showing that harmful results are rare over the long term would likely increase the number of living liver donors. Also, if donors develop serious medical problems years later that are shown to be donation-related, we need to decide how to compensate the donors and establish who has responsibility for doing so. Further, if some problems are serious and occur often, we need to decide whether our commitment to individual autonomy requires transplant centers to accept donors who are likely to develop diseases that will compel us to spend considerable public resources for treatment. Such considerations show that because we allow living donors, establishing a national living-donor registry is a compelling need. (For a similar recommendation, see United States Department of Health and Human Services Advisory Committee on Organ Transplantation 2002: 3.) The registry would keep track of donors and collect and preserve medical information about them over the years. The registry could take the form of a database operated and financed either by a federal agency or by an organization like UNOS, which works under a federal contract. Computers and the Internet make it possible for hospitals, transplant centers, and physicians to supply the information needed at relatively little cost. That the time has come

organ transplantation

231

to establish such a registry is a belief widely shared in the transplant community (Ochs 2002).

Donors of Last Resort ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

A basic rule of donor selection is that children and others incapable of consenting ought to be donors of last resort. Those able to consent are (by definition) capable of looking out for their welfare, but those incapable of doing so are open to exploitation. Hence, we have a duty to protect them. When primarily living donors were employed in the 1950s–1970s, whether it was legitimate to use a child as a donor was often a life-or-death issue. The situation has eased but not disappeared. Kidneys remain in short supply, and children benefit from being removed from dialysis as soon as possible. Thus, families continue to be pressured by circumstances to make wrenching decisions about risking the health and safety of one child to benefit another. The scope of the problem may also be increasing. While now only adults are accepted as liver-lobe donors, when the transplants become better established, children and other ‘incompetents’ may become regarded as potential donors for siblings or other family members. Circumstances in which a child might be the only available liver-lobe donor for a sibling are easy to imagine. The surviving parent, for example, might not have a blood type compatible with that of the child in need. Or the parent might be too ill to become a donor. More distant relatives, if any, might fail to qualify as donors or might refuse. The child’s sacrifice could be all that stands between her sibling and death.

Important Interest at Stake The fundamental requirement to be met in justifying a child’s becoming an organ donor, I suggest, is that the child must have something important at stake in the use made of the organ. (I will refer to children here, yet most considerations apply also to incompetent adults.) Becoming a donor must be in the child’s best interest, and this may require that the child suffer surgical injury and run some risk of death. The child’s best interest can be understood as the child’s having a significant stake in the welfare of the organ’s intended recipient. (As mentioned above, Leonard Marsden, with respect to his brother’s welfare, had at stake something affecting his own ‘health and physical well-being’.) No matter how slight the risk, a child (or other incompetent person) cannot be required to donate an organ to help a stranger, even if the organ would save the stranger’s life. The child has no direct stake in the stranger’s welfare, and thus the donation would not serve the best interest of the child. By contrast, an intimate

232

ronald munson

who is not a relative may be of crucial importance to the child’s welfare, as Anne Sullivan was to Helen Keller.

Reasonable Risk It is appropriate to subject a child to some risk to protect her best interest. Thus, in cases where the life of a person important in the child’s life is at stake, it is reasonable to put the child at risk for the benefit she may gain. We put children at risk for expected benefits in other medical contexts, even when their lives are not endangered—surgery to correct club foot, cleft palate, or amblyopia, for example. What we know of risks at present indicates that it is sometimes justifiable to make children into kidney donors, but not liver-lobe or lung-lobe or pancreas-segment donors. We do not yet know enough about the effects and risks of such donations to subject children to them, even when a child has an important stake in the life of a recipient. The American Medical Association’s Council on Ethical and Judicial Affairs puts the point tersely: ‘Children should not be used for transplants that are considered experimental or non-standard’ (American Medical Association 1996–7: 35). Where the chances of death or suffering serious harm are considerable or unknown, we lack justification to put a child at risk, even to save the life of a person important to the child. We are free to decide to risk our own lives for anyone, because we are able to understand our alternatives and the consequences of our actions. Children cannot. Hence, when we decide for them, we must take the most conservative stance compatible with their interest.

Deciding About Donors of Last Resort An asymmetry exists between those competent to consent and those who are not. Children are not competent to decide to become donors, but they are also not competent to decide not to become donors. (I will not address here issues of assent connected with older children.) A decision belonging to competent people belongs to someone else in the case of incompetent people. This asymmetry offers the potential for exploitation. Suppose Sue Crane needs a kidney transplant. High blood pressure eliminates her husband, Sam, as a donor, but their healthy 22-year-old son Bob, now in law school, has the same blood type and is a good antigen match. The Cranes’s retarded 16-year-old son, Tom, is also a good match, however. Bob is willing to be the donor, but he is the pride of the Crane family, and his parents do not want to interrupt his education and subject him to the risks of surgery. Tom is a constant source of difficulty. ‘Now he has a chance to do something to help the family,’ Sam says. Sam and Sue then instruct Bob to refuse

organ transplantation

233

to volunteer when he is interviewed at the transplant center. Sam is medically unacceptable, Bob refuses, and no one else steps forward. Thus, Tom becomes the donor of last resort. The duty to protect incompetent people from exploitation rests with whoever has the responsibility to decide what is in their best interest. The courts already decide for institutionalized and demonstrably incompetent adults. With respect to children, parents are the obvious candidates to make the decision, but two considerations rule them out. First, parents like the Cranes can conspire to sacrifice the weakest member of the family to protect a favored one. The person who needs the most protection thus becomes, ironically, the one who is the most vulnerable. The situation is not improved if, as Ross (1993) recommends, the family as a whole is given the power to decide. While this could, as she says, promote intimate relations and allow the family to draw upon its own values, religious beliefs, and sense of itself, it leaves children with no protection from family pressures. Indeed, Ross’s process of family decision making describes exactly the situation Starzl (1985) considered so inherently manipulative as to lead him to recommend against the use of even adults as living donors. A second difficulty is that parents can be forced into a Sophie’s-choice situation requiring them to help one child (or family member) only at the expense of another. This faces them with a conflict of interest, so that whatever decision they make will be suspect (even to themselves) and open to charges of unfairness and favoritism. Decisions about accepting competent adult candidates as donors are now made by committees at transplant centers, and this same approach might be taken with children. Williams (1995: 499) advocates the use of ad hoc groups to make decisions about children as potential bone marrow donors and describes how, at a Honolulu hospital, a staff committee interviews children in an informal way and determines if they understand ‘their role in the transplant procedure’ and if their willingness to be a donor is ‘free from duress and based on adequate information’. Depending on the judgment of the committee, a child is accepted or rejected as a donor. The committee process, Williams observes, is inexpensive, efficient, and offers a way to consider the best interest of a child. Despite these virtues, a committee approach has drawbacks so serious as to make it unacceptable. First, committees work effectively only when children are old enough to grasp what is being asked of them and assent to it. This leaves open the question of how we should deal with younger children. Committees are also limited in their powers to obtain data relevant to the decision they must make. If a family member withholds information or lies to the committee (claiming he has a close relationship with a child, for example), the committee can impose no sanctions and must make its decision on the basis of whatever data it can gather or surmise. More is at stake, furthermore, for an organ donor than for a bone marrow donor. Harvesting bone marrow involves discomfort and the risk of infection, but

234

ronald munson

no significant danger is associated with it. Being a kidney donor requires extensive surgery, greater risk of infection, and a chance of dying or long-term effects. Because more is at stake for organ donors, more protection for vulnerable potential donors is required.

Decision of the Court Williams’s observation that court proceedings can be time consuming and costly is correct, but protecting people from serious exploitation is sufficiently important to warrant additional time and money. The courts, more than any other institution, are in the best position to guarantee that stringent criteria for a child’s becoming a donor are satisfied and that the best interest of the child is served. Courts of law, unlike committees, however constituted, operate within a tradition of protecting the rights of individuals by invoking a variety of procedural and substantive safeguards. Should a 6-year-old girl contribute a kidney to her teenage sister? A court can conduct discovery proceedings and gather relevant medical and personal information, using its subpoena powers if necessary, and thus put itself in the position of answering the question. Experts can be called to offer opinions, and family members required to testify under oath. Rules of evidence, relevance, and proof can be brought to bear on the basic question. Most important, a court can appoint an attorney (a guardian ad litem) to represent the child to make sure everything recognized as relevant to her interest is brought forward for the court to consider. Because courts have powers committees lack, committees are never able to delve so thoroughly into issues affecting the welfare of candidate donors. At the end of hearings, when the evidence and arguments for and against a child’s becoming a donor have been presented, a court’s deliberations offer the best chance of getting an independent and objective decision. A committee might have arrived at the same decision, but where protecting the vulnerable is concerned, process and safeguards matter. The presiding judge of a Massachusetts court made this point forcefully in the 1977 Saikewicz decision: We take a dim view of any attempt to shift the ultimate decision-making responsibility away from the duly established court . . . to any committee, panel, or group, ad hoc or permanent . . . . questions of life and death seem to us to require the process of detached but passionate investigation and decision that would form the ideals under which the judicial branch of government was created. Achieving this ideal is our responsibility . . . (Superintendent of Belchertown State School, et al. v. Saikewicz, 417)

I have argued, to recapitulate, that a child (or other incompetent person) may become an organ donor when: it is in the child’s best interest; risk to the child is

organ transplantation

235

reasonable; the child is the donor of last resort; a court of law, rather than parents or any sort of committee, is making the decision.

Summary: Rules Regulating Liv ing Donors ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Perhaps the most useful way to summarize the above discussions is to state rules or guidelines. Yet because several important questions were not addressed and guidelines must always be interpreted, the following rules are not offered as either exhaustive or definitive. 1. A potential donor must be competent to make decisions. This includes being able to understand the nature and likelihood of the risks involved in becoming a donor. 2. A potential donor must be provided with information adequate for making the donation decision. The need to provide information about the nature and likelihood of risks and benefits is clear. Less obvious is the need to supply the candidate with information about the alternatives available to the potential recipient (e.g. dialysis, continued medical support, waiting for a deceased-donor organ, or waiting for another living donor). 3. Potential donors should not be solicited. A center may inform the patient and others that those who want to consider becoming donors should contact a designated person who is uninvolved with the patient. Russo and Brown endorse this rule (Russo and Brown 2003: 27), and Biller-Andorno and Schauenburg suggest that a volunteer should identify herself ‘without any action on the part of the physician’ (2001: 163). 4. A potential donor must be protected from pressures to volunteer. A willingness to become a donor ought to be considered a necessary condition for being a ‘suitable’ candidate. The assessment team should determine in a private interview if the candidate is willing. The candidate needs to be told that, no matter what he may have said to others nor what others may expect him to do, if he decides he is not willing to be a donor, this will remain confidential. If the candidate says he is unwilling, the assessment team will then declare him an ‘unsuitable’ candidate, with no details made public. This will protect the candidate from the anger, recriminations, or blame that might have been directed at him for publicly refusing to help the patient needing the transplant. A candidate must also be permitted to change his mind about becoming a donor until the last moment before surgery. This may result in great inconvenience and disappointment and even put the intended recipient at greater risk than if no apparent donor had become available, but it would be a serious violation of an

236

ronald munson

individual’s autonomy to remove one of his organs after he has withdrawn his consent. 5. Assessment of the suitability of a potential donor should not be done by physicians or others involved in the care of the potential recipient. This will eliminate the conflict of interest inherent in a relationship in which those caring for a patient needing a transplant also select a donor. The assessment should be done by a team (e.g. hepatologist or nephrologist, psychiatrist or psychologist, social worker, etc.) able to determine whether the candidate is medically and psychologically suitable to become a donor. The assessment should also consider a potential donor’s social and economic situation so that the candidate can be provided with information about the impact that becoming a donor might have on his or her life. 6. Potential donors should be provided with medical and legal advocates. Both advocates should advise a candidate before she makes a decision. If she decides to becomes a donor, the legal advocate should represent her interests in making contractual arrangements with the center and with the intended recipient. The medical and legal advocates should monitor her welfare after the transplant. 7. A registry should be established to gather longitudinal data about the health of living donors. The database in the United States could be operated by UNOS under a contract with the federal government. The information could be medically important to donors, and it would be relevant in informing potential donors about potential risks. 8. Donors incompetent to consent may become donors if it is in their best interest, the risk to them is reasonable, no other donors are available, and the decision permitting them to become donors is made by a court of law.

Conclusion ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Organ transplants save thousands of lives every year, yet thousands more die because of the shortage of organs. While increasing the number of organs from deceased donors would be of considerable value, the best hope for saving the lives of tens of thousands of people who would otherwise die is to increase the number of living donors. The autonomy of the individual legitimizes an individual’s decision to become a living organ donor. This does not relieve transplant centers of the responsibility for seeing to it that donors are genuine volunteers and have the information they need to assess their risks and options. Measures are needed to protect the autonomy of the individual in deciding whether to become a donor, but additional measures are needed to protect the welfare of living donors. These include appointing donor advocates and maintaining a registry of living donors.

organ transplantation

237

The prospect of saving so many thousands of lives requires us to take seriously the moral and the practical issues centering around the use of living donors. Yet we do not have time to discuss those issues indefinitely. The sooner some matters are settled, such as the moral legitimacy of paying donors, the more lives will be saved. The issues are urgent, for literally life and death are at stake.

References American Medical Association, Council on Ethical and Judicial Affairs (1996–7), ‘The Use of Minors as Organ and Tissue Donors’, in Code of Medical Ethics: Current Opinions with Annotations (Chicago: American Medical Association), 34–6. Biller-Andorno, N., and Schauenburg, H. (2001), ‘It’s Only Love? Some Pitfalls in Emotionally Related Organ Donation’, Journal of Medical Ethics, 27: 162–4. (2001), ‘Who Shall Be Allowed to Give? Living Organ Donors and the Concept of Autonomy’, Theoretical Medicine, 22: 351–68. Cameron, J., and Hoffenberg, R. (1999), ‘The Ethics of Organ Transplantation Reconsidered: Paid Organ Donation and the Use of Executed Prisoners as Donors’, Kidney International, 55: 724–32. Cohen, C. (2002), ‘Public Policy and the Sale of Human Organs’, Kennedy Institute of Ethics Journal, 12: 47–64. Crouch, R., and Elliott, C. (1999), ‘Moral Agency and the Family: The Case of Living Related Organ Transplantation’, Cambridge Quarterly of Healthcare Ethics, 8: 275–87. Curran, W. (1959), ‘A Problem of Consent: Kidney Transplantation in Minors’, New York University Law Review, 34: 891–8. Daar, A. (2002), ‘Strangers, Intimates, and Altruism in Organ Donation’, Transplantation, 74: 424–6. Dwyer, J., and Vig, E. (1995), ‘Rethinking Transplantation Between Siblings’, Hastings Center Report, 25: 7–12. Evans, M. (1989), ‘Organ Donations Should Not Be Restricted to Relatives’, Journal of Medical Ethics, 15: 15–20. Fox, R., and Swazey, J. (1978), The Courage to Fail: A Social View of Organ Transplants and Dialysis (Chicago: University of Chicago Press). Gill, M., and Sade, R. (2002), ‘Paying for Kidneys: The Case Against Prohibition’, Kennedy Institute of Ethics Journal, 12: 17–45. Gutmann, T., and Land, W. (1999), ‘Ethics Regarding Living-Donor Organ Transplantation’, Langenbeck’s Archives of Surgery, 384: 515–22. Joralemon, D. (2001), ‘Shifting Ethics: Debating the Incentive Question in Organ Transplantation’, Journal of Medical Ethics, 27: 30–5. Kaplan, B., and Polise, K. (2000), ‘In Defense of Altruistic Kidney Donation by Strangers’, Pediatric Nephrology, 14: 518–22. Lindsay, D. (2002), ‘An Organ, Stem Cells, and Blood Are Precious Gifts of Life’, Online Washingtonian Community Service,

. Meckler, L. (2003), ‘The Dark Side of Organ Donation’, Associated Press, <www.cbsnews. com/stories/2003/08/12/health>.

238

ronald munson

Merrill, J., Murray, J. E., Harrison, J. H., Guild, W. R.. (1956), ‘Successful Homotransplantation of the Human Kidney Between Identical Twins’, JAMA 160: 277–82. Miller, C., et al. (2001), ‘One Hundred Nine Living Donor Liver Transplants in Adults and Children: A Single-Center Experience’, Annals of Surgery, 234: 301–10. Morelli, M. (1999), ‘Commerce in Organs: A Kantian Critique’, Journal of Social Philosophy, 30: 315–24. Munson, R. (2002), Raising the Dead: Organ Transplants, Ethics and Society (Oxford: Oxford University Press). Najarian, J., Chavers, B. M., McHugh, L. E., and Matas, A. J. (1992), ‘Twenty Years or More of Follow-Up of Living Kidney Donors’, The Lancet, 340: 807–10. Ochs, R. (2002), ‘Live-Donor Procedures Carry Risks’, Newsday.com,

. Park, K., et al. (1996), ‘A 16-Year Experience with 1275 Primary Living Donor Transplants: Univariate and Multivariate Analysis and Risk Factors Affecting Graft Survival’, Transplantation Proceedings, 28: 1578–9. Phadke, K., and Anandh, U. (2002), ‘Ethics of Paid Organ Donation’, Pediatric Nephrology, 17: 309–11. Radcliffe-Richards, J., et al. (1998), ‘The Case for Allowing Kidney Sales’, The Lancet, 352: 1950–2. Rhodes, R. (1994), ‘A Review of Ethical Issues in Transplantation’, Mount Sinai Journal of Medicine, 61: 77–82. Riether, A., and Mahler, E. (1995), ‘Organ Donation: Psychiatric, Social, and Ethical Considerations’, Psychosomatics, 36: 336–43. Ross, L. (1993), ‘Moral Grounding for the Participation of Children as Organ Donors’, Journal of Law, Medicine, and Ethics, 21: 251–7. Glannon, W., Josephson, M. A., and Thistlethwaite, J. R., Jr. (2002), ‘Should All Living Donors Be Treated Equally?’, Transplantation, 74: 418–22. Russo, M., and Brown, R. (2003), ‘Ethical Issues in Living Donor Liver Transplantation’, Current Gastroenterology Reports, 5: 26–30. Spital, A. (1989), ‘Unconventional Living Kidney Donors: Attitudes and Use Among Transplant Centers’, Transplantation, 48: 243–8. (1996), ‘Do U.S. Transplant Centers Encourage Emotionally Related Kidney Donation?’, Transplantation, 61: 374–7. (2001), ‘Ethical Issues in Living Organ Donation: Donor Autonomy and Beyond’, American Journal of Kidney Diseases, 38: 189–95. and Spital, M. (1990), ‘The Ethics of Liver Transplantation from a Living Donor’, New England Journal of Medicine, 322: 549–50. Starzl, T. (1985), ‘Will Live Organ Donations No Longer Be Justified?’, Hastings Center Report, 15: 5. Superintendent of Belchertown State School, et al. v. Saikewicz, 370 NE 2d 417(MA), 1977. Tilney, N. (1986), ‘Renal Transplantation Between Identical Twins: A Review’, World Journal of Surgery, 10: 381–8. United Network for Organ Sharing (2004), . United States Department of Health and Human Services Advisory Committee on Organ Transplantation (2002), ‘Organ Donation: Recommendations to the Secretary’, .

organ transplantation

239

Williams, R. (1995), ‘Consent for Children as Organ Donors’, Hawaii Medical Journal, 54: 498–500. Williams, S. (1986), ‘Long-Term Renal Function in Kidney Donors: A Comparison of Donors and Their Siblings’, Annals of Internal Medicine, 106: 1–8. World Health Organization (1991), ‘Guiding Principles on Human Organ Transplantation’, The Lancet, 337: 1470–1.

chapter 10 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

BIOBANKING •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

louise irving and john harris

Introduction ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

The possession, storage, and display of human tissue has until recently been a relatively non-contentious issue. Henry Wellcome, founder of one of the largest pharmaceutical companies in the world and an inveterate collector of all things unusual, had in his display, shrunken heads, a tuft of hair from King George III, two fine tattoos complete with arm skin, and a piece of the philosopher Jeremy Bentham (Gosden et al. 2003). Peter the Great’s collection of monstrosities and malformations in organisms keep the visitors flocking to the Kunstkammer in St Petersburg. It is not just public collections that hold such an assortment of human body parts. There is currently a thriving market for shrunken heads, the most prized specimens being made by the Shuar people of Ecuador, so prized because of their method of carefully crushing the skulls before extracting bone fragments through the neck. Museums everywhere hold the bodies of the long dead. In ancient Egypt the belief that the journey to the afterlife could only be made if the body was preserved required ancient embalmers to remove the internal organs and store them in jars. The pharaohs presumably consented to this, although their poor attendants, accompanying the dead and themselves condemned to a lingering death, may not have consented so willingly. Indeed, the boy pharaoh Tutankhamun is currently having his remaining DNA studied by scientists, and while he may have consented to the original burial and mummification, it is unlikely that he envisaged the eventual use to which his remains have been put. (See Holm 2001 for an excellent

biobanking

241

discussion of how we might think about an ethical framework for applying new genetic testing techniques to ancient tissue samples.) From these ancient archives, medical anthropologists can tell us that much of the same afflictions that blight lives today were around in the past. In a similar vein, relics of the Christian saints and martyrs are to be found in cathedrals and churches across Europe; likewise it is claimed that the blood of Christ can be found on fragments of the true Cross. We are very familiar with the astonishing range of human tissue, bodies, bones, and bits that form much of our cultural, religious, and anthropological heritage, and are usually undisturbed by such collections. Collections with the moral purpose of attempting to elucidate the nature of health and illness can be dated to the time of the Italian anatomists in the fourteenth century, when human tissue started to be systematically collected, stored, and studied. This long history of the research of human biological materials is the basis of much of our knowledge and understanding of disease causation and progression. ‘The human tissue archive’ is a name that has been given to the sum total of collections of human tissue worldwide, and these have been, with some justification, called ‘a research resource that is rich, unique, irreplaceable, and virtually indestructible’ (Korn 1998: 41). The practice originated in Renaissance Italy, when physicians first began systematically to perform autopsies, as a method of completing their case records. Pathology, as it is undertaken now, is generally agreed to have originated in mid-nineteenth-century Germany, where Professor Rudolph Virchnow began the systematic study of diseased tissue through the use of light microscopy. The availability of tissue archives has meant that when a connection is made between a disease and its possible causation, the requisite samples are at hand to test the theory. If such speculative research required the collecting of relevant tissue samples de novo, then many of the breakthroughs in understanding disease causation and prevention could not have occurred. It is possible that this system, responsible for so much enlightenment in the history of medicine, may be under threat from two corners (Harris 2002). First, in an increasingly individualized society, proprietorial sensibilities are raised, encouraged by various scandals involving unauthorized organ and tissue retention. Secondly, with the rise of genetic medicine and the personally identifying nature of genetic information, citizens are likely to be more cautious regarding granting permission for the storage and undefined use of their tissue samples. We believe that any diminishing of tissue archiving or any threat to its continuation would be a retrograde and severely damaging trend. In this chapter we will look at some of the chance discoveries and elegant ideas that were borne out through the availability of archived tissue samples. We then discuss some of the planned changes to the method and purpose of tissue storage and collection. The changes are in the form of new types of tissue bank, or biobank as they are conceived. These banks are part of a trend to move towards a preventative approach to public health rather than the current costly interventionist

242

louise irving and john harris

model. This approach is not without its problems and it is these that threaten the unfettered continuation of the tissue archive. The sophistication of new research tools can uncover information about individuals that may have a detrimental effect on their well-being in various ways. We analyse these possibilities in the context of how health care might develop. Much of the disquiet centres on the fact that the information held has a genetic component. This leads us to consider the nature of genetic information. We discuss some social and cultural trends that have contributed to the idea of genetic ‘essentialism’ and ‘exceptionalism’. The former is the idea that persons are reducible to, or held captive by, their genetic components. Genetic exceptionalism is a corollary of this and insists that medical information with a genetic component is different in kind from other medical information and therefore requires different treatment. This ideology, if accepted, brings with it an assortment of individualistic concepts such as property rights in human biological materials and litigation avoidance strategies that will manifest themselves as consent hurdles. This is a clear threat to the centuries-old practice of the altruistic donation of human tissue for the advancement of the common good. We close with a brief discussion of an appropriate approach to biobanking and our moral obligation to participate in research.

The Legacy of Tissue Archives ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

In the United States the National Bioethics Advisory Commission was moved to consider the rights and welfare of human research subjects and the management and use of genetic information. These deliberations also occurred in the United Kingdom, where the issues were considered by both the Human Genetics Commission and the Nuffield Council on Bioethics. The catalysts for such deliberations are new technologies for the study of human biological materials. The sophistication of the new research tools can uncover information about individuals, which may impact upon their privacy in various ways. The need to increase knowledge about human disease in order to develop better diagnostic and prevention tools needs to be balanced with appropriate protection from unwarranted harms for those who participate in medical research by donating their tissue. To give an indication of the scale of the human tissue archive, the archive in the United States alone is thought to run to more than 282 million samples in laboratories, tissue repositories, and health care institutions. Tissue samples can be collected specifically for research purposes, as part of diagnostic procedures such as biopsies, appendectomies, or blood samples. The storage may be appropriate for secondary analysis, quality control, or research purpose. The astonishing worth of such enormous tissue archives is beautifully mapped out by David Korn in his commissioned paper for the NBAC (Korn 1998).

biobanking

243

This work is impressive in its scope and detail and we draw on it liberally in this first section. Atherosclerotic cardiovascular disease has long been the leading cause of death in the United States. In fact figures supplied by the World Health Organization state that cardiovascular disease kills an estimated 17 million people a year worldwide through its effect on the functioning of the heart and blood vessels. Before the mid-twentieth century it was generally believed to be a condition that accompanied middle through old age. In 1963, during the Korean War, military pathologists documenting chest wounds from artillery noticed something interesting. Approximately three out of every four young male American soldiers showed signs of atherosclerotic changes in their coronary arteries. Comparative studies on Korean prisoners of war and Japanese civilians showed a different pattern. The careful documenting and conclusions reached through tissue analysis led to a breakthrough in the understanding of the relationship between atherosclerotic lesions and cholesterol, diet, smoking, and blood pressure. Instead of being a late-onset disease, in Western subjects it develops at a significantly early age and progresses in both severity and extent. As a result of these chance findings and the use of stored human tissue samples, changes in both surgical and medical therapies and in individuals’ approach to their health, primarily through stopping smoking, and dietary modification, brought about a revolution in preventative medicine and public health. The collection and storage of biological samples has proved invaluable for the tracking and identifying of virus breakouts. In the American south-west during the early 1990s young people started dying from a pneumonia-like illness. Analysis of tissue from the archives of the Centers for Disease Control and Prevention, which contain global samples of viruses, serum samples, and proteins, enabled initial tests of seriological screening. Testing turned up the possibility of a Hantavirus. Then autopsy tissue samples were tested with relevant Hantavirus monoclonal antibodies and genetic probes to identify the presence of Hantavirus and its source. The above example relied on the combination of the tissue archives with the observations of ‘a suspicious clinician, an astute epidemiologist [and] observant Navajo elders’ (Korn 1998: 40). The suspicions of individuals are a common theme in Korn’s report. The existence of comprehensive and available tissue archives speeded up the recognition of the carcinogenic influence of certain chemicals used in pesticides. For example, recognition that a vinyl chloride monomer (MVC) was a carcinogen causing liver tumours started with the concern of the factory physician employed where polyvinyl chloride was manufactured. The physician’s suspicions were validated through the availability of pathologic archival materials collected from collections of the tumour type. Similarly, the tissue archives of underground uranium workers led to an understanding of the maximum allowable environmental radiation exposures for workers and to the encouragement of methods of cancer prevention.

244

louise irving and john harris

Understanding of the aetiology and pathogenesis of brain and muscle diseases is critically dependent on archives of pathological samples from the central nervous system and skeletal muscle samples. Rodent models were long used for research into multiple sclerosis (MS). The use of these models led to some misleading conclusions. Only recently have studies of acute and chronic lesions in human brain samples been able to illuminate the sequence of events in the causation and progression of MS. A more sophisticated understanding of muscular dystrophy has occurred through advanced techniques and the study of human tissue. The histopathological features can now be separated into different gene mutations. There are, then, countless examples of how tissue archives have benefited mankind. The availability of archived and accessible human tissue samples permits the rapid evaluation of disease. As Korn writes: To try to initiate prospective studies de novo for each new promising candidate marker for each of the many varieties of human neoplasia would not only be extraordinarily costly in dollars and human effort, but would require study periods of many years, or even decades before definitive endpoints could be reached. In contrast, being able to apply such new technologies to archival materials, where clinical course, therapeutic response and outcome are already known, represents an incredible collapse of time and money, to say nothing of the human suffering required to evaluate the technologies, launch the necessary corroborative community trials, and possibly bring entirely new screening strategies into general application. (Korn 1998: 11–12)

This fact is not particularly well appreciated or understood, and it may be that this under-awareness of the benefits attached to tissue banking may threaten its future.

Tissue Archiv ing Now: Biobanking ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

There are now changes affecting traditional tissue archiving. It has long been known, by patient groups at least, that medicines do not have set standards of efficacy for all patients. Currently, drugs for Alzheimer’s disease work in fewer than one in three patients, whereas those for cancer are only effective in a quarter of patients. Only half the sufferers of migraine, osteoporosis, and arthritis can hope to be helped by prescription medication. The answer to this problem is thought to lie in pharmacogenetics, which is the application of human genetics to drug development. That not all people respond in the same manner to prescription drugs is due to individual genetic differences—different susceptibilities to the effect of the drug. It is speculated that those who do benefit from drugs could be identified by a genetic test which could then be used to eliminate those people who would not respond; they, in turn, might be able to benefit from other medication. The inefficiency of many drugs and the ever-increasing cost of maintaining the health of populations mean that a more progressive and innovative approach to

biobanking

245

public health is desirable. This would ideally be towards a model of prevention. It has become apparent that a smart thing to do would be to collate genetic and lifestyle information on a large scale in an attempt to discern how genetic susceptibility combines with environmental influences to impact health and longevity. National biobanks have been set up in countries with typically small gene pools such as Estonia, Iceland, and Tonga. This idea is not limited to small gene pools. The United Kingdom, with its multicultural population, has the benefit of a National Health Service and therefore the possibility of tracking the health records of individuals, and is undertaking a similar venture. To call it ‘biobanking’ is no misnomer; the plan is truly to build a bank, a resource, rather than to undertake certain particular studies. The new biobanking seeks to hold the anonymized human tissue and lifestyle information of individual volunteers as a resource for multiple users and researchers. These new-style tissue banks, which have raised awareness about the importance of tissue archives, have also raised fears about the storage and use of such data. We shall briefly map out the expected potential of ventures such as national population biobanks; detail the possible downside or fears of such a project; and then consider how the balance of potential goods might be weighed against the concerns we have highlighted.

New-Style Biobanks ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

It is thought that the elusive nature of the interaction between nature and nurture might be elucidated by comprehensive analysis of individuals’ lifestyle habits, environmental influences, and a sample of some type of biological material—blood or tissue yielding the complete genome. Most, and perhaps all, conditions detrimental to well-being and our susceptibility to infectious diseases have some genetic component. Cancer, diabetes, asthma, and degenerative neurological diseases are prime candidates for investigation. Finding the genetic factors involved is naturally complicated by the myriad environmental influences. Biobanking projects are large-scale and long-term in order to be able to see correlations between lifestyle, susceptibility, and disease. The method is to collect information from volunteers on environmental and lifestyle factors and then link these to medical records and biological samples. Considerations will be of risk factors, diagnoses, what illnesses are suffered, any disabilities, which treatments were used, and which outcomes achieved. The samples will be stored in a central database so they can be analysed by scientists undertaking ethically approved research projects. There are some fears about what types of research might be undertaken. Projects of such scale and novelty cannot anticipate the identities of all the research users or estimate their purposes. Traditionally, tissue samples removed for research and archiving have been done under sparing consent language. Indeed, guidelines in

246

louise irving and john harris

the United Kingdom developed by the Nuffield Council on Bioethics proposed that, providing there were no adverse consequences for the patient, consent to remove tissue for therapeutic reasons implied consent to any subsequent ethical use of the tissue (Nuffield Council on Bioethics 1995). Nevertheless, concerns have been raised that research may be undertaken to try to find genetic causes for behavioural characteristics such as violence or antisocial behaviour. There is validity in these concerns. Past claims for genes linked to schizophrenia, manic depression, homosexuality, and alcoholism have all eventually been withdrawn, but without the press attention that accompanied their ‘discovery’. It may be that all characteristics or illnesses have a genetic component, but the tendency to link, in the mind of the public at least, a single gene and an affliction or characteristic contributes to the idea of ‘genetic essentialism’. The notion supporting essentialism is that human beings are reducible to, or held captive by, their genetic components. It is partly the lingering nature of this discredited notion that threatens tissue archiving, and we discuss genetic essentialism in more detail later in the chapter.

What Is Genetic Information? ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

In order to understand how genetic information might be thought to threaten individual privacy, it may be helpful to know how genetic information works at the basic level. There are different types of genetic information and different ways of obtaining it. The genotype itself is simply the genetic constitution of an organism. A gene is actually a section of sequence of the chemical DNA that goes into making a particular protein. In other words, a gene is the protein coding sequence. Proteins are the class of chemicals that largely determine the structure and function of the self. At fertilization, egg and sperm, which hold a single set of twenty-three chromosomes each, join to form the double set of forty-six chromosomes which are then replicated as each new cell is formed. Chromosomes are largely made up of DNA but only a small percentage of this DNA forms our genes. The rest is termed non-coding, or ‘junk’, DNA. The function of this junk DNA is unknown at this time. The genotype gives details, from the basic DNA or protein, of the precise variations inherited from both parents. The phenotype is how these variations are expressed, for example, height, eye colour, blood pressure. It is this pattern of inheritance of different phenotypes that also supplies the information about the families of individuals. Obtaining genetic information can be achieved by analysing either the DNA or proteins or blood. Genetic information is sensitive in a number of ways. Not only can it reveal information about the individual concerned, it can also reveal information about their family. It may be able to say what that person’s susceptibility to disease is—and there is lively and robust debate in the bioethics literature about whether one has a

biobanking

247

right not to know information about one’s future health (Takala 1999, 2001; Harris and Keywood 2001; Bennett 2001). Personal genetic information is thought to differ from other types of information in several respects. Most important is considered to be its uniquely identifying nature, which can confirm, deny, or reveal family relationships. Also, genetic information can be taken from the smallest amount of biological material. This capacity means that genetic material can be secured without the consent of the person. It is this potential of genetic testing to provide information about the individual that is of interest to others—family, insurers, or employers. The question that needs to be considered is: Does the fact that genetic information can affect others and that it can potentially be used to the detriment of the person mean that it is different in kind from regular medical information? What needs to be determined is whether the difference involved requires a change to the way human tissue is thought of and managed. To discover what might be different about genetic information as opposed to regular tissue we can look at some of the issues it has raised.

Trends Involv ing Genetics ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

It is not merely the nature of genetics that poses a threat to tissue storage and collection. There have been other social and cultural changes that influence how individuals perceive both their own biological materials and even their obligations to others in the form of a shared interest in scientific advance. We live in a climate where ownership is the fundamental framework for protecting interests. That this is stretched to human biological materials is probably somewhat natural. This is quite obviously seen in the idea and practice of patenting genes. The complexities of the necessary protection required by biotech and pharmaceutical companies who invest heavily in research and development are often subsumed to the notion that the importance of human tissue or the development of cell lines lie in their inherent value. The perceived importance of genetics and biology and the difference inherent therein are given a further boost by legislation supporting such perceived importance. For example, at the time of writing, the UK government has just announced a legislative change that will deny anonymity to sperm and egg donors. This is based on the primacy of the rights of individuals to know their genetic heritage. The ruling seeks to parallel the right of adopted children to this knowledge, despite the fact that children born naturally have no ‘right’ to know their genetic heritage. Should people have knowledge of their genetic origins? We know that there are significant non-paternity rates in the United Kingdom and other countries. Nonpaternity refers to births where the children of the family are not in fact genetically

248

louise irving and john harris

related to the person they believe to be their father and who usually believes he is their genetic father. Non-paternity rates are quoted with wildly differing values (from less than 1 per cent to more than 30 per cent). A modest, and probably reliable, figure is 2 per cent. However, even at a modest rate of 2 per cent, nonpaternity rates in the United Kingdom account for over 12,000 births registered annually to men who are not in fact the genetic father. Thus, if there is such a thing as a ‘need for children to know their genetic background and true identity’, then on the grounds of numbers alone we should start with normal families. This might imply an obligation for paternity testing in all families. Also, in a climate that stresses the importance of determination over one’s biological materials, the issue of broad consent to tissue collection and archiving is problematic. For example, it may be that many people would be happy to give their tissue samples for most research activities, but not for any research into human intelligence. But if nature and nurture cannot be separated, if they are inextricably combined and our knowledge of their relationship is in its infancy, then it may be that perfectly respectable, ethical, and innovative research into the complex nature of intelligence is legitimate. It is not possible or desirable, with all our prejudices and ill-informed suppositions, to attempt to exclude or determine in advance which research can proceed. In his report on the contribution of the human archive, David Korn relates the example of Kaposi’s sarcoma (KS), a strange spindle cell and vascular tumour now associated with the HIV virus. Previously endemic in parts of Africa, it had not been associated with any specific predisposition, including predisposing infectious diseases. However, KS was identified as an early defining feature among HIV patients. Researchers at Columbia University, led by Dr Yuan Chang, used analysis of archival tissue from HIV patients to discover a unique human herpes virus, HHV8, in KS cells. The discovery of this association led to many further studies on the molecular and cellular mechanisms by which HHV8 drives the precursor cells of the KS lesions into neoplastic proliferation. For Korn this research ‘demonstrates the remarkable utility of large human tissue archives, well characterized pathologically and clinically, in supporting novel kinds or research, not predictable at the time the tissue samples were originally collected, but of significant public benefit’ (Korn 1998: 38). These cultural and social changes emphasizing individual rights and the notion that genetic information is different in kind have fuelled the debate about privacy and discrimination. It is supposed by some that the interests of individuals regarding their tissue samples would be better served by some form of property rights over their tissue (Erin 1994; also, for a discussion of the arguments surrounding this issue, see Gold 1988). Property rights are an attractive framework in that they encompass several rights. They include the right of use, transfer, possession, management, and usually the right to receive any capital value and income generated by such property. Such a framework would permit individuals to sell their genetic material,

biobanking

249

and if the material was found to be of particular value then it could be sold to the highest bidder. This individualistic approach could see the replacement of altruistic donation by a system whereby researchers bid for access to tissues. This would severely impact academic researchers and increase the research and development costs of the private sector. The availability of genetic information has potentially serious adverse emotional, social, and financial consequences. It may threaten the ability of those, already unlucky in the genetic lottery, to access decent insurance coverage or employment contracts. It is these threats that support the growing lobby of those who seek to make genetic information a different class of medical information. This trend is known as ‘genetic exceptionalism’, and we shall discuss the merits, implications, and validity of such a concept in the next section.

Genetic Exceptionalism ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Genetic exceptionalism is the idea that genetic information is so importantly different that it deserves classification as exceptional. In her informative paper on the subject Lainie Friedman Ross (2001) tracks the history and content of the debate surrounding genetic exceptionalism from its origins in the early 1990s. It arose out of the early stages of mapping the human genome, when tests for dispositions were in development despite there being no possible interventions. The term ‘exceptionalism’ in relation to medical matters came into being as a result of HIV. New practices, such as pre- and post-HIV-test counselling, the development of new consent forms, and strict requirements of confidentiality, were brought into being. The rationale was that testing would prevent the spread of the disease and those at risk might be reluctant to be tested if complete confidentiality was not assured. Confidentiality was thought to be the only way to avoid the stigmatization associated with the illness. Many of the same issues, fear of discrimination and stigmatization, apply to genetics, hence genetic exceptionalism. Ross maps out the proposed justifications for genetic exceptionalism and considers their merits. The rationale of commentators who insist that genetic information is sui generis and thus deserving a separate and stringent legislative framework is that • genetic information is immutable; • it can be detrimental to the individual; • it poses implications for familial relationships. Taking these in turn, we will elaborate on the above and, in agreement with Ross, ultimately reject these justifications for genetic exceptionalism. The immutability of genetic information is actually a rather dated concept. It was supposed that genetic testing only had to be done once in order for one’s

250

louise irving and john harris

genetic status to be known for ever. But, as is often the case, the more that becomes known about genetics, the more we realize that this is less likely to be the case. The prophetic potential of genetic information is not cast in stone. Mutations occur, and little is known about the variations within gene combinations. Ross (2001: 141) explains that this means that ‘what geneticists share with their patients about their genetic make-up cannot be considered immutable’. Even if all the alleles are known, this would still leave the question of why the gene translates into illness. There are very few single gene disorders and myriad environmental influences. Genetic testing is not the accurate forecaster that reductionist models of genetic medicine or imprecise science journalism have implied. So how might the exceptionalists fare with the discrimination argument, which, at least according to measures taken to assess public sensibilities, has the most force? Most concerns centre on the possibility of an individual’s genetic information being somehow accessible by virtue of the fact that the information is stored. We will leave aside the obvious point that one does not imply the other and the relative ease with which this can be made unlikely by appropriate access regulations, the anonymization and encryption of data, and strict penalties for abuse. It is inconceivable that employers or the insurance industry could have access without consent to an individual’s medical information. Different rules may well apply in law enforcement, but that’s a debate for another time. It may be that employers wish to test employees for susceptibility, particularly if they will be exposed to known carcinogens. As nice as it would be if no one had to be exposed to carcinogens, it may be a prudent and preventative measure for both employees and companies alike. As for insurance companies, it may be that, in the paradigm shift from interventionist medicine to preventative medicine, different models of insurance need to be introduced. Again, possible alternatives to standard models of insurance coverage are being debated in the bioethics literature (Burley 1999; also see Knoppers 1999). After all, why should it be acceptable for those who have a good indication of their likely future health condition to be able to use that as an unfair advantage against their insurers? This aside, the main reason that the threat of discrimination, victimization, and stigmatization cannot be considered justification for genetic information to be treated differently is that these sorry conditions exist wherever there is difference. Age, gender, ethnicity, religion, social class are all characteristics by which people can be discriminated against. With the proper safeguards in place, genetic information should be considerably less likely to be used as a tool for discrimination than obvious characteristics such as gender. Finally, the implications for families and kin. The justification for exceptionalism here is that genetic knowledge is likely to reveal familial relationships and risk implications. Well, that is true, and perhaps one of the reasons that it is difficult for us to accept the responsibility for the kin and risk elements of genetic information, and thus to treat it differently, is that the sacred cow of individual autonomy results in a tendency to think one’s own interests are always paramount when

biobanking

251

it comes to medical matters. But medicine is not just about individual interests, as our first section on the value of tissue archiving testifies. Public health policy, particularly in countries where the state foots the bill, needs to have strategies not solely based on individual need. Ross (2001) points out that transplantation and public health issues such as vaccinating populations are testament to health care beyond individual patients. Instead of legislative frameworks that take individuals and their desires as a prime good, Ross (2001: 141) acknowledges that there is a need for ‘an ethic that can accommodate patients as members of families and communities’. That aside, the argument against exceptionalism is that implications for families are not limited to genetic information. Sexually transmitted diseases and infectious diseases require the disclosure of what may be very uncomfortable information to family members and spouses. How to assess risk within families, and issues of how and whom to tell, need to be approached through the appropriate genetic counselling mechanisms and the application of judgement and compassion. Issues of paternity have been problematic probably as long as families have existed.

Further Threats to Tissue Archives ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

It is not just the genetic factor that has changed our conception of human tissue. Those seeking a property framework have had their case strengthened by two very damaging scandals in the United Kingdom. In both cases, public inquiries were held into allegations that the organs and tissues of children were taken and stored without proper consent. Lawsuits abound. Ironically, although the scandals caused a dip in tissue donations, this was compensated by increased public awareness of the importance and need for tissue sample contributions (Dickson 2002). But such stories serve to confirm a link between human tissue and possession, and thus ownership. Furthermore, the emotive nature of the scandal serves to elevate the importance of tissue, in itself. For example, many families exhumed the bodies of their children, sometimes more than once, to bury the retrieved body and tissue parts. This had a direct impact on the availability of tissue and a perhaps more damaging secondary effect. The problem lies in the now over-cautious approach taken by medical intermediaries. The prominence of the principle of valid informed consent is taken as such a primary good that intermediaries such as pathologists and ethical review boards insist that ‘Removal of a piece of tissue during surgery requires that informed consent be obtained from the patient, that an independent research ethics committee approve the supply of tissue to researchers and companies, and that patient confidentiality be maintained’ (Dickson 2002: 543). It has resulted in risk-averse procedures within hospitals and other traditional centres of tissue collecting, and the vitiating of liability means that the consent hurdles are much greater.

252

louise irving and john harris

We commented earlier that the UK’s Nuffield Council on Bioethics, in its report on the ethical and legal issues surrounding human tissue, argued that, providing there were no adverse consequences for the patient, consent to remove a tissue for therapeutic reasons implied consent to any subsequent ethical use of the tissue. The rise of individual autonomy as the prime moral good in bioethics, combined with decreasing expectations of social responsibility and the organ retention scandals, mean that this presumption is no longer the case (Furness 2003). The position autonomy demands now is that the patient should have control over the use to which their excised tissue or blood is put. But it has been pointed out that patients who have given informed consent to their tissue being used for research or teaching should not have these wishes overridden by research ethics committees (Furness 2003). There is confusion caused by the tension between patient as arbiter of their tissue samples and the need for good access and availability of tissue samples. Much of the tension is between researchers and the research ethics committees that are required to approve their proposals. There are guidelines and mechanisms in place to permit research where consent is difficult to obtain—for example, in cases where the patients are unable to be contacted (Medical Research Council 2001). But the primacy of patient autonomy perceived by many research ethics committees has resulted in the ignoring of such mechanisms and guidelines. This confusion extends to the researchers themselves, with many imposing restrictions on their own work by ‘an unnecessary commitment to exclude samples from patients who had died or were lost to follow up’ (Furness 2003: 39). That research in the United Kingdom has been damaged in the wake of organ retention scandals is not in doubt. It created a substantial reduction in the numbers of post mortems undertaken with the consent of relatives. The effect of this is an increasing difficulty in obtaining tissue samples (Underwood 2001). The decrease has been unnecessary as many of the parents of children who died have spoken of their willingness to donate organs for use in research if only they had been asked.

The Nature of Genetic Information ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

These ideological positions undercut the social purpose and requirement of human tissue archiving. The very nature of genetic information is a denial of our individuality and separateness. In fact, it is about how very similar we are and in itself displays the connections and responsibilities we hold to others. All threats to privacy and non-discrimination can be secured against by the appropriate legislation. In fact when polls have been taken determining the public reaction to the holding, by anyone, of genetic information, there are very strong Rawlsian ‘veil of ignorance’ responses made. These intuitions of what would be just—i.e. no discrimination

biobanking

253

on top of being unlucky in the genetic lottery—are fairly unanimous. They should be used to make the foundation of appropriate legislative conditions under which altruism can flourish without penalty. It should also be borne in mind that demands for individual property rights come from persons who, like all of us, have been advantaged by past medical discoveries. Why they should think it appropriate that such benefits should not be extended to future generations is indicative of the mean-spirited nature of much individualistic thought. There should be no genetic exceptionalism—it is medical information like any other—and can be accommodated by appropriate moral principles and concepts. The very same familial consequences cause alarm should also remind us of our deep connection and therefore our responsibilities to each other. Personal genetic information, rather than being something detrimental or alarming, is actually beneficial for all of us. The sharing of genetic information within families can help people avoid serious illness. To be informed of one’s susceptibility to a genetic disorder may permit a therapeutic intervention or allow the individual to make lifestyle changes in order to lessen the possibility of contracting a disease or disorder. But, the main importance of research in genetics is the resulting therapeutic advances gained from such research. As this benefits all individuals, we perceive there to be a corresponding responsibility by individuals to continue with altruistic participation. When this question was raised most recently in the United Kingdom, it fell to the Human Genetics Commission to consider the implications of genetic information. Their consultation document ‘Inside Information’ canvassed public attitudes to personal genetic information and how it should be treated (Human Genetics Commission 2002). They concluded that there was strong public support for research into human genetics and the benefits it is expected to bring. It was widely held that public, rather than private, ownership of genetic knowledge is preferable, and the majority of people believe in the central role of consent for the obtaining and storing of genetic information. There was considerable opposition to the use of genetic information by insurance companies or employers. The Commission concluded that the public do not wish to see people disadvantaged by their genetic characteristics. We believe that genetic knowledge and the nature of genetic information creates a moral relationship between people. Concepts and moral principles have been proposed to give guidance to all parties in regard to genetics. The Human Genetics Commission has proposed the concepts of genetic solidarity and altruism that promote the common good. It states that We all share the same basic human genome, although there are individual variations which distinguish us from other people. Most of our genetic characteristics will be present in others. This sharing of our genetic constitution not only gives rise to opportunities to help others but also highlights our common interest in the fruits of medically-based genetic research. (Human Genetics Commission 2002: 2.11)

254

louise irving and john harris

As virtually all medical facts have a genetic component, genetic information should have the same robust protections that exist for all medical information. Finally, it may be that the fears surrounding the new type of biobanking is actually less of a threat to individual privacy. The goal of large-scale tissue banks is to predict the risk of disease in populations and subgroups rather than individuals.

How to Proceed ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

The need to continue tissue archiving is paramount. Korn writes that advances in therapeutics, diagnostics, and understanding do not obviate the need for the continuation of tissue archiving and research. The most striking example of the need for this is to be seen in respect to the central nervous system and neuromuscular disorders. This is for several reasons. There are still no tissue culture or animal models accurate enough to supply parallels to the human brain. Also the most effective way of trying to determine the aetiology and pathogenesis of the brain is still based on the ‘meticulous investigation of human tissue samples in correlation with equally meticulous clinical evaluation of patients over relatively long periods of time’ (Korn 1998: 17). As with the chance discoveries we noted at the start of this chapter, the application of new technologies or invention to the study of neurological and central nervous system disorders is completely dependent on the existing tissue archives to make the necessary experiments. Good concepts and moral principles are available to us to take us into a new era of preventative medicine. Furthermore, personal genetic information may or may not be significant for the individuals involved, therefore a ‘one size fits all approach’, such as property rights in tissue, is inappropriate. There is no reason for genetic information, in itself, to be treated as being particularly sensitive. Historically and culturally, it is understood that medical information is confidential. Medical information with a genetic component is not different. In both cases the information has the potential to disclose a patient’s vulnerability. Appropriate principles can be used to enlighten the frameworks for biobanking and the utilization of genetic information. Principles of respect for persons affirm the ‘equal value, dignity and moral rights of each individual. Each individual is entitled to lead a life in which genetic characteristics will not be the basis of unjust discrimination or unfair or inhuman treatment’ (Human Genetics Commission 2002: 2.20). This reflects the instincts and intuitions received from polls of public opinion. The principles identified as appropriate for guiding legislative and other frameworks are still concerned with safeguarding individuals—as they should be. Other secondary principles advocated by the Human Genetics Commission include the principle of privacy, the principle of consent, the principle of confidentiality, and the

biobanking

255

principle of non-discrimination. Instead of a persistent debate about the individual versus society, the nature of genetics and medical progress require individuals to be seen, and to see themselves, as members of a society with shared interests in the improvement of health.

Conclusion ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

When we consider the history of tissue archiving and appreciate the great gains it has brought individuals and society, we should appreciate that removing tissue samples for research and archive was done so under sparing consent language. We cannot foresee what research ideas will spring from the elegant ideas, inspiration, lateral thinking, or observations of medical researchers. Tissue archiving must remain a ‘rich, unique, irreplaceable, and virtually indestructible’ resource (Korn 1998: 41). Our fears of threats to the privacy and fair treatment of individuals must be tempered by our experience of the great good of medical research, which seeks to improve the lives of mankind as a whole. Medical information that is genetic in nature is neither necessarily nor fundamentally different, in itself. Vigilance and appropriate legislation can safeguard individuals. The thorny problems that arise from the obligations engendered by familial connections are an addition to the bioethical canon of problems to be dealt with through robust debate.

The Obligation to Undertake Research ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Finally, it must be emphasized that we, all humankind and all societies, have the strongest of obligations to pursue promising therapeutic research and that to fail to pursue research that might save these and many other lives would be both tragic and truly immoral. Two separate but complementary lines of argument lead to this conclusion. First, it follows from one of the most powerful obligations that we have, the obligation not to harm others. Where our actions will probably prevent serious harm, then if we reasonably can (given the balance of risk and burden to ourselves and benefit to others), we clearly should act because to fail to do so is to accept responsibility for the harm that will then occur (Harris 1980). This is the strong side of a somewhat weaker but still powerful duty of beneficence, our basic moral obligation to help other people. Most, if not all, diseases create unmet needs in those who are affected, and because medical research is often a necessary component of relieving those needs, furthering medical research becomes a moral obligation. We all benefit from living in a society, and indeed in a world, in which serious scientific research is carried out and which utilizes the benefits of past research. It

256

louise irving and john harris

is both of benefit to patients and research subjects and in their interests to be in a society that pursues and actively accepts the benefits of research and where research and its fruits are given a high priority. We all also benefit from the knowledge that research is ongoing into diseases or conditions from which we do not currently suffer but to which we may succumb. It makes us feel more secure and gives us hope for the future, for ourselves and our descendants, and others for whom we care. If this is right, then we all have a strong general interest that there be research, and in all well-founded research. The human tissue archive in the past, and biobanking in the future, together constitute one of the most powerful research tools available to humankind. To turn our backs on the research that might save so many lives is literally to acquiesce to participation in the sacrifice of those lives.

References Bennett, R. (2001), ‘Antenatal Genetic Testing and the Right to Remain in Ignorance’, Theoretical Medicine and Bioethics, 22/5: 461–71. Burley, J. (1999), ‘Bad Genetic Luck and Health Insurance’, in J. Burley (ed.), The Genetic Revolution and Human Rights (Oxford: Oxford University Press), 54–60. Dickson, D. (2002), ‘Human Tissue Samples More Difficult to Obtain for Academics’, Nature Medicine, 8/6: 543. Erin, C. (1994), ‘Who Owns Mo? Using Historical Entitlement Theory to Decide the Ownership of Human Derived Cell Lines’, in A. O. Dyson and J. Harris (eds.), Ethics and Biotechnology (London: Routledge), 157–78. Furness, P. N. (2003), ‘Use and Abuse of Consent’, Bulletin of the Royal College of Pathologists, 123 ( July 2003), 38–9. Gold, R. E. (1988), Body Parts: Property Rights and the Ownership of Human Biological Materials (Washington: Georgetown University Press). Gosden, C., Olsen, D., et al. (2003), Medicine Man: The Forgotten Museum of Henry Wellcome (London: British Museum Press). Harris, J. (1980), Violence and Responsibility (London: Routledge & Kegan Paul). (2002), ‘Law and Regulation of Retained Organs: The Ethical Issues’, Legal Studies, 22/4: 527–49. and Keywood, K. (2001), ‘Ignorance, Information and Autonomy’, Theoretical Medicine and Bioethics, 22/5 (Sept.), 415–36. Holm, S. (2001), ‘The Privacy of Tutankhamen: Utilizing the Genetic Information in Stored Tissue Samples’, Theoretical Medicine, 22: 437–49. Human Genetics Commission (2002), ‘Inside Information: Balancing Interests in the Use of Personal Genetic Data’,

. Knoppers, B. M. (1999), ‘Who Should Have Access to Genetic Information’, in J. Burley (ed.), The Genetic Revolution and Human Rights (Oxford: Oxford University Press), 39–53. Korn, D. (1998),‘Contribution of the Human Tissue Archive to the Advancement of Medical Knowledge and the Public Health’, in National Bioethics Advisory Commission

biobanking

257

(ed.), Research Involving Human Biological Materials: Ethical Issues and Policy Guidance, ii: Commissioned Papers (Rockville, Md.: NBAC). Medical Research Council (2001), Human Tissue and Biological Samples for Use in Research: Operational and Ethical Guidelines,

, accessed 19 June 2006. Nuffield Council on Bioethics (1995), ‘Human Tissue: Ethical and Legal Issues’, , accessed 19 June 2006. Ross, L. Friedman (2001), ‘Genetic Exceptionalism vs. Paradigm Shift: Lessons from HIV’, Journal of Law, Medicine and Ethics, 29: 141. Takala, T. (1999), ‘The Right to Genetic Ignorance Confirmed’, Bioethics, 13: 288–93. (2001), ‘Genetic Ignorance and Reasonable Paternalism’, Theoretical Medicine and Bioethics, 22/5: 485–91. Underwood, J. (2001), Vice-President of the Royal College of Pathologists in interview with the BBC, 31 Jan. 2001, , accessed 19 June 2006.

chapter 11 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

FOR DIGNITY OR M O N EY: F E M I N I S T S ON THE C O M M O D I F I C AT I O N O F WO M E N ’ S R E P RO D U C T I V E LABOUR •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

carolyn mcleod Rachel is a beautiful woman who is faced with a difficult choice. She is in love with a handsome man; yet if she chooses to be with him, she may be happy but she won’t be rich. She could choose to be without him, however, and take a million dollars instead, in which case she would be rich but alone. What to do? If you think women only face this sort of question on the hit American reality television I am grateful to Amanda Porter for her research assistance and to Bonnie Steinbock, Andrew Botterell, and Franc¸oise Baylis for their comments on drafts of the chapter. I would also like to thank the Stem Cell Network, a member of the Networks of Centres of Excellence program, for support in the way of funding, and the Lupina Foundation for giving me a fellowship at the Munk Centre for International Studies (University of Toronto), which gave me much-needed time to write the chapter. In many ways, this chapter is a companion piece to a paper I co-wrote with Franc¸oise Baylis (McLeod and Baylis 2006). One section of this chapter, namely ‘On Commodification and Alienability’, is a version of what appears in that other paper.

for dignity or money

259

show For Love or Money (which is the inspiration for the title of this chapter), think again. The premise of the show is more real than it may first appear. If, unlike most undergraduate students in North America, you have never seen the show and think the above question might be an idle academic one, think again also. Love—for men, for children, or for themselves—has often been something women could choose only if they were willing to sacrifice their financial independence. Good women put their partners’ careers first and do not pursue demanding careers if they have small children at home. Good women, dignified women, also do not sell their sexual or reproductive services (e.g. by becoming prostitutes or contract pregnant women), even if doing so could get them out of poverty or serious debt. For women, being ‘for money’ has often meant being against love or dignity. The dilemma of dignity or money presents itself to women in new ways in the current age of technological reproduction. In some legal jurisdictions,1 women can now sell forms of reproductive labour that in the past were non-existent: they can undergo oocyte retrieval for the purpose of selling oocytes, or they can engage in commercial, gestational contract pregnancy. Oocyte vendors respond to a demand for oocytes used in treating some forms of female infertility, or for research, particularly human embryo research done on embryos that are created for the research itself (which would include some embryonic stem cell research). As oocyte vendors, women commit themselves to performing the laborious task of oocyte retrieval, and consent to use this reproductive labour not for their direct reproductive benefit, but for the reproductive benefit of others, as women do with contract pregnancy.2 While women who sell such labour may get healthy sums of money in return, they also may sacrifice their dignity as women. Traditionally at least, dignified women did not treat their reproductive potential as a source of cash. At issue here is the moral permissibility of commodifying women’s reproductive labour, particularly given the double bind, or binds, that such commodification poses for women. Should women be able to treat their own reproductive labour as a commodity, that is, as something that can be traded for a price? Should others encourage them to do so, despite the difficulty many women would find in choosing whether to sell such labour? Assuming that ethical restrictions exist on what things can properly be commodified, it is an open question whether women’s reproductive labour is among these things. This question is both open and difficult, for a number of reasons. For example, while the financial independence that the selling of such labour could offer women is in some sense empowering, the work 1 These include many states in the United States, but exclude, for example, all areas of Canada, Britain, France, and Australia. 2 Following Donna Dickenson (2001: 211), I assume that when women sell or donate oocytes or embryos, they are engaged in a form of labour, since the oocyte retrieval process involves mental or physical exertion as well as physical pain on the woman’s part (see below for a description of this process). Labour is simply the ‘exertion of the faculties of the body or mind, especially when painful or compulsory’ (Oxford English Dictionary online).

260

carolyn mcleod

itself seems to be degrading. Feminists need to sort through such difficulties, and do it soon, because of the growing market in, and the growing pressure on women to provide, reproductive service through oocyte vending and commercial contract pregnancy. Some feminists have tried to provide answers to this problem; but unfortunately, their answers tend to conflict. Whether the subject is oocyte vending or commercial contract pregnancy, some feminists argue that being able to sell reproductive labour is empowering for women in general (i.e. not just financially); to deny women this right would be to treat them in a manner that is inconsistent with their status as autonomous persons. By contrast, other feminists claim that allowing women to sell reproductive labour is degrading, and hence ultimately disempowering, for women. The debate is reminiscent of feminist debates about pornography and prostitution. Are oocyte vending and commercial contract pregnancy reproductively liberating for women, just as pornography and prostitution are sexually liberating for women, according to some feminists (Vance 1984)? Or is the liberation really just disguised subordination? For ease of exposition, and borrowing some terminology from For Love or Money, I will say that those who claim the former are for money or in favour of commodification, which may or may not be compatible with being for dignity (i.e. women’s dignity) or against women’s subordination. My main purpose in this chapter is to lay out the ‘for money’ and ‘for dignity’ arguments that feminist ethicists have given about the reproductive labour women perform in providing oocytes or in getting pregnant for others.3 Feminist arguments about the morality of these two practices overlap significantly because, from a feminist perspective, the morally relevant facts about them are quite similar. Still, there are dissimilarities, stemming from the obvious fact that one practice involves giving up oocytes while the other involves giving up a baby after a pregnancy (Steinbock 2004: 255). Some arguments by feminists reflect this core difference, in that they apply specifically to one practice but not to the other. I shall highlight when the relevance of a particular argument differs for these different reproductive practices. The structure of the chapter is as follows. I begin with a discussion of the meaning of ‘commodification’ and of a related term, ‘alienability’, followed by a description of the commodification of women’s labour in providing oocytes and in undergoing contract pregnancies. I then elaborate on why having to choose dignity or money with respect to such labour is a double bind for women. In the next part of the chapter, which is the bulk of it, I explain how feminists have dealt with this dilemma of dignity or money. The chapter ends with a summary of the state of the feminist literature on this topic, along with recommendations for future feminist inquiry.

3 My exposition covers only the feminist literature, and thus excludes such non-feminist arguments as religious arguments against ‘sinful’ uses of one’s reproductive capacity (see Steinbock 2004: 256).

for dignity or money

261

On Commodification and Alienability ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

To begin: when exactly is women’s reproductive labour commodified? And is commodifying it a bad thing, given the nature of commodification? In other words, is commodification inherently bad? It turns out that commodification can be, morally speaking, malign or benign. Whether one commodifies something malignly depends in part on whether that thing is normatively alienable to persons (or to other beings with moral worth). Something is alienable to us if it is separable from us; and something is normatively (or benignly) alienable to us if it is separable without causing us harm or degradation. When we treat something that we possess as a commodity, we treat it as an ‘item of trade’ (OED), that is, as something that we can trade away and therefore separate from ourselves, to some degree at least. But we cannot separate everything from ourselves and remain intact as persons; therefore, we cannot commodify everything benignly. Understanding commodification and alienability, and how they connect up with one another, is crucial for navigating smoothly through the ethical debate on commodifying women’s reproductive labour.

Commodification First consider commodification: when we commodify something, we ‘take that which is not already a commodity and make it into, or treat it as though it were, a commodity. Simple enough, but what the heck is a commodity?’ (McLeod and Baylis 2006: 3). Marx (1867/1954: 43) wrote that ‘A commodity is, in the first place, an object outside us, a thing that by its properties satisfies human wants of some sort or another.’ The satisfaction of human wants is the use value of the commodity for Marx. As he notes, however, something can have use value without its being a commodity. For example, I might (poorly) design a coat and only wear it myself, to satisfy my desire for warmth, rather than sell it or trade it. In that case, the coat is not a commodity. ‘To become a commodity a product must be transferred to another, whom it will serve as a use-value, by means of an exchange’ (Marx 1867/1954: 48). In other words, a commodity is fundamentally an item of trade. As such, it has exchange value as well as use value. In his description of commodities, Marx (1867/1954: 43–87) speaks of commodities as though they were essentially objects, which is not obviously the case. Arguably, a service can be a commodity, and a service is not an object. With some services, the true commodity may just be the product of the service (e.g. a clean house in the case of cleaning as a service); but with other services, such as reproductive services, there may be no product in the end (i.e. no baby or no oocyte), yet the person who performs the service may be compensated nonetheless, which suggests that the service itself is a commodity. For example, some women

262

carolyn mcleod

who enter into contract pregnancies are remunerated to some degree for their reproductive service even if the pregnancy ends in a miscarriage or a stillbirth. Thus, a commodity is an object or service that one trades for something of equal value, typically money. To commodify something, then, is typically to turn it into, or treat it as though it were, an object or service that one trades for a price. As Margaret Radin (1996) emphasizes, this process can be complete or incomplete; one can commodify something only to a degree. Commodification can also be morally benign or malign (not unlike objectification; see Nussbaum 1995). Relevant factors in determining moral permissibility with commodification include the following: (1) ‘whether the thing commodified has intrinsic value that is incompatible with its being’ treated as a commodity, where an example might be a religious artifact; (2) ‘whether moral constraints exist on the alienability of the thing from persons,’ as they do in the case of lifesustaining organs; (3) ‘whether the consequences of making the thing alienable and of commodifying it are’ unfavourable, which may be true for non-life-sustaining organs, such as second kidneys (McLeod and Baylis 2006: 3). (2) and (3) both concern (normative) alienability, which is the focus of the feminist literature on the permissibility of commodifying women’s reproductive labour.4 Should the labour be commodified, given that women will have to alienate it from themselves, along with the products of their labour?

Alienability Alienability is related to commodification in that it shapes whether the latter is benign, as illustrated above. A legitimate commodity is an object or service that is normatively alienable to persons, meaning they could transfer or forfeit it without doing damage to their selves (Radin 1996: 17; Bartky 1990: 34). For example, people can transfer their savings into mutual funds without harming or degrading themselves. Yet they cannot exchange for money the protection afforded by their basic human rights without causing such harm. Alienating something from the self that is not normatively alienable to it brings on a state of alienation. Here, the self is so fragmented from what is constitutive of it that it cannot be itself, cannot be psychologically integrated, or cannot be truly human. Marx thought that alienation occurs when workers are forced to engage in productive activity that is neither free nor creative. Doing the opposite of such activity—that is, doing truly stimulating and imaginative work—was definitive of humanity for Marx (1844/1964). 4 (1) applies to whether the products of this labour (i.e. oocytes, embryos, fetuses, and babies) have such intrinsic value that they should not be commodified, which relates directly to the commodification of them, as opposed to the labour that goes into creating them. This chapter deals only with issues that directly concern the latter.

for dignity or money

263

In the Western philosophical tradition, our humanity resides primarily in our autonomous agency, and in what sustains that agency, although what does sustain it is a subject of considerable controversy. Feminists tend to disagree with mainstream moral theorists on this point about sustainability; they tend to define autonomous agents as more relational (i.e. sociopolitically constituted), more embodied, and more emotional than non-feminists do (Mackenzie and Stoljar 2000). What might count, therefore, as benign commodification based on what is normatively alienable to persons may differ for feminists compared with non-feminists. As we shall see, some feminists base their moral analyses of the commodification of women’s reproductive labour on feminist accounts of autonomous agency, and, more specifically, on whether women can maintain their autonomy, or personhood, while separating themselves from the emotional and embodied aspects of their labour.

The Reality of ‘Dignity or Money’ with Respect to Women’s Reproductive Labour ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Before turning to feminist arguments in favour of dignity or money, we should be clear on what the facts are concerning real practices of commodifying reproductive labour. What are the morally relevant facts, particularly from a feminist perspective? When women engage in oocyte vending or commercial contract pregnancy, what do they trade away, for what, why, under what conditions, and with what consequences to themselves? First of all, what do women trade away? The most obvious answer perhaps, assuming that their labour fulfills its purpose, is, in the case of oocyte vending, their own oocytes, and in the case of contract pregnancy, a baby that is their own in the sense that they gestated the baby (and would also be genetically related to it if their contract pregnancy were traditional or genetic, as opposed to gestational). But women also give away a lot of their own sweat and, possibly, tears when they perform reproductive labour for others. Even to get chosen in the first place to do such labour, women have to fill out forms that can be longer ‘than college applications’ (J. Cohen 2002); and they often have to undergo a series of medical tests, including psychological tests to rule out mental health problems (McShane 1996: 32; Serafini et al. 1996: 38, 39). The serious work begins, however, only if they succeed in getting chosen: they must have oocytes retrieved in the case of oocyte vending, or get pregnant, carry the pregnancy to term, and deliver a baby in the case of contract pregnancy. The process of oocyte retrieval involves uncomfortable daily hormone injections, frequent blood tests and ultrasound examinations, and

264

carolyn mcleod

the often painful procedure of retrieving the eggs, which proceeds by laparoscopy or vaginal ultrasound normally while the woman is under partial sedation (see Fielding et al. 1998: 274; New York State Task Force 2002). Gestational contract pregnancy includes weeks of hormone injections, as well as uncomfortable tests to assess potential blockage in Fallopian tubes, and the sometimes painful procedure of embryo transfer. What do women get in exchange for this labour? Aside from possibly feeling good about themselves for helping others, they can receive large sums of money. With oocyte vending, the going rate is anywhere from $1,500 to $5,000 in the United States, if the women are paid only for their ‘time, effort, and discomfort’ (New York State Task Force 2002). Yet sometimes the women are paid also for their oocytes, that is, if they provide them to couples who want a ‘designer baby’: often one who is intelligent, musical, and good-looking. Some of these couples are willing to pay thousands of dollars for oocytes (e.g. $50,000), at least according to advertisements in Ivy League college newspapers (Steinbock 2004: 259). Such payments resemble those given for gestational contract pregnancy, for which payments tend to run between $18,000 and $25,000 (again in the United States).5 Why do women perform the labour? Do they just really need the money? In general, women who vend their services of oocyte retrieval and pregnancy are not desperate for the money, although they could certainly use it. Oocyte vendors and contract pregnant women tend to be less well off financially than the couples who pay them. Sometimes, they are in university and face the prospect of large student debts. But often they are working class women, with little to no postsecondary education, who have children of their own to support (Fielding et al. 1998: 276; Ragone 1994: 54–5). Many reproductive labourers say that financial reward is an important consideration, but not their primary motive for performing the relevant service (Ragone 1994: 57; 1999: 78). More often than not, they want to do what they are doing because it helps people or because, in the case of contract pregnancy in particular, it makes them feel special (perhaps because they get to be pregnant, and being pregnant is special) (Anderson 1993: 180).6 Under what conditions are women performing this labour? Usually, with oocyte vending and contract pregnancy, a power differential exists between relevant parties, which does not favour the woman as reproductive labourer. The differential can be based on gender, class, or race. For example, traditional or genetic contract pregnancies often occur against a background of gender- and class-based inequalities, 5 Source: David Smotrich, reproductive endocrinologist and medical director of La Jolla IVF in San Diego, California; personal communication, 2 Oct. 2004. These amounts are what contract pregnant women get paid only if they produce a baby in the end. 6 Other possible reasons include, with oocyte vending, wanting ‘to know if [one’s] eggs are ‘‘good’’ ’ (Steinbock 2004: 258), and, with contract pregnancy, seeking to resolve guilt feelings about a past abortion (Anderson 1993: 180), presumably by carrying a fetus to term this time, rather than ending its life.

for dignity or money

265

since the contracts exist between the sperm provider, a man who is usually well off, and the contract pregnant woman, who is usually not so well off. With gestational contract pregnancy, racial inequality can occur, and in a way that benefits the paying couple, because of racial differences between them and the woman who bears a child for them (see Ragone 2000: 65–6). Such inequalities may not at all compromise the moral legitimacy of these contracts; but they could do so because of how vulnerable they can make the reproductive labourer. The conditions under which women exchange their reproductive labour for money can also involve manipulation or deception. Reproductive labourers are manipulated when information about the nature of the relevant medical procedures is withheld from them, and when the ‘downstream commodification’ (Holland 2001) of their labour is withheld. Recent studies show that withholding information about procedures (and consequently undermining informed choice) is common in oocyte ‘donation’ programs in the United States (Gurmankin 2001; New York State Task Force 1998). For instance, in preliminary attempts to gather information from many of these programs, prospective vendors will receive either no information about risks or inaccurate information (Gurmankin 2001). Often, they will not be told about the potential risk of ovarian cancer from the hormone injections, and about the risks of infertility or of ovarian hyperstimulation syndrome, which in severe cases is life-threatening (Fielding et al. 1998: 274; Serafini et al. 1996: 37). The motivation to manipulate or coerce women to sell their reproductive labour can be quite strong because of a shortage of ‘desirable’ labourers or simply because reproductive labour is big business for the intermediaries involved, that is, for those who run the programs.7 Enterprises in oocyte ‘donation’ are highly lucrative if the labour is commodified downstream from payment to the actual labourers (Mahoney 2000). Oocytes may be sold in this way to researchers or to infertile couples for significant profit, often without the woman who originally sold the oocytes knowing about that profit (Holland 2001: 266). What are the consequences for women as reproductive labourers? In particular, what are the emotional consequences? Reportedly, they can include regret or shame for agreeing just to perform the labour or for agreeing to specific acts outlined in the contract (e.g. genetic abortion). For example, a woman might regret her consent to be an oocyte vendor after she discovers how painful the process of retrieval can be.8 A woman who serves as a contract pregnant woman might wish that she had never consented to abort her pregnancy if the fetus has a genetic abnormality, to reduce her pregnancy selectively if it is a multiple one, or just to give ‘her’ child away.9 Of course, not all women who sell or donate their reproductive labour experience 7

See J. Cohen (2002) and Ragone (2000: 61) on the difficulties that couples can have with finding the labourer they want. 8 A student of mine had that experience. 9 Consents to both genetic abortions and selective reductions are common in pregnancy contracts. Selective termination is an issue with gestational contract pregnancy, because multiple pregnancies are

266

carolyn mcleod

regrets; presumably the ones who say they would do it again feel little to no remorse about doing it the first time.10 Women who respond to the market for reproductive labour can also experience rejection that threatens to undermine their dignity, or others’ dignity. While programs can reject a prospective labourer for reasons that are not morally troubling—for example, the woman’s health is truly poor—they can also reject a woman for reasons that are troubling—for example, she is not pretty enough, smart enough, or talented enough. Such standards may be appropriate for beauty contests or college admissions, but they are arguably inappropriate as standards for determining whose gametes to use when creating a new human being. The reason is that, in this context, the standards imply that some lives (including that of the gamete provider herself) were not, or are not, worth creating. The abovementioned facts suggest that in consenting to oocyte vending or commercial contract pregnancy, women might alienate from themselves more than just the physical act of reproductive labour. They might also lose some autonomy owing to manipulation, some integrity owing to regret, and some dignity owing to rejection. Since such qualities are normally deemed inalienable to persons, their loss is morally regrettable to say the least.

The Double Bind of Choosing Dignity or Money ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Given the negative impact that the market in reproductive labour can have on women, one might wonder why feminists are so divided about it. Why not just ban the market? What would we be sacrificing if we did that? It has to be something important, unless in disagreeing with one another so strongly, feminists are way off track. On the contrary, I think they are right on track in struggling so much with this issue, which is quite complex. Let me explain, drawing on Radin. In having to decide whether to endorse the commodification of women’s reproductive capacities, feminists are in a serious double bind. In other words, as feminists, they are damned if they allow the commodification of women’s reproductive labour and damned if they don’t. If they do the former, they risk allowing women to be exploited, especially if substantial downstream commodification occurs. Exploitation is an issue for all potential reproductive labourers, but particularly for poor or economically deprived women, who may find the payment common in such arrangements, since multiple embryos are transferred to the uterus of the contract mother during the IVF process (Ragone 2000). 10 In one study of oocyte donors, 60 per cent of them said they would donate again (Fielding et al. 1998: 279).

for dignity or money

267

for oocyte retrieval or contract pregnancy irresistible. A further problem with commodification is that it would ‘seem to treat [women’s reproductive capacities] as . . . fungible market commodities’, which ‘in this culture . . . [could] diminish the personhood of women’ (Radin 1991: 349). In other words, women could be objectified as breeders when their reproductive labour is commodified. Women may themselves go to market, as cows do. But banning commodification would not obviously serve women’s interests either. For one thing, being able to sell their reproductive labour gives women market power, which is power within capitalism. Yet women have ‘historically been denied [this power]’, and currently have a lot less of it than men, which suggests that barring women from it in the realm of reproduction could really harm women (Radin 1991: 349). In particular, it could harm poor women who might better their situation a lot by responding to the market for oocytes or for contract pregnancies. Sometimes, guidelines for the latter require ‘that the candidates not be poor women who need the money’ (Radin 2001: 309). Besides being paternalistic—perhaps ‘in the worst way’11 since the paternalism involves wealthy people telling poor people what they cannot sell—such measures force poor women to ‘remain in bad circumstances’ (Alpers and Lo 1995: 42). They prohibit ‘desperate exchanges’ (Walzer 1983) because exchanges founded on the desperation of one party are wrong; but they do nothing to ease the desperation. Moreover, if banning commodification meant encouraging more altruism (i.e. altruistic donation or contract pregnancy), then women would not necessarily benefit. Women have traditionally been the ‘care-takers of the world’ (Mahoney 2000: 188), which has probably brought them more toil and suffering than joy. Relying on women’s acculturated desire to help others as a way to ensure that oocytes and babies are available to infertile people would simply perpetuate sexism. Thus, various things that are bad from a feminist perspective could happen if we banned commodification, and if we did not. While a ban would do nothing to alleviate, and might even encourage, the disempowerment of many women, the absence of a ban leaves women open to exploitation and to breeder status. This dilemma of dignity or money exists largely because of women’s oppression. If women as a social group were not so oppressed that they were taught to be self-sacrificing nurturers, were excluded from powerful positions in the market economy, or were regularly forced into poverty upon divorce and becoming single mothers, the issue of commodification would not be so troubling for feminists. Oppression has this effect of creating double binds (Frye 1983); it keeps people down by ensuring that they repeatedly face choice situations in which the only available options are grim ones. One might argue that rather than choose one side of the bind of dignity or money, we could simply work toward eliminating the oppression of women—that 11

Alta Charo put it to me this way in conversation.

268

carolyn mcleod

is, toward removing the conditions that create the bind. Do we not just promote women’s oppression anyway if we accept the awful choice situations that it puts women in? Still, do we really have the choice not to respond to these situations? If, instead, we simply worked toward ‘ideal justice’—that is, no oppression but rather real reproductive and economic freedom for women—we might miss the chance to create ‘nonideal justice’, or a situation that is at least more just than our current one. Radin calls this broad dilemma of ideal versus nonideal justice ‘the double bind’ (1996, ch. 9; my emphasis). Feminist activists come up against it all of the time, in having to balance ‘their aspirations for a right world’ with their ability to right some wrongs ‘in the here and now’ (Shultz 1990: 337). Feminists who do embrace one side of the bind of whether to commodify women’s reproductive labour are sensitive to varying degrees to the strength of this bind and to the need to fight against the oppression that causes it. Let me outline their arguments, starting with those that are for money.

For Money ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Past feminist arguments for money focused on contract pregnancy (e.g. Shultz 1990; Shalev 1989), while present arguments deal mostly with oocyte vending (e.g. Mahoney 2000; O’Donnell 2000). Further, most of the arguments—past and present—apply equally well to both practices. Some of them are merely arguments for money, while others are arguments for both dignity and money—assuming that being for money is compatible with being in favour of women’s dignity. (Who would admit to being against it?) I deal with the latter sort of arguments in the next section, and focus on the former here. Arguments that are (merely) for money say that commodifying women’s reproductive labour is preferable to the alternatives. And the alternatives include: (1) a system of donation that relies upon altruism; (2) being paternalistic toward women and telling them that they cannot perform reproductive labour for others at all; and (3) having a black market crop up in response to a ban on commodification. Arguments for money exist that address each of these possibilities.

Altruism Feminists for money claim that a commercialized system is preferable to an altruistic one, because the latter would prey upon the socialization of women as ‘care-takers of the world’ (Mahoney 2000: 188, 192; Shultz 1990: 380). Most programs for oocyte vending and contract motherhood prey upon this socialization process now. They weed out women who apply because of the money rather than the opportunity to help others in need. For women who do become ‘donors’ or

for dignity or money

269

contract pregnant women, the programs tend to characterize the compensation they receive as ‘reimbursement’ rather than purchase (Mahoney 2000: 188), presumably to avoid the bad press the programs would get by admitting that they buy and sell women’s reproductive labour. According to feminists for money, the programs should be buying and selling the labour, and be open about that fact, because otherwise they exploit women (especially if they profit themselves from downstream commodification), and they promote sexist stereotypes about women. On the second point, Julia Mahoney writes that, ‘the implication that young women should desire to undergo a series of highly uncomfortable procedures that pose both short-term and long-term risks to their physical well-being for which they [may not or] will not collect the market clearing price threatens to reinforce stereotypes of females as generous rather than self-interested’ (2000: 188; my emphasis on ‘women’). The fact that we are asking women to be generous in this area is significant, given our history of relying on women to respond to others’ needs, especially surrounding childbearing and -rearing; and given that we do not have similar expectations of men (for whose sperm we pay).

Paternalism Some feminists would agree that we should not expect women to do reproductive work for free, especially when we pay men for much less arduous ‘work’ (as if masturbating in a cup was work!); but better yet, they would argue that we should not expect women to do the work at all. On their view, we should discourage women from becoming reproductive labourers because of how unlikely it is that women will choose to do so autonomously. Barriers to women’s autonomy in this area lie in sexist norms about women, the nature of the relevant work (particularly with contract pregnancy), and poverty. For example, norms about women’s worth residing in their reproductive potential, which could make a woman want to be pregnant even if she cannot keep the child she bears or even believe that the child is hers during the pregnancy (a common attitude amongst women in gestational contract pregnancies; Ragone 2000: 69), can profoundly shape a decision to undergo a contract pregnancy. Moreover, as some argue, entering into a contract to relinquish a child at birth before it has even been conceived is not something women can do freely, for they cannot predict how they will feel about a pregnancy once it occurs. In response to these objections about autonomy, feminists for money say two things: first, such arguments give too little credit to women as autonomous agents, and, second, they reinforce stereotypes that women lack autonomy. In other words, the objections support a form of paternalism that is profoundly disrespectful of women. Women are not dupes in the face of pronatalism, according to these

270

carolyn mcleod

feminists; women can rationally weigh the risks and benefits of contract pregnancy and of oocyte vending, and critique their own reasons for wanting to do either. Similarly, with contract pregnancies, women can anticipate and factor into their decision-making possible shifts in their perceptions of their pregnancies, especially if they have been pregnant before, which is true of most contract pregnant women (Ragone 1994: 54). The opposite idea—that women cannot enter into pregnancy contracts because they might have a change of heart—encourages a view of women as ‘unstable, as unable to make decisions and stick to them, and as necessarily vulnerable to their hormones and emotions’ (Shultz 1990: 384). The worry that they will change their minds in favour of keeping a baby that they gestate in a contract pregnancy also reinforces the view that ‘some kind of instinctive maternal bonding to the fetus’ occurs in pregnancy, which would be true only if women were maternal by nature (Shalev 1989: 121). But since most feminists reject reproductive essentialism for women, they should also reject paternalistic stances toward contract pregnancy in general. Or so say feminists for money.

Black Market In the view of these feminists, banning commodification would only lead to a black market in women’s reproductive labour anyway, which could not be good for women. With a black market, there may be little incentive, if any, to ensure that women’s participation is autonomous. There may instead be a strong incentive to coerce women into participating because of the ability either to reap large profits if one is an intermediary, or to get the oocyte vendor or contract pregnant woman of one’s dreams if one is a buyer. Moreover, a black market would almost inevitably heighten inequalities in power between women, including class- and race-based inequalities. While it is true that what the market will bear would determine payment, surely classism and racism would shape what the market will bear. Lowerincome women of colour would be paid less than middle-income white women either because the latter’s genes or wombs are deemed more desirable (because of classism and racism) or because the former may be desperate and take whatever money they can get.12 The net result of a ban on commodification, according to these feminists, is a situation that would be much worse than what we get with commodification. Feminists for money generally favour a regulated market in oocytes and contract pregnancies, compared to a black market or a system of donation. (Hence, what 12

With respect to oocytes specifically, Ann Alpers and Bernard Lo refer to a two-tiered market that may already exist: ‘one high-priced market for the eggs of white middle-class women to be sold to infertile couples seeking IVF treatment and another for the eggs of lower-income women of color whose ova are valued for research purposes only’ (1995: 42). In the latter case, the ova may be used to create embryos that will ultimately be destroyed, adding, perhaps, further insult to the injury of getting a relatively low price for one’s oocytes.

for dignity or money

271

they actually endorse is incomplete commodification. One does not commodify anything completely if one regulates the market for it. With regulation, non-market values are at play determining what the regulations are, which in turn limit property rights in the commodified item or service.) Feminists for money also tend to believe that regulation could solve many of the problems that people often see with commodification. For example, governments or medical associations could impose a cap on payments for reproductive labour so that no group of women gets considerably more than any other group and no woman finds the payments irresistible. The American Society of Reproductive Medicine already recommends a cap of $5,000 (an amount that may be too high) on compensation for oocyte ‘donation’ (ASRM 2000). In addition to policies that concern payment, we could have ones about informed choice that ensure oocyte vendors and contract pregnant women are always free, informed, and competent. The process of informed choice for either practice could always occur in the presence of a feminist counsellor even, who would be attuned to how pronatalist norms can (but do not necessarily or even frequently!) undermine autonomy for women in reproductive choice situations. With policies like these in place, women who are reproductive labourers would not be disempowered breeders. They would not be the handmaids of Margaret Atwood’s Handmaid’s Tale (1985). We have seen that a regulated market in women’s reproductive labour is preferable to various alternatives, according to feminists for money. We will now see that they defend such a market as well because it is a good or a just option, rather than simply the best amongst a host of possibly bad or mediocre options.

For Dignity and Money ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Feminists for money assert that receiving payment for reproductive labour is dignifying for women for three main reasons.13 (1) Within capitalism, being paid to do things for other people is a sign of respect. (2) Getting paid to do reproductive labour for others can also enhance women’s autonomy by fulfilling autonomous desires they may have to sell that labour. (3) Such payment disrupts patriarchal ideals of motherhood or womanhood. Let me deal with each of these points in turn.

Respect Financial compensation and respect are intimately connected, whether the person who receives the compensation desires it or not (Shultz 1990: 336; Mahoney 13

These feminists are therefore for dignity and money; but since they are the same feminists as those who we were just discussing, for simplicity’s sake, I will continue to refer to them as ‘feminists for money.’

272

carolyn mcleod

2000: 205). While according to one strand of thought, monetary exchanges are less valuable than gift exchanges, ‘In fact, there is a strongly competing truism suggesting that that which we reward with money is that which we value. In particular, the inability of women to gain monetary recognition for the things they uniquely or preeminently do is one of the core causal factors in the exploitation of women’ (Shultz 1990: 336). Paying for things suggests that we value them; but more to the point, paying people for what they do for us can show that we acknowledge them not as mere instruments for our use, but as people. Not paying others, especially when we reap substantial profits from their work (as some programs or clinics do that facilitate oocyte ‘donations’ or contract pregnancies) is exploitative. Marjorie Shultz implies that women are already exploited in many of the things they do for others. On her view, we should avoid adding reproductive labour to the list.

Autonomy We should also respect the autonomy of women who freely agree to relinquish their oocytes for a price or to bear a child for someone else, according to feminists for money. We should do so because women’s dignity demands it, not only because the alternative of being paternalistic is bad. Permitting the commodification of reproductive labour acknowledges that women can do what they want with their lives and with their bodies. Some theorists frame the moral issue of commodification as an issue of respect specifically for women’s bodily autonomy. Richard Arneson writes that ‘legal toleration of surrogacy presupposes that the woman’s body is hers and hers alone unless she consents to some particular use of it’ (1992: 162). In other words, if reproductive labour is commodified so that the women who perform it get to trade it for money, women’s bodies must be their own. Donna Dickenson (2001) and Kath O’Donnell (2000) make the same sort of claim but in the reverse: women’s bodies will be their own if we grant them some sort of property right in their bodies. Such a right would give reproductive labourers some control over who profits from their labour and some ability to protect themselves from exploitation. Whichever way they put the point, a strong connection exists for these feminists between commodifying reproductive labour and viewing women as the rulers of their own bodies specifically, and of their own fates more generally.

Disrupting Patriarchal Ideals Finally, we honour women’s dignity by paying them for reproductive labour performed for others, in the view of feminists for money, because payment, especially for contract pregnancy, helps to dismantle patriarchal ideals of motherhood and womanhood that are undignifying for women. According to these ideals, women

for dignity or money

273

undergo the ‘emotionally volatile condition’ of pregnancy (Shalev 1989: 121) for its own sake or in order to feel fulfilled as women. While pregnant, they develop ‘sacred bonds’ (Chesler 1988) with their unborn children that make it impossible for them to be separated from their children upon birth and for many years to come. Thus, with gestation comes motherhood, whether the pregnant woman intended to be a mother or not. In stark contrast to this picture, we have the practice of commercial contract pregnancy, in which women rationally choose to enter into and then honour a contract to bear a child for someone else in exchange for money, and not for mere feminine fulfillment. The contract pregnant woman acts rationally and need not bond so emotionally with her unborn child that she could not imagine life without it. (Gestation, therefore, does not entail motherhood.) Moreover, the mother of the child in the end (i.e. if the child is to have a mother, as opposed to one or two fathers) will not be the person who gave birth to it, but rather someone who paid someone else to give birth to it. Clearly then, with contract pregnancy, especially of the commercial variety, motherhood and womanhood cannot be what they are under patriarchy. For example, the ‘standard of motherhood’ must be ‘intent-based’ as opposed to gestational (Anderson 1993: 183), which is as it should be according to Shultz and Carmel Shalev. A parent of a child should be someone who intended to raise that child, or to bring him or her into existence. Shultz and Shalev object to the gestational standard because it does not allow women the dignity to choose their own life path. It binds women to children in circumstances in which women did not intend to be mothers, which is true of women in contract pregnancies and in unwanted pregnancies. Thus, feminists for money suggest that women not only maintain the level of dignity they currently have, but in fact gain some dignity when their reproductive labour is commodified. If commercial reproduction were normalized, less of a connection would exist in people’s minds between women’s reproductive activity and motherhood. The end result would be tangible freedom for women to pursue goals other than becoming a mother. While this outcome would benefit all women, the actual compensation for reproductive labourers would support them specifically. It would show respect for these women as subjects rather than objects to be used by others, and as autonomous beings, assuming that they autonomously choose to sell their labour.

For Dignity ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Underlying feminist arguments for money is the belief that we should deal with women’s reproductive labour as we do other forms of labour for which a market exists (e.g. university teaching, rough carpentry). As with other labour markets, the reproductive one should be regulated to prevent discrimination as well as forced or

274

carolyn mcleod

coerced labour. People who respond to that market by offering their reproductive services to others should be paid what the market will bear and what the relevant regulations will allow, according to feminists for money. To put a label on their view, these feminists defend a ‘symmetry thesis’ with respect to women’s reproductive labour: our treatment of it should be symmetrical with our treatment of other forms of labour.14 Feminists for dignity (alone) defend instead an asymmetry thesis. They contend that women’s reproductive labour is special, in that, unlike other forms of labour, no one should sell it, nor consent to perform it for others for free perhaps. The labour is special either because of inherent features of it, which is the view of some feminists for dignity, or because of contingent features that exist when the labour is performed in environments that oppress women, which is Satz’s (1992) view.

The Labour Is Inherently Special So some arguments for dignity (alone), or against the commodification of women’s reproductive labour, say that because of inherent features of such labour, the commodification of it is always malign. Let me consider two such arguments. One defends what I will call the ‘identity thesis’: that because a woman’s reproductive activity is so intimately tied to her identity, it can never be an item of trade. The other argument, which is relevant only to contract pregnancy, concerns a thesis that I discussed earlier: that a woman’s autonomous perspective on her pregnancy may evolve in such a way that she could no longer perceive the child within her as one that belongs to someone else; and the possibility that such a shift in perspective will occur suggests that she should not treat pregnancy as labour like any other form of labour. Let me call this thesis the ‘autonomy thesis’.

The Identity Thesis According to the identity thesis, reproductive activity is tied to the identity of women either as women or as persons. Feminists have used versions of this thesis to object first to contract pregnancy and then to oocyte vending and embryo donation. For example, Carole Pateman is described as asserting, in opposition to contract pregnancy, that ‘a woman’s self is intimately connected to her body in its reproductive function’ (Arneson 1992: 161; Pateman 1988). A woman’s identity as a woman is so tied to her reproductive activity, according to Pateman, that when she sells it, she ‘sell[s] herself in a very real sense’ (1988: 207; Satz 1992: 114). More recently, some feminists have stated, mostly in response to the growing market in oocytes and embryos, that women’s identity as persons is tied 14

This terminology is from Debra Satz (1992). She explores in her paper whether the opposite thesis, the asymmetry one, is true. She also discusses how feminists are split on the issue of symmetry or asymmetry.

for dignity or money

275

to their reproductive activity or to their reproductive bodily tissues, including oocytes and embryos. For example, Radin suggests that reproductive activity is not a ‘severable fungible object’, but an essential attribute of persons (1996: 127). Suzanne Holland maintains that gametes and embryos have such ‘an intimate connection to [our] personhood’ (2001: 265) that they are inalienable from us and are therefore non-commodifiable (see also C. Cohen 1999). Notice that if reproductive tissues or activity are inalienable to us, not only should they not be sold, they also should not be donated to others. The identity thesis rejects both donation and vending. Unfortunately, feminist claims about identity and women’s reproductive activity or tissues tend to lack sufficient argument. Sometimes feminists take the claims to be sufficiently obvious that they are stated without argument, as they are in Pateman and Radin. At other times, the arguments that feminists provide are deeply flawed. For example, Holland defends her version of the identity thesis by saying that gametes and embryos have a special connection to us as persons because they are parts of our bodies, which themselves are ‘intimately connected to . . . who we are’ (2001: 273). Just as we should not be treated as mere commodities, our gametes and embryos should not either. The flaws in this argument are serious. They include not distinguishing body parts that are somehow essential to persons from those that are not (surely ‘spit and fingernail parings’ do not count; C. Cohen 1999: 291), and failing to defend the view that gametes and embryos in particular are essential to persons. Feminist claims about why our reproductive activities or parts are inalienable to us tend to be not only unsound, but also pronatalist. If a woman’s identity is intimately connected to her reproductive potential, then she can never escape her fertility. On the contrary, she should embrace it and discover part of who she is by actualizing her reproductive potential (McLeod and Baylis 2006). But such a conclusion supports the status quo, according to which women and men (but women in particular) are not free to lead a life in which reproduction does not occur, or is a non-issue. Satz made this sort of objection against Pateman’s work in the early 1990s. Since then, feminists such as Holland and Radin have simply reproduced Pateman’s mistake.

The Autonomy Thesis A further attempt to establish that women’s reproductive labour, particularly in pregnancy, is inherently different from other forms of labour concerns the need to restrict the autonomy of contract pregnant women to ensure that they maintain a certain relationship to their ‘product’ (the child): that is, a non-parental relationship. Women in contract pregnancies will have to manipulate themselves, or be manipulated by others, into having, or continuing to have, a perspective on their labour that fits with their contractual obligations; otherwise, a couple will lose their child. Either contracts for other forms of labour (at least for morally legitimate forms) do not require for their fulfillment the same degree of control over the

276

carolyn mcleod

labourer’s perspective on what he is doing, or what’s at stake with their fulfillment is not important enough to warrant profound manipulation. It follows that the asymmetry thesis is true. Elizabeth Anderson makes this sort of argument in ‘Is Women’s Labor a Commodity?’ (1993). She explains that a woman in a contract pregnancy must agree at the outset not to view her pregnancy relationship as a parental one, which the woman may well do. But, Regardless of her initial state of mind, she is not free, once she enters the contract, to develop an autonomous perspective on her relationship with her child. She is contractually bound to manipulate her emotions to agree with the interests of the adoptive parents. Few things reach deeper into the self than a parent’s evolving relationship with her own child. Laying claim to the course of this relationship in virtue of a cash payment constitutes a severe violation of the mother’s personhood and a denial of her autonomy. (Anderson 1993: 178)

As noted above, according to feminists for money, this line of argument itself denies women’s autonomy by suggesting that their minds waver too much in pregnancy for them to enter into pregnancy contracts, or by implying that ‘sacred bonds’ inevitably develop during pregnancy. But there are responses that feminists for dignity could give to these objections. First, changing one’s mind about a decision at some point during a nine-month period does not indicate a lack of autonomy, since people who are autonomous will in fact reevaluate and often revise their decisions. Thus, a true autonomy thesis would respect a change of heart by a woman in a contract pregnancy. Second, it could do so while acknowledging that not all women bond with their fetuses. (‘Some women abort them,’ as Satz reminds us; 1992: 117.) Feminists could construct the autonomy thesis so that it says only that a woman may bond with her fetus, not because of some gender-wide instinct, but because she has cared for this being for months at a time. And whether a woman will bond with her fetus is not something of which she can be certain ahead of time, since women’s perspectives on their pregnancies do tend to evolve as their pregnancies progress, and the evolution need not occur in the same direction as previous pregnancies (see Mackenzie 1992). Further, when bonding does occur, it can reach so deeply into the self that forcing the woman to relinquish the child would be cruel. This autonomy thesis is weaker than Anderson’s, however (which assumes that the pregnancy relationship will reach deeply into the woman’s self); and it is not clear whether or not this weaker thesis supports the asymmetry thesis. Unless most pregnancy contracts required for their fulfillment that the labourer’s perspective on her labour be seriously manipulated (which would be the case only if sacred bonds were inevitable), such manipulation would not need to be common practice. And therefore perhaps the labour, in this respect, would not have to be treated differently from other (legitimate) forms of labour. But even though manipulation may not need to be common, one might argue that it would have to occur in some cases and it would not be known ahead of time which cases (i.e. if the weaker autonomy thesis were true), which itself could make pregnancy contracts unique and morally

for dignity or money

277

problematic. In other words, perhaps the weaker thesis does support the asymmetry thesis.

The Labour Is Contingently Special Satz maintains that both the autonomy thesis and the identity thesis are totally indefensible. Nevertheless, she contends that an asymmetry exists between women’s reproductive labour and other forms of labour. For Satz the asymmetry is contingent upon the reproductive labour being performed in an environment that is sexist, racist, and classist. Satz’s main criticism of pregnancy contracts is that they strengthen gender inequality. They feed on an environment in which women earn significantly less than men, live in poverty more often than men, are more confined to the home because of an unequal distribution of child care and other domestic work, and, in general, have less opportunity to better their lives (1992: 124). Gender inequality extends as well to reproduction, where men have historically had more control than women over when, and how, women reproduce. Contract pregnancy reproduces this pattern by having women relinquish significant control to others (sometimes specifically to men) over their own bodies in pregnancy (Satz 1992: 124, 125). The practice is troubling for this reason, and because it reinforces a pronatalist connection between women and reproduction, and a further connection between women and the home, since women in contract pregnancies tend to stay at home. Satz claims that while some women may prefer to be at home and to make money by selling their reproductive labour, ‘we need to pay attention to the limited range of economic opportunities available to these women and to the ways in which these opportunities have shaped their preferences’ (1992: 127). For Satz, contract pregnancy involves increased subordination rather than autonomy for women, in addition, because it shores up rather than dismantles the core foundation of parenthood under patriarchy, which is genetic rather than gestational. With a genetic standard for parenthood, men are at least equal parents to women, although, traditionally, their ‘seed’ meant more than a woman’s in defining whose child a particular child was. Women were simply ‘the incubators of men’s seeds’ (Satz 1992: 128). On the genetic model, gestation does little to shape whom a child will become, although it may do a lot to determine how a woman will feel toward that child. Women are mostly just the ‘maternal environment’ during pregnancy; and if their genes were no part of the fertilization process, then they are no parent at all. In settling disputes about pregnancy contracts, the courts in the United States have supported this picture; they will grant parental rights to a woman who bore a child as a result of a genetic (or traditional) contract pregnancy, but not as a result of a gestational contract pregnancy (Satz 1992: 127).

278

carolyn mcleod

Satz concludes that the loss of dignity for women is profound, but not inevitable with contract pregnancy. Many of us recall images of women as incubators or breeders when we contemplate this practice, and for good reason, according to Satz. Contract pregnancy reinforces sexist stereotypes that prescribe undignified social roles for women. But if these stereotypes did not exist—that is, if our society were nonsexist—the practice would be more benign, morally speaking, in Satz’s view. It may not be completely benign, only because it could still flourish as a result of inequalities that are class- or race-based, rather than gender-based (Satz 1992: 128, 129).15

Mov ing Forward ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

In broad summary, most feminists who are for dignity alone—that is, against commodification—are at the opposite pole of feminists who are for money, or in favour of commodification. Whereas the latter say, ‘Pay the women, not only because it is the best option given the circumstances (of poverty for women etc.), but because it is the most dignified option for women’, the former assert that it is simply undignified for women to sell their reproductive labour: they sacrifice too much of their identity or their autonomy in doing so. With her work on contract pregnancy, Satz is somewhere in between these two poles, for she claims that selling reproductive labour could be dignified, or at least not undignified, in an egalitarian society. Satz’s work is important because it cautions against the reproductive essentialism that is embedded in some feminist claims about women’s reproductive labour being inherently special. In discussing how contract pregnancy promotes gender inequality, Satz also exposes the naivety of some feminists for money in thinking that this practice could actually liberate women. Still, it is not clear that Satz responds adequately to the double bind of dignity or money that I outlined earlier. She says that we should discourage pregnancy contracts by making them legally unenforceable, for example (1992: 129), which is the logical conclusion to her argument. But, on its own, this solution does little to address the ‘limited range of economic opportunities available to [women for whom the contracts are attractive]’ (Satz 1992: 127), a range that Satz herself identifies. Are we justified in restricting this range, so that it excludes reproductive labour, while doing nothing to expand it? If not, could our plans be for long-term expansion alone, or would we have to introduce some short-term relief in exchange for eliminating the opportunity of selling one’s reproductive potential? Further, how serious is the risk of a black market arising in which female reproductive 15

A refusal to grant parental rights to a woman who underwent a gestational contract pregnancy because she lacked a genetic tie to the child (children in fact, since there were twins) has occurred in Canada as well; see J.R.v. L.H., [2002] OJ 3998 (QL).

for dignity or money

279

labourers have even less dignity than they do now? Feminists need to address these practical matters, or, in other words, confront head on the main oppressive aspect of commodification: that it creates a double bind for women. Moving forward with the feminist debate on commodification will also require that feminists learn from what other feminists, or non-feminists, have written on topics similar to their own. Even amongst feminists writing on women’s reproductive labour, there has been little cross-pollination of ideas. The newer debate on oocyte vending does not respond well to progress in the older debate on contract pregnancy. For example, just as reproduction should not be seen as inherent to women’s identities, perhaps neither should oocytes or embryos. But if that is true, the question remains whether oocyte vending is morally problematic; and if it is problematic, is the problem contingent on the practice occurring in certain sociopolitical environments, or is it inherent to the practice itself? Feminists writing on oocyte vending or commercial contract pregnancy could also draw valuable insights from feminists who have theorized not about these practices, but about others that raise similar moral issues for feminists. For example, feminists have done detailed work on what it means to say that women are autonomous in choosing cosmetic surgery (Morgan 1991) or medical interventions in pregnancy (Sherwin 1998; McLeod 2002). They have discussed how myths of beauty or of pronatalism can shape women’s preferences for these interventions such that the preferences are not fully autonomous, but may be rational nonetheless given how much easier it is to get by in a sexist world if one conforms to its expectations. Such subtle distinctions—for example, between autonomy and rationality (or a form of rationality)—are important for any nuanced discussion of the autonomy of women who engage in gendered pursuits such as selling their reproductive labour. Thus, while feminist debate on commodifying women’s reproductive labour is surely intricate, there is room for even more depth and sophistication. In particular, the debate could progress more toward concrete solutions that genuinely promote women’s autonomy if the situation of potential reproductive labourers was analogized with that of potential consumers of cosmetic surgery, for example, or potential sex trade workers. The promise of such a turn in the literature is a greater appreciation of how choosing between dignity and money with respect to women’s reproductive labour is a classic double bind.

References Alpers, A., and Lo, B. (1995), ‘Commodification and Commercialization in Human Embryo Research’, Stanford Law and Policy Review, 6/2: 39–46. Anderson, E. S. (1993), ‘Is Women’s Labor a Commodity ?’, in Anderson, Value in Ethics and Economics (Cambridge, Mass.: Harvard University Press), 168–89; first pub. in Philosophy and Public Affairs, 19/1 (1990), 71–92.

280

carolyn mcleod

Arneson, R. J. (1992), ‘Commodification and Commercial Surrogacy’, Philosophy and Public Affairs, 21/2: 132–64. ASRM (American Society of Reproductive Medicine), Ethics Committee (2000), ‘Financial Incentives in Recruitment of Oocyte Donors’, Fertility and Sterility, 74/2: 216–20. Atwood, M. (1985), The Handmaid’s Tale ( Toronto: McClelland & Stewart). Bartky, S. L. (1990), ‘Narcissism, Femininity, and Alienation’, in Bartky, Femininity and Domination: Studies in the Phenomenology of Oppression (New York: Routledge), 33–44. Chesler, P. (1988), Sacred Bond: The Legacy of Baby M. (New York: Times Books). Cohen, C. (ed.) (1996), New Ways of Making Babies: The Case of Egg Donation (Bloomington: Indiana University Press). (1999), ‘Selling Bits and Pieces of Humans to Make Babies: The Gift of the Magi Revisited’, Journal of Medicine and Philosophy, 24/3: 288–306. Cohen, J. (2002), ‘Grade A: The Market for a Yale Woman’s Eggs’, Atlantic Monthly, 290/5: 74–6. Dickenson, D. (2001), ‘Property and Women’s Alienation from Their Own Reproductive Labour’, Bioethics, 15/3: 205–17. Fielding, D., Handley, S., Duqueno, L., Weaver, S., and Lui, S. (1998), ‘Motivation, Attitudes and Experience of Donation: A Follow-Up of Women Donating Eggs in Assisted Conception Treatment’, Journal of Community and Applied Social Psychology, 8: 273–87. Frye, M. (1983), The Politics of Reality: Essays in Feminist Theory (Freedom, Calif.: Crossing Press). Gurmankin, A. D. (2001), ‘Risk Information Provided to Prospective Oocyte Donors in a Preliminary Phone Call’, American Journal of Bioethics, 1/4: 3–13. Holland, S. (2001), ‘Contested Commodities at Both Ends of Life: Buying and Selling Gametes, Embryos, and Body Tissues’, Kennedy Institute of Ethics Journal, 11/3: 263–84. Kittay, E. F. (1999), Love’s Labor: Essays on Women, Equality, and Dependency (New York: Routledge). Mackenzie, C. (1992), ‘Abortion and Embodiment’, Australian Journal of Philosophy, 70/2: 136–55. and Stoljar, N. (eds.) (2000), Relational Autonomy: Feminist Perspectives on Autonomy, Agency, and the Social Self (New York: Oxford University Press). McLeod, C. (2002), Self-Trust and Reproductive Autonomy (Cambridge, Mass.: MIT Press). and Baylis, F. (2006), ‘Feminists on the Inalienability of Human Embryos’, in R. Kukla (ed.), Hypatia, suppl. vol. 21: Maternal Bodies, 1–14. McShane, P. M. (1996), ‘Oocyte Donation Service at IVF America–Boston, Waltham, Massachusetts’, in Cohen (1996: 29–34). Mahoney, J. D. (2000), ‘The Market for Human Tissue’, Virginia Law Review, 86/2: 163–223. Marx, K. (1844/1964), The Economic and Philosophic Manuscripts of 1844, ed. D. J. Struik, trans. M. Milligan (New York: International Publishers). (1867/1954), Capital, ed. F. Engels, trans. S. Moore and E. Aveling (Moscow: Progress Press). Meyers, D. T. (1989), Self, Society, and Personal Choice (New York: Columbia University Press). Morgan, K. P. (1991), ‘Women and the Knife: Cosmetic Surgery and the Colonization of Women’s Bodies’, Hypatia (Fall), 25–53.

for dignity or money

281

New York State Task Force on Life and the Law (1998), ‘Assisted Reproductive Technologies: Analysis and Recommendations for Public Policy’, Executive Summary,

. (2002), Thinking of Becoming an Egg Donor ? A Guidebook, . Nussbaum, M. (1995), ‘Objectification’, Philosophy and Public Affairs, 24/4: 249–91. O’Donnell, K. (2000), ‘Legal Conceptions: Regulating Gametes and Gamete Donation’, Health Care Analysis, 8: 137–54. Pateman, C. (1988), The Sexual Contract (Stanford, Calif.: Stanford University Press). Radin, M. (1991), ‘Reflections on Objectification’, Southern California Law Review, 65: 341–54. (1996), Contested Commodities: The Trouble with Trade in Sex, Children, Body Parts, and Other Things (Cambridge, Mass.: Harvard University Press). (2001), ‘Response: Persistent Perplexities’, Kennedy Institute of Ethics Journal, 11/3: 305–15. Ragone, H. (1994), Surrogate Motherhood: Conception in the Heart (Boulder, Col.: Westview Press). (1999), ‘The Gift of Life: Surrogate Motherhood, Gamete Donation, and Constructions of Altruism’, in L. L. Layne (ed.), Transformative Motherhood: On Giving and Getting in a Consumer Culture (New York: New York University Press), 65–88. (2000), ‘Of Likeness and Difference: How Race Is Being Transfigured by Gestational Surrogacy’, in H. Ragone and F. Winddance Twine (eds.), Ideologies and Technologies of Motherhood: Race, Class, Sexuality, Nationalism (New York: Routledge), 56–75. Roberts, D. (1997), Killing the Black Body: Race, Reproduction, and the Meaning of Liberty (New York: Pantheon Books). Satz, D. (1992), ‘Markets in Women’s Reproductive Labor’, Philosophy and Public Affairs, 21/2: 107–31. Serafini, P. D., Nelson, J. R., Smith, S. B., Richardson, A., and Batzofin, J. (1996), ‘Oocyte Donation Program at Huntington Reproductive Center: Quality Control Issues, Pasadena, California’, in Cohen (1996: 35–48). Shalev, C. (1989), Birth Power: The Case for Surrogacy (New Haven: Yale University Press). Sherwin, S. (1998), ‘A Relational Approach to Autonomy in Health Care’, in Feminist Health Care Ethics Research Network, The Politics of Women’s Health: Exploring Agency and Autonomy (Philadelphia: Temple University Press), 19–47. Shultz, M. M. (1990), ‘Reproductive Technology and Intent-Based Parenthood: An Opportunity for Gender Neutrality’, Wisconsin Law Review, 297: 287–398. Steinbock, B. (2004), ‘Payment to Egg Donors’, Mount Sinai Journal of Medicine, 71/4: 255–65. Vance, C. S. (ed.) (1984), Pleasure and Danger: Exploring Female Sexuality (Boston: Routledge & Kegan Paul). Walzer, M. (1983), Spheres of Justice: A Defense of Pluralism and Equality (New York: Basic Books).

part iv •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

THE END OF LIFE •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

chapter 12 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

THE DEFINITION OF DEATH •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

stuart j. youngner

Introduction ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Until the invention of the stethoscope and the acquisition of knowledge about human anatomy in the early nineteenth century, physicians were unable to diagnose death with precision. The ability to do so provided them with great credibility from a public that had, until then, been concerned about premature burial. Significant uncertainty and debate about the definition and determination of death did not resurface until the latter half of the twentieth century, again owing to the state of medical science. But this time, physicians knew too much, rather than too little, about the pathophysiology of the dying process. In the modern intensive care unit (ICU), they have been increasingly able to break down the dying process, teasing apart each of its component parts and supporting some functions while providing technological replacement for others. In the intensive care unit, death approaches as much on the electronic screens of heart, brain, and blood pressure monitors as it does in the failing bodies of patients. These developments have had two important consequences. First, the cascade of events that previously led to death is no longer inevitable. Before, if one vital function ceased, the others quickly followed, removing the necessity of choosing one as more ‘vital’ than the other. Today, the mechanical ventilator, the cardiac pacemaker, drugs that maintain blood pressure, and many other interventions keep

286

stuart j. youngner

the body ‘going’ after loss of innate functions that formerly would have meant the cessation of them all. The second consequence is the incredible control such monitoring and support gives physicians, patients, and families over the timing of death. Once thought to be in the hands of God or fate, the time of death is now most often a matter of deliberate human decision. This understanding and control, while never able to defeat death, gives human beings the illusion of control over death and an uneasy sense of responsibility, however it arrives. Two factors, medical science’s growing control over the timing of death and the increasingly desperate need for organs, have led to a reopening of the debate about the definition of death and have forced consideration of aspects of the determination of death that had never been addressed before. Without the pressing need for organs, the definition of death would have remained on the back shelf, the conversation of a few interested philosophers or theologians. In the discussion that follows, we will examine some new questions raised by medical technology and the frantic search for new, morally acceptable sources of human organs over the past thirty years. This examination will lead us to conclude that death itself is a social construct and that, in a pluralistic society such as ours, a conclusive definition of death or determination of the moment of death is out of the reach of both medical science and philosophy.

Philosophical Definitions of Death ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

Philosophers have argued convincingly that the definition of death, while grounded in human biology, is, ultimately, a philosophical question. To define death, one must answer the question ‘What function(s) is (are) so essential that its (their) irreversible loss signifies (signify) the death of the human being?’ Once a correct definition of death is chosen, an operational criterion can be selected for determining that the definition has been fulfilled. At an even more operational level, specific clinical tests can be done to assure that the criterion has been fulfilled. Specific criteria and tests are valid only because they fulfill a philosophically defensible definition. So, for example, if one argued that consciousness and cognition were the unique functions that differentiated a living person from a corpse, the corresponding criterion would be loss of functioning of the higher brain. Specific tests to demonstrate that this criterion had been fulfilled might include non-responsiveness of the patient, a flat EEG, and a brain scan showing destruction of those parts of the brain responsible for consciousness and cognition. If this definition, loss of consciousness and cognition, is wrong, then the corresponding criteria and tests are ‘wrong’ as well. Conversely, no criterion can stand alone. It must be supported by a philosophically sound concept answering why it measures

the definition of death

287

the difference between a living human being and a corpse. For example, a claim that irreversible loss of all liver function was the correct criterion for measuring death would be absurd. Persons with no liver function may be dying, but they breathe, their hearts pump blood, and they can still talk, read, and suffer. In addition to loss of consciousness and cognition, there are two other philosophical notions of what it means to be dead. The first is loss of vital fluid flow —the circulation of blood and oxygen through the body. The corresponding criterion is irreversible cessation of cardiopulmonary function that might be tested by listening for breath sounds, feeling for a pulse, or performing an electrocardiogram. The vital fluid flow definition dominated from the early 1800s until the early 1970s. The other definition is loss of functioning of the organism as a whole. The corresponding criterion is irreversible loss of all brain function, including that of the brain stem. Tests to see that that criterion has been fulfilled might include examination of brain stem reflexes, EEG, or scans to see if all blood flow to the brain has ceased. Each of the three definitions has its adherents in the United States and the rest of the world, although the whole brain formulations have been enshrined in law in the United States and many European countries through recognition of so-called ‘brain death’. Before going on to discuss brain death in greater detail, an observation is in order.

The Implications of Stretching Out the Physiologic Events Surrounding Death Until the advent of the mechanical ventilator and the modern intensive care unit, when one vital function ceased, the others stopped quickly and, to the unsophisticated eye, simultaneously as well. It all happened at once. For example, if someone had a sudden heart attack and cardiac arrest, the person stopped breathing and lost consciousness (owing to lack of oxygenated blood to the brain). Similarly, if breathing ceased, loss of consciousness quickly followed, and the heart, also deprived of oxygen, stopped beating within minutes. If someone sustained a massive head injury (stroke, gunshot wound), he or she immediately stopped breathing because spontaneous respiration is controlled by a center in the brain stem. The heart, deprived of oxygen, stopped within minutes. Thus, we never had to choose between vital fluid flow, function of the organism as a whole, or consciousness and cognition as the key function that distinguished life from death. With medical intervention, these events no longer take place simultaneously. If a person has a heart attack, we can resuscitate them, put them on a mechanical ventilator, and shock or pump on the chest to restore circulation to the brain. Respiratory arrest can be treated with mechanical intervention. Most relevant here, following an acute and catastrophic brain injury, we can now mechanically ventilate a patient, preserving heart but not brain function.

288

stuart j. youngner

This latter category, patients who have lost all brain function, including that of the brain stem, but are maintained with life support, is called brain death. Braindead patients’ hearts continue to beat spontaneously, they breathe with the aid of a ventilator, and, in many other ways appear to be alive. Their kidneys produce urine, their pancreases produce insulin, their livers metabolize waste products in the blood, etc. In the early stages of pregnancy, such ‘dead’ patients can gestate fetuses for months until they are capable of living ex utero. Although initially considered very unstable, better technology, persistence, and demands of family members have kept some brain-dead patients ‘alive’ for up to ten years. As we shall see, the ‘discovery’ of so-called brain death was a direct consequence of the invention and employment of the mechanical ventilator in patients who had sustained catastrophic head injuries. Maintaining such patients provided both a problem and an opportunity. The problem was that when ventilators were first deployed in the late 1960s, we had no legal, professional, or cultural experience with turning them off. To many, it seemed that to do so would mean killing the patient. Coincidentally, the severely brain-damaged patients, now maintained on ventilators, were a potential source of organs for transplantation.

Brain Death in the United States ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

In 1968 an ad hoc committee at the Harvard Medical School proposed ‘a new criterion for death’: total and irreversible loss of functioning of the whole brain (Ad Hoc Committee 1968). This proposal came as a result of a new class of patients inhabiting intensive care units—patients with massive brain injuries who were being sustained by mechanical ventilators and aggressive ICU staff. Interestingly, the Harvard Committee proposed only a criterion of death with no corresponding definition. In other words, they did not ask or answer the question ‘What is the critical characteristic of brain-dead patients that makes them dead?’ Instead, they made it explicitly clear that they were trying to solve two practical problems: (1) to relieve the ‘burden’ imposed by severely brain-damaged patients; and (2) to quell the ‘the controversy in obtaining organs for transplantation’. Let us examine these practical problems in order.

Relieving the Burden Mechanical ventilators had just come into use in the 1960s. In many instances, they were able to stabilize patients not otherwise able to breathe, allowing critical medical treatment, removal from the ventilator, and discharge home from the hospital. In other cases, however, patients could not be ‘weaned’ from the ventilator. Examples included people at the end-stages of terminal illnesses such as metastatic cancer or

the definition of death

289

heart failure, or, more to the point here, patients with massive head injuries whose lungs and hearts were working fine. It may be hard to imagine now, fifty years later, when ‘terminal’ discontinuation of mechanical ventilation is an everyday event in modern ICUs, but in the 1960s our society had no legal, clinical, or cultural experience with turning off ventilators. There were no court cases, no laws, no professional guidelines, and no hospital policies. The Harvard Committee clearly believed that reclassifying some of these patients as ‘dead’ would avoid charges of homicide when they were allowed to die by ‘pulling the plug’.

Quelling the Controversy By the late 1960s organ transplantation was becoming a reality. One of the obstacles was (and remains) a shortage of potential donors. Patients who died of cardiorespiratory arrest were not adequate organ sources. After the heart and lungs stop, oxygenated blood no longer flows to the body’s organs and tissues. However, the cells in those tissues and organs continue to metabolize in the cooling, but still warm, body, but without the benefits of oxygen, nutrients, and a way to get rid of waste products. Thus, warm ischemic (without blood) damage occurs and, as a result, by the time they are removed, organs are no longer fit for transplantation. Until the invention of brain death, organs came primarily from live donors, and, at first (because of immunologic incompatibility), only from identical twins. As noted earlier, patients with massive brain destruction, maintain their heartbeats on ventilators, and have healthy and functioning hearts, kidneys, livers, lungs, and pancreases. It was precisely these characteristics that made them an attractive new source of organs. But, if they were not dead, removing vital organs would surely kill them, violating an unwritten but powerful rule that governs organ transplant to this day—the dead donor rule. The dead donor rule states that patients may not be killed by or for removal of their organs for transplantation. Violating this rule would have provoked the controversy the Harvard Committee proposed to ‘quell’. Amazingly, their radical new proposal was embraced by the legal and medical communities in the United States, all major religious groups, and, seemingly, the public—with very little controversy or even discussion. Within two decades, brain death became a legal standard of death throughout the United States and allowed a tremendous expansion in organ transplantation. Why this was so is an interesting question. In Japan, as we shall discuss in detail later, the opposite is true. Japan has only recently adopted a highly controversial and, largely, ineffective law linking brain death and transplantation. It is likely that a number of factors combined to ease acceptance in the United States. First is the practical or utilitarian bent of American society, with its great enthusiasm for the ‘advances of medical science’. Second, there was no natural political constituency to oppose it. As Courtney Campbell (1999: 199) has written:

290

stuart j. youngner

It can be questioned why such religious opposition did not emerge . . . The time frame is very important. One cannot speak of a politically mobilized and socially active fundamentalist movement until after the Roe v. Wade decision legalizing abortion in 1973, some 5 years after the report of the Harvard committee. Nor was euthanasia a realistic end-of-life option for patients until very recently.

Campbell notes that the current cultural climate is quite different. Currently, religious conservatives criticize today’s social ethos as an embrace of the ‘culture of death’ and apply a much more vigorous and carefully developed analysis to the issues of abortion and euthanasia than they ever did to brain death. Lastly, although brain death is, as we shall see, tremendously flawed from a philosophical and clinical perspective, it has practical characteristics that make it acceptable to different people for different reasons. First, unlike its cousin condition, persistent vegetative state (PVS), with which it is often confused, brain death is a relatively quick and easy diagnosis to make. Patients can reliably be pronounced brain-dead within hours of initial diagnosis. Moreover, once the diagnosis is made, the prognosis is entirely clear. Brain-dead patients will never wake up and they will never breathe on their own. In fact, they are notoriously unstable and, even maintained on ventilators, will often suffer cardiac arrest and cardiovascular collapse within hours or days of diagnosis. So, the diagnosis of irreversibility is certain and, even if one believed in a cardiopulmonary criterion of death, cardiac arrest would predictably come very soon on the heels of a brain death diagnosis. For the vast majority of people, the quality of life for brain-dead patients, ICU dependent and forever unconscious, is unacceptable. Finally, although the Harvard Committee attempted no definition of death with which to justify its criterion, the brain death criterion can actually satisfy two competing definitions, making it acceptable to a wider number of persons. The first definition to justify the whole brain criterion was proposed by Bernat and his colleagues in 1981—thirteen years after the whole brain criterion was first proposed by the Harvard Ad Hoc Committee, surely the case of a criterion in search of a definition (Bernat et al. 1981). Bernat said that the quality absent in brain-dead patients that made them dead was ‘permanent cessation of functioning of the organism as a whole’, that is, ‘the spontaneous and innate activities carried out by all or most subsystems’ and ‘the body’s capacity to organize and regulate itself’. Examples given included neuroendocrine control, temperature regulation, and the ability to maintain blood pressure and fluid and electrolyte balance—all functions of the brain stem. Thus, Bernat saw the critical brain function as the ability to integrate the body’s non-cognitive vegetative functions, thereby maintaining homeostasis. Although justifying a ‘whole brain’ criterion, Bernat gave no significance to another important brain function, consciousness and cognition. Bernat’s definition was quickly endorsed by a President’s Commission whose report facilitated the widespread acceptance of brain death (President’s Commission for

the definition of death

291

the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research 1981). In a simplified way, the brain has two major functions: the integrative functions of the brain stem so important to Bernat and the President’s Commission, and the capacity for consciousness and cognition. The integrative capacity resides primarily in the brain stem, the smaller, more primitive part of the brain buried beneath the much larger cerebral hemispheres where cognitive capacity largely resides. These are sometimes referred to as the lower brain (brain stem) and higher brain (cerebral hemispheres). Many persons think that permanent loss of consciousness and cognition is the critical one that makes brain-dead patients dead. However, when patients become brain-dead, they lose both consciousness and cognition and integrative capacity. Thus, a brain-dead patient satisfies both camps and does not force a choice between higher and lower brain functions as necessary and sufficient for death. There has been little intellectual support in the philosophical and bioethics literature for Bernat’s position. In fact, in the twenty-five years since Bernat’s whole brain formulations were introduced, philosophic thought and clinical experience have largely refuted them. However, because of its wide appeal for a variety of practical reasons, rejection of brain death has been confined primarily to academic circles; in the United States there have been no significant challenges to brain death as public policy.

Philosophical and Clinical Challenges to the Whole Brain Formulations Philosophers have savaged the formulations of Bernat and his colleagues on several counts. First, while they agree that the ability of the brain stem to integrate the body’s subsystems is necessary to maintain life and that without it the person would die very quickly, dying is not the same as death. Moreover, integrative function hardly defines what it means to be a living human being. Consciousness and cognition alone, they argue, are the necessary and sufficient conditions (Bartlett and Youngner 1988; Gervais 1986). Consider the following thought experiment. A patient has severe damage to the brain stem but not cerebral hemispheres and, therefore, has lost only the innate integrative brain functions so dear to Bernat. However, these brain stem functions are being carried out by the mechanical ventilator and a highly trained ICU staff. For example, the nurses monitor blood pressure; when it drops, the patient is given drugs to raise it. However, unlike a ‘brain-dead’ patient who has lost all brain function, this patient remains conscious and can respond to our questions. She has lost all the spontaneous and innate functions Bernat identifies as essential, yet she is aware and can communicate. So, while the patient does not meet Bernat’s

292

stuart j. youngner

criterion of death, since she still retains some brain function, she does meet his definition—the loss of integrating capacity. His formulations are inconsistent and illogical. He should either include consciousness and cognition as essential functions for his definition or drop the cerebral hemispheres from his criterion. Another problem with Bernat’s definition is that he insists the essential functions must be ‘innate and spontaneous’. But, why? During bypass surgery, the functions of circulation of blood and oxygen (vital fluid flow) are not spontaneous and innate, but rather done quite effectively by a so-called heart–lung machine. In the case of the brain-dead patient, the function of integration is not innate and spontaneous either. It is taken over by the machines and health professionals in the ICU. But, the critical function continues. Why does the critical function have to be innate and spontaneous in the brain-dead patient, but not so for the patient undergoing cardiac bypass? Clinical experience has identified a third problem with the brain death formulations. Some functions, identified as critical by Bernat, remain in the brains of many patients declared brain-dead. Physicians simply do not test for them. For example, routine tests for brain death do not include tests for an important neuroendocrine function—production of antidiuretic hormone. Presence of this hormone is integrative in the true sense that Bernat suggested. In fact, he specifically mentioned neuroendocrine function in his 1981 paper. There seems an easy way to reconcile this inconsistency. In addition to the current tests for brain death (that included testing for the reaction of pupils to light and a gag reflex) we could also test for adequate vasopressin levels in the blood. If they remained, the patient could not be declared dead. But this policy would add to the cost of care, and, more importantly, would exclude some patients who are currently organ donors. Bernat takes another way out. In response to criticism, he modified his definition to include only ‘critical’ functions and insisted that production of antidiuretic hormone is not one of them because ‘patients without such secretion can survive for long periods without treatment’ (Bernat 1998). But what about other functions that we currently do test for such as the gag reflex and the reaction of pupils to light? Surely, patients can survive even longer without these functions. Bernat’s defense gives no logical way to order brain functions as more or less critical. Moreover, all state laws require loss of all brain functions, and the word ‘all’ has without doubt been instrumental in public acceptance of brain death. Accepting Bernat’s position would require either changing all the state laws or maintaining the fiction that they were being followed.

Problems with Alternative Definitions The loss of consciousness and cognition has its adherents, but also its problems. Mayo and Wikler argue that the definition of death should include only those characteristics that are common to all living things (Mayo and Wikler 1979).

the definition of death

293

Another problem with the higher brain definition is that while it could explain why patients who had lost all brain function were dead, its application would logically require declaring PVS patients dead. If brain-dead patients retain many signs usually associated with life, PVS patients retain even more. PVS patients, as mentioned earlier, have sustained severe damage to their cerebral hemispheres and are permanently unconscious but have fully functioning brain stems. Thus, they breathe without assistance. They even have periods each day when they open their eyes. The only technology they require is a feeding tube. Thus, to many observers, calling them dead is counterintuitive, regardless of how logical the argument that they are dead.

Public Confusion and Disagreement About the Definition and Criterion of Death Bernat and his colleagues made an empirical claim (without any empirical evidence) that ‘what the layman actually means by death’ is loss of integrating function. However, empirical studies do not support their claim. When asked if a brain-dead patient was dead, 95 per cent of 195 physicians and nurses likely to be involved in organ procurement for transplantation said yes (Youngner et al. 1989). However, when asked why the patient was dead, 36 per cent gave loss of consciousness and cognition as the explanation while only 26 per cent chose loss of integrating capacity. Interestingly, 32 per cent gave answers such as ‘the quality of the patient’s life is unacceptable’ or ‘they will die soon no matter what you do’—indications that they believed the patient was as good as dead, rather than dead. Moreover, fully 38 per cent of respondents believed that a patient in PVS was dead. A more recent study of the general public had similar results. A survey of 1,351 randomly selected residents of the state of Ohio found that only 40.4 per cent personally believed that a brain-dead patient is dead, while 43.3 per cent thought they were ‘good as dead’ and 16.3 per cent thought they were alive (Siminoff et al. 2004). Similarly, to the health professionals in the earlier study, about 34 per cent of Ohio residents personally believed that a patient in PVS is dead. The findings of the two studies hardly support Bernat’s contention that loss of integration is ‘what is commonly meant by death’. To the contrary, the studies demonstrate the persistence of a variety of personal beliefs about the definition of death, including the traditional loss of vital fluid flow. In addition, both studies found a considerable level of ignorance about factual matters related to death. Among health professionals, for example, only 35 per cent were able to identify correctly the legal and medical criteria for determining death. Among the general public, only one third knew that brain death was legal death in Ohio, 28 per cent mistakenly thought that brain-dead patients can hear, and nearly 60 per cent

294

stuart j. youngner

incorrectly thought that organs were retrieved from brain-dead patients after, not before, the ventilator is turned off. This data suggests that, while brain death currently seems to be working as public policy, the public hardly has a monolithic view about what it means to be dead. Moreover, considerable confusion remains about the legal and clinical status of brain-dead patients.

Irreversibility: New Issues and New Debates As discussed earlier, brain-dead patients provided transplanters with a valuable new source of organs. However, the increased supply of organs and the growing success of organ transplantation spawned an even greater demand. Today, there are over 80,000 people waiting for organs. As a result, the transplant community revisited the idea of taking organs from patients who died of cardio-respiratory arrest by devising a strategy for removing organs before they had suffered warm ischemic damage. The strategy, formerly called non-heart-beating donation, is now called donation after cardiac death (DCD) and works as follows (Arnold et al. 1995). Often, patients with devastating head injuries still retain enough brain function to preclude a diagnosis of brain death. Yet, the prognosis for recovery is nil and families want to discontinue ventilatory support. Sometimes they are disappointed that they cannot donate organs. DCD works as follows. Instead of discontinuing ventilatory support in the ICU, patients are taken to the operating room, where they are draped, their skin is disinfected, and they are otherwise prepared for surgery. The patients, already unconscious from their injuries, are carefully monitored with electrocardiogram machines and blood pressure censors. When all is ready, the ventilator is turned off. Deprived of oxygen, the heart soon stops. Depending on the protocol, death is declared between two and ten minutes later by cardiopulmonary criteria. At this point the surgeons quickly open the patient and remove the desired organs—most often kidneys but sometimes liver. Once removed, the organs are quickly cooled on ice. The DCD protocols produce viable organs because they are structured to take the organs out as quickly as possible after the declaration of death in order to keep warm ischemic time to a minimum. However, there are no clinical guidelines upon which to make such a declaration. No law, textbook, or medical school lecture has indicated how long after cessation of breathing and heartbeat death occurs. It never used to matter because either death was declared after lengthy resuscitation efforts had failed, or at the end of a long dying process in patients for whom comfort measures only were being provided. In the latter cases, when there were no resuscitative efforts after cardiopulmonary arrest, there was a significant time interval when the patient was visited by the family and cleaned and prepared for transfer to the hospital morgue by the nursing staff. During this interval, if the

the definition of death

295

heart began beating again for a few seconds, no one knew and there were no bad consequences. With DCD, however, patients are closely monitored. It would not do for surgeons and nurses to see cardiac activity after they had begun to remove organs. There is some evidence, although incomplete, that the heart will not autoresuscitate (start beating on its own) after two minutes of stopping. More troublesome is the fact that many patients qualifying for organ removal under DCD protocols could have their hearts restarted if given external electric shocks. Of course, without ventilation, the hearts would soon stop again, but the fact remains that, however briefly it beats, a beating heart is not consistent with a pronouncement of death for patients who have not been pronounced brain-dead. This situation has forced physicians and bioethicists to revisit the meaning of ‘irreversible’. Although irreversibility is part of every clinical and legal determination of death, it has never been defined. In order to examine this issue comprehensively, let us consider the following case. Mr Smith is walking down the street and has a massive heart attack resulting in cardiopulmonary arrest and loss of consciousness at time T1 . Four minutes later, at T2 , he is discovered lying on the sidewalk, is not breathing, and has no pulse. Cardiopulmonary resuscitation is begun and a 911 call is placed. The EMS squad arrives and continues the resuscitation efforts as they transport the patient to the emergency room, where it is continued for a total of forty-five minutes following T2 when the patient was found on the street. After forty-five minutes of resuscitation, there is no resumption of spontaneous heartbeat so the emergency room physicians stop and declare the patient dead at T4 . Since there is no good data on how long resuscitative efforts must continue before declaring them a failure, it is likely that the situation became hopeless sometime earlier, between T2 and T4 . Let us call this time T3 . When did Mr Smith die? At T1 when he lost cardiopulmonary function? Four minutes later, at T2 , when it is determined that he has lost cardiopulmonary function? At T4 , when physicians decided to give up resuscitative efforts, or did he die at T3 , when his arrest really became irreversible? Again, the answer cannot be found in either the law or medical texts (Lynn and Cranford 1999). And, aside from the bizarre kinds of cases used to teach law students, nothing much has ever hinged on this decision. However, in DCD cases a great deal is at stake because there is a premium on getting organs out as quickly as possible without killing patients by violating the dead donor rule. Because DCD patients are being closely monitored, T1 and T2 are the same. The critical question is about T3 , the moment when irreversibility is determined. Supporters of the DCD protocols argue that even though the cessation of cardiac beat could be reversed with electric shocks at two or five minutes, the fact that a morally acceptable decision not to do it has been made means that the cessation

296

stuart j. youngner

is irreversible. Is this position defensible? Most philosophers who have tried to tackle this question say no. In fact, David Cole has argued that irreversibility is not part of the ordinary concept of death and should be dropped from the definition and criteria (Cole 1993). ‘Irreversibility’, argues Cole, is hopelessly ambiguous. It belongs to a class of modal terms that have always resisted analysis in logic and philosophy of language. More specifically, ‘irreversible’ invites such questions as: By whom? When? And under what circumstances? While it is unlikely that law and medical guidelines will abandon their insistence on irreversibility, Cole outlines three different construals of the term. The first and strongest construal of ‘irreversible’ is that ‘a lost function cannot be restored by anyone under any circumstances at any time now or in the future’. For example, if a person’s entire body is frozen upon death with the hope that sometime in the future they could be thawed and restored to life, the loss of function would not be irreversibly lost. Others reject this strongest construal of ‘irreversible’, arguing that ‘Magical, or futuristic scenarios are the sort of concepts to be contrasted with rather than being illustrative of ordinary ones’ (Bartlett 1995: 272). The philosopher David Lamb agrees with Bartlett. ‘In the real world,’ he says, ‘logical possibility without the check of plausibility is a worthless guide to action’ (Lamb 1992: 32). The second construal of ‘irreversibility’ is that loss of function cannot be reversed by those present. This construal would preclude taking organs in DCD protocols after two to five minutes because electric shock could restore heartbeat. The third construal, the weakest, is that a function is irreversibly lost if a morally defensible decision has been made not to try to reverse the loss. Tom Tomlinson has defended this construal, writing that patients in DCD protocols are dead after two minutes because ‘irreversible . . . is best understood not as an ontological or epistemic term, but as an ethical one’ (Tomlinson 1993: 157). This position has been effectively attacked by others. Bartlett argues that Tomlinson’s analysis justifies active euthanasia to obtain organs but does not support a determination of death. Youngner, Arnold, and DeVita argue that basing a declaration of death on a moral analysis rather than on ontological arguments or biological phenomena leads to several problems (Youngner et al. 1999). First, while removing a ventilator with family consent has been increasingly accepted as ethical in the United States, other countries do not turn off ventilators as a matter of social policy. Using Tomlinson’s standard, a patient who is considered dead in the United States would be alive in Spain or Italy. Second, moral judgments about the acceptability of stopping life-sustaining treatment often depend on secondary issues such as adequacy of informed consent that are themselves subjective and prone to controversy. Disagreement about non-medical details of a particular case could force discussion further and further from commonsense notions about

the definition of death

297

what it means to be dead. Consider the following case in the context of brain death. A 70-year-old patient with end-stage metastatic cancer is rushed to the hospital, comatose. She meets all other clinical criteria for brain death but is found to have high levels of barbiturates (a central nervous system depressant) in her blood. A note is found that says the patient took an intentional overdose and wants to be allowed to die. According to current medical standards, such a patient could not be declared dead. If she were supported while barbiturates cleared from her system, she could wake up with no damage whatsoever to her brain. Yet, according to Tomlinson, such a patient could be declared dead while the drug is still in her system if we thought her suicide attempt was morally justified. But our society is hopelessly divided about the issue of whether or not suicide, or assisted suicide, in such circumstances is morally acceptable. Tying the decision about whether someone is dead to public opinion or individual attitudes about the acceptability of various practices would hardly make for sound public policy. On the other hand, defenders of the DCD protocols note that when patients are not to be resuscitated, and their hearts stop beating, it is common practice to declare them dead right away. They also question whether if someone suffered cardiac arrest on a camping trip with no medical technology available, we would be obligated to wait five, ten, or more minutes to declare them dead simply because we would have had to wait longer in an ICU?

Is There a Solution? We have seen that there are at least three competing definitions of death that are accepted by health professionals and the general public. We have seen that philosophers and bioethicists continue to defend plausible, but competing, definitions of death. We have also seen that the medical criteria and tests for declaring brain death are applied inconsistently with each other. Moreover, the term ‘irreversible’, the reassuring adjective wedded to all definitions and criteria of death, is itself hopelessly vague and context-dependent. Is there a way to resolve these conflicts and contradictions? Will one philosophical view of what exactly constitutes death emerge that is so persuasive that competing views will fall by the wayside? Will medical science become so sophisticated that it can locate the moment of death when philosophy cannot? The answer to each of these questions is no. Proof of the correct definition will not arrive as proof that the world was round instead of flat. New medical technology and the knowledge it brings will only confound society’s understandable wish to have a definitive moment of death.

298

stuart j. youngner

However, death is a social construct. Medical science can better illustrate (or complicate) the biological context in which it takes place. Philosophy can present more or less cogent arguments to choose one loss of function as conceptually more important than another. But culture and context will always be the final arbiter. Brain death may have been ‘grandfathered’ into American culture, but a look at other modern societies illustrates that belief in and acceptance of brain death are not inevitable.

Defining Death in Other Cultures If one were to assume that the course of public acceptance of brain death in other highly industrialized liberal democracies would parallel that in the United States, the assumption would prove incorrect. For example, the Danish parliament passed a law recognizing brain death only in 1990, after long public debate and education encouraged by the Danish government (Rix 1999). Surveys showed, however, that, despite the education, the public still had trouble distinguishing between brain death and PVS. Moreover, open discussion about the ambiguities surrounding brain death may well have contributed to an actual decline in organ donation. Until the National Transplant Act in 1997, brain death in Germany was recognized de facto in clinical practice. However, the 1990s saw a tremendous debate about brain death and its inevitable connection to organ transplantation. Opposition came from several sources, but as Sch¨one-Seifert (1999: 267) has observed, ‘a core objection is the claim that brain death supports a reductionistic view of human beings’. This core objection was part of a more general ‘anti-bioethics’ sentiment that accused bioethics of an unacceptable utilitarian ethic that, argued the critics, resonated uncomfortably with Germany’s Nazi past. Of all the industrialized countries, the debate has been deepest and most bitter in Japan. In her book Twice Dead, the anthropologist Margaret Lock (2002) traces the roots of organ transplantation in the United States and Japan. Unlike the United States, problems in Japan emerged from the beginning, when, in 1968, a transplant surgeon, Wada Jiro, performed the world’s thirteenth heart transplant in Sapporo. Wada both pronounced the donor (an 18-year-old boy who had drowned) dead and performed the transplant surgery. The Japanese news media, involved from the beginning, questioned both whether the surgery was needed by the recipient and if the donor was really dead when his organs were removed. Wada was charged with intentional homicide and professional negligence. It came to light that Wada had tampered with the recipient’s discarded heart to make it look more damaged than it really was. It was also widely noted that Wada had obtained much of his training in the United States. A second heart transplant was not performed in Japan until 1999. In the interim, an incredible public debate ensued. Lock describes it as an unprecedented effort

the definition of death

299

to reach consensus about an issue that had caused ‘national angst’. More than 100 books for the general public were written on the topic. Brain death was the subject of numerous television news specials. The effort to debate and educate included children. Lock (2002: 143) writes: Manga (comics) have an enormous popularity in Japan; their influence on the young is arguably as great as that of television . . . The best-selling Black Jack series . . . read by millions of Japanese, has vividly depicted the living death of brain-dead and PVS patients and the moral dilemmas they pose, and the children’s page in major newspapers put out cartoon pages educating young readers about brain death.

Such a social practice is unimaginable in the United States. Lock paints a complicated picture of the roots of opposition to brain death in Japan, including: deep ambivalence about the ‘West’; the lingering presence of animism in Japanese culture; the absence of Christian charity as practiced in the West; mistrust of medicine; and that death is seen as more of a process in Japan, with the family more fully in control. As proof of its deep ambivalence, Japan did not legally recognize brain death until 1997, and then only grudgingly. Brain death must not only be rigorously established (in contrast to the United States, where the diagnosis is less and less rigorously conducted), but if the patient or family does not accept brain death, the medical diagnosis is not enough to take organs or even to declare death. Both donor and family must have signed donor cards. It is not surprising that the new law has yielded only a handful of donors in the seven years since its inception. Lock argues that while brain death is seemingly accepted in North America and Europe, ‘historical and multi-sited ethnographic research make plain the ambiguities and doubts that have surrounded living cadavers during nearly half of century of existence’ (Lock 2002: 42). The Japanese have an unnerving word for taking organs from brain-dead patients, autaguari, which means harvesting green rice. While no such word use exists in the United States, the stubborn persistence of the term ‘brain death’ may be a less jarring version of the same phenomenon. For two decades transplant surgeons and organ procurement personnel have been arguing that the adjective ‘brain-dead’ should be dropped in favor of the simpler ‘dead’. After all, the argument goes, if you call patients brain-dead instead of dead, it implies that they are different categories. Perhaps the implication is intentional, if not always conscious, and the persistence of the term ‘brain death’ represents a deep-seated sense among lay persons and health professionals that they are not the same. Our use of language often betrays other instances of the ‘ambiguities and doubts’ to which Lock refers. For example, it is commonplace for news reporters, families, and health professionals to talk about ‘brain-dead’ patients being ‘kept alive’ by machines. Perhaps, as the neurologist Alan Shewmon has written, we have simply not developed the proper language with which to understand death in the modern ICU. ‘Instead of proceeding via logical deduction from some preconceived,

300

stuart j. youngner

abstract ‘‘concept of death’’ ’, Shewmon (2004: 279) proposes ‘a phenomenological approach: to step back from linguistically constrained notions . . . and simply look afresh at the spectrum of realities surrounding what people call and have called ‘‘death’’—as though we had never heard of the term ‘‘death’’ and have no compulsion to label any particular phenomenon ‘‘death’’ ’. Shewmon argues that in the pretechnological era, when the events surrounding death occurred simultaneously, the single word ‘death’ sufficed. It no longer does, and our use and insistence on discovering the one correct meaning of death may grow from a delusion that our concepts of life and death are simply generalized from everyday experience and labeled by linguistic convention with the words ‘life’ and ‘death’ when in actuality those very concepts have been shaped by our language, and more importantly constrained by it . . . no wonder that the death-debate is so full of logical inconsistencies and failures to communicate, because people use the same word to express different concepts, not quite realizing (precisely because of the single word) that the concepts are different. (Shewmon 2004: 279)

Shewmon points out that in the west Greenlandic language there are at least forty-nine different words referring to snow and ice—snow on the ground, snow falling in air, air thick with snow, feathery clumps of falling snow, newly fallen snow—that are perceived and conceived by natives of western Greenland as significantly different things. These distinctions are important for a way of life and survival and should be reflected in linguistic distinctions; for those of us for whom snow is but a wintry nuisance or source of fun, the words ‘snow’, ‘slush’ and ‘ice’ probably suffice. Now imagine that pioneers from the tropics migrated to the arctic and suddenly had to learn how to survive there. Their very language, which allows them to see only ‘snow’ and ‘ice’ all around, would be detrimental to survival. (Shewmon 2004: 280)

We are today, says Shewmon, in a similar predicament regarding death. We have migrated through human history into the modern ICU, bringing with us the linguistic baggage of a relatively simple concept of death for which the one word has always sufficed. Now we find ourselves in a situation for which medically real and ethically critical distinctions lack words in the common vocabulary. The best we can do is to speak in awkward paraphrases, such as ‘the point in time beyond which cardiac autoresuscitation is impossible’. To ask which of these technological mumbo-jumbos is really death may perhaps be as linguistically and epistemologically inappropriate as asking an Eskimo which of sullarniq, aput, qantiit, nittaalaq . . . is really ‘snow’. (Shewmon 2004: 280)

Conclusions ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

It is clear that the pressing need for more organs will continue to test the limits of the dead donor rule. Public policy in this regard has several choices. The first is

the definition of death

301

a dramatic redefinition of death to include patients in PVS as a way to ‘quell’ the controversy that taking organs from them would cause. This is not likely. Vehement objections from religious fundamentalists and disability rights activists such as the group Not Dead Yet would make any major gerrymandering of the line between life and death impossible. Even the late Pope John Paul II declared there should be no judgments about the quality of life of PVS patients and that removing a feeding tube from patients in PVS is immoral and constitutes ‘true euthanasia by omission’. The implications of the Pope’s pronouncement for end-of-life care in PVS are uncertain, but is unlikely that the Catholic Church will embrace taking organs from PVS patients anytime soon. A second choice for social policy is to have a public discussion about whether the dead donor rule can be violated under circumstances in which: (1) the donors or their surrogates have authorized removal; and (2) no tangible harm (e.g. pain and suffering) can come to the donor as a result of organ removal. This course would be likely to galvanize the same religious and political forces mentioned above who are wedded to the notion of clear dichotomies—between life and death, killing and letting die, and good and evil. Furthermore, public acknowledgement that patients were being killed by organ removal would likely alienate other groups who already have great mistrust of the health care system. In fact, public discussion about either choice, redefining death yet again or abandoning the dead donor rule, is likely to erode what has appeared to be a workable social consensus, however flawed, about brain death. Most probably, the ongoing shortage of organs will force a gradual erosion of the line between life and death so that ambiguous states are increasingly and quietly included in the dead category or that little fuss occurs when severely compromised but living patients are killed by organ retrieval. The general acceptance of DCD protocols is an example of how this has happened already. If our society really insisted on maintaining a bright line between life and death and between killing and allowing to die, it would reject a two- to five-minute waiting period after heartbeat ceases because there is no proof or agreement about irreversibility at that point. There seems to be a passive acceptance that not much is lost either way. Perhaps, as Shewmon has suggested, we will begin adopting new words or phrases to describe such states and contexts. Perhaps these new words will allow us to avoid the traditional battles about who is ‘alive’ and who is ‘dead’ and whom we are ‘killing’ and whom we are ‘allowing to die’. The new words might improve on current phrases such as ‘good as dead’, ‘euthanasia by omission’, ‘morally irreversible’, and ‘critical function’, which, to some degree, reflect human feeling and practice, but, in the end, fail to capture the complexities of life and death as they are experienced in the modern intensive care unit.

302

stuart j. youngner

References Ad Hoc Committee of the Harvard Medical School (1968), ‘A Definition of Irreversible Coma’, JAMA 205: 337–40. Arnold, R. M., Youngner, S. J., Schapiro, R., and Spicer, C. M. (eds.) (1995), Procuring Organs for Transplant: The Debate Over Non-Heart-Beating Cadaver Protocols (Baltimore: JHUP). Bartlett, E. T. (1995), ‘Differences Between Death and Dying’, Journal of Medical Ethics, 21: 270–6. and Youngner, S. J. (1988), ‘Human Death and the Destruction of the Human Cortex’, in R. M. Zaner (ed.), Beyond Whole Brain Criteria (Dordrecht: Kluwer Academic Publishers), 199–216. Bernat, J. L. (1998), ‘A Defense of the Whole-Brain Concept of Death’, Hastings Center Report, 28/2: 14–23. Culver, C. M., and Gert, B. (1981), ‘On the Definition and Criterion of Death’, Annals of Internal Medicine, 94: 389–94. Campbell, C. S. (1999), ‘Fundamentals of Life and Death: Christian Fundamentalism and Medical Science’, in S. Youngner, R. Arnold, and R. Schapiro (eds.), The Definition of Death: Contemporary Controversies (Baltimore: JHUP), 194–209. Cole, D. J. (1993), ‘Statutory Definitions of Death and the Management of Terminally Ill Patients Who May Become Organ Donors After Death’, Kennedy Institute of Ethics Journal, 3: 145–55. Gervais, K. G. (1986), Redefining Death (New Haven: Yale University Press). Lamb, D. (1992), ‘Reversibility and Death: A Reply to David J. Cole’, Journal of Medical Ethics, 18: 31–3. Lock, M. (2002), Twice Dead: Organ Transplants and the Reinvention of Death (Berkeley: University of California Press). Lynn, J., and Cranford, R. (1999), ‘The Persisting Perplexities in the Determination of Death’, in S. Youngner, R. Arnold, and R. Schapiro (eds.), The Definition of Death: Contemporary Controversies (Baltimore: JHUP), 101–14. Mayo, D., and Wikler, D. (1979), ‘Euthanasia and the Transition from Life to Death’, in W. L. Robinson and M. S. Pritchard (eds.), Medical Responsibility: Paternalism, Informed Consent, and Euthanasia (Clifton, NJ: Humana Press), 195–211. President’s Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research (1981) Defining Death (Washington, DC: US Government Printing Office). Rix, B. A. (1999), ‘Brain Death, Ethics, and Politics in Denmark’, in S. Youngner, R. Arnold, and R. Schapiro (eds.), The Definition of Death: Contemporary Controversies (Baltimore: JHUP), 227–38. Sch¨one-Seifert, B. (1999), ‘Defining Death in Germany: Brain Death and Its Discontents’, in S. Youngner, R. Arnold, and R. Schapiro (eds.), The Definition of Death: Contemporary Controversies (Baltimore: JHUP), 257–71. Shewmon, A. (2004), ‘The Dead Donor Rule: Lessons from Linguistics’, Kennedy Institute of Ethics Journal, 14: 277–300. Siminoff, L., Burant, C., and Youngner, S. J. (2004), ‘Death and Organ Procurement: Public Beliefs and Attitudes’, Social Science in Medicine, 59: 2325–34.

the definition of death

303

Tomlinson, R. (1993), ‘The Irreversibility of Death: Reply to Cole’, Kennedy Institute of Ethics Journal, 3: 157–65. Youngner, S. J., Landefeld, C. S., Coulton, C. J., et al. (1989), ‘Brain Death and Organ Retrieval: A Cross-Sectional Survey of Knowledge and Concepts Among Health Professionals’, JAMA 261: 2205–10. Arnold, R. M., and DeVita, M. A. (1999), ‘When Is Dead?’, Hastings Center Report, 29: 14–21.

chapter 13 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

THE AGING SO CIETY AND THE EXPANSION OF SENILITY: BIOTECHNOLO GICAL AND TREATMENT GOALS •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

stephen g. post Of the many topics worthy of discussion regarding older adults and bioethics, two seem to provide an especially pointed opportunity for reflection on our aging society. First, is aging itself something that biomedical researchers should focus on as a deficit to be overcome through eventual anti-aging treatments? While aging may not fall neatly into the disease category, it is clearly the primary susceptibility factor for the innumerable diseases of older adults, and therefore its potential deceleration consistent with the compression of morbidity might constitute a salutary biomedical goal. The aging society is no panacea to those who suffer from a host of chronic illnesses and feel overwhelmed by the burden of years. Perhaps we must now take scientific steps that may allow an escape from widespread frailty and senility, or such costly demographics will drive us toward inordinate filial,

the aging society

305

economic, and social responsibilities. Such thoughts raise broad questions about endeavors to modify human nature along ‘posthumanist’ lines. Second, we must concentrate on the most challenging problematic of our current aging society, assuming that anti-aging technologies will only become available in future decades. One immense problem is the harsh reality of irreversible progressive dementia, which will serve here as an example of the rise of chronic illness, for which age itself is the primary risk factor. While this second focus overlooks the many older adults who age well and are functional until their final days, these examples of ‘successful aging’ pose no unique issues worthy of special bioethical consideration. And so this chapter is framed around the more unfortunate realities of an aging society, and the related question of whether we ought to strive to ameliorate these realities by developing anti-aging interventions.

Decelerated Aging and the Amelioration of Age-Related Disease ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

The possibility for significant deceleration of human aging can be ethically assessed in a variety of frameworks. The outcome of this assessment is a direct result of the ethical theory and assumptions that disparate camps bring to the debate. Two theories that currently shape a rather acrimonious discussion are natural law and equalitarianism, each of which presents a deeply critical perspective on the goal of prolongevity. Yet despite valid natural law cautions that human goods are violated by anti-aging goals, and meaningful equalitarian concerns that anti-aging treatments would create immense class disparities, there is one extremely powerful ethical argument that may outweigh such criticisms: i.e. the principle of beneficence, which supports the deceleration of aging so long as this significantly diminishes the onset of the many chronic diseases for which aging is the primary risk factor. This assumes that the deceleration of aging will achieve this purpose, rather than the harmful protraction of decrepitude. Ethical theories do have consequences. A resurgence of natural law ethics, which asserts the goodness of human nature and aging as we know it, has of late given rise to a significant set of criticisms of the goal of prolongevity, in particular through the work of Leon R. Kass and the President’s Council on Bioethics, for which he serves as chair. The equalitarians, on the other hand, do not engage in the moralization of human nature as we know it, but nevertheless cannot support the goal of prolongevity until the vast inequalities that currently plague the world—including access to basic healthcare—are corrected. These are both important frameworks for critique of decelerated aging, should this ever become possible.

306

stephen g. post

At first glance, however, it is difficult to see why the deceleration of human aging presents a major ethical concern. There are four reasons for this. 1. Aging may not be a ‘disease’ by most definitions, but it is the major risk factor for innumerable diseases of old age, and is therefore so closely associated with disease that it is not completely implausible to think of it as such. 2. While the concept of ‘aging as disease’ might be thought to reinforce existing stereotypes of older adults, and so contribute to ‘ageism’, there is no logical connection between ‘anti-aging’ research and prejudice. The goal of prolongevity through deceleration of aging has the more obvious inclusive aspect of making old age more enjoyable and acceptable by mitigating the frailties and infirmities that make the increase of years unwelcome. 3. The goal of decelerated aging and prolongevity does not obviously violate the ethical principle of nonmaleficence. Unlike abortion, infanticide, suicide, and euthanasia, the goal of prolongevity does not violate respect for, or sanctity of, life. It is difficult to see how the addition of healthy years to old age could be anything but ‘pro-life’. While moral criticism of therapeutic stem cell cloning is based on the unavoidable destruction of embryos in the process of procuring stem cell lines, there is no such destruction associated with prolongevity. 4. The goal of longer and healthier lives is hardly new and we have been acculturated to it. Life expectancy has risen dramatically over the past two centuries and has been embraced as progress. Public health, sanitation, reduction of infant mortality, antibiotics, and many other factors have contributed to our current level of prolongevity, which some might have deemed an ‘unnatural’ goal a century or more ago. In other words, the genie is in many respects already out of the bottle. This progress in life expectancy might be hampered by sedentary lifestyles, fast foods, obesity, and related diabetes, all of which are viewed as public health problems to be overcome. And yet deeply thoughtful bioethicists are concerned with the implications of anti-aging research and related prolongevity. Kass prepared a discussion paper for members of the Council on Bioethics prior to their meeting on 16–17 January 2003. Entitled ‘Beyond Therapy: Biotechnology and the Pursuit of Human Development’, it established the Council’s future agenda (Kass 2003). Biogerontologists, one of whom is with the National Institute on Aging, can embrace the ideas of ‘The Serious Search for an Anti-Aging Pill’ that would mimic the effects of caloric restriction as an approach to the eradication of disease (Lane et al. 2002). Aging is the prominent risk factor for so many diseases, and the best way to prevent such debilitation is to alleviate aging. The removal of this risk is not obviously contrary to human dignity. In fact, we would all agree that removing the major risk factor for diseases such as Alzheimer’s or osteoporosis would do quite a bit for the preservation of human dignity, for the control of health care costs,

the aging society

307

and for the alleviation of sometimes overwhelming adult filial duties. Still, Kass’s natural law ethics are to be taken seriously, a topic to which attention now turns.

Natural Law Criticisms Natural law places moral value on human nature as we know it, and represents a necessary reaction to post-humanism. One reason why natural law, a form of anti-posthumanism, has an immediate appeal is because many of the major defenders of posthumanism are rather cavalier (see e.g. www.betterhumans.com; www.transhumanism.org; www.foresight.org). Websites reflect the enthusiasm of the young convert to some new image of the human future that is liberated from biological constraints but lacking in wisdom. More reverence for human nature and aging itself as the product of millions of years of evolutionary selection would be the necessary corrective, and this is just what natural law theorists offer. The posthumanist embraces decelerated and even arrested aging, but only as a small part of a larger vision to re-engineer human nature, and thereby to create biologically and technologically superior human beings (Hayles 1999). Posthumans are the much more advanced models that we humans today will design for tomorrow, not unlike new models of automobiles. Genetics, nanotechnology, cybernetics, and computer technologies are all part of the posthuman vision, including the downloading of synaptic connections in the brain to form a computerized human mind freed of mortal flesh, and thereby immortalized. Posthumanists do not take biology to be destiny, and see human nature as an obstacle to be overcome (Hook 2004). They see the next great step in evolution as our own re-creation by human ingenuity. The posthumanist may be irreverent of human nature, but within the boundaries of the technology of the day, humans have been reinventing themselves anyway for millennia. And what is natural and what is unnatural, when the human condition has already been so deeply impacted by technological innovation? Where do we draw lines? As Freeman Dyson writes, ‘the artificial improvement of human beings will come, one way or another, whether we like it or not’, as scientific understanding increases, for after all, such improvement has always been viewed as a ‘liberation from past constraints’ (1997: 12). After all, the idea of human flight was once deemed heretical hubris in the light of eternity and nature. So it is that Gregory Stock writes a book entitled Redesigning Humans: Our Inevitable Genetic Future, in which he introduces the idea of ‘superbiology’ as we take full control of our own biology in turning toward perfectibility (2002). Thus did George Bernard Shaw, in his remarkable Back to Methuselah (1921), take up the variations in lifespans between species such as parrots and dogs, or turtles and wasps, and ask that science rise to the redemptive occasion by vastly increasing human longevity. Indeed, Francis Bacon, a founder of the scientific

308

stephen g. post

method four centuries ago, in his essay The New Atlantis (1996), set in motion a biological mandate for boldness that included both the making of new species, or ‘chimeras’, organ replacement, and the ‘Water of Paradise’ that would allow the possibility to ‘indeed live very long’. Three centuries before Francis Bacon the English theologian Roger Bacon argued that in the future the 900-year-long lives of the antediluvian patriarchs would be restored alchemically. Like many Western religious thinkers, both Bacons saw death as the unnatural result of Adamic fall into sin. These Western dreams of embodied near-immortality could only emerge against a theological background that more or less endorses them. The modern goals of anti-aging research and technology are historically emergent from a pre-modern religious drama of hope and salvation. Longings for immortality within a religious context are understandable in light of existential anxiety over finitude and mortality, and the goal of prolongevity may be—at least historically—shaped by such passions. Renaissance science transferred the task of achieving immortality from heaven to earth in the spirit of millennial hopes. The economy of salvation presented by Dante was replaced by the here and now. There is a vibrant millennialist optimism in the responsible biogerontologists, for they have proclaimed aging itself to be surmountable to degrees through human ingenuity. At the end of The New Atlantis, Bacon lists more specifically among the goals of science ‘the retardation of age’, which is taken with great seriousness by today’s biogerontologists who create scientific breakthroughs with fruit flies, roundworms, rodents, and monkeys, but with an eye toward the alleviation of senility in an already aging society plagued by chronic age-related diseases. But the natural law theorists attempt to stem such rash tides. One of the wiser minds of the last century, Hans Jonas (d. 1993), an intellectual inspiration for today’s natural law anti-posthumanists, first articulated deep questions about the prolongevity agenda. How desirable would this power to slow or arrest aging be for the individual and for the species? Do we want to tamper with the delicate biological balance of death and procreation, and preempt the place of youth? Would the species gain or lose? Jonas, by merely raising these questions, meant to cast significant doubt on the anti-aging enterprise. His later essays raising many of these same questions were published posthumously (1996). Many of these questions are echoed in the writings of Kass, who for the most part accepts biotechnological progress within a therapeutic mode; his issue is with efforts to enhance and improve upon the givens of human nature. There is arguably a tone of solipsism in grasping at extended life rather than accepting old age and celebrating youth in the lives of our offspring. Responsibility to future generations precludes a clinging to our own youthfulness. There is wisdom in simply accepting the fact that we evolved for reproductive success rather than for long-lived lives. Without such wisdom will we lose sight of our deepest creative

the aging society

309

motives? Possibly. And so it is written, ‘teach us to number our days that we may get a heart of wisdom’ (Psalm 90). The possibility of fundamentally modifying the rate of human aging and thereby altering the perennial process of generational transition could prove destructive of all the human propensities to nurture and elevate the young. Perhaps the improvement of the human condition lies not in modifying the human vessel, but in developing the treasures within, such as compassion, justice, and dignity. In summary, the natural law traditions represented by anti-posthumanists exhort us to live more or less according to nature, and warn that our efforts to depart from what we are will result in new evils that are more perilous than old ones. To use an analogy, we are like the sailor who climbs as high on the mast as one can in order to rise high above the waters of nature, but only to see the boat capsize under our weight tipping the mast into the waves below. How can we presume that the brave new world will be a better world? Should not the burden of proof be on the proponents of radical change? Who are we today to impose our arbitrary images of human enhancement on future generations? Thus are anti-posthumanist thinkers critical of biogerontology to the extent that it seeks to slow the aging process. Our focus, argue the anti-posthumanists, should be on the acceptance of aging rather than on its scientific modification. The intergenerational thrust of evolution, by which we are inclined toward parental and social investment in the hope, energy, and vitality of youth, provides the basis for a natural law ethic that requires us all to relinquish youthfulness. I agree with the anti-posthumanists that a decision for or against decelerated aging based on superficial thinking and a commitment to individual freedom is a formula for the easy destruction of what is good in being human. They are right to place an emphasis on Aristotelian ‘final causes’ and human goods. Unless we are radical relativists, it can and must be asked whether any scientific aspiration contributes to the human condition or detracts from it. There is truth in the claim of anti-posthumanists that a contemporary bioethics that does not ask such questions is inadequate. And yet the natural law criticism of the goal of prolongevity does not leave me quite convinced. In the context of posthumanism, the religion of technology, and scientific hubris, efforts to slow or arrest human aging appear morally ambiguous at best. And yet one aspect of human nature has always been freedom over itself. Freedom causes us to break the harmonies of nature and establish new harmonies, especially when confronted with severe risks that require a creative solution.

A Solution to the Chronic Illnesses Associated with Aging? If we bracket out this ancient debate between the natural law and its detractors, there is another context that is more immediate. The stark reality of our already

310

stephen g. post

aging societies is that, factoring out infant mortality, people can expect to live into their eighties. Many will experience chronic illness for which old age is the dominant risk factor, ranging from Alzheimer’s and Parkinson’s to osteoporosis and vascular disease. Our demographic transition to greatly increased life expectancy through means that do not involve the deceleration of aging has resulted in a scenario that is clearly not idyllic, and the solution may rest with advances in the basic science of aging that would achieve even greater prolongevity but without the perils of massive debilitation. In other words, to resolve the problems of senility and dementia, brought on chiefly by enhanced life expectancy absent anti-aging interventions, we must now develop those interventions or suffer the immense negative consequences. Among those consequences are future generations of young people extremely burdened with the direct and indirect support of tens of millions of citizens who require demanding, complex, constant, and expensive care. When one considers the goal of decelerated aging in this context, it begins to appear more rational, salutary, and necessary. No longer the bizarre dream of superficial technology zealots, decelerated aging appears to be a more reasonable aspiration. Indeed, the biogerontologists hard at work in serious research are much less the children of posthumanism than they are of a will to benefit a common humanity regrettably caught between the old world of relatively short and ‘natural’ life expectancy and the future world of the nonfrail and nonsenile. So, the individual confronted with the possibility of decelerating his or her aging process may not ask the question of the dignity of human nature as we know it, or of whether we humans are wise enough to control our future development, or of equalitarian justice. Instead, he or she may reflect on loved ones who have struggled with a syndrome that strips away memory and self-identity, such as dementia, and quickly declare that his or her true dignity lies in decelerated aging consistent with the retention of cognitive temporal glue between past, present, and future. It may well be that the only real progress we will see in ridding the world of such debilitating diseases as Alzheimer’s is through slowing the process of aging because old age is by far the most significant risk factor. Currently there is no reason for strong optimism about scientific breakthroughs to cure this disease, or to stabilize symptoms in the long term, although there are some compounds in use to ameliorate some symptoms for a limited period. Thus, basic scientific advances in the area of aging itself could, if successful, significantly reduce the incidence of diseases associated with old age that are too complex to be solved on their own terms. It is, then, unfair to responsible anti-aging researchers to suggest that they are engaged in an effort that will only radically protract the lives of the most deeply forgetful. Maybe greatly enhanced prolongevity is not all bad if a number of conditions can be met regarding preserved memory and other dimensions of health and rigor. It might also be that, given the realities of the old age boom, anti-aging research will prove curative where all else has failed.

the aging society

311

Will anti-aging researchers provide the solution? Two respected researchers from the University of London conclude in Nature that ultimately we will find the human lifespan to be quite plastic, and scientific progress in this area ‘may allow us to reduce the impact of ageing-related diseases as the limits on the human lifespan recede’ (Partridge and Gems 2002: 921). While I share natural law concerns over the unleashed powers of biotechnology to influence what increasingly appears to be a somewhat malleable human nature, and while I too cringe at immature millennialist enthusiasm over the remaking of the species, I nevertheless can appreciate the goal of eradicating our modern epidemic of chronic illness, which grows more extensive and costly with every elevation in average life expectancy. Indeed, it is because of the potential positive health consequences of anti-aging treatments that they will probably not be easily restrained in the name of natural law or equalitarianism. It may well be that, in the process of decelerating aging with the intent of ameliorating chronic diseases, the human lifespan will be extended. We do not have much choice but to move forward. The goal of prolongevity is rooted in the reality of the 80-year-old who has struggled with hip replacements, a diagnosis of dementia, retinopathy, and related depression. Perhaps Kass is correct that under conditions of prolongevity many of us would lose interest and engagement in life, seriousness and aspiration, and virtue and moral excellence centered on a generative relationship with the young. Yet more surely, a potential solution to the widespread problem of the death of the mind before the death of the body may rest with the anti-aging technologies of the future, and we must at least seriously grapple with this important and very real problem. A number of scientists convened in 1999, under the auspices of the National Institute on Aging, for a conference entitled Caloric Restriction’s Effects on Aging: Opportunities for Research on Human Implications, where they charted a bold new therapeutic research agenda focused on slowing or ‘retarding’ the aging process, which implies the extension of the lifespan (Masoro 2001). The driving motivations expressed, however, were focused on slowing aging in order to prevent the many diseases for which old age puts people at risk. This scholarly meeting sparked no significant public debate, although few things could be more deserving of critical reflection on human aspirations than this. A principal topic was the research to develop a treatment to prolong life and youthful vigor through imitating the remarkable effects of caloric restriction, which for more than sixty years has been documented for its dramatic ability to slow aging in nonhuman mammals and significantly extend their lifespans. More research on the basic science of aging will hopefully give rise to the ability to decelerate this natural process. People will be attracted to this technology because they know how burdensome are the diseases to which old age makes us all susceptible. Should they wish to select this medical option, as a way of insuring a better quality of life for themselves and for those loved ones who would otherwise

312

stephen g. post

have to provide long-term care, few would condemn them unless the problem of prolonged senescence arises. We simply cannot predict the future, but over the last century we have seen a dramatic rise in life expectancy, and it is hard to imagine that the addition of more healthy years would be widely deemed contrary to human dignity. But there is reason to urge extreme caution. The technology to slow the aging process may be coming nearer to reality, yet the goal of prolongevity has not been carefully considered (Juengst et al. 2003). In a time when biotechnology is allowing for the reconstruction of both nature and human nature, all thoughtful citizens must ponder the implications of potentially dramatic change. Of the many possible biotechnological goals on the horizon, which ones are likely to enhance the human condition, and which ones are likely to diminish human dignity? We think of the provocative developments in therapeutic cloning, in fertility and reproduction, in organ procurement and transplantation, in genetic testing and therapy, or in the treatment of a myriad of illnesses, and our collective breath is taken away by the pace of change. But we are also rightly haunted by the reality that, while biotechnological powers grow, human nature has in no obvious way progressed with regard to unselfish behavior, humility, peace, and equality. Thus, we raise the question of the very nature of goodness, and whether some biotechnological developments divert us from growth in virtue, or even tempt us to create a new class of an ageless elite that inevitably begins to look down upon the ordinary older adult as a misfit. Should we move forward in the twenty-first century as bold new ‘co-creators’ of our somewhat malleable human nature, or should we accept a more humble approach that endorses a caring and just stewardship over human nature more or less as it is, seeking therapies rather than transformations? At least in the area of decelerated aging, where therapy and enhancement merge, we will probably move forward. Progress takes creativity, and in the meanwhile, we must deal as humanely as possible with the major moral, social, economic, familial, and medical challenges of an aging society—the death of the mind before the death of the body.

Death and Dementia ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

The syndrome of dementia is an irreversible decline in cognitive abilities that causes significant dysfunction. Like most syndromes, dementia can be caused by a number of diseases. In the nineteenth century, for example, a main cause of dementia was syphilis. Currently, as a result of dramatic increases in average human life expectancy, dementia is caused primarily by a number of neurological diseases associated with old age. Dementia is distinguished

the aging society

313

from ‘pseudo-dementia’ because the latter is reversible; for example, depression, extreme stress, and infection can cause dementia but with treatment a return to a former cognitive state is likely. Dementia is also distinguished from ‘normal age-related memory loss’, which effects most people by about age 70 in the form of some slowing of cognitive skills and a deterioration in various aspects of memory. Although dementia can have many causes, the primary cause of dementia in our aging societies is Alzheimer’s disease (AD). Approximately 60 per cent of dementia in the American elderly and worldwide in industrialized nations is secondary to AD (US General Accounting Office 1998). This discussion will focus on so-called ‘Alzheimer’s dementia’ in order to illustrate ethical issues that pertain to all progressive dementias. Epidemiologists differ in their estimates of late-life AD prevalence, but most studies agree roughly on the following: about 1–2 per cent of older adults at age 60 have probable AD, and this percentage doubles every five years so that 3 per cent are affected at age 65, 6 per cent at age 70, 12 per cent at age 75, and 24 per cent by age 80. While some argue that those who live into their nineties without being effected by AD will usually never be affected by it, this is still speculative. According to a Swiss study, 10 per cent of nondemented persons between the ages of 85 and 88 become demented each year (Aevarsson 1996). There are very few people in their late forties and early fifties who are diagnosed with AD. Without delaying or preventive interventions, the number of people with AD, in the United States alone, will increase to 12 to 14 million by 2050. These numbers represent a new problem of major proportions and immense financial consequences for medicine, families, and society. Various stage theories of disease progression have been developed. However, in clinical practice, professionals speak roughly of three stages. In mild stage dementia, the newly diagnosed patient has significant cognitive losses resulting in disorientation and dysfunction, and often displays affective flatness and possibly depression. In moderate stage dementia, the patient forgets that he or she forgets, thereby gaining relief from insight into losses. Some patients will at this point adjust well emotionally to a life lived largely in the pure present, although some long-term memories are still in place. The recognition of loved ones is usually still possible. However, as many as a third of patients in the moderate stage will struggle with emotional and behavior problems, including agitation, combativeness, paranoia, hallucinations, wandering, and depression. The advanced stage of dementia includes a loss of all or nearly all ability to communicate by speech, inability to recognize loved ones in most cases, loss of ambulation without assistance, incontinence of bowel and/or bladder, and some weight loss due to swallowing difficulties. The advanced stage is generally considered terminal, with death occurring on average within two years. AD, however, is heterogeneous in its manifestations, and defies simplistic staging.

314

stephen g. post

A Fundamental Moral Question: Do People with Dementia ‘Count’? Despite the seriousness of dementia and the responsibilities it creates for caregivers, it is ethically important that the person with dementia not be judged by ‘hypercognitive’ values (Post 2000). The self is not cognition alone, but is rather a complex entity with emotional and relational aspects that should be deemed morally significant and worthy of affirmation (Sabat 2001). A bias against the deeply forgetful is especially pronounced in ‘personhood’ theories of moral status in which persons are defined by the presence of a set of cognitive abilities (Kitwood 1997). After discussion of the disparities in bioethical thinking about what constitutes a person, Stanley Rudman concludes, ‘It is clear that the emphasis on rationality easily leads to diminished concern for certain human beings such as infants . . . and the senile, groups of people who have, under the influence of both Christian and humanistic considerations, been given special considerations’ (1997: 47). Often, the personhood theorists couple their exclusionary rationalistic views with utilitarian ethical theories that are deeply incoherent with regard to life and death. As Rudman summarizes the concern, rationality is too severe a ground for moral standing, ‘allowing if not requiring the deaths of many individuals who may, in fact, continue to enjoy simple pleasures despite their lack of rationality’ (1997: 57). Of course, in the real world of families, love, and care, personhood theories have no practical significance. The philosophical tendency to diminish the moral status or considerability of people with dementia is also related to a radical differentiation between the formerly intact, or ‘then’, self and the currently demented, or ‘now’, self. The reality is that until the very advanced and even terminal stage of AD, the person with dementia will usually have sporadically articulated memories of deeply meaningful events and relationships ensconced in long-term memory. It is wrong to bifurcate the self into ‘then’ and ‘now’, as if continuities are not at least occasionally striking (Kitwood 1997; Sabat 2001). This is why it is essential that professional caregivers be aware of the person’s life story, making up for losses by providing cues toward continuity in self-identity. Even in the advanced stage of dementia, as in the case presented at the outset of this chapter, one finds varying degrees of emotional and relational expression, remnants of personality, and even meaningful nonverbal communication as in the reaching out for a hug. The fitting moral response to people with dementia, according to classical Western ethical thought and related conceptions of common human decency, is to enlarge our sense of human worth to counter an exclusionary emphasis on rationality, efficient use of time and energy, ability to control distracting impulses, thrift, economic success, self-reliance, self-control, ‘language advantage’, and the like. As Alasdair MacIntyre argues, we have made too much of the significance of language, for instance, obscuring the moral significance of species who lack

the aging society

315

linguistic abilities, or human beings who have lost such abilities (MacIntyre 1999). It is possible to distinguish two fundamental views of persons with dementia. Those in the tradition of Stoic and Enlightenment rationalism have achieved much for universal human moral standing by emphasizing the spark of reason (logos) in us all; yet when this rationality dissipates, so does moral status. Those who take an alternative position see the Stoic heritage as an arrogant view in the sense that it makes the worth of a human being entirely dependent on rationality, and then gives too much power to the reasonable. This alternative view is generally associated with most Jewish and Christian thought, as well as that of other religious traditions in which the individual retains equal value regardless of cognitive decline. As the Protestant ethicist Reinhold Niebuhr wrote, ‘In Stoicism life beyond the narrow bonds of class, community, and race is affirmed because all life reveals a unifying divine principle. Since the principle is reason, the logic of Stoicism tends to include only the intelligent in the divine community. An aristocratic condescension, therefore, corrupts Stoic universalism’ (1956: 53).

Diagnostic Truthtelling Diagnostic truthtelling in the context of dementia should be handled as it is in other medical contexts: be as truthful as information permits while attending to the patient’s need for social, emotional, spiritual, and practical support. Compassionate diagnostic disclosure is a moral act of respect for persons, an opportunity for human resilience and community, and a necessary practical step toward future planning. In serving on more than 120 ethics panels organized by Alzheimer’s Association chapters across the United States and Canada in the late 1990s, comprised of leading physicians and other health professionals caring for those with dementia, as well as of family caregivers and affected patients, I did not find a single individual among an estimated 880 who thought that compassionate honesty was not the best policy. There were disagreements about optimal emotional and relational supports, precise wording and approach, and whether to inform the patient together with the family or independently. There are some poor excuses for deceit and nondisclosure: Anxiety. The idea of denying the truth to protect a patient from anxiety underestimates the remarkable human capacity to deal creatively and resiliently with the implications of serious diagnoses, and denies the power of a caring family or community to provide emotional healing. Many patients, when provided with a diagnosis, are actually relieved of the anxiety that stems from uncertainty. The ways in which people cope with a diagnosis need to be empirically examined, but if we were to argue that anxiety-related behaviors justify nondisclosure or deceit, then in the final analysis no patient would be told of any serious diagnosis,

316

stephen g. post

with disastrous results. Clinically significant depression, however, indicates a more cautious and gradual approach than might otherwise be the case. Secondly, a diagnosis and its emotional challenges to the patient mobilize family and community to provide the care and acceptance without which only further isolation is possible. Culture. The physician may encounter some patients from cultures in which the model of nondisclosure to the patient is still in place, and families still operate in a highly ‘protective’ manner. However, it is well documented that individuals often want diagnostic information despite cultural pressures to the contrary. A physician should never presume to withhold information unless the patient specifically requests this. Professional commitment is to patients, not cultures. Moreover, cultures can have moral blind spots that are only gradually modified through interacting with other traditions, including the tradition of diagnostic truthtelling on the basis of human dignity. Lack of objectivity. The syndrome of dementia is comprised of a cluster of symptoms. The core feature of dementia is decline in cognitive abilities that causes significant dysfunction. Such cognitive dysfunction is reasonably objective. Like most syndromes, dementia can be caused by a number of diseases. All possible conditions—e.g. pseudo-dementia, dementia, and Alzheimer’s disease (AD)—can and should be supportively discussed with patients, and, when possible, treated with consent. Not to present reasonably clear diagnostic information is to disenfranchise the person who is experiencing an illness—and usually is well aware of some losses—and to create a climate of distrust that will ultimately serve no good purpose. Diagnostic truthtelling is the necessary beginning point for an ethics of ‘precedent autonomy’ for those who wish to implement control over their futures through advance directives such as durable power of attorney for health care, which allows a trusted loved one to make any and all treatment decisions once the agent becomes incompetent. This can effectively be coupled with a living will or some other specific indication of the agent’s material wishes with regard to end-oflife care. Unless the person knows the probable diagnosis in a timely way while still competent to file such legal instruments, the risk of burdensome medical technologies is increased. Even in the absence of such legal forms, however, many technologically advanced countries will allow next of kin to decide against efforts to extend life in severe dysfunction. This is important because many patients suffer incapacitating cognitive decline long before having a diagnostic work up; those who are diagnosed early enough to exercise their autonomy can become quickly incapacitated. In the United States the Alzheimer’s Disease Association does not support mandatory reporting of a probable diagnosis of AD to the Department of Motor Vehicles. There are a number of reasons for this caution, one of which is patient confidentiality. Reporting requirements might discourage some persons from

the aging society

317

coming into the clinic for diagnosis at a time early in the course of disease when drug treatments are most clearly indicated. Eventually all people with AD must stop driving when they are a serious risk to self or others. Family members must know that if a loved one drives too long and injures others, they may even be held financially liable and insurers may not be obliged to cover this liability. Ideally, a privilege is never limited without offering the person ways to fill in the gaps and diminish any sense of loss. An ‘all or nothing’ approach can and should be avoided. Compromise and adjustments can be successfully implemented by those who are informed and caring, especially when the person with AD has insight into diminishing mental abilities and loss of competence. The affected person should retain a sense of freedom and self-control if possible (Alzheimer’s Disease Association 2001). Truthtelling, which is clearly the accepted norm in modern medical care, has been encouraged by the emergence of new treatments such as cholinesterase inhibitors, and by an appreciation for the rights and needs of patients as they and their families plan for the future, financially and otherwise. The Association’s statement (2001) includes the important argument that disclosing the diagnosis early in the disease process allows the person to ‘be involved in communicating and planning for end-of-life decisions’. Diagnostic truthtelling is the necessary beginning point for responsible stewardship over one’s future through advance directives such as durable power of attorney for health care, which allows a trusted loved one to make any and all treatment decisions once the agent becomes incompetent. This can effectively be coupled with a living will or some other specific indication of the agent’s material wishes with regard to end-of-life care. Unless the person knows the probable diagnosis in a timely way while still competent to file such legal instruments, the risk of burdensome medical technologies is increased. With early treatments available, with options for research participation, with the availability of support groups where these exist, with many patients already having a significant knowledge base through informational technologies, with the need to be protected against the aggressive use of burdensome medical technologies as the disease progresses, and with the importance of financial, family, and career planning, it is simply impossible to defend the old paternalistic model of the physician–patient relationship. It is the job of the physician and the health care team, including pastoral care as requested, to disclose a diagnosis of dementia and, when possible, the disease cause underlying it, in a sensitive and supportive way. In individual cases, this may require different approaches, timing, and venues, but there is no alternative consistent with human dignity, resilience, and care. Where the patient can no longer retain information in a sustained and meaningful manner, diagnostic truthtelling is impossible and uncalled for. For those whose ability to retain information is unclear or fast fading, one trusts in the physician to improvise as the case merits.

318

stephen g. post

Quality of Life Emotional, relational, aesthetic, and symbolic well-being are possible to varying degrees in people with progressive dementia (Kitwood 1997). Quality of life can be much enhanced by working with these aspects of the person. Kitwood provides indicators of well-being in people with severe dementia: the assertion of will or desire, usually in the form of dissent despite various coaxings; the ability to express a range of emotions; initiation of social contact (for instance, a person with dementia has a small toy dog that he treasures and places it before another person with dementia to attract attention); affectional warmth (for instance, a woman wanders back and forth in the facility without much socializing, but when people say hello to her she gives them a kiss on the cheek and continues her wandering). In enhancing quality of life it is crucial to accept the reality of the person with dementia rather than try to impose one’s own reality. In the mild stage of AD there is much to be said for trying to orient a person to reality; at some point in moderate AD, however, it becomes oppressive to impose reality upon them. The aesthetic well-being available to people with AD is obvious to anyone who has watched art or music therapy sessions. In some cases, a person with advanced AD may still draw the same valued symbol, as though through art a sense of self is retained (Firlik 1991). The abstract expressionist de Kooning painted his way through much of his fourteen-year struggle with AD until his death in 1996. Good caregivers know how to work with remaining capacities in the person with dementia, rather than against them. In general, quality of life is a self-fulfilling prophesy. If those around the person with dementia see the glass as half empty and make no efforts to relate to the person in ways that enhance his or her experience, then quality of life is minimal. Steven R. Sabat, who has produced the definitive observer study of the experience of dementia, underscores the extent to which the dignity and value of the person with dementia can be maintained through affirmation and an existential perspective (Sabat 2001).

A Relatively ‘Good’ Dying AD is on the leading edge of the debate over physician-assisted suicide and euthanasia. The policies that emerge from this debate will have monumental significance for people with dementia, and for social attitudes toward the task of providing care when preemptive death is cheaper and easier. The Association affirms the right to dignity and life for every Alzheimer patient, and cannot condone suicide (Alzheimer’s Disease Association 2001). The Association asserts that the refusal or withdrawal of any and all medical treatment is a moral and legal right for all competent Americans of age, and this right can be asserted by a family surrogate acting on the basis of either ‘substituted

the aging society

319

judgement’ (what would the patient when competent have wanted) or ‘best interests’ (what seems the most humane and least burdensome option in the present). The Association concludes that AD in its advanced stage should be defined as a terminal disease, as roughly delineated by such features as the inability to recognize loved ones, to communicate by speech, to ambulate, or to maintain bowel and/or bladder control. When AD progresses to this stage, weight loss and swallowing difficulties will inevitably emerge. Death can be expected for most patients within a year or two, or even sooner, regardless of medical efforts. One useful consequence of viewing the advanced stage of AD as terminal is that family members will better appreciate the importance of palliative (pain medication) care as an alternative to medical treatments intended to extend the dying process. All efforts at life extension in this advanced stage create burdens and avoidable suffering for patients, who could otherwise live out the remainder of their lives in greater comfort and peace. Cardiopulmonary resuscitation, dialysis, tube-feeding, and all other invasive technologies should be avoided. The use of antibiotics usually does not prolong survival, and comfort can be maintained without antibiotic use in patients experiencing infections. Physicians and other health care professionals should recommend this less burdensome and therefore more appropriate approach to family members, and to persons with dementia who are competent, ideally soon after initial diagnosis. Early discussions of a peaceful dying should occur between persons with dementia and their families, guided by information from health care professionals on the relative benefits of a palliative care approach (Alzheimer’s Disease Association 2001). Avoiding hospitalization will also decrease the number of persons with advanced AD who receive tube-feeding, since many long-term care facilities send residents to hospitals for tube placement, after which they return to the facility. It should be remembered that the practice of long-term tube-feeding in persons with advanced dementia began only in the mid-1980s after the development of a new technique called ‘percutaneous endoscopic gastrostomy’ (PEG). Before then, such persons were cared for through assisted oral feeding. In comparison with assisted oral feeding, however, long-term tube-feeding has no advantages, and a number of disadvantages (Alzheimer’s Disease Association 2001). Gastrostomy tube-feeding became common in the context of advanced dementia and in elderly patients more generally after 1981, secondary to the development of the PEG procedure. The PEG procedure was developed by Dr Michael Gauderer and his colleagues at Rainbow Babies and Children’s Hospital in Cleveland (1979–80) for use in young children with swallowing difficulties. The procedure required only local anesthesia, thus eliminating the significant surgical risk associated with general anesthesia and infection (Gauderer and Ponsky 1981). Gauderer wrote, two decades later, that while PEG use has benefited countless patients, ‘in part because of its simplicity and low complication rate, this minimally invasive procedure also lends itself to over-utilization’ (Gauderer 1999). Expressing moral concerns about the proliferation of the procedure, Gauderer indicates that as the third decade of

320

stephen g. post

PEG use begins to unfold, ‘much of our effort in the future needs to be directed toward ethical aspects’ (1999: 882). PEG is being used more frequently even in those patients for whom these procedures were deemed too risky in the past. For over a decade, researchers have underscored the burdens and risks of PEG tube-feeding in persons with advanced dementia. The mounting literature was well summarized by Finucane et al., who found no published evidence that tube-feeding prevents aspiration pneumonia, prolongs survival, reduces risks of pressure sores or infections, improves function, or provides palliation in this population (Finucane et al. 1999; Gillick 2000; Post 2001). Families often perceive tube-feeding as preventing pneumonia or skin breakdown, and many assume that it extends survival. These perceptions are erroneous. The main benefit of PEG is that it makes life easier for the informal family caregiver who, for reason of competing duties or perhaps physical limitation, cannot find the time or energy to engage in assisted oral feeding. Yet PEG use is not really ‘easy’, because it has its technological complexities, and the recipient will usually have diarrhea. In some cases, physical restraints are used to keep a person from pulling on the several inches of tube that extend out of the abdomen. One wonders if assisted oral feeding is not easier after all. Regardless, purported technical ease and efficiency do not mean that these technologies should be applied. Should persons with advanced progressive dementia ever be provided with PEG? In general, assisted oral feeding and hospice are the better alternative to tube-feeding, although in practice there will be some cases in which the limited capacities of an informal family caregiver do justify tube-feeding as the ethically imperative alternative to starvation when the ability to swallow has begun to diminish. Ideally home health aides would make assisted oral feeding possible even in these cases, but this is not a priority in the current health care system. Institutions, however, should uniformly provide assisted oral feeding as the desired alternative to tube-feeding, a measure that would profoundly obviate the overuse of this technology. There will be many family caregivers who have no interest in PEG use and who feel that they are being loyal to their loved one’s prior wishes. A physician should expect this response. A study included in-person interviews of eighty-four cognitively normal men and women aged 65 years and older from a variety of urban and suburban settings (including private homes, assisted-living apartments, transitional care facilities, and nursing homes). Three-fourths of the subjects would not want cardiopulmonary resuscitation, use of a respirator, or parenteral or enteral tube nutrition with the milder forms of dementia; 95 per cent or more would not want any of these procedures with severe dementia (Gjerdingen et al. 1999). These subjects were adequately informed of the burdens and benefits of such interventions. Physicians and other health care professionals should recommend this less burdensome and therefore more appropriate approach to family members, and to persons with dementia who are competent, ideally soon after initial diagnosis. Early discussions of a peaceful dying should occur between persons with dementia and

the aging society

321

their families, guided by information from health care professionals on the relative benefits of a palliative care approach (Volicer and Hurley 1998). As we move forward to overcome the problematic of a partial mental death that precedes physical death, a degree of moral complexity arises that is quite distinct from death due to conditions that leave the autobiographical self intact.

A Concluding Call for Progress ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

While I have argued that we should care for the deeply forgetful by enhancing the quality of their lives, I have also suggested that openness to efforts to decelerate aging—the key risk factor associated with so many debilitating morbidities—should move forward to the extent that such deceleration is found to significantly reduce, rather than protract, said morbidities. Even for the antiposthumanists, the direct goal of anti-aging research and interventions should be the amelioration of those morbidities, including AD, while a secondary unintended effect is the further expansion of life expectancy, and even of maximum lifespan. Aging itself may not be easily categorized as a disease, although it is most certainly a process of natural deterioration. In a time when average life expectancy was low, people who survived birth and childhood could expect to live into their sixties, although always some individuals lived well beyond that. Now, as most people live into their late seventies and beyond, the implications of aging as a susceptibility factor for innumerable diseases and chronic conditions become more vivid. Our aging society becomes a complex of widespread deterioration, dysfunction, and dependence. Unless we create anti-aging interventions that also reduce morbidity and dysfunction to a considerable extent, we will continue on in the halfway house that we have created, in which many lives are burdened with the illnesses of old age, at great cost to their adult children and to society. Beneficence requires that we take the final bold step of freeing ourselves from decrepitude, should this become possible, for reasons of individual well-being and intergenerational covenant. For the adult children who sacrifice so much for a parent with dementia, and for a society that must care for the senile, the move toward a more long-lived society through unlocking the basic cellular mechanisms of aging and decelerating it consistent with the compression of morbidity must be welcomed.

References Aevarsson, O., and Skoog, I. (1996), ‘A Population-Based Study on the Incidence of Dementia Disorders Between 85 and 88 Years of Age’, Journal of the American Geriatrics Society, 44: 1455–60.

322

stephen g. post

Alzheimer’s Disease Association (2001), Ethical Issues in Alzheimer’s Disease (Chicago: Alzheimer Disease Association). Bacon, F. (1996), The New Atlantis, in Francis Bacon: A Critical Edition of the Major Works, ed. B. Vickers (Oxford: Oxford University Press), 457–89. Dyson, F. J. (1997), Imagined Worlds (Cambridge, Mass.: Harvard University Press). Finucane, T. E., Christmas, C., and Travis, K. (1999), ‘Tube Feeding in Patients with Advanced Dementia: A Review of the Evidence’, JAMA 282: 1365–70. Firlik, A. D. (1991), ‘Margo’s Logo’, JAMA 265: 201. Gauderer, M. (1999), ‘Twenty Years of Percutaneous Endoscopic Gastrostomy: Origin and Evolution of a Concept and Its Expanded Applications’, Gastrointestinal Endoscopy, 50: 879–82. and Ponsky, J. L. (1981), ‘A Simplified Technique for Constructing a Tube Feeding Gastrostomy’, Surgery in Gynecology and Obstetrics, 152: 83–5. Gillick, M. R. (2000), ‘Rethinking the Role of Tube Feeding in Patients with Advanced Dementia’, New England Journal of Medicine, 342/3: 206–10. Gjerdingen, D. K., Neff, J. A., Wang, M., and Chaloner, K. (1999), ‘Older Persons’ Opinions About Life-Sustaining Procedures in the Face of Dementia’, Archives of Family Medicine, 8: 421–5. Hayles, N. K. (1999), How We Become Posthuman: Virtual Bodies in Cybernetics, Literature and Informatics (Chicago: University of Chicago Press). Hook, C. C. (2004), ‘Transhumanism and Posthumanism’, in S. G. Post (ed.), The Encyclopedia of Bioethics, 3rd edn., 5 vols. (New York: Macmillan Reference). Jonas, H. (1985), The Imperative of Responsibility: In Search of an Ethics for the Technological Age (Chicago: University of Chicago Press). (1996), Mortality and Morality: A Search for the Good After Auschwitz (Evanston, Ill.: Northwestern University Press). Juengst, E. T., Binstock, R. H., Mehlman, M., and Post, S. G. (2003), ‘The Social Implications of Genuine Anti-Aging Interventions: The Need for Public Dialogue’, Science, 299: 1323. Kass, L. R. (2003), ‘Beyond Therapy: Biotechnology and the Pursuit of Human Development’,

, accessed 12 Mar. 2003. Kitwood, T. (1997), Dementia Reconsidered: The Person Comes First (Buckingham: Open University Press). Lane, M. A., Ingram, D. K., and Roth, G. S. (2002), ‘The Serious Search for an Anti-Aging Pill’, Scientific American, 287: 36–41. MacIntyre, A. (1999), Dependent Rational Animals: Why Human Beings Need the Virtues (Chicago: Open Court). Masoro, E. J. (2001), Caloric Restriction’s Effects on Aging: Opportunities for Research on Human Implications, Journal of Gerontology, ser. a, 56a, special issue 1. Niebuhr, R. (1956), An Interpretation of Christian Ethics (New York: Meridian). Partridge, L., and Gems, D. (2002), ‘A Lethal Side-Effect’, Nature, 418: 921. Post, S. G. (2000), The Moral Challenge of Alzheimer Disease: Ethical Issues from Diagnosis to Dying, 2nd edn. (Baltimore: Johns Hopkins University Press). (2001), ‘Tube-Feeding and Advanced Progressive Dementia’, Hastings Center Report, 31/1: 36–42.

the aging society

323

et al. (1997), ‘The Clinical Introduction of Genetic Testing for Alzheimer Disease: An Ethical Perspective’, JAMA 277/10: 832–6. Rudman, S. (1997), Concepts of Persons and Christian Ethics (Cambridge: Cambridge University Press). Sabat, S. R. (2001), The Experience of Alzheimer’s Disease: Life Through a Tangled Veil (Oxford: Blackwell). Singer, P. (1993), Practical Ethics (Cambridge: Cambridge University Press). Stock, G. (2002), Redesigning Humans: Our Inevitable Genetic Future (Boston: Houghton Mifflin). US General Accounting Office (1998), Alzheimer’s Disease: Estimates of Prevalence in the U.S., HEHS-98–16 (Washington, DC: GAO, 28 Jan.). Volicer, L., and Hurley, A. (eds.) (1998), Hospice Care for Patients with Advanced Progressive Dementia (New York: Springer).

chapter 14 •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

DEATH IS A PUNCH IN THE JAW: LIFE-EXTENSION AND ITS DISCONTENTS •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

felicia nimue ackerman

My hospice room is rose and blue. The blue is like the sky. They think that if you’re happy here, You’ll be content to die. They proffer comfort, warmth, and peace, All shining like the sun. They strive to meet your every need. They meet all needs but one. So now I have another scheme, I thank James Dreier, Sara Ann Ketchum, Christine Overall, and Bonnie Steinbock for extremely helpful discussions; James Dreier, David Estlund, and Bonnie Steinbock for extremely useful advice about background reading; Antonio Ramirez for extremely adept research assistance; and Zelda Sondack Ackerman, soon to embark on her second century, for extremely sharp-eyed proofreading.

death is a punch in the jaw

325

My object all sublime. I’ve gotten on a transplant list, And so I bide my time. (Ackerman 2005b)1

Graduate students in my department run a weblog that has solicited suggestions for ‘the worst philosophical position ever’ (Ichikawa 2004). When I suggested the position that ‘illness and mortality are good things [and] it is bad to do research on extending the human life span’,2 Jonathan Ichikawa wrote, ‘Do philosophers really hold [that view]? I find that so absurd that I have a hard time imagining anyone seriously thinking otherwise.’ I replied by citing the most powerful bioethicist in America, Leon Kass, then head of the President’s Council on Bioethics, whom the New York Times Book Review described as ‘among the most respected authorities in the field’ (Nuland 2002: 49).3 Kass has said that ‘the finitude of human life is a blessing for every human individual, whether he knows it or not’ (Kass 2002: 264). He also opposes attempts to extend the maximum human lifespan, i.e. ‘the outer limit of human longevity’ (Overall 2003: 7), which ‘has remained more or less fixed at about 120’ (Overall 2003: 10). This chapter will argue that Ichikawa’s common-sense intuition is basically right. The achievement of greatly extended human life and even (or especially) immortality would be like the discovery of electricity. It would bring many problems and dislocations and would even do some people more harm than good, but overall it would be an enormous boon to humanity. Like other secular philosophers who write on this topic, I will assume there is no afterlife and hence that death is ‘the unequivocal and permanent end of our existence’ (Nagel 1979: 1). Except where indicated, I use ‘immortality’ to mean earthly immortality rather than immortality in a religious sense. This assumption accords with what I believe, but my reason for making it here is to delimit the scope of this chapter. It is hardly surprising that, after seeing ‘a part of the adventures of the [Holy Grail]’, Sir Galahad ‘fell in his prayer long time to Our Lord, that at what time he asked, that he should pass out of this world’, because of the ‘great I dedicate this chapter to my mother, Zelda Sondack Ackerman (b. 1912), in anticipation of a time when nonagenarians will have most of their lives ahead of them. The chapter title comes from the following letter of mine in the Arts and Leisure Section of the New York Times (Ackerman 2002). ‘William Hurt, a star of the film ‘‘Tuck Everlasting’’, is quoted as saying, ‘‘Most people think of life as the story of life and death. Death is a bookend. If you take that bookend away, what have you got?’’ This is typical of attempts to sugar-coat death. Here’s an alternative metaphor: Death is a punch in the jaw. Take away the punch and what have you got? An [unpunched] jaw.’ 1 Also see ‘Flourish Your Heart in This World’ (Ackerman 1998b) and n. 10 below. ( This poem first appeared in Ragged Edge Online and is reprinted by permission.) 2 I actually wrote ‘so’ rather than ‘and’, but that was a poor formulation. 3 Many assessments of Kass are more critical, however. For example, see Mooney (2001) and McGee (2003). Even Nuland’s largely admiring review expresses reservations, acknowledging that ‘most [readers] will . . . recognize the need for [more] flexibility’ (Nuland 2002: 49) than Kass offers.

326

felicia nimue ackerman

joy [he anticipated at seeing] the Blessed Trinity every day, and the Majesty of Our Lord’ in the afterlife (Malory 1969: 367, 368). Admittedly, belief in an afterlife does not always bring willingness to leave earthly life. Believers may be unwilling to leave their familiar lives and loved ones, or they may fear ending up in hell. But when believers and unbelievers talk about death, they are talking about two entirely different things.4 Richard Momeyer points out that views that do ‘not regard death as final, as the complete cessation of consciousness and personhood’ should be ‘regarded as but sophisticated forms of death denial’ (Momeyer 1988: 8). Amusingly illustrating this point is the career of Elisabeth K¨ubler-Ross, who began by combating the ‘denial of death’ and finished by denying that death ends one’s conscious existence.5 Why is death bad? What Nagel calls the ‘natural’ answer is that ‘death is an evil [not] because of its positive features, but only because of what it deprives us of’ (Nagel 1979: 1), i.e. ‘all possibility of satisfying experience’ (Momeyer 1988: 50). Why, then, would anyone object to greatly extended human life? Such objections fall into two categories.6 Objections in the first category hold that greatly extended life would harm the person living it. Objections in the second category hold that greatly extended human life would bring some other sort of harm; for example, harm to society or to the human species. I will consider these types of objection in turn. My discussion will deal both with greatly extended finite life and with immortality and will use the term ‘greatly extended life’ to cover both. Except where indicated, I will proceed from some assumptions adapted from Christine Overall (2003: 96, 128 ff.). First, people would know the life expectancy in their society or would know (or at least have good reason to believe) that they were immortal. Second, everyone would have the opportunity to choose greatly extended life. Third, greatly extended life would not be mandatory; people would be able to opt out at any point.

Would Greatly Extended Life Benefit or Harm the Person Liv ing It? ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

If there is one position on which secular bioethicists approach unanimity, it is that life is worth extending only if the quality is satisfactory. As David Gems puts it, ‘Most of us would agree that it makes little sense to postpone death once our quality of life has diminished beyond a certain point’ (2003a: 34). This seemingly unproblematic statement covers up some crucial questions. What is that point? Who decides? The conventional wisdom that we should avoid ‘endlessly prolonging 4 6

5 See Overall (2003: 178–9). See Momeyer (1988: 9) and Rosen (2004). See Singer (1991); Overall (2003, ch. 1); President’s Council on Bioethics (2003a, ch. 4).

death is a punch in the jaw

327

the morbid phase of our lives’ (Gems 2003a: 34) should be assessed in light of the fact that many seriously ill people find the ‘morbid phase’ of their lives well worth prolonging. ‘Study after study has concluded that many healthy people often say they would never want to die in an intensive care unit, but may change their minds when they are very sick’ (Kolata 1997). One study found that ‘The majority of [a group of 244 patients with various serious illnesses] wanted to receive a life-sustaining treatment if it would prolong life for any length of time’ (Danis et al. 1996: 1813). Another study, emphasizing the disparity of values between the healthy and the ill, found that ‘Patients with cancer are much more likely to opt for radical treatment [a grueling course of chemotherapy] with minimal chance of [cure or if it offers as little as three months of life-prolongation] than people who do not have cancer, including medical and nursing professionals’ and also including healthy controls matched with patients for ‘age and sex, as well as for occupation and hence educational and social class’ (Slevin et al. 1990: 1458, 1460), i.e. a control group selected to represent the attitudes of the patients before they had cancer. And a study of patients with end-stage kidney disease concluded that ‘chronically ill people cherish their lives, despite the appearance [to outsiders] of inadequate quality’ (Gutman 1988: 899). Here it is crucial to recognize that expertise lies with the patients. They have first-hand knowledge of what it is like to be seriously ill. So their desire to extend the ‘morbid phase’ of their lives seems prima facie rational. This point will resurface in my discussion of resource allocation. For the present, note that even Overall—who perceptively warns against assuming that ‘life with chronic health problems or disabilities is not worth living’, and stresses that decisions about ‘the minimum conditions of quality of life that make existence worth sustaining’ should be made by competent patients, ‘not their families and not their health-care providers’—says, ‘If the available medical, social, and material resources are so limited that one’s extra stretch of time is lived in misery, then that is a problem, and no prolongevist advocates the extension of human life if it can be lived only in poverty, pain, and ill health’ (2003: 188, 191, 192, 41; italics added).7 She is mistaken. I advocate this (for people who desire it). Of course, I believe that wealth, comfort, and good health are better than poverty, pain, and ill health. But the blanket presumption that the latter states bring ‘misery’ that is worse than death is disrespectful to those who, having experienced them, disagree. (Perhaps they disagree because they have found that these states, far from entailing living in misery, do not preclude such pleasures as love, friendship, television, music, reading, and thinking.) Thus, the popular phrase ‘compression of morbidity’, i.e. ‘reduction of . . . the end-of-life morbidity characteristic of old age’ 7 Elsewhere in her book she offers the more moderate formulation ‘Prolongevists . . . advocate the extension of life only if it will not involve the severe illness, unrelieved pain, or disablement that undermine an individual’s capacity to pursue his or her life’s projects’ (Overall 2003: 65–6; italics added). This formulation is more respectful of individual differences in values. Also see Overall (2003: 96).

328

felicia nimue ackerman

(Battin 1987: 328), glosses over the difference between curing a sick person and denying him treatment that would extend his ‘morbid’ life.8 And Overall’s claim that ‘medical treatments that . . . extend an end-state of suffering, extreme dementia, or relentless deterioration must be recognized as pointless or even, at worst, as harmful’ (2003: 190; italics added) conflicts with her above-quoted views about patient self-determination, given that some competent people choose ‘relentless deterioration’ over death. This consideration is particularly important in view of some ideas behind the conventional wisdom about when life is worth prolonging. Thus, M. Pabst Battin identifies ‘increasing debility, dependence . . . loss of communication and affection . . . isolation’ as part of an ‘unfortunately realistic picture of old age’ constituting ‘human degradation’ that can lead to a sick old person’s ‘increasingly poor self-image’ and desire to die (1980: 274). By contrast, other philosophers, such as Susan Wendell, question independence as an ideal, pointing out that no one is completely independent and that ‘ ‘‘independence’’ . . . is defined according to a society’s expectations about what people ‘‘normally’’ do for themselves and how they do it’ (Wendell 1996: 145). Overall says, ‘There are varying patterns of dependence throughout the life span’ and ‘to the extent that [old people] are dependent, their dependence should not be interpreted as evidence that increased longevity is bad’ (2003: 188). Such views are more logical as well as more humane than the view that ‘dependence’ is characteristic of only some people’s lives or is inherently degrading. Three more points are worth noting here. First, the ‘increasingly poor self-image’ of the old, ill, and disabled, like the poor self-image some blacks may have in a racist society, arises ‘from the low value . . . society places on people of a certain sort, rather than from anything inherent in the person’s condition’ (Ackerman 1998a: 151–2). Second, present-day society over a quarter-century after Battin’s article appeared, offers the old, ill, and disabled a new opportunity to bypass ageism and ableism and to overcome isolation ‘in a better world, where there is neither young nor old, healthy nor crippled . . . On the Internet’ (Ackerman 2005c: 43)9 or in online relationships with people who have chosen to identify themselves online as old, ill, and/or disabled. Third, I am not claiming that all sick people want to extend their lives even if they cannot be cured, just that many bioethicists have shortchanged those who do. Overgeneralization is hardly rare in discussions of whether greatly extended life would benefit the person living it. For example, Kass, whose opposition to attempts to extend the maximum human lifespan rests partly on his ad hominem ‘For most of us . . . the desire to prolong the lifespan (even modestly) must be seen as expressing a desire never to grow old and die’,10 suggests the following ‘virtues of mortality’: 8

See the discussion in Battin (1987: 328–36). This reference is to a short story about a crippled old woman who does just that. 10 See also President’s Council on Bioethics (2003a: 162, 186). For a dissenting view, see Blackburn and Rowley (2004: 0420). (Of course, I also reject Kass’s view that there is something wrong with a 9

death is a punch in the jaw

329

interest and engagement, seriousness and aspiration, beauty and love, and virtue and moral excellence (Kass 2002: 263, 266–9).11 I will discuss these in turn. As to the first, Kass offers mainly not arguments but rhetorical questions designed to suggest that interest and engagement would decline as life proceeded endlessly. The most plausible answers, however, are far from what he supposes. Thus, his question ‘Would the Don Juans of our world feel better for having seduced 1,250 women rather than 1,000?’ (Kass 2002: 266) receives this apt reply from Gems: ‘on his deathbed, the extra 250 women might not seem that important to Don Juan. But if he was at only number 1,000 and was offered an extra 250, one can only imagine his joy’ (2003a: 35). Kass’s view that greatly extended life would bring diminishing interest and engagement pays scant attention to individual differences. The same problem is evident in his famous anti-cloning slogan ‘Shallow are the souls that have forgotten how to shudder’ (Kass 2002: 150)—a slogan that ‘invites this reply: Narrow are the souls that cannot comprehend that different souls shudder at different things. I shudder at the thought of being denied lifesaving technology because it violates Kass’s particular conception of dignity’ (Ackerman 2005a).12 Kass’s scant respect for individual differences is ironic in view of his criticism of bioethics as ‘