PLACE IN RErURN Box to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. ' DATE DUE DATE DUE DATE DUE arm clelRC/DateDmpBS-sz MORAL REASONING AND JUSTIFICATION IN MEDICAL ETHICS By Kyungsuk Choi A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Philosophy 2003 ABSTRACT MORAL REASONING AND JUSTIFICATION IN MEDICAL ETHICS By Kyungsuk Choi The method of wide reflective equilibrium (WRE) seeks coherence among our considered moral judgments, moral rules and principles, and relevant background beliefs or theories. I argue that WRE is a more plausible account of moral reasoning than either principle-based or case—based approaches. Two objections are commonly leveled against WRE, both of which can be answered. First, it responds to the problem of choosing between competing coherent belief systems by construing coherence as a matter of degree and by adding comprehensiveness as an additional criterion for evaluating belief systems. Second, it avoids the charge of making no contact with the world by incorporating moral experience. Because particular moral judgments made in the light of moral experience are often affective judgments, however, I stress that they should be appropriately “filtered.” WRE does not imply subjective, cultural, moral, or conceptual relativism. However, descriptive cultural relativism and descriptive conceptual relativism are compatible with reasonable pluralism — an important background belief of WRE —— which means that a variety of reasonable comprehensive philosophical or religious doctrines can coexist. But this does not imply normative conceptual relativism because WRE need not commit to the view that justified moral beliefs are necessarily true. In addition, WRE accepts the difference between moral beliefs justified in a particular reasonable belief system and moral beliefs justified in every reasonable belief system, that is, an overlapping consensus. Since certain fundamental moral beliefs will be supported by an overlapping consensus, these beliefs are not relative (in an objectionable sense) to a particular reasonable comprehensive belief system. I develop WRE into a group method of moral reasoning. Overlapping consensus provides beliefs mutually accepted by all the group members and establishes a group belief system. This system can be revised or broadened by complete consensus. Group WRE contributes to the reaching of those consensuses and, more importantly, guides a group in the search for integrity-preserving compromise. Complete consensus sometimes involves implicit compromise, which I distinguish from explicit compromise. The process of compromise requires the exercise of judgment, which is also required in revising beliefs in WRE. I stress that judgment is an inevitable part of practical moral reasoning. Judgment must be exercised when we make decisions using abstract and general concepts in unprecedented or unpredictable situations. Judgments are made in the interaction between universal and particular knowledge, and require justification. Finally, I illustrate the use of group WRE by dealing with the issues of voluntary active euthanasia and physician-assisted suicide. I distinguish their moral permissibility fi‘om their legalization. Using WRE, I then argue that there are situations in which both types of euthanasia are morally permissible. However, there might remain reasonable disagreement regarding their legalization. Using group WRE, I show that reasonable parties can reach a compromise: Patient Refusal of Hydration and Nutrition (PRHN) can be developed as a practice to assist patients who have decided to hasten their own death. For my teachers who taught me philosophy and made my life stimulating and enjoyable as a student of philosophy iv ACKNOWLEDGMENTS I am indebted to my advisor, Professor Martin Benjamin, who held weekly dissertation meetings for over a year and spent countless hours reviewing and commenting on drafts of each chapter. My project on wide reflective equilibrium could not even have begun without his Central Issues in Ethics course, which I took early in my doctoral program. I also took other courses from him, assisted in his medical ethics course, and had many valuable discussions with him, all of which were philosophically instructive and showed me what kind of moral philosopher I should strive to become. I am also grateful to Professors Tom Tomlinson, James L. Nelson, and Daniel Steel, who provided valuable comments on my dissertation. In addition, I am indebted to Professor Emeritus Richard J. Hall with whom I worked on an independent study centering on epistemological justification. I would also like to thank Professor Byung-Nam Oh at Seoul National University, who was my undergraduate and M.A. thesis adviser. He led me to the study of philosophy and has provided encouragement all the way. He taught and modeled a conception of a good scholarship and teaching that I should like to follow. He has been of great help to me as a student of philosophy. I am grateful for four years of teaching assistantships from the Department of Philosophy at Michigan State University and a Dissertation Completion Fellowship from the Graduate School, which enabled me to finish my doctoral study. In addition, I received enormous editorial help from fellow graduate student, Jason Matzke, in the later stages of writing. Finally, I am deeply grateful to my parents who have long supported my studies in various ways. I am particularly indebted to my wife, So Young Lee. Without her support, I could not have devoted so much time and effort to this dissertation. vi TABLE OF CONTENTS Introduction Chapter 1 Principlism, Case-Based Reasoning, and Wide Reflective Equilibrium 1. Three Methods of Moral Reasoning and Theories of Justification 1.1. Various Methods of Moral Reasoning in Medical Ethics 1.2. Principlism, Case-Based Reasoning, the Method of WRE and Their Relationship with Theories of Justification 2. Challenges to Principlism and Case-Based Reasoning 2.1. Three Desiderata for a Method of Moral Reasoning 2.2. Difficulties of Principlism 2.3. Difficulties of Case-Based Reasoning 3. The Method of WRE 3.1. Coherentism as a Model of Moral Justification 3.2. The Method of WRE and Its Characteristics 3.3. The Three Desiderata and the Method of WRE Chapter 2 Reflective Equilibrium and Foundationalism 1. The Method of Reflective Equilibrium and Foundationalism 1.1. Hare’s and Singer’s Claim 1.2. Daniels’s Defense 2. DePaul’s Defense and His Claim of Compatibility vii ll 12 12 14 l9 19 22 25 28 28 32 36 4O 4O 4O 43 47 2.1. DePaul’s Defense from the Charge of Foundationalism 2.2. DePaul’s Claim of Compatibility 2.3. No Compatibility 3. The Method of Reflective Equilibrium and Modest Foundationalism 3.1. Modest F oundationalism 3.2. Ebertz’s Claim 3.3. The Important Difference Between the Method of Reflective Equilibrium and the Modest Foundationalism Chapter 3 Moral Experience and Wide Reflective Equilibrium 1. The problem of Choosing the Better Belief System 1.1. Mutual Support and Circularity 1.2. The Problem of the Choice Among Coherent Belief Systems 1.3. Comprehensiveness as an Additional Criterion 2. The Problem of No Input from the World 2.1. Naive Inquirer’s Problem and Moral Experience 2.2. Emotion and Judgments 3. The method of WRE as an Interpersonal Epistemic Activity 3.1. Reasoning as an Interpersonal Epistemic Activity 3.2. An Additional Criterion for Choosing Between Competing Belief Systems 3.3. Implications and Limitations of the Three Criteria viii 47 54 57 61 64 65 69 70 7O 73 75 78 78 81 85 85 89 91 Chapter 4 Wide Reflective Equilibrium and Relativism 1. Subjective Relativism and Cultural Relativism 1 .1 . Subjective Relativism 1.2. Cultural Relativism 2. Moral Relativism and Conceptual Relativism 2.1. Moral Relativism 2.2. Conceptual Relativism 2.3. Conceptual Relativism Revisited 3. Reasonable Pluralism and Conceptual Relativism 3.1. Reasonable Persons and the Burdens of Judgment 3.2. Reasonable Pluralism and Implication of Toleration 3.3. Moral Justification and Morality Chapter 5 The Group Method of Wide Reflective Equilibrium 1. The Group Method of WRE 1.1. The Method of WRE as Group Reasoning 1.2. The Reasonable for Group Reasoning 1.3. Group Belief and Group Justification 2. The Group Method of WRE and Consensus 2.1. Overlapping Consensus 2.2. Complete Consensus 3. Compromise and Integrity in the Group Method of WRE ix 95 96 96 97 101 101 104 107 110 110 114 116 121 122 122 126 127 132 132 135 141 3.1. Discrepancies Between Group and Personal Beliefs 3.2. Implicit and Explicit Compromise 3.3. The Group Method of WRE and Personal Integrity Chapter 6 Wide Reflective Equilibrium and Judgment 1. Moral Reasoning and Judgment 1.1. Judgments in the Methods of WRE 1.2. Calling for Judgment 2. The Exercise of Judgment 2.1. Judgments in Principlism and Casuistry 2.2. Judgments: Discussible Decisions 2.3. Indispensable Principles and Values 3. Justification of Judgments and a Web of Beliefs 3.1. Rules of Thumb for the Use of the Methods of WRE 3.2. A Web of Beliefs Chapter 7 Wide Reflective Equilibrium and the Issues of Euthanasia and Physician-Assisted Suicide 1. The Moral Pennissibility of VAE and PAS 1.1. Background: Conceptual Distinctions Among Euthanasia 1.2. The Standard Method of WRE and the Issue of the Moral Permissibility of VAE and PAS 2. The Group Method of WRE and Legalization of VAE and PAS 141 144 148 151 153 153 154 158 158 164 166 168 168 172 176 177 177 180 192 2.1. Background: The Loss of Equilibrium 2.2. Back-and-forth Discussion 2.3. The Possibility of Abuse 3. Compromise 3.1. Reasonable Disagreement: Different Judgments 3.2. A compromise Position: PRHN 3.3. Consideration of Possible Objections to PRHN Bibliography xi 192 194 198 201 201 204 207 209 INTRODUCTION When we face moral questions such as “What course of action should be taken in this situation?” or more specifically “Should physician-assisted suicide be allowed?” we try to reach a conclusion or a decision. But, how should we as moral reasoners arrive at conclusions? Moral reasoning is important in two ways. First, it is used to evaluate moral beliefs, which vary according to generality — they can be at the levels of ethical theory, moral principles or rules, or particular moral judgments. Second, moral reasoning is employed to provide answers about what we should do, and as such is a form of practical reasoning. In addition to the above functions, we need a method of moral reasoning that will satisfy the following three desiderata. First, our moral reasoning should be able to cope with new ethical issues. As medical knowledge and technology have advanced, we have, for example, been faced with new ethical issues related to physician-assisted suicide, organ transplantation, In Vitro fertilization, and genetic therapy/engineering. Answers to these issues are elusive because the problems are so novel. However, we must often make decisions. Second, various cultures, religions, ways of life, and value systems coexist within a single society. Our society has become less homogeneous and more pluralistic. We can see disagreement on ethical issues even as we work to overcome prejudice, selfishness, and bad reasoning. Nonetheless, we must reach agreement on certain matters because morality is for the most part social and not private. Thus, we need a method of moral reasoning that can resolve certain questions in a pluralistic society. Third, many ethical issues are problems for society as well as individuals. We need to, for example, develop health care policy. Thus, moral reasoning must be able to be used by a group like an ethics committee or commission. There are three competing methods of moral reasoning: principle-based reasoning (or principlism), case-based reasoning (or casuistry), and the method of wide reflective equilibrium (a coherentist method of moral reasoning). I argue that the method of wide reflective equilibrium (WRE) satisfies the foregoing desiderata better than the other two methods. The term ‘reflective equilibrium’ was first used by John Rawls. Norman Daniels clarified the notion by emphasizing the difference between wide and narrow reflective equilibrium (NRE). NRE pursues coherence between moral principles and considered moral judgments while WRE includes relevant background beliefs or theories as a third set of important considerations. The resulting coherence achieved by the method of WRE is also called WRE. Thus, WRE (or NRE) refers both to a method as a process and to the resulting state of beliefs in equilibrium. I will explicitly refer to either the method Or the state of WRE only when a potential confusion makes it necessary to do so. Rawls originally used the method of WRE to justify the choice of theories of justice and Daniels used it to justify the choice of ethical theories. Tom L. Beauchamp and James F. Childress have introduced it as a method of moral reasoning aimed at moral justification of moral judgments in medical ethics. In Chapter One, I discuss the other two methods of moral reasoning (principlism and case-based reasoning) and argue that the method of WRE is more plausible than either of them. In Chapters, Two through Four, I clarify and characterize the method of WRE by responding to objections to coherentism before finally developing it as a group method of moral reasoning in Chapter Five. I will call this the group method of WRE to distinguish it from the standard method. Considering some practical questions related to the use of both methods of WRE, I claim in Chapter Six that practical moral reasoning cannot be done without the exercise of judgment. To illustrate the both methods, in Chapter Seven, I use the standard method of WRE to arrive at answers to debates over voluntary active euthanasia (VAE) and physician-assisted suicide (PAS), and I use the group method of WRE to demonstrate how we can arrive at collective answers regarding these debates. Tom L. Beauchamp and James F. Childress, in their Principles of Biomedical Ethics, and John D. Arras and Bonnie Steinbock, in their Ethical Issues in Modern Medicine, consider principle-based reasoning, case-based reasoning, and coherentist reasoning (the method of WRE) as competing methods. Principlism, which is ofien characterized as a “top-down approach” or “deductivism,” maintains that we can arrive at solutions to individual ethical issues by deducing them from moral principles. Case-based reasoning, which is often called “casuistry,” obtains particular moral judgments from careful review of particular cases or analogies among particular cases. By reminding us of the importance of looking at particular cases and of the important role particular moral judgments play in moral reasoning, this method has contributed to the discussion of moral reasoning. Case-based reasoning is often called inductivism because a particular moral judgment is obtained by induction from similar cases without direct appeal to moral principles. Principlism, however, faces the difficulty of not being able to handle conflicts between moral principles or conflicts between what a moral principle and an intuitive moral judgment recommend. This method also ignores the difficulties of metaphysical controversies, which should be resolved before principlists appeal to a particular moral principle. In addition, this method retains the problematic one-directional structure of moving from moral principles to particular moral judgments, which is why I see principlism as a form of foundationalist moral reasoning. Case-based reasoning faces similar difficulties even though it has been offered as an alternative to principlism. Case-based reasoning is unable to handle conflicts between particular moral judgments. Moreover, identifying similarities and dissimilarities among particular cases requires generalizations such as moral principles or moral theories as well as an overarching perspective by which to locate relevant similarities and differences. Because case-based reasoning is also a foundationalist method, it too ignores bi-directional relationships between moral principles and particular moral judgments. The method of WRE avoids the above problems. It not only rejects one- directional relationships between moral principles and particular moral judgments, but also aims at equilibrium among three elements: particular moral judgments, moral principles, and relevant background (including scientific and metaphysical) beliefs and theories. If there are any inconsistencies or conflicts among the three main elements, the method of WRE tries to eliminate them through a back-and-forth process of revision. No moral principles, particular judgments, or background beliefs are, in principle, immune to revision. Equilibrium obtained through a back-and-forth process of revision may be challenged when some of our beliefs change or when we encounter new circumstances. In these situations, we will have to make revisions to reach another state of WRE. The method of WRE should, then, be understood as an ongoing process. Interestingly, however, there are challenges to construing the method of WRE as a coherentist method of moral reasoning. R. M. Hare and Peter Singer, for example, claim that the method of reflective equilibrium is committed to intuitionism. Norman Daniels and Michael DePaul take this to imply a commitment to foundationalism, and have proceeded to argue against it. Daniels claims that WRE is not committed to foundationalism, though NRE is. DePaul argues that WRE is compatible with foundationalism. By focusing on the similarities between prima facie foundational beliefs and the role of considered moral judgments in WRE, Roger P. Ebertz claims that WRE is not actually a coherentist model of justification, but can be understood as a modest foundationalism. Thus, the issue is whether considered moral judgments have epistemic privilege in the justification of moral beliefs. Any view that sees considered moral judgments as enjoying special justificatory status, I argue, is based on an incomplete understanding of considered moral judgments and WRE. In fact, neither WRE nor NRE is committed to, or compatible with, foundationalism. When WRE is understood as a coherentist method of moral reasoning, it must confront two objections to coherentism: (l) the problem of choosing between competing coherent belief systems or choosing between conflicting moral beliefs belonging to different coherent belief systems, and (2) the problem of a coherent ethical framework’s having no contact with the world. Responding to the first objection, I emphasize that coherence should be construed as a matter of degree, and hence that moral justification should be as well. And, in addition to coherence, comprehensiveness should be adopted as a positive criterion for the justification of beliefs and for the evaluation of belief systems. I introduce the number of incorrect moral or non-moral beliefs as a third, but negative criterion. Incorrect beliefs refer to beliefs that are not justified in a more comprehensive and coherent actual or “working” belief system, though they may be justified within a less comprehensive and coherent belief system. Thus, although a moral belief might achieve coherence within one’s narrow belief system, it can be said to be a morally incorrect belief in light of a more comprehensive and coherent working belief system. Coherence, therefore, is neither the sole criterion of moral justification nor the sole criterion for choosing among belief systems. More importantly, I do not value coherence at any cost. Responding to the second objection, I acknowledge that moral experience is necessary for moral reasoning because a naive inquirer who lacks moral experience can maintain the coherence of a defective belief system without revising any of the incorrect moral beliefs or broadening the belief system. The method of WRE can incorporate moral experience in its framework. Particular moral judgments belonging to a larger set of considered moral judgments play the role of receiving input from one’s moral experience because they are judgments of whether particular actions are morally right or wrong. However, particular moral judgments may sometimes be made be infused with emotions or feelings, which will sometimes distort judgment. Aflective particular moral judgments can be made. The method of WRE must, however, filter initial affective particular moral judgments. But this does not mean filtering out emotions, but rather, limiting unconsidered or snap judgments. Thus, the method of WRE does not ignore the roles of moral experience, emotions, or feelings in our moral reasoning. Although moral experience plays an important role in providing an opportunity to revise or broaden a belief system, so too does the experience of reasoning. Reasoning has an interpersonal, social aspect. Constructing and evaluating arguments are not merely carried out within an individual person, but in interaction with others through actual or hypothetical interpersonal epistemic activities. By interpersonal epistemic activities, I mean presenting one’s reasoning, considering others’ views, and responding to the strongest objections. Thus, one’s belief system must be revised or broadened by both moral and reasoning experience. In this respect, I construe the method of WRE as an interpersonal epistemic activity. However, there often exist conflicting moral beliefs, each of which achieves equilibrium in its own reasonable comprehensive coherent belief system. Does this possibility imply that the method of WRE leads to relativism? More importantly, does WRE lead to an objectionable form of relativism? I subdivide ethical relativism into subjective, cultural, moral, and conceptual relativism and consider whether the method of WRE leads to any of these. I conclude that it does not. As versions of descriptive relativism, however, cultural and conceptual relativism may be compatible with a variety of reasonable comprehensive philosophical or religious doctrines, which Rawls calls reasonable pluralism. The method of WRE includes reasonable pluralism as an important background belief, but this inclusion does not imply normative conceptual relativism. To imply normative conceptual relativism, the method of WRE must assume that justified moral beliefs are identical with true moral beliefs. But WRE need not be committed to such a metaphysical view. More importantly, those who follow the method of WRE must admit that there is a difference between two kinds of justified moral beliefs. One is a moral belief that is justified in a particular person’s reasonable comprehensive belief system. The other is a moral belief that is justified in all reasonable comprehensive belief systems. The latter are part of what Rawls calls an overlapping consensus. Thus, the method of WRE will not claim that morality is relative only to a particular reasonable comprehensive coherent belief system. Some moral beliefs, such as the belief that murder is wrong, will be parts of all reasonable comprehensive coherent belief systems. Overlapping consensus has been seen in medical ethics as shown in agreement on the middle level of moral principles such as autonomy, utility, beneficence, nonmaleficence, and justice -— even though the justification of these principles may differ among people according to their ethical theories or belief systems. Overlapping consensus provides beliefs that are mutually accepted by all the members of a group. With this in mind, I develop the method of WRE into a group method of moral reasoning for group decisions, such as those needed for public policy-making. The development of a method of group reasoning is urgent when we consider the variety of very difficult ethical issues in heath care. Overlapping consensus will provide some beliefs that are mutually shared by all the group members. The group method of WRE attempts to reach coherence among the beliefs accepted by all of the members. The group method presupposes that all members value mutual respect, which is the moral virtue of a reasonable person. Group beliefs will establish a group belief system. To allow the group method of WRE to be used by those who follow the other methods, however, I must neutrally define the justification of group moral beliefs. A moral belief is justified when all the members of a group mutually recognize that it is either justified in all reasonable belief systems (though it may be justified in various ways) or there is a substantial reason for everyone to support it. Overlapping consensus is one of the ways to produce group beliefs. A second way is complete consensus, which means that a group has reached consensus on both a belief and its supporting reasons. The group belief system will be revised or broadened by complete consensus as well as overlapping consensus. The process of the group method will consist of actual interpersonal epistemic activities such as give-and-take discussions of a given issue and the members’ deciding upon group beliefs and their justification. This process, which is basically the same as the revision process of the standard method of WRE, aims at reasonable, informed, and uncoerced agreement that finally coheres with other beliefs of the group belief system. The group method of WRE contributes to the achievement of overlapping consensus and complete consensus. More importantly, the group method guides how a group can arrive at an integrity-preserving compromise. Although complete consensus is one of the most desirable types of agreement, it is hard to reach in a pluralistic society. Furthermore, even complete consensus involves implicit compromise, which I distinguish from explicit compromise. The group method of WRE provides some requirements for a well-grounded and integrity-preserving compromise. But the process of reaching compromise and of revising beliefs, whether personal or group beliefs, requires the exercise of judgment. Judgment is inevitable in practical moral reasoning. When we must make decisions using abstract concepts in unprecedented or unpredictable situations, we cannot help but call for judgment. Judgments made in the interaction between universal and particular knowledge cannot be algorithmic because of the abstractness and generality that are inherent to the relevant concepts, values and principles. But this does not mean that judgment is arbitrary. Despite the indeterminacy of principles and values, they guide and regulate the use of judgment, and provide reasons for judgments. Judgments thus are not arbitrary preferences but should be justified by relevant reasoning. Finally, I illustrate the standard and group methods of WRE by dealing with the issues of voluntary active euthanasia and physician-assisted suicide. 1 distinguish the issue of their moral permissibility from that of their legalization. Using the standard method of WRE, I demonstrate how I reach my answer that there are situations in which both types of physician-assisted death are morally permissible. Nonetheless, there is reasonable disagreement as to whether to legalize them. Using the group method of WRE, I show that reasonable parties can reach compromise on this matter. One possibility is to refine Patient Refusal of Hydration and Nutrition as a practice used to help patients who have decided to hasten their own death. 10 Chapter One Principlism, Case-Based Reasoning, and Wide Reflective Equilibrium We are now faced with a variety of complex ethical issues related to health care. Resolutions are elusive not only because many of the issues are unprecedented, but also because the diversity of values, ways of life, religions, and worldviews in a liberal society makes it difficult to reach consensus. But consensus is important because many ethical issues in health care are problems of social ethics as well as personal ones. Many issues then are related to health care policy-making. We therefore need a method of moral reasoning that is: (I) able to cope with new ethical issues generated by the development of medical knowledge and technology, (2) able to work in a pluralistic society, and (3) able to facilitate health care policy-making. Tom L. Beauchamp and James F. Childress in their Principles of Biomedical Ethics, and John D. Arras and Bonnie Steinbock in their Ethical Issues in Modern Medicine consider the following three methods of moral reasoning as competing methods in medical ethics; principlism, case-based reasoning, and the method of wide reflective equilibrium (WRE).I Different terms are often used to refer to each method. Principlism is also called the ‘top-down approach,’ ‘principle-based reasoning,’ or ‘deductivism’; case-based reasoning is often called the ‘bottom-up approach,’ ‘casuistry,’ ‘inductivism,’ or ‘particularism’;2 and the method of WRE is characterized as coherentist reasoning. In this chapter I argue that the method of WRE is more plausible than principlism ' Tom L. Beauchamp and James F. Childress, Principles of Biomedical Ethics, 4th ed. (Oxford University Press, 1994), 13-28. John D. Arras and Bonnie Steinbock, eds., Ethical Issues in Modern Medicine, 5‘” ed. (Mayfield Publishing Company, 1999), 34-40. 2 There are sometimes differences between casuistry and particularism. These differences are not important 11 and case-based reasoning. The method of WRE not only satisfies the foregoing desiderata, but for a method of moral reasoning also avoids the difficulties inherent in the other methods. 1. Three Methods of Moral Reasoning and Theories of Justification 1.1. Various Methods of Moral Reasoning in Medical Ethics To better see how the various methods of moral reasoning work, let us consider the following case used in a medical ethics course and compare five students’ answers to the question: Mat course of action should be taken, and why? Dr. Healer is in a situation where he wonders if he should tell the whole story of his patient’s diagnosis. The patient, Lisa, is 54 years old. She came to the emergency room two weeks ago because she had a severe pain and a mass in the right lower quadrant. The previous December the patient had had a severe attack in the same area. During a brief conversation with Lisa, Dr. Healer came to know that she was afraid she had cancer and was concerned about her condition because her daughter would be married within a month. After she had been examined, the diagnosis was made of a degenerating fibroid. She went to surgery. Finally, Dr. Healer found her having stage IV cancer of the cervix, the most advanced stage. He cleaned out all the tumor he could see. But the tumor had spread to the pelvic wall. All he could do was to try chemotherapy and radiation. The five-year survival rate of stage IV cancer is O to 20 percent. Chemotherapy and radiation has to begin so that she can live her life more or less normally for a short while.3 Dr. Healer is faced with the moral dilemma of whether or not to tell the truth to his patient. The answers given by the five students from the class represent different approaches to moral reasoning. Kelly argued that Dr. Healer should tell the truth because for my present purposes, but I briefly explain them on p.7 in this chapter. 3 This case is modified from that presented by Robert M. Veatch in his article ”The Dying Cancer Patient,” 12 telling the truth maintains patient autonomy. In order for Lisa to make an autonomous decision about her future treatment and her life, she must know her diagnosis. In addition, if Dr. Healer does not disclose this information to her, he cannot properly obtain her informed consent for medical treatment. Adam argued that Dr. Healer should not tell Lisa the whole story of her condition because doing so in that situation would harm her. In particular, she would fear dying. In addition, if Dr. Healer discloses information about Lisa’s diagnosis, her daughter may not go through with her planned wedding. Her family, or at least her daughter, may blame the doctor for divulging her diagnosis at an inappropriate time. Telling only part of, rather than the whole story will bring the greatest happiness to both Lisa’s family and herself. A physician has a duty, after all, to do good rather than harm a patient. Anne agreed with Adam’s idea that a physician’s duty is to act with beneficence and not maleficence with respect to a patient. However, she argued that not telling Lisa the whole story harms her. If Dr. Healer does not disclose information about her diagnosis, her anxiety will be exaggerated because she had already been afraid that she might have cancer. If she knew her diagnosis, it could allow her to prepare for the possibility of death and use the remainder of her life to do what she wants before dying. Therefore, Anne argued that Dr. Healer should tell the full truth. Cathy argued that Dr. Healer should tell the whole story of Lisa’s diagnosis. She based this belief on her own experience. Her aunt had stage IV cancer of the cervix at the age of fifty-seven and she did want to know the true diagnosis. Presumably Lisa would also want to know. The only potentially relevant difference between the two cases was one of age. Cathy decided, however, that age should not influence our assessment of such in Ethical Issues in Modern Medicine, 51-52. 13 situations. In addition, she herself would want to know the truth if she were in the similar situation. Therefore, the doctor should tell the truth in Lisa’s case too. Wendy argued that the doctor should tell the truth because it maintains patient autonomy and because it fits her judgment about real-life cases she has experienced. For example, she thought it wrong that the physician did not tell her grandmother the truth that she was suffering from cancer. The consequences of not telling the truth may sometimes look less troubling than those of telling the truth. But, Wendy concluded, our obligations to respect a patient’s autonomy outweigh our concerns regarding consequences. In part this is because the physician and the patient may disagree as to the value of different ends or outcomes. In order to make decisions about her treatment and remaining life, Lisa must be given a full account of her diagnosis. Moreover, Wendy has heard that a high percentage of laypeople, between eighty-two and ninety-eight percent, said that they would like to be truthfully told their diagnosis even if they are in the last stages of cancer. Dr. Healer should, therefore, tell the truth. Most important for our concerns here is not the right or wrongness of each answer, but the reasons each student gave for his or her answer. We can see the various methods of moral reasoning employed by looking at the relationships between the conclusions and reasons in each case. 1.2. Principlism, Case-Based Reasoning, the Method of WRE and Their Relationship with Theories of Justification We can see that Kelly, Adam, and Anne appeal to moral principles like autonomy and beneficence (or utility). Cathy appeals to a particular moral judgment she formed 14 about a similar situation. Wendy appeals to the principle of autonomy, her particular moral judgment regarding her grandmother’s case, and considerations such as laypeople’s opinions. But what we should focus on when considering Wendy’s reasoning is the relationship among her particular moral judgment, the principles used, and her conclusion. For Wendy, the principle of autonomy and her particular moral judgment that her grandmother’s physician should have told her the truth are not the only sources of justification. If we ask five students why the principle, the particular moral judgment, or the intuition to which each of them appeals is justified, the differences in their answers will be apparent. Kelly, Adam, and Anne may believe that moral principles are self- evidently true or justified on other grounds. Therefore, they appeal to moral principles as in justifying their conclusions. Cathy may believe that her particular moral judgment is self-evidently true. Wendy may think that her conclusion is justified because it squares with sound moral principles, her particular moral judgment, and the survey she mentioned. She may think that her conclusion is justified by this coherence rather than by moral principles or her particular moral judgment alone. The method of moral reasoning represented by the first three students is called principlism,4 or principle-based reasoning.5 Kelly and Anne reach the same conclusion even though they appeal to different moral principles -— the principles of autonomy and utility, respectively. Adam and Anne reach different conclusions even though they both appeal to the principle of utility. All of them, though, appeal to some specific moral principle to justifying their conclusion. A particular moral judgment as a conclusion of ‘ Arras and Steinbock use the terms ‘principlism’ and ‘principles approach’ in Ethical Issues in Modern Medicine, 35—38. 5 Tom Tomlinson uses the expressions ‘principle-based method’ and ‘principle-based approach’ in “Casuistry In Medical Ethics: Rehabilitated, or Repeat Offender?" Theoretical Medicine 15 (1994): 5. 15 moral reasoning, such as “Dr. Healer should tell Lisa the truth,” they believe, must be based on a moral principle. In other words, the justification of a conclusion ultimately comes from a moral principle.6 In this respect, this method of moral reasoning is often called deductivism because conclusions are deduced from moral principles.7 This method of moral reasoning is implicit in the term ‘applied ethics,’ which is often used to characterize medical, business, and environmental ethics. This term suggests that solutions to moral issues are obtained by applying moral principles to them. The main feature of this method is that moral principles play a pivotal role in the moral reasoning. This is evident in Kelly’s, Adam’s, and Anne’s reasoning where the principle of autonomy or beneficence (or utility) serves as the basis, which does not in this context require further justification.8 The method represented by Cathy is called case-based reasoning or casuistry. Case-based reasoning has been suggested as an alternative to principlism in Albert R. Jonsen and Stephen Toulrnin’s The Abuse of Casuistry.9 In contrast to principlism, this method stresses the importance of looking at particular cases and the important role particular moral judgments play in our moral reasoning. This method is often called inductivism10 because conclusions are obtained by looking at similar individual cases, 6 In fact, the justification of a conclusion here is obtained from at least one premise including a moral principle and another premise that states facts, such as that “truth telling maintains a patient’s autonomy” or “telling the truth improves relevant people’s happiness.” But we can say that the first is the main premise because it contains the “ought” statement required for the justification of a conclusion also containing an “ought.” 7 Beauchamp and Childress use the term ‘deductivism’ for this type of moral reasoning in Principles of Biomedical Ethics, 14-17. Arras and Steinbock also say, “principlism is deductivistic” even though they do not use directly “deductivism” in Ethical Issues in Modern Medicine, 38. 8 I do not argue here that Kantianism or utilitarianism claims that the principle of autonomy or utility is the basic belief. They may present other principles or beliefs to justify these principles. 9 Albert R. Jonsen and Stephen Toulmr'n, The Abuse of Casuistry: A History of MoralReasoning (University of California Press, 1988). ‘0 Beauchamp and Childress use the term “inductivism” interchangeably with “the individual-case model.” See Principles of Biomedical Ethics, 17-20. 16 that is, by induction from similar cases. According to this method, we start with some agreed particular moral judgments regarding previous cases without appeal to moral principles. For example, the well-known Karen Quinlan and Nancy Cruzan cases involving patients in persistent vegetative state (total and permanent loss of consciousness) can serve as paradigm cases by which to evaluate similar situations. Case- based reasoning in ethics is thus similar to the common law tradition that emphasizes the comparative study of legal cases. 11 Therefore, the main feature of this method is that it emphasizes the particularity of individual cases rather than general moral principles. In this sense, it can be called particularism. Radical particularists such as Jonathan Dancy argue that each individual case is, in principle, unique. Reasons used to support a certain conclusion in one case can be used to deny the same conclusion in a different case.12 But casuists, who may be called moderate particularists, do not reject the use of some moral principles. They regard moral principles as summaries of particular moral judgments, that is, as what are inductively generalized from particular moral judgments. Therefore, moral principles are justified by particular moral judgments. Particular moral judgments related to individual cases play, for casuists, a pivotal role in justifying a conclusion. Moral principles can be used, but only to provide provisional guidance. The method represented by Wendy can be called coherentist reasoning.” More specifically, this method is now called the method of wide reflective equilibrium (WRE). H Arras and Steinbock, eds., Ethical Issues in Modern Medicine, 39. '2 Jonathan Dancy, “The Particularist’s Progress,” in Moral Particularism, edited by Brad Hooker and Margaret Little (Oxford University Press, 2000), 130-156. '3 Arras and Steinbock do not use any specific name to indicate this method of reasoning. But they say, “justification resides in the coherence or “fit” among the whole network of our considered judgments and the principles and rules that emanate from them.” See Ethical Issues in Modern Medicine, 40. Beauchamp and Childress deal with this method of reasoning under the title of coherentism. See Principles of 17 The notion of WRE was presented by John Rawls in characterizing the method of reasoning employed in his theory of justice. It was more fully developed by Norman Daniels for the evaluation of ethical theories, and has been focused on as a method of moral reasoning for biomedical ethics. The method of WRE seeks coherence among moral judgments, moral principles or rules, and relevant background beliefs or theories (the three main components of the framework of WRE). A particular moral judgment as a conclusion from moral reasoning, such as that Dr. Healer should tell Lisa the truth, should be justified based on the coherence among the three elements including the conclusion, rather than on moral principles or analogies with particular moral judgments in similar cases alone. The important feature of this method is that it does not give justificatory priority to any of the three elements. As the above considerations show, the three different methods of moral reasoning differ in terms of what enjoys justificatory priority. Principlism gives priority to moral principles while case-based reasoning gives it to particular moral judgments. The method of WRE does not give priority to either moral principles or particular moral judgments, but emphasizes their coherence. This difference among the three methods highlights theoretical differences among them in relation to theories of justification. Principlism presupposes that moral principles enjoy a justificatory priority over particular moral judgments. A particular moral judgment as a conclusion should be justified by moral principles. Principlism is, thus, committed to foundationalism. Case-based reasoning is also committed to foundationalism because this method presupposes that particular moral judgments obtained from careful review of a particular case or analogies among particular cases are Biomedical Ethics, 20-23. 18 self-evidently justified. '4 The method of WRE, unlike principlism and case-based reasoning, involves coherentist moral reasoning, as shown earlier. According to Laurence BonJ our, “the relation between the various particular beliefs is to be conceived, not as one of linear dependence, but rather as one of mutual or reciprocal support.”15 This is true of moral beliefs. Coherentist moral reasoning seeks judgments that will cohere with other moral and non-moral beliefs better than any alternative. The method of WRE sorts the relevant beliefs into the three kinds: Considered (particular) moral judgments, moral principles or rules, and non-moral relevant background beliefs or theories. Of course, coherentists acknowledge that inferential justification may sometimes take a linear appearance at a “local level of justification.”16 But inferential justification is essentially holistic at the global level of justification. In this way, the justification of a conclusion in moral reasoning comes essentially from its coherence with other beliefs, though it may seem to appeal to moral principles, like principlism, or to particular moral judgments, like case- based reasoning at local levels of justification. 2. Challenges to Principlism and Case-Based Reasoning 2.1. Three Desiderata for a Method of Moral Reasoning Before exploring the difficulties of principlism and case-based reasoning, we must '4 For foundationalists, “basic beliefs” refer to beliefs that justify other beliefs but do not require any other beliefs for their own justification. '5 Laurence BonJour, The Structure of Empirical Knowledge (Harvard University Press, 1985), 91. ‘6 BonJour, The Structure of Empirical Knowledge, 91. 19 consider what roles a method of moral reasoning should play in handling ethical issues in health care. First, as medical technology has advanced, new ethical issues have arisen related to abortion, In Vitro Fertilization (IVF), organ transplantation, human experimentation, physician-assisted suicide, genetic therapy, and the like: Who can make use of NF technology? Should it be allowed only for heterosexual couples? If so, must they also be married? How should we distribute the limited supply of transplantable organs? Should abstinent alcoholics compete equally with non-alcoholics for the limited supply of transplantable livers? How much information should be given to obtain consent for experimentation? In what circumstances may physician-assisted suicide be morally permissible? Should physician-assisted suicide be legalized? When should we provide genetic therapy and to whom? Past moral reasoning and moral theories have not considered many of these issues. We nevertheless must arrive at conclusions. We need, then, a method of moral reasoning that can cope with new such ethical issues. Second, as the world has become smaller, various cultures, religions, ways of life, and value systems have come to exist side by side within single societies. With the development of transportation and communication technology we now more easily and more often meet people with widely different beliefs and worldviews. Society has become less homogeneous and more pluralistic. What I mean by ‘pluralistic’ needs to be more explained. According to John Rawls, there are in liberal democratic societies diverse reasonable comprehensive religious or philosophical doctrines or worldviews. This is “a permanent feature of the public culture of democracy.”'7 He calls this feature reasonable '7 John Rawls, Political Liberalism (Columbia University Press, 1996), 36. or course, we should distinguish reasonable pluralism from pluralism as such including unreasonable views or doctrines as 20 pluralism. Following Joshua Cohen’s distinction, Rawls distinguishes reasonable pluralism from pluralism as such.18 Pluralism as such allows for diversity to include unreasonable views or doctrines that come from “selfishness, prejudice, ignorance, bad reasoning, and so on.” On the contrary, reasonable pluralism holds that diverse reasonable comprehensive views or doctrines can still remain even if we have overcome “selfishness, prejudice, ignorance, bad reasoning, and so on.”19 Rawls identifies a number of sources of reasonable disagreement. Even among reasonable persons, for example, different people may make different judgments if (1) the evidence is conflicting and complex, (2) the weight or overall assessment of relevant considerations differ because our total experiences, culture, and upbringing differ, (3) concepts and their interpretations differ because they are vague and indeterminate, and (4) good and important values and principles are themselves in conflict.20 The last consideration deserves more attention. It states the view that there is a plurality of values that are occasionally conflicting. This view is often called moral pluralism. Virginia Held presents the stronger view that there are inherently diverse values, obligation, virtues, ideals, or fundamental moral principles that cannot be reconciled into one harmonious scheme of morality.21 Conflicts among values, obligations, and moral principles are often seen when we Rawls emphasizes. '8 Rawls, Political Liberalism, 36. '9 Martin Benjamin, Philosophy and This Actual World: An Introduction to Practical Philosophical Inquiry (Roman & Littlefield Publishers, Inc., 2003), 125. Isaiah Berlin also presents the same point, comparing relativism with pluralism in “On the Pursuit of the Ideal,” in The Crooked Timber of Humanity, edited by Henry Hardy (Princeton University Press, 1990), 1-19. 20 Rawls, Political Liberalism, 56-57. 2' Virginia Held describes this view in Encyclopedia of Ethics, 2"d ed., edited by Lawrence C. Becker and Charlotte B. Becker (Routledge, 2001), 1138. 21 are faced with moral dilemmas.22 The diversity of reasonable comprehensive views or doctrines becomes apparent when we struggle with societal ethical problems. What we need is, then, a method of moral reasoning that can work in a pluralistic society characterized by reasonable moral pluralism. Finally, many ethical issues in health care are societal moral problems of social ethics as well as an individual ethics. Questions such as whether I should respect my patient’s refusal of life-sustaining treatment, for example, present not only a moral problem for the physician (or the patient), but also problem of a health care policy regarding society’s position on the relative value of life and autonomy. In addition, ethical questions such as “How should we distribute limited organs?” or “In what circumstances should genetic therapy be allowed?” directly involve health care policy matters. Therefore, a method of moral reasoning should be able to be employed by a group as well as an individual in order to serve health care policy-making. In the following subsections, I evaluate principlism, case-based reasoning, and the method of WRE to see which of the three is the most plausible. 2.2. Difficulties of Principlism Principlists assume that moral principles such as that of autonomy, utility, beneficence, and justice, or the Kantian categorical imperative, etc., have a privileged epistemic status. They think we can reach a conclusion by applying moral principles to particular cases. Therefore, they decide what moral principle (or principles) is to cover a 22 The existence of moral dilemmas does not mean that the set of moral principles held by an individual person is inconsistent. See, Ruth Barcan Marcus, “Moral Dilemmas and Consistency,” The Journal of 22 given case and then appeal to it to justify their conclusion. But as controversial issues in medical ethics show, the mere application of principles does not suffice. Our moral world is too complex to merely apply moral principles to particular cases. Principlism faces several difficulties. First, before even considering applying moral principles, we must often deal with difficult metaphysical issues. For example, the metaphysical issue of when a human life starts complicates the issue of abortion. Likewise, questions of personal identity through time often arise as a key problem in addressing ethical issues related to advance directives. Many moral disagreements come from different religious positions or from different descriptions of the situation rather than fi'om disagreement on the application of moral principles. As mentioned with reasonable pluralism, we often encounter various metaphysical or religious worldviews that sometimes conflict. This is why the project of applied ethics cannot easily work to solve our controversial moral issues. Many issues are so metaphysically or morally complicated that they cannot be resolved by merely applying moral principles. Second, we sometimes face a conflict between what a moral principle recommends and what our intuitions (or particular moral judgments) tell. For example, according the Kantian principle, telling the truth should be always preserved. But should we tell the truth if we are faced with an angry abusive husband asking the whereabouts of his wife and we know where she is? For a Kantian, the answer will be "yes," even if our intuition says “no.” What should we do when our antecedently selected moral principles conflict with our intuitions or particular moral judgments? Should our moral principles still enjoy justificatory priority? Third, and more importantly, we are often faced with what seem on the face of it Philosophy 77 (1980): 121—136. 23 to be conflicts between moral principles themselves as moral pluralism holds. The conflict between the principle of utility and the Kantian categorical imperative is at the heart of issues related to human experimentation. Debates about the allocation of transplantable organs also involve conflicts between the principles of utility and justice. How can conflicts between moral principles be solved within principlism? One possible answer is that we can solve such conflicts by developing a hierarchy of principles. But can we have a systematic moral theory covering all our moral principles and putting them in a hierarchy? To have a hierarchy of principles would mean that we could assign priorities to each before looking at particular cases. It does not sound plausible because the choice of principles that will be applied should be decided with the consideration of particular cases. There seems to be no a priori way by which to decide a hierarchy of principles. Fourth, the conflict of moral principles raises the worry that moral reasoners following principlism may arrive at their conclusions too hastily. When a reasoner does not pay attention to what moral principles may conflict in a given case, he/she may move too quickly to a conclusion by appealing to the only moral principle he/she holds or happens to consider. In this way, the reasoner may too easily ignore other moral principles that could lead to competing conclusions. Fifth, foundationalism, which underlies principlism, presents its own theoretical problem to principlism. Foundationalism maintains a one-directional justification. Therefore, the direction of justification in principlism is also one-directional. Particular moral judgments should be supported by moral principles. But why should we give a justificatory priority to moral principles rather than particular moral judgments? In 24 addition, we end up with a whole variety of particular moral judgments as the conclusions of our moral reasoning. If the justification of each particular moral judgment comes from a moral principle, it would mean that all particular moral judgments could be traced back to a certain limited number of moral principles. But I wonder if they actually could be. I know of no plausible account that shows they can. 2.3. Difficulties of Case-Based Reasoning Casuists recommend that we look at individual cases carefully without having in mind any moral principles. Their emphasis on careful review of individual cases, as John D. Arras points out, has educational merit.23 Case-based reasoning, unlike principlism, reminds us of the importance of looking at particular cases and circumstances more carefully. Furthermore, it makes us reconsider the importance of particular moral judgments in moral reasoning. However, case-based reasoning has the following difficulties. First, how can we judge whether one case is similar to others or to a paradigm case without appealing to moral principles or some kind of generalization? For casuists, case study is important because they think that moral reasoning is similar to the method of common law. Particular moral judgments established from previous cases may help us draw conclusions in cases similar to them. But finding similarities between cases seems to involve some level of generality. We have to know first what kind of issue is involved in a given case. This requires us to know what values are in conflict in a given case. 23 John D. Arras, “Getting Down to Cases: The Revival of Casuistry in Bioethics,” The Journal of Medicine and Philosophy 16 ( 1991 )2 29-51. 25 Thus, we cannot find similarities between cases without appealing to or using generalizations such as independent moral concepts, principles or theories. In fact, similarities and differences are dependent on perspectives. They can vary depending on what points of a given case we are focusing on. For example, voluntary passive euthanasia and voluntary active euthanasia can be distinguished according to who or what is the proximate physical cause of a patient’s death, or found to be similar when we focus on a patient’s will and respect for patient autonomy by a health care provider. If we do not have moral principles or moral theories to inform our perspectives, finding similarities or differences is likely to be merely arbitrary. Second, can case-based reasoning explain how we can get a particular moral judgment all of us agree on? Is it due to our intuition? Merely looking at an individual case itself does not always lead us to a particular moral judgment all of us agree on. The case-based reasoning view looks too optimistic. Could not a person make a biased particular moral judgment even after looking at a given particular case? How can case— based reasoning correct this person’s judgment? Case-based reasoning sounds like saying “Look at cases — they will lead us to the right moral judgment.” In this respect, case- based reasoning seems to collapse into intuitionism and all its problems. Third, what is the solution if different particular moral judgments conflict with one another even after reasoners look at a particular case carefully? What is the solution if there are competing paradigm cases that a given problematic case is considered to belong to?24 Conflicting particular moral judgments or paradigm cases can occur because of different people having different worldviews, religious views, or different perspectives 2" Tom Tomlinson, “Casuistry In Medical Ethics: Rehabilitated, or Repeat Offender?” Theoretical Medicine 15 (1994): 13. 26 in looking at the same problematic case. Of course, we should not overlook the fact that many moral beliefs enjoy a broad consensus regardless of culture and religion. But we should also not miss the fact that there exist reasonable moral disagreements. If casuists cannot provide an explanation for how our particular moral judgments are formed from our looking at particular cases and how we might correct or modify these judgments, we are left with no way to solve conflicts of judgment within case-based reasoning. Furthermore, case-based reasoning has a weakness in that it cannot explain why we have various reasonable moral disagreements. Fourth, case-based reasoning, like principlism, has some theoretical problems that arise from its commitment to foundationalism. Case-based reasoning is involved in one- directional justification. The justification of moral principles or moral theories comes from particular moral judgments. As mentioned before, moral principles or theories are considered, for casuists, to be summaries of particular moral judgments. But are all moral principles summaries of particular moral judgments? Must we always have a particular moral judgment before establishing a moral principle? Does the moral rule of keeping a promise require a particular moral judgment for its justification? In fact, this rule can be justified on the basis of the notion of a promise itself. Case-based reasoning, like principlism, ignores the possibility of bi-directional justification between particular moral judgments and moral principles, or multi-directional justification among various moral beliefs and non-moral beliefs. Fifth, case-based reasoning does not seem to successfully serve our health care policy-making needs. It is more oriented to the evaluation of particular cases. But we also 27 have to reach conclusions on ethical issues which can contribute to developing health care policy. Policy-making requires establishing some set of rules or principles covering future possible cases as well as diverse current particular cases. It requires us to work at a level of generality, rather than at a level of particularity of individual cases. Finally, case-based reasoning involves “the danger of uncritical conventionalism and conservatism”25 because it is based on established paradigm cases. New cases may involve ethical aspects not yet seen in the settled paradigm cases. This method may miss these new ethical aspects. 3. The Method of WRE 3.1. Coherentism as a Model of Moral Justification The difficulties related to foundationalism, which underlies both principlism and case-based reasoning, leads us to consider the virtues of coherentism. The difficulties of foundationaist moral reasoning are as follows. First, how can a certain type of moral beliefs — moral principles for principlism and particular moral judgments for case-based reasoning — enjoy a privileged epistemic status? Principlists may say that moral principles function in practical moral reasoning as basic beliefs not requiring any further beliefs for their justification. Casuists may say that particular moral judgments are basic beliefs. However, why should we accept the idea that the principle of utility or the principle of autonomy is a basic moral belief? The same 25 Tomlinson, “Casuistry In Medical Ethics: Rehabilitated, or Repeat Offender?” 14. He points out that the settled paradigms themselves might not be challenged. 28 question arises with respect to particular moral judgments such as “Killing an innocent person who is not involved in this terror is wrong.” If moral principles or particular moral judgments were self-evident, we would not in practical reasoning ask about their justification. But this is not the case. We often ask for the justification of moral principles and particular moral judgments — especially when they conflict with one another. Therefore, we can conclude that moral principles and particular moral judgments are not self-evident. There may be two possible responses. First, moderate foundationalists might respond that a basic moral belief, whether it is a moral principle or a particular moral judgment, does not have to be a self-evident belief. They argue that beliefs are basic beliefs if they have merely “a prima facie direct justification.” According to this view, moral principles or particular moral judgments may be defeasible, but they are still basic beliefs. Even if we can ask why lying is wrong or why the principle of utility is justified in some particular cases, the moderate foundationalist will argue that this just shows that these basic moral beliefs are not feasible in those cases. Second, principlists may deny that moral principles are basic beliefs saying that they should be justified by an ethical theory.26 Thus, a set of basic beliefs justifies moral principles. However, both responses still have the following difficulty, which is true of both basic beliefs and prima facie basic beliefs. How can we know whether a given belief is a basic belief (or prima facie basic belief)? If some criteria are provided, the justification of 27 the basic beliefs results in their being based on the belief that they satisfy these criteria. Therefore, basic beliefs are no longer basic. If no criteria can be provided, we will be 2° See Beauchamp and Childress, Principles of Biomedical Ethics, 15. 27 BonJour presents a similar argument: A person should be aware that a basic belief satisfies some features 29 unable to discern basic beliefs from non-basic beliefs, and more importantly, unable to distinguish basic beliefs from beliefs mistakenly considered to be basic beliefs. Unfortunately, therefore, we cannot exclude the possibility that some wrong beliefs may be presented as basic beliefs. We are faced, then, with the problem of not knowing how to argue against wrong beliefs or bad intuitions. A second difficulty of foundationalist moral reasoning is this. Why should the justificatory relationship between moral principles and particular judgments be understood in only one-directional terms, from moral principles to particular judgments or vice versa? F oundationalist moral reasoning holds a misguided view of the structure of moral justification. According to foundationalism, the chain of justification runs one- directionally from basic beliefs to non-basic beliefs. Foundationalist moral reasoning thus ignores multi-directional relationships among beliefs, which is, as I hope to show, a more plausible picture of moral justification. When we try to support some moral principles, we sometimes appeal to relevant particular moral judgments. When we justify particular moral judgments, we sometimes appeal to moral principles or moral theories. For example, when we explain the Kantian categorical imperative, “respect persons as ends in themselves, not as a mere means,” we can appeal to a particular moral judgment such as “Forcing some one to have sex or treating some one as a slave is morally wrong.” On the contrary, to justify particular moral judgments that rape or slavery are morally wrong, we can appeal to the Kantian categorical imperative. So, the idea that the only available basic resources of justification are either moral principles or particular moral judgments seems to give us too simplified a picture of moral justification and ignores more dynamic relationships between these of being a basic belief. See BonJour, The Structure of Empirical Knowledge, 30-32. 30 two kinds of moral beliefs. A similar example can be drawn from Nelson Goodman, who tries to show a bi- directional relationship between deductive inferences and valid general rules: “Deductive inferences are justified by their conformity to valid general rules, and general rules are justified by their conformity to valid inference.”28 He admits that this looks circular, but claims that this circularity is virtuous.29 He continues by saying, The point is that rules and particular inferences alike are justified by being brought into agreement with each other. A rule is amended if it yields an inference we are unwilling to accept; an inference is rejected if it violates a rule we are unwilling to amend. The process of justification is the delicate one of making mutual adjustments between rules and accepted inferences.30 So, mutual justificatory relationships between rules or principles and what they are supposed to regulate are not limited to moral beliefs. Coherentism, unlike foundationalism, emphasizes mutual relationships in justification, holding that justification is obtained when a belief coheres with other beliefs.31 For coherentism, justification is a matter of mutual supports among beliefs. The structure of justification is not one-directional, but multi-directional. When this idea is applied to the relationship between moral principles and particular moral judgments, we find that they should be mutually supportive. Furthermore, coherentism does not give justificatory priority to either moral principles or particular moral judgments. There are no basic beliefs. The method of WRE 28 Nelson Goodman, F act, Fiction, and Forecast (Harvard University Press, 1983), 64. 29 Goodman, Fact, Fiction, and Forecast, 64. 3° Goodman, Fact, Fiction, and Forecast, 64 3‘ The notion of coherence requires more than logical consistency. More detailed work will be needed about what is meant by coherence. BonJour provides some features of coherence. See Chapter 5, in The Structure of Empirical Knowledge, 87-110. Norman Daniels also says that coherence involves not only more than mere logical consistency, but also inference to the best explanation, plausibility, and simplicity. See his Justice and Justification: Reflective equilibrium in Theory and Practice (Cambridge University Press, 31 as coherentist moral reasoning can avoid some of the difficulties involved in the justification of so-called basic beliefs. This method also stresses multi-directional justificatory relationships among particular moral judgments, moral principles, and relevant background beliefs, thereby moving beyond even a bi-directional relationship between the first two elements.32 3.2. The Method of WRE and Its Characteristics The structure of WRE consists of the following three main elements: considered moral judgments, moral principles or rules, and background beliefs or theories. Norman Daniels expresses the method of WRE as follows: The method of wide reflective equilibrium is an attempt to produce coherence in an ordered triple of sets of beliefs held by a particular person, namely, (a) a set of considered moral judgments, (b) a set of moral principles, and (c) a set of relevant background theories. .. We can imagine the agent working back and forth, making adjustments to his considered judgments, his moral principles, and his background theories. In this way he arrives at an equilibrium point that consists of the ordered triple (a), (b), (”.33 In the framework of WRE, the set of considered moral judgments covers judgments that occur at all levels of generality “from those about particular situations and institutions up 1996), 2. 32 Before I move on the detailed explanation of the notion of WRE, I must make clear that there may be ethical certainties that are similar to Wittgenstein’s epistemological certainties presented in his 0n Certainty, ed. G. E. M. Anscombe and G. H. von Wright (Harper & Row, 1969). Ethical certainties are not a matter of moral knowledge. They thus do not need moral justification. But they are ethical convictions that are certain and that stand fast, serving as the hinges for the various language games in which we state, defend, criticize, and revise our ethical beliefs and knowledge claims. Ethical certainties are not first and foremost propositional. They are revealed in what we do rather than what we say. Examples of ethical certainties include “It is wrong to cut off someone ’5 arms just to see what he or she would look like without them” and “It is wrong to torture someone for no other reason than to watch him or her suffer.” Our certainty that these actions are wrong is stronger than any possible justification we can give for why they are wrong. Such certainties neither need nor admit of meaningful justification. 33 Norman Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” The Journal of 32 through broad standards and first principles to formal and abstract conditions on moral conceptions.”34 So, among the set are particular moral judgments such as: “Hitler’s ”35 “The terrorists’ deliberately targeting civilians during the Blitz on London was wrong, attack that occurred in America on September 11th was wrong,” and “The Bush administration’s decision to go to war against Iraq was wrong.” In addition, the set contains more general moral judgments such as “Religious intolerance and racial discrimination are unjust.”36 Considered moral judgments are, according to Rawls, judgments “in which our ”37 He does not include moral capacities are most likely to be displayed without distortion. as a considered moral judgment “those judgments made with hesitation, or in which we have little confidence.”38 Building on this, Daniels describes considered moral judgments as judgments in which we are “relatively confident and [that] have been made under conditions conducive to avoid errors of judgment.”3 9 Considered moral judgments should be distinguished from merely intuitive judgments because we have to “filter initial moral judgments to arrive at considered moral judgments.”4O As Rawls says, “In deciding which of our judgments to take into account we may reasonably select some and exclude ”41 other. (My italics.) The set of moral principles covers a variety of moral principles and rules:42 Kant’s Philosophy 76 (1979): 258-259. 3" John Rawls, “The Independence of Moral Theory,” Proceedings and Addresses of the American Philosophical Association 48 (1974-75): 8. Cited by Beauchamp and Childress, Principles of Biomedical Ethics, 21. 35 Benjamin, Philosophy and This Actual World, 113. 36 John Rawls, A Theory of Justice, revised ed. (Harvard University Press, 1999), 17. 37 Rawls, A Theory of Justice, 42. 3’ Rawls, A Theory of Justice, 42. ’9 Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” 258. ‘0 Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” 266. “ Rawls, A Theory of Justice, 42. ’2 Daniels is not clear about moral rules. He considers a maxim to be a considered moral judgment. He 33 categorical imperatives, the principle of utility, the principle of beneficence, Rawls’s principles of justice, and moral rules such as “Don’t murder,” “Keep your promises,” and “Keep patients from harm and injustice.” A wide variety of beliefs constitute the set of background beliefs or theories. The inclusion of these beliefs makes reflective equilibrium wide. When equilibrium is pursued only between considered moral judgments and moral principles, it is a narrow equilibrium. Norman Daniels mentions the following beliefs as belonging to the set of background beliefs or theories: Our beliefs about the conflict between consequentialist and deontological views ; about the nature of persons; about the role or function of ethics in our lives: about the implication of game theory, decision theory, and accounts of rationality for morality; about human psychology, sociology. and political and economic behavior; about the ways we should reply to moral skepticism and moral disagreement; and about moral justification itself.43 As the above passage shows, background beliefs or theories will include a variety of non- moral beliefs or theories as well as meta-ethical beliefs or theories. Therefore, the method of WRE suggests that we should seek coherence among various moral and nonmoral beliefs. So, when we are faced with a moral question, we should seek an answer (that is, a moral belief as a conclusion of our moral reasoning) that achieves equilibrium among the above three elements, making revisions to one or more if necessary. The pursuit of coherence is the main characteristic of the method of WRE. Rawls in his description of the original position and the resulting principles of justice says that: says, “suppose the considered moral judgment is about what is right or wrong, just or unjust, in particular situations, or is a maxim that governs such situations.” See, Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” 266. 43 Norman Daniels, Justice and Justification: Reflective equilibrium in Theory and Practice (Cambridge 34 In arriving at the favored interpretation of the initial situation there is no point at which an appeal is made to self-evidence in the traditional sense either of general conceptions or particular convictions... A conception of justice cannot be deduced from self-evident premises or conditions on principles: instead, its justification is a matter of the mutual support of many considerations, of everything fitting together into one coherent view.44 (My italics) In the above quotation, we can see Rawls as a coherentist. More importantly, we can also see his belief that neither moral principles nor particular moral judgments serve as self- evident basic moral beliefs. Neither moral principles nor particular judgments can be justified independently of the other. The method of WRE is, in this way, unlike principlism and case-based reasoning. Another important characteristic of the method of WRE is its view that no beliefs are immune to revision. As Daniels points out in the above citation, we can “imagine the agent working back and forth, making adjustments to his considered judgments, his moral principles, and his background theories” in order to get the strongest mutual support among the three main elements. As long as the revisions improve coherence, any belief can be revised because none serves as a basic belief. Therefore, even considered moral judgments that sometimes play the role of providing starting points or are taken to be “provisional fixed points”45 can be revised or replaced by another moral judgment. Rawls puts it this way: We can either modify the account of the initial situation or we can revise our existing judgments, for even the judgments we take provisionally as fixed points are liable to revision.46 None of considered judgments, moral principles, or background beliefs has justificatory University Press, 1996), 6. 4‘ Rawls, A Theory of Justice, 19. ‘5 Rawls, A Theory of Justice, 18. 4" Rawls, A Theory of Justice, 18. 35 priority over others. Each of them is subject to revision. Finally, we should not miss another important characteristic of WRE. When we reach a state through the back-and-forth process of revision in which all the relevant considered moral judgments, moral principles, and background beliefs cohere with one another, a state of WRE is obtained. But, as Rawls points out, “this equilibrium is not necessarily stable.”47 In fact, the state of equilibrium is not a once-and—for-all affair. Equilibrium may be challenged when some of our beliefs change or when we encounter new circumstances calling for revision. In these situations, we will have to make adjustments to reach another state of WRE. The method of WRE should, then, be understood to be an ongoing process. 3.3. The Three Desiderata and the Method of WRE The above characteristics of the method of WRE show that it is more likely to satisfy the three desiderata for a method of moral reasoning than the others. For the first desideratum, the ability of coping with new ethical issues, the method of WRE incorporates new beliefs, whether they are moral or non-moral, into its framework and then seeks new equilibrium. For example, the development of medical technology may bring about new particular moral judgments and/or background beliefs. They may sometimes result in a loss of equilibrium. In such situations, the method tries to develop a new equilibrium. The back-and-forth process of revision may show that some of previous considered moral judgments, moral principles, or background beliefs should be revised to ‘7 Rawls, A Theory of Justice, 18. “This equilibrium” here refers to narrow reflective equilibrium. But the same idea is no doubt true of wide reflective equilibrium. 36 cope with the new situation and beliefs. This occurred when we revised the conception of death. People were previously determined to be dead if there was cessation of respiration and circulation of the blood. Respirators were now allowing bodies of those whose brains had been totally and permanently destroyed to breathe and circulate the blood. On the criteria — breathing and heartbeat - these patients were alive. But did this make sense if they were totally and permanently unconscious due to the loss of all brain function? Many people would have new particular moral judgments when they observed their friends or family members in such a circumstance. They might think, “this treatment should be stopped; it is useless and meaningless.” We developed a new considered moral judgment that a physician’s providing a respirator to a patient whose brain is permanently destroyed is wrong. This particular moral judgment upsets our equilibrium. This equilibrium consisted of (1) a particular moral judgment that it is wrong to withdraw life-saving treatment, (2) a moral rule that a physician should preserve his/her patient’s life, and (3) a background belief that individuals with heartbeat and respiration are alive. Thus, we had a new issue of whether a physician should continue to respirator treatments for patients whose brain function had been totally and permanently destroyed. But we retained equilibrium by revising the conception of death without throwing out the physician’s moral rule and the old and new particular moral judgments involved. Our conception of death was changed to the cessation of the function of a whole brain. Patients whose brain function was totally and permanently lost were now to be regarded as dead. If we just tried to deduce an answer from moral principles alone or if we tried to find an agreed intuition through looking at individual cases, without revising the relevant background belief, we would 37 not have had a plausible answer to the issue. We would have had either to weaken the physician’s obligation to preserve life or our intensive care units would be filled with patients who were totally and permanently unconscious. For the second desideratum, the method of WRE is more likely to work with moral pluralism and reasonable pluralism than the others. We have an example of conflicting moral principles from the issue of human experimentation. Getting an informed consent can be understood as an outcome of the revision process the method of WRE stresses, rather than from deducing from moral principles, or from just careful review of individual cases. If we just tried to deduce an answer to the issue from the relevant moral principles, we would easily find the conflicting two moral principles; the principle of utility and Kant’s categorical imperative (“treat people as ends-in- themselves, not as a mere means”). By appeal to the former, human experimentation may be justified even though it may put some people in situations they will lose their life or healthy. On the contrary, human experimentation itself should not be allowed by appeal to the latter even though it may produce benefits to all of us. But we did not merely apply these rules to human experimentation. We establish a rule to require an informed consent, which restricts a researcher’s unlimited pursuit of utility and keeps Kant’s categorical imperative with its appropriate interpretation. We should pursue utility, but the principle of utility does not always win other principles. So long as a person participates in human experimentation with his/her informed autonomous decision, it is not considered to treat the person merely as a means whereas the Tuskegee experimentation treated persons merely as a means. Thus, the revision of the relevant moral principles and their interpretation made us achieve equilibrium among our particular moral judgments on the 38 Tuskegee experimentation, the principle of utility and autonomy, and background beliefs about the importance of acquiring scientifically proved medial knowledge. The above two examples illustrate how the method of WRE has already been used in medical ethics. As Martin Benjamin says, “though the label of ‘wide reflective equilibrium’ may be new to you, the process is probably not.”48 He continues to say, “If you’ve deliberated about real-life ethical problems, my hunch is you’ve already engaging in this kind of thinking, even if you were not explicitly aware of it.”49 The method of WRE can work with reasonable pluralism because equilibrium should be sought among beliefs the society has as well as among beliefs an individual person has. The former can be called the interpersonal method of WRE if we call the latter the intrapersonal method of WRE. The aspects of the interpersonal method of WRE will be dealt with in Chapter Three. The method of WRE can take into account reasonable disagreement among reasonable comprehensive views, which will be dealt with in Chapter Four. For the third desideratum, the method of WRE is more likely to contribute to developing heath care policies than the others. The method of WRE can be accommodated to group reasoning. As applied to group reasoning, the method of WRE will be used to achieve equilibrium among the beliefs the members of the group hold. In some cases this will require compromise. I will try to show how in some cases members of a group can agree to a compromise on an ethical question without compromising their personal integrity. 1 will deal with the use of the method of WRE as group reasoning in Chapter Five. 48 Benjamin, Philosophy and This Actual World, 119. 49 Benjamin, Philosophy and This Actual World, 119. 39 Chapter Two Reflective Equilibrium and Foundationalism The method of reflective equilibrium, whether it is wide or narrow, has usually been understood as a coherentist moral reasoning and justification. However, R. M. Hare and Peter Singer hold that the method of reflective equilibrium is committed to foundationalism because it is committed to intuitionism. Michael R. DePaul claims that the methods of wide and narrow reflective equilibrium while not committed to foundationalism, are compatible with it. More strongly, Roger P. Ebertz argues that the method of reflective equilibrium can be explained as a modest foundationalism rather than coherentism. In contrast to Hare, Singer, DePaul, and Ebertz, I argue in this chapter that the methods of reflective equilibrium and foundationalism are not compatible. These authors have an incomplete understanding of Rawls’s notion of “considered moral judgments.” I hope here to correct some misunderstandings of the method of reflective equilibrium by responding to their claims. 1. The Method of Reflective Equilibrium and Foundationalism 1.1. Hare’s and Singer’s Claim Hare and Singer claim that Rawls’s reflective equilibrium is committed to 40 intuitionism and subjectivism.l They do not, however, directly claim that reflective equilibrium is committed to foundationalism. Rather, their claim that reflective equilibrium is committed to intuitionism has been construed by Norman Daniels and Michael R. DePaul to imply a corresponding commitment to foundationalism. DePaul says, “It is surprising to find that some leading critics of [the method of reflective equilibrium], e.g., R. M. Hare and Peter Singer, have concluded that the method ’92 is committed to foundationalism. DePaul is simply assuming that intuitionism implies foundationalism. According to him, Hare and Singer charge that reflective equilibrium is “a thinly disguised form of intuitionism, and hence, is committed to foundationalism”3 (My italics). Daniels reads Hare and Singer in the same way. He says that their charge of intuitionism “seems unfounded” because “intuitionist theories have generally been foundationalist”4 — intuitionists will pick out some set of beliefs “as basic or self- warranting?” Hare and Singer both reference the following passage from Rawls to make their claim that he is committed to intuitionism and hence to foundationalism: There is a definite if limited class of facts against which conjectured principles can be checked, namely our considered judgments in reflective equilibrium.6 (My italics) If considered judgments are taken to be intuitions, the passage seems to commit Rawls to ' R. M. Hare, “Rawls’ Theory of Justice-l,” Philosophical Quarterly 23 (1973): 144-155. Peter Singer, “Sidgwick and Reflective Equilibrium,” The Monist 58 (1974): 490-517. 2 Michael R. DePaul, “Reflective Equilibrium and Foundationalism,” American Philosophical Quarterly 23, no. 1 (1986): 60. 3 DePaul, “Reflective Equilibrium and Foundationalism,” 60. He adds, of course, that what Hare and Singer find objectionable is the intuitionistic element of reflective equilibrium, not foundationalism per se. " Norman Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” The Journal of Philosophy 76 (1979): 264. 5 Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” 264. 6 John Rawls, A Theory of Justice, revised ed. (Harvard University Press, 1999), 51. Hare cites this passage on pp.144-145 in his “Rawls’ Theory of Justice-I.” Singer cites it on p.493 in his “Sidgwick and Reflective Equilibrium.” 41 intuitionism. The words ‘facts’ and ‘checked’ also make it sound like considered judgments are to serve as foundations. Singer criticizes the method of reflective equilibrium as follows: The reflective equilibrium conception of moral philosophy, then, by leading us to think of our particular moral judgments as data against which moral theories are to be tested, is liable to mislead it slides easily into the view that moral theories are to be tested against the moral judgments made by some group or consensus ...7 Singer complains that for Rawls the moral judgments made by a group function as data or facts against which moral theories are tested. DePaul adds that Hare and Singer think that “Rawls’ method takes the wrong beliefs as the foundation for moral theory.”8 Considered moral judgments, Hare and Singer think, do not deserve such privilege.9 Similarly, Daniels summarizes Hare’s and Singer’s position by saying that the charge of intuitionism, rests on two distinct complaints: (1) that reflective equilibrium merely systematizes some relatively determinate set of moral judgments; and (2) that the considered moral judgments are not a proper foundation for an ethical theory.10 (My italics) In fact, however, the method of reflective equilibrium does not merely “systematize” some relatively fixed set of moral judgments because, according to Rawls, these judgments should be revised when they do not fit with justifiable moral principles. However, the words ‘facts’ and ‘check’ in the above passage from Rawls may give the impression that considered moral judgments are to serve as foundations, or enjoy a privileged epistemic status as Hare and Singer suppose. But this understanding is not 7 Singer, “Sidgwick and Reflective Equilibrium,” p.517. 8 DePaul, “Reflective Equilibrium and F oundationalisrn,” 60. 9 As mentioned in Chapter One, considered moral judgments are made at all levels of generality. So, particular considered moral judgments belong to a set of considered moral judgments. I do not here, however, pay attention to this distinction between particular moral judgments and considered moral 42 accurate; it considers only one direction of bi-directional justificatory relationship between considered moral judgments and moral principles. I will return to this after dealing with Daniels’s response to Hare’s and Singer’s claim. 1.2. Daniels’s Defense Daniels defends reflective equilibrium from Hare’s and Singer’s charge by emphasizing that a considered moral judgment is a “provisional fixed point,”11 or a starting point for the process of reflective equilibrium. Considered judgments are not firmly fixed foundations, but are open to revision. To highlight this point, he compares narrow reflective equilibrium (NRE) with wide reflective equilibrium (WRE), arguing that the latter “allows far more drastic theory-based revisions for moral judgments” than does the former.12 Suppose we are comparing ethical theories following Hare’s, Singer’s, and Daniels’s interest in using of the method of WRE for deciding whether to accept an ethical theory rather than for arriving at a particular moral belief as the conclusion of moral reasoning. We would start with some considered moral judgments that have been filtered from initial moral judgments. Suppose we have competing sets of moral principles that fit with these considered moral judgments. We have now achieved the state of NRE. The method of WRE requires a further step. As Daniels says, in this case, “we advance philosophical arguments intended to bring out the relative strengths and judgments because Hare, Singer, Daniels, DePaul, and Ebertz did not make use of it. '0 Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” 265. ” Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” 267. '2 Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” 266. 43 weakness of the alternative sets of principles (or competing moral conceptions)” ‘3 These arguments are “inferences from some set of relevant background theories.”l4 A set of moral principles will be accepted if the set of arguments supporting it is deemed stronger than those for the competing set. But suppose we find that none of the competing sets quite fit with our background beliefs or theories. We would have to revise the moral principles in each set. In this situation, Daniels says that a considered moral judgment is a judgment relevant to establishing partial reflective equilibrium with general moral principles. Consequently, we must revise it if background theories compel us to revise our general principles.15 This passage makes little sense, however, if it means that we must revise only considered moral judgments, and not moral principles. The revision of considered moral judgments should go hand in hand with revisions of moral principles. The moral principles, after all, are supposed to have been in equilibrium with the considered moral judgments until we find them incompatible with our background beliefs or theories. There may be, of course, a different reason for revising moral principles when background beliefs or theories compel us to revise our considered moral judgments. These various reasons for revision show, Daniels says, “the additional ways in which a considered moral judgment is subject to revision in wide equilibrium.”16 Of course, background beliefs or theories can be also revised. We can imagine a more complex back-and-forth process in which adjustments are made among all of considered moral judgments, moral principles, and background theories. As Daniels puts '3 Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” 258. '4 Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” 258. '5 Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” 266 '6 Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” 266. 44 it, the method of WRE “permits extensive revision.”l7 Therefore, the charge of intuitionism and thus of foundationalism is unfounded. Daniels, however, goes one step further. He believes that although WRE is not committed to intuitionism (and thus not to foundationalism), NRE is: One reason philosophers have thought reflective equilibrium “intuitionist” is a failure to distinguish narrow and wide equilibria. A more obvious source lies in Rawls’s remark, cited by nearly everyone who makes the charge of intuitionism, that “There is a definite if limited class of facts against which conjectured principles can be checked, namely our considered judgments in reflective equilibrium.” It is tempting to read the remark as follows: “to arrive at a reflective equilibrium, treat considered judgments as a ‘definite if limited class of facts’ which is to determine the shape and content of the rest of the theory.” R. M. Hare and Peter Singer read Rawls’s remark this way. But the remark can and should be taken to mean that “the small but definite class” emerges only when reflective equilibrium is reached, and still is revisable in the light of further theory change. I 8 I wonder, though, how ‘checked’ may be said to refer to ‘checked in the state of after- equilibrium’ rather than ‘checked in the state of before-equilibrium.’ Even if we set this question aside, Daniels’s defense of WRE from the charge of intuitionism, and thus foundationalism, remains problematic. In particular, the distinction between NRE and WRE seems irrelevant to the defense of reflective equilibrium from charges of intuitionism and thus foundationalism. The problematic reading of the passage from Rawls does not result from “a failure to distinguish narrow and wide equilibria,” but from an incomplete understanding of the role of considered moral judgments. So long as Hare and Singer ignore the notion of reciprocal support, they will continue to understand the words ‘checked’ and ‘data’ in a foundationalist sense even in the following passage: There is a definite if limited class of '7 Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” 266. '8 Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” 267-268. See footnote #17. 45 facts against which both moral principles and background beliefs can be checked, namely our considered judgments in wide reflective equilibrium.19 Therefore, the problem with Hare’s and Singer’s contention does not lie in their failure to distinguish between NRE and WRE, but in their focus on just one side of the story. Their position faces two difficulties. First, Hare and Singer focus on only one direction of support. The context of the quoted remark from Rawls is that of the acceptance of ethical theory. When we are concerned with the reliability of an ethical theory, we are wondering how well it explains our considered moral judgments. But this need not imply that our considered moral judgments are to be the foundations for the theory. Mutual or reciprocal support means that each element acts sometimes as a justifying belief and other times as a justified belief. As James Blachowicz says, “each part is in some way both fixed and variable.”20 80, the method of reflective equilibrium may begin with a set of considered moral judgments, but this is not the only possible starting point. Depending on the context, we can begin with moral principles or relevant background beliefs. For example, when considering the Nazis judgment that subhuman Jews should be exterminated, we can start first with some biological and psychological knowledge as relevant background beliefs or theories. The Nazis’s judgment can be tested against biological and psychological beliefs to see whether it fits with them. Second, contrary to Hare’s and Singer’s interpretation, Rawls does not intend his considered moral judgments to be privileged basic or foundational beliefs. He says, “We can revise our existing judgments, for even the judgments we take provisionally as fixed '9 This is my modification of the quoted remark from Rawls. 20 James Blachowicz, “Reciprocal Justification in Science and Moral Theory,” Synthese 110 (1997): 449. 46 points are liable to revision.”21 Therefore, “some of the so-called fixed points may not really be fixed.”22 He says in addition, “Justification is a matter of the mutual support of many considerations, of everything fitting together into one coherent view.”23 Therefore, the single passage from Rawls on which Hare and Singer focus should be balanced with the rest of what he says. 2. DePaul’s Defense and His Claim of Compatibility 2.1. DePaul’s Defense from the Charge of F oundationalism DePaul argues that the method of reflective equilibrium, whether it is narrow or wide, is not committed to, but is nonetheless compatible with foundationalism. He also criticizes Daniels’s view that although WRE is not committed to foundationalism, NRE is. This view, DePaul thinks, does not “reveal the actual relationships between the two forms of the method of reflective equilibrium and the various forms of 9924 foundationalism. He argues, “There is not a necessary connection between following the method of reflective equilibrium and whether one’s moral beliefs have a foundational structure.”25 Let us consider first DePaul’s definitions of the various forms of foundationalism. He offers five types: pure epistemic, historical, psychological, epistemic, and methodological foundationalism. 2‘ Rawls, A Theory of Justice, 18. 22 Rawls, A Theory of Justice, 508. 23 Rawls, A Theory of Justice, 507. 2‘ DePaul, “Reflective Equilibrium and Foundationalism,” 63. 47 Pure epistemic foundatinalism is defined as follows: (PEF) A person S’s beliefs in a given area, i.e., a non-empty set G of S’s beliefs, have a purely epistemicly foundational structure if, and only if, (1) there is a non-empty subset F of G such that (a) the beliefs in F have a certain positive epistemic status E for S, and (b) the beliefs in F do not have E for S solely in virtue of their standing in some relation to any subset of S’s beliefs; and (2) there is some relation R such that (a) each belief in G that is not in F stands in R to some subset of the beliefs in F, and (b) each belief in G that is not in F has a positive epistemic status E for S in virtue of its standing in R to some subset of the beliefs in F .26 (My bolds) In the above definition, the roles of (1)(a) and (1)(b) are to define the beliefs in F as basic or foundational beliefs, which do not require any other beliefs for their justification. A basic belief has a certain positive epistemic status E independent of any relationship with other beliefs. The role of (2) is to define justificatory relationships between basic beliefs and non-basic beliefs. As (2)(b) shows, non-basic beliefs can be justified because of the relation R. DePaul defines historical foundationalism in this way: (HF) A person S’s beliefs in a given area, i.e., a non-empty set G of S’s beliefs, have a historically foundational structure if, and only if, (1) there is a non-empty subset F of G such that S did not form any belief in F because of any other beliefs; and (2) there is a relation R such that (a) each belief that is in G, but not in F, stands in R to some subset of the beliefs in F, and (b) S formed each of the beliefs in G that is not in F because it stands in R to some subset of the beliefs in F .27 (My bolds) Historical foundationalism pays attention to how beliefs come to exist. DePaul presents two kinds of beliefs. One is a belief that came to exist independently of other beliefs. I will call this kind of belief historically independent belief. The other is a belief that came into existence on the basis of a relation to historically independent beliefs. I will call this 25 DePaul, “Reflective Equilibrium and Foundationalism,” 60. 26 DePaul, “Reflective Equilibrium and F oundationalisrn,” 61. 27 DePaul, “Reflective Equilibrium and Foundationalism,” 61. 48 kind of belief historically dependent belief. It should not be overlooked that DePaul did not include in the above definition the requirement that a certain positive epistemic status E exits for historically independent beliefs. I do not construe historically independent beliefs to be identical to basic beliefs. DePaul says, “Most persons change the reasons they have for holding their beliefs over time.”28 We need, then, a form of foundationalism that is not sensitive to the history of a person’s belief structure, but to the present structure. This type of foundationalism DePaul calls psychological foundationalism: (PF) A person S’s beliefs in a given area, i.e., a non-empty set G of S’s beliefs, have a psychologically foundational structure if, and only if, (1) there is a non-empty subset F of G such that S does not now hold any belief in F because of any other belief S now holds; and (2) there is a relation R such that (a) each belief that is in G, but not in F, stands in R to some subset of the beliefs in F, and (b) S now holds each of the beliefs in G that is not in F because it stands in R to some subset of the beliefs in F.29 (My bolds) Here too, DePaul presents two kinds of belief. One is a belief that is held independently of other beliefs. I will call this kind of belief psychologically independent belief The other is a belief that is held on the basis of a relation to psychologically independent beliefs. I will call this kind of belief psychologically dependent belief In cases where I have no need to distinguish historically independent beliefs from psychologically independent beliefs, I will call them simply independent beliefs. Similarly historically dependent and psychologically dependent beliefs will sometimes be referred to generically as dependent beliefs. Psychological foundationalism differs from historical foundationalism because there are now psychologically independent beliefs and a relation R between psychologically independent beliefs and psychologically dependent beliefs 2" DePaul, “Reflective Equilibrium and Foundationalism,” 61. 49 such that the former justifies the latter. DePaul understands epistemic foundationalism as a combination of psychological foundationalism and pure epistemic foundationalism: (EP) A person S’s beliefs in a given area, i.e., a non-empty set G of S’s beliefs, have an epistemicly foundational structure if, and only if, (1) there is a non-empty subset F of G such that (a) S does not now hold any belief in F because it stands in some relation to any subset of S’s other beliefs, (b) the beliefs in F have a certain positive epistemic status E for S, and (c) the beliefs in F do not have E for S solely in virtue of standing in some relation to any subset of S’s other beliefs; and (2) there is some relation R such that (a) each belief in G that is not in F stands in R to some subset of the beliefs in F, (b) S now holds each belief in G that is not in F because it stands in R to some subset of the beliefs in F, and (C) each belief in G that is not in F has a positive epistemic status E for S in virtue of its standing in R to some subset of the beliefs in F .3 0 (My bolds) This definition’s (2)(a) is shared by all the forms of foundationalism. (l)(a) and (2)(b) come from psychological foundationalism. (l)(b), (l)(c), and (2)(c) come from pure epistemic foundationalism. Epistemic foundationalism thereby states that there are currently held basic and non-basic beliefs, and a relation R holding between basic and non-basic beliefs such that the former justifies the latter. The term ‘basic belief’ is appropriate here because there is an assertion of ‘a certain positive epistemic status E.’ Lastly, DePaul understands methodological foundationalism as a combination of historical and pure epistemic foundationalism: (MF) A person S’s beliefs in a given area, i.e., a non-empty set G of S’s beliefs, have an methodologically foundational structure if, and only if, (1) there is a non-empty subset F of G such that (a) S did not form any belief in F because it stands in some relation to any subset of S’s beliefs, (1)) the beliefs in F have a certain positive epistemic status E for S, and (c) the beliefs in F do not have E for S solely in virtue of standing in some relation to any subset of S’s other beliefs; and (2) there is some relation R such that (a) each belief in G that is not in F stands in R to some subset of the beliefs in F, (b) S formed each belief in G that is not in F because it 29 DePaul, “Reflective Equilibrium and F oundationalism,” 61. 3° DePaul, “Reflective Equilibrium and Foundationalism,” 62. 50 stands in R to some subset of the beliefs in F, and (C) each belief in G that is not in F has a positive epistemic status E for S in virtue of its standing in R to some subset of the beliefs in F .3 I (My bolds) As in the previous definition, (l)(a) and (2)(b) come from historical foundationalism while (l)(b), (l)(c), and (2)(c) come from pure epistemic foundationalism. Methodological foundationalism states that there are basic and non-basic beliefs a person formed, and a relation R between basic and non-basic beliefs such that the former justify the latter. DePaul shows that, in contrast to Daniels, even the method of NRE is not committed to foundationalism. To make this point, he demonstrates how a set of considered moral judgments can change with the process of revision. Let us consider first whether the method of NRE is committed to historical foundationalism. A person S can employ the method of NRE with a set of initial moral judgments, {IMJ}.32 According to DePaul, “The first step of the method would be for S to screen out some of the questionable beliefs in {IMJ} to obtain her initial set of ”33 This is the process by which we filter moral considered moral judgments, {CMJI }. judgments and subsequently collect considered moral judgments. S will formulate a set of moral principles, {MP1}, as the so-called “systematization” of the beliefs in {CMJi }. But this is just the first attempt. It may not be fully satisfactory. S will continue to revise some of her considered moral judgments and some moral principles through the back-and-forth process of revision. In this way, S will arrive at the pair of sets, <{CMJn} and {MPn}>,3'4 which will be in the state of NRE. S will have formulated, during the process, several 3‘ DePaul, “Reflective Equilibrium and Foundationalism,” 62-63. 32 I use “{ . . . }” where DePaul used “(. . .)” to indicate a set of beliefs. 33 DePaul, “Reflective Equilibrium and Foundationalism,” 64. He says here, “It is likely that both {IMJ} and {CMJ i} will consist of beliefs in moral propositions of all levels of generality.” 51 {CMJ} and {MP} pairs: <{CMJl}, {MP1}>, <{CMJz}, {MP2}>, and so on. S’s beliefs may have a historically foundational structure when {CMJ 1} is the same as {CMJn}. As DePaul points out, however, it is unlikely that this situation would arise: “{CMJl} cannot serve as the foundation for the person’s moral beliefs, since it is most likely that he would no longer hold all the beliefs in this set, indeed, may not hold any of the beliefs in this set.”35 Furthermore, “some of the beliefs in {CMJn} would have been formed precisely because of their relation to some of the beliefs in {MPn}?36 Therefore, the charge of historical foundationalism is possible “only if one fails to distinguish {CMJI} from {CMJn}.”37 Another possibility of a historically foundational structure is if the intersection of {CMJI} and {CMJn} serves as a foundation. But, DePaul says, “although it is possible, and perhaps even likely, that the intersection of these two sets of beliefs would be non- empty for some persons, it could just as well be empty for others.”38 (My italics) Thus, the existence of such a non-empty set cannot be guaranteed. Second, with respect to psychological foundationalism, DePaul thinks, “it is quite probable that the person’s moral beliefs would have a psychological foundational structure” because “there is likely to be a central core of moral beliefs that the person will hold for no other reason which will also serve as reasons for the other beliefs that the ”39 person has. He thinks that the beliefs in {CMJn} may provide this core, but concludes that “this fact will have little to do with the person’s having followed that particular 3" The subscript “n” refers to narrow reflective equilibrium. 35 DePaul, “Reflective Equilibrium and Foundationalism,” 65. 3° DePaul, “Reflective Equilibrium and Foundationalism,” 65. 37 DePaul, “Reflective Equilibrium and Foundationalism,” 65. 3" DePaul, “Reflective Equilibrium and Foundationalism,” 65. 39 DePaul, “Reflective Equilibrium and F oundationalism,” 65. 52 method.”40 In other words, using the method of NRE need not result in a psychological foundational structure. In fact, I am sure that, for the person following the method of reflective equilibrium, there is no set of beliefs held for no other reasons. Coherentists cannot accept or justify a belief without considering whether it coheres with other beliefs. DePaul’s thoughts regarding psychological foundationalism are tied to his belief that the beliefs of the person following the method of NRE might possibly have a psychologically foundational structure. But this does not imply that the method of NRE is committed to psychological foundationalism. His point is that although there are no necessary relationships between the method of reflective equilibrium and the various forms of foundationalism, they are compatible. The beliefs of a person following the method of NRE are unlikely to have a methodologically foundational structure because her beliefs are unlikely to have a historically foundational structure. In addition, a methodologically foundational structure assumes that “a person’s considered moral judgments must have some epistemic privilege on their own?"1 But, DePaul says, “one need not make this assumption in order to provide a plausible account of the epistemic status of a person’s beliefs in narrow ”42 (My italics) reflective equilibrium. Similar points can be made with respect to epistemic foundationalism. The beliefs of a person following the method of NRE will not necessarily have a psychological foundational structure, as explained earlier. In addition, to have an epistemicly foundational structure, a set of beliefs must be held now due to no other beliefs, but serve ’0 DePaul, “Reflective Equilibrium and Foundationalism,” 65. 4' DePaul, “Reflective Equilibrium and Foundationalism,” 66. ‘2 DePaul, “Reflective Equilibrium and Foundationalism,” 66 53 as reasons for other beliefs, and have some epistemic privilege.43 But a person following the method of NRE will not say that a certain set of beliefs have some epistemic privilege. Instead, she will say that, because all beliefs are related one another, there is no set of beliefs that are held for no other beliefs. DePaul concludes, then, that even the method of NRE does not necessarily fall into some form of foundationalism. The above considerations even more obviously apply to the method of WRE. Therefore, Daniels’s claims based on the distinction between WRE and NRE may mislead us into thinking that although WRE is not committed to foundationalism, NRE is. DePaul’s defense of reflective equilibrium from the charge of foundationalism is valuable because it effectively shows how the back-and-forth process works. This process is one of the most important characteristics of the method of reflective equilibrium. It is important to note, however, that DePaul’s defense of reflective equilibrium stops short of denying compatibility between the two forms of reflective equilibrium and his various forms of foundationalism. We can see in the above exposition of his views that he is open to the possibility that the beliefs of a person following the method of NRE has a historically, psychologically, methodologically, or even epistemicly foundational structure. What he denies is the necessary relationship. 2.2. DePaul’s Claim of Compatibility If {CMJ 1} is the same as {CMJn}, the beliefs of someone following the method of NRE will have a historically foundational structure. Or, if the intersection of {CMJ 1} ‘3 DePaul uses ‘epistemicly’ instead of ‘epistemically.’ See his definition of epistemic foundationalism. 54 and {CMJn} is not empty and if all other beliefs are justified by their relation to the beliefs in this intersection, one’s beliefs would also have a historically foundational structure. If there is a set of core beliefs that are psychologically independent beliefs and if other beliefs are now held because of their relation to these core beliefs, one’s beliefs will have a psychologically foundational structure. If some one thinks that {CMJ 1} or the intersection of {CMJ1} with {CMJn} has some epistemic privilege, their beliefs will have a methodologically foundational structure. Finally, if the core beliefs that are held for no other beliefs and serve as reasons for other beliefs have some epistemic privilege, one’s beliefs will have an epistemicly foundational structure. DePaul illustrates how the method of WRE can be compatible with epistemic foundationalism: In order to hold that a person’s moral beliefs in wide reflective equilibrium have an epistemicly foundational structure with her considered moral judgments as the foundations one must assume that (1) the members of {CMJw} have a privileged epistemic status in their own right and the propositions in {MTw} have their epistemic status in virtue of their systematizing the proposition 5 in {CMJw}, and (2) the person now does not believe the members of {CMJw} because of any relation they bear to other propositions she believes, but now does believe the members of {MPw} because they best systematize the members of {CMJw} .44 In this passage, (1) is needed for pure epistemic foundationalism and (2) for psychological foundationalism. DePaul acknowledges that “assumption (2) does seem outlandish” because typically “a person’s moral beliefs in reflective equilibrium are interdependent” due to the rocess of revision.45 He tries, however, to resent the most lausible wa the beliefs of a P P P Y person following the method of WRE could have epistemicly foundational structure — 4" DePaul, “Reflective Equilibrium and Foundationalism,” 67. The subscript “w” refers to wide reflective equilibrium. 55 that is by introducing a set of beliefs drawn from each of {CMJw}, {MPw}, and {BTW} as foundations, instead of from simply {CMJw}. DePaul says: Surely there will be a set of propositions that a person holds in wide reflective equilibrium composed of members from each of {CMJw}, {MPw} and {BTW} that is the psychological foundation for the union of these three sets. This will be a set of beliefs that seems to the person to be more likely true than most of that person’s other beliefs in reflective equilibrium and, more importantly, consist of the reasons for which the person holds her other beliefs.46 His introduction of this set sounds as outlandish to me as assumption (2) above did to DePaul. But what he wants to say is that “there was no necessary connection between following the method of reflective equilibrium and forming moral beliefs with a foundational structure.”47 He finally says, “foundationalism and reflective equilibrium are not really positions on the same topic?“ This is revealed in the following passage. Others have been led to hold that there is some sort of simple necessary connection between employing the method of reflective equilibrium and foundationalism because they have failed to appreciate the richness of foundationalism and the freedom afforded to one by coherence methods of moral inquiry.49 (My italics) However, can this freedom be provided without degrading the method of reflective equilibrium? What gives a certain class of beliefs privileged epistemic status for one employing the method of reflective equilibrium? ‘5 DePaul, “Reflective Equilibrium and Foundationalism,” 67. 4" DePaul, “Reflective Equilibrium and F oundationalism,” 67. ‘7 DePaul, “Reflective Equilibrium and Foundationalism,” 64. ‘8 DePaul, “Reflective Equilibrium and Foundationalism,” 68. 56 2.3. No Compatibility DePaul nicely articulates how a set of beliefs changes with the process of revision within the framework of NRE or WRE. This effectively shows that the two methods of reflective equilibrium are not committed to foundationalism. But his claim of compatibility has the following difficulties. First, psychological foundationalism and historical foundationaism appear compatible with the two methods of reflective equilibrium, but they cannot be considered to be genuine foundationalisms. I did not use the term ‘basic belief’ in my explanation of DePaul’s historical and psychological foundationalism because he made no claim regarding a positive epistemic status E for the beliefs I am calling historically independent or psychologically independent beliefs. Although the definitions of these forms of foundationalism include a relation R holding between independent beliefs and dependent beliefs, it is not obvious that this relation R is the same as that found in foundationalist justificatory relationships. The relationship between independent and dependent beliefs is not identical to the relationship between basic and non-basic beliefs because independent beliefs are not identical to basic beliefs. There is no reason to think that independent beliefs must be basic beliefs because DePaul’s definitions do not include a description of their privileged epistemic status. We can legitimately question whether beliefs that were formulated or are now held independently of other beliefs are justified. Therefore, neither historical nor psychological foundationalism can be considered to be a full-fledged foundationalism. Second, historical and psychological foundationalism cannot guarantee ‘9 DePaul, “Reflective Equilibrium and Foundationalism,” 68. 57 foundational structures. DePaul’s notion of a relation R in the phrase, “some relation R such that (a) each belief in G that is not in F stands in R to some subset of the beliefs in F,” is not clearly defined as a foundational relationship. To have a foundational relationship, we must state additionally “the beliefs in F have a certain positive epistemic status E.” A relation R may be a relationship of mutual support without such a statement. His definitions, then, cannot prohibit a coherentist interpretation of a relation R. There is no reason to think a person’s beliefs must have a foundational structure if they satisfy DePaul’s definitions of historical or psychological foundationalism. Only when a form of foundationalism specifies a positive epistemic status E for independent beliefs, will it become methodological or epistemic foundationalism, either of which is genuine foundationalism. The issue, then, is whether someone following the methods of reflective equilibrium can say that a particular set of beliefs has a positive epistemic status E without appealing to other beliefs — that is, that it has a privileged epistemic status. I argue that such a person cannot. This leads us to the next difficulty. Third, for the methods of reflective equilibrium to be compatible with methodological and epistemic foundationalism, as explained just above, it must be assumed that a set of beliefs has a privileged epistemic status in its own right. But this requires us to give up the idea — important to the very concept of reflective equilibrium — that no beliefs are immune to revision. This idea implies that the methods of reflective equilibrium do not have any privileged beliefs. This assumption that some beliefs enjoy a privileged epistemic status, therefore, is inconsistent with the methods of reflective equilibrium. Therefore, DePaul’s claim of compatibility cannot be made without radically transforming the methods of reflective equilibrium. 58 Let us ask once more whether my conclusion is correct by considering the most plausible case of compatibility suggested by DePaul. When DePaul argues that a set of propositions composed of members from each of {CMJw}, {MPw}, and {BTW} can be psychological foundations for the beliefs of someone following the method of WRE, he must assume two things: (1) that this set of beliefs may be held for no other beliefs but serves as reasons for other beliefs, and (2) that the beliefs in this set have an epistemic privilege. Suppose Wendy is wondering whether a physician should respect his/her patient’s autonomous decision to refuse medical treatment. One belief she holds is the principle of autonomy, which states roughly “rational individuals should be permitted to be self- deter'minning.”50 This belief might have been formed because of her belief that individuals should have the right to shape their own lives according to their own values. But she does not remember if this was the reason for her holding the principle of autonomy. Instead, she accepts now the principle of autonomy because she believes that moral responsibility is impossible unless we are permitted to be self-determining. As seen in the distinction between DePaul’s historical foundationalism and psychological foundationalism, the history of belief formation may not always be reflected in the current structure of beliefs. It is possible that we may hold a belief for reasons other than those that led to our originally formulating it. Wendy followed the method of WRE to answer to the above question. She concluded that a physician should respect his/her competent patient’s refusal of treatment when the decision of the patient is autonomous. This conclusion fit with her background belief that human beings are happiest and most 50 Ronald Munson, ed., Intervention and Reflection: Basic Issues in Medical Ethics, 6th ed. (Wadsworth, 2000), 40. 59 dignified when they are allowed to pursue their own wishes, and with her particular moral judgment that Julie, an adult, mentally competent Jehovah’s Witness with no dependents, should be allowed to refuse a blood transfusion based on her religious beliefs. The conclusion cohered also with the belief that if a physician forces treatments upon someone without his/her informed consent, the physician is committing battery under current law and with the belief that a mentally competent, informed patient him/herself, and not the physician, is the person who can best decide what his/her values are. Now suppose Wendy has a core set of beliefs that includes the principle of autonomy, the particular moral judgment regarding the Jehovah’s Witness, and the belief about the connection between autonomy and happiness (or dignity) — that is, the set of beliefs from each of {CMJw}, {MPw}, and {BTW}. Following DePaul’s claim, suppose further that this set of beliefs is held for no other beliefs but provides reasons for other beliefs. This set may support the belief that a physician should get informed consent for human experimentation, the belief that physician-assisted suicide is sometimes morally permissible, and the belief that we should treat people as ends-in-themselves and not as a mere means, etc. I am not quite sure how to say that this set is now held for no other reasons. But let us suppose that Wendy has forgotten the beliefs that supported each of them. She has forgotten even the recent reason supporting the principle of autonomy — that is, that moral responsibility is impossible unless we are permitted to be self- determining. But she must say that this set of beliefs enjoys epistemic privilege for my belief system to have an epistemic foundational structure as DePaul’s assumption (2) shows. But, now Wendy finds a serious difficulty. How can she maintain that the 60 principle of autonomy has a privileged epistemic status? There seem to be two ways to respond. First, she might say that the principle of autonomy is self-evidently justified. But this is not consistent with the methods of reflective equilibrium, which says that there are no beliefs with privileged epistemic status. In addition, this belief was formed in relation with other beliefs, although she does not remember the process. Second, Wendy might say that the set of beliefs itself has epistemic privilege, rather than each belief of the set. On this view, although each belief can be justified by its coherence with other beliefs in the set, the set itself is held for no other beliefs and yet serves as reasons for other beliefs. But there remains the question of why this set of belief has special epistemic status. Wendy may say that this is because this set serves as reasons for other beliefs. But coherentists will say that because Wendy’s reason is also a belief, the set of beliefs cannot be held without any other beliefs. Therefore, Wendy’s set of beliefs is not basic or self-evidently justified. 3. The Method of Reflective Equilibrium and Modest Foundationalism 3.1 . Modest Foundationalism Roger P. Ebertz argues that the method of reflective equilibrium is in fact best construed as involving modest foundationalism.51 His claim is interesting because, as he points out, it is even stronger than DePaul’s claim.52 Ebertz has reissued the claim that the method of reflective equilibrium is committed to foundationalism, albeit a weaker form of 5' Roger P. Ebertz, “Is Reflective Equilibrium a Coherentist Model?” Canadian Journal of Philosophy 23, no. 2 (1993): 200. 61 foundationalism; one that attempts to weaken the status of foundational beliefs and to include coherence.53 Let us consider features of a weaker form of foundationalism before dealing with Ebertz’s claim. Robert Audi contrasts his moderate foundationalism with strong foundationalism as follows: Foundationalism can acknowledge a significant role for coherence in relation to justification I have in mind a kind of moderate foundationalism: a foundationalist view of knowledge or justification which (1) takes the justification of foundational belief to be at least typically defeasible; (2) is not deductivist, that is, does not demand that principles governing the inferential transmission of knowledge or justification be deductive; and (3) allows a significant role for coherence by requiring, not that inferentially justified beliefs derive all their justification from foundational ones, but only that they derive enough of it from the latter so that they would remain justified if any other justification they have were eliminated." Audi’s moderate foundationalism, then, incorporates coherence as one element of justification. He contrasts his moderate foundationalism with strong foundationalism, which, according to him, holds that inferentially justified non-foundational beliefs derive their justification from foundational beliefs, which are indefeasibly justified.55 In contrast to strong foundationalism, the moderate version allows for coherence to play two roles. First, it works negatively in relation to the justification of foundational beliefs in the sense that their justification may be defeasible when incoherence defeats the justification of a directly justified foundational belief.56 Second, coherence “raises the level of justification originally drawn from other sources to a level higher than it would 52 Ebertz, “Is Reflective Equilibrium a Coherentist Model?” 200. 53 Robert Audi uses ‘foundational’ and Ebertz uses ‘directly justified’ to refer to basic beliefs. 54 Robert Audi, Epistemology (Routledge, 1998), 205. 55 Audi, Epistemology, 206. 56 Audi, Epistemolog, 205. 62 have if those sources were not mutually coherent.”57 Coherence may improve justification. But moderate foundationalism denies that coherence is a basic source of justification. Therefore, Audi’s moderate foundationalism retains the asymmetry of justificatory direction because the ultimate source of justification for non-foundational beliefs comes from defeasible, but foundational, beliefs. Ebertz’s modest foundationalism is another attempt to weaken the status of foundational beliefs. He defines his modest ethical foundationalism this way: We can define modest ethical foundationalism as the view that ethical beliefs are justified when (i) some of these beliefs have a prima facie direct justification and (ii) all of the other beliefs are justified in a way that depends on their relationship to these directly justified beliefs."8 This kind of foundationalism can be called “modest” because it holds that there are prima facie justified foundational beliefs. “To say that a moral belief is prima facie directly justified, is to say that either (a) its justification is not derived solely from inferences from other beliefs, or (b) it is justified by inference from some non-moral belief(s).”59 A prima facie justified belief is held only if there are no other factors that undermine or defeat it. Despite this feature, this kind of foundationalism is still foundationalism because it retains an asymmetrical justificatory model. All other justified beliefs depend upon the prima facie directly justified beliefs. Therefore, Ebertz’s modest foundationalism retains a significant one-directional structure of justification. 57 Audi, Epistemology, 206. 58 Ebertz, “Is Reflective Equilibrium a Coherentist Model?” 201. 59 Ebertz, “Is Reflective Equilibrium a Coherentist Model?” 201. 63 3.2. Ebertz’s Claim Ebertz’s modest and Audi’s moderate foundationalisms have the following similarities. First, both weaken the status of foundational beliefs. Just as Ebertz’s prima facie foundational beliefs are held only if there are no other factors which undermine or defeat them, so Audi’s defeasible foundational beliefs are held only until incoherence defeats justification of the directly justified foundational beliefs. Second, both envision a largely asymmetrical justificatory structure, which moves from foundational beliefs to non-foundational ones. The only difference between them is that Audi carefirlly delineates the roles coherence plays in justification. Ebertz’s modest foundationalism might nonetheless include coherence. He says that if coherence is necessary for justification, it can be made consistent with modest foundationalism.60 Because their foundationalisms are very similar, and the differences are largely irrelevant when comparing them to the method of reflective equilibrium, 1 will focus from here out on modest foundationalism. Ebertz argues that the method of WRE is a model of modest foundationalism. He says, “As a person begins the reflective process[,] his considered moral judgments have a privileged status in the evaluation of other beliefs?“ Considered moral judgments as provisionally fixed points are said to function like prima facie foundational beliefs. In this respect, they “have a prima facie privileged justificatory status in the structure, a status which is not derived merely from their relationship to other beliefs.”62 Ebertz also stresses that in the reflective process, considered judgments may be defeated or thrown 60 Ebertz, “Is Reflective Equilibrium a Coherentist Model?” 201-202. 6' Ebertz, “Is Reflective Equilibrium a Coherentist Model?” 202. out.63 But if they do survive the process, according to Ebertz, they play a special justificatory role even in the resulting state of reflective equilibrium.64 Ebertz also appeals to the passage from Rawls cited by Hare and Singer as discussed earlier in this chapter. Rawls had said that there is a definite class of considered judgments against which conjectured principles can be checked.“ With this in mind, he argues that considered moral judgments serve as modest foundations because they play a special justificatory role in reflective equilibrium. Finally, Ebertz argues that, The reflective equilibrium model does take at least one step in the coherentist direction, by suggesting that coherence between beliefs is necessary for justification. But as we have seen this claim is consistent with modest foundationalism. Thus reflective equilibrium is a model not of coherentism but of modest foundationalism combined with the claim that coherence between beliefs is an additional necessary condition for justification."6 Ebertz, however, misses important differences between modest foundationalism and the method of WRE. 3.3. Important Differences Between the Method of Reflective Equilibrium and Modest F oundationalism The complaints I raised earlier regarding interpretations of Rawls that focus on merely one side of reciprocal support are true also of Ebertz’s claim. I contend that the foundationalist interpretation is problematic because Rawls clearly emphasizes mutual support. There are several important differences between modest foundationalism and the 62 Ebertz, “Is Reflective Equilibrium a Coherentist Model?” 202. 63 Ebertz, “Is Reflective Equilibrium a Coherentist Model?” 202. 64 Ebertz, “Is Reflective Equilibrium a Coherentist Model?” 203. ‘55 Rawls, A Theory of Justice, 51. Ebertz cites this passage on p.203 in his “ls Reflective Equilibrium a Coherentist Model?” 66 Ebertz, “Is Reflective Equilibrium a Coherentist Model?” 206. 65 method of reflective equilibrium. First, even if the method of reflective equilibrium may begin with considered moral judgments, it is not because they are foundations. Considered moral judgments are simply provisional starting points. If they were foundations, the method of reflective equilibrium would always start with considered moral judgments. As mentioned before, however, we can just as well start with moral principles or background beliefs if the context demanded. As long as the method does not always require considered moral judgments as starting points, they cannot be considered to be foundational beliefs. If we follow Ebertz’s view, all beliefs may be defeasible or prima facie foundations for coherentism. Foundationalists, however, will be unsatisfied with this —- they want inequality among beliefs. Second, if we start with considered moral judgments and they fail to fit with other beliefs, they will be revised or replaced. But defeasible or prima facie foundational beliefs will not be revised or replaced even as they are defeated in certain circumstances. They are still considered to be foundational beliefs and will not lose any justificatory status. So, foundational beliefs enjoy a special epistemic status that considered moral judgments do not. Third, considered moral judgments and any other beliefs as serving provisional starting points have merely initial credibility. When equilibrium is lost, beliefs considered for revision will have justification temporarily withdrawn until equilibrium is regained. Although considered moral judgments may enjoy some credibility as starting points, this is just at least temporary. We cannot say they are justified until we achieve equilibrium. When we achieve equilibrium, the justification of these considered moral judgments will, 66 if they survive revision, be based on their coherence with other beliefs, and not because of the credibility we gave them as starting points. There seems to be a difference in degree of justification between considered moral judgments as starting points and considered moral judgments after they survive the dialectical process of WRE. The justification of defeasible or prima facie foundational beliefs, on the other hand, does not change when they are either defeated or, when not defeated; they have their own justificatory status independent of other beliefs. Fourth, the above discussion assumes that considered moral judgments survive without revision. But if they are revised, they are not the same as before. What are justified in the newly achieved equilibrium are not the initial considered moral judgments, but new or revised considered moral judgments. Therefore, the view that considered moral judgments firnction like defeasible or prima facie foundational beliefs can no longer be held. This contrasts the foundationalist and the coherentist views regarding justification. Coherentists consider the process of moral justification to be more dynamic than do foundationalists. Finally, Ebertz says of considered moral judgments that, “If they do survive the process, we have no reason to believe they somehow lose their direct justification?"7 This reveals his view that considered moral judgments have direct justification. But proponents of the methods of reflective equilibrium need not agree. Considered moral judgments (or any other beliefs serving as starting points) may be beliefs that have already been held in equilibrium. Their credibility generally comes from their previous or current coherence with other beliefs, not from “direct justification.” The claims that the method of reflective equilibrium is committed to or 67 compatible with foundationalism are not plausible. I conclude that this method must be understood as coherentist moral reasoning. 67 Ebertz, “Is Reflective Equilibrium a Coherentist Model?” 202. 68 Chapter Three Moral Experience and Wide Reflective Equilibrium As a coherentist method of moral reasoning, the method of wide reflective equilibrium (WRE) must confront four issues standardly raised against coherentism. First, coherentism is criticized for being circular. Second, it is accused of being incapable of telling us which of two competing beliefs to accept when they achieve equilibrium in their respective belief systems. This question leads us to wonder how we could decide which of the competing belief systems is better. The problem of choosing between two internally coherent belief systems seems, then, also to be a problem for the method of WRE. Third, coherentism is criticized by those epistemological foundationalists for whom observation plays a special justificatory role regarding beliefs about the world. This is sometimes characterized as the problem of having “no contact with the world.” Relatedly, in the context of moral beliefs, we might ask whether the method of WRE can accommodate moral experience. More specifically, it is unclear whether the method of WRE can incorporate new moral experience or new moral beliefs. Fourth, coherentism is criticized for allowing the justification of even morally wrong beliefs so long as they cohere with other beliefs within a particular person’s belief system. An acceptable method of WRE must be able to distinguish incorrect from correct moral beliefs. By way of responding to the above difficulties, 1 will identify two criteria for moral justification in addition to coherence and will show how moral experience can be incorporated into the method of WRE. These adjustments are further developments of the notion of the method of WRE. I conceive the method of WRE as an interpersonal 69 epistemic activity. In the first section, I argue that mutual support, if circular, is not necessarily vicious circularity, and that the method of WRE can use comprehensiveness as a criterion for choosing between coherent belief systems. In response to the third difficulty, I suggest in the second section that the method of WRE can take moral experience into account as an input from the world in the formation of particular moral judgments. The inclusion of moral experience provides the method of WRE with a way to change or broaden moral understanding. Finally, I deal with the fourth problem in the third section. Construing the method of WRE as an interpersonal epistemic activity provides a way of denying the legitimacy of certain moral beliefs, even though they may cohere with a person’s other beliefs. This construal provides for another criterion of justification, in addition to coherence and comprehensiveness. It takes into account the number of moral beliefs that are considered to be unjustified in a more comprehensive and coherent working belief system, though they may be justified in a less comprehensive belief system. 1. The Problem of Choosing the Better Belief System 1.1. Mutual Support and Circularity If belief A is justified on the basis of belief B, if B is justified on the basis of C, and if C is justified on the basis of A, then this justification is circular. Foundationalists favoring one-directional justification understand mutual support in this way. When we are committed to a linear structure of inference, this interpretation cannot be avoided. 70 Coherentism, however, involves multi-directional justification, even though it may allow linear structures at the local level of justification. Belief A can be justified because B and C both support it, because B and C support each other, and because A supports B and C. In other words, when A, B, and C cohere with one another, A is justified by the acceptance of the group of beliefs, A, B, and C. Therefore, the complete justification of any one belief cannot ultimately be given without the justification of the whole belief system to which the belief belongs. According to Laurence BonJour: Inferential justification, despite its linear appearance, is essentially systematic or holistic in character: beliefs are justified by being inferentially related to other beliefs in the overall context of a coherent 1 system. He distinguishes a local level from a global level of justification: The epistemic issue on a particular occasion will usually be merely the justification of a single empirical belief, or small set of such beliefs, within the context of a cognitive system Whose overall justification is (more or less) taken for granted; we may call this the local level of justification. But it is also possible, at least in principle, to raise the issue of the overall justification of the entire system of empirical beliefs; we may call this the global level of justification.2 When the justification of a single belief is at issue, we identify premises that are acceptable in a particular context. Justification in this case is at the local level. However, if we cannot identify any acceptable premises and if beliefs offered as premises continue to be challenged, the challenge is to the entire belief system. The justification will then have to occur at the global level. In this case, the justification of a particular belief will finally depend upon the justification of the entire belief system - that is, its coherence.3 As BonJour says, “The apparent circle of justification is not in fact vicious because it is ' Laurence BonJour, The Structure ofEmpiricaI Knowledge (Harvard University Press, 1985), 90. 2 BonJ our, The Structure of Empirical Knowledge, 91. 3 Coherence is, in fact, a matter of degree. One belief system may be more coherent than another. I will deal 71 not genuinely a circle.”4 The same point can be made in a different way. Suppose that a belief turns out to be false. If we follow one-directional justification, any resulting revision would be fairly uncomplicated because we would be required to simply trace back the thread of justification. But the infeasibility of this scenario becomes apparent once we consider the difficulties of revising a scientific theory. Quine says this: Total science is like a field of force whose boundary conditions are experience. A conflict with experience at the periphery occasions readjustrnents in the interior of the field. Reévaluation of some statements entails reevaluation of others, because of their logical interconnections... But the total field is so underdetennined by its boundary conditions, experience, that there is much latitude of choice as to What statements to reevaluate in the light of any single contrary experience. No particular experiences are linked with any particular statements in the interior of the field, except indirectly through considerations of equilibrium affecting the filed as a whole.5 This passage shows the difficulties of picking out a single statement for correction. This is because beliefs are interwoven and mutually supporting. The structure of our beliefs, in my view, is like a web, or rather, like multiple webs. Our acceptance or rejection of a particular belief is sometimes related to the acceptance or rejection of the whole belief system to which it belongs. That is, the justification of a particular belief may be challenged at the global rather than the local level. For example, whether or not you accept the belief that God exits may be related to whether or not you accept an entire basically theistic belief system. with this topic in the subsequent subsection. 4 BonJour, The Structure of Empirical Knowledge, 92. 5 Willard Van Orrnan Quine, From a Logical Point of View, 2"d ed. (Harvard University Press, 1980), 42-43. 72 1.2. The Problem of the Choice Among Coherent Belief Systems When we accept the view that justification ultimately depends on the coherence of an entire belief system, we may be faced with the second problem raised above, that of choosing between coherent belief systems. This problem is not only a problem between belief systems. Suppose two different beliefs regarding the same state of affairs achieve WRE in their respective belief systems. Should we say that both are equally justified? Or can we still compare the two? This problem of choosing between two competing beliefs, each of which achieves equilibrium in its particular belief system, is related to the problem of choosing between two internally coherent belief systems. If we can solve the latter, we will be able to solve the former. By what criteria can we say that one belief system is better than another, and thus that one moral belief is better than another? An answer is comparative coherence. We can comparatively assess the coherence of belief systems. Because coherence is a matter of degree, justification is also a matter of degree. BonJour says the following regarding the notion of coherence: First, coherence is not to be equated with mere consistency; second, coherence has to do with the mutual inferability of the beliefs in the system; third, relations of explanation are one central ingredient in coherence, though not the only one; and, fourth, coherence may be enhanced through conceptual change.6 The last point has already been discussed in terms of the back-and-forth process for revision in the method of WRE. From the second and third, we can see that inference and explanation are important relationships between beliefs, which improve the degree of coherence. Because coherence is thought to be a function of inference and explanation, 73 the first point is obviously true. To show this, BonJour says the following regarding two sets of propositions, A and B: ,9 ‘6 A contains “this chair is brown, electrons are negatively charged,” and “today is Thursday.” B contains “all ravens are black,” “this bird is a raven,” and “this bird is black.”7 Both sets of propositions are free of contradiction. However, the consistency of set A results from the fact that its component propositions are almost entirely unrelated to one another, whereas the consistency of set B results from that fact that the component propositions significantly fit together and reinforce one another. Because of this, set B is much more coherent than set A. As BonJour says, “Coherence must involve some sort of positive connection among the beliefs in question, not merely the absence of conflict.”8 He adds: The coherence of a system of beliefs is increased by the presence of inferential connections between its component beliefs and increased in proportion to the number and strength of such connections.9 Although he is speaking here specifically about inferential connections between beliefs, this must be true also of explanatory connections. Furthermore, because coherence is increased or decreased in proportion to the number and strength of inferential and explanatory connections among beliefs, this will be true also of justification. Therefore, we will be able to compare one belief system with another if we can assess the number and strength of such connections among beliefs. When faced with two competing moral beliefs, we may also compare one with the other in terms of justification.10 6 BonJour, The Structure of Empirical Knowledge, 95. 7 BonJ our, The Structure of Empirical Knowledge, 96. 8 BonJour, The Structure of Empirical Knowledge, 96. 9 BonJour, The Structure of Empirical Knowledge, 98. '0 I acknowledge here that the suggestion for assessing inferential and explanatory connections among beliefs is quite abstract and theoretical. We will ultimately have to develop more practical, concrete, and 74 But one question remains. Suppose two competing beliefs regarding a single state of affairs achieve equilibrium in their respective belief systems. Let us stipulate that the two different belief systems enjoy the same degree of coherence. Must we say that both beliefs are equally justified or, can we still compare one with the other? 1.3. Comprehensiveness as an Additional Criterion The problem of being unable to compare the equally coherent belief systems comes from the view that coherence is the only criterion of justification. But this view is too narrow because coherence is not the sole criterion of justification. The distinction between pure coherentism and impure coherentism may help here. If a cohemtist accepts coherence as the only justificatory criterion, we can call this view pure coherentism. If a coherentist accepts other criteria in addition to coherence, we can call this view impure coherentism. I argue for the addition of the criterion of comprehensiveness here, and will present another criterion in the final section. Suppose a moral inquirer, Kelly, considers the issue of experimentation on humans and concludes that no human experimentation should be allowed. Let us suppose firrther that his conclusion coheres with his other beliefs, including: “The case of the ,9 6‘ Tuskegee syphilis was morally wrong, the Nazi’s human experimentation cannot be justified,” “human experimentation uses persons as mere means,” “the principles of nonmaleficence and autonomy as middle level moral principles should be observed in medical ethics,” and so on. But suppose that the principle of utility is not part of his belief system. So, even if he employed the method of WRE, he might not take into account the contextual ways of assessing degrees of coherence. 75 considerable utility of human experimentation. In addition, if he does not have a relevant background belief —— that a regulation exists requiring some proof of safety before a new treatment or medication can be made available to the public, then he would have no idea of the necessity of human experimentation. Even though Kelly’s conclusion fits nicely with his other beliefs, his belief system is not sufficiently comprehensive to be plausible. We would not be satisfied with his conclusion because the conclusion enjoys equilibrium in his narrow belief system. When two beliefs conflict, the one that coheres in the more comprehensive belief system will be more justified than the one that coheres in the less comprehensive belief system. The use of comprehensiveness as a criterion for the choice of belief systems is not new. Thomas Kuhn, for example, mentions “scope” as a criterion of theory choice in science.11 According to him, a good scientific theory should have broad scope. This point can also be applied to the choice of belief systems including moral beliefs, and thus to the choice of moral beliefs themselves. The coherence of a belief system, as BonJ our says, will be increased in proportion to the number of connections among its beliefs.12 More comprehensive belief systems will usually have more beliefs than less comprehensive ones. While it is true that a system having more beliefs will not always have more connections among its beliefs than a system with fewer beliefs, such will more likely be the case assuming each does not have contradictories. Of course, there may sometimes be conflicts between two belief systems, one of which is more coherent, the other more comprehensive. However, the problem of choosing between two competing coherent belief H Thomas S. Kuhn, “Objectivity, Value Judgment, and Theory Choice,” in Philosophy of Science, edited by Martin Curd and J. A. Cover (W.W. Norton & Company, 1998), 103. 76 systems raises the question of whether we can compare one belief system with the other when they are equal in terms of degree of coherence. I must show that comprehensiveness can be used as an additional criterion of choice when there is no difference in degree of coherence between the more and less comprehensive belief systems. If a belief is justified in a more comprehensive coherent belief system, it can be thought to be withstanding some possible defeaters. The less comprehensive belief system may exclude beliefs that can be defeaters. Thus, choosing the more comprehensive coherent belief system is less risky when we consider the possibility of defeaters. That a belief withstands possible defeaters does not directly contribute to the degree of coherence, but it can be considered an important value in justification. The same point seems to be implied by the comparison between the methods of wide and narrow reflective equilibrium. Beliefs may have more chances to be revised and tested against one another in the method of WRE than in the method of narrow reflective equilibrium (NRE). The method of WRE exposes a belief to the possibility of more extensive revision than the method of NRE. Because of this, a belief that is justified in WRE can be considered to be more reliable, and then more justified than a belief that is justified in NRE. Similarly, a belief that is justified in a more comprehensive coherent belief system is more justified than one in the less comprehensive coherent system. Therefore, the degree of justification will be increased or decreased in proportion to the degree of comprehensiveness. '2 BonJour, The Structure of Empirical Knowledge, 98. 77 2. The Problem of No Input from the World 2.1. Naive Inquirer’s Problem and Moral Experience Another objection to coherentism is that there is no input from the world or contact with the world. This is often called the “isolation objection.” When coherence and comprehensiveness are the only criteria of justification, coherence can be maintained in a more or less comprehensive belief system even though it remains isolated from the world. This problem can be articulated in the context of moral reasoning as follows. An inquirer may stick to some limited number of moral beliefs — some or all of which might be wrong — so long as they all cohere, thereby not changing or broadening his/her narrow belief system in the light of contact with or experience of the real world. Michael R. DePaul describes this problem through the example of a naive inquirer.l3 A sheltered naive inquirer (in the sense that he/she has a very limited experience) may remain naive even after he/she has followed the method of WRE. This sheltered naive person may remain isolated from experiences that would otherwise force him/her to change or broadening his/her belief system. With this example, DePaul intends to show the limitations of the method of WRE and to emphasize the importance of formative experience, which is one type of experience leading us to form or alter our beliefs. According to DePaul, both formative experience and reasoning experience can lead a person to form or alter moral his/her beliefs.l4 Reasoning experience involves “such things as constructing an argument from propositions one already believes, ‘3 Michael R. DePaul, Balance and Refinement: Beyond coherence methods of moral inquiry (Routledge, 1993), 148-160. 78 appreciating a counterexample to a moral principle one accepts, or seeing a proposition ”15 (My italics) This experience is, as he says, “the engine as obvious upon considering it. that drives reflective equilibrium, and every other philosophical method as well.”16 (My italics) In contrast, formative experience is “experience that leads to alteration of belief, but does not fit the mold of reasoning experience?” He admits that he cannot give a more explicit definition.18 But he presents examples of experiences that can either cause formative experience or are themselves formative. These include direct real-life experiences such as a person’s: “actually performing an action he considers right or wrong, living with a person who has a certain virtue or vice, and closely interacting with people who value markedly different things.”19 He also lists indirect (or vicarious) experiences such as reading a novel, watching a play or film, memorizing a moving poem, listening to a piece of music, or viewing a painting.20 Through indirect experience, we may better understand someone who must make a difficult moral decision or who is strongly committed to certain moral principles. DePaul himself did not mention “moral experience” in his discussion of formative experience. The experiences he lists above can, though, bring about moral experiences. By ‘moral experience’ I mean experience that leads us to produce moral beliefs, views, or emotions. Thus, DePaul’s formative experience may be called moral experience. A naive inquirer who has had reasoning experience that brings his/her beliefs into equilibrium remains naive, according to DePaul, because he/she lacks the formative H DePaul, Balance and Refinement, 140. '5 DePaul, Balance and Refinement, 140-141. '6 DePaul, Balance and Refinement, 141. '7 DePaul, Balance and Refinement, 144. '8 DePaul, Balance and Refinement, 145. '9 DePaul, Balance and Refinement, 144. 2° DePaul, Balance and Refinement, 144. 79 experiences (or moral experiences) that might correct or change his/her moral beliefs. A na'ive inquirer’s beliefs remain inadequate because his/her initial formative experiences have been inadequate. Therefore, DePaul argues that the naive inquirer must consider the matter of formative experiences and then must engage in them where appropriate.“ DePaul’s claim is important in two aspects. First, moral experience is one of the important ways of changing or forming moral beliefs. Reasoning experience alone may not be sufficient for the task. DePaul seems to give more weigh to formative experience over reasoning experience. He says, “A person is more likely to change one of her beliefs as a result of a formative experience.”22 Second, moral experience plays an important role in providing contact with the world, as does experience in general. As observation and experimentation play important roles in science, so moral experience is important in ethics. Virginia Held says that moral inquiry “should be seen as a process of continual adjustment of theory in the light of moral experience, as well as of particular judgments and actions in the light of theory.”23 I agree with the necessity of moral experience in moral reasoning. The method of WRE faces a challenge in incorporating moral experience. When the method does try to incorporate moral experience, arises another problem that moral experience is often accompanied by emotions or feelings. As DePaul says: While emotions are somehow connected with formative experience, and they are crucially involved in our responses to theater, film, and the like, the connection between formative experience and emotion must be 2' DePaul, Balance and Refinement, 173-174. DePaul calls his method of reasoning which incorporates formative experience the method of balance and refinement. He thinks this method goes beyond coherence method of moral inquiry. 22 Michael R. DePaul, “Naivete and Corruption in Moral Inquiry,” Philosophy and Phenomenological Research 48, no.4 (1988): 621. 23 Virginia Held, Feminist Morality (The University of Chicago Press, 1993), 29. 8O complicated and contingent.24 Although the connection is contingent, moral judgments often involve emotion in some way or other. Emotions like sympathy, empathy, compassion, hopefulness, pity, anger, fear, hatred, desire for revenge, and indignation are ofien components of complex moral j udgrnents. I will call such judgments aflective moral judgments to indicate that they have emotional as well as more purely cognitive aspects or components. Thus, the second question is whether affective moral judgments should be included in the framework of WRE. 2.2. Emotion and Judgments The method of WRE can, I believe, incorporate moral experience. Particular moral judgments, which in the framework of WRE belong to a set of considered moral judgments, play a role in making contact with the world. Particular moral judgments may often be experiential judgments — those made in the light of first-hand experience. But the issue now is whether affective judgments (or beliefs) should be included in the framework of WRE. Held thinks that affective moral judgments will be left out of the set of considered moral judgments. She notes that Rawls distinguished considered judgments from judgments “given when we are upset or frightened, or when we stand to gain one way or the other ..?25 Held infers from this that Rawls excludes affective moral judgments. So, she says, “He counsels us to admit into our pool of ‘considered judgments’ only those 2‘ DePaul, Balance and Refinement, 145. 25 John Rawls, A Theory of Justice, revised ed. (Harvard University Press, 1999), 42. 81 which have been laundered of their entanglements with self-interested or other emotional colorings.”26 Emotion has often been assumed to be irrational. As Catherine Z. Elgin says, “That reason and passion are antithetical has long been an article of philosophical faith, less often argued than assumed.”27 Beliefs have been considered to be cool, calm, settled convictions whereas emotions or feelings have been thought to be visceral, volatile, violent agitations.28 Kenneth R. Hammond, a psychologist, expresses this view as follows: “Emotions are detrimental to good judgment, and good judgment demands cool reason devoid of emotion.”29 But, Hammond notes that some neuroscientists argue for the interdependence of emotion and reason. According to him, emotions depend upon reason because “the emotion we experience is dependent on our judgment of the situation; first we judge, then we experience emotion.”30 Emotions involve beliefs as well as judgments. Elgin provides an example: “Fiona’s fear of frogs is cognitive because it imbeds her belief that frogs are dangerous?“ This shows how reason (or cognition) affects emotion. If her belief about frogs being fearsome does not change, neither would her emotion. But, does reason depend on emotion? Antonio Damasio, a neuroscientist, says: Reason may not be as pure as most of us think it is or wish it were, that emotions and feelings may not be intruders in the bastion of reason at all: they may be enmeshed in its networks, for worse and for better.32 26 Held, Feminist Morality, 28. 27 Catherine Z. Elgin, Considered Judgment (Princeton University press, 1996), 146. 28 Emotions cannot be identified with feelings even if emotions are often called feelings. For example, the two emotions of pride and admiration can feel the same. See Elgin, Considered Judgment, 147-148. 29 Kenneth R. Hammond, Judgments Under Stress (Oxford University Press, 2000), 17. 3° Hammond, Judgments Under Stress, l9. Harmnond attributes this to Richard S. Lazarus and Bernice N. Lazarus, Passion and Reason: Making Sense of Our Emotions (Oxford University Press, 1994), 199. 3' Elgin, Considered Judgment, 146. 32 Antonio R. Damasio, Descartes'Error: Emotion, Reason, and the Human Brain (G. P. Putnam, l994), 82 Damasio noticed that Elliot, a patient of his with frontal lobe damage, performed normally on intelligence tests, but could no longer make rational choices or prioritize tasks.33 Elliot, it was discovered, was unable to feel. He did not react to pictures of gruesome accidents and natural disasters. Damasio found that other patients with similar brain damage had the same combination of impaired reason and impaired affect. This observation seems to show that practical reasoning involves emotion. At the very least, we can see that emotion and reason are interconnected. Elgin explains how emotion affects reason. Emotion fulfills a cognitive function in that it makes us pay attention to circumstances in a way we did not before. According to her, emotions affect “both the configuration and the constitution of a system of thought” by “highlighting some aspects of the domain, obscuring others, [and] engendering relations of relevance and irrelevance,” and emotions often “heighten awareness, redirect attention, and sensitize their subjects to factors that had previously eluded them (and others)?34 From the above considerations, I conclude that emotions depend upon beliefs and judgments of circumstances, but also affect other judgments and beliefs. But, do all emotions contribute to our cognition in a positive way? As Held mentions, emotions such as empathy, concern for others, hopefulness, and indignation may be crucial in developing appropriate moral positions.35 But “there are,” she continues, “morally harmful emotions, such as prejudice, hatred, desire for revenge, blind egotism, and so x11. 33 My description of his observation is based on “I feel, Therefore I am,” New York Times, 19 April 2003. His findings are originally in Descartes 'Error, 34-79. 34 Elgin, Considered Judgment, 149-150. 35 Held, Feminist Morality, 3o. 83 forth.”36 Even these emotions, however, can function cognitively because they also lead us to pay attention to factors we might otherwise overlook. Our worry is, however, whether such emotions negatively affect our moral understanding. Morally harmful emotions may often provide misleading orientations and misinterpretation, as sometimes do anxiety, fear, and excitement. But the potential negative cognitive effect of some emotions should not induce us to eliminate all emotion from moral reasoning. Our visual and auditory systems may sometimes mislead us; they do not always function correctly. But it is unreasonable to exclude all observational reports in epistemology. The same is true of emotions. Therefore, our moral reasoning should incorporate affective judgments because they may deliver correct information. But we want to make sure we incorporate correct judgments. When Daniels says that we collect initial particular moral judgments and filter them to yield considered moral judgments,37 he is making the same point. When Rawls identifies considered moral judgments with judgments “in which our moral capacities are most likely to be displayed without distortion,”38 he may intend not to get rid of all affective judgments, but to exclude the incorrect affective judgments. What are in need of being filtered out are incorrect judgments, not emotion. Snap or visceral judgments are likely to be incorrect judgments. Considered judgments are those that do not, on the surface, unduly favor ourselves and that are made in circumstances in which factors that usually distort our judgment — our being frightened, upset or enraged — are absent. This is not to say that considered judgments 3° Held, Feminist Morality, 30. 37 Norman Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” The Journal of Philosophy 76 (1979): 258. 38 Rawls, A Theory of Justice, 42. 84 must be free of emotion but rather that they must not be judgments distorted by feelings and emotions. Considered judgments, however, have at best only a prima facie claim on us. Thus, we often consider whether or not our moral judgments are justified. Held says, “Once one has arrived at a particular judgment, one needs to bring it into coherence with other related judgments.”39 One’s considered judgments obtained through filtering initial affective judgments must also be justified to become a new moral belief of his/her belief system. According to Elgin, deliverances of emotion are “initially tenable but must promote reflective equilibrium to qualify as firlly tenable.”4O The inclusion of affective moral judgments in the framework of WRE thus does not mean that they are foundational beliefs. The justification of these judgments entirely depends upon their coherence with other beliefs. Any justificatory privilege is not given to these considered moral judgments. The method of WRE includes initial affective particular moral judgments, but their moral justification will be withheld until they achieve equilibrium. 3. The Method of WRE as an Interpersonal Epistemic Activity 3.1. Reasoning as an Interpersonal Epistemic Activity DePaul’s distinction between reasoning experience and formative experience is valuable. This shows the importance of moral experience in moral reasoning. In order to 39 Held, Feminist Morality, 27. 40 Elgin, Considered Judgment, 155. 85 have contact with the world, the method of WRE as moral reasoning had to incorporate moral experience. We should note, though, the limitations of moral experience and DePaul’s formative experience in changing or forming moral beliefs. Even though a naive moral inquirer may have a certain moral experience (or formative experience), this experience still may not bring about the needed correction or broadening of his/her beliefs. Such experiences cannot guarantee that one will correct or broaden one’s beliefs. My reasons are as follows. DePaul seems to presuppose that we can have common moral experiences that lead us to correct particular moral judgments or moral beliefs. In other words, he seems to presuppose that we can universally share the same moral experience. I do not deny this possibility or the possibility of correcting or revising some beliefs through moral experience. For example, if we watch a film, for example, Schindlers list, showing the atrocities of Nazi death camps, we might all see the cruelty that was involved. But our experience is not likely to be formed independently of beliefs or perspective. If we had different background beliefs, ways of life, or worldviews, our moral experience may differ. I acknowledge that we may have the same formative experience — however it may not result in the correction or formation of the same belief. Therefore, there are limitations to formative experience (or moral experience), though I acknowledge the important roles it plays. This means that we must find another way of correcting or broadening an inquirer’s belief system. To this end, I place more emphasis on the role of reasoning experience than does DePaul. He explained reasoning experience as “constructing an argument from propositions one already believes, appreciating a counterexample to a 86 moral principle one accepts, or seeing a proposition as obvious upon considering it?“ (My italics) DePaul thereby focuses on an individual’s previously held beliefs. But reasoning experience cannot and should not be restricted to these. I think of reasoning experience as experience one has when trying to construct and evaluate arguments, whether they are one’s own arguments or someone else Is. A reasoner should take seriously competing arguments and possible objections, and seek responses to them in order to make his/her conclusions the most plausible. This attempt will often lead us to discover new information from other people’s arguments, and to then broaden or revise our belief system. Thus, reasoning should be understood to be an interpersonal rather than intrapersonal epistemic activity. Providing reasons has an interpersonal aspect. Although people do not always ask us to provide reasons for our beliefs, we often conceive others as hypothetical critics. Thus, there are social aspects to reasoning. When we construe reasoning as an interpersonal epistemic activity, reasoning experience includes not only considering competing views and responding to possible objections, but also actually participating in interpersonal activities such as give-and-take discussions, conferences, and publication. The latter can be called “actual” interpersonal epistemic activities whereas the former are “hypothetical” interpersonal epistemic activities. In hypothetical interpersonal activities, one alternately plays the roles of arguer and defender. Thus, both actual and hypothetical epistemic activities are interpersonal. We try to correct or revise our beliefs by engaging opposing views through either “actual” or “hypothetical” interpersonal activities. The consideration of other competing views and possible objections pushes a moral inquirer to overcome his/her narrow view in order to respond to objections. 4' DePaul, Balance and Refinement, 140- 14 l. 87 Therefore, the method of WRE should be understood to be an interpersonal rather than a purely intrapersonal epistemic activity. Let me call tentatively the method of WRE construed as an interpersonal activity the interpersonal method of WRE and the method construed as an intraperosnal activity the intrapersonal method of WRE. When one following the interpersonal method of WRE tries to give an answer to a certain ethical issue, one tries to reach a conclusion that is more coherent, from one’s point of view, with one’s belief system than any alternative, by reviewing other’s views, attempting to respond to the possible strongest objections, and revising or broadening his/her belief system. One following the intrapersonal method of WRE may easily arrive at a conclusion that reaches equilibrium within his/her narrow belief system without engaging interpersonal epistemic activities. The intrapersonal method of WRE may preserve equilibrium obtained within a narrow system of beliefs already held. Thus, DePaul’s na’r’ve inquirer who follows the intrapersonal method remains na'ive not merely because he/she lacks formative experience that would correct the wrong belief, but also because that he/she lacks reasoning experience that would be brought about by the interpersonal method of WRE. The interpersonal method of WRE will make a reasoner keep his/her belief system as coherent and comprehensive as possible. This method requires a reasoner to improve the degree of justification of his/her conclusion, by making his/her belief system as coherent and comprehensive as possible, because justification is increased or decreased in proportion to degree of coherence and comprehensiveness. A conclusion that reaches equilibrium through the intrapersonal method may not cohere with a more comprehensive coherent belief system, and then will not be justified in this system. 88 This leads me to an answer to the fourth issue raised above concerning coherentism. The worry was whether we must take even wrong or bad moral beliefs to be justified so long as they cohere with other beliefs within one’s personal belief system. I will provide an additional criterion for choosing among competing belief systems as an answer to this question. (From now on, I will call the interpersonal method of WRE simply the method of WRE.) 3.2. An Additional Criterion For Choosing Between Competing Belief Systems Can we say that a person’s belief is not justified by appealing to incoherence with commonly accepted beliefs? Can we say that some moral beliefs are wrong, bad, or unsatisfactory even though they cohere with other beliefs in that particular person’s belief system? Beauchamp and Childress say that a substantive body of judgments and principles that cohere could be morally unsatisfactory.42 For example, the so-called Pirates’ Creed of Ethics or “Custom of the Brothers of the Coast,” is a coherent set of rules governing mutual assistance in emergencies and penalties for prohibited acts. This set of rules, although coherent, is morally unsatisfactory. This demonstrates, according to Beauchamp and Childress, the importance of considered judgments with which the method of WRE begins. But on what ground can we say that these questionable beliefs are unsatisfactory? Viewing the method of WRE as an interpersonal epistemic activity seems to provide an answer. Even if a moral belief coheres with other beliefs within a less comprehensive 42 Tom L. Beauchamp & James F. Childress, Principles of Biomedical Ethics, 4th ed. (Oxford University Press, 1994), 24. 89 belief system, we may say that it is wrong, bad, or unsatisfactory when it is not justified in a more comprehensive coherent belief system. By a more comprehensive coherent belief system I mean an actual or “working” comprehensive and coherent belief system. Without this specification, one may simply build a more comprehensive belief adding some irrelevant but logically consistent beliefs to the less comprehensive belief system that holds a problematic moral belief. Again coherence is not, then, the only criterion of justification. I present as another criterion of choice among belief systems the number of incorrect moral and non- moral beliefs in system A as identified in a more comprehensive and coherent working belief system B. By incorrect non-moral beliefs I mean false beliefs. Incorrect moral beliefs refer to beliefs that are not justified in a more comprehensive coherent working belief system. We can choose between competing belief systems by comparing them with a third, more comprehensive and coherent working belief system if one is available. Kuhn mentions “accuracy” as a criterion for scientific theory choice."3 He says, “Consequences deducible from a theory should be in demonstrated agreement with the results of existing experiments and observations?“ Similarly, with respect to the justification of a moral belief, agreements between the belief system of which it is a part and a more comprehensive coherent working belief system can be considered to be of epistemic value in moral justification. If one belief system has more beliefs than another in agreement with a more comprehensive coherent working belief system, the former system will be better. Furthermore, the number of incorrect moral and non-moral beliefs can serve as a "’3 Kuhn, “Objectivity, Value Judgment, and Theory Choice,” 103. ‘4 Kuhn, “Objectivity, Value Judgment, and Theory Choice,” 103. 90 criterion for choosing among competing beliefs. We should consider one belief to be more justified than another if the former is part of a belief system that has fewer incorrect moral and non-moral beliefs than is the other. Suppose two sets of beliefs, A and B, have almost the same degree of coherence and comprehensiveness. But suppose fiirther that set A has more incorrect beliefs than set B in the light of a third working belief system that is more comprehensive and coherent than sets A and B. The conclusion that coheres with those incorrect beliefs in set A will be said to be less justified than the conclusion in set B. If some background beliefs or some moral beliefs are incorrect in the light of a more comprehensive and coherent working belief system, a conclusion that coheres with those beliefs will be less justified. 3.3. Implications and Limitations of the Three Criteria The two criteria (comprehensiveness and the number of incorrect moral and non- moral beliefs) may make it possible to compare two competing moral beliefs that have each reached their own state of WRE in their respective systems. We need not always say that both are equally justified. We may still distinguish one from the other on the basis of the two additional criteria. I admit however that together with coherence these two additional criteria do not exhaust the list of possible criteria for choosing among beliefs systems. But I can say at least that coherence is neither the sole criterion of moral justification nor the sole criterion for choosing among belief systems. As the addition of comprehensiveness indicates, I am not saying that coherence is the supreme or only value. More importantly, 91 I do not value coherence at any cost. We will not value an oversimplified coherent view over a more complex and comprehensive, but less coherent belief system. For example, there are two different, overarching ways of describing human behavior. One is a description in terms of “folk psychology” including the vocabularies of moral responsibility and right and wrong. The other is purely scientific and deterministic. We talk about freedom and responsibility in the first and the various causal determinants of human behaviors in the second. The two general kinds of description do not always mesh or cohere. If we valued only coherence, we would feel forced to choose once and for all between one or the other. As practical beings we cannot do this, however, because each is in certain contexts indispensable. Martin Benjamin says this: Looking backward, the impersonal or scientific perspective is useful for understanding the causal determinants of our own and other people’s past behavior. We are then, however, presented from the personal or pragmatic perspective with a new, forward-looking choice. “What do we make of what we have learned? What should we do in the light of our newly acquired causal understanding?” These are open, forward-looking questions that cannot be conclusively answered by backward-looking considerations."5 No one has yet come up with a way to integrate the two ways of talking. They do not perfectly cohere. Despite unsettling juxtaposition of the two ways of talking, we would not reject one or the other simply to achieve greater coherence. Furthermore, there may be conflicts among the above three criteria in some circumstances. One belief system might be preferable due to its degree of coherence while another might be deemed better because of comprehensiveness. Or, one system might have a couple of incorrect beliefs, but may be much better than its competitors in terms of coherence and comprehensiveness. Thus, I do not say that the use of the three ’5 Martin Benjamin, Philosophy and This Actual World: An Introduction to Practical Philosophical Inquiry 92 criteria will always lead to mechanical or algorithmic choices among belief systems. Judgment, as I argue in Chapter Six, is indispensable. Even if the method of WRE were to include these two additional criteria, therefore, there will be some competing reasonable comprehensive belief systems that cannot readily be compared with one another. If, for example, different standards of coherence are intrinsic to two competing belief systems, the systems may be incommensurable. Thus, I acknowledge that there may be some difficulty in comparing the coherence of competing reasonable belief systems. In addition, there will be many comprehensive belief systems that contain different reasonable philosophical or religious doctrines. We are sometimes faced with examples of different competing reflective equilibria among well-formulated arguments grounded in two competing reasonable comprehensive coherent belief systems, especially when we deal with ethical issues in health care. For example, we continue to be faced with individuals with different reasonable competing reflective equilibria regarding abortion and physician-assisted suicide. I acknowledge that we may not always be able to decide which belief or belief system is better, even if we have the three criteria discussed above. There may exist competing reasonable belief systems that are almost the same with respect to the three criteria or are hard to compare one another with the use of the three criteria. I have not argued that the three criteria can always tell us which belief or belief system is preferable. This aspect is captured by reasonable pluralism, which states that there are many, occasionally conflicting, reasonable philosophical or religious doctrines. This kind of pluralism is not to be identified with objectionable forms of relativism. I will deal with (Roman & Littlefield Publishers, Inc., 2003), 94. 93 this topic in the next chapter. 94 Chapter Four Wide Reflective Equilibrium and Relativism Some moral beliefs can be understood to deliver judgments of the tightness or wrongness of human conduct, whether particular actions or social customs. Other moral beliefs express moral principles or values.I Together these concerns are captured as the moral judgments and the moral principles or rules in the framework of wide reflective equilibrium (WRE). Ethical or meta-ethical theories (as meta-beliefs about moral beliefs) belong among the background beliefs or theories in the framework. When the justification of moral beliefs is supposed to depend upon their coherence with other beliefs in a belief system, we may find that there are incompatible or conflicting moral beliefs which are each in equilibrium within a different belief system. These belief systems may be held by individual persons or by particular groups. One might conclude that the lightness of our conduct or the truth of our moral beliefs depends upon a belief system. However, there are many different belief systems. Thus, one may ask further: “Does not the method of WRE lead to relativism?” Kai Nielsen puts it this way: “The method of appealing to considered judgments in Wide Reflective Equilibrium has been thought to have unwelcome relativistic or ethnocentric implications.”2 The question of relativism, however, should be more carefully specified. After all, there are several kinds of relativism with various implications, not all of which are necessarily unwelcome. Thus, a more accurate set of questions will include: Does the N 0' fl " l Some moral beliefs are over such moral content as "we ought to do x, x is morally right, x is morally pernrissible," "we should not do x," "x is morally wrong," and "x is not morally permissible." The first three expressions have to do roughly with the rightness of x whereas the last three are roughly about the wrongness of x. 95 method of WRE lead to relativism? If yes, what kind of relativism does the method of WRE lead to? If no, what kind of relativism does the method avoid? Relativism amounts to at least the following three claims: (1) Judgments of X differ according to Y; (2) There are no objective criteria to decide which is better among different or competing Y’s; (3) therefore, X is relative to Y (X depends upon Y, or X is determined in relation to Y). Relativisms can be distinguished according to what X and Y stand for. Cognitive relativism and ethical (or moral) relativism can be distinguished on the basis of whether X is of a cognitive matter (e.g., truth) or a moral one (e.g., moral right/wrong or moral truth). If X is of an aesthetic matter, we would have aesthetic relativism. What I want to deal with here are the relationships between various ethical relativisms and the method of WRE. Ethical relativisms vary according to whether Y stands for individual persons (subjective relativism), cultures (cultural relativism), moral frameworks like a set of values and principles (moral relativism), or conceptual schemes (conceptual relativism.) 1. Subjective Relativism and Cultural Relativism l . l . Subjective Relativism Subjective relativism holds that: (SR) Judgments of right or wrong differ according to individual subjects. Therefore, right/wrong is relative to an individual subject. According to subjective relativism, a certain action may be morally right for one person 2 Kai Nielsen, “Relativism and Wide Reflective Equilibrium,” Monist 76, no. 3 (1993): 316. 96 but morally wrong for another. Emotivism, which is often called subjectivism,3 is a type of subjective relativism. I understand the emotivist version of subjective relativism (SRe) as follows: (SRe) (1) Judgments of right or wrong differ according to individual persons. (2) Judgments of right or wrong are, at bottom, merely the expression of the emotions. (3) There are no cognitive or objective criteria for deciding whether, in a particular situation, one expression of emotion is better or more correct than another. (4) Therefore, right/wrong is relative to the emotions or feelings of individual subjects. For emotivists, a moral judgment is merely an expression of a subject’s emotions, as seen in the expressions “Hurrah!” “Alas!” or “Yecch!” More importantly, emotivists are not interested in moral justification. As James Rachels points out, the fact that emotivism cannot account for the place of reason in our ethical thinking is its main problem.4 It is clear that the method of WRE does not lead to this type of subjective relativism. 1.2. Cultural Relativism Cultural relativism can be defined as follows: (CR) (1) Judgments of right or wrong differ according to cultures. (2) There are no objective criteria by which to decide which cultural perspective is better than another. (3) Therefore, right/wrong is relative to culture. Anthropologists have reported that different customs have been developed in different cultures. The Eskimos perform infanticide and the Callatians (a tribe in India) ate the bodies of their dead fathers.5 Drawing from these sorts of observation, cultural relativism 3 In fact, subjectivism itself does not logically imply relativism. For example, Kant’s epistemology can be called subjectivism but not relativism. ‘ James Rachels, The Elements ofMoral Philosophy, 3“1 ed. (McGraw-Hill College, 1999), 45. 5 Rachels, The Elements ofMoral Philosophy, 20-22. 97 claims that right/wrong is determined in relation to culture because different cultures have different moral codes. Rachels calls this argument the Cultural Difference Argument: (1) Different cultures have different moral codes. (2) Therefore, there is no objective “truth”. in morality. Right and Wronbg are only matters of opinion, and oprmons vary from culture to culture. I think the argument contains some suppressed premises. These become more apparent when we reforrnulate the above argument (the suppressed premises are indicated by square brackets): (1) Different cultures have different moral codes. [1-1] If (1) is true, judgments of right or wrong differ according to cultures. [1-2] There are no objective criteria by which to decide which cultural perspective is better than others. (2) Therefore, there is no objective “truth” in morality. The Cultural Difference Argument has the following problems. First, the argument draws conclusion (2), regarding the status of morality, from premise (l), which talks of differences between different culture’s moral codes.7 Even if premise (1), along with the suppressed premises, is true, however, conclusion (2) still may be false because there may be shared moral norms or considerations underlying the different moral codes. Different moral codes might have arisen due to different ways of life or circumstances. But these various moral codes might have been developed using the same moral principles or values. If so, a logical gap might exist between the above argument’s premises and conclusion. When we try to discover the reasons behind the moral codes of other societies, we in fact find less disagreement than we often expect. For example, Eskimo infanticide 6 Rachels, The Elements of Moral Philosophy, 24. 7 If conclusion (2) is understood to deny moral truth itself, cultural relativism may lead to moral nihilism because moral nihilism claims that there are no moral facts, no moral truths, and no moral knowledge. 98 might be justified by their harsh living environment. According to Rachels, as nomadic people, the Eskimos must move about with their infants when searching for food. But, for nutritional reasons, Eskimo mothers must nurse their infants for a much longer period of time than mothers of other societies — often four up to four years. There are limits, then, to the number of infants that one mother can sustain. Therefore, infanticide, though a drastic measure, is sometimes needed to ensure the family’s survival, and need not demonstrate a fundamentally different values or attitudes toward children than found in other cultures.8 We can see from this example that even apparently different moral codes can often be based on common moral principles or considerations. Different moral codes may be simply different actualizations of the same moral principles or values. But as alternatives to infanticide or changes in circumstances arise, Eskimo infanticide loses its justification. Second, suppressed premise [1-2] states that there are no objective criteria by which to decide which cultural perspective is better than others. Different cultures or moral codes are not, however, always beyond moral criticism. Moral codes or cultural customs may often be criticized from within a particular culture, and sometimes from without. We ask for reasons for moral codes or customs and criticize them when they lack reasonable grounds. We do not respect them merely because they are accepted in a different culture or society. For example, we criticized Nazism because it lacked acceptable supporting grounds. The method of WRE, because it aims to arrive at justified moral beliefs, will request reasons to justify moral codes held by different cultures. Cultural relativism, on the other hand, assumes the justification of moral codes without any criticism merely 8 Rachels, The Elements of Moral Philosophy, 28-29. 99 because they have been accepted and practiced in different cultures. Nielsen points out that the method of WRE can enable us to criticize and correct culturally based views through the revision process using moral principles and relevant background beliefs or theories.9 In addition, the method of WRE allows for the criticism of cultures if belief systems imbedded in those cultures are overly narrow or if they contain incorrect moral or non-moral beliefs in the light of a more comprehensive coherent belief system. Therefore, the method of WRE can say that some aspects of culture should be corrected or revised, though it will be hard to say one entire culture is preferable to all others. As the method of WRE encourages individual moral inquirers to revise or broaden their belief systems, so too it will recommend to societies or groups that they correct or broaden the belief systems imbedded in their cultures. In this respect, subjective relativism and cultural relativism are similar — they can both be the sort of relativism that allows anything to go. Both may lead to moral nihilism, denying the existence of any moral facts or truths. Any relativism based merely on the fact that moral judgments or codes differ according to individual subjects or cultures, and that gives up requiring moral justification, must allow nearly anything to be acceptable. I will call this kind of relativism naive or vulgar relativism. But more modest conceptions of cultural relativism are not subject to the same criticism and are of positive value. First, cultural relativism may be merely descriptive. This sort of cultural relativism, as Rachels points out, tells us that “many (but not all) of our practices are merely peculiar to our society,” and that our feelings with respect to the customs or practices of other cultures may be “the result of cultural conditioning.”IO 9 Nielsen, “Relativism and Wide Reflective Equilibrium,” 326-327. '0 Rachels, The Elements of Moral Philosophy, 34 and 36. 100 Cultural relativism interpreted this way encourages us to maintain an open mind. Second, although the moral codes of different cultures are not likely to be entirely different, some elements of them may differ because of cultural differences reflecting aspects of moral frameworks, worldviews, or conceptions of the good life. I will discuss these issues in the next section. 2. Moral Relativism and Conceptual Relativism 2.1. Moral Relativism I will simply use the term ‘moral relativism’ to refer to one version of ethical relativism. Gilbert Harman argues for moral relativism this way: I am going to argue that moral right and wrong (good and bad, justice and injustice, virtue and vice, etc.) are always relative to a choice of moral framework. What is morally right in relation to one moral framework can be morally wrong in relation to a different moral framework. And no moral framework is objectively privileged as the one true morality.11 (My italics) His argument can be reformulated as follows (his words are in parentheses): (MR) (1) Judgments of right or wrong differ according to moral fi'ameworks (What is morally right in relation to one moral framework can be morally wrong in relation to a different moral framework). (2) There are no objective criteria by which to decide which moral framework is better than another (No moral framework is objectively privileged). (3) Therefore, right/wrong is relative to a moral framework (moral right and wrong are always relative to a choice of moral framework). My premise (1) is not exactly the same as his way of putting it. His way, however, already contains his conclusion. What he should say is my premise (1). Harman claims 11 Gilbert Harman and Judith Jarvis Thomson, Moral Relativism and Moral Objectivity (Blackwell, 1996), 101 that there are no objective criteria by which to decide between frameworks. What does he mean by a moral framework? He often draws on analogy between the relativity of morality and that of motion. He says, “Something can be in motion relative to one system of spatio-temporal coordinates and not in motion relative to another system?” Similarly, he says, “a given act can be right with respect to one system of moral coordinates and wrong with respect to another system of moral coordinates.”l3 What is meant by a moral system of coordinates? He answers: “By ‘a moral system of coordinates’ I mean a set of values (standards, principles, etc.).’”4 Thus, an action is right or wrong in relation to a set of values and principles. Let us consider an example. Suppose principle set A places a priority on the principles of utility and beneficence, whereas set B places a priority on the principles of justice and nonmaleficence. Forty-six year-old Susan needs a liver transplantation and has been waiting for six months. But forty-five year-old Amanda was hospitalized two weeks ago and also needs a liver transplantation. Suppose neither is an alcoholic but both will die within the month without a transplant. Although Susan can be expected to live an additional five years with the transplant, Amanda would live ten. Because their physician, John Smith gives priority to medical utility in light of the limited number of organs available, he decided to provide transplantation surgery to Amanda. His action would be right in relation to principle set A. But the action would be wrong in relation to set B because Susan was on the waiting list first and principles of justice usually translate to “first-come, first-served,” in the context of organ transplantation. According to the 3. '2 Harman and Thomson, Moral Relativism and Moral Objectivity, 12-13. '3 Harman and Thomson, Moral Relativism and Moral Objectivity, 13. H Harman and Thomson, Moral Relativism and Moral Objectivity, 13. 102 second line of thought, the physician’s decision was unfair. Each person’s life is equally precious. For the moral relativists, neither set A nor set is objectively privileged. Therefore, right/wrong is relative to the choice of a set of moral principles (or moral framework.) The method of WRE will not lead to the same conclusion. The method requires a comprehensive and coherent set of moral principles (values or standards) Therefore, since both sets A and B include important principles, both must be included in the framework of WRE. When both sets are included, the above example will be reformulated as a moral dilemma or a difficult moral question, rather than a demonstration of the relativity of morality. We often face moral dilemmas in which our moral values and principles conflict. Moral relativism, however, drew the relativity of the rightness from the above example that could be understood as moral dilemma. Conflicts between moral principles will be taken by the method of WRE as evidence of the need for revisions or a balancing of the moral principles involved, the modification of other relevant beliefs, or the acquisition of more information or new moral beliefs. For example, we can consider whether or not it should be a physician who must decide who should get transplants. We can perhaps develop a procedure to take into account medical utility and fairness, based on public consensus. The method of WRE, of course, will not always provide an answer to every moral dilemma. But the difference between the method of WRE and moral relativism is apparent. The method of WRE has no need to conclude that right/wrong is relative to a moral framework. A person following the method will say that he/she cannot resolve this problem now, but a solution may arise when we better understand the situation in 103 question and then come to know which principle should enjoy priority. Nonetheless, moral relativism has some aspects worth noting. It tells us that there is no single absolute moral framework or set of moral principles. As moral nihilism is hard to accept, so too is moral absolutism, which claims that there is only one absolute moral framework. More importantly, there is an important difference between moral relativism and the vulgar relativism. Moral relativism argues that different sets of moral principles support different conclusions. Judgments of right/wrong are not arbitrarily, but must be supported by reasons grounded in moral principles. In contrast, vulgar relativisms make their claims by the mere appeal to the fact that different individuals or cultures have different moral codes or judgments. In addition, moral relativism may contain an aspect of conceptual relativism if the decision of which set of principles is salient to a certain situation is said to depend upon different views about the situation due to different philosophical or religious perspectives. 2.2. Conceptual Relativism Conceptual relativism has been discussed as a version of cognitive relativism. It claims that truth is relative to a conceptual scheme. Conceptual relativism, as Nielsen ”15 points out, has “important implications for morality. He explains conceptual relativism as follows: Conceptual relativism is the claim that different cultures see the world differently in certain crucial respects. It is the further claim that there can be no neutral or culturally ubiquitous way in which the world can be described against which these different and incommensurable conceptual systems or schemes can be assessed or measured. There is no one true '5 Nielsen, “Relativism and Wide Reflective Equilibrium,” 328. 104 description of the world; there is no right way of seeing the world or viewing the world. '6 However, what is meant by a conceptual scheme? When Donald Davidson deals with conceptual relativism as cognitive relativism, he explains conceptual schemes in this way: Conceptual schemes, we are told, are ways of organizing experience; they are systems of categories that give form to the data of sensation; they are points of view from which individuals, cultures, or periods survey the passing scene.17 As this passage shows, a conceptual scheme is a way of organizing experience with a system of categories or perspectives. There is no reason that moral experience must be excluded. Various people sometimes experience an identical occurrence differently and judge it differently. If these differences are due to different conceptual schemes, conceptual relativism may be a version of ethical relativism. Conceptual relativism can be described in this way: (CR) (1) Judgments of right or wrong differ according to conceptual schemes. (2) There are no objective criteria by which to decide which conceptual scheme is better than another. (3) Therefore, right/wrong is relative to conceptual schemes. Does the method of WRE lead to conceptual relativism? Kai Nielsen argues that the method of WRE would be impossible if conceptual relativism were true, but adds that conceptual relativism is not intelligible. Nielsen says as follows. Considered judgments, principles, theories, descriptions of the world where they belong to different conceptual schemes could not even in principle be compared, if conceptual relativism were true, and thus, because we cannot make such comparisons, we could not, even in principle, get our considered judgments into wide reflective equilibrium.‘8 (My italics) '6 Nielsen, “Relativism and Wide Reflective Equilibrium,” 328. '7 Donald Davidson, Inquiries into Truth and Interpretation (Clarendon Press, 1984), 183. '8 Nielsen, “Relativism and Wide Reflective Equilibrium,” 328-329. 105 Why would we not be able to get our considered judgments into the state of WRE? According to Nielsen, if conceptual relativism is true, beliefs that belong to different conceptual schemes will not be able to be compared. We would have, then, two or more reflective equilibria. “Where we had two or more reflective equilibria,” he says, “we ”19 would not yet have attained WRE. Therefore, “with conceptual relativism,” he says, “there could be various narrow conceptual-scheme-dependent reflective equilibria, ”20 (My italics) In other words, if conceptual relativism is mutually incommensurable. true, narrow reflective equilibrium, but not wide reflective equilibrium, is possible. If we look closer, we can see that Nielsen’s use of the words ‘wide’ and ‘narrow’ are different from Daniels’s use, which I have been following. My distinction between narrow and wide reflective equilibrium has to do with whether or not background beliefs enter into the framework of reflective equilibrium. But Nielsen distinguishes them according to whether equilibrium is obtained in one of different competing belief systems, or in the most comprehensive or adequate belief system with no competitors. He says: Where a situation obtains in which there are X believers and Z believers, there would not, with the conflicting beliefs of X and Z, be a WRE. Neither X nor Z beliefs could be in WRE. At least where the worlds are in any way in contact there can only be one WRE.21 We may, of course, need to distinguish WRE obtained in one of various competing comprehensive coherent belief systems from WRE obtained in the most comprehensive or adequate coherent belief system which has no competing belief systems. If we have to distinguish them following Nielsen, we could call the latter ideal wide reflective '9 Nielsen, “Relativism and Wide Reflective Equilibrium,” 327. 20 Nielsen, “Relativism and Wide Reflective Equilibrium,” 328. 2' Nielsen, “Relativism and Wide Reflective Equilibrium,” 327. 106 equilibrium. Nielsen’s expression, “narrow conceptual-scheme-dependent reflective equilibria, mutually incommensurable,” in the above passage, will be more appropriate stated as "conceptual-scheme-dependent wide reflective equilibria, mutually incommensurable.” Thus, if conceptual relativism is true, there are conceptual-scheme-dependent wide reflective equilibria, mutually incommensurable. Nielsen argues, of course, in agreement with Donald Davidson, that conceptual relativism is unintelligible. Davidson conceives of conceptual schemes as languages, whether natural or not. He claims that the total failure of translation between languages is impossible, and that, hence, the notion of incommensurable conceptual schemes is unintelligible.22 Although I do not deal with this claim here, I acknowledge that the matter of the intelligibility of incommensurable conceptual schemes is an important philosophical issue. But, the more important point is what conceptual relativists intend to say about incommensurable conceptual schemes. Their main concern would not be that of incommensurability itself, but of incomparability. This is because if incommensurable conceptual schemes are comparable, relativism cannot stand at all. Thus, if there are incomparable conceptual schemes, or something similar to them, conceptual relativism may retain its “important implication for morality.” 2.3. Conceptual Relativism Revisited We do not yet have a clear answer to whether or not the method of WRE leads to conceptual relativism. To get an answer to this question, we must first explore two others. 107 First, are there different conceptual schemes, or anything similar to them, in the context of ethics? One possibility is that there are competing reasonable comprehensive coherent belief systems. Although such belief systems are not identical to conceptual schemes, they may contain conceptual schemes or worldviews. Competing reasonable comprehensive coherent belief systems are likely to include different philosophical or religious worldviews. These different worldviews may play a role in organizing our moral and non-moral experience in different ways. Competing moral judgments may arise due to these different worldviews. If these various philosophical or religious worldviews are as incomparable as conceptual schemes, however, conceptual relativism may still stand as a version of ethical relativism. In ethical issues related to abortion or reproductive technology, for example, we can find competing positions that may come from different worldviews. There are, for instance, alternative metaphysical views regarding the status of fetus. This is not merely due to our lack of biomedical knowledge. Current reproductive technology makes it possible to produce a baby in a variety of ways. A sperm or egg donor and a surrogate mother can be parents of a single baby. Some gay or lesbian couples as well as a few menopausal women want to make use of this technology. Should we prevent these people from using the technology? This forces us to reconsider our notion of family, motherhood, parenting, and the limits of procreative liberty. Differences of view regarding these notions may be responsible for differences in people’s willingness to use reproductive technology. Thus, conceptual relativists may claim that moral judgments regarding abortion can be justified in relation to a reasonable comprehensive coherent belief system. Each 22 See Davidson, Inquiries into Truth and Interpretation, 183-198. 108 view on the issue may be in equilibrium within its reasonable comprehensive coherent belief system. Thus, competing reasonable comprehensive coherent belief systems might be incomparable. Thus, if competing reasonable comprehensive coherent belief systems are not comparable, there may be a version of conceptual relativism described as follows: (CRr) (l) Judgments of right or wrong differ according to reasonable comprehensive coherent belief systems. (2) There are no objective criteria by which to decide which belief system is better than another. (3) Therefore, right/wrong is relative to reasonable comprehensive coherent belief systems. The success of this claim depends upon the possibility of there being incomparable reasonable comprehensive coherent belief systems. Are there such belief systems? This raises the second question relevant to the larger issue of whether or not the method of WRE leads to conceptual relativism. Even if we can suppose that there are incomparable competing reasonable comprehensive coherent belief systems, does this possibility mean that the method of WRE implies conceptual relativism? As mentioned in the discussion of cultural relativism, there seems to be a logical gap between the existence of different moral judgments and the metaphysical conclusion that there is no objective moral truth. Conceptual relativists must show that there is no such a gap in the method of WRE to argue that the method leads to conceptual relativism. 109 3. Reasonable Pluralism and Conceptual Relativism 3.1. Reasonable Persons and the Burdens of J udgrnent Are there incomparable reasonable comprehensive coherent belief systems? One possibility can be seen in Rawls’s notion of reasonable pluralism, which states that there are many reasonable philosophical or religious doctrines. The reasonable, according to him, is distinguished from the rational: In everyday speech we are aware of a difference and common examples readily bring it out. We say: “Their proposal was perfectly rational given their strong bargaining position, but it was nevertheless highly unreasonable, even outrageous.”23 Martin Benjamin’s explanation can help here. “Rationality is,” he says, “for the most part, an intellectual virtue having to do with the selection and pursuit of the most effective means to a set of carefully selected ends.”24 However, “reasonableness,” he continues, “requires giving equal or fair consideration to the reasonable ends or viewpoints of others for their own sakes.”25 (My italics) Egoists might be rational, though not reasonable. The reasonable, according to Rawls, has two basic aspects. One is “the willingness to propose fair terms of cooperation and to abide by them provided others do.”26 The other is “the willingness to recognize the burdens of judgment and to accept 23 John Rawls, Political Liberalism (Columbia University Press, 1996), 48. 2" Martin Benjamin, “The Value of Consensus,” in Societys Choices: Social and Ethical Decision Making in Biomedicine, edited by Ruth Ellen Bulger, Elizabeth Meyer Bobby, and Harvey V. Fineberg (National Academy Press, 1995), 251. 25 Benjamin, “The Value of Consensus,” 252. 2" Rawls, Political Liberalism, 54. 110 ”27 In the discussion of the fundamental principles of society, Rawls their consequences. explains the first aspect of the reasonable this way: Persons are reasonable in one basic aspect when, among equals say, they are ready to propose principles and standards as fair terms of cooperation and to abide by them willingly, given the assurance that others will likewise do so. Those norms they view as reasonable for everyone to accept and therefore as justifiable to them; and they are ready to discuss the fair terms that others propose.28 (My italics) As the italicized words show, reasonable persons consider fair terms of cooperation with other persons and are ready to discuss such terms. But other reasonable persons may have different religious or philosophical views about the world and the good life. Therefore, reasonable persons concerned with fair terms of cooperation will propose norms that are justifiable to other reasonable persons affected by the proposed norm. Thus, reasonable persons would give equal or fair consideration to the perspectives of others. In this way, respect for others is the key notion in Rawls’ sense of the reasonable. Irnportantly, this will be mutual respect because reasonable persons would respect one another.29 However, there will be disagreement among reasonable persons, which is called reasonable disagreement.3o This will occur due to a variety of sources that produce different judgments. That informed, clear thinking, good reasoning individuals will sometimes come to different conclusions on matters of morality Rawls calls the burdens of judgment. Rawls identifies a number of sources of reasonable disagreement. These should, however, be distinguished from sources of unreasonable disagreement such as 27 Rawls, Political Liberalism, 54. 28 Rawls, Political Liberalism, 49. 29 In this respect, the reasonable is a moral virtue of a person. 3° Rawls, Political Liberalism, 55. 111 “prejudice and bias, self- and group interest, blindness and willfulness.”31 Reasonable disagreements are those that remain even after these are overcome. Rawls describes six sources of reasonable disagreement (a through f). The first three are: a. The evidence - empirical and scientific — bearing on the given 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 reply on judgment and interpretation (and on judgments about interpretations) within some range (not sharply specifiable) where reasonable persons may differ.32 The first burden of judgment arises because of the limitations of our knowledge. The second may be related to differences of background beliefs, growing out of different cultures or upbringing. For example, both discipline and care are necessary in raising a child. But even reasonable persons differ in their judgments of whether in certain situations discipline or care should be given more weight. The third comes from some extent of vagueness inherent in our concepts. Different interpretations of concepts may occur among reasonable persons. For example, our concept of death refers to the cessation of life. We may differ, however, in our interpretations of “the cessation of life.” There is disagreement as to when a life can be said to have ended. The criterion of the whole brain death is legally accepted. But some have argued that the criterion of the higher brain death is preferable. They think that a person is dead when there is irreversible loss of cerebral function, which is key to personality, conscious life, personal uniqueness, remembering, judging, reasoning, and enjoying, even though the brainstem 3' Rawls, Political Liberalism, 58. ’2 Rawls, Political Liberalism, 56. 112 may be fully functioning.” The next two burdens of judgment are more directly related to different background beliefs or theories. d. To some extent (how great we cannot tell) 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.34 As these passages show, different total experience and normative considerations may provide sources of reasonable disagreement. Different cultures, ways of upbringing, or careers will lead to different courses of life that can, in turn, influence how one weighs evidence. With respect to the issue of different normative considerations, let us consider, for example, the principles of utility and justice. Normative considerations based on each of these principles are valuable for moral reflection. But they may sometimes give conflicting guidance, as shown in the earlier example of liver transplantation. The following burden of judgment addresses the difficulties of choosing one among incompatible competing values. f. Finally, as we note in referring to Berlin’s view (V:6.2), 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 realized. This is because any system of institutions has, as it were, a limited social space. In being forced to select among cherished values, or when we hold to several and must restrict each in view of the requirements of the others, we face great difficulties in setting priorities and making adjustments. Many hard decisions may seem to have no clear 33 Rober M. Veatch, “The Impending Collapse of the Whole-Brain Definition of Death,” in Ethical Issues in Modern Medicine, 5th edition, edited by John D. Arras and Bonnie Steinbock (Mayfield Publishing Company, 1999), 153. 3" Rawls, Political Liberalism, 56-57. 113 answer.35 (My italics) As the italicized portions emphasize, we must often choose one of various competing values. If values are comparable and compatible, or incomparable but compatible, our decision may not be difficult. But as Isaiah Berlin claims, there are good and important values that are occasionally conflicting and incompatible, such as liberty and equality or planning and spontaneity. Such conceptual conflicts are also among the sources of reasonable disagreement. Reasonable persons who give equal or fair consideration to the perspectives of others would be willing to recognize the above burdens of judgment and to accept the consequences of the burdens of judgment. These consequences include reasonable disagreement — moral disagreement that cannot be attributed to ignorance, selfishness, bias, bad reasoning or other remediable faults. 3.2. Reasonable Pluralism and Implication of Toleration Why do reasonable persons affirm many different comprehensive reasonable doctrines? Rawls says, “Religious and philosophical doctrines express views of the world and of our life.”36 Each doctrine “organizes and characterizes recognized values” and does this “in ways that distinguish it from other doctrines, for example, by giving certain values a particular primacy and weight.”37 (My italics) Therefore, various comprehensive doctrines will differently affect our ways of weighing relevant evidence, the interpretation of concepts, and the choice of values from several incompatible values. Thus, reasonable 35 Rawls, Political Liberalism, 57. 36 Rawls, Political Liberalism, 58. 114 disagreements are due to the burdens of judgment, and different judgments are due to the variety of reasonable comprehensive doctrines. Rawls says that reasonable persons will not all affirm the same comprehensive doctrine, but will recognize that many reasonable comprehensive doctrines are affirrned because all persons are subject to the burdens of judgment.38 Rawls calls this feature reasonable pluralism.39 However, can we identify a reasonable comprehensive doctrine with a particular comprehensive belief system? A comprehensive belief system will contain a comprehensive doctrine. The background beliefs or theories in the framework of WRE include such a doctrine. Comprehensive belief systems may be differentiated according to which doctrine they contain. But are reasonable comprehensive doctrines incomparable? We can compare and criticize the doctrines of others in the light of our own. It is unclear, however, whether we can do this from outside of the perspective of our own doctrine. An answer to this issue seems to be found in another consequence of the burdens of judgment — the need of toleration. Rawls says, “These burdens of judgment are of first significance for a democratic idea of toleration.”40 According to him, reasonable ”4' We must disagreement can be eliminated “only by the oppressive use of state power. tolerate reasonable disagreement, a variety of reasonable comprehensive doctrines, and thus a variety of occasionally conflicting reasonable comprehensive coherent belief systems. Toleration, I think, implies that we should not insist that our belief system is objectively better than any other equally reasonable one. If reasonable persons are to tolerate other reasonable doctrines, they should acknowledge that there are no objective 37 Rawls, Political Liberalism, 59. 38 Rawls, Political Liberalism, 60. ’9 Rawls, Political Liberalism, 36. Of course, we should distinguish reasonable pluralism from pluralism as such, which includes unreasonable views or doctrines, according to Rawls. 115 criteria outside of their own doctrines by which to make evaluations between reasonable doctrines. If a variety of reasonable comprehensive doctrines is a feature of our democratic society, and if it results in the need to be tolerant, then this is also true of reasonable comprehensive coherent belief systems. I conclude that there are reasonable comprehensive coherent belief systems that cannot be compared outside of their own viewpoints. In this respect, reasonable pluralism leads us to the same lessons as did cultural relativism that is interpreted as a descriptive theory. But does our acceptance of reasonable pluralism imply conceptual relativism? 3.3. Moral Justification and Morality To show that the method of WRE leads to conceptual relativism, as mentioned earlier, there must be no logical gap between the existence of different judgments and the metaphysical conclusion that there is no objective moral truth. To repeat, conceptual relativism was described as follows: (CRr) (l) Judgments of right or wrong differ according to reasonable comprehensive coherent belief systems. (2) There are no objective criteria by which to tell which belief system is better than another. (3) Therefore, right/wrong is relative to reasonable comprehensive coherent belief systems. Even if we take premise (2) to be true, a problematic gap exists between the combination of premises (1) and (2) and conclusion (3). To show this more clearly, I have put premise (l) and conclusion (3) in this way: [1] Justified judgments of right or wrong differ according to reasonable comprehensive coherent belief systems. ’0 Rawls, Political Liberalism, 58. 4' Rawls, Political Liberalism, 54. 116 [3] Right or wrong differ according to reasonable comprehensive coherent belief systems. If judgments of right or wrong are in wide reflective equilibrium, premise (l) is intended as premise [l]. The difference between [1] and [3] is that premise [1] is about justified judgments of right or wrong. Premise [1] indicates a relativity of justified judgments while conclusion [3] is about the relativity of morality. Therefore, the issue is whether relativity of justification implies relativity of morality. To show that this implication holds, a justified moral belief must be guaranteed to be true. In other words, if a moral judgment that action A is morally right is justified, then action A must in fact be morally right. However, a justified belief is not always guaranteed to be true. A theory of justification is not identical to a theory of truth. For this reason, coherentism as a theory of justification should be distinguished from coherentism as a theory of truth.42 Coherentism as a theory of justification holds that if a belief belongs to a coherent belief system, it is justified, but not necessarily true. On the other hand, coherentism as a theory of truth holds that if a belief belongs to a coherent belief system, it is true. This distinction, however, does not deny an important role for justification in seeking truth. Justification helps lead us to true belief. Moral justification is similarly a way of leading us to true moral belief. Therefore, the justification of moral judgments must not be understood to be identical to their truth. In this way, the method of WRE need not commit to the relativity of morality due to a relativity of justification. Therefore, it does not necessarily lead to conceptual relativism. But one may object that a theory of moral justification must also be a theory of ’2 Laurence BonJour, The Structure of Empirical Knowledge (Harvard University Press, 1985), 88. 117 truth — what else, after all, can make a moral belief a true one? Thus, one may claim that the method of WRE must also serve as a theory of truth. Suppose some proponents of the method of WRE identify a justified moral belief as a true moral belief. However, would they also say that morality is relative to a particular reasonable comprehensive coherent belief system? If they are reasonable persons, I think that they would not. Let us consider two types of reasonable persons who follow the method of WRE. The first type of persons may identify moral justification with moral truth. However, this type of persons would affirm a variety of reasonable comprehensive coherent belief systems for the same reason as Rawls’s, such that “reasonable persons see that the burdens of judgment set limits on what can be reasonably justified to others.”43 Therefore, a reasonable person will observe the difference between a belief that is justified only to him/herself and a moral belief that is justified to other reasonable persons as well. The first is a belief that is justified in one reasonable comprehensive belief system. The second is a belief that is justified in all reasonable comprehensive belief systems. The second belief would be part of what Rawls calls an “overlapping consensus.” Thus, this type of reasonable persons will not ascribe the same moral status to the two kinds of justified moral beliefs. If a reasonable person following the method of WRE identifies a justified moral belief with a true one, he/she would say that moral beliefs obtaining overlapping consensus are true. For a reasonable person who identifies overlapping consensus with a true moral belief, a moral belief justified in only one reasonable comprehensive belief system may have a limited moral status. Such a moral belief would be a kind of candidate for true moral belief. If two moral beliefs from different reasonable belief systems conflict, they 118 compete against one another like scientific theories. The existence of competing scientific theories does not necessarily imply relativity of truth. We withhold our final judgments of them until we can acquire more evidence while particular scientists continue to present evidence to show their theories to be true. In the same way, when reasonable persons are confionted with competing moral judgments, they will withhold their final judgments even while individual reasonable persons continue to argue that their judgments are true. These persons must provide reasons that all reasonable persons can accept. This attempt is valuable because it provides an opportunity for reasonable persons to give equal or fair consideration to the perspectives of other reasonable persons. This will broaden the scope of overlapping consensus just as a scientist’s attempt contributes to our knowledge of the world. In this way, if a reasonable person identifies a justified moral belief that has obtained overlapping consensus with a true belief, he/she will say that the rightness of an action cannot be determined until we see agreement on the judgments of all reasonable persons. Therefore, reasonable persons who follow the method of WRE and identify moral justification with moral truth will conclude that the method of WRE does not commit us to conceptual relativism. The method of WRE as a theory of truth may require equilibrium in a single reasonable comprehensive belief system shared among all reasonable persons. This belief system will provide a shared basis for the determination of moral right or wrong. These reasonable persons will say that competing moral judgments indicate only disagreement in justification, not in moral truth. The second type of reasonable persons who follow the method of WRE may identify instead a belief justified in even one reasonable comprehensive belief system ’3 Rawls, Political Liberalism, 61. 119 with a true moral belief. In this case, they seem committed to conceptual relativism. But this is where the difficulty arises. How can they explain the difference of moral status between a moral belief justified in one reasonable belief system and a moral belief obtaining overlapping consensus? If they ignore the difference, they are not reasonable because their ignorance results in denying reasonable pluralism and toleration. If they acknowledge the difference, they must admit further that all true moral beliefs are not equal in their moral status. Some moral beliefs are true for all reasonable persons, but some are true only to persons who think the beliefs are justified within their own reasonable belief systems. We would be unable to call this view relativism. This is because this view must say that some judgments of right or wrong do not differ according to reasonable comprehensive belief systems. Therefore, the second type of reasonable persons would conclude that the method of WRE does not lead to conceptual relativism. 120 Chapter Five The Group Method of Wide Reflective Equilibrium Ethical issues in health care are problems for both individuals and society. Abortion and euthanasia, for example, are moral problems waiting not only for individuals to decide how to act, but also for society to develop public policy. Policy- making may be understood to make group decisions. We must often make various group decisions as members of a physician-patient pair at one extreme or as members of society as a whole at another extreme. However, as revealed in Chapter Four, group decisions are not always easy. There are reasonable disagreements and reasonable comprehensive belief systems. How should we attempt to reach group decisions in such situations? This question is inevitably raised in the context of the method of wide reflective equilibrium (WRE) because reasonable disagreements and reasonable pluralism are important background beliefs in its framework. Reasonable disagreement and pluralism, as John Rawls claims, are facts of our pluralistic society. If the method of WRE must accept reasonable disagreement and pluralism, however, how is it to handle reasonable disagreement when we are in need of agreement? It suggests that we should tolerate reasonable disagreements, but we are often in need of an agreement on an answer to an ethical question now rather than later when making public policy. In this chapter, I will show how the method of WRE can be used in the development of group-based decisions. In this way, I will treat the method of WRE as a method of group reasoning. There are several types of agreements that can result from group decisions. The method of WRE can contribute to our decision processes in the case 121 of each type of agreement. 1. The Group Method of WRE 1.1. The Method of WRE as Group Reasoning As discussed in Chapter One, Norman Daniels describes the method of WRE as follows: The method of wide reflective equilibrium is an attempt to produce coherence in an ordered triple of sets of beliefs held by a particular person, namely, (a) a set of considered moral judgments, (b) a set of moral principles, and (c) a set of relevant background theories.l (My italics) Coherence here is pursued within a particular person’s comprehensive belief system. Such a person will try to revise or broaden her belief system by participating in actual or hypothetical interpersonal activity because reasoning involves an interpersonal social element. In Chapter Three, I thus construed the method of WRE articulated by Daniels as an interpersonal epistemic activity. But coherence achieved by this method is one within a particular reasonable comprehensive belief system held by a particular person. This coherence results from a process of revision or broadening from this person s perspective. But the moral reasoning necessary for group decisions does not aim at moral beliefs that are justified within a particular person’s reasonable comprehensive belief system. Moral reasoning related to group decisions seeks moral beliefs that will be justified in all reasonable comprehensive belief systems held by the various group members. Thus, a reasonable person following the method of WRE should attempt to achieve as much ‘ Norman Daniels, “Wide Reflective Equilibrium and Theory Acceptance in Ethics,” The Journal of 122 coherence as is possible with the beliefs held by reasonable persons in addition to him/herself. Martin Benjamin stresses this point by distinguishing personal and interpersonal coherence: Our aim, as we discuss and debate questions, is personal and interpersonal coherence. By personal coherence, I mean answers that, from your own point of view, satisfy the conditions of wide reflective equilibrium better than any alternative. By interpersonal coherence, I mean that your answers and the answers of others strive for as much agreement or congruence with each other as can be obtained.2 He describes interpersonal coherence as agreement among the answers proposed by different persons. We miss his point, however, if we simply identify interpersonal coherence with such agreement. Let us consider what we should do to anive at an agreement. We must know the reasons others use to support their answers to the question at hand. To do this, we need to understand their perspectives. We might begin by asking about some further beliefs and their justifications. As the conversation goes deeper, we would come to know where people agreed and disagreed. We may find that we agree on certain matters, which can then serve as a basis for seeking agreement about matters on which we are divided. Charles Larmore, quoting from Rawls, puts it this way: Justification is not merely “valid argument from listed premises.” Instead, it “is addressed to others who disagree with us, and therefore it must always proceed from some consensus, that is, from premises that we and others publicly recognize as true.”3 Final agreement should be based on those areas of agreement shared with those with Philosophy 76 (1979): 258-9. 2 Martin Benjamin, Philosophy and This Actual World: An Introduction to Practical Philosophical Inquiry (Roman & Littlefield Publishers, Inc., 2003), 122. 3 Charles Larmore, “Public Reason,” in The Cambridge Companion to Rawls, edited by Samuel Freeman (Cambridge University Press, 2003), 376. The quotation from Rawls is from John Rawls, Collected papers, edited by Samuel Freeman (Harvard University Press, 1999), 394 and 426-427. 123 whom we otherwise disagree. The others involved may hold different reasonable comprehensive belief systems. Thus, at the beginning of the process, we are likely to find disagreement. But our final goal is to mitigate the initial disagreement by finding areas of agreement. Therefore, the process of achieving agreement on a final answer is one in which each member tries to provide reasons that others can accept from a mutually shared perspective. In this way, a member of a group following the method of WRE will try to achieve coherence with beliefs shared by all involved. I will call this application of the method of WRE to group-based reasoning the group method of WRE. It can be, building on Daniels’s original formulation above, described like this: The group method of wide reflective equilibrium is an attempt to produce coherence in an ordered triple of sets of beliefs mutually accepted by persons who have different reasonable comprehensive belief systems, namely, (a) a set of mutually accepted considered moral judgments, (b) a set of mutually accepted moral principles, and (c) a set of mutually accepted relevant background theories. The group method of WRE can serve as a method of reasoning for group decisions. I will call this reasoning simply group reasoning. This method will have the following features distinguishing it from the method of WRE articulated by Daniels. First, we can expect that a belief system mutually held by all the members of a group will usually be smaller than a particular reasonable comprehensive belief system held only by an individual person. But the first belief system will involve a mutually shared point of view coming from those moral and non-moral beliefs that are shared by all the members. I will call this set of shared beliefs a group belief system. Beliefs that are mutually accepted by all the members of a group will be called by group beliefs to be distinguished from personal beliefs of individuals. 124 Second, group belief systems may vary depending upon the purpose of a particular group. In the above discussion, I had in mind the whole society as the relevant group. The group method of WRE can be used, though, by ethics committees, political parties, groups of priests, and small groups composed of a patient, his/her family members, and health care providers. Group belief systems may contain unique sets of beliefs, as long as they are mutually accepted by all the members of the group. In my discussion of the group method of WRE, however, I will primarily have in mind an ethics committee or commission. Such groups have the feature that their members usually hold or represent various reasonable comprehensive belief systems. Third, group beliefs are not merely those beliefs that are collected on the basis that all the members of a group simply hold a certain belief. If these were so, there would be no used of the method of WRE at all. Instead, group reasoning, including the process of revision, will play an important role in reaching agreement in the group method of WRE. In this way, personal beliefs will become group beliefs when the former cohere with other group beliefs. To be group beliefs, thus, personal beliefs proposed as new group beliefs may often be revised during the revision process. Fourth, the main aim of group reasoning will be to achieve justified moral beliefs that are mutually accepted by all members. Moral beliefs differ in their level of generality. Members may seek agreement on particular moral judgments, moral codes, rules, or principles. But they may also seek agreement regarding specific procedures or safeguards. Assuming that moral justification is identical with true moral belief, one may want to use this method to achieve moral truth, as discussed in Chapter Four. It is not clear, however, whether we can identify moral justification (even moral justification 125 obtaining overlapping consensus) with moral truth. Proper use of the group method of WRE demands that its users be reasonable. Rawls’s notion of the reasonable can give us some indication of how to best proceed. 1.2. The Reasonable for Group Reasoning The process of group reasoning requires us to acknowledge several relevant background beliefs. Participants in group reasoning will often hold different and occasionally conflicting reasonable comprehensive coherent belief systems. Let us simply assume that all of the participants are reasonable persons, and as such, as explained in Chapter Four, are willing to (1) give equal and fair consideration to the perspectives of other reasonable persons, (2) recognize the burdens of judgment, and (3) accept the need of tolerating reasonable disagreements and the diversity of reasonable comprehensive belief systems. Mutual respect among reasonable persons would be considered an important moral virtue. To understand one anther, members can and should freely express their personal beliefs and perspectives, but they should also be ready to discuss them in light of the group belief system. Participants should be ready to present reasons to support their proposed beliefs. They also should consider whether their supporting reasons could be accepted by all of the other reasonable members. Therefore, the process of group reasoning is one where reasonable persons exchange and discuss personal beliefs coming from their various reasonable comprehensive belief systems and perspectives, and as a result, come to understand one another more deeply. In this way, any agreement achieved 126 would be informed and uncoerced. Thus, the group method of WRE will be a process of presenting personal beliefs, identifying agreements, revising presented personal beliefs, and forming new group beliefs with the aim of achieving interpersonal coherence. This process will lead to the establishment of as coherent and comprehensive a group belief system as possible. However, a question arises with the group method of WRE. If mutual respect among group members is to be observed, how can this method be used by those who follow the principlist or casuist methods of reasoning? This question is important because group members may have different views regarding the structure of justification. Although from a coherentist perspective the group method of WRE is the most plausible method of group moral reasoning, it might be objected that it is wrong to propose a method of group reasoning that prevents others from using their own preferred method of reasoning. In response, I will have to show both that the group method of WRE can be used by members who follow a principlist or casuist approach and also that this method is better than the alternatives. 1.3. Group Belief and Group Justification The foregoing description of the group method of WRE states that it seeks coherence among three elements: mutually accepted considered moral judgments, moral principles, relevant background beliefs or theories. Thus, to show that the group method of WRE can involve reasonable persons who use other reasoning methods, we need a neutral definition of the justification of group beliefs that all members of a group can 127 accept. I will adopt a definition that is presented by Frederick F. Schmitt. First, Schmitt defines group beliefs: A group G believes that p just in case the members of G jointly accept that p, where the latter happens just in case each member has openly expressed a willingness to let p stand as the view of G, or openly expressed a commitment jointly to accept that p, conditional on a like open expression of commitment by other members of 0.4 A group belief p can be simply said to be a belief accepted jointly by the group members to stand as the position of the group. The expression, ‘conditional on a like open expression of commitment by other members,’ in the above passage, is important because group members must affirm whether a belief is jointly accepted. In practice, there are various ways to express a commitment to a belief. Schmitt next defines justification of a group belief: A group G is justified in believing p just in case G has good reason to believe p, where G has a reason r to believe p just in case all members of G would properly express openly a willingness to accept r jointly as the group’s reason to believe p.5 The justification of a group belief is achieved when all of the group members jointly accept another belief as good reason to believe the group belief in question. Let us call the justification of a group belief or the justification of a group holding a belief group justification. But, there remains the issue of what counts as a good reason for a group to hold a belief. We usually think that if a belief has an inferential or explanatory relationship with supported beliefs, then we have good reason to hold it. Let us refer to this as our having a substantial reason to hold a belief. If we define a good reason as a substantial reason, 4 Frederick F. Schmitt, “The justification of Group Beliefs,” in Socializing Epistemology: The Social Dimensions of Knowledge, edited by Frederick F. Schmitt (Rowman & Littlefield Publishers, 1994), 262. 5 Schmitt, “The justification of Group Beliefs,” 265. 128 however, we will have a definition that is too strong to be useful in group justification. Under this definition, group justification will be achieved only when all members accept jointly a substantial reason to hold a proposed belief. This will exclude from group belief systems many otherwise correct or justified beliefs held by individual persons. Furthermore, it will not be easy for a group to have a jointly accepted substantial reason regarding a belief nevertheless shared by all. Stephen Toulrnin, when he served as a member of the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, observed that members of the group often reach consensus even though they have different reasons for accepting the shared belief.6 That is, all of the members of the group can hold an identical belief, but may do so for different substantial reasons. If alternatively we identify group justification with all members holding a belief, an incorrect belief could be adopted as a justified group belief merely because all members hold an identical belief. I will call this nai've overlapping consensus. What we should aim to achieve is agreement obtained when all reasonable members recognize that a proposed belief is substantially justified in light of the perspective of each comprehensive belief system. I will call this agreement reasonable overlapping consensus because it is achieved among reasonable persons who recognize and respect different supporting reasons, which may be due to different methods of justification. Thus, to adopt Schmitt’s definition of group justification, we have the following interpretation of a good reason: (GR) A reason r that all members accept jointly for group justification is 6 Martin Benjamin, “The Value of Consensus,” in Society is Choices: Social and Ethical Decision Making in Biomedicine, edited by Ruth Ellen Bulger, Elizabeth Meyer Bobby, and Harvey V. Fineberg (National Academy Press, 1995), 243. 129 mutual recognition of a proposed belief p being substantially justified in light of the perspective of each available reasonable comprehensive belief system among all the members. This notion of a good reason for a group to accept a proposed belief can be condensed in this way: (GRl) A reason r is mutual recognition of each member having a substantial reason. But this definition of a good reason will not cover every situation in which members of a group come to decide their group beliefs. Reasonable overlapping consensus may be more common than we expect, but it will not always be available. Members will often face disagreement. When they do, they will have to adopt one of the personal beliefs in question and then revise it to be accepted as a group belief. Or, they must work to form an entirely new group belief. In these circumstances, members will likely seek a jointly accepted substantial reason. Of course, a revised belief might be accepted as a reasonable overlapping consensus if all the members accept it as their personal belief and recognize that each member has a substantial reason for that belief. However, the process of revision will usually require discussion about why and how a proposed personal belief is to be revised. If so, the members of the group are already seeking a substantial reason. Therefore, not every group belief can be justified under the above definition of a good reason. In addition, a group’s members will seek a substantial reason when they must achieve a complete consensus, which means that all of the members agree not only on a belief, but also on its supporting reasons.7 In contrast, overlapping consensus can be called a partial consensus because all members can agree on a belief without agreeing on 7 Complete consensus is often called fully deep consensus. 130 its supporting reasons. Therefore, if members are faced with a situation in which they must agree on the reason for a belief that has already obtained reasonable overlapping consensus, they must also identify a mutually accepted substantial reason for it. This may occur when a group attempts to explain its agreement to persons outside of the group or to the public. Thus, we often need to interpret a good reason as a substantial reason. (GR2) A reason r is a substantial reason. Thus, we can give the definition of group moral justification by adding (GRl) and (GR2) to Schimitt’s definition: A group G is justified in believing p just in case G has good reason to believe p, where G has a reason r to believe p just in case (1) all members of G would properly express openly a willingness to accept r jointly as the group’s reason to believe p and (2) a reason r is a substantial reason or the mutual recognition of all the members having their own substantial reason. This definition of group justification has the important merit of respecting both the coherentist and foundationalist positions regarding moral justification. It does not indicate what kind of substantial reason must be adopted. That determination is up to a group’s members. So, when a group is to decide whether a moral judgment can be accepted as a group moral judgment, members who follow principlist reasoning will provide moral principles as reasons; casuists will present other particular moral judgments; and, interestingly, followers of the method of WRE may provide moral principles, other particular moral judgments, or both with relevant background beliefs. What reason the follower of the method of WRE finally provides will depend upon the situation. If the situation requires a local level of justification, he/she need not provide all of the relevant beliefs that cohere with the pr0posed belief. But if the situation demands a global level of justification, he/she will do so. It seems, then, that all reasonable members 131 of a group can accept the above definition of group moral justification. If principlists or casuists argue that we must specify the kind of substantial reason required in advance — whether that be moral principles or other particular moral judgments — they are not being reasonable because they do not respect alternative methods of reasoning. It is highly doubtful that all reasonable members will ever agree that all reasons for a group moral judgment must always consist of principles or other moral judgments. Members should be allowed to come to the same conclusion by what they take to be the best conception of moral reasoning whatever it be by appeal to moral principles, appeal to other moral judgments, or some other reasonable conception of moral reasoning. I have shown that the above definition of group moral justification incorporated into the group method of WRE can respect other reasonable persons who follow different methods of reasoning. It remains to be shown, however, that the group method of WRE is a more plausible form of group reasoning than are the other methods. It is to this that I now turn, via a discussion of how the group method of WRE works with respect to different types of group decision. I will consider overlapping consensus, complete consensus, and compromise. 2. The Group Method of WRE and Consensus 2.1 . Overlapping Consensus Rawls’s “overlapping consensus” refers to agreement regarding basic principles 132 of justice among persons holding different, occasionally conflicting, reasonable comprehensive philosophical or religious doctrines.8 As the word ‘overlapping’ indicates, we can sometimes find beliefs on which different reasonable comprehensive belief systems overlap. This concept has useful implications for medical ethics. Benjamin says, “Individuals arguing from different moral, religious, philosophical, and empirical premises may nonetheless reach the same conclusion with respect to positions or policies ”9 in bioethics. Jonathan D. Moreno lists four moral principles that reach overlapping consensus in medical ethics: autonomy, beneficence, nonmaleficence, and justice.10 Overlapping consensus must be taken to mean reasonable overlapping consensus, which is distinguished from naive overlapping consensus in the previous section. It was important to distinguish them because na’r‘ve overlapping consensus may contain incorrect beliefs as group beliefs. More importantly, the process of mutual recognition of different substantial reasons for a shared belief plays an important role in group reasoning for several reasons. First, to reach overlapping consensus, every member of a group must recognize that members may have different substantial reasons for the same belief. The recognition process will improve mutual understanding of different perspectives. Mutual understanding can help members predict which beliefs (or reasons) will or will not be accepted mutually by all members. Without this, the process of reaching agreement will not likely go smoothly. Members must understand one another’s perspectives. If they fail to understand the various perspectives in play, they will be unable to know whether or not 8 John Rawls, Political Liberalism (Columbia University Press, 1996), 134. 9 Benjamin, “The Value of Consensus,” 243. '0 Jonathan D. Moreno, Deciding Together: Bioethics and Moral consensus (Oxford University Press, 1995), 61. 133 a given disagreement is reasonable disagreement. Thus, reasonable overlapping consensus is preferable to naive overlapping consensus for group reasoning. Second, the process of reaching reasonable overlapping consensus will better enable to members to discover the breadth of considered moral judgments, principles, norms, values, and background beliefs shared by all members. Thus, the process helps establish a group belief system. It will provide a shared perspective. When members know where a shared point of view is, they will have fewer difficulties reaching consensus even on novel issues. Furthermore, coherentists do not consider belief formation to be built up from nothing or fi'om several self-evident beliefs. Rather, they accept new beliefs or revise old beliefs on the basis of their coherence with other beliefs. Therefore, the existence of a group belief system plays an important role in the group method of WRE. The method can begin with beliefs belonging to a group belief system. Third, a group belief system established through the process of mutual recognition is not likely to have its beliefs hierarchically arranged. Finding or reaching overlapping consensus will not proceed systematically simply from principles to particular judgments or from particular judgments to principles. Beliefs obtaining overlapping consensus are likely to be collected somewhat randomly and recognized by chance, as occurs in our real life. Group belief system obtained through reasonable overlapping consensus will not significantly differ from any particular reasonable comprehensive belief system, though it may, in the interest of practical or workable agreement, exclude controversial philosophical or religious beliefs. 134 2.2. Complete Consensus Although overlapping consensus might be reached more often than expected, we are sometimes in need of complete consensus, as mentioned above. When an ethics committee must develop a health care policy or when a government committee must explain its position regarding a current ethical issue, we expect it to explain its views with reasons and not just assert their conclusions. In addition, even when members reach overlapping consensus, they may need to specify substantial reasons in order to explain their agreement to the public or others outside their group. How should a process of reaching a complete consensus proceed? When seeking agreement on both substantive position and the reasoning for it, group members must identify substantial reasons that are jointly accepted by all members. When they discuss, they will try to provide reasons that are as acceptable as possible to all other reasonable members. They must also be ready to discuss their proposed belief and its supporting reasons in terms of the group belief system. They will present an argument. This argument may differ somewhat from their own personal arguments. Some of the premises accepted in their personal reasoning may be revised to be made acceptable to other members because they can in part predict which beliefs have potential to belong to the group belief system. Or, they will try to find an alternative premise in the group belief system if possible. But if members must present new beliefs, they must evaluate the likelihood of these beliefs being accepted as group beliefs. The process may look complicated, but it can be categorized into two reasoning activities. The first involves presenting an argument, the second evaluating an argument (which is 135 also important for the review of alternative positions regarding an issue). Presenting and evaluating arguments are in fact interpersonal, social epistemic activities. The group method of WRE should also be considered a social epistemic practice, as it is aimed to produce socially justified moral beliefs. Alvin I. Goldman recommends fourteen rules of good argumentation as a social epistemic practice aimed at knowledge.11 These rules are also valuable for moral argumentation, and in particular, to the group method of WRE. I will emphasize only three of the rules here. These three are primarily related to defeaters, which deserve our attention in the process of providing and evaluating arguments. The first is this: (R1) Critical argumentation may challenge or rebut a speaker’s argument by either (A) presenting a defeater, (B) denying the truth of some premises, or (C) denying the strength of the premises-conclusion relationship.” This may sound plain, but it shows which parts of an argument are the best targets for evaluation. If other members do not have defeaters, a proposed argument will be successful. The existence of defeaters will undermine a conclusion even when the relationship between premises and conclusion is strong. The second is as follows: (R2) In selecting responses to critics, a justifiably undissuaded proponent should weigh both the perceived seriousness of the criticisms by the audience and the audience’s comparative receptivity to the potential responses available to him.'3 This rule encourages an arguer to weigh objections from the audience and the audience’s receptivity to his/her response. Lastly, the third rule points out that good argumentation " Alvin I. Goldman, Knowledge in a Social World (Clarendon Press, 1999), 134-144. He presents fourteen rules. Most of them can be also applied to moral argumentation. '2 Goldman, Knowledge in a Social World, 140. ‘3 Goldman, Knowledge in a Social World, 143. 136 must include responses to existing or possible objections: '4 (R3) When there are existing or foreseeable criticisms of one’s main argument, 3 speaker should embed that argument in an extended argumentative discourse that contains replies to as many of these (important) criticisms as is feasible.15 To summarize his points, a good argument should have no defeaters, no false beliefs, and a strong relationship between its premises and conclusion. In addition, good argumentation should include responses to existing or possible objections, taking into account the seriousness of the objections and the audience’s receptivity to the responses to those objections. Most of these points were implicitly captured in my discussion of the method of WRE as an interpersonal epistemic activity. When we take the method of WRE to be an interpersonal epistemic activity, we should consider competing views and possible objections and attempt to respond to them. Argumentation within the group method of WRE will mainly be “actual” interpersonal epistemic activity. The interpersonal reasoning activities of the group method of WRE will be different from those of WRE as a method of personal reasoning. Defeaters will differ between the two. A successful defeater in group reasoning will be a belief accepted jointly by all members. In WRE as personal reasoning, one may have a richer set of defeaters found from one’s perspective. To reach complete consensus, however, the reasons as well as the proposed belief must be accepted jointly by all members. Defeaters must also be accepted by all members. When one’s defeaters or reasons are initially grounded in explicitly religious terms, they should be expressed in more secular terms. If '4 I experienced the importance of this point when assisting in Martin Benjamin’s classes. He requires students to include the possible strongest objections to their own view and to respond to them in their papers. 137 one’s reasons cannot be accepted without commitment to one’s religious views, he/she will, as a committee member in a religiously pluralistic society, have to give up those reasons or try to find some alternatives from the group belief system. This does not imply, however, that beliefs can be easily accepted as group beliefs. Less controversial beliefs will more likely be accepted. A more or less restricted version of the principle of utility, for example, must be accepted. But, the view that all morality can be explained by the principle of utility or that all human conduct should be assessed only from the utilitarian perspective will not be accepted. A version of Kant’s categorical imperative will likewise be accepted. But members will not hold it as absolute or the only ethical principle. We must also take consequences into account. But the view that the justification of our actions is only a matter of their consequences will not be accepted. This will result in conflicts between moral principles, but it is better than giving one ethical theory supreme power. Therefore, the four moral principles of autonomy, utility, beneficence, nonmaleficence, and justice may be more balanced in this group belief system than in a particular reasonable comprehensive belief system. There will be fewer socially justified moral beliefs than personally justified ones, but the former will be more acceptable to all of the group members than the variety of personal beliefs. Finally, beliefs in a group belief system are likely to be equal in their epistemic status, much like those of a coherentist belief system. Therefore, the group method of WRE pursues coherence in the light of a system of mutually accepted beliefs, while the standard method of WRE seeks coherence from one’s personal perspective. Both methods involve interpersonal epistemic activities, but the perspectives from which coherence is sought are different. '5 Goldman, Knowledge in a Social World, 144. 138 With this in mind, members of a group must decide whether a proposed belief should become a group belief. Complete consensus may extend a group belief system by a process in which personal beliefs are adopted, revised through a give-and-take discussion, and then accepted as group beliefs. In addition, members may face difficulties in interpreting moral principles, lessons of paradigm cases, or relevant concepts already accepted as group beliefs when their interpretations conflict. Members must choose together the most appropriate interpretation. This shows that a group belief system should be revised or broadened, as are particular personal belief systems. Thus, members try to resolve or adjudicate disagreements using agreements already reached. Particular belief systems will be revised or broadened through facing objections and responding to them in a give-and-take discussion. Therefore, interpersonal reasoning activities provide a place for interaction between the group method of WRE and the method of WRE as personal reasoning. Would reasonable persons following other methods of reasoning, however, use the above process outlined by the group method of WRE? The group method of WRE, in fact, incorporates two other methods. In the process for complete consensus, members who follow principlist or casuist methods of reasoning will contribute to the overall discussion, which finally proceed to coherence among the three elements of the framework of WRE. Persons who follow the group method of WRE will hope that their own method and the other methods converge on the same conclusion. If they fail to converge, equilibrium has not been achieved. Therefore, persons following the group method of WRE contribute to the group discussion from the perspectives of the other two methods and try to learn from those using the other methods. They will, however, play 139 the role of balancing the other two methods and of emphasizing coherence with relevant background beliefs. Those following the principlist or casuist methods may, however, object on two grounds. First, they may complain about writing a group statement. Suppose the principlists insist that the final explanation of a group moral judgment be made by an appeal to mutually accepted moral principles. But, coherentists would be glad to accommodate their request because any conclusion must, for coherentists, cohere with accepted moral principles. The principlists are simply focusing on the final stage of the process rather than the whole procedure. The explanation they favor may simply be their preferred way of viewing the matter. The justification of a group statement will not usually occur at the global level of justification. The reasons given for a group agreement will usually be at a local level of justification. Thus, coherenists will not see any problem in accommodating the complaint. Casuists may insist the use of particular moral judgments from paradigm cases in the final statement. But this complaint can also be accommodated. Second, principlists and casuists might complain that the group method of WRE cannot accommodate, in the final statement, the conflicts between principlist and casuist ways of providing reasons if each party insists on having its own way at the same time.16 If the conflict is not mediated, members might be able to reach reasonable overlapping consensus, but not complete consensus. If both types of reasoning are included in the final explanation, coherentists will be happy — the issue is not a serious one to them. But if the final statement must be made by choosing one type of reasoning (principlist or '6 Of course, if their disagreement is merely regarding a substantial reason, the supported belief can be accepted as a reasonable overlapping consensus. But the issue is raised when members must reach complete 140 casuist) with the concession of the other party, those who yield will have to acknowledge the chosen reason as a group reason, but they will not personally accept it as a substantial reason. They will not be fully satisfied with the group reason jointly accepted. This kind of discrepancy is in fact foreseeable in group reasoning in more serious ways. There can be, for example, conflicting answers to issues even between people following the same method of reasoning. Not all personal beliefs can be accepted as group beliefs. This will cause some odd situations. 3. Compromise and Integrity in the Group Method of WRE 3.1. Discrepancies Between Group and Personal Beliefs Group reasoning may produce some examples of discrepancies arising between a particular member’s personal beliefs and his/her group beliefs. Let us consider the following three situations. In the first case, a belief personally held is not accepted by others as a group belief, though there is no serious conflict between them. If a personal belief is justified by particular philosophical or religious commitments, the personal belief may not be accepted as group beliefs. For example, Paul, a philosopher, considers higher brain activities to be essential for personhood, such that if these activities are absent, even though a fiinctioning brainstem maintains respiration and circulation, a living human body but not a living person is present. Thus, Paul believes that personhood is essential for being alive and that the criteria of death should be higher brain death. Suppose further that Paul is a member of an ethics committee that is considering how the C ODSCDSUS. 141 current medical system can improve organ procurement. Paul might argue that we should change the criteria of death from whole brain death to higher brain death. But he did not propose this to the other committee members. He thought that his belief would not be accepted by them all. All the members including Paul discussed the issue accepting the current criteria of death as their group view. Therefore, Paul’s group belief regarding the criteria of death is different from his personal belief. Second, sometimes group beliefs contradict personal beliefs. Suppose the members of a group are trying to decide whether a patient, who is involuntarily pregnant, should be allowed to abort. Suppose further that most of the members believe that abortion is permissible, though one, Susan, personally believes that abortion is morally wrong. She is Catholic and thinks that she herself would not abort even if she were involuntarily impregnated. However, Susan decides not to share her personal views and accepts jointly the final judgment that the patient should be allowed to abort. Susan thinks that her personal position is supported mainly by her religious beliefs, and she is not ready to discuss these with others. She does not want to force the other members to accept her religious beliefs. In addition, she came to know that the patient in question was not Catholic. Susan’s group belief regarding abortion conflicts with her personal belief. She will not, however, change her personal belief. A third situation is one in which members cannot reach consensus because reasonable disagreements exist between personal beliefs. Suppose in the above situation that some of the members hold the same personal belief as Susan and are ready to discuss their reasons. Some of the reasons may be related to religious beliefs, but others of them are not entirely religious. These members are able to present alternative reasons drawn 142 from a system of mutually accepted beliefs. These members argue that abortion is morally wrong. But other members of the group have different position. They argue that abortion is morally permissible. The group is unable to reach consensus. Both parties see the other side as not entirely wrong and they all understand what is preventing them from aniving at a consensus. Despite this, they must reach consensus and explain their decision to the public. A similar situation can occur when a group identifies substantial reasons although all of the members reach the same conclusion. The conflict between principlists and casuists is of this sort. Suppose that all the members of a group are physicians. They reach an overlapping consensus that physician-assisted suicide is not permissible. People outside of the group ask them to explain their decision. They must present a unified voice to justify their group belief. During their discussion, some of the members claim that assisting a person’s suicide is a violation of the physician’s professional moral code. But others claim, pointing to Jack Kevorkian as an example, that physician-assisted suicide should not be permissible. Of course, the final statement will be able to accommodate both kinds of explanation. In this respect, this is not as serious a conflict as was the above situation regarding abortion. But if principlists and casuists insist on only one substantial reason, this situation may similarly problematic in that there is no solution except for compromise. The above three circumstances raise several questions. In the first two cases, have Paul and Susan lost their personal integrity? Should they express and defend their personal beliefs even though they expect their personal beliefs would not be accepted as group beliefs? In addition, can their actions thought of as compromises? In the third 143 situation, what should the group members do? If they consider a compromise, on what grounds are they to decide to compromise? Furthermore, if they compromise, how can they preserve their personal integrity? Because these situations may occur often, the group method of WRE must respond to these questions. 3.2. Implicit and Explicit Compromise Compromise may be the only option for the two parties regarding abortion in the third situation above so long as they must reach consensus. Can Paul and Susan be seen as making a compromise? I consider their actions to represent compromises that are implicitly made. This kind of compromise seems to be inevitable when all the members of a group try to reach a complete consensus. Compromise may itself be a type of consensus. The two notions underlie compromise. First, as Benjamin says, “Central to compromise is the idea of mutual concession for mutual gain.”l7 Second, it places value, as does consensus, on a single voice from a group. He says, compromise “resembles consensus insofar as the group speaks with one voice in making and supporting a particular recommendation?”8 With compromise, however, there is a discrepancy between one’s personal beliefs and the beliefs expressed by the compromise position. According to Benjamin, “Those supporting [a] compromise retain personal moral views that are more or less at odds with the osition the endorse in their roles as committee members.”19 This shows that eo le P Y P P making a compromise are not fully satisfied, in the light of their personal belief system, '7 Benjamin, “The Value of Consensus,” 244. '8 Benjamin, “The Value of Consensus,” 245. 144 with a compromise position. This feature can be found in Paul and Susan. Even though all the members of a group might agree on their final group belief and its supporting reasons and hence reach complete consensus, they may still personally hold a different view and different reasons, as do Paul and Susan. Compromise, according to Moreno, “suggests that parties to a controversy have entered the scene with more or less fixed preferences” whereas “consensus carries no such baggage; it suggests an openness to unanticipated possibilities and points of view.”20 Paul and Susan entered their respective discussions with more or less fixed preferences even though they did not present their views to the group. They would not have been fully satisfied with their group beliefs. In this respect, they compromised. They did, though, place value on a unified group voice as do people in consensus. If Paul and Susan are taken to make a compromise, what was the mutual concession and gain in their situations? They made the concession of not presenting their personal views and gained in the sense of furthering the group’s single voice. But there was no concession by the others. There was no mutual concession or gain. Paul’s and Susan’s concession and gain take a place within an individual person. Therefore, they did not make a compromise explicitly. However, their decisions not to raise their respective personal beliefs do involve some aspects of compromise. It is for this reason that I call their actions implicit compromise to distinguish it from compromise explicitly made. Implicit and explicit compromise permeate complete consensus. Moreno might object to my view. Members may, however, face situations needing compromise even though they enter the discussions without more or less fixed preferences. During a '9 Benjamin, “The Value of Consensus,” 245. 2° Moreno, Deciding Together: Bioethics and Moral consensus, 45. 145 discussion, members may reach different positions. Each party will be unable to yield its position. So long as openness does not always result in a consensus that satisfies every member, explicit or implicit compromise cannot be avoided in the process striving towards complete consensus. More importantly, if members modify proposed personal beliefs a bit to become group beliefs, compromise may be observed more often than we think. Suppose that Susan is ready to discuss her reasons against abortion. She presents her personal beliefs and other members come to agree with parts of her view. Upon reaching mutual understanding, they might reach an agreement such that abortion is morally permissible only in the case of involuntary pregnancy. There is a consensus, but also compromise because each member may continue to personally hold the same belief as before. If the group belief were the same as each member’s personal belief, each member would be happier. The conflict between two parties of a group in the third situation above shows how difficult it is to reach a complete consensus. Benjamin points out two difficulties that might arise when an ethics committee seeks a complete consensus. First, the issues that it must consider are usually contested.21 We usually ask such a committee to consider an ethical issue when we are uncertain or have conflicting views regarding that issue. Second, “committee members,” Benjamin says, “usually represent differing social or ethical viewpoints or differing areas of biomedical, social scientific, or other types of expertise, or both?22 It is not easy to produce a complete consensus without compromise due to this diversity. Therefore, if we must acknowledge the possibility of discrepancies 2' Benjamin, “The Value of Consensus,” 242. 22 Benjamin, “The Value of Consensus,” 242. 146 between group beliefs and personal beliefs in the process of reaching a complete consensus, and also must admit difficulties in reaching it, it will be useful to ask when members of a group should consider compromise. However, there are reasons why compromise has been ignored in ethical debates. For example, compromise has been thought to diminish personal integrity. It is often understood as making a deal or bargaining, and bargaining is not considered to be a good way to reach justified moral beliefs. Generally speaking, bargaining is a matter of business, rather than a matter of truth or morality — whether a belief about the world is right or wrong is not decided by bargaining. The same point can be applied to morality. Compromise, however, may not always harm integrity. Compromise in the group method of WRE should be toward a justified moral belief mutually accepted. The mutual gain is the acquisition of a justified moral group belief. The mutual concession is the withdrawal of proposed personal beliefs or the revision of them to be made group beliefs. During the process of group reasoning, then, members may give up or revise their proposed personal beliefs as a mutual concession for the mutual gain of group beliefs. The group method of WRE will contain some requirement of acceptable compromise. First, compromise will be made when members recognize reasonable disagreement among members. To know that their disagreement is reasonable, they must know why they cannot say which proposed belief is better. They must recognize that the proposed view is not based on ignorance, bad reasoning, selfishness, bias, etc., and also that it is not the only reasonable view of the matter. All members thus must recognize why they make different judgments. Without this recognition, the group decision whether to compromise could not be an informed and uncoerced agreement. 147 Mutual understanding provides a perspective by which members can devise a compromise position. Therefore, a compromise position generated by the group method of WRE should cohere with the group belief system. This will prevent compromise from violating at least the integrity of the group. 3.3. The Group Method of WRE and Personal Integrity The remaining question is how personal integrity can be preserved despite compromise. The group method of WRE will include some important background beliefs regarding compromise. First, as Rawls’s reasonable pluralism shows, we must tolerate some conflicting reasonable beliefs that are justified in their respective reasonable belief systems. Of course, either of two competing beliefs may be taken to be more justified in terms of one’s preferred reasonable belief system. But the relevant issue is which belief is more justified from the perspective of the group belief system. If neither is better from this shared point of view, compromise will be the best option available. One of the competing views cannot be accepted as the group view without suppressing the other in such a situation. Therefore, members must recognize that their personal beliefs will not always be the same as the group beliefs. Second, members recognize that they must sometimes present a unified voice, and come to appreciate its values. Benjamin puts it this way: Members of the committee holding either of these positions who nonetheless agreed to the compromise, must be presumed, for one reason or another, to have valued the committee’s speaking with one voice on this matter more than they valued the committee’s endorsing their own view at 148 lllllllllillllllilllillilillllilll Fill-ll 41‘1““ l‘ the price of continued impasse.23 If they do not value a unified voice, compromise may not be necessary. However, most of the ethical issues in health care are related to health care policy-making, which needs a single voice. Without a new policy, the status quo may be maintained and will likely favor one position over another regardless of our intention. For example, if we do not decide whether physician-assisted suicide is legalized, we are left favoring patients who suffer unbearably and want to bring their lives to a close by refiising treatment that keeps them alive over patients who are in similar situation, but do not involve such treatment. Therefore, because reasonable members understand reasonable disagreement and the need of tolerating reasonable disagreement, rejecting all compromise may hurt the integrity of a reasonable person who must make a group decision immediately. Third, most decisions will have to be made by a designated time. A committee will usually be requested to reach an agreement and to submit a report by a certain date. This is our reality. Group reasoning cannot always aim at the most justified moral beliefs. When a group reaches complete consensus on a compromise position, this does not mean that the group has reached permanent equilibrium. Fourth, members, from their own personal perspectives, may not be fully satisfied with their final group belief that a compromise position holds. But this does not mean that the group belief is not a justified moral belief. The group method of WRE requires a compromise position to cohere with the group belief system. As our knowledge improves and as mutual understanding is extended, members should try to reach new group beliefs with which they will be more satisfied than the current compromise position. Therefore, when the group method of WRE includes the above considerations in 23 Benjamin, “The Value of Consensus,” 244-245- 149 its background beliefs, compromise will not be considered harmful to personal integrity. The above background beliefs provide guidance when members should seek compromise. However, compromise does not imply that members must change their personal beliefs. They may continue, apart from their roles as committee members, to argue for their personal views and to persuade others to accept them even though they mutually accept the compromise position. Rather, they should argue for their personal beliefs outside the group. Such activities will improve further mutual understanding and encourage firrther give-and-take discussion. A group belief system must be broadened by encountering beliefs from the most justified personal beliefs. The group belief system cannot be made richer without adopting and revising personal beliefs that are justified in particular reasonable belief systems. I will illustrate in Chapter Seven how the group method of WRE explained here can be used in order for a group to arrive at answers to debate over the legalization of voluntary active euthanasia and physician-assisted suicide. 150 Chapter Six Wide Reflective Equilibrium and Judgment In previous chapters, I defended the method of wide reflective equilibrium (WRE) from objections to coherentism, clarified that the method can incorporate moral experience, and extended the method to group deliberation. When it was first presented by John Rawls and developed by Norman Daniels, WRE was thought of primarily as a method for evaluating theories of justice (Rawls) or ethical theories (Daniels). Tom L. Beauchamp and James F. Childress then considered WRE as an explicit methodology for biomedical ethics — that is, moral reasoning for the justification of moral judgments. Moral reasoning aims epistemologically at justified moral belief. Moral beliefs vary in their generality, from ethical theories to particular moral judgments. Thus, the method of WRE can be used to evaluate ethical theories, moral principles, moral rules, and particular moral judgments. The evaluation of a particular moral judgment is in fact the same as the moral evaluation of a particular action because a particular moral judgment is about whether a particular action is morally obligatory, permissible, or prohibited. From a practical perspective, thus, moral reasoning addresses questions of what to do. As our beliefs and actions are interwoven, so are the evaluations of moral beliefs and actions. Thus, I characterize the method of WRE as practical moral reasoning. I add ‘practical’ because the methods provide answers about what to do. If I call the method of WRE the standard method to distinguish it from the group method of WRE, the standard method of WRE can serve as a method of personal practical moral reasoning while the group method can provide a group with a method of practical moral reasoning for 151 deciding matters of policy. As practical moral reasoning, the standard and group methods of WRE include two kinds of moral evaluations of an action. One is an evaluation of an action already taken. The other is an evaluation of an action yet to be performed. When we wonder what to do, we consider which action is morally preferable and then choose the preferable one. The question arises, however, as to whether the methods of WRE can give us determinate answers about what to do. The process of reaching a conclusion using the methods of WRE is characterized as a back-and-forth process of revision aimed at coherent comprehensive personal or group belief systems without incorrect beliefs. But there must be different ways of revising beliefs. When we follow the methods of WRE, we are faced with a variety of questions: With which belief should the method begin? Which beliefs should be selected and how should they be revised to increase coherence? When should the revision process stop? When must we compromise? How should we devise a compromise position that all members can accept? I argue in this chapter that we arrive at our answers to those questions through the exercise of judgment,1 rather than by any algorithmic decision procedure. Some may expect a mechanical decision procedure by which to reach answers to the above questions, but this is misconceived. ' I use ‘judgment’ to indicate a faculty of thinking or a mode of thinking. I use ‘a judgment’ or ‘judgments’ to refer to a belief that judgment produces. 152 1. Moral Reasoning and Judgment 1.1. Judgments in the Methods of WRE Many practical questions can arise in the course of using either the standard or group method of WRE. We may wonder, for example, which belief should be chosen as a provisional fixed point. In principle, we can start with any considered particular moral judgment, moral principle, or background belief. But how should we decide which belief should be selected? Our choice of provisionally fixed belief may be important because it can affect the efficiency of the rest of the reasoning process. In addition, when faced with a loss of equilibrium, which beliefs should we revise to improve the overall coherence of the belief system? One may answer these questions by simply saying that our criterion is coherence. However, to know that coherence is an important criterion is one thing, to decide how a problematic belief should be revised is another. This is because we will face various circumstances requiring belief revision and cannot say in advance exactly how to revise problematic beliefs. As shown in current ethical issues related to health care, we now face various circumstances in which our equilibrium has been upset. These are usually unprecedented novel situations brought about, for example, by IVF, cloning, and genetic engineering. In some cases, we cannot predict exactly what the results of a new procedure will be. We may be able to utilize genetic engineering, but cannot know the long-terrn consequences. We must decide what to believe and what to do in the light of factual uncertainties. In some cases, such as abortion, we must make decisions about what to believe about a 153 given circumstance with metaphysical uncertainties. Thus, our decisions about which beliefs to select and how to revise them must be made in these unprecedented or unpredictable situations. Similar questions can also be raised with respect to the group method of WRE. The members of a group may sometimes have to compromise. They must decide whether they are in a situation requiring compromise — that is, whether it is best to compromise or to refrain from issuing a final conclusion. But again, there will be various situations that lead the members of a group to consider compromise. If they decide to compromise and develop a compromise position, they also must decide what beliefs they must craft into a compromise position. They will aim at a compromise position that is more well-grounded and integrity-preserving than any alternative. But, to know what requirements are relevant is one thing, to decide whether a proposed position in a particular circumstance satisfies those requirements is another. I claim that there is no algorithmic decision procedure by which to arrive at determinate answers to the above questions raised in the context of using the methods of WRE. Answers to those questions are in fact a matter of judgment. As Martin Benjamin says, “so long as we must make decision in unprecedented or unpredictable situations and circumstances, we will have to rely on judgment.”2 1.2. Calling for Judgment Why can we not have an algorithmic decision procedure if we must make 2 Martin Benjamin, Splitting the Difference: Compromise and Integrity in Ethics and Politics (University Press of Kansas, 1990), 112. 154 dCCisions in unprecedented or unpredictable situations? Benjamin presents a comparison of the evaluation of the performance of speed skaters and the performance of figure skaters.3 When we evaluate the performance of speed skaters, our decision of who is the fastest may be mechanized. A computerized machine can evaluate their performances more accurately than can we. When we evaluate the performance of the figure skaters, however, can we let a computerized machine replace our decisions? One element that a panel of judges must evaluate is the artistry of the performance. Regarding this, Benjamin says: One cannot antecedently specify determinate criteria for evaluating this component of a figure skater’s performance without denying his or her artistic creativity. To presume to do so is to imply that the skater will be unable to come up with anything of artistic merit that the judges have not already conceived.4 The performance of a figure skater must be evaluated in an unprecedented situation. Each performance may be unique. Some criteria that judges had before will be unable to be determinately applied. Those criteria must be interpreted in evaluating a given performance or new criteria may be found. When we decide who is the fastest, we must know about the concept of “the fastest.” Our knowledge about this concept tells us which property we must note in the particular performance of speed skaters. This is the single property (or criteria) — who reaches the finish line first. This property can be identified in any particular performance at any time or place. There are no unprecedented or unpredictable situations with respect to identifying the skater with that property. Thus, we can evaluate mechanically the 3 Benjamin, Splitting the Difference, 112-113. 4 Benjamin, Splitting the Difference, 113. 155 performance of speed skaters. We can build a computerized machine by programming it to identify that property. In this case, the relevant concept is simple. It indicates a single definite property that must be identified in particular performances. Among the relevant concepts in the evaluation of a figure skater’s performance, however, is artistry. The concept of artistry is so abstract and general that we cannot in advance precisely describe the particular properties that we must identify in particular performances. We cannot determine exactly what concrete properties constitute artistry before we see particular performances. We must identify the properties that satisfy the concept of artistry by looking at particular performances. The properties we must identify in particular performances are like moving targets rather a fixed one. Our evaluation in terms of artistry must rely on judgment and cannot be mechanized. The example of evaluating a figure skater’s performance demonstrates two points. First, if our decisions must be made in unprecedented or unpredictable situations, we must rely on judgment. We cannot in advance determine what properties of particulars will satisfy the concept relevant to certain decisions. Second, if the concept relevant to decisions is very abstract and general, we may know roughly which properties are relevant to the concept, but we carmot determine in advance and in detail what concrete properties of particulars will satisfy the concept. Therefore, our decisions in such circumstances call for the exercise of judgment and cannot be made through a mechanical or algorithmic decision procedure. As discussed earlier, situations in which the methods of WRE are used are unprecedented or unpredictable. Just as a panel of figure skating judges do not know exactly what performance will be presented, so we do not know in what situation the 156 methods of WRE will be used. We must decide how to revise our beliefs or to devise a compromise in an unprecedented or unpredictable situation. This is one of the reasons we must rely on judgment. Just as the panel of judges has only the concept of artistry, so we have only some relevant concepts like coherence and integrity-preserving compromise. Most of the relevant concepts involving the use of the methods of WRE are abstract and general. We must have knowledge about coherence, integrity, morality, moral rightness, autonomy, justice, competence, informed consent, and the best interest. However, those concepts are as abstract and general as the concept of artistry. Abstract concepts and unprecedented or unpredictable situations are often part of practical moral reasoning. Suppose we are considering whether a severely impaired newly born baby should receive aggressive treatment. “The best interest for the baby” may be an important relevant concept. But whether providing or withholding a treatment furthers the best interests of the baby is difficult to answer. Because the concept of best interest is abstract, we do not know exactly which properties we will have to pay attention to in any given situation. Properties that contribute to the best interest of the baby will vary according to the informed judgment of the physician, the parents of the baby, and third parties. The greatest difficulty is that they cannot predict how the situation will develop after providing or withholding a treatment. Thus, the situation requires the exercise of judgment. In addition, revising beliefs in WRE or devising a compromise position requires a kind of creativity. These tasks require more than just putting together the current beliefs. Benjamin stresses this point as follows. 157 A well-grounded, integrity-preserving compromise will often require a good bit of creative imagination. Parties seeking compromise must be also be sensitive to the situation and feelings of those involved. Accepting a moral compromise usually involves relinquishing, at least temporarily, part of one’s world view and way of life. We should in negotiating compromise be sensitive to this loss by acknowledging and attempting to ameliorate it.5 Even though he mentions creative imagination and sensitivity to the possibility of strains on integrity in the context of compromise, the point is also true of the revision of personal beliefs. When we attempt to revise our beliefs, we require not only creative thinking for revision, but we must also be sensitive to the possible loss of integrity. More importantly, the concepts or values relevant to our decisions will not always be a single concept or value. They may sometimes conflict. Thus, we must decide how to balance them. But again, we cannot determine in advance how to accomplish such balancing without looking at a particular situation. Therefore, answers to practical questions raised in the use of the methods of WRE are often a matter of judgment. The methods cannot be employed without the exercise of judgment. 2. The Exercise of Judgment 2.1. Judgments in Principlism and Casuistry Judgment is not unique to the methods of WRE. Other methods, such as principlism and casuistry, also rely on judgment. One may expect that principlists could deduce a particular action from a moral principle. But even principlists require the faculty of j udgrnent. When principlists attempt to apply moral principles to a particular case, they 5 Benjamin, Splitting the Difference, 124. 158 must decide which of their moral principles covers the case, just as a judge would have to decide which law or regulation is relevant to a given case. Because principles are abstract and general, they must be interpreted in the light of the details of the particular case. In addition, the application of moral principles becomes more difficult when the relevant principles and values conflict. More importantly, the application of principles to a given case requires knowledge of whether the situation has the features necessary for the application of those principles. Suppose we are wondering whether a physician should allow a patient to refirse life- sustaining treatment. If the patient is competent and his/her decision is informed and autonomous, the refusal of life-sustaining treatment would be justified under the principle of autonomy. But this does not answer the questions of whether the patient is in fact competent and whether his/her decision is autonomous. Tom Tomlinson puts the point this way: Some implicit judgment must always be made whether this patient is competent and informed enough to refirse treatment. The gap between principle and application is unbridgeable except by some non-principled mode of thinking.6 J udgrnent, then, is needed in the operation even of the principlist method of reasoning. As Onora O’Neill says, principles “must be complemented by judgment.”7 This point leads us to consider what is problematic in deductive models of moral reasoning. Benjamin explains this model of moral reasoning as follows. Dubbed by Arthur Caplan the engineering model of applied ethics, this approach requires first that we acquire knowledge of ethical theories and then that we “apply” them by: “a) deducing conclusions from theories in light of relevant empirical facts and descriptions of circumstances and b) 6 Tom Tomlinson, “Casuistry in Medical Ethics: Rehabilitated, or Repeat Offender?” Theoretical Medicine 15 (1994): 6. 7 Onora O’Neill, “Practical Principles and Practical judgment,” Hasting Center Report 31, no. 4 (2001 ): 18. 159 analyzing the process of the deduction.”8 As the passage shows, we need knowledge about ethical theories. Ethical theories are about moral concepts, value, principles, and virtues, etc. Most of relevant concepts are abstract and general. Let us call this knowledge universal knowledge because knowledge is about universal entities like concepts and properties. We can find another type of knowledge in the above passage. That is knowledge of “relevant empirical facts and descriptions of circumstances.” Facts and circumstances are particulars. Thus, let us call this kind of knowledge particular knowledge. Particular knowledge must tell us whether relevant facts and circumstances have properties addressed and related by universal knowledge. In the above example of the refusal treatment, universal knowledge centers on the concepts of autonomy, competency, and informed consent. Particular knowledge is about whether the particular patient is competent and whether he or she makes an informed and autonomous decision. Universal knowledge and particular knowledge are found in a practical syllogism. R. M. Hare says, “There are two factors which may be involved in the making of any decision to do something,” and he continues, “they correspond to the major and minor ”9 (My italics) The major premise premises of the Aristotelian practical syllogism. involves universal knowledge while the minor premise involves particular knowledge. The engineering model can be understood to be based on a practical syllogism. If we put the above example of the refusal of treatment into a syllogism, we will have the following deductive argument: Major: A physician should respect the refusal of treatment (even life- 8 Benjamin, Splitting the Diflerence, 107-108. The quotes in the passage comes from Arthur Caplan, “Carr Applied Ethics Be Effective in Health Care and Should It Strive to Be?” Ethics 93 (1983): 314. 9 R. M. Hare, The Language of Morals (Oxford University Press, 1952), 56. 160 sustaining treatment) based on an informed and autonomous decision of an adult competent patient. Minor: John is refusing the treatment. He is competent. His decision is informed and autonomous. Conclusion: John’s refusal of treatment should be respected by his physician. As seen, the argument is deductive. As the engineering model argues, if we have universal knowledge, the conclusion seems to be deduced in light of particular knowledge. However, this view just focuses on the final outcome of our moral reasoning. Actually, the conclusion is not deduced from universal knowledge, that is, ethical theories or moral principles. We arrive at conclusions from the interaction between universal knowledge and particular knowledge. Suppose John is at the age of fifty-seven, in the last stage of stomach cancer, and will live six month longer if he receives the suggested very expensive treatment. Suppose further that he feels a moral obligation to die.10 He thinks that the suggested treatment will be a big financial burden to his family. If he receives the treatment, his family will lose the money that has been saved for his children’s university tuitions. Thus, he decides to refuse the treatment. If we have this piece of practical knowledge, we will come to wonder if his decision is autonomous. We must in turn move on universal knowledge. What is the concept of an autonomous decision? Is his decision autonomous or coerced? I do not intend to deal with this issue here. But to say that his decision is autonomous, we must make clear the notion of autonomy, and then determine whether his refusal of treatment is, in this circumstance, autonomous. What I want to show here is that the conclusion is '0 John Hardwig uses the expression of ‘a duty to die’ in “Is There a Duty to Die,” in Ethical Issues in Modern Medicine, 5th edition, edited by John D. Arras and Bonnie Steinbock (Mayfield Publishing Company, 1999), 292-301. 161 not simply deduced from the universal knowledge, but involves the interaction between universal knowledge and the relevant particular knowledge. John McDowell, from whom I have borrowed the terms ‘universal knowledge’ and ‘particular knowledge,’ explains a practical syllogism of a virtuous person as follows: The virtuous person’s reliably right judgments as to what he should do, occasion by occasion, can be explained in terms of interaction between this universal knowledge and some appropriate piece of particular knowledge about the situation at hand; and the explanation can take the form of a “practical syllogism,” with the content of the universal knowledge, or some suitable part of it, as major premise, the relevant particular knowledge as minor premises, and the judgment about what is to be done as deductive conclusion.” (My italics) Therefore, although the appearance of the argument shown in a practical syllogism is deductive, the conclusion (or the practical judgment) is not straightforwardly deduced from the universal knowledge contained in the major premise. Particular knowledge requires our decision about whether a particular instantiates properties conceived as universal knowledge. This decision must sometimes rely on judgment and cannot be made deductively or algorithmically because this decision also involves interaction between the content of the universal knowledge and the relevant particular knowledge, as shown in the above John’s case. Thus, the conclusion of a practical syllogism cannot be made without the interaction between universal knowledge and relevant particular knowledge. The engineering model of moral reasoning merely focuses on the final logical structure of a practical syllogism. This model ignores the interactive relationship between ethical theories and descriptions of circumstances (or relevant empirical facts). So-called empirical facts or descriptions are not determined mechanically and sometimes require ” John McDowell, “Virtue and Reason,” The Monist 62, no. 3 (1979): 336. 162 the exercise of judgment. Therefore, judgment may often be needed for practical moral reasoning in two ways: First, we must judge which principle should be applied to a given case. Second, we must judge whether the case falls under the selected principle. Therefore, principles cannot tell us in any determinate way which particular action should be undertaken. The exercise of judgment is required. The hope of establishing an algorithmic method of moral reasoning by appealing to principlism cannot be successful. One might expect that casuists would call for the use of judgment. After all, they object to the deductive reasoning of principlism and argue instead for case-based reasoning. Casuistry emphasizes a carefirl review of each case in question. This requires an interpretation of each case. With this in mind, casuists usually suggest the use of analogical thinking employing paradigm cases. Similarities must be sought between a given case and paradigm cases. Irnportantly though, the recognition of similarity is not a mechanical procedure — it requires judgment to determine which features of two cases being compared are relevant. Thus, casuists must decide which paradigm case is best suited for the job. This means that they must judge which perspective is to be adopted. Tomlinson says, “One is now in the position to make the critical analogical judgment (rather than a deductive inference) which assesses which of the paradigmatic cases most ”'2 (My italics) closely resembles the case under debate. So, the exercise of judgment is not unique to the method of WRE. Two other common methods of moral reasoning must also rely on judgment. The sorts of judgments required in principlism and casuistry are in fact also needed in the method of WRE since it must also determine whether a proposed particular moral judgment coheres with moral 163 principles or with other moral judgments obtained from paradigm cases. Therefore, we can conclude that our practical moral reasoning cannot be done without the exercise of judgment. 2.2. Judgments: Discussible Decisions The need for judgment implies that there are no determinate answers for resolving a conflict between two arguers following same method of reasoning. If the revision process is not algorithmic, the method of WRE will be unable to provide a determinate answer. As shown by Rawls’s notion of the burdens of judgment, even informed reasonable people may have different answers to the same question because of different judgments. The same point is applicable to other methods so long as they must also rely on judgment. Two reasoners employing the same moral principles may reach different decisions because principles themselves cannot determine our actions. O’Neill explains the reason as follows: Practical principles, including ethical principles cannot, hence do not, require uniform action. Because the act descriptions incorporated in practical principles are invariably indeterminate they must underdetennine the action of those who confirm to them. Even if principles or rules are elaborated to incorporate many details and exceptions, indeterminacy is ineliminable and invariably limits the specificity of prescription.13 As she says, the indeterminacy of principles is ineliminable. The room between principles and particular actions is the space in which judgment must be exercised. But that we may not have a determinate practical answer to a question due to different exercises of judgment does not mean that judgments are merely arbitrary. We '2 Tomlinson, “Casuistry in Medical Ethics: Rehabilitated, or Repeat Offender?” 8. 164 can discuss and debate them. We often ask for justification of people’s judgments. Moral judgments in particular are often required to be justified. Suppose Paul is driving a car and is nearing an intersection when he notices that a traffic light has just turned from green into yellow. He must judge whether he can keep going or must stop. To decide whether to go or stop, however, he must consider his situation. He notices that he is almost reaching the intersection, judges that he is moving fast enough to pass through it while the light is still yellow, but notices that the car behind him is following too closely, and he thinks that the following car may hit his car if he stops because the road is slippery. Thus, he decides to continue to go even though this act may not be allowed if the current regulation is strictly interpreted. But suppose further that a police officer catches his keeping going and stops him. Paul would try to defend his judgment. He will try to provide reasons to justify his decision. As seen in the above example, we may argue for, defend, and justify our judgments. The exercise of judgment is not a matter of mere taste or arbitrary preference. Judgment is not like the choice between the colors of cars. Mere taste is not subject to debate or discussion, such that we can blame others for being wrong. We especially debate moral judgments since morality is a social matter. Marcus George Singer characterizes moral judgments this way: Moral judgments are judgments, not deductions; they are not themselves deduced; they can be supported, defended, argued for or against, justified or established, but not deduced.14 Judgment is not a mechanical or algorithmic process or arbitrary preference, but it sometimes requires reasons. Benjamin characterizes this aspect of judgment this way: '3 O’Neill, “Practical Principles and Practical judgment,” l7. '4 Marcus George Singer, “The Ideal of a Rational Morality,” Proceedings and Addresses of the American 165 “judgment falls somewhere between algorithmic decision procedures, on the one hand, . 15 and expressrons of mere preference or taste, on the other hand.” 2.3. Indispensable Principles and Values Why is judgment not arbitrary? As discussed earlier, judgments are a feature of interaction between universal and practical knowledge. If we do not have universal knowledge, the faculty of judgment cannot work. Our decisions then would be made arbitrarily. Although knowledge about values, concepts, principles, or properties is abstract and general, it guides and regulates the exercise of judgment. To demonstrate the need for value or principles, Hare presents the following example: Let us suppose that a man has a special kind of clairvoyance such that he can know everything about the effects of all the alternative actions open to him. But let us suppose that he has so far formed for himself, or been taught, no principles of conduct. In deciding between alternative courses of action, such a man would know, fully and exactly, between what he was deciding. Suppose that we were ask such a man ‘Why did you choose this set of effects rather than that? He might say ‘I can’t give any reasons; I just felt like deciding that way; another time, faced with the same choice, I might decide differently’. On the other hand, he might say ‘It was this and this that made me decide; I was deliberately avoiding such and such effects, and seeking such and such’.‘6 The first answer is an example of an arbitrary decision much like mere taste or preference. The second answer is one where reasons have been provided. It shows, as Hare says, that “he has started to form principles for himself; for to choose effects because they are such and such is to begin to act on a principle that such and such effects Philosophical Association 60 (1986): 26-27. Cited by Benjamin, Splitting the Diflerence, 107. '5 Benjamin, Splitting the Difference, lll. '6 Hare, The Language ofMorals, 58-59. 166 are to be chosen.”17 The principle he begins to form will contain values even if he cannot exactly express or explain them. Suppose further that he must decide how to act in other situations. He will see if such effects are brought about by each alternative action, or may discover different kinds of effect that should be valued. As he gains experience in making decisions, he will discover and come to hold a variety of principles and values. Accumulated values and principles will tell him what to look for in his considerations of effects. The principles and values he develops will guide his decisions. We are not as fortunate as the clairvoyant man. We cannot know how things will go. We can sometimes predict roughly what will happen, but not exactly. Due to this limitation of our ability, we are in greater need of principles and values than is the clairvoyant man. The indeterminacy of principles does not mean, however, that they are useless. We must distinguish guiding from determining. Suppose we are sailing a boat on the open sea. But suppose that we are not terribly skilled at sailing and the only navigational aids we have are a not-so-detailed map and a compass. We may not be able to reach the exact place we want, but we would be lost if we failed to use even these instruments. We must judge the direction in which we are headed using our tools. As this analogy suggests, we need principles and values to guide the exercise of judgment, though they may not be clear or definite. '7 Hare, The Language of Morals, 59. 167 3. Justification of Judgments and a Web of Beliefs 3.1. Rules of Thumb for the Use of the Methods of WRE Which principles and values should guide and regulate the exercise of judgment required in the methods of WRE? We need to make judgments during the revision process. Earlier, three questions were raised: With which beliefs must the method begin, which beliefs should be selected, and how should they be revised? The principles or values I list here may be imprecise and sometimes conflicting. Relevant practical judgments will require practical knowledge that is obtained contextually and by practice. But when we earlier characterized the method of WRE and defended it from objections to coherentism, some important values and criteria were found and developed. The following values or principles will guide, but not firlly determine, the exercise of judgment needed in a revision process. First, coherence, comprehensiveness, and the number of incorrect beliefs are not only criteria for comparing competing belief systems, but will also be values for a revision process. The first two values are positive values while the last is a negative one. As discussed earlier, coherence and comprehensiveness may sometimes conflict. Second, we will pursue the maximization of coherence and comprehensiveness while minimizing revision, by revising peripheral beliefs rather than core beliefs in our belief system. By core beliefs, I mean those beliefs that have lots of inferential or explanatory connections with other beliefs. I do not mean foundational beliefs because the justification of core beliefs depends entirely upon coherence with other beliefs. Some 168 beliefs enjoy many more connections to other beliefs in our belief system than do others. Belief systems can be differentiated due to the reasonable comprehensive philosophical or religious doctrines they express. Beliefs constituting those doctrines are likely to be core beliefs. By peripheral beliefs, I mean those beliefs that have relatively fewer inferential or explanatory connections with other beliefs. The revision of our peripheral beliefs will minimize the number of further revisions. Revising a core belief will require further revisions of other relevant beliefs. The preference for peripheral beliefs over core beliefs as candidates in revision can be explained in terms of the integrity and identity of the belief system. This integrity and identity in turn may be relevant to one’s personal integrity and identity. Core beliefs will contribute to the establishment of our basic understanding of the world and of ourselves. Core moral beliefs may be closely related to personality. But the principle of maximizing coherence with minimizing revisions should not be taken to be more important than the three above main values. If one of our core beliefs is not correct or does not cohere with our other beliefs, it must be revised even though lots of further revisions are expected. Core beliefs are not immune to revision. The distinction between core and peripheral beliefs does not deny that all beliefs are equally susceptible to revision. The revision of core beliefs, however, requires stronger evidence of the need to revise than that of peripheral beliefs because, by definition, core beliefs enjoy greater coherence with other beliefs than peripheral beliefs. Coherence is a function of the strength and number of inferential or explanatory connections among beliefs. Third, the efficiency of a revision process may be one of the important considerations tied to the choice of provisionally fixed beliefs. By efficiency I have in 169 mind how long the revision process runs and how many beliefs are under review in the back-and-forth process. But different choices may not necessarily affect the overall coherence finally achieved. Different choices may not necessarily show differences in beliefs that are eventually revised because any selected provisional belief can also be checked against other beliefs and should itself be revised if needed. In fact, the choice of provisionally fixed beliefs will vary according to context. We cannot determine in advance and detail what kind of belief is best as a starting point. In the operation of the group method of WRE, an important practical question concerns what principles can guide the process of compromise. A compromise position (the product of compromise) will be devised by following the principles for the revision process discussed above. But the process of compromise should be regulated by other values and principles. By the process of compromise I mean a process in which the members of a group decide whether at a certain point it is the best time to make a compromise, what attitude is required from members, and from what perspective the content of compromise must be devised. We must know when we need to compromise. We need to compromise when we are confronted with rationally irreconcilable conflict, but must reach agreement — in other words, when agreement on a reasonable position is better than continued disagreement and irresolution. When these conditions are met, the following three values suggested by Benjamin will guide the process of compromise: The principle values framing or guiding the exercise of judgment in the circumstances of compromise are individual integrity, overall utility (including social integrity), and equal respect.18 These values have already been mentioned indirectly in our discussion of how the group 170 method of WRE can preserve personal integrity despite implicit or explicit compromise. First, compromise should not infringe personal integrity. To reach integrity-preserving compromise, the above conditions for compromise must be satisfied. In addition, as discussed in Chapter Five, members should understand moral pluralism and reasonable pluralism as relevant background beliefs and theories. They must also understand the notions of group belief and justification in the group method of WRE. Social integrity is a second value mentioned by Benjamin. This integrity guides the judgments employed in the process of compromise. In addition to the consideration of overall utility, the members should preserve the integrity of their group. To do so, a compromise position should be devised in light of a shared perspective, which is to be provided by the group belief system. Third, equal and mutual respect are values that should be preserved in the overall process of the group method of WRE as well as in the process of compromise. These values are also moral virtues of reasonable persons who participate in group reasoning. Furthermore, as Benjamin points out, “the likelihood of arriving at a maximally integrity- preserving, mutually satisfactory outcome will be improved by candor and hindered by coercion or deception.”19 This value was also stressed as one of the conditions in which the group method of WRE should operate. An agreement including compromise should be an informed, uncoerced, and reasonable agreement. Those values or principles that guide judgments needed for either a revision or compromise process cannot be as perfectly clear and definite as, for example, the concept of “the fastest.” Most concepts and principles used in practical moral reasoning are so '8 Benjamin, Splitting the Difference, 122. '9 Benjamin, Splitting the Difl‘erence, 137. 171 abstract and general that they will need to be interpreted in particular situations. But if they were definite and clear, we would not enjoy the flexibility, diversity, and openness to new situations that we do. We might lose the opportunity to exercise freedom. The room for the exercise of judgment is the very place from which our freedom sprouts. David Wiggins puts it this way: The unfinished or indeterminate character of our ideals and value structure is constitutive both of human freedom and, for finite creatures who face an indefinite or infinite range of contingencies with only finite powers of prediction and imagination (NE 1137b), of practical rationality itself.20 But just as our freedom does not mean that just anything we want is allowed, so too the exercise of judgment does not mean that any decision will be justified. We should in general respect other people’s differing judgments, though not all judgments should be respected. If not all judgments are acceptable, how then should they be evaluated? 3.2. A Web of Beliefs The values and principles that guide the exercise of judgment will, in part, provide reasons or criteria to justify or evaluate judgments. For example, our traffic regulations guide our decisions while driving, but they also justify them and can be used to evaluate other’s decisions as well. But what should we do when our practical judgments and the traffic regulations are in conflict? We need reasoning to evaluate practical judgments and the regulations. In the intersection example above, Paul’s judgment was that he should keep going. A police officer watching what Paul is doing 2° David Wiggins, “Deliberation and Practical Reason,” in Essays on Aristotle's ethics, edited by Amelie Rorty (University of California Press, 1980), 234. Cited by Benjamin, Splitting the Diflerence, 113-114. ‘NE’ in the citation indicates Aristotle’s Nicomachean Ethics. 172 said to him that he violated the current traffic regulation because he had not reached the intersection soon enough when the light changed. Suppose Paul tries to defend and justify his practical judgment. One of his main reasons would be his judgment about the closely following car and the slippery road condition. His judgment in the case might be justified. The judgment would in fact cohere with other general social principles beyond the scope of ordinary traffic regulations. The decision would be to avoid more harmful consequences — a collision and its resulting harm to someone or himself. If traffic regulations aim to make travel safe, Paul’s judgment would be more appropriate than the alternatives in this situation. This example illustrates that the evaluation of judgments cannot be restricted to the coherence between judgments and the traffic regulations or the mere observance of the regulations. Considerations of this kind have already been made in the context of focusing on the method of WRE as a method of moral reasoning aimed at the evaluation of moral judgments. Moral reasoning serves two purposes. One is to evaluate moral beliefs. When this is our intent, we can use the method of WRE to see if a particular moral judgment is justified. If it does not cohere with other beliefs, it is not justified and is not defendable. Another aim is to produce justified moral beliefs. When the method of WRE produces a justified particular moral judgment, it tells us what action should be taken, which is what I mean by practical moral reasoning. Individual practical moral judgments are formed in the interaction between universal knowledge about ethical theories or moral principles and particular knowledge of relevant empirical facts and descriptions of situations. However, practical moral judgments should be justified through the method of WRE that serves as a process of evaluating moral judgments. 173 Can the same point be applied to judgments made during the processes of revision or compromise? The principles or values presented for the revision process are usually discovered and developed in light of coherentism as a theory of justification. If these principles or values do not cohere with our theories of beliefs or justification in epistemology, theories in psychology, or theories in cognitive science, they cannot work as practical principles or values. Similarly, principles or values provided for the process of compromise should cohere with other relevant background beliefs or theories, such as with our understandings of a pluralistic society and a good society, theories about good discussions, and theories of human relationships developed in psychology. Principles or values are discovered and established with relevant background knowledge. Therefore, judgments made during the revision process should be justified by relevant practical reasoning in the light of relevant background beliefs. Relevant practical reasoning evaluates whether the judgment in question coheres with our other beliefs including relevant background beliefs or theories. If our belief system consists of a web of belief, or of webs of small groups of beliefs, the complexity revealed by the standard or group method of WRE should not be surprising. The standard method of WRE as practical moral reasoning should provide an answer about which act to undertake. To do this, a practical moral judgment regarding an action is made with consideration of the alternatives. This practical judgment will be an outcome of the relationship between relevant universal knowledge about moral principles or ethical theories and particular knowledge of relevant facts and circumstances. To obtain particular knowledge, we must know whether particulars have properties that relevant universal knowledge indicates. To do this, we must sometimes rely on judgment 174 because relevant concepts are abstract and general and also because situations we face may be unprecedented or unpredictable. Thus, judgments about particular situations must involve interaction between universal knowledge about relevant concepts and particular knowledge about relevant particulars. A practical judgment, as a candidate for a justified moral belief, then should be evaluated through the method of WRE. In this case, if this practical moral judgment coheres with other beliefs, it is a justified practical moral judgment. We will, then, get an answer about what to do. But if it does not easily cohere with other beliefs, we may need to begin a process of revision, though only if this practical moral judgment is a considered judgment. This situation is a state of disequilibrium. A revision process will require judgments about which beliefs are to be revised. These judgments should also be justified by relevant reasoning process. The exercise of judgment is inevitable in practical moral reasoning because we must act here and now with some uncertainty and the guidance of abstract and general moral values or principles. Because judgment is not an algorithmic decision procedure, neither is practical moral reasoning. As shown in the above general description of the relevant processes of practical moral reasoning, the processes are as complicated as they are due to the complexity of our webs of beliefs. 175 Chapter Seven Wide Reflective Equilibrium and the Issues of Euthanasia and Physician-Assisted Suicide Although medical technology now extends our life spans, it does so in some cases with unbearable suffering and pain, keeping people alive who might otherwise have died. Some of those who are terminally ill and unbearably suffering wish to hasten their own deaths. In some cases individuals who are adult, conscious, and mentally competent may hasten their deaths simply by refusing life-prolonging medical treatment. Refusing treatment for the purpose of hastening death is voluntary passive euthanasia (VPE). In other case, conscious, competent adult patients who are terminally ill and unbearably suffering may also wish to hasten their deaths but they cannot do so by refusing life- prolonging medical treatment, for they need no such treatment to remain alive. In this case they may request voluntary active euthanasia (VAE), a physician’s taking positive steps to hasten their deaths or physician-assisted suicide (PAS), a physician’s providing the means and instruction for their taking their own lives. But we face a variety of ethical questions regarding the moral difference between VPE and VAE and between PAS and VAE. There are different, but related questions about legalizing VAE and/or PAS. Dealing with the above questions in this chapter, I have two purposes. First, I intend to demonstrate how the standard method of wide reflective equilibrium (WRE) can be used, trying to arrive at an answer to the issue of the moral permissibility of VAE/PAS. I seek an answer that achieves equilibrium in my belief system. I focus mainly on the following two questions: (1) Are VAE or PAS ever morally permissible, and (2) If they are, in what circumstance are they morally permissible? 176 I begin by distinguishing between sorts of euthanasia to provide some relevant background information. I then demonstrate my use of the standard method of WRE to answer the above questions. I explain the loss of equilibrium and seek an answer through reviewing some main arguments for and against VAE and PAS. I try to show how using WRE I arrive at my answer. Second, I attempt to show how the group method of WRE can work by imaging a hypothetical ethics committee trying to reach an agreement as to whether VAE and/or PAS should be legalized. After showing that there will remain reasonable disagreement on the issue, I suggest how members of such a committee might reach a compromise position. 1. The Moral Permissibility of VAE and PAS 1.1. Background: Conceptual Distinctions Among Euthanasia Before I employ the standard method of WRE to consider the moral permissibility of VAE and PAS, an explanation of the relevant conceptual distinctions will provide needed background. As John D. Arras and Bonnie Steinbock explain, “Euthanasia (literally, a “beautiful death”) means an easy or painless death, but has come to stand for 9,1 deliberately bringing about such a death through action or inaction. Two important distinctions have been used for the moral assessment of euthanasia. The first is between active and passive euthanasia. Arras and Steinbock explain this difference as follows: ' John D. Arras and Bonnie Steinbock, eds., Ethical Issues in Modern Medicine, 5‘h ed. (Mayfield Publishing Company, 1999), 139. 177 The first [distinction] is between active euthanasia (deliberately bringing death about through some action, such as administering a lethal injection) and passive euthanasia (deliberately allowing death to occur through some form of inaction, such as refraining from performing corrective surgery). Often this distinction has been characterized as the difference “killing” and “letting die?2 The conceptual distinction between active and passive euthanasia, which is based on the distinction between killing and letting die, has been pervasive in biomedical ethics. But the expression ‘some form of inaction’ is not accurate if this means only withholding treatment since both withholding and withdrawing treatment are considered passive euthanasia. Ronald Munson uses instead the terms ‘commission’ and ‘omission.’ He says that active euthanasia is “direct killing and is an action of commission,” whereas passive euthanasia is “an action of omission.”3 The second distinction is made in relation to a patient’s wish. Arras and Steinbock explain as follows: A second distinction is made between voluntary euthanasia (actively requested by the patient) and nonvoluntary euthanasia, in which the patient (for example, a PVS patient) lacks the capacity to consent.4 We can add a third category of involuntary euthanasia to get a better conceptual understanding of euthanasia. Involuntary euthanasia is when a patient’s death is against his or her wish — a clear case of murder. Involuntary euthanasia is not, of course, morally permissible. So, we have conceptually six different types of euthanasia when the above two distinctions are put together: voluntary active, voluntary passive, non-voluntary active, non-voluntary passive, involuntary active, involuntary passive euthanasia. I deal with 2 Arras and Steinbock, eds., Ethical Issues in Modern Medicine, 140. 3 Ronald Munson, ed., Intervention and Reflection: Basic Issues in Medical Ethics, 6th ed. (Wadsworth, 2000), 204. 178 only voluntary active and passive euthanasia in this chapter. Non-voluntary euthanasia, whether active or passive, involves some additional moral and practical complexities because of the inclusion of surrogate decision-making. It should be noted that even when we focus on voluntary active and passive euthanasia, the real question is in fact whether VAE is morally permissible. The moral permissibility of VPE can be explained by the right to refuse treatment, which is already protected under constitutional law. But it does not mean that the Supreme Court says that there is a right to VPE. The right to refuse treatment is not specifically a right to die or to be dead. It is a right to be free of burdensome interference — treatment that, in the patient’s view, is such that its potential benefits do not outweigh its burdens. However, it is true that as a matter of fact some people refuse treatment expressly to bring about their own deaths. In practice some people may use the right to refuse treatment for the sole purpose of hastening their deaths. If a competent patient who knows that his/her refusal of treatment will result in death refuses life-sustaining treatment and if he/she intends to hasten death through the refusal of treatment, it can be considered VPE. Of course, merely knowing that the refirsal of treatment will result in death cannot constitute VPE. For example, a competent, adult Jehovah’s Witness who refuses a blood transfusion cannot automatically be thought to want to hasten his/her death even though he/she knows that such a refusal will likely bring about death. When a patient makes use of the right to refuse treatment in order to hasten his/her death, however, there is no way to force him/her to accept treatment. As Arras and Steinbock say, “A doctor who continues treating a competent patient who has refirsed treatment is theoretically liable for battery.”5 " Arras and Steinbock, eds., Ethical Issues in Modern Medicine, 140. 5 Arras and Steinbock, eds., Ethical Issues in Modern Medicine, 140. 179 Therefore, hastening death through withholding or withdrawing treatment cannot be legally prohibited even though there is no explicit constitutional right to die or to be dead. So the more important issue is whether VAE is morally permissible or not. PAS is another option that has been considered as a way to bring about death in the medical context. In this case, the physician prescribes a drug with instructions on how to use it for the patient who in turn takes it to end his/her life. That is the reason we called this physician-assisted suicide. When Ronald Munson distinguishes self-administered euthanasia from other-administered euthanasia, he seems to take PAS to be a self- administered active euthanasia. Oregon is the only state that currently allows PAS. PAS is allowed in Oregon only for competent patients who are terminally ill with six or fewer month to live. The patient must make “two oral requests (at least forty-eight hours apart) for drugs to use to terminate his/her life” and “wait at least fifteen days after the initial oral request, then make a written request to the physician.”6 The issues now are whether PAS should be legalized in other states and whether there is any moral difference between VAE and PAS. 1.2. The Standard Method of WRE and the Issue of the Moral Pennissibility of VAE and PAS In this section I will deal with whether, and in what circumstances, VAE or PAS are morally permissible. I will address the issue of their legalization when 1 demonstrate the group method of WRE in the next section. Moral permissibility and legalization need to be dealt with separately because the moral permissibility of an act (or a type of act) does not always imply that it should be legalized. 180 In discussing the moral permissibility of VAE or PAS, I will present how I would follow the standard method of WRE and my resulting answers. My aim is to arrive at answers that cohere with the other beliefs of my belief system. Prior to being confronted with an argument in favor of VAE, my beliefs about VAE and VPE were in equilibrium and relied on the distinction between killing and letting die. This distinction seemed to suggest that VAE is morally wrong while VPE is morally permissible. My equilibrium was disrupted, however, when I read James Rachels’s attempt to show that there is no moral difference between killing and letting die.7 He presents the following hypothetical cases: In the first, Smith stands to gain a large inheritance if anything should happen to his six-year-old cousin. One evening while the child is taking his bath, Smith sneaks into the bathroom and drowns the child, and then arranges things so that it will look like an accident. In the second, Jones also stands to gain if anything should happen to his six-year-old cousin. Like Smith, Jones sneaks in planning to drown the child in his bath. However, just as he enters the bathroom Jones sees the child slip and hit his head, and fall face down in the water. Jones is delighted; he stands by, ready to push the child’s head back under if it is necessary, but it is not necessary. With only a little thrashing about, the child drowns all by himself, “accidentally,” as Jones watches and does nothing.8 The only difference between Smith’s and Jones’s acts is that Smith killed the child whereas Jones let the child die. But they share the desired end of gaining a large inheritance and have the death of the child in view when they acted. My particular moral judgment about the above the two cases is that both are morally wrong. These cases illustrate that there is no moral difference between the bare acts of killing a person and 6 Munson, ed., Intervention and Reflection, 192. 7 This just explains my experience. Experiencing the disruption of equilibrium will vary according to individuals. One’s equilibrium might be upset by directly reading an article that argues for VAE or in other various ways. 8 James Rachels, “Active and Passive Euthanasia,” in Intervention and Reflection, 213-214. 181 letting a person die. The following points can be drawn from the cases. First, the cases show that if active and passive euthanasia are distinguished in terms of killing and letting die, there is no moral difference between active and passive euthanasia. I do not say that there are no legal differences. The causes of the child’s death difier in the physical causality of each case. But the issue I am interested in here is whether there are any moral differences because I am asking whether VAE is morally permissible. Second, if VAE is morally equivalent to VPE, and we formerly accepted VPE, but rejected VAE, we have two choices: either reject both VAE and VPE or accept both them. My judgment about Rachels’s cases upset my prior equilibrium maintained on the basis of the distinction between killing and letting die. To restore equilibrium to my belief system, I then decide to review arguments for and against VAE. Of course, I might reconsider my acceptance of VPE. As discussed earlier, however, VPE is not only supported by the right to refuse treatment, but also is hard to prohibit in practice. Thus, I undertake hypothetical interpersonal epistemic activities in reviewing arguments for and against VAE. Reasoning experience from these activities will give me an opportunity to seek and include new correct beliefs in my belief system. Arguments for each position can be summarized as follows. The main arguments for VAE/PAS consist of the beliefs that (1) self-detennination regarding when and how to die should be respected, (2) VAE/PAS are merciful acts aimed at relieving unbearable pain and suffering, (3) VAE/PAS are not incompatible with the goals of medicine, and (4) there is no moral difference between VPE and VAE because there is no moral difference between killing and allowing to die (If this argument can be extended to PAS, there is no 182 moral difference between VAE and PAS.) On the other hand, the main arguments against VAE/PAS consist of the beliefs that (l) VAE/PAS are beyond the limits of self-determination, (2) VAE/PAS are not compatible with the goals of medicine, (3) there is a slippery slope from VAE/PAS to involuntary euthanasia, and (4) life and death are not meant to be decided by us as human beings. I have considered VAE and PAS together because I have come to know that PAS is similar to VAE. The only difference lies in who takes the last action resulting in death, which may be legally important. But the physician plays a necessary role in both, whether it is direct in VAE or indirect in PAS. More importantly, the goal of requesting a physician’s help is the same in both. As Arras says, “The logic of justification for active euthanasia is identical to that of PAS.”9 But people usually seem to believe that PAS is more justifiable than VAE because it seems less susceptible to abuse or a slippery slope to involuntary active euthanasia. Thus, if VAE is shown to be morally permissible, then PAS can be taken to be morally permissible. As can be seen, I have found disagreements and conflicting views regarding the same topics each position concerns. I must review each belief and then choose or revise beliefs to arrive at a coherent answer. In this way, I am trying to regain equilibrium of my belief system. With respect to self-determination, the argument in favor of VAE is as follows. According to Arras, “There is the claim of autonomy, that all of us possess a right to self- determination in matters profoundly touching on such religious themes as life, death, and 9 John Arras, “Physician-Assisted Suicide: A Tragic View,” in Ethical Issues in Modern Medicine, 276. 183 the meaning of suffering.”10 This appeals to the principle of autonomy. An individual has the right to direct his/her life according to his/her own values or conceptions of the good life. Self-detennination is so important in our lives because, as Dan W. Brock says, “a central aspect of human dignity lies in people’s capacity to direct their lives.”ll Thus, the first reason to support VAE can be found in the idea of self-determination. The notion of self-determination is already part of my belief system. Not everyone, however, agrees with this appeal to self-determination. Callahan, for example, says that euthanasia is “no longer a matter only of self-determination, but of a mutual, social decision between two people, the one to be killed and the other to do the killing.”12 He continues: “How are we to make the moral move from my right of self- determination to some doctor’s right to kill me — from my right to his right?”13 The distinction between refusal and request may support his point. VPE takes place in the context of a patient’s refirsal whereas VAE occurs through his/her request. His point sounds right in that a patient’s right cannot theoretically imply a physician’s right to kill the patient. Self-determination may be just a negative right — a right to non-interference - rather than a positive right. I must then consider whether the distinction between negative and positive rights is meaningful in situations in which there are some physicians who are willing to help patients hasten their own death if requested to do so by the patient. In fact, proponents of VAE/PAS do not claim that all physicians, when they are requested to provide such help, should be compelled to do so. They see VAE and PAS as merciful acts, and may respond '0 Arras, “Physician-Assisted Suicide: A Tragic View,” 274. ” Dan W. Brock, “Voluntary Active Euthanasia,” in Intervention and Reflection, 216. ‘2 Daniel Callahan, “When Self-Determination Runs Amok,” in Intervention and Reflection, 226. '3 Callahan, “When Self-Determination Runs Amok,” 226. 184 to Callahan by saying that a physician’s right to help patients who desperately wish to die can be justified by the principle of beneficence. Thus, I come to wonder why we should prohibit a physician’s right to help such patients if the physician is willing to do so. Callahan’s point looked sound at first, but does not fit well this thought of mine. Callahan’s point nonetheless leads me to think about the issue in terms of the goals of medicine. Another important argument in favor of VAE/PAS is that they are merciful acts that release dying patients from unbearable pain and suffering.14 Even though successful control of pain and suffering can reduce the number of people who want to hasten their own death, there will remain some unfortunate patients whose pain and suffering cannot be alleviated. Theses patients see life to be no longer a benefit but rather a burden to them. In these cases, VAE/PAS may be more merciful than VPE. As Brock points out, “Once death has been accepted, it is often more humane to end life quickly and peacefillly, when that is what the patient wants.”15 Thus, I cannot see in the proponents’ arguments, anything incoherent with my other beliefs. As to the goals of medicine, opponents of VAE/PAS argue as follows. The involvement of physicians in VAE/PAS is incompatible with medicine’s goal of caring for patients and protecting their lives. But proponents of VAE argue that VAE/PAS are compatible with the goals of medicine. According to Brock, “The relief of pain and suffering has long been, quite properly, one of the central goals of medicine?“5 Therefore, as Arras and Steinbock say, “if the purpose of medicine is not simply to prevent death but to alleviate suffering, then perhaps euthanasia is not entirely foreign to " See, for example, Arras, “Physician-Assisted Suicide: A Tragic View,” 274. '5 Brock, “Voluntary Active Euthanasia,” 218. 185 ”'7 I agree with Brock and Arras because it is hard to deny good, ethical, medical practice. that alleviating suffering is one of the goals of medicine. But Callahan distinguishes suffering coming from illness and dying which are biological phenomena, from suffering arising from anguish or despair at the human condition. He says: It is not medicine’s place to lift from us the burden of that suffering which turns on the meaning we assign to the decay of the body and its eventual death. It is not medicine’s place to determine when lives are not worth living or when the burden of life is too great to be bome.... Medicine should try to relieve human suffering, but only that suffering which is brought on by illness and dying as biological phenomena, not that suffering which comes from anguish or despair at the human condition.18 But Callahan’s view does not cohere with my other beliefs. First, we seem to be unable to clearly distinguish suffering arising from illness from that of anguish or despair related to the human condition. Second, if medicine should take care only of suffering from illness and dying as biological phenomena, why does medicine take care of psychiatric problems? Psychiatric problems do not seem to come only from illness as a biological phenomenon. Rather, the problem may begin with the anguish or despair at the human condition. Third, a patient’s request of VAE is due to his/her suffering from dying as a biological phenomenon. Therefore, the view of the proponents’ of VAE looks like it better fits with my other beliefs. It coheres with the principle of autonomy, its relevant considerations, and the purposes of medicine that I hold as relevant background beliefs. But I have not considered particular moral judgments about cases of VAE. My judgment about VAE has been made analogically from Rachels’s example. Thus, I wonder what particular moral '6 Brock, “Voluntary Active Euthanasia,” 218. '7 Arras and Steinbock, eds., Ethical Issues in Modern Medicine, 139 186 judgments I would make if I faced an actual case. It is hard to find actual cases of VAE because it is legally prohibited. The case of Dr. Timothy Quill and his patient Diane may with some modifications help.” In the actual case, Dr. Quill had taken care of Diane for eight years and had come to deeply understand her suffering, wishes, and values. She was diagnosed with acute myelomonocytic leukemia. She very clearly expressed her wish not to receive chemotherapy and to live whatever time she had left outside the hospital. Despite Dr. Quill’s persistent attempts to persuade her otherwise, Diane clearly wanted to die and requested him to provide her with barbiturates. He finally decided to accommodate her decision and prescribed the drug. In fact, he knew that the last four patients with acute leukemia at his hospital had died very painful deaths. Diane’s family also understood her wishes. My moral judgment about this case is that it was morally permissible. Suppose that the situation were the same except that Diane could not swallow anything and hence could not take the drug herself. Suppose further that Dr. Quill finally decided to fulfill her request that he inject her with a lethal drug. The justification of Dr. Quill’s helping Diane in this case would be not different from that in the previous case. My particular moral judgment is that both cases are morally permissible. A possible objection to my judgment could be made on the basis that PAS is a more secure way than is VAE to confirm a patient’s autonomous decision. Opponents may think that the difference between a physician’s direct involvement and indirect involvement is morally important. To consider whether this is the case, I modify the case '8 Callahan, “When Self-Determination Runs Amok,” 229. '9 Timothy E. Quill, “Death and Dignity: A Case of Individualized Decision Making,” in Ethical Issues in 187 further. Suppose that there is a suicide machine that is so well-designed that it can identify a particular movement or voice command of a patient. Using a specific movement or voice command, the patient can command the machine to inject a lethal drug. Suppose Dr. Quill decides to fulfill Diane’s request that she wants to use this machine due to her disability to swallow lethal pills. He would set up the machine for Diane instead of prescribing lethal drugs because she cannot swallow. She would hasten her death by using this machine. It is clear that there is no moral difference between Diane’s using the suicide machine and her swallowing the drug, since both are cases of physician-assisted suicide. Now suppose further that Dr. Quill pushes, in the same circumstance as the above, the button on the machine when signaled by Diane instead of the machine acting on its own. Is there any moral difference between Dr. Quill’s simply providing the suicide machine and his pushing the button of the same machine? Everything seems to me be the same as the original case. The only differences are the methods used to terminate Diane’s life. But his pushing the button would be a case of VAE. I fail to see any good reasons for thinking that letting the machine act on its own is morally better than letting Dr. Quill push the button. The intended action and foreseen result are the same in both cases, as well as in the case of simply prescribing lethal drugs. Thus, my particular moral judgment regarding even the last modified case is that VAE for Diane is morally permissible. Rachel’s example stresses our consistency in our views about VPE and VAE, and my examples show that we must also maintain consistency in our views about PAS and VAE. If we think that VAE is morally wrong, we should say the same about VPE. In circumstance in which PAS is morally permissible, VAE will be also morally permissible. Modern Medicine, 250-253. 188 If we think that there are cases in which VPE is morally permissible, we must say that there are cases in which VAE is morally permissible and in such cases PAS is also permissible because there is no moral difference among VPE, VAE, and PAS. Therefore, I arrive at the answer that there are some cases in which VAE and PAS are morally permissible. As shown in my flow of reasoning, this claim better coheres with the other beliefs of my belief system than the alternative view. But I wonder if there are objections that I have not yet considered. A serious objection to VAE/PAS seems to be the slippery slope argument that is, according to Arras, based on “the close logical resemblance of concepts or justifications”20 If self-determination is crucial to the moral justification of VAE and PAS, he argues, the requirement of unbearable pain or suffering will be hard to justify.21 It might be easy to shift from terminally ill, through merely incurable, to just suffering without disease if we only focus on the notion of self-determination. If pain and suffering is crucial, he continues, the shift from current consent to best interests may be easy. If so, the shift may be made from voluntary euthanasia to non-voluntary or involuntary euthanasia. But I wonder why we should think of the two requirements separately. Rather, both requirements must be combined as shown in the earlier considerations of both self- detennination and merciful acts. Circumstances in which VAE and PAS could be morally permissible were ones in which these acts were supported by the principle of autonomy (the patient’s self-determination) and the principle of beneficence (the physician’s merciful act). The combined requirements rule out non-voluntary and involuntary 2° Arras, “Physician-Assisted Suicide: A Tragic View,” 276. 189 euthanasia. If we emphasize only one of them, equilibrium will be lost again. There are two controversial cases from my opponents’ perspective. One is the case of a patient who has never yet been competent such as an infant or a child. The other is the case of a patient who has once been competent, but now is not. If euthanasia should be considered in either case, it would be generally classified as a case of non-voluntary euthanasia. But the latter will be more controversial because patients might have written an advance directive requesting either VPE or VAE in certain circumstances. But the slippery slope argument makes me raise a question of whether the moral justification of non-voluntary euthanasia will be the same as that of VAE and PAS, more importantly, whether the justification of the latter sorts will compel us to justify non-voluntary euthanasia. If we need other reasons that replace one of the two reasons supporting VAE and PAS in order to justify non-voluntary euthanasia, this in itself tells us that it is not the case that the slope is so slippery that we cannot help swallowing the unpleasant outcome of admitting non-voluntary active euthanasia. Thus, the slippery slope argument looks implausible. I arrive at an answer to the question of what circumstances both VAE and PAS can be morally permissible. They are circumstances in which (1) an adult patient is dying. (2) his/her pain and suffering is unbearable, and (3) the patient is competent and requests VAE or PAS on the based of his/her uncoerced, autonomous decision. One may object to the involvement of a physician in VAE/PAS. In my view, however, a physician is the best person to deliver medical information to a patient and to make sure that his/her decision is informed and autonomous. Using the standard method of WRE, I have reached equilibrium among my beliefs 2' Arras, “Physician-Assisted Suicide: A Tragic View,” 276. 190 regarding the moral permissibility of VAE/PAS. Particular moral judgments about Rachels’s example, the actual case of Dr. Quill and Diane, and my modified cases cohere with moral principles such as the principle of autonomy (a patient’s self-determination), the principle of beneficence (a physician’s merciful acts), and the goals of medicine as background beliefs. I would not have reached such equilibrium had I not modified my moral belief about VAE based on the distinction between killing and letting die, and had I not accepted the new understanding of the purpose of medicine. Lastly, the method of WRE may not be the only method of reasoning by which to reach the same conclusion as mine. Principlists and casuists may reach the same conclusion. Principlists can appeal to the two moral principles. But the conclusion, I stress, cannot be justified simply by appeal to the two principles and also cannot be directly deduced from them. This is because the justification of a physician’s merciful act cannot be obtained without considering that the act requires a revised understanding of the function of medicine. In addition, we can observe conflicts between the prinCiple of beneficence and the professional ethical code. The document accepted by most of the doctors of the American Medical Association says, “The intentional termination of the life of one human being by another — mercy killing — is contrary to that for which the medical profession stands and is contrary to the policy of the American Medical Association.”22 A person following the principlist method of reasoning must show why he/she deduces the conclusion from the principles of autonomy and beneficence rather than from the professional ethical code. This shows that the person must rely on the exercise of j udgment about the choice of relevant moral principles. Casuists can arrive at the conclusion that there are some circumstances in which 191 VAE/PAS are morally permissible by appealing to moral judgments about the case of Dr. Quill and Diane or to the difference between this case and the case of Dr. Jack Kevorkian. The differences between these two cases may help us to understand what the features of morally justified cases are. But casuists cannot generalize the lesson learned from the differences without appealing to moral principles or values revealed in the case of Dr. Quill. Casuists must explain why they looked at the differences of the process to confirm a patient’s autonomous decision in the two cases rather than the similarity of a physician’s involvement in them. They could not arrive at the conclusion without the perspective that led them to focus the differences. 2. The Group Method of WRE and Legalization of VAE and PAS 2.1. Background: The Loss of Equilibrium I now move on the issue of the legalization of VAE and/or PAS. Before I demonstrate how the group method of WRE can be used, I should explain why this issue is raised in our society. The proponents of legalization think that there is no moral difference between VPE and VAE, and/or no moral difference between VAE and PAS.23 But the law permits doctors to withhold or withdraw life-sustaining treatment, and thus allows VPE. It prohibits, though, both VAE and PAS. United States District Judge Barbara F. Rothstein views this discrepancy in Washington’s law as untenable, saying “from a constitutional perspective, the court does 22 Rachels, “Active and Passive Euthanasia,” 212. 2’ Some opponents of the legalization of PAS may think that there are relevant differences between PAS 192 not believe that a distinction can be drawn between refusing life-sustaining medical treatment and physician-assisted suicide by an uncoerced, mentally competent, terminally ill adult”24 Under the current law, a patient who is unbearably suffering wishes to end his/her life, and is on life-sustaining treatment is more fortunate than a patient who is in the same situation but not dependent on such treatment. Under the Oregon law, patients not dependent on such treatment, but who are able to swallow lethal pills, are more fortunate than patients in the same situation who cannot do so because of some physical impairment. The current law favors the first patient over the others and even the Oregon state law favors the second patient rather than the third. The issue of j ustice is thus raised. The justice issue is not only a problem for proponents of the legalization of PASNAE, but also a challenge to opponents. So long as opponents to legalization do not object to VPE, they would acknowledge that the second and third kinds of patients in the above are not treated equally by current regulations. Thus, our society must confront the issue of justice as well as the legalization of VAE/PAS. The legalization of VAE/PAS may be a solution to the issue of justice, but there are conflicting views on legalization. But we must resolve the issue of justice. Suppose that an ethics committee is formed to make a recommendation on these issues. And suppose that all the members of the committee are reasonable and that they understand why these issues are addressed to the committee. By examining the views of proponents and opponents of the legalization of VAE/PAS, I will demonstrate the likely give-and— take discussions that would constitute the members’ group reasoning. Although I cite the and VAE and oppose legalization of VAE. 2’ Compassion in Dying v. Washington 850 f. Supp. 1454 W. D. Wash. 1994. Cited by Arras and Steinbock, 193 words of some medical ethicists, these can be understood as views expressed by committee members. Some members may want to morally distinguish VAE from PAS. I will not distinguish them because the differences, if any, do not affect the overall discussion. Let us assume that all the members of the committee reach overlapping consensus on the five moral mid-level principles of autonomy, justice, utility, beneficence, and nonmaleficence. Suppose that they know about how group beliefs are decided and how group beliefs are justified. Group beliefs are defined as beliefs that are accepted by all the members of the committee. Group moral beliefs are justified when all the members of the committee recognize that they are justified in all reasonable belief systems, though for different reasons, or when they mutually accept the same substantial reasons to support them. 2.2. Back-and-Forth Discussion Proponents of the legalization of VAE/PAS might begin the discussion by arguing for legalization on the basis of the moral permissibility of VAE/PAS. It may seem a natural move to try to legalize VAE/PAS because they are morally permissible in some circumstances from the proponents’ perspective. But some members may have a different view even though they argue for the moral permissibility of VAE/PAS. They might think that the moral permissibility of a certain type of acts does not always mean that it should be legalized. The following example may support their view. Ethical Issues in Modern Medicine, 141. 194 Suppose you find a person injured lying alone on the sidewalk as you are driving by. The injured person looks to be in bad shape and is bleeding badly. You know a hospital is located five miles away and judge that you must get the person to the hospital quickly. But on the way to the hospital you come to a red traffic light. The intersection is not busy, so you decide to look carefiilly and then drive through the intersection despite the red light to get the hospital as soon as possible. Even though you end up violating a current traffic law, your act can be morally justified. Saving a person’s life is more urgent than observing the traffic regulation. This does not lead us to think, however, that we must revise the traffic regulation. We would not add, for example, a clause specifying that “you can violate the current traffic law when you must save a person’s life.” This is because we cannot describe in an accurate way all of the particular circumstances in which one can be allowed to violate a traffic regulation. Having the law without any mentioned exceptions will do more to prevent unjustifiable exceptions than mentioning some exceptions. For example, utility will be maximized by having the law without mention of exceptions. Mention of exceptions may lead to a slippery slope which in turn will lead to many unjustifiable violations of the law. Our judgments regarding the urgency of the circumstances depend on the particular circumstances and their various characteristics. In the above situation, it might be better to call 911 than to bring the person to the hospital by yourself because an ambulance could better provide first-aid. Most of the members of the ethics committee are likely to agree on the belief that the moral permissibility of a certain kind of acts does not always imply its legalization. All the members may accept this belief as a group belief if they can recognize that the 195 belief is justified for all the members or they accept the above example as a substantial reason. Proponents may present another reason to support the legalization. Some psychological consideration supports their view. As Brock says, “the strong concern about losing control of their care expressed by many people who face serious illness likely to end in death suggests that they give substantial importance to the legalization of euthanasia as a means of maintaining this control.”25 The existence of a means of control seems important because, as Arras also says, people “fear a loss of control at the end of life.”26 Most people who request and are provided with the means for PAS in Oregon do not use them. They take comfort in the fact that if things became too bad — if pain and suffering became unbearable — they would have a way to end them. Sometimes it is said that just knowing that they can end their lives if necessary puts patients in a good frame of mind, which in turn lessens pain and suffering. In addition, as Brock says, “we care about how we die in part because we care about how others remember us, and we hope they will remember us as we were in ‘good times’ with them and not as we might be when disease has robbed us of our dignity as human beings.”27 But opponents may disagree due to different understandings of the existence of an option to use VAE/PAS. They may argue that if the option exits, there will be psychological pressure to request it. According to Brock, for example, David Velleman says that if patients have this option, they may feel that they must justify why they want to remain alive or at least a heavy psychological burden to this effect may be placed on 2’ Brock, “Voluntary Active Euthanasia,” 218. 2’ Arras, “Physician-Assisted Suicide: A Tragic View,” 274. 27 Brock, “Voluntary Active Euthanasia,” 218. 196 the patient.28 However, proponents may object to this view. First, we can make the option purely additional in the sense that health care providers or we should not recommend or suggest VAE/PAS until a patient requests one of them. Second, proponents may dismiss Velleman’s worry as not well grounded because the same thing could be said about the right to refuse life-sustaining treatment. But it would be obviously unreasonable to eliminate this right simply due to the possibility of psychological burdens. With respect to the possible psychological pressure, proponents and opponents would be unable to reach agreement. They have different judgments about such pressure because they cannot predict what will follow the legalization. But both positions might have a shared view about the importance of autonomous decisions because they are concerned about whether the psychological pressure in part prevents a patient from making an autonomous decision. An answer to this question, however, would depend on how serious and heavy such a pressure occurs in our health care system. Similarly, both positions would show different judgments about how our health care system will change after the legalization. Opponents may argue that if VAE/PAS is legalized, physicians will be less likely to treat or care for their dying patients. This would result in patients losing trust in their physicians. Proponents, on the other hand, may claim that legalization would improve the current palliative care because physicians would be highly motivated to try all reasonable alternatives short of VAE/PAS. The above discussion demonstrates that committee members would have conflicting judgments about the possible psychological effects and the overall impact on palliative care. The more serious conflict between proponents and opponents, however, 2” Brock, “Voluntary Active Euthanasia,” 219-220. 197 has to do with possible abuse. 2.3. The Possibility of Abuse Proponents who worry about the foreseen psychological pressure and the loss of trust in physicians may provide the following safeguard suggested by Brock. His procedural safeguards are designed to ensure that a patient’s decision of VAE or PAS is a “fiilly informed, voluntary, and competent, and so true exercise of self-determination.”29 These safeguards will help us identify morally permissible cases of VAE/PAS: (1) The patient should be provided with all relevant information about his or her medical condition, current prognosis, available alternative treatments, and the prognosis of each. (2) Procedures should ensure that the patient’s request for euthanasia is stable or enduring (a brief waiting period could be required) and fully voluntary (an advocate for the patient might be appointed to ensure this). (3) All reasonable alternatives must have been explored for improving the patient’s quality of life and relieving any pain or suffering. (4) A psychiatric evaluation should ensure that the patient’s request is not the result of a treatable psychological impairment such as depression.30 Opponents will ask whether those safeguards will work effectively. Because legalization must include practically feasible practices, the issue may depend upon whether we can succeed in developing a reliable system to regulate VAE and PAS even if all members were to agree on their moral permissibility. There will, however, be reasonable disagreement on their moral permissibility. Opponents will worry that we cannot set up such a system and will point to the likelihood of abuse. They may object, as Callahan points out, “unbearable” suffering is “a 2’ Brock, “Voluntary Active Euthanasia,” 221. 3° Brock, “Voluntary Active Euthanasia,” 221. 198 highly subjective matter admitting of no objective standards.”3| Even though safeguards may be in place requiring that a patient be terminally ill, it is hard to decide when an illness is actually terminal. In fact, measuring pain faces the same problem: Pain and suffering are subjective. Proponents may argue in response that the lack of an objective standard does not mean that we cannot know or understand that a patient’s pain and suffering is unbearable. Some of the committee members may argue that the real concern has to do with the practical difficulties of knowing whether a patient’s decision is autonomous, rather than about the objective measuring of pain and suffering. But opponents may worry about how we can know whether a patient’s decision is autonomous. As Callahan says: Suffering is as much a fimction of the values of individuals as it is of the physical causes of that suffering. Inevitably in that circumstance, the doctor will in effect be treating the patient’s values...The doctor would have to decide, on her own, whether the patient’s life was “no longer worth living.” But how could a doctor possibly know that or make such a judgment?. .. And if it is difficult to measure suffering, how much more difficult to determine the value of a patient’s statement that her life is not worth living?32 From Callahan’s perspective, members may seriously consider whether there are practical difficulties in a physician’s confirming a patient’s decisions. Thus, members might consider whether, in the case of Dr. Quill and Diane, Dr. Quill evaluated Diane’s decision carefully and deliberately. He took into consideration her values and judgment that her life was “no longer worth living.” Although there may be disagreement on moral judgments about the case, members will have a shared point of view on the difference between the case of Dr. Quill and that of Dr. Kevorkian. Unlike Dr. Quill, Dr. Kevorkian sometimes helped his patients terminate their lives without 3 ' Callahan, “When Self-Determination Runs Amok,” 228. 199 sufficient enough time to understand their values and judgments. So, the members cannot reach consensus, but they will agree on the importance of making sure a patient’s decision is autonomous. The practical question that remains is that of whether the current health care system can provide sufficient time for a good relationship to develop between a physician and patient. Some members may argue that under the current system only a few American can enjoy such a developed relationship. Even some of proponents of legalization may admit that the current health care system is not ready to provide good physician-patient relationships for all Americans because around 40 million Americans lack access to even decent primary care, let alone to good patient-physician relationships. Expert palliative care is expensive. Under the current health care system, it appears impossible that all reasonable alternatives to improve a patient’s quality of life and relieve pain or suffering can be explored because of the financial burden on both individuals and the government. Although patients might in fact want to prolong their lives, they may decide to hasten death because they believe that the money for treatment might be better used by their families. Some committee members will worry that cases that do not satisfy the safeguards might be accepted because the present health care system cannot provide all possible palliative care for the poor. Or, only the few who are able to buy good health insurance will be able to satisfy the safeguards. Some members might point to the differences between the current American health care system and those of other countries. The Netherlands, for example, legally ’2 Callahan, “When Self-Determination Runs Amok,” 227. 200 accepts some VAE and PAS.33 Germany does not legally prohibit assisting suicide, though physician involvement is legally prohibited.34 But the significant difference, as Arras and Steinbock point out, is that the US. has no national health insurance, and thus the consideration of cost can unduly influence health care decisions.35 From this consideration, some members may judge that the current health care system in the US. is likely to bring about abuse. But not all the members may agree. Some pr0ponents may still argue that worries about abuse are exaggerated. They may argue that Oregon’s experience provides some evidence against opponents’ speculation. As Brock points out, “opponents of voluntary euthanasia on slippery slope grounds have not provided the date or evidence necessary to turn their speculative concerns into well-grounded likelihoods?”6 Therefore, the members may not reach consensus regarding the legalization of VAE/PAS due to their differing judgments about the possibility of abuse under the current health care system. What decisions should they make? 3. Compromise 3.1. Reasonable Disagreement: Different Judgments Members of the committee will agree that they hold different judgments about the possibility of abuse, psychological pressure, and impact on palliative care. They will ’3 Margaret P. Battin, “Euthanasia: The Way We Do It, The Way They Do It,” in Ethical Issues in Modern Medicine, 282. 3’ Battin, “Euthanasia: The Way We Do It, The Way They Do It,” 284. ’5 Arras and Steinbock, eds., Ethical Issues in Modern Medicine, 142. 201 recognize that these are reasonable disagreements because of the factual uncertainty. The various judgments are not based on prejudice, bias, and bad reasoning, but they are made differently due to unpredictable situations and limited factual data. Oregon might provide more useful data concerning possibilities of abuse. In addition, the legalization will require a reconceptualization of the function of medicine. Although I did not discuss how this might play out in terms of group beliefs, there will likely be conflicting views as shown in my earlier discussion of the moral permissibility of VAE/PAS. The notion that medicine is not simply to prevent death but to alleviate pain and suffering can support VAE/PAS, but this notion may require, as Martin Benjamin points out, “a reconceptualization of the function of medicine that the profession has, to this point, been unwilling to undertake.”37 Therefore, it is not clear that this notion will be accepted as a group belief of the committee. In this respect, there must be conflicting views as to the moral permissibility of VAE/PAS. From my personal perspective, I argued for the moral permissibility. But it is highly predictable that all the members will not be able to reach agreement on this issue as shown in the above conflicting views on the purpose of medicine. If all the members accept that there is reasonable disagreement on the legalization of VAE/PAS, they cannot reach either overlapping consensus or complete consensus. Thus, the committee may explain to the public why they could not reach agreement about the issue of legalization by showing the existence of reasonable disagreement. Alternatively, the committee may decide to reconsider the issue when there is more data by which to evaluate the possibility of abuse. 3° Brock, “Voluntary Active Euthanasia,” 221. 37 Martin Benjamin, Philosophy and This Actual World: An Introduction to Practical Philosophical Inquiry 202 Although members cannot reach consensus on the issue of legalization, their back-and-forth discussion would have provided a valuable opportunity to discover what safeguards must be developed for legalization to be acceptable. In addition, some members might change their personal views. Disagreement will be found among the members who follow the same method of reasoning. However, the failure to reach consensus means that the committee has not finished its assigned task. Our equilibrium is still disrupted because of the justice issue as discussed in the three types of patients, who are unbearably suffering and wish to hasten their own death, but are treated unequally under the current law. If the committee simply stops discussion at the point where they cannot reach consensus on the issue of legalization, the result will simply favor one position over the other despite the fact that they all agree that both positions are reasonable. The committee must struggle with this justice issue recognizing the existence of reasonable disagreement. Thus, the committee will may look for a compromise. What compromise position can be devised? Is there an alternative position that could be jointly accepted while maintaining every member’s personal integrity? As mentioned in Chapter Five, members of the committee, with mutual respect, should look for an informed and uncoerced compromise position because they cannot come to an overlapping or complete consensus although they must present a unified voice. All the members of the committee must acknowledge that the current legal system prohibiting VAE and PAS (except for Oregon) favors one position over the other about the issue. (Roman & Littlefield Publishers, Inc., 2003), 143. 203 3.2. A Compromise Position: PRHN Some members may propose Patient Refusal of Hydration and Nutrition (PRHN) as a compromise position. According to Benjamin, “the possibility of a patient’s refusing food and water while receiving medical relief for pain and discomfort may be a bedside compromise — one worked out by doctors, nurses, and patients and their families -— that can be accepted by those on both sides of the debate without compromising their integrity.”38 He explains PRHN as follows: PRHN permits mentally competent, terminally ill patients for whom palliative care is either ineffective or unacceptable to hasten their death by refirsing hydration and nutrition with the assistance of physicians and nurses who relieve associated pain and discomfort.39 Although Benjamin restricts PRHN to “terminally ill patients,” the committee members can revise this requirement if all the members agree that there is no strong reason to deny PRHN to a patient who is not terminally ill, but whose pain and suffering are unbearable and for whom palliative care is either ineffective or unacceptable. As discussed before, some members may point out that it is hard to decide whether a patient is terminally ill. The decisions of whether the proposed revision will be accepted as a group belief must follow the definition of group justification. For a proposed belief to be a justified group belief, all the members reach an overlapping consensus or they mutually accept substantial reasons for that belief. Some members may object that PRHN is not humane and even cruel. But some empirical data can be provided. Following physicians James Bernat and R. Peter Mogielnicki and philosopher Bernard Gert, Benjamin says, “Death by starvation and 38 Benjamin, Philosophy and This Actual World, 142. 204 dehydration need not be accompanied by intolerable physical suffering.”40 Benjamin adds: “Depending on the patient’s overall condition, the process will take one to three weeks or longer if he or she continues to take some fluids.”41 During this time, care professionals should comfort the dying patient providing skillful treatment of symptoms, including pain, dyspnea, and dryness of mouth. This will be the same process as provided for a competent patient who refiises life-sustaining treatment. The merit of PRHN is that it can help health care providers and family members become more certain as to how resolute a patient’s resolve to hasten death. All the members may reach agreement that PRHN will pose less concern about whether a patient’s decision is autonomous than PAS/VAE. PRHN will be supported by the principle of autonomy that is already part of the committee’s group beliefs. But some members may wonder how PRHN is allowed within the current law. One possible issue may be whether the refusal of hydration and nutrition can be included in the refusal of medical treatment. Benjamin says, “the established ethical and legal right of competent informed patients to refuse life-prolonging medical interventions, including medically administered hydration and nutrition, may reasonably be interpreted as to ‘natural’ drinking and eating as well?"2 Members may consider the following scenario. If a patient receives hydration and nutrition through a feeding tube, refirsing hydration and nutrition may pose little controversy because such a tube is medically implanted. More controversial will be cases in which a patient does not need a feeding tube. But how can heath care providers force 39 Benjamin, Philosophy and This Actual World, 145. 4° Benjamin, Philosophy and This Actual World, 145. See James Bemat, Bernard Gert, and R. Peter Mogielnicki, “Patient Refusal of Hydration and Nutrition: An Alternative to Physician—Assisted Suicide and Voluntary Euthanasia,” Archives of Internal Medicine 153, no. 24 (1993): 2725-2726. 205 the patient to take hydration and nutrition if the patient refuses them? When the patient strongly refuses hydration and nutrition, a feeding tube may be the only way to force it upon him. But this action is already blocked because patients can refuse the implantation of feeding tubes. So, all the members will share the belief that PRHN can be allowed under the current law. Furthermore, PRHN will not bring about debates over the goals of medicine. Resolving the justice issue, therefore, PRHN achieves equilibrium in the group belief system that includes the principle of autonomy, beliefs about legitimacy of refusing a feeding tube, and background beliefs about the goals of medicine. In this respect, PRHN is a well-grounded position. What mutual concession and gain are provided by that position? First, as Benjamin says, proponents of the legalization of VAE/PAS “must accept a slower death, and perhaps more psychological suffering, than they would like,” while opponents “must agree to the provision of comfort care to individuals engaging in PRHN.”43 Second, PRHN provides mutual gain. For proponents, the provision of comfort care is better than the status quo, which favors patients who can refuse life-sustaining treatment. For opponents, as Benjamin points out, “the mechanism of death is provided and controlled by the patient alone.”44 The possibility of abuse will be slim because, in PRHN, a patient’s voluntary decision will be more surely known. This compromise position preserves each party’s integrity. The reason, as Benjamin says, is that PRHN permits both parties “to retain much of what is central to " Benjamin, Philosophy and This Actual World, 145. ’2 Benjamin, Philosophy and This Actual World, 145. ’3 Benjamin, Philosophy and This Actual World, 146. ’4 Benjamin, Philosophy and This Actual World, 146. 206 their respective positions.”45 In addition, as reasonable persons, all the members have acknowledged that both competing positions are reasonable and they must tolerate one another’s view. In such a situation, neither position can be accepted as the group position without violating the integrity of those on the other side. Thus, all of the member will accept the compromise position. 3.3. Consideration of Possible Objections to PRHN Members may consider the following possible objection to their new group position: Why is PRHN a better compromise position than terminal sedation? Terminal sedation, justified by appeal to the principle of double effect, is a procedure in which a physician sedates a patient permanently and then stops providing hydration and nutrition. Proponents of sedation argue that the intended end is not to bring about a patient’s death, but to relieve his/her pain and suffering. However, it is suspected that the real intention of terminal sedation is to hasten a patient’s death rather than to relieve pain and suffering. On what grounds is the subsequent step of not providing hydration and nutrition justified? If a physician has already scheduled what he/she takes to be the next step after terminally sedating his/her patient, the intention of sedation seems disingenuous. The genuine intention is not merely to relieve pain and suffering, but to end the patient’s life through the ceasation of hydration and nutrition. If the reasons justifying terminal sedation include respecting the self-determination of competent patients and doing a merciful act, they are the same as those supporting VAE and PAS. Opponents of the legalization of VAE/PAS will have the ‘5 Benjamin, Philosophy and This Actual World, 146 207 same concerns about the feasibility of safeguards for terminal sedation as for VAE/PAS. For them, there will be the same likelihood of abuse with terminal sedation. In fact, their concerns may be greater because proponents of terminal sedation consider sedation to be a way of relieving pain and suffering as if it were a kind of palliative care. Therefore, it is not clear that terminal sedation would improve current palliative care. Responding to the above objection, members may devise an alternative compromise position. They may combine PRHN with sedation. Let us call this the “PRHN-Sedation” option. The process would be that after a competent patient follows PRHN, sedation may be provided on the basis of the patient’s request only when that patient is competent, and his/her pain or suffering is unbearable. In this situation, sedation may genuinely serve to relieve unbearable pain and suffering. The refusal of hydration and nutrition would already have taken its course. So, this will be continued. But the remaining issue will be whether sedation can be considered as one of the ways to relieve pain and discomfort following PRHN. With this issue in mind, the committee may discuss whether PRHN or PRHN-Sedation should be its final compromise position. 208 BIBLIOGRAPHY Alston, William P. “Concepts of Epistemic Justification.” In Human Knowledge: Classical and Contemporary Approaches, 2"d edition, edited by Paul K. Moser & Arnold Vander Nat, 321-340. Oxford University Press, 1995. Aronovitch, Hilliard. “Reflective Equilibrium or Evolving Tradition?” Inquiry 39, no. 3-4 (1996): 399-419. Arras, John D. “Getting Down to Cases: The Revival of Casuistry in Bioethics.” The Journal of Medicine and Philosophy 16 (1991): 29-51. Arras, John D., and Bonnie Steinbock, eds. Ethical Issues in Modern Medicine, 5th edition. Mayfield Publishing Company, 1999. Arras, John D. “Physician-Assisted Suicide: A Tragic View.” In Ethical Issues in Modern Medicine, 5th edition, edited by John D. Arras & Bonnie Steinbock, 274-280. Mayfield Publishing Company, 1999. Audi, Robert. Epistemology. Routledge, 1998. . Moral Knowledge and Ethical Character. Oxford University Press, 1 997. . The Structure of Justification. Cambridge University Press, 1993. Battin, Margaret P. “Euthanasia: The Way We Do It, The Way They Do It.” In Ethical Issues in Modern Medicine, 5th edition, edited by John D. Arras & Bonnie Steinbock, 280-292. Mayfield Publishing Company, 1999. Beauchamp, Tom L., and James F. Childress. Principles ofBiomedical Ethics, 4th edition. Oxford University Press, 1994. Benjamin, Martin. “Philosophical Integrity and Policy Development in Bioethics.” The Journal of Medicine and Philosophy 15 (1990): 375-389. . Splitting the Dzflerence: Compromise and Integrity in Ethics and Politics. University Press of Kansas, 1990. . “The Value of Consensus.” In Society 3* Choices: Social and Ethical Decision Making in Biomedicine, edited by Ruth Ellen Bulger, Elizabeth Meyer Bobby, and Harvey V. Fineberg, 241-260. National Academy Press, 1995. . Philosophy and This Actual World: An Introduction to Practical 209 Philosophical Inquiry. Roman & Littlefield Publishers, Inc., 2003. Berlin, Isaiah. “On the Pursuit of the Ideal.” In The Crooked Iimber of Humanity, edited by Henry Hardy, 1-19. Princeton University Press, 1990. Bemat, James, Bernard Gert, and R. Peter Mogielnicki. “Patient Refusal of Hydration and Nutrition: An Alternative to Physician-Assisted Suicide and Voluntary Euthanasia.” Archives of Internal Medicine 153, no. 24 (1993): 2723-2728. Blachowicz, James. “Reciprocal Justification in Science and Moral Theory.” Synthese 110 (1997): 447-468. Blackburn, Thomas S. “On the Very Idea of a Formative Experience: DePaul’s Challenge to Coherence Theories in Ethics.” Philosophy and Phenomenological Research 49 (1988): 139-144. BonJ our, Laurence. The Structure of Empirical Knowledge. Harvard University Press, 1 985. Brandt, Richard B. “The science of Man and Wide Reflective Equilibrium.” Ethics 100, no. 2 (1990): 259-278. Brock, Dan W. “Voluntary Active Euthanasia.” In Intervention and Reflection: Basic Issues in Medical Ethics, 6"1 edition, edited by Ronald Munson, 215-222. Wadsworth, 2000. Callahan, Daniel. “When Self-Determination Runs Amok.” In Intervention and Reflection: Basic Issues in Medical Ethics, 6th edition, edited by Ronald Munson, 225-230. Wadsworth, 2000. Chang, Ruth, ed. Incommensurability, Incomparability, and Practical Reason. Harvard University Press, 1997. Cooper, W. E. “Taking Reflective Equilibrium Seriously.” Dialogue 20 (1981): 548-555. Curd, Martin, and J. A. Cover, eds. Philosophy of Science. WW. Norton & Company, 1998. Damasio, Antonio R. Descartes ’Error: Emotion, Reason, and the Human Brain. G. P. Putnam, 1994. Dancy, Jonathan. “A Defense of Coherentism.” In The Theory of Knowledge: Classic and Contemporary Readings, 2nd edition, edited by Louis P. Pojman, 236-244. Wadsworth, 1998, . “The Particularist’s Progress.” In Moral Particularism, edited by Brad 210 Hooker and Margaret Little, 130-156. Oxford University Press, 2000. Daniels, Norman. “Wide Reflective Equilibrium and Theory Acceptance in Ethics.” The Journal of Philosophy 76 (1979): 256-282. . Justice and Justification: Reflective equilibrium in Theory and Practice. Cambridge University Press, 1996. Davidson, Donald. Inquiries into Truth and Interpretation. Clarendon Press, 1984. DeGrazia, David. Taking Animals Seriously. Cambridge University Press, 1996. DePaul, Michael R. “Reflective Equilibrium and Foundationalism.” American Philosophical Quarterly 23, no.1 (1986): 59-69. . “Naivete and Corruption in Moral Inquiry.” Philosophy and Phenomenological Research 48 (1988): 619-635. . Balance and Refinement: Beyond Coherence Methods of Moral Inquiry. Routledge, 1993. . “Liberal Exclusions and Foundationalism.” Ethical Theory and Moral Practice 1, no 1 (1998): 103-120. Ebertz, Roger. “Is Reflective Equilibrium a Coherentist Model?” Canadian Journal of ‘ Philosophy 23, no 2 (1993): 193-214. Elgin, Catherine Z. Considered Judgment. Princeton University Press, 1996. Feldman, Richard. “Epistemic Obligations.” In The Theory of Knowledge: Classic and Contemporary Readings, lSt edition, edited by Louis P. Pojman, 544-555. Wadsworth, 1993. Fumerton, Richard. “Critique of coherentism.” In The Theory of Knowledge: Classic and Contemporary Readings, 2nd edition, edited by Louis P. Pojman, 245-250. Wadsworth, 1998. Goldman, Alvin 1. Knowledge in a Social World. Clarendon Press, 1999. Goodman, Nelson. Fact, Fiction, and Forecast. Harvard University Press, 1983. Haack, Susan. “A Foundherentist Theory of Empirical Justification.” In The Theory of Knowledge: Classic and Contemporary Readings, 2"d edition, edited by Louis P. Pojman, 283-293. Wadsworth, 1998. . “Reply to BonJour.” Synthesis 112, no 1 (1997): 25-35. 211 Hammond, Kenneth R. Judgments Under Stress. Oxford University Press, 2000. Hardwig, John. “Is There a Duty to Die.” In Ethical Issues in Modern Medicine, 5th edition, edited by John D. Arras and Bonnie Steinbock, 292-301. Mayfield Publishing Company, 1999. Hare, R. M. The Language of Morals. Oxford University Press, 1952. . “Rawls’ Theory of Justice-I.” Philosophical Quarterly 23 (1973): 144-155. Harman, Gilbert. Change in View. The MIT Press, 1986. Harman, Gilbert, and Judith Jarvis Thomson. Moral Relativism and Moral Objectivity. Blackwell, 1996. Haslett, D. W. “What Is Wrong With Reflective Equilibria?” Philosophical Quarterly 37 (1987): 305-311. Heath, Joseph. “Foundationalism and Practical Reason.” Mind 106, no. 423 (1997): 451- 473. Held, Virginia. Feminist Morality. The University of Chicago Press, 1993. Holmgren, Margaret. “Wide Reflective Equilibrium and Objective Moral Truth.” Metaphilosophyl 8 (1987): 108-124. . “The Wide and Narrow of Reflective Equilibrium.” Canadian Journal of Philosophy 19 (1989): 43-60. Hooker, Brad, and Margaret Little, eds. Moral Particularism. Oxford University Press, 2000. J onsen, Albert R., and Stephen Toulmin. The Abuse of Casuistry: A History of Moral Reasoning. University of California Press, 1988. Kelley, Leigh B. “Anti-intuitionism and Reflective Equilibria Revisited.” Pacific Philosophical Quarterly 69 (1988): 201-221. Kuhn, Thomas S. “Objectivity, Value Judgment, and Theory Choice.” In Philosophy of Science, edited by Martin Curd and J. A. Cover, 102-118. W.W. Norton & Company, 1998. Larmore, Charles. “Public Reason.” In The Cambridge Companion to Rawls, edited by Samuel Freeman, 368-393. Cambridge University Press, 2003. 212 Lazarus, Richard S., and Bernice N. Lazarus. Passion and Reason: Making Sense of Our Emotions. Oxford University Press, 1994. Marcus, Ruth Barcan. “Moral Dilemmas and Consistency.” The Journal of Philosophy 77 (1980): 121-136. McDowell, John. “Virtue and Reason.” The Monist 62, no. 3 (1979):331-350. Moreno, Jonathan D. Deciding Together: Bioethics and Moral consensus. Oxford University Press, 1995. Munson, Ronald, ed. Intervention and Reflection: Basic Issues in Medical Ethics, 6th edition. Wadsworth, 2000. Nielsen, Kai. “Reflective Equilibrium and the Transformation of Philosophy.” Metaphilosophy 20, no. 3-4 (1989): 235-246. . “Relativism and Wide Reflective Equilibrium.” The Monist 76, no. 3 (1993): 316-332. Olsson, Erik J. “Coherence and the Modularity of Mind.” Australasian Journal of Philosophy 75, no. 3 (1997): 404-411. O’Neill, Onora. “Practical Principles and Practical judgment.” Hosting Center Report 31, no. 4 (2001): 15-23. Quine, Willard Van Orrnan. From a Logical Point of View, 2"d edition. Harvard University Press, 1980. Rachels, James. The Elements of Moral Philosophy, 3rd edition. McGraw-Hill College, 1999. . “Active and Passive Euthanasia.” In Intervention and Reflection: Basic Issues in Medical Ethics, 6th edition, edited by Ronald Munson, 212-215. Wadsworth, 2000. Rawls, John. A Theory of Justice, revised edition, The Belknap Press of Harvard University Press, 1999. . Political Liberalism. Columbia University Press, 1996. . “The Independence of Moral Theory.” Proceedings and Addresses of the American Philosophical Association 48 (1974-75): 5-22. . Collected Papers. Edited by Samuel Freeman. Harvard University Press, 1999. 213 Raz, Joseph. “The Claims of Reflective Equilibrium.” Inquiry 25 (1982): 307-330. Schmitt, Frederick F. “The justification of Group Beliefs.” In Socializing Epistemology: The Social Dimensions of Knowledge, edited by Frederick F. Schmitt, 257-287. Rowman & Littlefield Publishers, 1994. Singer, Marcus George. “The Ideal of a Rational Morality.” Proceedings and Addresses of the American Philosophical Association 60 (1986): 15-38. Singer, Peter. “Sidgwick and Reflective Equilibrium.” The Monist 58 (1974): 490-517. Tomlinson, Tom. “Casuistry In Medical Ethics: Rehabilitated, or Repeat Offender?” Theoretical Medicine 15 (1994): 5-20. Veatch, Robert M. “The Dying Cancer Patient.” In Ethical Issues in Modern Medicine, 5th edition, edited by John D. Arras and Bonnie Steinbock, 51-55. Mayfield Publishing Company, 1999. . “The Impending Collapse of the Whole-Brain Definition of Death.” In Ethical Issues in Modern Medicine, 5th edition, edited by John D. Arras and Bonnie Steinbock, 152-160. Mayfield Publishing Company, 1999. Wiggins, David. “Deliberation and Practical Reason.” In Essays on Aristotle (9 ethics, edited by Amelie Rorty, 221-240. University of California Press, 1980. Wittgenstein, Ludwig. 0n Certainty, edited by G. E. M. Anscombe and G. H. von Wright, translated by Denis Paul and G. E. M. Anscombe. Harper & Row, 1969. “I feel, Therefore I am,” New York Times, 19 April 2003. 214