Research question and annotated bibliography

article3.pdf

Home >English homework help >Article writing homework help >Research question and annotated bibliography

CO-OPERATION DESPITE DISAGREEMENT: FROM POLITICS TO HEALTHCARE*

NOAM J. ZOHAR

ABSTRACT

Political interaction among citizens who hold opposing moral views com- monly requires reaching beyond toleration, toward actual co-operation with policies one opposes. On the more personal level, however, regarding (e.g.) interactions between healthcare providers and patients, several authors emphasise the importance of preserving integrity. But those who oppose any ‘complicity in evil’ often wrongly conflate instances in which the other’s position is (and should be) totally rejected with instances of legitimate, although deep, disagreement. Starting with a striking example from the context of a particular tradition, I argue generally that in the latter sort of disagreements, talk of ‘complicity’ should be largely replaced with a more co-operative moral stance, grounded in a pluralistic frame- work. Co-operation Despite Disagreement (CDD) should be sought either for institutional reasons – akin to the political – or for relational reasons. CDD involves sharing another’s perspective and sometimes calls for adopt- ing another’s moral judgements in preference to one’s own. I seek to iden- tify some of the conditions and circumstances that would justify such a shift, particularly in scenarios involving assistance, such as physician- assisted suicide (PAS) or the role of an anaesthesiologist in abortion. This discussion is meant to provide examples of the kind of second-order reasons

Bioethics ISSN 0269-9702 Volume 17 Number 2 2003

* My work on this essay was facilitated by my fellowship, during 1999–2000, at Harvard University’s Center for Ethics and the Professions; a draft was pre- sented at the Faculty Fellows seminar there, chaired by Dennis Thompson. Earlier versions of the argument presented here were presented also at Princeton University (Center for Human Values), at the Philosophy Department colloquia at Tel Aviv University and Bar Ilan University, and at the fifth World Congress of the International Association of Bioethics. For useful critiques and many helpful suggestions, I am grateful to many of the participants in these various forums.

appropriate for determining the terms for CDD – in distinction from first- order considerations (e.g., the much-contested ‘active/passive’ distinction) which are likely to be the subject of the initial disagreement and hence cannot serve to resolve it.

INTRODUCTION: ‘COMPLICITY IN EVILDOING’

Citizens who hold opposing moral views are often exhorted to embrace toleration as a mode of peaceful co-existence. Yet passive toleration is not enough: political interaction commonly calls for actual co-operation with policies one opposes. On the more personal level, however, regarding (for example) interactions between healthcare providers and patients, several authors em- phasise the importance of preserving integrity and strict avoid- ance of ‘co-operation with evil.’ In this paper I will argue that in this, they wrongly conflate instances in which the other’s position is (and should be) totally rejected with instances of legitimate, although deep, disagreement. Starting with a striking example from the context of a particular tradition, I shall argue generally that in the latter sort of disagreements, talk of ‘complicity’ should be largely replaced with a more co-operative moral stance.

In an essay published in Asya, a Journal of Jewish Medical Ethics, Rabbi A. Sofer-Abraham published an exchange of letters between himself and a certain New York anaesthesiologist.1 This anaesthesiologist was called upon to serve in the performance of abortions, and wanted to know the extent of his obligation to resist. Should he request to be released from this duty? Must he be prepared to take a cut in pay for this resistance, or even risk losing his job? The response goes into a detailed halakhic2 dis- course about the duty to avoid complicity in evildoing, or ‘aiding transgressors.’ Throughout, it is taken for granted that abortion is a heinous deed; at issue are only:

(1) Determining the extent of the anaesthesiologist’s moral liability for the sin he is facilitating, though not directly committing; and

(2) Deciding the implications of this liability – what he must do, and what price he must be prepared to pay, in order to avoid culpability.

122 NOAM J. ZOHAR

1 Abraham Sofer-Abraham. Anesthesia for the Sake of Performing an Abortion. Asya 1989; 47/48: 40–50. (Hebrew)

2 ‘Halakhah’ denotes the Jewish religious tradition of normative discourse.

What is entirely missing is an assessment of whether, and to what degree, abortion is indeed an evil – first, in abstraction, and second, in whatever particular circumstances pertain in each instance. This is especially striking since there is in fact a wide disparity of positions regarding abortion within the halakhic tradition. Although abortion is generally frowned upon, and significant sources suggest that it is a serious offence, few indeed would equate it to murder. For example, the late Lord Rabbi Immanuel Jakobovits – in his classical Jewish Medical Ethics – writes: ‘While, according to the consensus of rabbinic opinion [the foetus’s] life is not protected by any definite legal provisions, the artificial termination of pregnancy is strongly condemned on moral grounds, unless it can be justified for medical or, possibly, other grave reasons.’3 Moreover, even if abortion is assumed to be generally wrong, it is naive (at best) to assume that the initiative for having the abortion lacks any countervailing moral weight. Yet the costs to the woman – and possibly to others as well – of not having the abortion are not mentioned in Sofer-Abraham’s dis- cussion; as long as they are not accounted for, how can it be taken for granted that the deed is wrong, all things considered?

These questions have considerable force, I believe, even when examined wholly from the perspective of the (anti-abortion) physician asked to participate in providing abortion services. Yet why should that be the only perspective employed? After all, there are other people involved, first and foremost the woman seeking the abortion. Most likely, she believes that – in her situation, here and now – the abortion is morally permissible (or even requisite!). Yet Sofer-Abraham’s analysis treats the physician’s predicament exactly as though he were being requested to par- ticipate, say, in stealing patients’ kidneys (while they are under sedation for some legitimate operation) for sale on the black market. The failure to distinguish between that sort of request and the one at hand depends upon totally ignoring the moral per- spectives of the other agents involved. These include not only the woman seeking abortion, but also the physician who is prepared to perform the operation. It seems wrong to simply ignore these moral perspectives – from which the abortion is, we may presume, fully legitimate.

CO-OPERATION DESPITE DISAGREEMENT 123

3 Immanuel Jakobovits. 1975. Jewish Medical Ethics (second, enlarged edition). New York. Bloch: 190. See generally, 182–191. See more extensively: David M. Feldman. 1998. Birth control in Jewish law: marital relations, contraception, and abor- tion as set forth in the classic texts of Jewish law (second edition). Northvale. Jason Aronson Publishers.

A similar exclusive focus on one agent’s perspective charac- terises many discussions of this issue and of similar ones in con- temporary Bioethics, especially – though by no means exclusively – discussions written from a religious viewpoint. Joseph Boyle, for example, speaks of a physician co-operating with a ‘patient’s sui- cidal will’, which the physician recognises (or ought to recognise) as ‘objectively bad.’ A recognition of the patient’s good faith and sincerity can yield, at most, ‘rationalising confusion.’4 Against this, I will explore here a pluralist approach that calls for communica- tion and for a sharing of perspectives. Where two or more people face a decision together, they ought to seek a joint decision, and for this it is not enough that each agent address the issue at hand from his or her individual perspective alone. Rather, this should be expanded to include the differing moral views of the other parties involved. Even if this does not ultimately produce a consensus, each individual’s decision will be informed by an enhanced, dual (or multiple) perspective.

It is worth emphasising that this pluralist approach is quite dif- ferent from the commonplace notion that, in weighing what is right for me to do, I should consider the interests of all impacted parties. This I should, of course, do; but even if I employ perfect empathy, and appreciate the interests of others as they themselves see them, I might still form my moral assessment from my per- spective alone. Indeed, the Kantian tradition of moral individu- alism – and to many, the very notion of moral integrity – seems to require that I do just that, excluding from my reasoning the other parties’ differing moral views. Given such an exclusivist stance, any suggestion that I co-operate with others (even partly) on their terms must indeed come under the unfavourable heading of ‘complicity.’ Instead, I will argue here for a pluralistic, inter- personal mode of moral thinking.

But what does such pluralistic thinking amount to? What does it mean to say that, in co-operation between individuals with

124 NOAM J. ZOHAR

4 Joseph Boyle. 1999. Collaboration and Integrity: how to think clearly about moral problems of cooperation. In Issues for a Catholic Bioethic. L. Gormally, ed. London. The Linacre Centre: 195 (187–199). See also: H. Bouma, D. Diekema, E. Langerak, T. Rottman & A. Varhey. 1989. Christian Faith, Health, and Medical Practice. Grand Rapids, MI. Eerdmans. These authors work out a scheme for ‘Respect, Tolerance, and Co-operation’ (pp. 95–101), yet in fact leave little moral space for co-operation despite disagreement. Sofer-Abraham himself, in consul- tation with some prominent rabbis, went on to extend his discussion to prenatal testing, to conducting diagnostic procedures that might be used to determine brain-death (which in his view is not truly death), and to abiding by a DNR order where the doctor believes the patient should be rescued: A. Sofer-Abraham. Do not place a stumbling-block before the blind. Asya 1990; 49/50: 62–67.

deeply differing moral views, each should be prepared to include the other’s perspective in his or her own judgement and practice? My main purpose here will be to sketch an answer to this ques- tion, through examining a set of issues in healthcare ethics. First, however, it is necessary to say something about the scope and the basis of co-operation between those who disagree.

I CO-OPERATION AND DISAGREEMENT: FROM THE POLITICAL TO THE PERSONAL

How should people co-operate when they have a deep moral dis- agreement? The question, thus formulated, assumes that they will co-operate. One response might be to reject this assumption, asking instead: why should people with deep disagreements try to co-operate? Of course, they need not disagree about everything, and where they agree they will have good reasons to co-operate despite disagreeing on other matters. But on the very matter on which they deeply disagree, is it not best that each should go his or her own way? Indeed this seems to be the classical and vener- ated liberal approach, encapsulated in the motto ‘live and let live.’ Is there, nevertheless, a reason to assume that co-operation ought to be sought – reinstating the question of how it should be managed?

A call for co-operation-despite-disagreement (CDD) is prob- lematic, for it seems that one or more of the parties will have to compromise his or her values or even integrity. Indeed, ‘failure to act in accordance with one’s values’ – even (perhaps: espe- cially) if this is due to interaction with others – defines lack of integrity in one of its common senses.5 We might begin, then, with a presumption against such co-operation. Hence any discussion of how to manage CDD will be significantly shaped by the nature of the reasons for countering this presumption. The nature of such reasons will vary significantly depending on who the ‘people’ in our original formulation are taken to be, e.g., two strangers whose paths cross briefly, members of a family, or the citizens of a country.

CO-OPERATION DESPITE DISAGREEMENT 125

5 ‘Integrity’ is a protean concept; for an illuminating analysis, see: Damian Cox, Marguerite Lacaze & Michael P. Levine. Should We Strive for Integrity? The Journal of Value Inquiry 1999; 33: 519–530. The authors consider integrity in this sense (‘steadfast devotion to principles’, p. 520), but go on to examine several other meanings as well. There is, for instance, a very different sense of ‘lack of integrity’, meaning dishonesty; that is not an issue in the context of my discus- sion here.

If we are thinking of citizens, defining the terms for CDD is – in one view – a central concern of political theory. I add the caveat ‘in one view’ because there is a view of politics that seeks to avoid the necessity of CDD, namely, the liberal theory of state neutral- ity. According to that theory, state power – the instrument of enforced political co-operation – should decidedly not be employed against the deeply held values of individual citizens. This can be achieved either by keeping the state completely out of certain realms (e.g., the separation of state and religion), or by ensuring that when the state does act in areas where disagreement prevails, it acts as a neutral facilitator, enabling each individual or group to follow their own conscience (e.g., protecting or even enabling free speech).

Some critics of the state-neutrality theory claim that it can never work, that is, the state can never really be neutral, since whatever it does (or refrains from doing) will always favour the values of some citizens more than others. I am not sure about this claim; examining it is outside the scope of the present discussion. In any case, I am convinced by the more moderate claim that – even if the state can sometimes act neutrally – there is also a significant realm where a choice must be made regarding state action and no option is neutral among citizens’ values.6 In decid- ing how their state should act in this realm, democratic citizens are perforce seeking CDD.

In this sense, CDD is fundamental to the nature of politics. Insofar as we are not prepared to opt for anarchy (doing away with politics altogether), we are thereby committed to CDD. If we posit a democratic political framework, it implies important fea- tures of the appropriate process for achieving co-operation, and some of the terms of reasonable agreement. It should also help specify the limits of reasonable disagreement, for even in a uni- verse of discourse that is morally pluralistic, some positions are deemed wholly illegitimate and call for rejection and utter opposition rather than co-operation.7 Although defining what

126 NOAM J. ZOHAR

6 In the context of abortion, the US legal doctrine (based on ‘privacy’) claims neutrality – leaving the decision to individual citizens. The validity of this claim is debatable, but even if it is granted, the federal government cannot avoid taking a stand in the realm of policy, either for or against abortion. This is most clearly evident with regard to policies of public funding; one particular example will be discussed below.

7 Some of the minimal requirements for legitimacy are defined, for example, by Gutmann and Thompson within the framework of ‘deliberative democracy’ in terms of ‘reciprocity’: Amy Gutmann & Dennis Thompson. 1996. Democracy and Disagreement. Cambridge, MA. Harvard University Press.

is out-of-bounds is notoriously problematic, in the discussion that follows I postulate that agents must be (and in fact are) able to distinguish between legitimate positions with which they strongly disagree, and those that are beyond the pale. Absent a capacity to draw this distinction, CDD is ruled out and with it any moral vision of a polity.

It is essential to recognise that the severity one ascribes to the policies proposed or carried out by one’s opponents cannot suffice, in itself, for branding them ‘illegitimate.’ If the realm of political co-operation is characterised by legitimate disagree- ments, these emphatically extend to matters of the utmost sever- ity, specifically including matters of life and death. Political decisions by their very nature routinely impact the life and death of numerous persons. This is true first of all in the context of dis- tributive decisions, whether these address matters of life and death explicitly (e.g., expenditure for and delivery of healthcare services), or implicitly (e.g., which roads to improve, how many police officers to hire). But there are also several state actions and policies that involve direct killing, not only in war but also in peace. For even where there is no death penalty, the state holds the ultimate power to kill those who defy its authority (i.e., felons resisting arrest or escaping from custody).

Given the range of actual disagreements, there will be many citizens who disagree with the state’s policy in some of these contexts, yet do not withdraw from all political participation (e.g., they continue to pay taxes). In doing so, they support actions of the state’s agents through which innocents are left to die or are directly killed. There appear to be three main approaches to justifying such support, which I shall very briefly sketch without trying to argue properly for any of them. In this, my purpose is merely to illustrate possible grounds for accepting as legitimate, certain (though not just any) views and actions of one’s fellow citizens, even such as involve – in one’s own view – killing the innocent.8

First, (echoing Hobbes’s argument), support for even severely immoral acts may be deemed the lesser evil, insofar as the alter- native is anarchy with its host of evils. Second, there may be epis- temic uncertainty. I may be convinced that a particular energy policy will cause numerous unnecessary deaths from pollution or cancer; those who implement it are actually, in my view, slayers of

CO-OPERATION DESPITE DISAGREEMENT 127

8 It may be that there is some better justification or way of tracing the limits of legitimate disagreement. As argued above, my main point is that the notion of a moral polity assumes that some such justification is possible.

the innocent. But, listening to their good-faith reasons for viewing their deeds otherwise, I should be less certain in my judgement of their actions than I am regarding, say, a war of blatant aggres- sion. Finally, it may be posited that valid moral knowledge does not reside in the isolated individual, but in jointly constructed practices and understandings.9 Hence, what I know to be right or wrong cannot be simply the basis for my interaction with my fellows, but rather also the result of such interaction and its concomitant compromises.

When citizens are periodically urged to endorse compromises in the political context, it makes little sense for them to turn around each time and say, ‘But that departs from my moral posi- tion!’ – or even, ‘But in my view that involves (wrongful) killing!’ Pointing to the disagreement itself, even if it involves killing, cannot rule out co-operation – that is the point of the exercise. In his instructive discussion on compromise and integrity, Martin Benjamin aptly quotes the common aphorism that ‘Politics is the “art of compromise.” ’10 Every political agent will necessarily deviate to some significant degree from his or her moral ideals; criticising him for this is in general misguided,11 reflecting a failure to appreciate that compromise is, as it were, ‘analytic’ to the very concept of politics. Defining the proper conditions or procedures for the compromises between citizens, or the terms of their co-operation, is a central concern of contemporary political philosophy.

Hence, it is somewhat bewildering that, with regard to certain issues – most notably abortion – it is often asserted as a matter of course that citizens should not be expected to co-operate (even indirectly), through state-endorsed policies, with actions they strongly oppose. Such a claim was posited, for example, as a central assumption in the deliberations of the HFTTR Panel on US federal funding for the use of foetal tissue (produced through abortions and otherwise discarded foetuses) in research

128 NOAM J. ZOHAR

9 Margaret Walker. 1998. Moral Understandings. New York and London. Routledge.

10 Martin Benjamin. 1990. Splitting the Difference: Compromise and integrity in ethics and politics. Lawrence, KA. University Press of Kansas: 1.

11 In specific cases, such criticism may be warranted, and there might be a call for conscientious objection or civil disobedience. But these will be the excep- tion; the rule must allow for discrepancies between each citizen’s ideals and the policies with which she will go along with. For an illuminating analysis, see: Arthur I. Applbaum. 1999. Ethics for Adversaries: The Morality of Roles in Public and Professional Life. Princeton. Princeton University Press: Chapter 9 ‘Democratic Legitimacy and Official Discretion’ (pp. 207–239), particularly 226 ff.

and therapy.12 The minority, representing a staunchly anti- abortion position, held that by promoting the use of foetal tissue, the government (and thereby the anti-abortion citizens) would become complicit in the evil of the abortions themselves. The majority basically granted that such enforced complicity is to be avoided, but maintained that it is possible to install sufficient pro- visions for a complete separation between the decision to abort and the subsequent use of the foetal tissue. This majority included some anti-abortion representatives, who relied crucially on these provisions. However, it included others who were themselves pro- choice, yet apparently recognised the validity of the concern over complicity, and hence joined in requiring the same provisions. A decade later, a similar approach informs much of the debate over stem cell research.

One may question, however, whether such a stringent effort to avert citizens’ complicity is indeed called for. After all, these abor- tions are legal. And in view of the prolonged and complex debate – moral as well as legal – concerning abortion, it seems clear that neither side has reason to regard its opponents’ position as illegitimate, beyond the pale of normal CDD among citizens. As argued above, the fact that some individuals believe that abortion involves killing the innocent is not in itself grounds for them to brand its performance by those who believe otherwise, as totally illegitimate. Interestingly, although the commission recom- mended a right for conscientious objection on the part of indi- vidual researchers, it did not extend such a right to citizens as taxpayers: political co-operation routinely pools the tax dollars of all citizens, who of necessity thus participate in whatever policies have been jointly adopted.13

Still, while the necessity of co-operation may hold true for all political life, it is also true that not all life is political. Surely there are significant portions (some of us moderns might say, pace Aristotle: the most significant portions) of human endeavour that are predominantly personal, not political, and in which the pre- sumption of mandatory co-operation does not hold. By this I do not intend an atomistic account of human existence in which the primary subjects are self-sufficient individuals. Rather, I wish to

CO-OPERATION DESPITE DISAGREEMENT 129

12 James F. Childress. Ethics, Public Policy, and Human Fetal Tissue Research. Kennedy Institute of Ethics Journal 1991; 1: 93–121.

13 This is recognised by Gutmann and Thompson, although they seek to minimise such participation as far as possible: op. cit. note 7, pp. 89–90. Boyle cites the moral duty to pay taxes as an instance of ‘constraining circumstances’ that may generate ‘mixed voluntariness’ and possibly allow co-operation with evildoing (op. cit. note 4, p. 190); I will say more on this approach below.

emphasise that universal co-operation in the non-political realm is not mandatory. However crucial and pervasive co-operation may be for all persons, they can often choose with whom and on what terms to co-operate. Indeed, the freedom of such choices is a highly cherished part of personal autonomy. When the moral cost of co-operation becomes too high, they can consider dis- association in a sense that is basically unavailable in the political sphere. Absent the mandatory force of political co-existence, CDD appears to be an avoidable evil.

However, is it true that, once we move down from the level of universal co-operation between the citizens of a state, CDD is avoidable, and – no less significant – ought we, in general, to strive to avoid it? Classical liberal thought seems to answer with a resounding ‘Yes’: even if the political framework sometimes forces us to go beyond negative toleration, people’s personal choices must be premised on a complete freedom to disengage. But I believe this is wrong: even on the non-political level, we often have compelling reasons to seek co-operation despite disagreement. Continuing, for the time being, with the example of abortion, I will explore two kinds of reasons for co-operation: institutional and relational.

II INSTITUTIONAL AND RELATIONAL REASONS FOR CO-OPERATION

The contested issue of abortion illustrates the difference between the political and non-political contexts for considering CDD. At the level of the polity as a whole, there is a necessity of co-ordi- nating between citizens with sharply conflicting positions. This involves determining both the law (or, as in the US, the limits of law) and various aspects of public policy. Let us suppose now that the law permits abortion in (at least) certain cases, and shift our attention to the context of individual interactions. P, believing that abortion in her circumstances is justified, is unable to perform it herself and thus seeks help from healthcare workers, some of whom believe that what she is seeking is wrong. Does an agent A, one of these workers, have reason to co-operate with her despite their deep disagreement?14

In the liberal-democratic tradition, the main resource for dealing with difference has been the principle of toleration. Let

130 NOAM J. ZOHAR

14 This is generically similar to the case from which we started above, except that there it was not the gynaecologist but only the anaesthesiologist who had qualms about abortion. I will come back to that below.

us assume that A adheres sincerely to an ethic of toleration, and shows full respect for the woman’s position.15 He does not try to manipulate her away from abortion, nor in any way to constrain her pursuit of her chosen goal. This same ethic of toleration does not appear, however, to furnish him with any reason for co- operation.16 He might say to her: ‘Look, although I think you are very wrong, I will not interfere in your actions; but I myself shall certainly have nothing to do with them. Let us each go our own way!’ Is there anything more to say?

Benjamin defines what he calls the ‘circumstances of compro- mise’, in which people have reasons to work toward a compro- mise. He describes his concern as being ‘mainly with the circumstances of compromise for comparatively simple, two party conflicts’;17 yet, the circumstances he describes crucially place the two parties within an institutional setting. The persons contem- plating whether or not to compromise cannot just avoid dealing with one another, because – somewhat like the citizens of a state – they co-exist within a framework requiring their continued co- operation. In his leading example, Benjamin describes an argu- ment between a nurse and a doctor in an intensive care unit about (dis)continuing treatment of a particular patient. For the unit to carry on its vital functions the team must co-operate; when they disagree, a resolution must be found:

As co-workers in the intensive care unit, Nurse Chapman and Dr. Lehman are in some sense stuck with each other [. . .] and neither of them can be fully effective without the other’s assistance and cooperation. [. . .] This is also true for other continuing, cooperative relationships even if not so precisely defined. The desirability of preserving continuing, cooperative relationships among members of a family or citizens of a nation [. . .] also counts as a circumstance of compromise.18

CO-OPERATION DESPITE DISAGREEMENT 131

15 I am assuming throughout, that despite their strong disagreement, the agents view each other’s positions as lying within the bounds of legitimacy. Ronald Dworkin has offered a powerful philosophical account of how they can have such views while strongly disagreeing about abortion itself: Ronald Dworkin. 1993. Life’s Dominion. London. HarperCollins. For the halakhic context, see note 3 (above).

16 This holds even under an expanded notion of toleration, such as the ‘Affir- mative Tolerance’ advocated by Apel: Karl-Otto Apel. Plurality of the Good? The Problem of Affirmative Tolerance in a Multicultural Society from an Ethical Point of View. Ratio-Juris 1997; 10: 199–212.

17 Benjamin, op. cit. note 10, p. 26. 18 Ibid. pp. 30–32.

CDD here is not optional, as it might be between individuals who are not bound together by such a demanding framework. Cases where co-membership in such a framework provides the main reason to seek a compromise, form a distinct class of CDD prob- lems, bearing a close affinity to political CDD. There are impor- tant differences, of course, between the political level and the level of a particular ICU. But (as Benjamin notes) the reasons for co-operation are basically the same; I shall call them institutional reasons.

But in our example, P is not part of the healthcare team, so there are no institutional reasons that directly impel A to co- operate with her. Let us suppose that A is a physician; does this role imply any reason to seek CDD? This question must be broken down into two separate queries, depending upon whether P has been A’s patient or is a newcomer to him.

Admittedly, the division between patients and newcomers is by no means a simple dichotomy. P may have never seen A before, yet she might be ‘his patient’ in a morally significant sense by virtue of some professional or institutional arrangements. For the sake of analysis, however, let us consider the two clear-cut cat- egories of ‘patient’ versus ‘newcomer’, where the latter can stake no special claim to being cared for by a particular physician. If P is a newcomer in this sense, then from A’s individual perspective, it seems that he may – indeed, given his moral disapproval of what she is requesting, even should – turn her away.19

But, what if she is already his patient? Healthcare delivery is organised in many different ways, so there seems little point in describing exactly how the doctor-patient relationship became established prior to the request for an abortion. Moreover, our inquiry should extend to other healthcare workers who (are asked to) contribute less directly to performing the abortion. Their involvement normally arises in a secondary fashion, following that of the patient’s physician. For these persons there is even less room for wondering how they came to have any carer-patient obligations requiring an act they deem forbidden. But let us begin with the basic scenario, in which a doctor opposed to abortion is asked by his patient to perform this service for her.

This introduces the second kind of reasons for seeking CDD. These derive from specific commitments pertaining between

132 NOAM J. ZOHAR

19 Beyond the individual perspective, P may have some claim to become A’s patient, grounded in a broader societal or professional perspective. The ques- tion whether that claim could generate an obligation to engage in CDD calls for further reflection.

individuals, and I shall call them relational reasons. A may have no institutional reason to co-operate with P given their deep disagreement, yet he is not simply free to turn her away, since that would violate his commitment toward her implied by their relationship as doctor and patient. In general, such a relation- ship provides strong reasons to try to seek a way for CDD.

Nevertheless, it seems to be a commonplace that in this case, the doctor is at liberty to refuse (perhaps, indeed, he ought to). The reason for this lies in the gravity of the offence (as he per- ceives it): certainly the doctor-patient obligation cannot mandate committing murder – if that is how he regards abortion. After all, there are also limits to the CDD mandated by institutional and political reasons: the necessity of co-operation, whether between co-workers or between citizens, would not justify an individual in committing what he regards as murder. Conscientious objection defines the limits of political obligation.

This parallel suggests that the above-mentioned commonplace, positing a liberty to refuse, is straightforward only where A, asked to perform the abortion, believes it to be murder: let us call this case ‘Direct Murder.’ A should refuse, and is at full liberty to do so despite his prior commitment to P as a patient.20 Neither insti- tutional nor relational reasons for CDD appear strong enough to affect A’s severe perspective. This should not be generalised, however, into a judgement that a person should refuse to co- operate with anything he regards as wrong. ‘Direct Murder’ is not a model for ‘similar’ cases, but rather a limiting case, whose sim- plicity serves as a backdrop for the dilemmas of more complex scenarios.

The simplicity of ‘Direct Murder’ is due to a conjunction of three factors. First, the nature of the wrong (as seen from A’s per- spective) is so grievous as to preclude almost any kind of appeal to countervailing interests or rights of the patient. Second, the agent is being asked to perform the act himself – not merely to allow or facilitate it. Third, the wrong (as he sees it) consists in harming a third party, not the patient. The significance of each of these factors will be explored in the following sections. At this point, it is perhaps worth emphasising that we are looking at these factors from the perspective of the agent, not from a disengaged or ‘neutral’ perspective. That is to say, the question is not what stance other parties (say, A’s superiors) should adopt in the face of such a deep division between P and A. Nor is it what the proper

CO-OPERATION DESPITE DISAGREEMENT 133

20 The question of a duty to refer naturally arises; basically, however, this is another way of contributing to the same deed.

public policy should be, given that numerous citizens agree with P and numerous others with A. Rather, the question pertains to A himself: is it morally right for him to refuse this request despite his long-term commitment to his patient? And since, as seen from his perspective, he is being requested to directly kill a third party (i.e., someone other than the patient making the request), he should refuse.

As mentioned above with respect to the halakhic tradition, many who oppose abortion regard it as an evil lesser than murder. The parameters for CDD regarding such lesser evils call for a separate discussion. Here, however, I shall focus on what seems to be the greater and crucial challenge, namely, that involving CDD with respect to acts of killing, and will accordingly alter only factors #2 and #3, while holding the first factor constant. Even where the wrong consists in ending a human life, once either of these two factors is missing it is no longer manifest that the doctor’s perspective should prevail.

III ACTION, ASSISTANCE AND COMPLICITY

Let us consider the following scenario. A physician is asked by her end-of-life patient to help him commit suicide, where the patient – but not the physician – regards the suicide as justified in his par- ticular circumstances. The life of no third party is at stake. And since the deed is to be performed by the patient himself, her pro- viding (say) a subscription of a lethal drug does not constitute direct killing.

In a volume devoted to the PAS (Physician-Assisted Suicide) debate, I have offered a tentative analysis of such a scenario, describing two alternative approaches that the physician might take: a ‘hands-off’ approach versus a ‘co-operative’ approach.21 The ‘hands-off’ approach, grounded in preserving the physician’s integrity, advocates refusing any co-operation with a (self-) killing that she regards as wrong. But under the ‘co-operative’ approach – grounded in a relational stance – the physician should set aside her own (prohibitive) perspective and go along with the patient’s view.

In that essay, I left off without deciding between the two approaches. In light of my analysis here regarding reasons for co- operation, I am inclined to endorse the ‘hands-off’ stance where

134 NOAM J. ZOHAR

21 Noam Zohar. 1998. Jewish Deliberations on Suicide: Exceptions, Tolera- tion and Assistance. In Physician Assisted Suicide: Expanding the Debate. M.P. Battin, R. Rhodes & A. Silver, eds. New York. Routledge: 362–372.

we cannot assume a doctor-patient relationship. But where such a relationship exists, the physician is not morally at liberty to simply walk away. This patient is part of her moral universe, and his moral perspective has a claim on her moral commitments. First, these two people should communicate and – in the spirit of Habermas’s ‘communicative action’ – seek to forge a consensus for their interaction. But consensus is not always possible; as argued by Iris Marion Young, an ethic of communication and deliberation must allow for the persistence of difference.22 Assum- ing the basic difference holds, I believe that the physician has a relational reason to seek CDD. But, given that the disagreement persists, on whose terms should they co-operate? Is it not true that the physician, should she yield to the patient’s request, would become complicit in (what she regards as) evildoing?

One way of addressing such a situation is in terms of conflict- ing obligations. The physician has an obligation not to promote the illicit taking of life, but also an obligation to help her patient. Boyle argues that co-operation motivated by such a countervail- ing obligation can be viewed as constrained by (moral) necessity, and might thereby be justified – although it is by no means directly evident which of the two obligations should prevail.23 Where the conflict is construed as obtaining between two first- order obligations of the same agent, we arguably have the makings of a genuine moral dilemma.

I propose, instead, that the physician’s duty toward her patient is better viewed as an obligation to seek CDD, allowing a second- order decision to shift perspectives and adopt the patient’s (moral) perspective as her guide for action. The relational frame- work calls upon the physician to transcend her stance as an iso- lated individual; she might then say, ‘In general, I oppose suicide – even for a terminally ill person. But in dealing with this par- ticular patient, who happens to hold that his (proposed) suicide is appropriate, I shall adopt his valuation and act to facilitate his decision.’24

CO-OPERATION DESPITE DISAGREEMENT 135

22 Iris Marion Young. 1996. Communication and the Other: Beyond Delib- erative Democracy. In Democracy and Difference: Contesting the Boundaries of the Political. S. Benhabib, ed. Princeton. Princeton University Press: 120–135.

23 Boyle, op. cit. note 4. 24 Compare the argument advanced by Benhabib, that a ‘communicative

ethic’ requires engagement with the concrete other (Seyla Benhabib. 1987. The Generalized and the Concrete Other. In Women and Moral Theory. E.F. Kittay & D.T. Meyers, eds. USA. Rowman & Littlefield: 154–177). More radically, Noddings has argued that in a ‘one-caring’ relationship, an agent must fully take on the other’s perspective (Nel Noddings. 1984. Caring: A Feminine Approach to

Naturally, there arises the question of symmetry. If the rela- tionship requires sharing of perspectives, why shouldn’t the patient adopt the doctor’s perspective? Perhaps the patient should say, ‘In general, I believe suicide by the terminally ill is morally legitimate. But in dealing with this particular physician, who happens to hold that my (proposed) suicide is forbidden, I shall adopt her valuation and refrain from seeking help in dying.’ This would not bar the patient from seeking help from another physician, but it would arguably cancel out any claim that the physician ought to adapt to the patient’s outlook.

Is the situation truly symmetrical? We might endeavour to assess the matter either in terms of the relationship or in terms of the concrete situation. The doctor-patient relationship is decid- edly non-symmetrical in more than one aspect; two opposing con- siderations come to mind. On the one hand, there is the fact that the doctor is acting in her professional role, and the profession (here: the medical association) may have defined a stance reject- ing all suicide.25 On the other hand, it may be thought that the doctor’s special duty to care compassionately for her patient extends even to supporting his choice regarding the manner of his death.

Apart from these considerations, however, looms the difference in existential significance for each of the agents involved. Through sharing perspectives, the physician may come to see that the moral stakes are higher for the patient. She might end up saying, ‘For me, causing the ending of a human life is a serious transgression. Yet in this particular instance, the life in question is that of this patient, for whom ending his life as he desires is a far more momentous matter.’ This might be enough to tip the scales of symmetry and to instruct the doctor to adopt her patient’s valuation. A relational response along such lines – combined with a complementary notion of physicians’ pro- fessional duty – prevails, apparently, in the Netherlands, provid- ing grounds not only for PAS but also for active, voluntary euthanasia.

136 NOAM J. ZOHAR

Ethics and Moral Education. Berkley. University of California Press). Davion points out that, if indeed relational commitments entail such a response, one must beware of involvement with persons whose projects one considers evil, so as not to become a contributor to their evildoing. This seems right as far as it goes; but Davion does not consider the kind of cases with which I am concerned here, where moral disagreement is coupled with a recognition of the other’s view as legitimate (Victoria Davion. Autonomy, Integrity and Care. Social Theory and Practice 1993; 19: 161–182).

25 This actually appears to be the case in many countries.

I do not expect that everyone will respond this way – particu- larly not every physician. Relational sharing of perspectives is not a recipe for a predetermined conclusion. For many, the balance between the two perspectives – that of the patient and that of the physician – is likely to be weighted toward the agent who is called upon to perform the act, rather than toward the subject whose life-and-death is at stake. Thus factor #3 alone will probably not decide the issue. Even though – unlike abortion – active euthana- sia does not constitute harming a (putative) third party, it is likely to remain a matter of personal determination between physician and patient. This may occur explicitly – as in the Netherlands – or instead, in the more common, veiled form of administering large doses of pain-relieving medications.

For the focus to shift decisively toward the patient’s perspec- tive, factor #2 must be changed as well. That is to say, the patient must be not only the sole object of the deed in question, but also the agent for its performance, with the physician’s role reduced to assistance only. For a physician who accepts that there are rela- tional reasons to seek CDD with her patient, I propose the fol- lowing: even if, after engaging in perspective-sharing discourse with her patient, she is not prepared to adopt his permissive stance and to directly bring about his death, she has a strong reason to adopt that same stance with regard to assisting him in ending his life by his own hand.

At his point, more than one reader is likely to respond with exasperation: ‘Oh, that same old active/passive, direct/indirect sophistry! If, despite the patient’s suffering and pleading for an early exit, it is wrong for the doctor to kill him, how can it be okay for her to assist him in killing himself ?’

My rejoinder is that an analysis in terms of perspective-sharing can yield a uniquely convincing answer here, independent of one’s position about the aforementioned distinctions in general. It may well be that, if one gives primary weight to considerations of personal autonomy, then condoning the patient’s suicide should lead also to condoning a lethal injection. Similarly, if con- sequentialist calculations weigh in favour of an early and painless demise, there may be no sense in distinguishing between differ- ent paths that lead to the same result. On the other side, a deon- tological judgement may yield a sharp distinction between killing and suicide. All these are first-order positions, and we already assume that the doctor and the patient each subscribe to one or another of them. But, given a commitment to CDD, they now face the second-order question of determining through whose perspective to make the final call. It is in this that the difference

CO-OPERATION DESPITE DISAGREEMENT 137

between performing a deed and merely assisting in it makes a world of difference. For an agent’s guilt for facilitating the deed of another is in principle derivative; hence, since the doctor recognises that the patient’s deed is justified from his perspective, her assistance in its performance is equally permissible.

In his classical discussion of complicity, S. Kadish seeks to ground the law of complicity in a moral frame of reference.26 As he explains, when X facilitates the deed of Y, our basic notions of personal responsibility bar us from attributing guilt to X in terms of causation, since Y’s responsibility rests on his having free will in choosing whether or not to perform the deed. Instead, we turn to the notion of complicity: X may be held accountable as an accomplice to Y’s transgression. From this doctrinal foundation it follows that X’s guilt is necessarily derivative: X can be held as an accomplice only if Y’s action is indeed a crime.

As Kadish shows, this leads to some problematic results. Some- times Y is found not guilty by virtue of a strong excuse, or cannot be prosecuted altogether because of some agent-specific reason, such as diplomatic immunity. In such cases, it still makes sense to regard Y’s action as evil, even if Y cannot be held accountable for it; and thus it is still possible to attribute derivative guilt to X.27 But the most perplexing cases are those in which the primary agent Y acted rightly, yet we feel that X ought to be held account- able. Kadish cites a case in which two felons are trapped in a build- ing, surrounded by the police. Felon X tells felon Z that, ‘the coast is clear’ at the back entrance, knowing full well that the police are waiting there; felon Z, believing this, runs out with his gun in hand and is killed by police officer Y. Now X did not shoot Z, and seemingly he can be held accountable only as an accomplice to Y’s shooting; but Y’s deed was perfectly justified.

In this type of case, we certainly want to condemn X morally, and perhaps also legally, even in the face of a doctrinal paradox. Should the doctor, perhaps, similarly see her own contribution to the patient’s suicide as wrong, even though from his perspective the act is permissible? I think the comparison is unconvincing. The police officer’s deed was justified in itself, but felon X’s inten- tion of getting felon Z slain is not justified from any perspective.

138 NOAM J. ZOHAR

26 Sanford Kadish. Complicity, Cause and Blame: A Study in the Interpreta- tion of Doctrine. California Law Review 1985; 73: 323–410.

27 This analysis is not substantially altered even if we adopt the alternative view regarding complicity, advanced by M. Gur-Aryeh. Miriam Gur-Aryeh. 1987. Comment on ‘A Theory of Complicity.’ In Issues in Contemporary Legal Philosophy. Ruth Gavison, ed. Oxford. Oxford University Press: 304–310.

We hold him morally accountable for the death because we regard him as having manipulated another to bring about his evil goal. He is guilty for using officer Y to get Z killed, even though technically he might be (at most) an accomplice, since he merely spoke some words, while another did the shooting.

Applying a similar analysis to the PAS scenario, we would have to say that the doctor brought about the patient’s death – some- thing she regards as a wrong – through ‘manipulating’ the patient into killing himself. Clearly, that would be a gross mis-description of the relational sharing of perspectives we have been positing. The police officer in Kadish’s case is an unwitting agent; the patient, by contrast, is acting for himself with full knowledge and responsibility. The doctor is merely a participant in his deed, and as such she should judge herself according to his moral perspec- tive – even while maintaining that in her own deeds she should remain true to her own perspective. The patient’s moral agency precludes ascribing to her any causal responsibility for his freely chosen act.

Thus, what makes the moral difference is not some (first order) distinction between direct and indirect action or causation. Rather, the crucial difference between euthanasia and PAS lies in the agent-specific attribution of the deed, which in turn governs the appropriate outcome of relational perspective-sharing.

IV ASSISTING IN ABORTION IN AN INSTITUTIONAL SETTING

We can now return to the case from which we began, that of the anaesthesiologist – let us designate him ‘A’ – asked to facilitate an abortion. Let us assume that the woman seeking abortion is not his patient, so that he has no relational reason to seek CDD with her. Nevertheless, he does have institutional reasons to co-operate with his colleagues – primarily with the woman’s doctor D, who (we posit) sees this abortion as justified and intends to perform it. If the above analysis of assistance is correct, A should engage in perspective-sharing with D. Ideally this can lead to agreement; alternatively, A might be able to arrange for another anaesthesi- ologist to take his place. The latter option is akin to referral – an option often advocated for a physician who has moral reservations about a course of treatment sought by his or her patient. One may wonder, of course, why an agent who is seeking to avoid facilitat- ing an action would be more comfortable arranging for its per- formance through someone else’s support. A plausible answer might be that not all modes of facilitation are the same: adding

CO-OPERATION DESPITE DISAGREEMENT 139

an intervening agent creates more distance between the facilita- tor and the evil deed.28 In any case, this solution is not always avail- able; with regard to abortion, policies allowing for ‘opting out’ have apparently greatly curtailed its availability in numerous loca- tions.29 When the chips are down, A arguably ought to accept D’s perspective concerning assisting in the latter’s action.

But here we must consider a major objection, for A’s situation is only partly similar to that of the physician in the PAS scenario. The similarity is with regard to factor #2: both these doctors are being asked to assist in the deed of another agent, with whom they are bound to seek CDD. But with regard to factor #3, A’s sit- uation is significantly different, for – in his view – there is a third party involved, namely the foetus.

In response to this objection, it is possible to enlist factor #1, stipulating that the duty of assistance applies only to an anaes- thesiologist who – although strongly opposed to abortion – does not regard it as murder. In terms of Sofer-Abraham’s essay, from which we began, even this would be a major correction. However, should it be conceded that, for an agent who views abortion as murder, a refusal to engage in CDD is warranted not only for ‘Direct Murder’ but also for all manner of facilitation? After all, the PAS scenario also involves an act of killing, yet I argued that factor #2 (indirect participation) should decisively lead to adopt- ing the perspective of the agent performing the deed.

One part of an answer may be found in the fact that the very existence of a third party is the issue about which the two agents (A and D) disagree.30 If we follow the analysis of assistance and complicity offered above, the moral assessment of A’s (facil- itating) actions is derivative, wholly dependent upon the assess- ment of D’s actions. If no alternative arrangement can be found, they should seek CDD and mutually share their differing per- spectives. If this does not produce agreement, it is D’s perspective that should prevail, even for A – with regard to assessing D’s

140 NOAM J. ZOHAR

28 For an illuminating discussion, see: Judith L. Kissel. 1997. A Comprehensive View of Complicity as Positive Collaboration and Toleration-of-Evil. PhD Dissertation. Georgetown University.

29 See: Rebecca S. Dresser. Freedom of Conscience, Professional Respon- sibility and Access to Abortion. The Journal of Law, Medicine & Ethics 1994; 22: 280–285.

30 If D’s support of abortion in the case at hand is grounded in a position like that expounded by J.J. Thomson, then this line should be amended: they disagree about whether there is a third party with a valid claim against being killed through removal from the woman’s body. Judith Jarvis Thomson. A Defence of Abortion. Philosophy and Public Affairs 1971; 1: 47–66.

actions, including A’s contribution to them. And from that per- spective, there is no third party to be concerned about. This is decisively different from ‘Direct Murder’, where the physician – even if he has a relational reason to seek CDD with his patient – has no compelling reason to adopt her perspective (denying a third-party concern) over and against his own recognition of a third party.

This conclusion is supported by the analogy between institu- tional and political reasons for seeking CDD (see section II above). As noted above, in the political realm the necessary co- operation extends as a matter of course to issues of life and death. Still, it seems clear that an individual commanded by state author- ity to directly perform what he or she regards as murder may (and should) refuse. Thus if we are to endorse CDD between citizens, this must include a crucial distinction between directly commit- ting what one regards as murder, and facilitating the deeds of one’s co-citizens (through the state’s actions) that they legiti- mately regard as appropriate,31 even when one regards those deeds as murder.

It seems that a moral justification of such facilitation would have to proceed along the lines suggested here with respect to the institutional setting, granting primacy to what I called factor #2. ‘Direct Murder’, and perhaps even ‘Direct Killing’ (an apt caption for the euthanasia case), would be rightly resisted. But wherever there are reasons for seeking CDD (whether political, institu- tional or relational), and agreement cannot be reached, an agent asked to facilitate the deed of another should adopt the per- spective of the main actor in assessing the morality of his or her contribution.

Noam J. Zohar Department of Philosophy Bar Ilan University Ramat Gan 52900 Israel [email protected]

CO-OPERATION DESPITE DISAGREEMENT 141

31 The qualifying word ‘legitimately’ is intended to reflect the limits of co- operation, as mentioned at the outset.