Joseph S. Ellin
Special Professional Morality and the Duty of Veracity
(From Business and Professional Ethics Journal 1.2 (1982): 75- 90.

TWO THEORIES OF SPECIAL PROFESSIONAL MORALITY
Are there special rules and principles which govern professionals in their professional conduct? This question has been called (by Alan Goldman) "the most fundamental question for professional ethics." 1 But what exactly is being asked? Are we supposed to imagine the possibility of a fundamental conflict between ordinary morality and the special standards of the professions?2 It certainly seems as if the standards of professional morality do not necessarily correspond with those of ordinary morality: lawyers, for example, are required to defend vigorously the interests of unsavory characters who by ordinary moral standards are entitled to no assistance from anyone. But the question is not whether professionals subscribe to such special standards, but whether such standards would amount to a departure, within certain limited circumstances, from ordinary moral rules. But departures from ordinary moral rules, it seems, can only be justified by reference to other, more basic, principles of morality. We must have good moral reasons for allowing such exceptions to our general principles. Where there are apparent conflicts between ordinary and special morality, these conflicts can generally be adjudicated by moral reasoning based on large moral considerations. This view, which we may call "the priority of ordinary reflective morality, gives to ordinary morality a double function. First, ordinary reflective morality imposes the rules and standards which govern all of us in our ordinary, that is non-special, life encounters. Second, ordinary reflective morality plays an adjudicating role, resolving apparent conflicts between obligations in ordinary contexts and those in special contexts, such as those of professional life.

To hold the priority view is not, of course, to hold that in cases of apparent conflict, one's ordinary obligations always override, as if professionals could have no obligations inconsistent with those imposed on everybody by ordinary morality. There might be good moral reasons why we would want to impose special obligations on professionals in their professional life. Nor is it necessarily to hold that there can not be fundamental or irreconcilable moral conflicts which arise in the course of one's professional activities. The crux of the "priority" view is that whether there are such special obligations, and such irreconcilable conflicts, is to be determined by ordinary moral considerations. At bottom, no conflict can occur between an ordinary moral norm and a norm drawn from some special context; such conflicts that do occur are conflicts between the norms of ordinary morality itself. Suppose, for example, a professional finds a conflict between an obligation to tell the truth and an obligation to protect a client from a foolish blunder. The professional might conclude that he or she is forced simply to decide between inconsistent norms. On the priority view, such a conflict might be possible, but when it occurs, it does not arise because ordinary morality supports one norm and professional morality the other, but rather because ordinary morality supports both norms and provides no deeper principle of reconciliation.

The priority view holds that, since ordinary reflective morality is the only source of moral obligation, there is no such thing as a morality that is "internal" to special contexts such as professional life. The morality of the professions, rather, is imposed on them from above, by our usual values and common principles. Professional morality is, thus, [comprised of] a set of rules which is, in a sense, purely external to the professions. [These rules] do not grow out of any conception of professional life. The obligations of professionals must be assessed and determined strictly in accordance with our usual moral standards.

The priority view, then, gives us our first theory of special professional morality. According to it, there may well be special rules which govern professionals, and which impose duties inconsistent with the duties imposed on everybody in ordinary life. But the sole justification for imposing such special duties is that, judged by the norms of ordinary morality, better moral consequences' , (as Goldman puts it) will result.3 Since every apparent conflict between ordinary and special morality can, by reflection, be resolved into a conflict within ordinary morality itself, the professional faces no greater moral difficulties than anybody else.4

However, this reduction of special professional to ordinary morality may seem to professionals themselves to distort their moral conflicts. Lawyers, for example, who defend the interests of clients whose interests are morally indefensible might think they are doing their duty as lawyers, and that in so doing they act in the face of ordinary morality, not with the ultimate sanction of ordinary morality. Doctors might believe that as doctors they have a special allegiance to the norm of health, a norm which they must respect even to the extent of sacrificing other moral interests a patient might have. It is such conflicts which make professionals even more morally uncomfortable than the rest of us. Since their obligations are derived from a certain conception of their profession, there may well be cases in which a conflict between two norms would be resolvable, were it simply a matter of resolving conflicting norms within ordinary morality; yet this does not settle the conflict between the professional and the ordinary norm.5

This gives us our second theory of special professional morality, according to which there are moral obligations which derive not from ordinary morality but from the nature of the professions. We may call this the parallel view, because ac- cording to it, professional morality is parallel with, not subordinate to, ordinary morality. The profession itself is a source of special moral obligations; hence fundamental conflicts with ordinary morality are possible, dilemmas of professional morality are not dissolvable into dilemmas of ordinary morality, and professionals are faced, at least potentially, with moral perplexities which are different than those faced by ordinary people.

...I propose to defend the "parallel" view. To do so, what is needed is some conception of professional ethics which makes fundamental conflicts possible. Of course it is not enough to point out that professions exist for certain ends, for example, medicine for health, as if this disposed of the question whether professionals are thereby obligated to pursue those ends even in violation of usual moral standards. What I suggest instead is that special professional morality internal to the professions may be derived through certain conceptions of what a profession is. Different conceptions are possible, and each different conception suggests different moral principles. More specifically, we can produce certain models of professional activities and relationships and examine these models to discover what obligations they suggest. For example, to conceive of a doctor on the model of a parent suggests certain obligations which would not be suggested were we to conceive of a doctor as a friend or counselor (a parent has paternalistic duties; a counselor gives advice). The obligations so suggested would be internal to the profession alone since they would depend on the nature of the model. Since these obligations would not be imposed on the professions by reflective ordinary morality, there might well be fundamental conflicts which ordinary morality could not resolve.

I intend to go further, however, and argue that there actually are such fundamental conflicts, at least if we accept a certain conception of a profession. My argument shall be relatively narrow in scope. I shall consider only one professional relationship, that of professional and client. I shall consider only one model of that relationship, the fiduciary .I shall use a modeling scheme according to which five models of the professional-client relationship are presented. Each of these models suggest somewhat different conceptions of the duty of veracity. I concentrate on one of the models, the fiduciary model, in order to contrast its conception of the duty of veracity with that of ordinary morality. These conceptions are not only significantly different, they are inconsistent. Hence, I will argue that if the relationship of professional to client is thought of as a fiduciary relationship, there is a fundamental conflict between ordinary and professional obligations.

THE ORDINARY DUTY OF VERACITY
In order to substantiate this claim, it is necessary to examine exactly what ordinary reflective morality tells us about veracity. Of course not every principle need be discussed, and I will consider only the distinction between lying and deception, and their relative moral wrongness. My main point is that ordinary morality provides reasons not only for condemning lying and deception, but also for considering deception a lesser wrong then lying. When we turn to the fiduciary relationship between professional and client, however, we find the "moral gap" between lying and deception much greater than in ordinary morality: Professional morality considers lying a more serious wrong than does ordinary morality, yet does not consider deception morally wrong at all. Hence, there might be situations in which, from the point of view of ordinary morality, it would be justifiable to lie, but not from the point of view of professional morality; there are also situations in which ordinary morality would, but professional morality would not, prohibit deception.

The distinction between lying and deception is fairly obvious, although the definition of lying is a question of some philosophical complexity .6 We may consider a lie to be a statement which the speaker believes to be false.7 What I call a deception, on the other hand, consists either in true statements which are nonetheless misleading, or in actions which convey a false impression, or in the deliberate withholding of information where the person not informed is misled into drawing a false conclusion. Deception can be inadvertent, but where it is deliberate, the agent must want someone to draw a false conclusion. Now most of us see a clear moral difference between lying and deception: Where we think it necessary to plant false beliefs, most of us would prefer to do so without actually lying, if we can.8 Nonetheless this interesting moral distinction is commonly overlooked by philosophers, who seem to think that the duty of veracity is simply the duty not to deceive.9 When, however, philosophers do acknowledge the distinction, they are apt to think that it is morally invalid; or else they accept its moral force but fail to explain it in any very satisfactory way. For example, Benjamin and Curtis, in their book Ethics in Nursing,10 citing a case in which a nurse conveys a false impression about a patient's medication without actually lying to the patient, assert that the nurse would be "compounding deception with self-deception if she were to believe that there is a significant ethical difference." The authors leave the impression that in their view there can never be a "significant ethical difference" between deceiving and lying, although the example shows no more than that in the given case, in which the patient is assumed to have a clear right to the information, the wrong to the patient is so great as to make insignificant any difference in the way in which the deception is accomplished. But this does not show that in cases where deception is justified, there is no significant difference between deceiving by lying and deceiving by evasion or by withholding information; nor does it show that there are no cases in which it might be justified to deceive by evasive or misleading statements but not justified to deceive by telling a lie.

What would be an argument against the distinction between lying and deception? There might be many. It might be pointed out that lying and deception are equally harmful to the person deceived, since they deprive him of information rightfully his. It might be argued that whether one plants false beliefs by means of lying, or by means of evasive or misleading statements (or actions), is a question of means, whereas morality judges by intentions and consequences, and lying and deception are done with the same intention and (if successful) have the same consequence, namely, that someone is wrongfully made to have a false belief. Or it could be said that to deceive someone is to show that you do not respect that person, for you claim for yourself the right to give him false beliefs and thus to manipulate him: Hence deception is a form of contempt for others. Thus deception, no less than lying, harms a person's interest in having true beliefs, in having the information necessary to make intelligent decisions, in not being manipulated, and in being regarded with respect. What these arguments show, however, is that there are very good reasons why we should object to deception, and that some of them are the same reasons we have for objecting to lying. What they do not show is that there are no additional reasons for objecting to lying, which are not reasons for objecting to deception.

Let us consider some philosophers who acknowledge the distinction between lying and deception. Peter Geach, for instance, who counsels "total abstinence from lying," tells the amusing story of one St. Athanasius, who "was rowing on a river when the persecutors came rowing in the opposite direction: "Where is the traitor Athanasius?" "Not far away," the Saint gaily replied. Here we have the concurrence of a contemporary moral philosopher with the opinion of a Christian Saint that when deception is justified, the gain in producing it without actually lying is great enough to be a cause for gaiety. No less a moralist than Kant, who even in the Lectures on Ethics 12 seems to take a position absolutely prohibiting lying (or what Kant calls lying), gives the following example: "It is possible to deceive without making any statement whatever. I can make believe, make a demonstration from which others will draw the conclusion I want, though they have no right to expect that my action will express my real mind. In that case, I have not lied to them. ...I may, for instance, wish people to think that I am off on a journey, and so I pack my luggage. ..."13 This is a case in which deception, at least in Kant's view, is not even morally problematic, although lying (suppose I told my neighbors, "I will soon be off on a journey") would be excluded.

The philosophers Chisholm and Feehan offer an explanation for the distinction. They write:

Why is it thought wrong to lie? And why is lying thought to be worse, other things being equal, than other types of intended deception? The answer would seem to be this. It is assumed that, if a person L asserts a proposition p to another person D, then D has the right to expect that L himself believes p. And it is assumed that L knows, or at least that he ought to know that, if he asserts p to D, while believing himself that p is not true, then he violates this right of D's. But analogous assumptions are not made with respect to all other types of intended deception. When the man of Kant's example packs his luggage, he does not thereby give his friends the right to assume that he is about to make a journey. Lying, unlike the other types of intended deception, is essentially a breach of faith.14

What this explanation fails to tell us is both why D has such a right against L when L asserts a proposition, and why D does not have an equivalent right when L does something without asserting a proposition. (Why do I not have a right to assume that if you pack your bags, then you are going on a journey?) Clearly Chisholm and Feehan, like Kant himself, are far too tolerant of "other forms of intended deception." Kant's neighbors would complain that they had been deceived, and there would be some merit in their complaint; though we might indeed concede that there is some merit in their friends' defense that he had not given them "the right to assume" that his action of packing his bags would in this case have its natural and expected consequence.

These authors fail to explain what is wrong with deception, and why deception, though wrong, is not as wrong as lying. They leave the impression that if deception is wrong, it must be wrong for some other reason altogether than the reason (breach of faith, according to Chisholm and Feehan) that makes lying wrong. But we must not defend the distinction in such a way that deception turns out either to be not wrong at all, or wrong for entirely different reasons than the wrong of lying. Ordinary morality holds that lying is worse than deception, not that it has another moral character altogether . Fortunately, we have already seen excellent reasons why deception should be considered morally wrong, and analogous to lying. I will now give three arguments to show why lying is a greater wrong. The first two distinguish between lying and deception by degree only: Lying is a greater violation of a principle which deception also violates. The third argument, drawn from Kant, does, however, introduce a new idea: that lying violates the social contract in a way that deception does not.

The first argument is that the liar takes advantage of weakness more than does the deceiver. Consider a typical case of deception in the professions: The surgeon, who smilingly enters the patient's room and reports that the operation went very well, but fails to mention that the findings were devastating. The deception occurs (the false belief is formed) only if the patient assumes that when the surgeon said the operation went well, he meant well from the patient's point of view rather than from the surgeon's (i.e., there were no complications or unexpected difficulties). This may be a natural inference, but it is an inference: If deception harms the victim's interest in the truth, this harm can occur only if the victim draws an inference grounded on what he has been told (or has observed, in Kant's example). We can defend ourselves against this harm by adopting the following maxim of prudence governing belief: "Believe everything you are told but draw no inferences unless supported by independent evidence." If we followed such a rule no one would ever be harmed by deception. Such a rule would impose far fewer burdens on life than a rule which protected against lying, to wit, "Believe nothing you hear unless it is supported by independent evidence." Most of us find the cost of following either rule excessive, hence we are vulnerable to the deceiver as well as to the liar. But the costs are not equal. Each rule may be regarded as a defense, where the cost of the defense against lying is greater than the cost of the defense against deception. Since weakness may be measured by the costs of defense (the more it costs to protect yourself against something, the weaker you are with respect to that thing), and since we are therefore weaker with respect to lying than we are with respect to deception, we can say that the liar takes advantage of our weakness to a greater degree than the deceiver, and is consequently morally worse, even though the harm produced by each is the same.

The second argument follows easily from the first. The liar is more responsible for the harm caused than is the mere deceiver. The reason for this is that in mere deception, the person deceived participates in his own deception, hence is in part responsible for causing it. This is usually obvious: In the case of the surgeon, the patient could have unmasked the deception simply by asking the surgeon a direct question about the findings. The patient is at fault for failing to ask. Even where there is a lie, of course, the victim must bear some responsibility for being deceived, since he has imprudently trusted the liar and failed to verify the statement made to him. But usually it is unreason- able for the victim to seek such verification: In the absence of reason to think otherwise, it is unreasonable not to accept for truth direct statements made to you. But the victim of deception does not simply believe what is said: He draws an inference which is based on, but not verified by, the evidence offered, and then fails to ask the speaker to confirm the inference. The victim of the lie fails to verify a direct statement, but the victim of deception fails to verify a conclusion of his own and fails to seek confirmation from the speaker, and hence is more responsible for the ensuing harm (his coming to hold a false belief).

We now come to the third difference between lying and deception. Suppose we adopted a social contract point of view and postulated that the duty of veracity depends on an original undertaking not to deceive. A lie would then be a violation of this implicit agreement. Such a view was held by Ross, who, however, confined the duty of veracity to the duty to use language truthfully: "Yet the peculiar stringency of the duty of veracity seems to spring from an implicit understanding that language shall be used to convey the real opinions of the speakers ..." 15 We could even make the strong claim that unless there were such an implicit understanding, speech itself would be frustrated (and society as we know it impossible), since words establish their meaning only by being applied in standard situations, that is, by being spoken when they truly apply. Now the social contract point itself does not establish a difference with deception, since the undertaking is not to deceive, not merely not to lie. To establish a difference we would have to make one or both of two further points. The first is that the promise to speak the truth is more important than the promise not to deceive, since speech is necessary to any human society, whereas non-deception is necessary only for a tolerable or decent society. Truth we might say is an enabling condition for society, whereas non-deception is but an enhancing condition. This point, though powerful, rests on the assumptions above connecting meaningful speech with truth telling and with human society, assumptions clearly beyond our present scope.

The second point depends on our interpretation of the social contract, that is, how we affirm and reaffirm the promise to be truthful. Suppose we held that in addition to the underlying' 'implicit understanding, " there is also a more explicit promise made every time we speak, so that to speak at all is virtually to warrant that our words are true. A lie would then amount to a violation of the very promise made by the speech act in which the lie is stated. If we further supposed that it would be implausible to make a parallel claim about deception, then a clear moral difference would emerge: A lie violates a warrant given by the very act of speech, whereas deception violates at most only the underlying "implicit understanding." Such a view would explain our feeling that a liar is less trustworthy than a mere deceiver, since the liar violates his promise in the very act of making it, whereas the deceiver violates only the remote understanding of the original agreement.

VERACITY AND THE PROFESSIONAL-CLIENT RELATIONSHIP
I now consider that I have both elucidated and defended what ordinary reflective morality has to say about veracity. According to it, lying and deception are both wrong for a number of basically similar reasons; but there are also good reasons for considering lying morally worse than deception. I now propose to argue that the situation is different with respect to professional morality. But first we must resolve a difficulty: What is professional morality? From what does it come, given that we are prevented by our previous rejection of the "priority" approach from simply deriving its precepts from ordinary morality. We have already indicated that some "conception" of the profession is needed, a conception according to which it might be possible to hold, for example, that the overriding obligation of the doctor is to the patient's health, or of the lawyer to the client's strictly legal rights. I propose to solve this problem by presenting models of a profession, or more precisely, models of the professional-client relationship. Modelling is a frequently used device, which enables us to understand basic relationships within the subject matter being modelled, by representing these relationships as something else, presumably more easily understood.

A number of interesting modelling proposals have been put forward in professional ethics, no- tably with respect to the doctor-patient relationship. I shall just note one, that of Szasz and Hollender .16 They distinguish three models, which they call activity-passivity, guidance-cooperation, and mutual participation. Although they do not state explicitly what principle governs the derivation of the models, the idea seems to be the location of the control of the relationship: In the first two models the doctor controls, in the third, doctor and patient "mutually participate." (They do not consider the possibility that the patient might control the relationship.) The difference between the first and second models seems to be the kind of control exercised: The "cooperative" patient "is expected to. ..'obey' his doctor. ..he is neither to question nor to argue or disagree with the orders he receives."17

Szasz and Hollender's principle, to distinguish models according to the location of the control of the relationship, is provocative, and I follow it in my own proposals. If we consider the ability to control the relationship, then we have the following possibilities. Either the position of the parties is one of equal power and control, or it is not. If the position is one of equality, then we have two possibilities: Either the relationship is competitive or it is cooperative. If competitive, then we have an adversary model characterized by arm's-length bargaining and contractual agreements; that is what both Veatch and May, in their analysis of doctor- patient relationships, call contract, but with explicit recognition that contract involves mutual wariness based on the need to compromise conflicting inter- ests.1s A cooperative relationship, on the other hand, also based on equality, is characterized by partnership, mutual trust and other-reliance; this is equivalent to the models of mutual participation which Szasz and Hollender advocate.19

If the position is not one of equality, then again we have two possibilities: either the professional is superior to the client, or the client to the professional. In the latter case, we have an agency model: The professional merely carries out the wishes of the client. The client determines the ends to be attained, and the professional acts to achieve these ends, being an instrument of the client's will. But if the professional is superior, then we have two further possibilities: Either the professional uses the client for his or her ends, which we call exploitation (this arises, for example, where the pa- tient is an ' 'interesting case' , potentially suitable for write-up in the journals; or where the legal client has a novel problem which the lawyer may use to achieve a landmark legal victory); or the professional works for the client's ends, in which case we have a fiduciary model. Hence our principle gives us five models: adversary , cooperative, agency, exploitation, and fiduciary .

It is not my intention to discuss these models in detail, since I am primarily interested in the fiduciary .20 The important distinction is that between the fiduciary model and the agency model. In each case the professional works entirely for the client, so it may seem that there is no difference. But there are two senses of "for the client." In the agency model, the agent works at the direction of the principal, that is, he does only those things the principal himself would do, if he had the ability, knowledge or inclination. When the agent acts, it is as if the principal were acting. In the fiduciary model, however, the fiduciary acts in the interests of the beneficiary. Here the fiduciary does not necessarily act as if he were the beneficiary nor does he do only those things which the beneficiary would do, were he in a position to act for himself. Hence the fiduciary has an independent responsibility for his acts on behalf of the beneficiary .He is not a mere instrument of the will of another , confined only to judgments about means and methods but never allowed to determine ends and principles. In a fiduciary relationship, the professional services the true interests of the client, not the client's immediate wishes or desires.21

When we reflect on the duties a fiduciary relationship might entail, we are struck with a paradox. On the one hand, a fiduciary relationship rests on trust. On the other hand, a fiduciary relationship encourages deception. I propose to resolve this paradox by widening the moral gap between deception and lying.

That professional-client relationships rest on trust is an oft-noted platitude; but it is in fact not true of any of our models but the fiduciary .It is only in this model that the professional uses his or her superior position to serve the client's interests, possibly contrary to the client's wishes or desires and possibly beyond the client's ability to under- stand the procedures the professional recommends.22 If the professional merely carries out the client's wishes (agency) there is no need of trust, since the client can determine for himself whether the professional acts as desired. A contract model involves arm's-length bargaining between equals, which enables adversaries to reach agreements fair to each; trust is unnecessary .Partnerships indeed generate trust, but do not rest on it: If each partner is an equal, respectful of the other's knowledge and abilities, trust will naturally develop, but the partnership can carry on without it. Exploitation perhaps requires that the victim-client be trusting, but not that the professional-exploiter be trustworthy. Only in the fiduciary relationship must the professional act to protect the beneficiary , s interests as he alone understands them. In the absence of trust, the client is unlikely to submit to the professional' s authority, and will prefer an agency or contract model, in which the client may make use of the professional's services without the necessity of submitting to the professional's control. If the professional is not trustworthy, the relationship cannot be said to be fiduciary at all. It is not my intention to debate the merits of the models, but only to draw conclusions about one professional duty, that of veracity, within the fiduciary model.

VERACITY IN THE FIDUCIARY RELATIONSHIP
But why should the fiduciary relationship condone deception? The reason is that the relationship is governed by certain strictly defined ends. People enter into relations with professionals for specific and limited purposes: to improve their health, to protect their legal rights, to enhance their financial condition. If the relationship is fiduciary , the professional pledges to use his or her superior position to protect only those interests for the protection of which the relationship exists. One consults a doctor to protect one's health, not one's finances: It is not within the professional competence of a doctor to make judgments about financial matters. (This is not to say the doctor should not provide information about costs, but that he is not obligated to make judgments about them. The doctor tells me what, in this opinion, is the medically desirable treatment; my financial advisor tells me whether he thinks I can afford it.)

If we take this strict view of the fiduciary con- text, then we find a significant difference between ordinary morality and the special (professional) morality which governs the fiduciary relationship. Ordinary morality is designed to protect people's total package of interests which, as interests, are considered equally worthy of protection; but professional morality protects only those interests for which the professional relationship exists. Professional morality is thus spared the necessity of assigning weights to various interests in order to balance them in cases of conflict. Deception is thus not a violation of professional morality, since professionals are not mandated to protect the client's interest in having true beliefs, in not being manipulated or in being treated with respect. Nor can there be said to be any underlying contract beyond the pledge by the professional to use his or her superior position to further the goals of the relationship: health, legal rights, and so on. A patient's interest in the truth, for example, is exactly as relevant to the doctor's professional concern as the patient's interest in friends or enjoyable leisure: That is, all these interests are relevant to the physician's responsibility only as they might affect the patient's health. It is a medical judgment whether a person's health might be affected by the possession of certain information, and therefore, within the doctor's responsibility; but it is not a medical judgment, and so not within his responsibility whether a person's other interests ought to be respected at the price of some risk to the person's health. Insofar as information might affect a patient's health, what to tell the patient, whether to tell the patient, and how to tell the patient become purely medical questions, to be answered on medical grounds. Since the interests which deception harms are not otherwise relevant to the fiduciary medical context, there is no reason within that context to prohibit deception. To say this is at bottom only to say that a physician, acting as such, should make medical judgments, not financial judgments, not legal judgments, and not moral judgments either. ...

It is important to be clear about what I am advocating. My point is that professional morality as it emerges from a certain conception of the professional-client relationship does not prohibit deception; but this is not to say that the professional may deceive for whatever reason he or she chooses. Ordinary morality governs all human relationships, and ordinary morality does not normally condone deception. Professional morality, however, as I understand it, allows deception where necessary to protect the client's relevant interests, but these interests only. It is at the point where, in the professional's best judgment, possession of certain information might harm the client's relevant interests, that professional and ordinary morality may conflict.

Let us consider two examples. Suppose a patient who has been in a serious accident has undergone surgery to restore mobility in an injured leg. After some days, the surgeon concludes that the operation has failed, and that chances of recovered mobility are slight. It is unlikely the patient will ever walk again. This news may be expected to depress the patient badly. Should the patient be informed? Assuming that the depression will not affect the patient's physical condition in a negative way, then ordinary morality must govern. It is not the doctor's professional duty to protect patients from depressing news, even if health-related. (It may, however, be his, that is anyone's, moral duty in these circumstances.)

Suppose another patient has undergone major heart surgery, from which he is recovering. While the patient is in the hospital, the patient's business suffers serious reverses. The patient's doctors fear this news will sufficiently upset the patient so as to impede or perhaps preclude complete recovery. Should they advise the family not to inform the patient? Although the information in question is not itself medical, the judgment about its effects is a medical judgment, and their duty is to so advise. Now suppose (implausibly, in my opinion) ordinary morality might hold that the patient's right to know and to make all decisions concerning his own life, including his business affairs, outweighs the patient's interest in health, as Goldman seems to think.23 If so, there is a conflict between ordinary and professional morality and professional morality must rule.

But given that professional morality has a significantly greater tolerance for deception than has ordinary morality, how can we resolve the paradox that the professional relationship depends on trust? My answer is to require a significantly greater preclusion of lying. Indeed, it is probably best if professional morality excluded lying altogether. We have already seen some reasons why lying should be considered a greater violation of trust than deception. First, deception cannot occur without some blame, even complicity, on the part of the person deceived. In the case of deception, we might say, trust has not been pushed to the limit, since the direct question which would prevent the deception has not been asked. To this we may add (second), that lying vitiates, but deception does not, what may be called the maxim of trust, "You may believe what you are told. " Where this maxim is respected, then it, together with the prudent maxim, "Draw no inferences unless supported by independent evidence," will protect anyone from any kind of deliberate deception. Third, as we have seen, lying violates a promise made in the very act of speech, and so is a greater breach of trust than deception, which violates only the original under- standing. And fourth, we may consider that the fiduciary relationship depends on a pledge, by the professional, to protect the interests of the client. But if the professional is capable of lying, how can the client expect that this pledge will be respected? If the professional will lie to the client in order to protect the client's interests, perhaps he will also lie to the client about whether he will protect his interests.

Finally, since lying is a great violation of trust, there is a strong reason for requiring a greater prohibition of lying in a relationship governed by trust, than in ordinary morality. The reason is that trust is not as important in the ordinary relation- ships of life for which ordinary morality is de- signed, first, because to a large extent ordinary relationships depend on the recognition by the parties that everyone's self-interest is served best in the long run if everyone obeys the rules of morality (this is the contract element underlying ordinary morality); and second because in ordinary life we do not consign our interests to the care of another. What distinguishes the fiduciary situation from ordinary life is that the client-beneficiary, being unable to control the relationship, must trust the fiduciary-professional to act in his, the beneficiary's, best interest, even where this contradicts the beneficiary's own judgment. This significant difference between ordinary life and the fiduciary relationship is sufficient to explain why the fiduciary relationship must impose a more stringent prohibition against lying than does ordinary morality.

There will, therefore, be situations in which a professional should tell the truth when ordinary morality would condone lying. A woman has been seriously injured in an automobile accident in which one of her children was killed, the others badly crippled.24 She regains consciousness in the hospital and immediately inquires after her children. The nurses on duty are afraid that evasion will only arouse her suspicions and cause mounting anxiety. Ordinary morality might surely condone a lie in this situation. But if the woman insists on a straight answer, the nurses will be placed in a dilemma arising from their fiduciary responsibility not to lie. Only in an extreme case-for example, where they are certain beyond reasonable doubt that severe medical consequences would follow full disclosure-can a lie be justified. In a less clear case- where, for example, they are unable to judge how detrimental to the patient's subsequent recovery full disclosure would be-they are obligated to answer questions truthfully, even if ordinary morality might counsel lying. To do otherwise would jeopardize the fiduciary relationship which under- lies the provision of future care of the patient.