Content uploaded by Michael Davis
Author content
All content in this area was uploaded by Michael Davis on Aug 13, 2016
Content may be subject to copyright.
Authority Revisited: NOMOS XXIX (New York University Press: New York, 1987), pp. 302-337 Michael Davis
THE MORAL AUTHORITY OF A PROFESSIONAL CODE
For many, the claim that a professional code can have moral authority is perplexing. Either
a code is enforceable or it is not. If it is, it certainly has authority but the authority is law-like, that
of a command backed by force, not moral authority. On the other hand, if the code is not
enforceable, there are other problems in explaining its moral authority. It seems that a code must
either state what any decent person knows (for example, "Don't lie"), or seek to impose
obligations beyond what ordinary morality requires (for example, "Don't advertise"), or--most
often--now do one, now another. Insofar as a code simply states what any person knows, there
does not seem to be room for its authority. Its authority for me can begin only where my
knowledge ends. But, insofar as a code goes beyond what everyone knows, whatever authority it
has cannot, it seems, be moral. Morality is a set of rules, principles, or the like binding on rational
beings insofar as they are rational, not insofar as they are lawyers, nurses, engineers, or the like. If
ordinary people can see nothing wrong with, say, advertising, why should the fact that a
professional code says otherwise carry any moral weight?1 The title of this chapter, "the moral
authority of a professional code," seems to be an oxymoron.2
Explaining how a professional code can have moral authority is plainly a philosophical
undertaking. The problem is not so much lack of some fact as an inability to make sense of the
facts we have. Yet the problem is not, or at least not merely, a philosopher's problem. It can easily
end up as part of an important public controversy. Let me give one example.
In the mid 1970s, two young lawyers representing an accused murderer learned from him
where he had left the corpse of a victim with whose murder he had not been charged, a young
woman whose parents were still looking for her. The lawyers' professional code forbade them
knowingly to reveal a confidence or secret of their client. So, when the parents of the woman
came to them, begging for information about their daughter, all the lawyers thought it right to tell
them was that any pertinent information would come out at trial. It did--nine months later--when
their client broke down on the witness stand. The public was appalled. Leaders of public opinion,
especially newspaper editors and politicians, seemed agreed that the two lawyers should have
revealed the information nine months earlier. The bar responded that the lawyers could not, as
lawyers, have done so. Yet the controversy never really became a debate. The bar emphasized the
importance of having a rule of confidentiality. Public opinion was more concerned that lawyers
should think their rule took precedence over ordinary moral considerations.3
2
My subject, the relation of professional ethics to ordinary morality, has received
considerable attention.4 Indeed, a substantial literature has grown up just on the moral justification
of a lawyer's duty of confidentiality.5 Yet I can find nothing treating the subject as one of
explaining the moral authority of a professional code. I find that surprising. Putting the problem as
one of moral authority seems to clarify it and suggests an approach more promising than those
offered till now.
That, anyway, is what I try to show in this chapter. Section I summarizes critically our
understanding of authority. Section II uses that understanding to explain how, and in what sense,
a professional code can have authority. Section III explains how a code's authority can be moral
authority. Section IV distinguishes this explanation from some important alternatives. Section V
concludes with some remarks concerning the relation between professional codes and society.
I. Three Kinds of Authority
Authority is never authority in the abstract. It is always qualified by both subject and
audience. An authority is an authority on something and for someone. I am an authority on logic
for the students of my logic class, but not for anyone who knows much about logic.
Authority can be (merely) apparent ("de facto," "received," or "accepted") or true ("de
jure," "proper," or "real"). An apparent authority is one that, upon closer examination, could turn
out not to have what would provide the basis of true authority (and mere apparent authority does
in fact lack such a basis).6
All authorities, true and apparent alike, "speak" whether they are, like books or omens,
"silent" and so in need of some person to "read" them or, like living experts or courts, literally
speak for themselves. What authorities say may be expressed in "propositions" like "The plural of
genius is genii if what one is referring to is a demon," in "commands" like "Love thy neighbor" or
"Don't smoke," or in less easily defined formulas like "I wouldn't do that if I were you" or "You
ought not to do that." An authoritative expression of the first sort is sometimes described as an
expression of "epistemic authority;" one of the second sort, as an expression of "performatory
authority." Expressions of the third sort seem to be unnamed, but perhaps "advisory" would be
appropriate, since that is their tone.
Writers often describe authority (or, more exactly, an authoritative expression) as "more
than advice but less than command." That description cannot be correct if what is referred to is
the form in which authority expresses itself. Authority can express itself in advice or command as
well as in simple propositions. Usually, however, context makes it clear that what is meant when
3
authority is so described is that, whatever the form in which authority expresses itself, the
expression has a status in our reasoning distinct from that both of "advice" (understood as
guidance that convinces simply because of its content, its plausibility) and of "command"
(understood as a directive to be obeyed merely because its source has power to enforce it).
What then is the status of authority in our reasoning? An "unrecognized authority" is only
a possible authority, much as an unrecognized legal system is only a possible legal system. For
those who recognize an authority, whether true or merely apparent, what it says provides a reason
for believing or doing as the authority says (in part at least) simply because the authority says it.7
An appeal to authority is an appeal to reason (or, at least, to reasons of a certain sort). True
authority differs from mere apparent authority because appeal to true authority provides a good
reason for the belief or action in question, rather than just seeming to. I may be justified in
believing a certain proposition because I have it on good authority (and know that I have). I may
be justified in doing something because I have been so ordered by competent authority (and know
that I have) or so advised by one who knows about such things (and know that I have been so
advised). An authority may give reasons; but if its reasons are themselves decisive, there is no
place for the authority to operate as authority. In this sense, authority does not "advise." An
authority may also threaten as, for example, many suppose the criminal law to do. But if the
threats are themselves decisive, there is again no place for authority to operate. This is the sense
in which authority is said not to "command."
That brings us to the central puzzle about authority: how can one have a rational basis for
a certain belief or action when one does not know enough first - hand to be justified in believing the
proposition or choosing the action and the reasons the authority offers are insufficient, apart from
the appeal to authority, to justify the belief or choice in question? The solution seems to depend a
good deal on the kind of authority in question. We need to distinguish at least three: "material
authority," "formal authority," and "inherent authority." The terminology is new. Its main
advantage is that it is both unfamiliar enough and opaque enough not to suggest conclusions I do
not want to draw. Let us now examine these three kinds of authority, considering how each in
turn can serve as the basis of rational belief or action.
Material authority is a kind of substantiated reliability, a deserving to be believed or
followed that is supported by inductive inference. We are justified in treating what a material
authority says as a reason to believe or act accordingly for the same reason we are justified in
expecting rain when we see dark clouds and lightning. Relying on material authority is often
convenient, but the convenience is itself a function of its reliability, of our being justified in
believing that we are not likely to be better informed if we take the trouble to find out directly or
4
better off if we think things through ourselves.
Material authority can be the reliability of a "black box." We can recognize such an
authority without understanding how it works. For example, I consider the Oxford English
Dictionary an authority on correct formal English. Let us suppose it is. Now, I know practically
nothing about how that dictionary was assembled, the credentials of those who wrote it, or the
like. I have come to accept it as an authority (in part at least) because I have found its description
of formal English usage to be generally accurate.
Material authority can also be like the reliability of a gauge the workings of which we
understand. For example, those who know how to make a good dictionary themselves might still
treat the Oxford as an authority because it is the product of methods they have learned to trust
and because they have not yet duplicated the work that went into making it. The induction
underlying such authority is not an induction on what the Oxford actually says but an induction
from experience with similar "sayings" by things having a structure analogous to it.
Material authority often rests on a mix of those two ways of establishing reliability. We
recognize most authorities in part by their success and in part by their methods. Reliable
procedures (such as "scientific method") promise success. But if, while the procedures seem
good, the results are not, we are justified in giving the authority little credence. Similarly, if, while
the procedures seem unreliable, the results are regularly good, we might be entitled to conclude
that "there must be something to it" and give significant weight to the results (as, for a time, the
police did with the advice of psychics).
Though most examples of material authority seem to be epistemic, not all are. Directives
can prove themselves in much the same way a dictionary can. For example: I may be justified in
now doing as a certain law says, even though its command makes no sense to me, because the law
says to do it. This law has commanded such seemingly senseless things before and each time I
later found what it commanded to be what I would have wanted done if I had been better
informed, less distraught, or more disinterested. The law's past success has made it rational for me
to give any conclusion derived by applying it to particular cases an important place in my
deliberations just because such conclusions derive from that law. It has material authority with
me.
As the examples given so far suggest, material authority is not necessarily the authority of
a person. A law can have material authority with me though I know nothing of its origin or,
indeed, even if I know the legislature that made it to be a club of scoundrels from which this good
law seems to have emerged by accident. I might similarly trust a dictionary to tell me about usage
but not trust its maker on the subject. He might be one of those people who is trustworthy only
5
when what he says is subject to review by his peers. Still, individual persons (for example, a
recognized scholar) or organizations of persons (for example, Amnesty International) can be
material authorities. They have only to be beings we have good reason to believe will be relatively
reliable on the topic in question.
In principle, then, the subject matter of a particular material authority is limited only by the
ignorance of the audience and the ability of the authority to prove itself reliable. An omniscient
and benevolent god could, on this analysis, have material authority on every subject except those
too close to us to leave room for ignorance. In practice, however, because material authority rests
on an inductive argument, the subject of its authority is certain to be quite limited. For example, I
early learned that the Oxford English Dictionary is not a good guide to American slang.
Because material authority depends on an induction and inductions vary in strength,
material authorities should differ in the weight they (justifiably) carry in our deliberation. Some
authorities should outweigh even relatively weighty arguments of other kinds. But no material
authority can outweigh an argument that would otherwise be decisive. If a material authority is
like a gauge used to make measurements that could in principle be made without it, a decisive
argument is like a direct measurement known to be accurate. Because a decisive argument is itself
"knowledge," it forecloses appeal to authority.
It may seem then that while some arguments can preempt material authority, material
authority should not be able to preempt other arguments, only weigh against them. That, in fact, is
the conclusion I wish to draw. But Joseph Raz has argued that all authority (or at least all
nonepistemic authority) is characterized in part by its preempting some arguments that would
otherwise be relevant. If material authority is authority, then (according to Raz) it must make
inappropriate consideration of some arguments that, but for the existence of that authority, would
have been appropriately considered.
Raz gives one example to illustrate that claim. He asks us to imagine a financial expert
known to reach the "right" decision about stocks that Raz should buy in twenty percent more
cases than Raz does when not relying on the expert's advice. "Should I," Raz asks, ". . . when
confronting such decisions, carry on as before but take his advice as a factor counting in favor of
the decision he recommends?" His answer is no: "If another's reasoning is usually better than
mine, then comparing on each occasion our two sets of arguments may help me detect my mistake
and mend my reasoning . . . [but] if neither is sufficient to bring my performance up to the level of
the other person, then my optimific course is to give his decision preemptive force."8 Raz takes
giving "preemptive force" to the authority to be equivalent to denying certain reasons any weight
in the deliberation.
6
Raz is, I think, right that he should do as the analyst recommends rather than try to
compromise between the analyst's recommendation and his own. That, however, is consistent
with the analyst's advice having weight in Raz's reasoning rather than preempting some part of it.
Raz's example seems to support his thesis only because it is too simple to permit us to distinguish
being outweighed from being preempted. We have only to add a little detail to see the distinction.
Consider two variants of Raz's example: in one, Raz's own reasoning merely inclines him
in a direction different from that the analyst recommended; in the other, it is strongly against the
analyst's. Raz would, I think, certainly be entitled (all else equal) to take the analyst's advice with
more assurance in the first case than in the second. That difference in assurance is a measure of
the weight Raz's own reasoning still has in his deliberations.
The question then is whether that weight can manifest itself directly in the decision to do
as the analyst says. The answer, I think, is that it can but that no example is wholly immune from
being interpreted as involving preemption rather than just outweighing. The best we can hope to
provide are examples that make such an interpretation look decidedly strained.
Suppose that Raz decides to test his doubts rather than decide right away. He might check
to see what other respected analysts suggest. If most agree with the first, Raz should certainly go
along with the first's advice. If most of them disagree with the first, Raz probably should not. But
if the authorities are more or less evenly divided, Raz might be entitled to throw his own opinion
into the balance. Whether the decision Raz is making is worth the trouble of getting several
"second opinions" is, of course, another matter. For our purposes, what is important is that taking
such trouble seems neither unusual in a world where experts are willing to give an opinion when
others have already spoken, nor inconsistent with continuing to recognize the first analyst as a
material authority on the very question on which we seek a second opinion or take our own into
account. The most natural interpretation of this example is, I think, that the weight of authority
has reduced the relative weight of other reasons rather than excluding them altogether. The other
reasons can still be decisive when the weight of authority is not overwhelmingly on one side.
Recognizing a particular material authority does not necessarily preempt other reasons, even
those the authority presumably considered.
That brings me to the second kind of authority I wish to distinguish. If we think of
material authority as a proved ability, we should think of formal authority as a conferred ability.
To be a formal authority is to be someone we can be justified in believing or obeying because she
occupies a certain place within a practice, convention, or the like, that place making it rational to
rely on what she says. For example, in baseball the umpire behind the plate is the formal authority
on whether a pitch is a strike or ball. His authority does not rest on a proved ability to make
7
correct calls. He is someone the players must rely on (while he remains their umpire or until the
rules of the game are amended or ignored) because that is how, under the rules, an umpire's call is
supposed to be treated.
Formal authorities, unlike material ones, do not simply weigh against other arguments.
They preempt them, just as Raz says, even some that would otherwise be decisive. Thus, our
umpire's call decides that the players should treat the pitch as a strike even if everyone with a
good view says that the pitch was "in fact" a ball. Formal authorities cannot be weighed against
one another. To be the umpire is to have exclusive authority concerning certain matters, not
simply to be an authority on them. Formal authority is not "authority with", as material authority
is, but "authority over".9
As the umpire example suggests, formal authority can be authority on facts as well as on
what should be done. It may seem, however, that a formal authority should not be believed just
because of its formal authority. We may, for example, go on relying on our umpire even though
we can see that he is "blind as a bat" but (it might seem) we need not therefore give credence to
what he says. We need only act as what he says requires. Formal authority is merely authority
over actions.
That, I think, is the correct conclusion to draw; but, as stated, it is also misleading. Saying
that formal authority is only authority over actions suggests that formal authorities need not have
any material authority, that their authority is entirely a matter of what is conferred on them. But,
except under extraordinary conditions, formal authority presupposes material authority, indeed,
material authority concerning what should be believed. Consider this relatively pure case of formal
authority.
Someone yells "Follow me" in a disorganized crowd trying to escape a fire. Recognizing
her as someone to be obeyed without question may be rational as soon as others seem to be
obeying. We should do as she says because she seems to be the solution to a difficult problem of
coordination. We are likely to prevent each other's escape unless we organize our exit and, for
now, she is the only organization we have. Once some people begin to follow her, asking about
her prior qualifications or the wisdom of her choice of direction would, in the smoke and disorder,
be irrational. That is so not because she is any more likely to be right than anyone else but because
her being followed makes it more likely that doing as she says will get us all safely out than trying
to convince people to follow some other leader, however wise, or letting each follow his own
nose. She seems to be a leader whom we have decisive reason to obey but no reason to believe.
Yet a little reflection shows that not to be so. True, our self-appointed leader has no
special authority apart from her position as leader. Her formal authority is nonetheless not solely a
8
matter of what her position confers. We are justified in following her in part at least because her
apparent likeness to us gives us good reason to believe that she would not knowingly lead us to
our death. Our trust rests on an inductive argument (a relatively weak one given how little we
know of her and situations like this, but still an argument from analogy not unlike those we have
already discussed). So, had our leader said "This is a way out" instead of "Follow me," we would,
it seems to me, have had good reason not only to follow her lead but to believe what she said.
If that is so, our incompetent umpire is not, as he may have seemed, a paradigm of formal
authority. He is much more like a limiting case. The same incompetence that shows his authority
to be clearly not material threatens to make doing as he says irrational. After all, as the would-be
leader example suggests, the justification of formal authority is generally the need to have certain
questions decided so that we can get on with this or that activity. What justifies such authorities is
their usefulness. If a formal authority's decisions threaten the very activity that makes it useful to
have the authority, the justification of the authority is also threatened.
We might call any justification of formal authority resting on the authority's usefulness for
the activity over which it has authority an "internal justification." Not all justifications for formal
authority are of this sort. Sometimes it is rational to recognize an authority for reasons external to
the activity itself. We could have a reason to do what a formal authority asks, for example,
because we had promised to obey it, or because we would do great harm if we did not obey, or
because another formal authority ordered it. Such external justifications can, I think, make it
rational to accept some authorities that an internal justification cannot. But, like internal
justifications, these justifications cannot justify anything whatever. A promise to go along with
our blind umpire's calls may be enough to justify going along with him. But a promise to obey him
no matter what he said on or off the field is probably such an unreasonable promise that it can
justify nothing but its own neglect.10
That brings us to the third kind of authority, inherent authority ("charisma"). This is
certainly the most controversial, a kind of self-evident deserving to be believed or obeyed, a
deserving for which further argument is as impossible as it is unnecessary. Some people just are
natural leaders. Their followers think it obvious that one should do as they say even if the only
obvious merit doing it has is that the leader said to. Similarly, some propositions just seem so
obviously to demand belief that one is obliged to accept whatever follows from them.
Inherent authority is often assimilated to mere de facto authority. It is said to be an
appearance of some other kind of authority, an appearance more information will confirm or
dissolve. This seems to be a mistake. We can, for example, easily imagine someone finding a
proposition like "All events have causes" to be convincing. Further information (a course in
9
quantum mechanics) may eventually lead him to give up the proposition. But giving up a
proposition because of new information is not the same as recognizing that its authority was only
apparent. Even material authority must give way before a decisive argument. Giving up the
proposition is consistent with continuing to recognize in it an attraction that, but for contrary
considerations, would convince. That enduring appeal, if there is one, is what constitutes inherent
authority--or, rather, that is what constitutes inherent authority if it is rational to take such appeal
into account.11
Now, taking such appeal into account seems to me to be rational whenever there is no
better reason for belief or action. I say that not because I think believing or acting is more likely,
all else equal, to lead to truth or good than not believing or not acting (as William James did).12 I
doubt there is a decisive argument one way or the other. I certainly do not have one. And, if I did,
it would destroy my defense of inherent authority. Resting its defense on the likely consequences
of accepting "inherent authority" would amount to making inherent authority much more like
"material authority" than I think it is. So, I do not rest its defense on likely consequences but on
there being no reason to believe, all else equal, that suspension of belief or action is more likely to
lead to truth or good than belief or action is. If suspension of belief or action is not likely to have
better consequences than belief or action, reason cannot, it seems, forbid the belief or action in
question. And that being so, it is rational (reason permits us) to believe as one is inclined to
believe and to act as one is inclined to act.13
Because inherent authority is controversial, perhaps it is worth observing that, while I
have argued that appeal to it can provide a good argument, I have not argued that the argument
so provided is strong. On my analysis, inherent authority can neither preempt other arguments, as
formal authority can, nor outweigh them, as material authority. Its status in argument is like that
of an initial probability assignment in statistics. Accepting it is rational only until the first hard
information gives us more to go on.
II. Which Kind of Authority for a Professional Code?
We must now classify the authority of a professional code; or, rather, we must now
determine which kind of authority we shall take to be primary. The task is one of determining
primary authority because even a cursory examination shows professional codes to be capable of
several kinds of authority. Our concern here is the authority a code has with respect to individual
practitioners, not its authority with respect to their professional organization or those outside the
profession.
10
Let us begin with inherent authority. The American Bar Association's Model Code of
Professional Responsibility, adopted in 1969 and replaced in 1983, provides a good example of a
code claiming inherent authority. The Code is divided into "Canons," "Ethical Considerations,"
and "Disciplinary Rules." The Preamble declares the Canons to be "axiomatic norms expressing in
general terms the . . . concepts from which the Ethical Considerations and Disciplinary Rules are
derived."14 Calling the Canons "axiomatic" suggests their authority is inherent (as does the
absence of any argument for them anywhere in the Code). An examination of the nine Canons
suggests the same. For example, Canon 6 asserts, "A lawyer should represent a client
competently." While not an analytic proposition, it seems undeniable.
Unfortunately, the project of explaining the authority of most professional codes as
primarily inherent does not look promising. Even the ABA's present Model Rules of Professional
Conduct does not fit the pattern. The Rules lack the distinction between Canons, Ethical
Considerations, and Disciplinary Rules. In its place are "Rules," "Comments" which are to help
interpret the Rules, and a "Preamble" which also seems designed to help interpret the Rules.
While the Preamble and Comments do contain most of the former Canons, the Canons are not
distinguished, described as axiomatic, or made the source from which the Rules derive. So,
though the authority of some provisions may still be inherent, analyzing the authority of the
present Rules as primarily inherent seems unpromising.
For somewhat related reasons, we should not try to analyze the authority of the Rules as
primarily material. Because the ABA adopted the Rules only in 1983, their material authority
cannot, even now, rest on much direct experience. Nor can it rest on the Rules' similarity to the
old Code. The Code was discarded only fourteen years after its adoption because it was widely
considered to be seriously flawed both in form and in substance. The new Rules were supposed to
be a new beginning. Insofar as they are, their authority cannot rest on analogy with the Code. The
Rules are also not a return to the ABA's original Canons of Legal Ethics, which were much more
like the Rules' Preamble than like the Rules themselves. In form, the Rules most resemble ordinary
legislation; but, in substance, they are sufficiently different to make arguments from analogy with
ordinary legislation quite weak. So, given the few years in which the new Rules replaced the old
Code as the standard of professional conduct for lawyers in most jurisdictions, it seems unlikely
that the Rules' authority could derive from direct experience with them.
Nor does the authority of the Rules seem to derive primarily from the material authority of
the body proposing them. The committee that drafted them was not well-known even among
those with a strong interest in legal ethics. An anonymous body hardly seems likely to be a
material authority for most lawyers. The only group the drafting committee may have had much
11
authority with was the ABA's legislative body. That is to say, that body probably would not have
adopted the new code if its drafting committee had not so recommended. Resting the authority of
the Rules on the material authority of the ABA's legislative body still does not seem likely to give
the Rules much material authority. That body had not proved itself a reliable judge of codes. The
Rules were, after all, the ABA's third professional code in twenty years.
That leaves formal authority. Formal authority may be justified by either external or
internal considerations. The Rules certainly seem capable of both. Many jurisdictions require law-
yers to take an oath to obey the Rules as a condition of being admitted to practice. In these
jurisdictions, the authority of the Rules could rest on something like a classic contract. In some
jurisdictions, the Rules have been enacted into law, either by the legislature as statute or by the
state's highest court as "rules of court." So, insofar as the law as such has formal authority in a
jurisdiction in which the Rules have been enacted in that way, so must the Rules. And, in most
jurisdictions, the formal authority of the Rules might be traced back to the ABA or state bar
association, both of which would have approved the Rules through their representatives at the
appropriate meeting. Insofar as membership in a voluntary association can bind one to obey the
authorized enactments of that association, the Rules could have formal authority (at least over
that minority of lawyers belonging to the ABA or state bar association) because the Rules had
been so enacted.
I nevertheless hesitate to let the argument for the authority of a professional code rest
primarily on an external justification for two reasons. One is that the authority of the Rules could
be much the same without oath and law. Indeed, the nineteenth century provides some evidence
for that. While states did not generally enact codes of ethics, and the ABA did not exist, learned
individuals did propose standards of conduct that lawyers seem to have generally followed and
held one another to. Two of these proposals, David Hoffman's A Course of Legal Study (1836)
and George Sharswood's A Compend of Lectures on the Aims and Duties of the Profession of
Law (1854), are still cited as material authority on certain questions of legal ethics.
The other reason I hesitate to let the argument for the formal authority of a professional
code rest on an external justification is related to the first. Because external justifications seem to
rest on accidents of history, they seem likely to miss what is "essential" in the authority of any
professional code.
So, we must look for an internal justification of the formal authority of a professional
code. Since the justification is to be internal, we should be able to work it out by determining
what the point is of an activity in which a professional code is central. A good way to start is by
imagining what life among lawyers would be without a professional code and then ask what
12
reason they could have for adding an activity in which such a code is central. Imagining that
should not be hard. Some contemporary occupations, though analogous to lawyering in important
ways, do not have any code beyond what law (including nonprofessional custom), market, and
ordinary morality provide. Ordinary business managers are a good example of such a
"nonprofessional" occupation. What benefits do lawyers get by having a professional code that
business managers do not get without one?
Consider what happens when lawyer Brown answers lawyer Jones' question concerning a
case in which they are on opposite sides. Should Jones trust the answer she gets from Brown?
Under Rule 8.4 of the Model Rules, a lawyer acts unprofessionally if he engages "in conduct
involving dishonesty, fraud, deceit, or misrepresentation." Absent such a professional standard,
the relations between Jones and Brown would be governed only by those standards governing
relations between any two strangers whose interests are adverse in some important way. Jones
and Brown need only meet the minimum standard of conduct set by law, market, and ordinary
morality. These may allow misleading silences, "puffery," and the like. Only if Jones and Brown
expect to have frequent dealings with one another might they seek to establish a more demanding
relation. And doing that would require some time, explicitness, and risk.
Absent a professional standard, professionals seem to have that special problem of
coordination Thomas Hobbes (unfortunately) labeled "the state of nature" and game theorists
(even more unfortunately) have labeled "the prisoners' dilemma." Both Jones and Brown would be
better off in the long run if each could justifiably expect the other to be guided by a higher
standard than law, market, and ordinary morality impose. But each would be worse off than she is
now without such a code if she acted according to that higher standard while the other did not.
Absent a professional code, Jones and Brown cannot even know whether they agree on what the
higher standard is. Misunderstanding, and the suspicion of misunderstanding, would threaten
whatever coordination they could manage or force them to rely on such expensive coordinating
procedures as a legally enforceable contract. Each is likely to make the other worse off than she
would have been with the cooperation of the other.15
Coordination problems are not hard to solve when relatively few people are involved,
relations are continuing, no single transaction is of overwhelming importance, and each has good
information about what others are doing. Perhaps these conditions were approximated for lawyers
in the United States during much of the nineteenth century. That would certainly explain why the
profession of law could so long remain primarily an activity of unorganized individuals for whom
writers like Hoffman and Sharswood were the authorities on professional conduct.16 In the small
towns of nineteenth-century America, individual lawyers might each do as Hoffman or Sharswood
13
said they should, reasonably sure the other lawyers were doing the same. They might be
reasonably sure of that in part because the excellent communications of a small town made it easy
to learn whether most lawyers were doing as they should most of the time. Those same
communications would also have made it easy for each lawyer to assess the character of the
others, and so easy to make a reliable judgment about whether the others could be trusted where
they could not be watched.
But, by the end of the nineteenth century, lawyering was becoming a big-city occupation.
Thousands of lawyers might work in the same locale. Transactions were often not repeated. Some
transactions were much more important than others. And information was likely to be harder to
get. By the end of the nineteenth century, lawyers had a much harder coordination problem. The
only workable solution (apart from leaving everything to the market) was what Hobbes called "a
sovereign," that is, an institution that both lays down standards of conduct and makes it
reasonable for each to adhere to them.
One form of that solution is simply to have the government pass laws governing the
relations among lawyers and use ordinary civil or criminal procedures to get compliance. Another
is for lawyers to form an organized profession to set standards and then get compliance some way
or other. The American solution has been to combine these. In most jurisdictions, the ABA's
model code will be adopted by the state legislature or supreme court (more or less as the ABA
recommended), but obtaining compliance will be left in large part to state or local bar associations
(which, in most states, are private voluntary associations). Bar grievance committees will have the
power to investigate an allegation of unprofessional conduct, whether committed by a member of
their association or not, to recommend discipline where that seems appropriate, and to forward
any recommendation of serious discipline to the state supreme court, which retains the ultimate
power to punish but in general follows bar recommendations.
Other means are also available to a professional organization to help get compliance. One
way is screening for what has come to be known as "character and fitness." Insofar as an
individual's past conduct and reputation indicate ability to live up to a professional code, a
profession should be able to increase compliance by barring those whose conduct shows inability
to comply.
Another way to get better compliance is to provide interpretations of the code. If a lawyer
wants to do the right thing but cannot tell from the code what that is, he can put a question to the
local, state, or ABA ethics committee. The committee will then give him an opinion. Such
opinions make it unnecessary for disagreements between lawyers to arise just because they cannot
agree on what the code means. Many of these opinions are then published, making it unnecessary
14
for the committee to answer the same question again.
Our analysis suggests that a professional code is central to solving a coordination problem.
The members of a particular profession trade certain liberties others have for the advantage of
being regulated by their own code. This may seem an oddly inward-looking view of professions,
one that largely ignores the people professions are supposed to serve. This view is not, I think, as
odd or inward-looking as it may seem--so long as we suppose professionals in general to believe
that being able to satisfy a certain moral ideal is itself an advantage to be aimed at. That does not
strike me as an unlikely supposition. One way we have of distinguishing professional
organizations from trade unions or the like (even when, as with teachers, many of the same people
belong to both organizations) is by the professional organization's express commitment to serving
a certain moral ideal as its primary purpose. Absent proof of widespread hypocrisy, we must
accept organizational commitment as an expression of widespread individual commitment.
The following seems to be a rough summary of the declared purposes of some recognized
professions. Curing the sick, easing the pain of the dying, and protecting patients from disease is
the purpose of private medicine; protecting the public from disease, of public health; educating the
young, of teaching; designing safe and useful structures, of civil engineering; reporting financial
information in a useful and accurate form, of accounting; and informing the public of important
events, of journalism. The profession of law also seems to be defined in part by an ideal, that of
providing all who seek justice within the law with the help they need to get it.
Such idealistic declarations are not mere lip service. For example: No ordinary business
person considers it a duty to provide her services free to those who need it and cannot pay. Nor
do we expect her to. When a business does serve the needy gratis, it considers the service an act
of private charity or civic responsibility deserving special praise (as indeed it does). In contrast,
any lawyer can be asked to provide free legal services to those who cannot pay. To refuse without
good cause would be a serious violation of professional ethics; but to perform is to deserve no
more praise than anyone else deserves for doing his duty.
This sort of example does not have its counterpart in every profession. Some professions,
such as medicine, teaching, and social work, seem, like law, to understand membership in their
respective professions to include a duty to serve even when the patient, student, or client cannot
pay. Other professions, like engineering or accounting, seem to recognize no such duty. But even
they require their members to do things for the public good that we would not expect of ordinary
business people. For example, an engineer should refuse a commission if the client intends to build
a structure that, though technically within the law, would pose a serious risk to public safety.
Public service can be part of the job only so long as some institution makes it that. If some
15
professionals do their part while others do not, those who do theirs will have higher overhead
than those who do not. They will either have to live more cheaply or market their services at a
higher price. Those who live up to the ideal would then be at a competitive disadvantage. Absent
a saintly commitment, they could not rationally go on doing their part. Individual professionals
deserve less praise for serving their chosen ideal than individual business people because their
profession assures that they, unlike business people, do not have a free-rider problem.
On this analysis, a professional code is primarily a formal authority because it solves a
certain coordination problem, the problem of how to pursue, without unnecessary hardship, the
ideal that distinguishes the profession.
III. From Formal Authority to Moral Authority
I have so far argued only that a professional code can have formal authority over the
members of the profession in question and that that authority can rest on an internal justification. I
have not argued that this authority can be moral authority. I must now do that. I shall argue that
whenever the ideal around which a profession is organized is a moral ideal and certain other
reasonable conditions are met, the professional code has formal moral authority.
By "ideal" I mean a possible but not ordinary state of affairs that is good without
qualification, whether good as means or end, whether a great or small good. An ideal may be
thought of as the natural end of a certain process of improvement. Though ideals are often
difficult to achieve, they need not be impossible. A certain room provides ideal working
conditions if conditions in it are unusually good and no change could make them better.
Ideals are moral only if they meet three conditions: First, achieving the ideal must be
possible without doing anything morally wrong. Second, failing to achieve the ideal must also not
be morally wrong. And third, the achieved ideal must be a state of affairs that (all else equal) any
rational person would favor (over alternative nonideal states) even if favoring it means some
restriction on what he or others could otherwise do.
An "ideal" that fails to satisfy the first condition would be morally flawed; an "ideal" that
failed to satisfy the second would be part of minimally decent conduct; but an ideal that failed only
the third condition would be a nonmoral ideal, one toward which a moral agent could rationally
be indifferent.
Consider the ideal of a torturer. Such an ideal is not impossible. The ideal torturer would
presumably be able to empathize with his victims enough to detect weakness but not enough to be
bothered by their suffering. He would be capable of any moral outrage serving his employer.
16
And so on. Still, the ideal torturer is, though a possible ideal, not a moral one. Morality plainly
forbids torturing (except perhaps under extraordinary circumstances).
Because torturing is morally wrong, abstaining from torture cannot be a moral ideal any
more than ideal torturing can. If torturing is morally wrong, then abstaining from torture is simply
doing what ordinary moral decency requires, not (according to morality) achieving an ideal.
Perfect baseball playing, on the other hand, is merely a nonmoral ideal. Nothing about
perfect baseball makes it rational for a moral agent to favor anyone's playing well, badly, or not at
all. Only if one has an interest in baseball to begin with would it be rational to favor his or others'
pursuing the ideal of perfect playing. And nothing in the concept of moral agent requires such an
interest.
Behind this conception of a moral ideal is a conception of rational goods. Though all
ideals are rational guides to action in the sense that pursuing what one considers good is rational
(all else equal), some ideals are also rational in the stronger sense that the supposed good would
actually be good for the one pursuing it. We might call these "prudential goods." Certain
prudential goods are rational in an even stronger sense, that is, they are (all else equal) what any
rational person has an interest in. Among these, presumably, are health, a good education, safe
and useful structures, accurate financial information, knowledge of important events, and justice.
Pursuing these--which, following John Rawls, we might call "primary goods"--is, however, still
not necessarily pursuing a moral ideal. Moral ideals have a connection with morality that a
particular primary good need not have.
We might usefully picture this connection by thinking of morality as the work of a "moral
legislature" consisting of all rational persons laying down rules binding on all.17 Achieving the
desired distribution of certain primary goods may be possible without adopting any particular rule.
These distributions, though they can be ideals, cannot be moral ideals. Other primary goods may
be such that achieving the desired distribution, while possible only by adopting a rule, is best
approached by adopting a rule expressly requiring the desired distribution. This distribution of
primary goods would be a moral requirement, not an ideal. Still other primary goods may be such
that the desired distribution cannot be directly legislated. For example, the necessary rule might
impose too great a burden, or something about the good itself might make requiring the
distribution self- defeating (as requiring everyone to trust everyone else equally might be). The
desired distribution of such goods may still be approximated by adopting auxiliary rules, that is,
rules making it easier than it would otherwise be to engage in activity tending to produce the
desired distribution. For such goods, the desired distribution is a moral ideal. The connection
between the rule and the ideal provides a reason for making the rule part of morality.
17
On this analysis, a moral ideal is a distribution of goods every rational person might want
enough to accept a significant moral burden but not enough to accept the burden that morally
requiring the good to be provided would entail. So, for example, achieving good health for
everyone is a moral ideal in this sense. Health is certainly a primary good. Yet we are unwilling to
require each of us to help the sick, to avoid all conduct that might cause disease, and to do
whatever else might be necessary to provide that good for everyone. The moral rules therefore
include no requirement that each do what he can to assure his own health or anyone else's, only
such auxiliary rules as that against maiming others.
The moral authority of professional codes seems to rest on such an auxiliary rule. We
might state it briefly as: Obey your professions code. The rule's justification is packed into the
three words "your profession's code." We need to unpack all of them to see what rational persons
would gain from the rule. Since we defined "profession" in Chapter 1, we need only unpack
"your" and "code" here.
"Your" indicates membership. "Your profession" is the one you are a member of. Given
our understanding of profession, we have at least three criteria of membership: (a) being
considered to be engaged in serving a certain moral ideal (in a certain way), (b) claiming to be
someone engaged in serving it (in that way), and (c) benefiting from so claiming.
These criteria are not necessarily independent. They are also not necessary or sufficient for
membership in a profession. We can certainly imagine cases in which someone benefited from
membership even though she did not claim it. ("She just must be a lawyer, whatever she says.")
Such a person at least seems someone who might be held to the code of "her profession" and so
to our proposed rule. We can equally well imagine someone who, though satisfying all three
conditions, was technically not a member of the profession because no one noticed that she failed
to satisfy some requirement, for example, having filed the appropriate papers. She might not be
subject to her profession's code simply because the code expressly excluded persons who failed to
satisfy such technical conditions.
These cases seem to depend on the existence of formalized procedures of admission about
which our criteria say nothing. That is as it should be. For professions as we know them, our
imagined cases will be exceptional. One cannot, for example, become a lawyer without state
recognition and the state will generally not recognize someone as a lawyer without proof both
that he intends to serve the appropriate ideal and that he has the ability to do it. Where a
candidate meets these substantive requirements, admissions committees, or the courts, can usually
cook up a way to cure any merely technical failing. Actual procedures tend to approximate the
ideal, that is, tend to pick out just those people who will serve the appropriate moral ideal. But, as
18
in most formal procedures, form can sometimes get in the way of substance. Our three criteria (a-
c) should therefore be read as defining only the central case of membership in a profession, the
one that following actual procedures of admission and retention should realize as often as feasible.
That brings us to the last of the terms needing to be unpacked. By "profession's code," I
mean those rules that are (a) justifiably thought to govern members of the profession in question
because they are members and (b) reasonably related to serving the moral ideal the profession is
supposed to serve. The code may have grown up as custom, perhaps with someone like Hoffman
or Sharswood giving it authoritative form, been imposed by some organization outside the group,
for example, a state legislature, or been the conscious work of the profession itself or of an
organized part. What makes it the profession's code is simply that people (both members of the
group and nonmembers) are justified in expecting members to act as if the code generally guides
their conduct. Ordinarily, people will be so justified because members of the profession generally
give the code lip service and generally seem to act as the code requires. A code's being a certain
profession's is a reasonable inference from such ordinary facts.
Given this understanding of "your profession's code," we can easily see why every rational
person should want something like the proposed rule among those governing the conduct of
everyone else, even if that means having to obey it too. The rule is likely to benefit everyone by
facilitating provision of a primary good for everyone, while leaving all those who do not claim to
provide that good as free of restraint as if the rule had not been adopted.
The rule facilitates provision of a primary good for everyone by giving moral support to
those attempting to provide the good through their coordinated work. Those who belong to the
profession are morally obliged to act according to rules which, being reasonably related to
achieving the ideal in question and justifiably thought to govern the conduct of members,
constitute a solution to the problem of coordinating relations among them in a way permitting
them all to serve their common ideal without undue cost.
The proposed rule nevertheless leaves all who do not want its protection as free as before
because, to come under a particular professional code, one must claim membership in that
profession. To be as free as before, one need only avoid doing what would make one a member of
a profession.
Now, it might be objected that professional codes often leave nonprofessionals less free
than before. For many professions, not being a member means not being legally able to do what
members can legally do (or at least not being able to do it for pay). Though that is true, it is not
relevant to the argument for giving moral support to the professional code.
Normally, the law must grant a profession a "monopoly" if there is to be one. A
19
professional code cannot grant it because (by definition) such a code applies only to members of
the profession and the monopoly requires restraining nonmembers. Someone who engages in the
unauthorized practice of the profession does not engage in unprofessional conduct. Of course,
professionals are likely to think serving their favored moral ideal includes protecting the unwary
from the incompetent. A professional code may even include provisions requiring members to
work for legislation giving their profession an exclusive right to serve the ideal. The code may
also bar members from cooperating with nonmembers performing the same service. But all that is
consistent with the code itself restraining only members, leaving all nonmembers as free as they
would be if the code had no formal moral authority.
"Obey your profession's code," when understood as I propose, seems to be a special case
of what is now commonly called the principle of fairness ("Don't cheat"). That principle has
received a good deal of criticism. So, perhaps it is worth pointing out that in two respects at least
the rule requiring obedience to one's professional code is significantly more plausible than the
most plausible general versions of the principle of fairness.18
First, the proposed rule concerns not all just cooperative activity or even all just mutually-
advantageous cooperative activity, but only cooperative activity (obeying the code) reasonably
related to serving a moral ideal. If any just cooperative activity providing us with a net benefit can
have a moral claim on our cooperation, one serving a moral ideal should.
Second, the proposed rule does not, as the principle of fairness would, govern all those
who benefit overall from the professional activity, or even all those who, like clients, voluntarily
accept benefits, but only those who claim membership in the profession and benefit from so
claiming. The proposed rule resembles the principle of consent in making formal authority over a
person contingent on an act both voluntary and easy to avoid. That resemblance is certainly part
of its appeal. Still, the proposed rule is not merely a disguised version of the principle of consent.
The voluntary act by which one comes under the authority of a professional code is independent
of promising, vowing, giving one's bond, or any other convention of consenting. Even in a society
that had no convention by which one could pledge one's future conduct, one could still claim to
serve this or that moral ideal and benefit from so claiming. Claiming membership in a profession in
that way under the appropriate circumstances is, though not itself consent, its moral equivalent.
Given these two respects in which the proposed rule is more plausible than the principle of
fairness, little of the criticism of that principle could give us reason to reject the proposed rule. On
the other hand, given that the proposed rule is an especially plausible version of the principle of
fairness, almost any defense of that principle should also be a defense of the proposed rule. I
therefore take the rule's moral status to be relatively uncontroversial. "Obey your profession's
20
code" is a moral rule (a special case of "Don't cheat").
IV. How Good an Explanation is This?
We have explained the (formal, internal) moral authority of a professional code as a
function of membership in a profession and that membership as a function of openly serving a
moral ideal in cooperation with others. Because serving a particular moral ideal is something any
of us can do, we have, like Charles Fried,19 located the authority of one's professional code in
conduct (serving a moral ideal) both professionals and nonprofessionals can understand. We have,
however, departed from Fried by stressing a relation among professionals rather than the relation
between professional and client. For us, being a professional is less like being the client's friend
than Fried makes it seem. This, I think, is a difference in our favor. Some professions, like public
health, seem to lack the personal relation making analogy with friendship easy, while others, like
auditing, seem to direct the professional's loyalty away from those with whom he has personal
relations. Even some lawyers, for example, those advising a government or large corporation,
may find Fried's analogy with friendship misleading.
We have also departed from Fried in treating membership in the profession rather than
taking on a client as central. For us, a professional's first loyalty must be to the profession, not to
clients. Service is a by-product of that membership, not membership a by-product of service. This
difference between us and Fried, though perhaps only one of emphasis, seems important. Fried
has some difficulty with the commanded "friendship" that occurs, for example, when a judge
orders a lawyer to represent a poor client. Indeed, it seems he must regard such a command as an
infringement of the professional's "moral autonomy" (that is, of the professional's right to choose
whom to serve.) For us, in contrast, to join a profession is to come under the formal moral
authority of its code and, if the code provides, to serve the poor. While one remains a lawyer, one
is not as free as one would otherwise be to decline to help those seeking justice within the law. A
professional is, on our analysis, not so much a client's "special purpose friend" as a person doing
her moral duty (though a duty voluntarily acquired by joining the profession). She exercises her
moral autonomy by performing that duty as it should be performed.20
This explanation of professional ethics may seem too close to a view Robert Veatch
satirized as "the code of British gentlemen, philanthropically bestowing benefits on patients who
ought to show gratitude[,] rather than duties negotiated mutually between physicians and
patients."21 Our explanation certainly belongs to the same family as the "gentlemen's code." Still,
even brief examination will show our explanation to be free of the serious faults Veatch attributes
21
to that Victorian conception of professional ethics. Our explanation leaves room for all of the
negotiation Veatch thinks should go on between professional and client, for example, over
services to be performed and price to be paid. All our explanation excludes is professional and
client negotiating to have the professional do what members of that profession refrain from doing
(in part at least) because each is reasonably sure other members of their profession will refrain as
well as part of cooperating to serve their common moral ideal. I see no fault in such gentility.
I also see no fault in thinking of a profession as a kind of philanthropy. Some professions,
especially medicine and law, are so profitable that one easily forgets "the love of man", that is, the
service of a moral ideal, that entering even such a profession entails. We are likely to think that
the financial rewards are so great that such professionals do not deserve our gratitude as a further
reward. Perhaps they do not. These examples are, however, not decisive. Other professions, for
example, teaching or nursing, remind us that one can still practice a profession while earning less
than a plumber. Members of these professions do seem to earn our gratitude in something like the
way Veatch seems to think they cannot. Perhaps Veatch's understanding of professional ethics
suffers here from too much attention to medicine. I think it a strength of our alternative that it
reminds us of professions differing in this respect from both medicine and law.
Because humans lack "world enough, and time," serving one moral ideal is necessarily to
neglect others. Indeed, given the way moral ideals jostle one another in this less than ideal world,
we must recognize that someone's serving one ideal may mean interfering with the achievement of
others. The lawyer who helps acquit a child molester of a theft he did not commit thereby risks
the health of any child her client might then molest. Yet we need not think of professionals, as
Benjamin Freedman does, as exhibiting "fanatical adherence to an idea."22
"Fanatical adherence" at least suggests that the professional would let nothing stand in the
way of serving her chosen ideal, not even the commands of ordinary morality. Our explanation of
professional ethics makes no such suggestion. We have defined professions as serving moral
ideals, not just any idea. A moral ideal is, by definition, one that can be served without doing
anything morally wrong. So, fanaticism is not necessary to serve a moral ideal. We have also
required the content of a profession's code to be reasonably related to serving the appropriate
ideal. If moral ideals can be served without doing anything morally wrong, a particular provision
of a code could not both require something morally wrong and be reasonably related to serving
the appropriate moral ideal. So, on our analysis, fanaticism seems to be ruled out of professional
ethics.
Why then have some writers, including Freedman, said the opposite? The answer, I think,
is that they have confused a code's requiring immoral conduct (which none do) with one of three
22
unproblematic states. They have missed at least one of three important distinctions, suggesting
that what lies behind their view is an impoverished moral theory.
One distinction often missed is that between what we might call "basic morality"--what we
are morally required to do or not to do--and "extended morality"--what we should do or should
not do. Basic morality is the domain of moral right and wrong, of the moral and immoral;
extended morality, of moral good and bad, of mere moral virtue or vice. We are morally required
not to kill (with certain exceptions), but we are not morally required to give some of our surplus
to the poor. Giving to the poor is, however, something we should do. Both not killing and charity
are part of ordinary morality (decent conduct), but not killing, like obeying one's professional
code, is part of basic morality, while giving charity is, like serving some moral ideal or other, part
of extended morality.
Many examples of a professional code allegedly requiring one to do something against
ordinary morality are in fact only examples of a code requiring something "against" extended
morality. In this category belongs the situation of our two lawyers. Certainly, they should have
told the parents where the body was if their code allowed it. But it did not and, assuming the code
was reasonably related to serving the appropriate moral ideal, morality required them to do as it
says. So, not telling the woman's parents where her body was could not be immoral. The situation
was much like that you would be in if you refused to give a poor man the money in your wallet
because you had already promised it to a friend to help expand her business. A prior obligation
can make doubtful what would otherwise be the only decent thing.23
That brings me to the second distinction often missed. Some examples of a professional
code allegedly requiring immoral conduct are analogous to examples philosophers use to show
that moral rules have exceptions. They are, that is, situations involving an enormous disproportion
between the good to be achieved by following the rule and the good to be achieved by breaking it.
The classic example is that of choosing between breaking a trivial promise to meet someone and
saving a drowning child. Such situations are not examples of the immorality of some moral rule,
for example, "Keep your promises," or of the particular promise, for example, "I'll meet you at ten
sharp." They are, rather, simply situations in which breaking the promise is at least excusable. The
duty the moral rule imposes remains (and so, in a sense, the rule remains exceptionless). For
example, even if you break a trivial promise to me in order to save the child, you owe me some
recompense, at the least an apology for not showing.
The situation of our two lawyers is similar. Their client would have suffered little harm
had they surreptitiously informed the prosecutor where the body was (at most, another charge of
murder). On the other hand, the victim's parents, like the drowning child, would have been saved
23
great harm. This disproportion between the harm avoided by not disclosing and the harm allowed
seems to have been what most concerned the public. The bar's response was to defend the general
rule of confidentiality. The bar seems never to have considered that it might also need to defend
obeying an admittedly justified rule in the extraordinary situation that concerned the public.
The bar may have been right about the merits of the general rule. Certainly, meditation on
cases such as this did not lead the ABA to revise the rule as one might expect. The Model Rules
actually reduces somewhat the number of situations that, under the old Code, permitted
disclosure.24 But what is important here, and what was important in the situation the lawyers
faced, is that showing a general rule to be justified does not necessarily justify doing as the rule
says. Extraordinary situations at least raise the possibility of justified departures from even the
most justified rules.25 The controversy over what the two lawyers did never really became a
debate because the bar simply did not discuss that possibility.
That brings me to the third distinction often missed. Some (supposed) examples of a code
requiring (or, at least, allowing) immoral conduct seem to confuse a state of affairs in which the
existence of a profession makes it morally wrong to do something if one is not a member, with the
state of affairs in which doing the act is morally wrong whether or not the situation includes the
appropriate profession. The underlying moral rule is something like "Don't engage in conduct that
is unreasonably dangerous." What conduct is unreasonably dangerous depends in part on the
situation, but the existence of a safer alternative always makes an otherwise reasonable alternative
unreasonable. So, for example, morality permits a surgeon to remove shrapnel from a brain even
though it forbids me. That is not because surgeons are free from some rule of basic morality while
I am not, only that the circumstances make their poking about in someone's brain not
unreasonably dangerous. Under the circumstances, it is better, though not much better, that I try
to save his life, because the only alternative, doing nothing, is worse. If no surgeon were available
and removing the shrapnel now were the only way to save the poor fellow's life, I might be
justified in trying to remove the shrapnel. Under the circumstances, my poking about would not
be unreasonably dangerous. Differences between what morality permits professionals and
nonprofessionals need not show that professionals are exempt from any moral rule governing the
rest of us, only that circumstances can change what the moral rules require or permit us to do.
Those are the three distinctions I think we need to make when considering alleged
examples of professional codes requiring immoral conduct. Once they are made, the claim that no
code can contain an immoral provision should seem much less daring. The claim may, however,
still seem too daring. After all, can we not imagine a code containing an immoral provision?
The answer to that question is yes and no. Yes, it is possible (in a sense) to imagine such a
24
code. Any particular code can, as a matter of brute fact, include such a provision. Yet this
possibility is morally irrelevant. When an immoral provision appears in an actual code, it is,
strictly speaking, not part of it (not, that is, what "Obey your profession's code" commands
obedience to). A professional code must by definition be reasonably related to serving the
appropriate moral ideal. A code is not reasonably related to serving the appropriate moral ideal
unless every element is so related. Since an immoral provision cannot be so related, the code as a
whole is reasonably related to serving the appropriate moral ideal only if immoral provisions are
considered void (or just outside the code)--just as a promise to do what is clearly immoral is,
though a promise, morally stillborn.26
That answer may leave a residual doubt. Surely, it may be said, an entire code could
somehow be reasonably related to serving a moral ideal even if, and perhaps just because, some
provision requires immoral conduct. Since I know of no examples, the only way I can think to
dispose of this residual doubt is to restate the answer already given in terms of authority. As we
have understood formal authority, it is a means to an end. The appeal to authority is an appeal to
reason because, and only so long as, we have good reason to believe doing as the authority says is
serving the end in view. So, for example, if the would-be leader of our earlier example were
leading us into the flames rather than away, her formal authority would only be apparent. As soon
as we saw what she was doing, we would be justified in disobeying her even if disobeying meant
throwing the crowd back into disorder. Similarly, once we see that a provision of the code is
unconnected with serving the appropriate ideal, we are no longer internally justified in recognizing
it as a formal moral authority. We have discovered that its internal formal authority is only
apparent.
One may, of course, still have moral reasons to go along even with an immoral provision,
for example, because of the harm disobeying it would do the innocent. But all such reasons will be
external, contingent, at least partially undercut by contrary considerations, something quite
different from the reason one has when a provision, being reasonably related to serving the
appropriate ideal, speaks with the full moral authority of one's profession.27
The term "reasonably related" does not, of course, entail that there is only one morally
permissible way to serve the moral ideal in question, which the code of ethics merely puts in
writing. On the contrary, a distinctive feature of codes is that they involve a choice among
morally permissible strategies for serving the moral ideal in question. There is always an element
of convention, a point I shall expand on in the next three chapters.
V. Concluding Remarks
25
On this analysis, the formal authority of a professional code is by definition inherently
parochial. Only a professional is subject to the code's formal authority. While ordinary moral
agents can understand why a professional has a special moral obligation that most moral agents
do not, they are not, for that reason, bound to do anything they would not be bound to do
anyway. A society need not, for example, guarantee a profession a monopoly anymore than it
need provide for the enforcement of certain promises as contracts. Granting special powers to a
well-regulated profession may, of course, be reasonable. But ultimately, the claim a profession has
on society must rest not on the profession's status as a profession but on the good its members do.
A professional code is, at best, part of what can make a profession deserve the trust on which
society's support should rest.
26
NOTES
I presented an early draft of the first half of this paper (in absentia) at IIT's Center for the Study of
Ethics in the Professions on December 2, 1985. I should like to thank those present, especially
Robert Ladenson, Fay Sawyier and Vivian Weil, for transmitting a number of useful comments to
one forced by a negligent baby-sitter to miss his own paper. I should also like to thank the
members of Russell Hardin's Workshop on Ethics and Public Policy, January 27, 1986, for a lively
discussion of the penultimate draft of what then appeared in Authority Revisited: NOMOS
XXXIX, J. Roland Pennock and John W. Chapman, eds. (New York: New York University
Press, 1987), pp. 302-337 (after significant changes suggested by those two august editors). This
chapter has also benefited from the intervening years.
1.For the conventional answer to this question, see, for example, Edmund G. Seebauer and Robert
L. Barry, Fundamentals of Ethics for Scientists and Engineers (Oxford University Press: New York,
2001), p. 9: "Such codes remind society members of the high ethical standards expected in the work
place. Also, codes lay out those standards to new workers who have little experience. Finally, as public
documents, codes help professional societies take formal or legal disciplinary action against flagrant
violators." All this, though true, ignores what I shall argue is the central function of such codes, to set
those "high ethical standards" in the first place.
2.For two extreme versions of this argument, though ones not cast in terms of moral authority,
see: Philip Shuchman, "Ethics and Legal Ethics: The Propriety of the Canons as a Group Moral Code,"
George Washington Law Review 37 (December 1976): 244-269; and John Ladd, "Collective and
Individual Responsibility in Engineering: Some Questions", in Beyond Whistleblowing, Vivian Weil, ed.
(Illinois Institute of Technology: Chicago, 1983), pp. 90-113.
3.If I am to judge from my own experience teaching professional responsibility to senior law
students at the time, even many lawyers must have found the bar's response inadequate. My students
knew what they were supposed do as lawyers. They also knew what they should do if they were not
lawyers. What they did not know was what to do as ordinary decent persons who were also lawyers.
And that, of course, was the question. For more of the excruciating details of this case together with a
good example of the lawyer's defense, see Michael S. Callahan and Hal C. Pitkow, "The Propriety of the
Attorneys' Actions in the Lake Pleasant Case," in Lawyers' Ethics, ed. Allan Gerson (New Brunswick,
NJ: Transaction Books, 1980), pp 156-181.
4.See, for example, Monroe H. Freedman, "Personal Responsibility in a Professional System,"
Catholic University Law Review 27 (Winter 1978): 191-205; Thomas L. Shaffer. "Advocacy as Moral
Discourse," North Carolina Law Review 57 (May 1979): 647-670; Paul R. Camenisch, "On Being a
Professional, Morally Speaking," in Moral Responsibility and the Professions, eds. Bernard Baumrin
and Benjamin Freedman (New York: Haven Publications, 1983), pp. 42-60: and Alan Gewirth,
"Professional Ethics: The Separatist Thesis," Ethics 96 (January 1986): 282-300; as well as the works
cited in notes 17-19 and 24 below.
5.See, for example, Monroe H. Freedman, "Professional Responsibility of the Criminal Defense
Lawyer: The Three Hardest Questions," Michigan Law Review 64 (June 1966): 1469-1484; Robert P.
Lawry, "Lying, Confidentiality, and the Adversary System of Justice," Utah Law, Review 1977 (1977):
653-695; Charles W. Wolfram, "Client Perjury," California Law Review 50 (July 1977): 809-870;
Bruce M. Landesman, "Confidentiality and the Lawyer-Client Relationship," Utah Law Review 1980
(1980): 765-786; Robert Sampson, "Client Perjury: Truth, Autonomy, and the Criminal Defense
Lawyer," American Journal of Criminal Law 9 (November 1981): 387-403: and David Luban, "The
Adversary System Excuse," in The Good Lawyer: Lawyers' Roles and Lawyers' Ethics, ed. David
Luban (Totowa, NJ: Rowman & Allanheld, 1984), pp. 83-122.
6.Cf. E.D. Watt, Authority (New York: St. Martin's Press, 1982), p. 30: "If . . . [legitimate or
approved authority] means no more than accepted, then it is hard to see what it adds to this [de facto]
meaning of the word authority, which as we have seen, has the notion of acceptance built into it; if, on
the other hand, it means justified, then the enterprise of considering authority de facto, is at an end . . ."
Since our concern is not with what might make a professional code merely appear to be a moral
authority but with what could make it truly so, we need say little more about "apparent authority."
7.Cf. Richard T. De George, "The Nature and Function of Epistemic Authority," in Authority: A
Philosophical Analysis, ed. R. Baine Harris (University, AL: University of Alabama Press, 1976), p. 81:
"Involved in the very notion of [epistemic] authority is y's holding [proposition] p to be true or more
probably true than before x enunciated it."
8.Joseph Raz, "Authority and Justification," Philosophy & Public Affairs 14 (Winter 1985): 3-29.
Quotation, pp. 28-29.
9.The distinction between material and formal authority as I have drawn it bears a certain
resemblance to the distinction between legal "rules" and "principles" that Ronald Dworkin makes in
Taking Rights Seriously (Cambridge: Harvard University Press, 1977), especially pp. 22-39. I don't
think that's accidental. Rules, as Dworkin understands them, seem to have formal authority; principles,
to develop material authority by proving themselves case after case.
10 .See Raz, "Authority and Justification", especially pp. 18-22, for a good discussion of this point.
11 .Cf. Watt, Authority, p. 39: "It might seem that the kind of authority called charismatic would be
difficult to fit into this insistence on the subordination of authority to reasoning or rules . . . . But even
here the priority of rules or reasons over authority holds . . . . It may be thought inappropriate, even
improper, to ask an exceptional person to give reasons for his pronouncements; but he could not be
recognized as such without reasons of some kind. There must be some way of identifying an
authoritative utterance . . ."
12 .William James, "The Will to Believe", in Essays on Faith and Morals, selected by Ralph Barton
Perry (World Punishing Company: New York, 1962), pp. 32-62.
13 .For a more extensive discussion of the related question of the rationality of treating one's
emotions as reasons, see my "Interested Vegetables, Rational Emotions, and Moral Status,"
Philosophical Research Archives 11 (March 1983): 531-550.
14 .American Bar Association, Code of Professional Responsibility (Chicago: National Center for
Professional Responsibility, 1980).
15 .For a useful discussion of coordinating problems in general, see David Lewis, Conventions
(Cambridge: Harvard University Press, 1969), especially pp 5-36. I am not, however, suggesting that a
professional code is a pure convention in Lewis's sense. The exact provisions of a code may make
important differences in the practice of a profession. What 1 am suggesting is that, because of the cost
of giving a profession a code and the benefits to be derived from any code reasonably related to serving
the relevant ideal, the argument for the formal authority of the actual code will be pretty much what it
would be even if the code were only a pure convention.
16 .For a description of law as a profession in nineteenth-century America, see James W. Hurst, The
Growth of the American Law: The Lawmakers (New York: Little, Brown, 1950), pp. 249-294; or
Roscoe Pound, The Lawyer from Antiquity to Modern Times (St. Paul, MN: West Publishing, 1953).
pp. 175-249.
17 .For a fuller explanation of this technical device, see my "The Moral Legislature: Morality
without an Archimedean Point", Ethics 102 (January 1992): 303-318.
18 .See, for example, Robert Nozick's well-known criticism of the principle, in Anarchy. State, and
Utopia (New York: Basic Books, 1974), pp. 90-95, and of restricted versions of the principle in A.
John Simmons. Moral Principles and Political Obligations (Princeton: Princeton University Press, 1981),
pp. 118-136. For a defense, see Richard J. Arneson "The Principle of Fairness and Free-Rider
Problems," Ethics (1982): 616-633. For a criticism of the method both Nozick and Simmons use to
criticize the principle (in whatever form), see my "Nozick's Argument FOR the Legitimacy of the
Welfare State", Ethics 97 (April 1987): 576-594.
19 .Charles Fried, "The Lawyer as Friend: The Moral Foundations of the Lawyer-Client Relations,"
Yale Law Review 85 (July 1976): 1060-1089; and the critical response of Edward A. Dauer and Arthur
Allen Leff, "Correspondence: The Lawyer as Friend," Yale Law Review 86 (January 1977): 571-576.
20 .For a full defense of this claim, see my "Professional Autonomy: A Framework for Empirical
Research", Business Ethics Quarterly 6 (October 1996): 441-460.
21 .Robert M. Veatch, "Professional Ethics and Role-Specific Duties," Journal of Medicine and
Philosophy 4 (March 1979): 1-19. Quotation, p. 6.
22 .Benjamin Freedman, "A Meta-Ethics for Professional Morality," Ethics 89 (October 1978):
1-19, and "What Really Makes Professional Morality Different: Response to Martin," Ethics 91 (July
1981): 626-630.
23 .For a related example of the importance of this distinction, see Chapter 5.
24 .DR 4-101(C)(3) of the old Code permitted a lawyer to reveal "the intention of his client to
commit a crime and the information necessary to prevent the crime." Rule 1.6(b)(1) of the new Model
Rules permits a lawyer to reveal information only when he believes it is necessary to prevent the crime
and the crime "is likely to result in imminent death or substantial bodily harm."
25 .For a good discussion of the structure of such justified departures though, unfortunately, not
one concerned with departures from moral rules, see Mortimer R. Kadish and Sanford H. Kadish,
Discretion to Disobey (Stanford: Stanford University Press, 1973). Cf. Raz, "Authority and
justification," pp. 25-26.
26 .See, for example, Mike W. Martin, Meaningful Work: Rethinking Professional Ethics (Oxford
University Press: New York, 2000), p. 39: "For the most part he [Davis] focuses on actual codes
--"conventions" among professionals. But in responding to the challenge that some entries in actual
codes are not justified, he quickly shifts to ideal codes as yet unwritten." I do not switch to an ideal
code yet to be written. I merely void provisions morality rules out. The difference is as important as
the distinction judges make between striking from a contract or will a provision that is against public
policy (something they will do) and rewriting a provision of a contract or will (something they will not
do). There is a similar distinction in constitutional law. If a provision declared unconstitutional can be
separated from the rest of a statute, the statute remains law--while the provision is "void". If the
provision cannot be separated, the statute as a whole is void. The court will not second-guess the
legislature.
27 .This is not an answer to Alan Goldman's claim that judges, prosecutors, police, and other
law-enforcement officials but not ordinary professionals like lawyers or doctors have "strongly
differentiated roles," that is, roles giving them a moral duty to do what they know to be morally wrong,
for example, ordering a poor widow out of her house so that a prosperous bank can sell it to redeem her
debt. Goldman's claim, though (I believe) mistaken like those discussed here, is not mistaken for the
same reasons. His claim seems to belong to the philosophy of law rather than to the philosophy of
professions. See Alan Goldman, The Moral Foundations of Professional Responsibility (Totowa, NJ:
Littlefield Adams, 1980), especially pp. 38-49.
