Commentator:
Robert N. Johnson, University of Missouri-Columbia
email: [email protected]
Posted 08/10/01

o one said it more clearly than C.L. Stevenson: Any adequate account of "good" should comply with three requirements "which appeal strongly to our common sense…(1) goodness must be a topic for intelligent disagreement; (2) it must be 'magnetic'; and (3) it must not be discoverable solely through the scientific method." ("The Emotive Meaning of Ethical Terms", reprinted in Ethical Theory. Sellars & Hospers, eds., p. 418) Stevenson argued that while many implausible accounts could at least meet these requirements, interest theories (such as Hobbes') could not unless they rejected cognitivism. As it happens, it is to the second requirement that ethicists most often appeal, that "a person who recognizes X to be 'good' must ipso facto acquire a stronger tendency to act in its favor than he otherwise would have had". (p. 421) This is the internalism requirement.

Svavarsdóttir challenges this requirement. But it is not merely because she challenges internalism that her article is important. Her leading idea is to lay bare internalism's pretensions to be a requirement on any adequate theory of moral judgment, and then attack its claim to this status, a status which internalists have often used as a theoretical cudgel to beat their enemies into submission. Svavarsdóttir begins by arguing that internalists carry the burden of proof that they are entitled to wield internalism in this way. She then defends externalism's claim to be a live hypothesis. Svavarsdóttir's arguments are of particular interest because she embraces what many find to be the least promising strategy for the externalist, that is, defending the hypothesis that "the desire to be moral underlies the motivational impact of moral judgments". (p. 215) Her arguments are subtle, wide-ranging and detailed-enough so that I will have to concentrate on what I take to be a core move in the paper. Although I myself think internalism has unsolved problems, I am, as I explain in what follows, unconvinced by Svavarsdóttir's arguments.

My main focus will be on Svavarsdóttir's burden-shifting argument. First, I want to say something about the overall strategy itself. In outline, her argument this: Presented with an example of someone who genuinely makes a moral judgment but remains unmotivated, an internalist will reject the example as incoherent taken at face value. But in so doing he is ruling out a plausible explanation of this phenomenon, namely, the externalist's. It is a reasonable methodological principle to require those who would restrict the available hypotheses for explaining a given phenomenon to carry the burden of argument for this restriction. But internalists have not, so far, typically done this, or where they have tried, they fail.

Now I agree that many internalists might well reject the example as incoherent (though more on this below). But that would not violate any methodological principle governing the handling of explanatory hypotheses. In empirical investigation, hypotheses are not evaluated by their capacity to explain a single phenomenon. They are evaluated by their capacity to explain patterns of phenomena. Given one's hypothesis has been shown to fit a pattern better than the alternatives, the production of an exceptional example should not reintroduce the alternative hypothesis as a contender. Svavarsdóttir's issue, therefore, must lay elsewhere. In particular, the issue must be whether, were internalism and externalism to be represented as two explanations of observed patterns of moral judgment and motivation, internalists would insist on ruling out externalism and hence restricting the available hypotheses. But Svavarsdóttir gives us no reason to think any internalist has ever or would ever suggest the former. So even if "it is motivational internalists who are restricting the range of hypotheses for Patricke's conduct" (p. 181), that does not show that internalists have wrongly restricted the range of hypotheses for a pattern of moral judgment and motivation. Were the issue presented in these terms, internalists would surely argue that internalism better fits the pattern. This is, for instance, Michael Smith's strategy, whose arguments against externalism as an explanation of motivational impact Svavarsdóttir spends considerable time examining. Of course, once satisfied that it is the best supported hypothesis, internalists will seek alternative ways to explain seeming exceptions other than giving up internalism. But that is as it should be.

A different, though related, complaint might be that the externalist hypothesis cannot get so much as a foothold as a possible explanation of a pattern of observations if the only observations internalists will accept as parts of that pattern are those which are "coherent"-i.e., conform to the internalist thesis. It would be in the process of selecting data, then, that the externalist hypothesis gets ruled out, not in the explanation of that data. But although this sort of objection might eventually be shown to have some bite to it, Svavarsdóttir gives us little reason to think that internalists have engaged in this sort of chicanery.

Stevenson was, in any case, right about this much: It is clearly a condition on the adequacy of any theory of moral judgment that it must account for the pervasive motivational impact of moral judgment, even if internalism itself is not such a condition. It may therefore appear that internalists have surreptitiously and without argument presented their own stronger claim as itself the condition of theoretical adequacy-indeed, even as a platitude-by implicitly appealing to the uncontroversial condition. But this isn't true of Smith, nor even of Stevenson, though the latter comes as close as anyone to wielding internalism as a rhetorical club. It is the "magnetism of the good" that gets top billing in Stevenson's views, and that the good "must be magnetic" means, not the internalist doctrine, but that when all is said and done, any theory must explain the general observation of the motivational impact. I say this about magnetism because the idea of a dynamic function of language is prominent in Stevenson's pragmatism-quite apart from his views on moral language. So although the above passage reveals Stevenson sliding apparently without argument from an agreeable observation of magnetism to a contentious "conceptual truth", this is not to say that he never carried the burden of argument. The internalist thesis falls out of his general pragmatic view that language has to be understood in terms of the uses to which it is put-what he took to be the observed pattern of linguistic phenomena. He simply tried to come up with a general theory that would account for as much of the observable landscape as possible. Again, as with any theory, that will require explaining apparent counterexamples that do not fit. But theories should be evaluated in terms of their general explanatory power, not piecemeal.

Although I think Svavarsdóttir strategy may involve a portrayal of internalism that is not entirely fair, it is not the reason I remain unconvinced. I am unconvinced by the burden-shifting argument itself, which turns on the example of Virginia and Patrick. It seems to me that this example will show in the end quite the reverse of what Svavarsdóttir wants it to. Briefly, we are to imagine that Virginia tries without success to get Patrick to help with victims of political oppression. She does this by appealing to "explicit moral considerations". Patrick unambiguously agrees with those "moral considerations", but says that he just doesn't care. Patrick should not be described as dissembling, nor as having a weak will or not fully possessing moral concepts. He genuinely believed it right to help "but could not have cared less". Here is where there are supposed to be conflicting intuitions about the case: Internalists supposedly find the description of Patrick incoherent. But, again, this requires excluding at the get-go externalism as an explanation. If so, the burden of proof rests with the internalist to offer an argument that this is not coherent.

There is, indeed, no incoherence in this information regarding Patrick's behavior, so far as Svavarsdóttir's example goes . And an internalist who rejected it as incoherent at this stage would be jumping the gun. For Svavarsdóttir only reports Patrick's behavior; she makes no attempt to depict it. We are given the report of Patrick's agreement with Virginia's judgments-not the exchange itself. We are given the report of his own moral claims and lack of certain reaction while possession of others-but not the claims themselves and the behavior that reveals his reactions. We cannot conclude from this alone that there is no incoherence in the example. So internalists, rather than rejecting the example, should ask, in the manner of a detective, just how exactly this is all supposed to go together. An example's coherence depends on more than whether such a string of reports seem coherent. (It is no accident that Lewis Carroll left out any depictions when he wrote that all but the Cheshire Cat's smile disappeared.)

I assume, then, that we are to accept these reports only as schemata within which we are to imagine details. But whether such an example would turn out to be coherent depends entirely on how one fills in those details. I don't mean that the example is incomplete or incoherent unless it comes out favoring internalism. Nor need one go so far as a novel-length treatments. But this is certainly very far from representing the kinds of observations from which one could, as does Svavarsdóttir , draw conclusions about how an internalist will react and whether that reaction is methodologically unreasonable. For instance, what were the terms Virginia used and how did she use them? What were the "explicit moral considerations" to which she appealed? Did she speak in vague platitudes or march through the horrifying details? What did Patrick say in response? How did that conversation go?

The main difficulty I see with Svavarsdóttir's handling of the example is that she does not consider the perspective of Virginia. Yet it only takes one or two questions asked about her for things to start looking quite good for internalism. Let me explain: She is, as Svavarsdóttir would put it, "morally committed" while Patrick is clearly not. But she is also obviously quite rational. This is shown less in what she is fighting for than it is in how she handles Patrick. Any reasonable person would talk to Patrick in the way Virginia does. Virginia cares about those who are politically persecuted; she wants to get others, others such as Patrick, to care about and to help with the cause; she wants others to do something about it. That makes sense, since the facts are shocking and there is no doubt who is to blame and what is to be done. Obviously, Virginia will launch into a blow by blow description of injustice, wickedness and so on. So far, so good.

But why is it so obvious and plain that, despite the fact that Patrick lacks any desire to be moral, Virginia would "appeal to explicit moral considerations"? It is very difficult to come up with any explanation other than that magnetism-the very idea enshrined in Stevenson's requirements-is built right into her conception of moral language and judgment, that moral judgments themselves motivate. Otherwise, why would one bring up moral considerations with people such as Patrick? Why would a reasonable person expect appeals to moral considerations to affect his behavior? Indeed, why is it reasonable to continue trying to convince someone such as Patrick that X is wrong even after he voices agreement but has yet to betray a corresponding disapproval of X? Stevenson has already answered the question: "If in the end you do not succeed in getting him to disapprove of [X], you will feel that you've failed to convince him that [X-ing] is wrong." (p. 419)

Virginia's behavior is not an exception. People typically use moral considerations to affect attitudes and behavior. We do not know to what end Patrick puts such considerations, except that when he uses moral language, he expresses no desire to be moral. Indeed, I see no reason why any internalist would want to deny that Patrick lacks any desire to be moral. Quite the contrary: An internalist will embrace this explanation. The problem is that Svavarsdóttir must rely on this to explain not only Patrick's behavior, but Virginia's, and she must do this without saying that Virginia thinks that moral considerations carry their own motivation with them. Her externalist explanation will entail that, for instance, Virginia need not be seen as having failed to convince Patrick that helping is right. Rather, Patrick, who is supposed to be a competent judge of moral matters, simply differs psychologically from most of us. This is the externalist hypothesis: What explains the motivational effect of moral considerations is this pervasive, though unfortunately by no means universal, desire to be moral. Changes in attitude will follow upon change in moral belief, then, to the extent that a person has a desire to be moral. Other things equal, without that desire, changes in moral belief will not effect attitudes.

This hypothesis, however, makes it quite impossible to make sense of Virginia or anyone else in her position. If her conception of moral considerations does not contain the idea of their intrinsic motivational impact, how do we make sense of her in this example? She cannot be appealing to a desire to be moral in the face of a man who entirely lacks such a desire. What would she think she was doing? Surely not trying to instill a desire in him to be moral by blatant appeals to moral considerations. (Indeed, how does one instill such a desire in someone such as Patrick? Where would she begin?) I don't mean to suggest that this is the only desire a good person needs on the externalist view. As Svavarsdóttir rightly makes plain, the externalist is not committed to the claim that a good person needs one and only one desire, the desire to do what is right. My point is that it is not sheer bias or a mere difference of intuition that is behind the idea that internalism is required to make moral judgment intelligible. It is a pressing question indeed how an externalist can make sense of people such as Virginia-people such as ourselves-who conceive of moral considerations as capable of affecting attitudes and behavior in such circumstances.

I assume that Virginia sees, just as you or I see, that Patrick had no desire to be moral, but appeals to explicit moral considerations anyway. Or are we to think that Virginia does not see this? But if she does not see it, why can we? It cannot be that it is covertly stipulated that he has no such desire, as an aside to the reader, since that would be to rig the example. So we must assume that Virginia could see just as easily as you or I that Patrick had no desire to be moral. Still, she appealed-indeed, as would we-to moral considerations. It therefore must be part of her-and our-conception of moral considerations that they carry motivating power themselves.

Or should Virginia just give up with people such as Patrick-people she believes have no desire to be moral, and concentrate instead only on those who do, for whatever reasons, have this desire? Do we, in fact, give up with them? The answer to both questions is quite clearly "No". Virginia does not and should not give up. Notice that Svavarsdóttir cannot say that Virginia does not and should not give up because she cannot be sure that Patrick still has some secret buried desire to be moral waiting to be uncovered if she badgered him enough. Svavarsdóttir intends Patrick to be a pretty obvious example of someone who entirely lacks a desire to be moral. Virginia must then have persisted with Patrick for some other reason. But then what is it?

True, if Virginia's "explicit moral considerations" were only explicit appeals to the desire to be moral, she would have failed to move Patrick. This much is clear. But this would stop no one who is serious about her moral views-no one such as Virginia. Virginia would appeal, not merely to Patrick's desire to be moral, but to the evils of the political oppression, the cruelty and inhumanity involved, the degrading and hateful way in which these people are dealt with, and so on. This would undoubtedly get more graphic and detailed as she went. Sure, Patrick has no interest in being moral. Virginia must see that, if we can. But that is exactly why only the assumption of internalism can now account for the rationality of Virginia's appealing to moral considerations. It is rational only if it is part of her conception of moral considerations that they carry motivational power. Without that assumption, subsequent appeal to moral considerations in such cases does seem incoherent. She would be about as intelligible as if she had argued with a stone.

Things seem more difficult still for the externalist if Virginia also knew the background information-that Patrick was honest, strong willed, and so forth. Would it have made sense, knowing this, for Virginia to appeal to explicit moral considerations to get him to change his motivations? Surely not if she conceived of these considerations as moving Patrick only if he had a desire to be moral. Patrick had no such desire and Virginia can see this. But appealing to explicit moral considerations did make perfect sense even knowing as we all did that Patrick entirely lacked any interest in being moral. The externalist hypothesis really has been utterly eliminated here as a potential explanatory hypothesis of Virginia; it is not even in the running. That leaves Stevenson's: Convincing Patrick of the badness of the situation, such as the inhumane treatment and so on, entails getting him to disapprove of it. And convincing Patrick of the rightness of helping, because it would stop that inhumane treatment, will engender approval of helping. No approval of helping, no belief that helping is right. That Patrick has no desire to be moral was no reason for Virginia not to appeal to the rightness of helping, the evil of the situation and other explicit moral considerations. It was rational to continue to appeal to moral considerations until Patrick is moved or she concludes that he is suffering from some sort of irrationality. And it is only rational to continue given it was part of her very conception of moral language and judgment that it has motivational power. Externalism is a complete non-starter. It must say that Virginia is deluded or in error. Indeed, it must deny that ordinary rational people regularly appeal to moral considerations with those they know have no interest in being moral.

It is no accident of course that I have turned attention away from Patrick and focused on Virginia. It is her perspective that non-cognitivists such as Stevenson have typically relied on to support their contention that it is part of our conception of moral judgment that it is essentially motivating. Of course, given one person's modus ponens is another's modus tollens, one might try to insist that if the discussion failed to move Patrick, then it is possible that moral judgments don't, in fact, carry motivational power. But that would miss the point, as would removing Virginia from the example. The point is that unless it is a part of Virginia's, and our own, conception of moral judgment that it carries its own motivating power, we cannot understand what she would be doing-indeed, what we would be doing-with someone such as Patrick. If Patrick is a problem for internalism, Virginia is doubly so for externalism.

Let me summarize. Whatever else might be said against internalism, its strength is surely that it makes sense of people such as Virginia. It makes sense to appeal to explicit moral considerations to affect a person's attitudes and motivations, even when he is known to have no desire to be moral, understood in the externalist's sense. But it would not make sense unless built-in motivational power were a part of our conception of moral considerations. If the motivational impact of moral considerations came instead from the desire to be moral, appealing to explicit moral considerations in such circumstances would make no sense. Thus, contemplating examples such as that of Virginia and Patrick seems to me to do exactly what Svavarsdóttir thinks it will not do, namely, rule out externalism entirely.