Do Moral Judgments Express Complex Attitudes?

Josh's post about Russ Shafer-Landau's moral nonnaturalism and a strand in Michael's post about Mackie and disagreement brought back to me something that I have been thinking about for awhile: whether moral attitudes are (logically) complex. As far as I have been able to tell, I have not seen any work suggesting that this might be the case. Rather, almost all discussion that I have seen concerning moral attitudes assume that such attitudes are either (logically) simple beliefs or (logically) simple pro- or con-attitudes. I think there may be reason to think otherwise. But even if there is not, I think that recognizing this possible view is important for other reasons concerning a good deal of the dialectic that goes on in metaethics.

By 'logically complex' psychological attitude, I just mean an attitude that is composed of more than one attitude. So, the view I am suggesting is this: moral attitudes are complex psychological states consisting of a belief and a desire. For example, if I judge that donating to charity is right, that judgment expresses a complex psychological state consisting of a belief that donating to charity has a certain property (viz., rightness) and a pro-attitude toward donating to charity or toward the property rightness, or something. So, on this view, moral attitudes are not merely beliefs, nor are they merely desires, but a complex consisting of both. (Also, if a "besire," assuming there are such things, is a simple psychological state that has both directions of fit, then neither, on the suggested view, are moral attitudes besires.)

The reasons to think that moral attitudes may be complex in this way are not very novel; they are just most of the standard reasons to think that moral attitudes are beliefs and most of the standard reasons to think that moral attitudes are pro- or con-attitudes. For example, that ethical statements are in the indicative mood that is intuitively reserved for conventional use to express beliefs, that some phrases, such as 'good intentions', appear to refer to moral properties and, hence, to have descriptive content, and that many of our actions, such as arguing, reflecting, deliberating, and convincing others about ethical matters provides some evidence that moral attitudes have representational or descriptive content and, hence, are beliefs. On the other hand, that a change in one's pro- or con-attitudes reliably follows in the wake of a change in one's moral judgments and that there appears to be a very strong link between a person's acceptance of a moral claim and the person's motivation provides some evidence that moral attitudes are motivational and, hence, are desires, broadly construed as some kind of pro- or con-attitude. Of course, all of this evidence for moral attitudes being either beliefs or desires is defeasible, but there is no doubt that these considerations seem to lead us to the conclusions that moral attitudes are beliefs and that moral attitudes are desires.

Thus, we find much of the literature arguing over whether moral attitudes are beliefs or desires. But, such an argument seems pressing only on the assumption that moral attitudes are simple psychological states; otherwise, there would seem to be no reason to argue for one side or the other. But all of the evidence just mentioned is not only consistent with, but also suggestive of, the claim that moral attitudes are both beliefs and desires--or, more accurately, that they are complex psychological states consisting, in part, of a belief and, in part, of a pro- or con-attitude. Perhaps the only initial, intuitively strong reason to reject such a view is a correspondingly strong intuition that an amoralist is conceptually possible. Nor can it be claimed that the existence of complex psychological attitudes is uncommon. For example, it seems quite plausible to think that fears and hopes are also complex psychological states. For example, if I fear that there is a black widow in the shoe that I am about to put on, my attitude seems to be a complex consisting of a belief that a black widow may be in the shoe that I am about to put on and a desire that there not be.

I think the question of whether moral attitudes are simple or complex psychological attitudes is important. For example, the assumption that moral attitudes are simple states is precisely the wedge in current metaethics that drives what Michael Smith thinks is the central organizing problem in contemporary metaethics, what he calls "The Moral Problem," the problem of reconciling three intuitively plausible, but apparently mutually inconsistent claims. Here are the three claims, modified slightly (The Moral Problem, p. 12).

(1) Moral judgments of the form 'It is right that I ___' express a subject's belief about an objective matter of fact, a fact about what it is right for her to do.
(2) If someone judges that it is right that she ___, then, ceteris paribus, she is motivated to ___.
(3) An agent is motivated to act in a certain way just in case (a) she has an appropriate desire and a means-end belief, where (b) belief and desire are, in Hume's terms, distinct existences.

The idea is that these three claims appear to be mutually inconsistent. (1) and (2) indicate that moral attitudes are necessarily, motivational beliefs, but (2), (3a), and (3b) indicate that only desires, and not beliefs, can be necessarily motivational. Smith then goes on to describe, justly I think, the last century of metaethics as trying to resolve this apparent problem. However, notice that this problem is only a problem on the assumption that moral judgments express either simple beliefs or simple desires. That is, (1) and (2) indicate that moral attitudes are necessarily motivational beliefsonly if we also assume that the moral attitude expressed in moral judgment is a simple belief. If we are not assuming this, then all we can really conclude from (1) and (2) are that moral judgments express attitudes that are in part beliefs and in part motivational--and this is perfectly consistent with moral attitudes being complex states consisting of both a belief and a desire. The recognition that moral attitudes can be complex attitudes consisting of a belief and a desire provides a swift, powerful dissolution to The Moral Problem.

I think the recognition that moral attitudes can be complex psychological states is also important for other kinds of metaethical dialectic. For example, there seem to be at least two important things at stake in Shafer-Landau's recent discussion of what he calls the Non-cognitivist argument against moral realism (Moral Realism, p. 121). First, by ignoring the distinction between simple and complex attitudes, SL misses an opportunity to rather easily refute the Non-cognitivist argument--at least as an argument against moral realism. Second, by ignoring the distinction, SL's first and third objections to Humeanism fail (or, more precisely, these objections will remain unsettled at least until the issue of the possibility of an amoralist is settled).

As SL understands it, the Non-cognitivist argument is an argument against realism. (I think, historically, this argument should be taken as an argument for Expressivism, rather than as one against Realism, but that is another matter.) Taking now into account the distinction between simple and complex attitudes, the Non-cognitivist Argument as stated on p. 121 is ambiguous between the following two arguments. (I'll use 'simple' and 'complex belief' rather than the more cumbersome 'simple' or 'complex attitude consisting, in part, of a belief'.)

Argument 1:
(1) Necessarily, if one sincerely judges an action right, then one is motivated to some extent to act in accordance with that judgment (Motivational Judgment Internalism).
(2) When taken by themselves, neither simple beliefs nor complex beliefs motivate (Motivational Humeanism?).
(3) Therefore, moral judgments are neither simple nor complex beliefs.

Argument 2:
(1) Necessarily, if one sincerely judges an action right, then one is motivated to some extent to act in accordance with that judgment (MJI).
(2) When taken by themselves, simple beliefs do not motivate (Motivational Humeanism?).
(3) Therefore, moral judgments are not simple beliefs.

If the Non-cognitivist argument is to be taken as Argument 1, then it is unsound, because 2. is obviously false. Consider again, for a counterexample, the complex psychological state of fearing that there is a black widow in the shoe that I am about to put on. The psychological state of fearing that ___ can reasonably be considered a complex state consisting of a (simple) belief that a black widow might be in the shoe that I am about to put on and a (simple) desire that it not be. Since this state is certainly motivational, 2. is obviously false, and Argument 1 is unsound. On the other hand, if the Non-cognitivist argument is to be taken as Argument 2, then, even if it is sound, it does not tell against realism at all, since its conclusion is consistent with moral judgments being in part simple beliefs.

So, by ignoring the distinction between simple and complex attitudes, SL misses what appears to be an easy opportunity to refute the Non-cognitivist Argument. But, SL doesn't use these objections against the Non-cognitivist Argument. Instead, he offers three other objections to Motivational Humeanism (and then some other objections (1)). By ignoring the distinction between simple and complex beliefs, I think SL's first and third objections against Humeanism fail. Let's just look at his first objection, which is an appeal to our intuitions about some apparently evaluative beliefs:

Before theorizing about motivation, most of us would find it plausible to suppose that some evaluative beliefs . . . (c)an motivate all by themselves…. Sometimes a desire simply doesn't appear on the scene--'I saw that it needed to be done, and did it. I don't recall wanting anything at the moment'." (p. 123)

Now if one is assuming that these examples provide examples of (i) beliefs that are (ii) evaluative, (iii) simple, and (iv) had by persons who are motivated as described, then one should draw the conclusion that these are evaluative, simple beliefs that are sufficient for motivation and, hence, that Humeanism is false.

But what reason has SL given us for supposing that these beliefs are evaluative and simple? If evaluative beliefs are complex psychological states consisting of a simple belief and a simple desire, then if these examples provide examples of evaluative beliefs, then these beliefs consist in part of a desire, and hence, Humeanism has not been shown to be false. On the other hand, if evaluative beliefs are complex, then if these examples provide examples of simple beliefs that are sufficient for motivation, then these beliefs are not evaluative (because they are simple).

The upshot is that, in order to use this objection effectively against Humeanism, Shafer-Landau has to first show that evaluative beliefs are simple beliefs. So, I think there are two things at stake for SL's failure to appreciate the possibility that evaluative judgments are complex. First, he does not avail himself to what appear to be very good objections to the Non-cognitivist Argument as an argument against realism, objections he has ready-to-hand. Second, at least two of the objections he does offer to the Non-cognitivist argument seem to fail--at least until it can be shown first that moral attitudes are simple. (I want to reiterate that I am not intending here to pick on Shafer-Landau. I am just using his discussion of the Non-cognitivist Argument as representative of much of the discussion in the literature concerning moral attitudes.)

So: I want to throw out the suggestion that moral attitudes are complex psychological states consisting in part of a belief and in part of a pro- or con-attitude. The reasons to think so are most of the standard reasons to think that moral judgments are beliefs and most of the standard reasons to think that they are pro- or con-attitudes. The only reason to think not, it seems to me, is if one also has a correspondingly strong intuition that an amoralist is possible. Moreover, until the issue is resolved, it seems that a lot of the current metaethical dialectic should be massaged to include this possibility. I am really interested to hear what others think about these matters. And if there have been others that have suggested that moral attitudes are complex, please let me know.

Comments

[uriah]

Dan, I think the view of cognitivist internalists, such as McDowell, McNaughton, and Dancy, can be understood as holding that moral judgments are complex attitudes. In this respect, they are like, say, gladness (an example given to me by Jamie Dreier way back when). To be glad that p involves both believing that p and desiring that p, or more generally, having both a thetic (mind-to-world) and a telic (world-to-mind) direction of fit to p. A more subtle view is developed by Horgan and Timmons in a series of recent articles, some of which haven't come out yet (e.g., "Nondescriptivist Cognitivism," "Cognitivist Expressivism," and others). On their view, moral judgments are attitudes that are similar in some crucial respect, but not in all respects, to beliefs, and also to desires. Finally, there is the view sketched by Millikan in "Pushmi-Pullyu Representations" (*Mind and Morals*, ed. L. May et al.) that moral judgments are attitudes more primitive (evolutionarily and cognitively) than either beliefs or desires but feature some crucial aspects of both.

[Dan Boisvert]

Hi Uriah. Thanks for the pointers, which is exactly what I was hoping for with the post. I think, however, that all of the folks you mention above view moral attitudes as (logically) simple, unitary states, not complex or composite ones as I was suggesting. For example, I have always read McDowell, McNaughton, Dancy, and Millikan as holding a "besire" theory of moral attitudes. That is, I have always read them as holding that a moral attitude is a (logically) simple, unitary psychological state that has both word-to-world and world-to-word directions of fit; and these directions of fit are the respects in which moral attitudes are like beliefs (word-to-world) and also like desires (world-to-word). I'll look them over again to make sure I've read them correctly. Of the Horgan and Timmons articles you mention, I have read only "Nondescriptivist Cognivitism," but, as I understand them, they also want to hold that a moral attitude is a simple psychological state, not a complex one (of course, they go on in that article to describe complex psychological states that are composed, in part, of these moral attitudes, such as psychological states that are expressed by complex sentences that embed atomic ethical sentences). More specifically, as I understand them, they want to hold that a moral attitude is a special kind of simple belief, namely, one that is not in the business of representing or describing the way the world is and, so, does not have a word-to-world direction of fit. Like the others, I'll read this article over to make sure that I have got their view correctly.

So, again, the view I am suggesting is one in which moral attitudes are complex and consist, in part, of a belief and a desire. I also want to make clear that the view I am advocating is not really novel. The idea that psychological attitudes can be complex has been recognized, though I haven't been able to find that it has been seriously entertained as a possibility for moral attitudes. For example, Smith very briefly mentions the possibility that moral attitudes can be complex as a defense from a particular objection by besire-theorists (The Moral Problem, p. 117), but other than for that purpose, seems just to ignore this possible view that, as I see it, would completely dissolve The Moral Problem.

[Kyle Swan]

Dan,

First, I think that your reading of McDowell, McNaughten, et al. concerning this issue is the right one.

Second, perhaps characteristic moral emotions – guilt, anger, resentment, indignation, shame, contempt, derision – are complex attitudes of the kind you have in mind. Emotions: An Essay in Aid of Moral Psychology (Cambridge, 2003) by Robert Roberts presents an account of emotions according to which they are "concern-based construals". For Roberts, construals are like perceptual beliefs, involving the way things appear to the subject. But emotions are construals that are based on concerns – desires, interests, or aversions – the subject has.

Roberts has in the works a companion volume to the 2003 book where the account of the emotions is directly implemented in an account of moral judgment. Until this appears, you might look at Linda Zagzebski, “Emotion and Moral Judgment” PPR, (Jan 2003) and Sabine Doring, “Explaining Action by Emotion” PQ (April 2003), which both seem to suggest something similar. Zagzebski does write as if she thinks this approach dissolves the moral problem.

[Dan Boisvert]

Hi Kyle. Thank you so much for the suggested readings. Also, the suggestion that the characteristic emotions of morality might also be complex attitudes is interesting. If they are, then the idea that the more general moral attitudes are complex would really not be all that surprising.

Thanks again.

[uriah]

Dan, I think I misinterpreted your original post. But now I'm a bit worried about how to go about distinguishing complex from simple attitudes. Suppose Comp and Simp are two psychologists working on gladness, and they have this disagreement:

Comp: Gladness that p is a complex attitude that has belief that p and a desire that p as "parts" or "constituents".

Simp: Gladness that p is a simple attitude that has a belief-ish-that-p and desire-ish-that-p aspects to it.

How would they go about settling the dispute? More specifically, what is the psychologically real feature that is supposed to distinguish gladness as conceived from Comp from gladness as conceived by Simp?

(Also, Dan, I realize I use "judgment" differently from you, to cover not only linguistic acts, but also mental states.)

[Dan Boisvert]

Hi Uriah. Your good question is a good one. (Here's another: "Who cares? What's the practical upshot of moral attitudes being complexes of beliefs and desires rather than simple states with two directions of fit?") I haven't been able to think about these questions very much, so perhaps others will have some suggestions. But I will make what, I'm sure, is a rather obvious point: even supposing that we are not able to discover any real difference between one who has a complex attitude consisting of a belief and a desire and one who has a single attitude with both directions of fit, there is still the metaphysical question of whether moral attitudes are complex or simple. And on this question, I think there is more support for the idea that moral attitudes are complex (assuming that the debate is between only these two positions). There is nothing mysterious at all about having a belief that p and some kind of pro-attitude toward p, so there is nothing mysterious about a complex attitude consisting of the two. On the other hand, I think besires are quite mysterious kinds of mental states, the arguments for their existence are few, and, to me, are not very persuasive. It seems to me that anything the postulation of besires is intended to explain can be explained equally well by appealing to complex attitudes consisting of a belief and a desire. Perhaps others, especially defenders of the existence of besires, can shed a more charitable light on these questions for us?

[Neil Sinhababu]

You raised the possibility of an amoralist -- someone who has the epistemic part of moral judgments but not the practical part -- as a potential objection. I'm trying to think about what a person who had the practial but not the epistemic part would be like. I'm thinking that such a person could be moved to action by experiences that we would call 'feelings of obligation'. However, if others weren't moved to action in the same circumstances that moved him, he wouldn't regard them as making any kind of mistake. I don't regard people as making mistakes if they fail to desire foods that I enjoy, and that's how he would see other people's failure to be morally engaged.

I don't have very strong intuitions about whether this person's attitudes count as moral, though I lean slightly towards seeing his attitudes as nonmoral, which is good for your view. Do you have any intuitions about this case? And do you know of any literature on cases like this?

[Dan Boisvert]

Hi Neil. Your question seems to be, "What makes an attitude a moral one?" A good place to start might be Chapter 1 of Simon Blackburn's Ruling Passions. He claims there, plausibly, that what makes one's attitudes moral depends in large part on our second-order attitudes towards our first-order attitudes which, in turn, are directed towards some salient feature(s) of a situation, and then on our third-order attitudes towards those second-order attitudes, and so on. For example, suppose a person were to delight in injuring or lying to others for the fun of it. My initial reaction (i.e., my first-order attitude) is repulsiveness at such acts. Moreover, I approve of this first-order attitude, which is a second-order attitude, and this second-order attitude may cause me to have a different first-order attitude of desire to instill in my children a repulsiveness toward such acts, and so on. Blackburn's story may explain why you do not take the person in your example to be making a mistake. That is, perhaps you do not have the appropriate first-, second-, third-, etc. order attitudes towards his failure to desire certain foods. For example, you may not care at all whether that person desires foods that you enjoy, you may not care that you do not care about such things, and you may not have any desire whatsoever to instill in your children the desire to eat certain foods that you enjoy, and so on. If so, then your attitudes towards eating foods that you enjoy are not moral ones, and so you may not regard this person as making a mistake. (Compare your example with the example above in which a person delights in injuring or lying for the fun of it. If you are repulsed by this person's delight, if you desire to be repulsed by this person's delight, if you have a desire to instill in your children a repulsiveness towards such actions, etc., then you may have a moral attitude towards such acts, and, in turn, may be inclined to view this person as making some kind of mistake, namely, the mistake of having an inapt attitude.)