I’m a moral expressivist. There. I’ve said it. I know its implications, and I know what it commits me to: amongst other things, the denial of the existence of moral truths (and moral falsehoods for that matter), a seemingly eminent slope towards relativism, and the stark realization that the objectivity and authority of morality may be nothing more than smoke and mirrors. But I do think I have good reasons to hold this view; I have been convinced by what I consider to be sound arguments, and I believe I have effectively argued for a certain kind of non-cognitivism elsewhere. As for those irksome consequences of holding such a view, I can accept that sometimes, we need to simply bite the bullet. For all it’s worth, I think we all should be moral expressivists. But that’s just my opinion.
I’ve been spending some time lately thinking about various forms of non-cognitivism, of which moral expressivism is but one kind. Another sort is what is sometimes referred to as motivational non-cognitivism. In the effort to prepare a short manuscript about its relationship with moral motivation, I figured it would be a good idea to familiarize myself with some common arguments against motivational non-cognitivism. While the literature certainly contains its fair share, none of them I found extremely effective. In fact, most of them seem to beg the question against the motivational non-cognitivist. But more surprisingly, nowhere could I find a solid argument for its rival view; i.e. motivational cognitivism.
In the following, I attempt to present what I conjecture to be the most charitable argument for motivational cognitivism and demonstrate that, even in its most charitable form, the position for motivational cognitivism is rather shaky. I would be very interested to hear what you have to say, both about the argument for motivational cognitivism and my argument against it.
For those that are not familiar with the distinction mentioned above, let me broadly set the stage. The debate between motivational cognitivism and non-cognitivism should not be confused with the debate between moral cognitivism and non-cognitivism. The latter is a debate concerning the ontological status of moral claims. The moral cognitivist claims that moral judgments are a certain kind of belief which can either be true of false. Since a true moral belief must be true in virtue of corresponding to reality, this is often understood to lend support to moral realism. Moral non-cognitivism, on the other hand, holds that moral claims are, in their very nature, not cognitive (they are merely conative) and hence not the right sort of things to be true or false. One form of moral non-cognitivism is moral expressivism; the view holding that a moral claim, like any other evaluative claim, is nothing more that an expression of one’s sentiment.
The debate between motivational cognitivists and non-cognitivists centers on the relation between evaluative claims (of which moral claims are but one kind) and motivation. The motivational cognitivist claims that certain sorts of beliefs (e.g. moral beliefs) can motivate an agent to act. For example, it is often believed that the belief that it would be wrong to cheat on your taxes can motivate an agent to refrain from doing so. Of course, motivational cognitivists are not committed to the view that moral or other evaluative beliefs always motivate; indeed, they must allow for the possibility of some form of irrationality. All they are committed to is the view that certain evaluative beliefs can potentially motivate; i.e. they are of the right sort of things to motivate agents to act.
The argument for motivational non-cognitivism consists primarily in making a distinction between two different sorts of psychological states. On the one hand, we have what are called conative states; states like desires, wants, values, and other pro-attitudes. These can be distinguished from cognitive states by identifying their particular direction of fit. Desires and other pro-attitudes are psychological states in which the world must fit. As Michael Smith writes, one does not stop desiring P upon the realization that ~P. On the other hand, cognitive states like beliefs must fit the world. If rational, one stops believing P upon the realization of ~P. Because of this difference in direction of fit between beliefs and desires, it is argued that the only sort of psychological state that can motivate an agent to act is a conative state, i.e. a desire or some other pro-attitude. According to the motivational non-cognitivist, simply believing that P is what one should do is not enough to motivate an agent to act. There are plenty of facts out there that play no role in motivating an agent to act. So what makes certain putative facts and not others have motivational force? By virtue of what is it true that some putative facts can motivate an agent to act while some cannot? According to the motivational non-cognitivist, it is because the agent must also, e.g., care about the fact that it is what she ought to do.
The argument for motivational cognitivism usually appeals to one simple but arguably shaky premise. According to the motivational cognitivist, there are certain beliefs – for example moral beliefs – that can motivate an agent to act. Support for this claim usually appeals to the apparent paradox of an agent holding the belief (or even knowledge) that e.g. it would be wrong to torture children for fun but does not care about that. Indeed, holding such a pair of psychological states is sometimes understood to be equivalent to Moorean paradoxes in which an agent e.g. claims that it is raining outside but does not believe it.
Now, I agree that there is something quite odd about an agent holding the belief that it would be wrong to torture children for fun while not caring about that. But the reason that makes it odd need not in any particular way lend support to motivational cognitivism. In fact, I believe it does the exact opposite; it rather lends support to its rival view.
In order to come to see how this is so, one first has to accept some form of reasons-internalism, which, although still controversial, is nevertheless understood to be the default position in the domain of practical reasons. According to this thesis, a reason for action is determined in part by an agent’s subjective motivational set (or more generally, by an agent’s desires). Combining this thesis with moral expressivism, one can come to understand that the belief that it would be wrong to torture children for fun can only mean that the agent (perhaps everybody else as well) has conclusive reason to refrain from torturing children for fun. But for the reasons-internalist, if it is true that the agent truly has conclusive reason to refrain from torturing children for fun, then it must be the case that the agent has a desire or something else in his subjective motivational set that would be furthered or promoted by refraining from torturing children for fun. And if that is true, then the agent already has the desire present in his psychology; a conative state that has motivational force that can also explain why it is the case that the agent has the reason for action in the first place. What is odd about an agent holding the belief that it would be wrong to torture children for fun while not actually caring about that is that the moral judgment is already loaded with the implicit commitment to the welfare of children and so cannot also fail to care about it at the same time. The agent here would be desire P and ~P at the same time, and this is, if not a mark of irrationality, a definite mark of oddity.
So it seems that moral judgments, when understood as a claim about what one has reason to do (or refrain from doing) are already conatively loaded, so to speak, and it would be inaccurate for the moral cognitivist to claim that certain cognitive states of an agent can motivate them to act. Moral judgments can motivate, yes. But that is because there is already a relevant pro-attitude in place that explains the presence of a motivational force.
Any thoughts?
