David and I seem to agree that, for at least certain sorts of normative beliefs or judgments, a kind of moderate internalism is true. For example, Bri’s judgment that she ought to tie her shoe (her belief that this is what she has most reason to do presently or what she has compelling reason to do presently) tends, as a non-accidental matter, to produce in her the intention to tie her shoe.
(Arguably, she would also be irrational not to form this intention upon making this judgment. Set that aside. For reasons that I won’t go into, I don’t think this is essential to there being a necessary connection of some sort between normative belief and motivation.)
I proposed to explain this sort of connection by taking relevant belief-states to be partially constituted by their functioning to motivate. However, partly due to David’s points in our earlier discussion, I’m now less convinced that the partial constitution claim is true. For one thing, the idea that the belief necessarily but not constitutively has the motivational function (and this coming along with a tendency to produce the relevant motivating attitude) does the same explanatory work. And this latter idea fits better with a Humean belief-desire motivational psychology than does my original proposal (the original proposal seems to fit better with the idea that some one attitude-state can function cognitively and motivationally – so-called “besire” views).
So here’s how this might go in our spotlighted type of case. First element: the speaker/agent has a belief of the form ‘I ought to PHI’. We might interpret the ‘ought’ concept here as a special way of referencing some natural, descriptive property (or perhaps other sort of property) that would vary depending on your substantive theory of reasons, normativity, etc. We are concerned, then, not simply with the speaker predicating a certain property of herself or her prospective actions, but of her doing so while using a particular “mode of presentation” (and that’s what the ‘ought’ concept is). Second element: this belief, along with (a) a normal-human-psychology-given antecedent desire to do whatever it is that one ought (has most reason, has compelling reason) to do and (b) relevant normal-human-psychology-given unconscious or conscious mechanisms of rationality or rational response tend, as a non-accidental matter of human psychology, to produce in the speaker the intention to PHI (or produce in the speaker the specific desire to form the intention to PHI). In this way, the relevant necessary or non-accidental connection between belief and motivation (specific desires, intention) is explained. In other words, a kind of “moderate” internalism (defeasible but necessary connection to motivation) regarding this sort of normative judgment is explained. And in a way that jibes well with standard belief-desire psychology.
I’m not sure you need functions, as opposed to just non-accidental tendencies, to get the same explanatory work done. It is just that, for independent reasons, the functional language strikes me as appropriate. (This sort of picture, functions included, worked out in greater detail and applied to moral thought and language specifically, might explain why Hare’s famous Missionaries and Cannibals Argument provides strong reasons for accepting moderate internalism about specifically moral language.)
Expressivism, and hybrid expressivism, are better explanations for “strong” internalism or the view that normative judgments necessarily always motivate. Since strong internalism strikes me as false, and because strong internalism has to be true if expressivism or hybrid expressivism is true, it seems to me that these views are false. They are too strong or explain too much. They are prima facie inferior to both my “belief-constitutive” and “non-belief-constitutive” motivational functionalism about normative belief. In a way, though, these two sorts of pictures do not compete because they are not taking as true and trying to explain precisely the same thing. The most fundamental division is that between accepting moderate (defeasible) and strong (indefeasible) internalism about normative (or specifically moral) judgment. I don’t know why so many philosophers accept strong internalism – or expressivism (or hybrid expressivism) given that they imply strong internalism (and are super-complicated and counter-intuitive to boot).