This is all interesting. I think habit is a good argument against the idea that ALL behavior, however rational, stems directly from what we desire. Even Searle doesn’t seem to take habit into account. He questions the philosophical relevance of the sub-Humean model, which holds that desire motivates ALL action. He says, of course we do everything we do because on some psychological level we want to do it. But that’s not saying much. All it basically means is that we choose (or at least appear to choose; practical-reason philosophers tend to assume free will as a sort of pragmatic construct) to do everything we do. But Searle tries to define desires more narrowly, as those motivations which stem from a sort of non-rational Id: a guy watches porn because he WANTS to; that’s desire. For Searle, things like obligations are fundamentally different. Of course people WANT to fulfill their moral obligations, but this is a very different kind of desire from the kind that causes a person to watch porn.
Habits seem to form an interesting exception though. I think we can account for habits by saying that habit-forming actions stem from desire, so by extension the actions performed under the influence of developed habits stem from those initial desires; not necessarily from what we want at the moment of action. I don’t think the sub-Humean view has an answer for this.
But I suppose I failed to clarify that Searle is talking primarily about RATIONAL action: what things qualify as GOOD reasons for acting? And I don’t think he ultimately defeats the Classical view that all rational motivation must reduce to desires.
Desires remain in the background of practal reasoning; often we don’t engage in full causal reasoning about our actions–which is why desire is not always INVOKED when a rational agent explores her motives. Often it is sufficent for someone to say, “I said x because x is true.” But the desire ALWAYS remains the most primary consideration. If the agent were to engage in FULL causal reasoning she would have to come up with something like,
-I want to tell the truth.
-x is true.
I want to say x.
The desire is the most basic consideration, and when coupled with a true fact about the world, that desire becomes a good reason for rational action. Searle of course admits that desire can be construed in a broader sense like this, but he wants to say that a desire to tell the truth is different from a desire to–say–eat ice cream. However, I tend to think that desires end up reducing to a few ultimate ends; and fulfilling those ultimate ends is what constitutes rationality. My contention is that the end of physical pleasure is no different from the end of fulfilling moral obligations, so long as they are both ULTIMATE ends. Of course, ultimate ends conflict all the time, and much of practical reasoning consists in optimizing and prioritizing in such a way that all of one’s primary ends are maximally satisfied.
Let me just say one more thing about habit here: if the agent can specify a primary end, it will be rational for her to form habits that will consistently meet that end. We may not be directly responding to our desires in habitual action, but if the habit is rational, it has been formed and therefore indirectly motivated by desire.