I’ve recently been thinking a little bit (and only a little bit!) about the relation between intention and intentional action. Presumably, an intention is a mental state. And an intentional action is an action with a certain property, namely, the property of being intentional. According to the simple view of intentional action, intentional action requires the corresponding intention—that is,
SVI: x intentionally A-s only if x intends to A.
Nearly everyone thinks that this principle is just obvious the first time they see or hear it. But most philosophers of action reject it; they believe, instead, that it is a necessary condition for intentional action only that the actor has some intention or other. The standard way to argue against SVI is to offer putative counterexamples, such as variants on Harman’s sniper case (in which a sniper, in knowingly firing his gun within earshot of his enemy, is said to intentionally alert the enemy to his presence, though he does not intend to do so; rather, he simply intends to shoot his target) or Knobe’s harm case (in which a businessman, in implementing a program that he knows will harm the environment, is said to intentionally harm the environment, though he does not intend to do so; rather, he simply intends to make a profit).
One (in my opinion) very good philosopher sympathetic to SVI (who shall remain nameless) has responded to these not uncontroversial cases in conversation in an interesting manner. The sympathizer made the provocative suggestion that we shouldn’t take our arguably unclear intuitions about the putative counterexamples all that seriously, since those intuitions might be tainted by any one of a number of factors (e.g., fallacious reasoning, affect-inspired misattributions, conceptual confusion, etc.). Instead, the sympathizer continued, we should privilege our intuitions about the principle that x intentionally A-s only if x intends to A (or has an intention to A). Why? Well, look at it: it just seems obvious—analytic, even!
Setting aside the prima facie contentious assumptions about intuition and analyticity implicit in this suggestion, I think the suggestion can be seen to be problematic simply by comparing SVI with other prima facie similar principles.
Consider, first, what we might call the simple view of deliberate action: acting deliberately requires the corresponding deliberation—that is,
SVD: x deliberately A-s only if x deliberates about A (or engages in deliberation with respect to A).
This principle might initially seem plausible, if not obvious. If Joe kills his neighbor deliberately, it seems like he must, then, have engaged in the corresponding deliberation. But SVD is actually quite bad. Consider expert birders, who can name the species of particular birds with a surprisingly high degree of accuracy just on the basis of a few simple notes: “a quick glance, hearing the song, may be all that is needed for immediate identification” (Samuels & Flor, “The importance of automaticity for developing expertise in reading”). Novice birders, on the other hand, must “focus their attention deliberately and laboriously on noticing the identifying characteristics…and mapping them onto a particular species” (ibid). Surely, expert birders deliberately name the species of the birds they name. But they don’t, and need not, deliberate about naming the species. Because they are experts, they don't, and need not, engage in deliberation. This, I take it, is an uncontroversial counterexample to SVD.
Consider, second, what we might call the simple view of attentive action: acting attentively requires the corresponding attention—that is,
SVA: x attentively A-s only if x attends to A.
This might seem plausible. For instance, driving attentively seems to require attending to one’s driving. But consider: Joe’s mother asks him to mow the lawn, which is looking quite unruly. She warns him to watch out for the many, many rocks hidden in the grass. If he doesn’t avoid all of the rocks, pebbles and all, he might hurt himself or damage the machine. Joe heeds his mother’s advice: though it takes all the effort he can muster, he is very careful and successfully avoids all of the rocks. In this case, Joe mows the lawn attentively. But Joe does not attend to mowing the lawn; rather, his attention is on avoiding the rocks. There are perhaps clearer (surely uncontroversial) counterexamples to SVA: Typically, when one listens to music attentively one does not attend to listening to the music (rather, one attends to the music). Nor need one attend to watching the performance in order to watch the performance attentively (typically, one just attends to the performance). And so on.
SVI, SVD, and SVA appear to be instances of what we might call the ‘simple schema’, which (abstracting from certain grammatical contingencies of one sort or another) would go something like this:
SS: x phi-ly A-s only if x phi-s with respect to A.
Of course, SS seems plausible (obvious? analytic, even?). But the foregoing shows that it is generally quite implausible, since many of its instances are open to uncontroversial counterexamples. In these cases, we don’t privilege our intuitions about the principle and reject our intuitions about the putative counterexamples. I see no reason to think that we should treat SVI any differently.