Here is one way to formulate what is at stake in the quarrel between disjunctivists and conjunctivists in the philosophy of perception. According to conjunctivists, the mental act in which I am engaged when I see a red apple and when I merely seem to see a red apple is the same act. There is a difference, of course. In the first case there is truly a red apple, in the second case there is not. (Perhaps it is a yellow apple in red light, or perhaps I am simply hallucinating and there is no object at all.) But this just shows that successful perception consists of two things: the act of perception, and the favourable circumstances that make the act successful. Unsuccessful perception also consists of two things: that same act of perception, and the unfavourable circumstances that make it a failure. A full picture requires us to add two things together — hence the term ‘conjunctivism’.

Disjunctivists, on the other hand, believe that in the successful and the unsuccessful case there are two different acts. In the first case, it is the act of perceiving a red apple. In the second case, a different act; perhaps an act that is best described as an unsuccessful attempt at perception, or, depending on exactly what is taking place, an act that is best not described as perception at all.

Why is this relevant? Take the following two cases. In case A, I am really seeing a red apple. In case B, I am hallucinating a red apple — but in such a way that the experience is subjectively identical to that of case A. In both cases do I conclude that there is a red apple in front of me. Now there is Cartesian argument that goes as follows: unless I know that I am in case A rather than case B, I do not know that there is a red apple in front of me. So even when I am in case A, unless there is some guarantee available to me that that is the case I am in, I still don’t have knowledge. But I cannot know that I am in case A, since it is subjectively indistinguishable from case B. Therefore the mere possibility of hallucination is enough to render perceptual knowledge impossible.

Disjunctivists tend to believe that conjunctivists are especially vulnerable to this Cartesian argument. The conjunctivist believes that the subject in A and the subject in B perform the same act of perception. But it is clear that the subject in B does not have knowledge. So the conjunctivist must admit that every act of perception falls short of giving us knowledge. It at least doubtful that one can ever square this admission with a robust account of perceptual knowledge.

The disjunctivist, on the other hand, can simply point out that the acts in these two cases are very different and that this explain the difference in knowledge. The person in case A has knowledge because the act of perceiving a red apple gives — if nothing intervenes, which by hypothesis nothing does in this case — knowledge of there being a red apple. The person in case B does not have knowledge because the act of hallucinating a red apple does not give knowledge of there being a red apple. Perception typically does give knowledge. It sometimes does not, but those are the atypical cases which don’t undermine the general rule.

(To be clear, there is much more to be said about these issues. Many other paths might be walked.)

Should we accept disjunctivism? According to the conjunctivist, we should explain seeing a red apple in terms of a conceptually prior act of seeming to see a red apple (which we can perform even if there is no apple) and favourable circumstances that turn this act into a success. According to the disjunctivist, we should take seeing a red apple as the conceptually prior act, and explain merely seeming to see a red apple in terms of something going wrong in the perceptual act. Which explanation makes more sense?

Well, consider the act of drinking. Should we explain drinking in terms of the conceptually prior act of attempting to drink, which turns into drinking only when the circumstances are favourable? Or should we explaining merely attempting to drink in terms of the conceptually prior act of drinking and negative circumstances that make it fail? It seems pretty obvious that the latter is better than the former. You don’t spend your day attempting to drink, attempting to sit down, attempting to read a book, and so on. Those are not your most basic acts. You spend your day drinking, sitting down, reading books, and so on. And then, once in a while, something goes wrong — you pick up a glass and realise too late that it’s empty. What were you doing? You tried to drink, that is, you embarked on the action of drinking, but something went wrong which prevented you from performing it.

The disjunctivist theory of perception has the great advantage that it treats the act of perception like all other acts; arguably, in the only way that acts can be treated. You don’t spend your day trying to perceive the world. You perceive it. And then, once in a while, something goes wrong with perception, and your act falls short of perceiving; that’s when a special explanation is needed.

So I would say that disjunctivism is clearly right… if perception is an act. Now anyone who knows the literature will have noticed how idiosyncratic this presentation of the matter is; in the Stanford Encyclopedia article on disjunctivism, perception isn’t called an act even once. It talks about perception in terms of experiences, events, and states. None of these come with success criteria; for none if them does it make sense to distinguish between doing them and merely attempting them. Now I think that perception is to be thought of as an act of human spontaneity. But that’s a Kantian claim that I cannot defend here and now — except, perhaps, by pointing out that it fits so well with disjunctivism, and that disjunctivism is such a powerful aid against the sceptic.

(I was convinced to think about perception as an act by James Conant’s book The Logical Alien, though having immersed myself into Kant before that certainly helped.)