I
So we have seen what Burge needs to be claiming, in order to hit McDowell's actual position (instead of the one he misleadingly attributes to him). He needs to be claiming that the perceptual state which characterizes one's subjectivity/point of view/access to the world is held in common between the three cases(Good, Bad, and Other) since that is what McDowell is concerned to deny. The question is whether the considerations Burge brins against disjunctivism have force against this position, or only against the one he misleadingly attributes. Let us see what he says.
II
Burge introduces the "Proximality Principle", which states that, assuming identical antecedent psychological state, two perceivers which receive identical proximal stimulation (e.g. same stimulation of retinas), as well as the same afferent and efferent input will produce a given type of perceptual state, assuming no malfunction or interference.
Burge claims that explanations of perceptual states in the science assume the Principle. It is "basic to the method of the science". The science studies perceptual states in artificial scenarios, ones in which there are illusions and such, and assumes that in doing so it is shedding light on the perceptual states we are in in cases which are not unusual. This is how the science studies perceptual states.
Burge claims that if the Principle is denied for a state, this is tantamount to the claim that any "normal causal psychological explanation is possible" for that state, since the normal methodology of psychological explanation takes the individual "as a unit", and denies action at a distance. This requires that the states of an individual be explained by proximal and internal things. Do these claims have force against the disjunctivist position McDowell is defending? Indeed they do, if true. The state McDowell is interested in, which characterizes one's subjectivity/access/whatever, is indeed a perceptual state. As such, if the Proximality Principle is true, it must be shared between the three cases. But McDowell claims it is not shared between the three case. If the Proximality Principle is true about the perceptual state that characterizes access/etc, McDowell is wrong about this perceptual state.
III
But what is the argument for the Proximality Principle? Burge simply says that perception science is committed to it, by its methods. Burge wants McDowell to be in opposition to perceptual science, which is why he is constantly emphasizing that McDowell is talking about a perceptual state, and perception science is too. What he needs is this.
- Perception science is committed by its methods to the Proximality Principle's being true of its subjects. (From Burge's examples of how the science proceeds)
- The state McDowell is talking about is a subject of perception sceince. (Burge keeps insisting this)
Therefore,
3. Perception science is committed to the Proximality Principle being true of the state McDowell is talking about. (From 1 and 2)
And only if McDowell were coming up against vision science in just such a way would the substance of Burge's accusation be accurate. In particular, his claim that not since Hegel claimed that there are only six planets for some (presumably bad) philosophical reason have we had such a brazen denial of science from philosophy.
IV
So, is this right? Is McDowell being that irresponsible?
I think Burge is probably right about the methods of perception science. I think it is the second claim, that this is a topic for perception science, taken as being constituted by its methods (and their committment to the Proximality Principle) where McDowell would disagree, and where the really interesting issue is. I think McDowell really is committed to saying that this perceptual state cannot be characterized by the methods of perceptual science Though, of course, perception science can say lots of helpful things about it, as long as their methods are committed to the Proximality Principle they cannot fully characterize it.
One way to put this would be to say that they will never get beyond the visual system, or perhaps the brain. They can characterize what it is that the brain is doing, but cannot characterize what we ourselves are doing, on McDowell's view. McDowell has indeed defended claims like this in the past (the "Content of Perception" paper, for example). He says that the methods of perception science are limited to descriptions of "subsystems", rather than "whole animal" states.
Burge denies this quite strongly. He says that perception science is committed to aiming at "whole animal" states because of its constituitive connection to whole animal actions such as eating, running away from stuff, etc. Indeed, somewhere (I can't remember quite where) he even disclaims on behalf of perception science a brain-mapping goal. He points out that this is not the goal–rather, perception science is after an adequate account of whole animal action and how it relates to perception as well as accounting for subjective reports of what one is perceiving, and so on.
At this level, Burge is surely right–perception science isn't talking about just something eyes do, for example. But there is a gap between establishing that perception science is describing a "whole animal" phenomenon and saying that it is describing the "whole animal" phenomenon that McDowell is interested in when he talks about "subjectivity". I do not think the "whole animal" vs. "subsystem" approach is a useful way to look at this question.
One thing which would serve to distinguish the two would be consciousness. Suppose McDowell were essentially interested in states which are essentially conscious (or characterize a subject's consciousness) and perceptual psychology were explicitly not about such states, then we would have a situation in which the states McDowell are talking about are not those perception science is studying.
But I do not think the issue can be seen this way. For one thing, McDowell doesn't say that he is constituitively interested in consciousness (or at least nowhere I can find). He does say that conscious states are states of this kind. But I don't think he wants to commit himself to the claim that all these states are by their nature conscious. So he thinks that some but not necessarily all of these states he is interested in are conscious, but all conscious states are states of this kind. Burge says similar things. He doesn't weigh in on what makes a state conscious, but he describes some of the states of perception he is interested in as themselves conscious states. As he sees it, I think, perception science really is trying to characterize conscious states. This is the point of his insistence on their taking subjective reports as their data–they are after representations which will account for all the data of consciousness.
Even if McDowell did think that the states in question were all conscious, this would not suffice to show that they are not the subject matter of perception science, since Burge (speaking on behalf of the science) claims that some of the states of perception are conscious. This means that some of the states studied by perception science (and therefore subject to their commitments) are part of the class McDowell is interested in.
V
So, for now, I think I do see McDowell as being in the position Burge is accusing–namely that of having his views conflict with those to which perception science is commmitted. Some perceptual states that are within the subject matter of perception science, the ones that characterize our "point of view", including the conscious ones, are said by McDowell to not be subject to the proximality principle. If perception science is indeed committed to this principle in studying those states, they are in conflict with one another.
But it is not clear that this is that bad a position for McDowell to be in. When it comes to the realm of access and especially of consciousness it is notoriously difficult to see how more of the kind of science we are engaging in now can provide us with a satisfying explanation. McDowell is certainly able to take on board much of the business of perception science as characterizing something important, but he is committed to saying that if they take themselves to be characterizing the states that constitute our subjectivity, as Burge says they are, then they are mistaken.