Download Russellianism and Self-Representation: The Problem of Egocentric Indexicals - Prof. Jeffre and more Study notes Introduction to Philosophy in PDF only on Docsity!
Russellianism and self-representation
phil 93507
Jeff Speaks
November 2, 2009
We’ve already seen that it is plausible to take perceptual experiences to represent things as having a certain distance or orientation relative to a perceiver. But that means that when I have a perceptual experience I am representing myself as standing in certain relations to things.
On a Russellian theory of content of the sort we have been discussing, there is no difference between my representing a tree as a certain distance from me, and my representing the tree as that distance from Jeff Speaks.
This gives rise to a standard problem for Russellianism, which is sometimes called ‘the problem of the essential indexical’ (a phrase due to Perry). This is actually a cluster of problems, each of which can be thought of as an argument for the conclusion that a pair of sentences which differ only by the substitution of ‘I’ for ‘Jeff Speaks’ can differ in content (when uttered by me). Some versions of this problem:
- ‘I am JS’ is a posteriori/cognitively significant, whereas ‘JS is JS’ or ‘I am I’ are a priori/trivial.
- ‘I exist’ is a priori, whereas ‘JS exists’ is not.
- ‘I am on fire’ has a direct connection to action which ‘JS is on fire does not.’
Standard versions of Russellianism, which take indexicals like names to be devices of direct reference, must come up with responses to this sort of problem. (Fregeans can say that the content of ‘I’ as used by me is a special first-personal content which rigidly designates me, and is not the content of any name. Fregeans, however, run into their own problems with indexicals; the classic discussion of this is Perry (1977).)
There is also an interesting variant on this problem of first-personal contents which arises specifically in perception. This is brought out nicely by the discussion of the perceptual representation of egocentric directions in ch. 3 of Peacocke (1992):
“Take first the construal on which seeing something to be in egocentric direc- tion D involves merely seeing it as having a certain direction in relation to object x, where x is in fact the perceiver himself. This reading is too weak to capture what is wanted. This is because one can see something as having
a particular direction in relation to an object x which is in fact oneself while not realizing that the object to which one sees it as bearing that relation is in fact oneself. Examples of persons seen in mirrors suffice to make the point.”
Suppose that I see someone in a mirror (who turns out to be me) and I visually represent that person as having a book to their right. It then follows that I visually represent that: there is a book to the right of JS. This is the same thing, for the Russellian, as visually representing that there is a book to the right of me. But intuitively in this sort of case I need not visually represent that there is a book to the right of me.
(This argument assumes that the Russellian should treat visual representation of objects in mirrors as having contents which are singular propositions involving the relevant objects. This is not completely obvious, but I think that it is plausible, so I will just ignore this assumption.)
The Russellian is used to denying intuitions about differences in content of this sort, so perhaps this case does not introduce any new worries. But something like Peacocke’s case can be used to raise a kind of problem for a Russellian who is also an intentionalist.
The most straightforward way to run such an argument would be to try to construct a pair of cases of the following sort: in Case 1, I represent myself as surrounded by such- and-such features of the world. In Case 2, I represent someone (who turns out to be me) in a mirror as surrounded by just those features of the world. Case 1 and Case 2 will clearly differ dramatically in phenomenal character. So, if we can find a way to construct a pair of cases of this sort which are also the same in content, then it looks like we will have a counterexample to Russellian intentionalism.
Unfortunately (or fortunately) it is not obvious that we can construct a pair of cases of this sort. In the most obvious ways of constructing the cases, Case 2 will include some ‘extra content’ which the intentionalist can use to block the counterexample. (For example, in Case 2 I will visually represent JS as some distance from me — and in Case 1 I will not.)
Even if we don’t have a counterexample, there is some weirdness here — after all, on the present construal, Case 2 (at least a version in which I don’t know that the object I am seeing is in a mirror) involves me representing JS as some distance from JS — but intuitively Case 2 is not a case in which I represent myself as bilocated.
But I think that we can also get a bit closer to a counterexample by changing the cases. Consider instead two variants on Case 2: two cases in which I am viewing myself in a mirror. In one case, let the mirror-representation be as of that object having a book to his right. Now imagine a case in which in addition to this, I also (intuitively speaking) represent the book as to my right. We will then have no difference in content between the two experiences, since both represent a book as to the right of JS.
(The second one represents this twice over, but it is hard to see how this could make a difference. If you think it does make a difference — e.g. because you think that in the second case I represent two objects as having a book to their right — then subtract the book from the mirror-representation. You can suppose that the mirror is so constructed as not to reflect objects to the right of where I am sitting. Though this arguably introduces a new difference in content, since now I will not represent there being a book at the location
