Stanley Fish often plays the equivocation game.  This game consists in solving a real philosophical problem by erasing or denying  some of its critical semantic and conceptual distinctions.  It's more parlor trick than intellectual move.  I heard him do this the other day on NPR's "Talk of the Nation."  Near the end of the section a caller had made the claim that one ought not to vote narrowly on his or her own interests.  Some minutes later, even after another caller had spoken about a different issue, Stanley Fish returned to make the point that it's just false that you cannot vote your own "interests."  It's impossible, in other words, not to vote your interests.  

For those familiar with a little bit of Plato and Aristotle, this sounds a lot like the following: it's impossible, so said Socrates, to know the good and not do it.  If you don't do the good, you don't know what it is.  Those in philosophical land will recognize this as the problem of akrasia.  They will also notice that there might be any number of plausible interpretations of Socrates's position.  Let me draw on one for the purposes of illustration.  You will always do, Socrates seems to say, what you view as the good.  Even if its bad, you view it as the good.  It's the good because all actions aim at the good.  Even if the good is bad.  For you it's the good.  See?

Nor did Aristotle.  He said this view "contradicts the plain phenomena."  People do all the time what they know they shouldn't be doing. 

So on one reading, the Socratic position plays on a semantic ambiguity in order to claim that "doing the good" is a "definitional"  or "analytic" truth. 

Having said all that, let's return to Stanley Fish.  In his column in the New York Times he makes a similar point:

We should distinguish, I think, between two forms of identity politics. The first I have already named “tribal”; it is the politics based on who a candidate is rather than on what he or she believes or argues for. And that, I agree, is usually a bad idea. (I say “usually” because it is possible to argue that the election of a black or female president, no matter what his or positions happen to be, will be more than a symbolic correction of the errors that have marred the country’s history, and an important international statement as well.) The second form of identity politics is what I call “interest” identity politics. It is based on the assumption (itself resting on history and observation) that because of his or her race or ethnicity or gender a candidate might pursue an agenda that would advance the interests a voter is committed to. Not only is there nothing wrong with such a calculation – it is both rational and considered – I don’t see that there is an alternative to voting on the basis of interest.

The last claim–there is no alternative to voting on the basis of interest–has that "analytical" ring to it.  Notice, however, how Fish uses that broad analytical sense of interest to make the more narrow claim that one must vote for one's identity interests. Fish ought to know that these are two rather different senses of the term "interest."  But he doesn't.  Following this he asserts:

The alternative usually put forward is Crouch’s: Vote “for human qualities” rather than sectarian qualities. That is, vote on the basis of reasons everyone, no matter what his or her identity, will acknowledge as worthy.

That really isn't the real alternative.  The real alternative would return to the sensible discussion of interest.  If we grant that it's analytically true that everyone votes his own interest, we can put aside the question of interest as telling us nothing interesting, and return to the discussion we were having before–which of my many interests ought to be the deciding factor in voting in a democracy?  My economic interests?  My racial identity interests?  My religious interests?  My professional interests, my family interests, my friend's interests, my leisure interests, my civic interests?  Knowing that I must vote for one, because that's the nature of reality, doesn't help me figure which one.