Monthly Archive

I don’t think this is philosophically very important, but it struck me as interesting.

I was reading Jerry Fodor’s Psychosemantics and came to the following claim: that the possibility of making an error requires the possibility of being correct. My immediate thought was that, in form anyway, this is pretty much the converse of a famous thought of Wittgenstein’s. He holds, in Philosophical Investigations for instance, that it can only make sense to be correct about something or other if it makes sense to incorrect about it. As I said, I don’t think much comes of this but it was thought provoking.

I’ve been hearing a lot about functionalism recently. It’s in Frank Jackson’s fascinating From Metaphysics to Ethics. It was mentioned in an M Litt seminar on the Philosophy of Film given by Hans Vodder today. Of course it crops up in the Philosophy of Mind now and again. I’ve had a few thoughts about these occurrences. Here they are, in no particular order.

It seems to me that ‘functionalism’ can be used to gesture towards several sorts of idea. Of course it gets turned into specific well-worked theories as well, but I’m not so much interested in the details. The first thing that it’s used to gesture to is the identification of the elements of some system by the role they play in that system. Whatever it is that plays the role of a X is a X, or has the property that allows it to realise Xs role. I think that this is the main thought that goes on in the cases I’ve recently come across. It certainly seems to be what supports the ‘multiple realisability’ claims that are taken to be the main selling point of functionalism. At least they were in Hans’ talk.

There is another thought that often seems to be operating, and that is functionalism applied not to parts of systems but to whole systems. That is the thought that if a system behaves like a X, then it’s an X. This is what is being got at if functionalism is associated with the Turing test for intelligence.

Both of the above are just applications of the term to slightly different things. They’re not likely to be confused and so there’s no problem.

Finally there is the thought that functionalism implies a level of mutual interconnectedness. This is part of what is going on in Jackson. I think this is the furthest from the classical usage. These cases might be better understood in terms of ‘reflective equilibrium’ or ‘accommodation and negotiation’. At least I think so.

As noted on Ian Church’s ‘Reformed Philosophy’ blog, there is a prima facie problem for foundationalists: how to resolve disagreement. The problem would arise because both sides would accuse the other of begging the question by appealing to their foundational beliefs. Let us distinguish between base and superstructure, the base beliefs are held non-inferentially and the superstructure is inferred from base beliefs and rules of inference. In the cases that lead us to be having this discussion the base is some theological system and the superstructure is an epistemological one, but that’s not essential to the debate.

The problem is understanding how two people with different bases can have a meaningful debate about superstructures. It seems that each party believes what they believe for reasons that the other will not accept. Any arguments they offer will be met by a charge of begging the question. What to do?

Ian’s suggestion is to appeal to the following method of resolving conflicts. A conflict is resolved if one side can show that the other is internally inconsistent. I think it’s implicit in what Ian says that we imagine this process to be iterated and that one belief system will emerge from the process as undefeated. As he notes, this relies on the premise that there is at most one consistent base. I think that, if the base is assumed to contain or imply every atomic sentence or its negation (or something of the sort), this is right.

There are a couple of points to make about this, which I think make the view look pretty costly. The first is that it only works if we make the world views extremely comprehensive. This is required by the condition sketched at the end of the previous paragraph. If this is accepted then the other assumption of the optimistic view looks threatened. Why shouldn’t there be more than one internally consistent set of base beliefs plus inference rules? Presumably the world is consistent, but there could be more than one system of reasoning from what is given by it. This problem would be resolved if the inference rules were fixed. But trying to do so will bring back the charge of question-begging. Whoever has their favoured inference rules questioned will cry ‘foul!’