next up previous
Next: Date: Sat1 Jun Up: Discussion Previous: Reference

Date: Sun, 12 May 1996

Some while ago, the organisers of the Monist discussion groups suggested that we all close down for a while to assess progress and re-organise. It turns out that some groups have been quite active, while others have had almost no discussion. After reviewing the archives of the other groups, ours seems to fall about in the middle level of activity.

Here is how we are going to proceed from now on. It seems obvious from the discussion so far that we should focus on issues concerning science, logic and philosophy of mind. Anyone wanting to add something in other areas is welcome, but a diffuse focus doesn't seem to generate much discussion, so we'll concentrate on these harder-edged areas of philosophy.

Let me pick up the discussion by responding to some of the things that Jim Fetzer wrote in his piece on non-propositional thought processes. It was certainly valuable to remind us of Peirce's distinction between iconic, indexical, and symbolic signs, and to note the strange obsession that contemporary philosophy has had with the last. I fully endorse his critical remarks about Nelson Goodman's ``worldmaking'' views. The (surely by now fading) view that everything we know has to be encoded in some language has led to many a reductio, Goodman's being one of them. Another version of this view, which gained a certain amount of notoriety in the 1980's when Putnam popularised it, repeats the most straightforward consequence of so-called basis theorems in first-order logic, which is that the intrinsic ontology of theories is completely under-determined by the formal structure of the theory. (The basic argument can be found in Putnam's ``Models and Reality'' paper - Journal of Symbolic Logic 45, 1980, pp. 464-482, - but as Bill Demopoulos and Michael Friedman pointed out,- Philosophy of Science 52, 1985, pp. 621-39 - Max Newman had made essentially the same point back in the 1920's. It is also a staple of Carnap's Aufbau (1928), and Carnap is, in my view, the source of most of Quine's (and hence Goodman's and Putnam's) views on ontological relativity).

What is relevant for our purposes here is that Putnam had to claim that causal relations were also linguistically and theory dependent, and hence subject to the same relativistic arguments. So if you want to introduce indexical signs, as Fetzer does, you not only have to deny this claim but to provide some grounds for the claim that we know causal relations other than through theories of causation. Fetzer, I'm sure, is capable of giving such grounds. My own preferred way of dealing with this is to simply point out that any empiricist who is willing to admit perceptual evidence as grounds for knowledge ought to seriously consider allowing our ability to discriminate between causal and non-causal relations (at least in simple cases) as legitimate grounds as well. This discriminatory ability is acquired very early, perhaps even pre-linguistically, and in any case well before any serious theory or even terminology of causation becomes part of a human's linguistic apparatus.We can tell from experience itself which are causal relations and which are not, and this is not just practical knowledge (a.k.a. ``knowing how'') but what is misleadingly called ``theoretical knowledge'' (a.k.a. ``knowing that''). (Caution: let's avoid the simple-minded trap of thinking that anything that can be put into a propositional attitude mode, such as ``knowing that_____'' is ipso facto necessarily propositional in form. All that approach gives is that you could represent the knowledge that way, not that you must, nor that such a representation is correct.) I know that fire causes burns on my skin, and if pre-linguistic hominids knew anything, they surely knew that. If this sounds plausible, and it should, indexical signs are a distinctively different source of knowledge than symbolic signs, and are not subject to the well-known relativistic deficiencies of the latter.

One further point along these lines. The above line of argument does not give us an argument for any sort of representational realism, if only because nobody expects effects to resemble their causes. But this is irrelevant to the fact of the existence of non-propositional knowledge. Indexical knowledge is (or can be) non-propositional knowledge.

pwh2a


next up previous
Next: Date: Sat1 Jun Up: Discussion Previous: Reference

hh
Fri Jul 25 22:00:35 MEST 1997