blog.kennypearce.net

Representative Realism, Phenomenalism, and "Physical-Talk"

It has been over a month since my last post, and for this I apologize. I doubt if I will be posting any more frequently in the near future as I am getting married on August 2 and moving from Philadelphia to Los Angeles immediately after the honeymoon. I'm sure the Internet will get by just fine without me.

Right now, however, I do have a bit of time, and I want to discuss an argument for phenomenalism about the physical world. When I wrote a while back about the idealist strategy, I said that the second step was to "argue that our physical statements - both ordinary statements about physical objects and statements about the discipline of physics - are best construed as talking about perception." What I want to do here is to unpack this statement. First, let's examine what the argument is supposed to do, and then we'll look at the argument as it appears in a brief section of Berkeley's Three Dialogues.

This piece of the argument is a reductio against representative realism. The first step of the idealist strategy is supposed to eliminate direct realism (the view that the very same things we experience in sense perception exist mind-independently and are known by us directly). I will assume this has already been accomplished. This leaves representative realism, the view that our perception are representations of mind-independent reality.

There are effectively two flavors of representative realism, both of which are, I think, fairly popular among philosophers today. The first is causal representation, which claims that our mental states come to represent things in the world in virtue of having been caused by them. This view has been supported by Fred Dretske. It has some problems which many philosophers have tried to shore up by a variety of strategies. The most important problem for it is the possibility of misrepresentation - e.g., how can we mistake a cow for a horse (from a distance, in the dark) if horse-thoughts represent horses precisely because they are caused by horses (but this one was caused by a cow)? I will not dwell on this objection, but there is a vast literature on it.

The second flavor is primitive or mysterian representation. This view takes representation as a primitive -i.e. one of the fundamental concepts of the theory, which does not admit of further analysis. The main objections to this view have to do with (1) whether you can adequately define the formal properties of representation in a coherent fashion, and (2) whether representation makes a good primitive. The latter is probably the most important, but the question of what makes something a good or bad primitive is extremely complex.

For the idealist's purposes, what matters is that when I perceive a table, there are two things: the 'real' table, and my perception or representation of the table. These are not the same thing. This much is conceded by the representative realist. It is customary to refer to the mental tokening which represents the table as a 'table', after the way we discuss words in philosophy of language, but this is going to get really confusing in this particular argument, so from here on out I will use table_i to refer to mind-independent table objects, table_m to refer to mind-dependent table-representations, and 'table' to refer to the English word spelled t-a-b-l-e. (I'm not sure how much less confusing that will be, but I'm hoping it won't be too difficult to follow.)

Suppose the phenomenalist grants, for the sake of argument, that there is such a thing as a table_i and that, under ordinary circumstances, there is a one-to-one correlation between tables_i and tables_m. Now listen to Berkeley:

Ask the gardener, why he thinks yonder cherry-tree exists in the garden, and he shall tell you, because he sees and feels it; in a word, because he perceives it by his senses. Ask him, why he thinks an orange-tree not to be there, and he shall tell you, because he does not perceive it. What he perceives by sense, that he terms a real being, and saith it is, or exists; but that which is not perceivable, the same, he saith, hath no being. (Three Dialogues Between Hylas and Philonous, 234)

The disagreement between the Hylas and Philonous at this point is that Hylas supposes that 'cherry-tree' refers to the cherry-tree_i, whereas Philonous believes it refers to the cherry-tree_m. The argument, in other words, is that representative realism provides a poor analysis of "physical-talk". By "physical-talk" I mean the ordinary statements of non-philosophers about physical things, whether in the context of physics or everyday life.

The realist needs to argue that 'table' refers to the table_i. Now, Berkeley's principal target is Locke, and this argument immediately overcomes Locke. Consider:

Let us then suppose the mind to be, as we say, white paper, void of all characters, without any ideas; how comes it to be furnished? ... To this I answer, in one word, from experience ... Our observation employed either about external, sensible objects; or about the internal operations of our minds, perceived and reflected on by ourselves, is that, which supplies our understandings with all the materials of thinking. (Essay Concerning Human Understanding, 2.1.2, emphasis original)

Locke isn't a representative realist. Rather, he holds a hybrid view according to which some properties ('primary qualities') are perceived directly and others ('secondary qualities') mediated by a representation. If direct realism falls (and we are assuming that it has already fallen), then no object can be both external and sensible, so Locke cannot claim that our "physical-talk" refers to external objects.

More recently the cause has been taken up by Kripke:

When I refer to heat, I refer not to an internal sensation that someone may have, but to an external phenomenon which we perceive through the sense of feeling; it produces a characteristic sensation which we call the sensation of heat. Heat is the motion of molecules. (Naming and Necessity, 129)

Now, recall that under Kripke's theory of reference, an object or natural kind is named by an initial 'baptism' in which someone points to it (literally or figuratively) and gives it a name. The name is subsequently passed on by the speaker community from one person to another. So, under Kripke's theory, how would 'heat' acquire the referent heat_i? Nobody can point to heat_i, because nobody perceives heat_i ("molecular motion"), we all perceive heat_m ("the sensation of heat"). Kripke says: "In the case of a natural phenomenon perceptible to the senses, the way the reference is picked out is simple: 'Heat = that which is sensed by sensation S'." (ibid. 136) This, however, is simply a statement, not an argument. To say "that which is sensed by sensation S" is the same as to say "that which is represented by sensation S".

The phenomenalist wants to argue that this is not a good analysis of 'heat'. Heat_i isn't a sensation. It can't be felt. If you ask the gardener to define 'cherry tree', he will describe a cherry tree_m: something that is seen, felt, smelled, etc. If you ask an ordinary person to define 'table', she will describe something that looks and feels (and therefore is) flat, that you can set objects on, etc. No one who has not been reading Aristotle, Locke, and friends will say anything about a "material substratum." No one will say "the object that causes my table perceptions." The table doesn't cause something to feel flat, the table itself feels flat.

Physicalists tend to be very adamant about believing only in the objects of their senses, but then begin describing things that can't be sensed at all, and claiming that those are the objects of their senses. If the phenomenalist can make this case that physical-talk is best understood as referring to objects_m, then matter will be superfluous to metaphysical explanations of the world we experience. Furthermore, if Kripke's "pass-through" reference fails, then his theory will make it impossible to refer to objects_i, for the same reason it is impossible for Putnam's brains in vats to wonder whether they are brains in vats.

Posted by Kenny at June 5, 2008 6:12 PM