Putnam's Pragmatic Realism

Putnam originally called it internal realism, which suggests Carnap's internal/external distinction. Like Carnap, Putnam thinks that what is real depends on what language you adopt, that is, in Carnappian terms, your framework.

However, unlike Carnap, Putnam is a realist (though, he says, not a Realist). That is, roughly, Putnam is an internal realist but not an external Realist. That also sounds like Carnap: Carnap thinks external Realism is incoherent, and, like Putnam, Carnap thinks that once we have adopted a language, that determines, almost trivially, what is real. So where do they differ? Carnap thinks that internal realism doesn't deserve to be called a form of realism, he thinks it is just a pragmatic adoption of a language; Putnam thinks that internal realism is a genuine form of realism, indeed the only tenable form of realism. What they differ about is "only" what counts as realism.

To put it starkly, Putnam's realism is almost identical to Carnap's antirealism. What they differ about is not how they take the world to be, but what they take the term "realism" to imply. (I'm exaggerating.) This makes very clear what I said last time, that there is little agreement, and too little explicit discussion, of what realism is, before arguing for or against it. (I don't really think Putnam is guilty here, he is, in large part, using his advocacy of pragmatic realism to argue for a certain view of what it is for a view to be a form of realism. It is very hard to talk about such matters directly, and so sometimes it is easier just to use an example as a stalking horse.

In saying he should call his view "pragmatic realism" Putnam makes himself sound even more like Carnap. There are a couple of telling differences in the examples Putnam and Carnap choose.

Carnap: In one framework, there are physical objects, but, in another, only sense data.

Putnam: "Goodman vs Lesniewski": Given three "individuals," are there three objects or seven?

There are several striking differences between the two examples:

1. The Carnap one is one that people have deeply cared about, while the Putnam one is one in which the two views are virtually interchangeable.
2. Carnap thinks we choose a language on pragmatic grounds, while Putnam thinks that, at least sometimes, there isn't much to distinguish the languages—even God couldn't tell you which one is better.
3. Carnap seems to think of his alternative frameworks as "describing" different worlds; while Putnam seems to think of his as being different ways of describing the same world.

In the sense data vs. physical objects case, Putnam might reasonably say that these are two different descriptions of a common world: after all, when, in the first, you sense some chairlike sense data, in the second you see a chair. The case of mathematical objects vs. none seems like a harder case for Putnam to handle.

How can such a difference in emphasis make the difference between a realist and an antirealist view? To make sense of that, I, at least, need to turn to Dummett.

Putnam claims to be, in appropriate circumstances, a realist about ordinary causes (the valve getting stuck caused the pressure cooker to explode), a realist about mereological sums (Lesniewski's seven objects), colors, and intentions. In other circumstances (when doing particle physics), he claims not to be realist about any of them. So, when is he a realist about an ordinary cause? Do Dummett's criteria apply?

The first thing to note is that Putnam's questions are very different from those of Dummett. He is trying to make sense of how anything short of the objects of particle physics can, not be real, be have any status at all given standard realist views about physics. The usual story developed by Descartes, Hobbes, and Newton is that the objects of physics (henceforth, physical objects) do not have, for example, colors, only "powers to produce colors," or, as we say today, dispositions to appear colored in suitable circumstances. The problem with that is that the powers, as it happens, are no more candidates to be physical properties than the colors were, and we are left without any way at all to talk about colors. The problem is much more serious than a mere failure of realism about colors. The inevitable conclusion seems to be that there aren't any colors at all in any sense. (Same with causes, dispositions, and intentions.)

Dummett isn't concerned with a particular realism and he is only concerned with realism, not the further, more serious problem Putnam raises. Look at what I said: "we are left with no way at all to talk about colors." Putnam, while talking about realism is driven by epistemological questions: not whether or not something exists, but how we could come to characterize it or know that it exists. Dummett only discusses that kind of issue briefly on the last two or three pages, where he says that an acceptable metaphysical theory, realist or not, must make it possible to see how we could come to know it.

Let's apply Dummett's tests to, say, Putnam's realist account of ordinary causes. The second test is easy: Putnam's explanation of what makes the valve getting stuck a cause certainly refers to the valve. The hard test is bivalence. Does Putnam think that statements about ordinary causes are either true or false? In addition, if he is arguing for realism, then, according to Dummett, he should be arguing for bivalence. What kind of example does Putnam consider? Could (a random part of the surface of the pressure cooker) be the cause instead? From basic physics, it probably is also a cause, with the result that there is no single cause. It is our interests and expectations, our knowledge of the function of the valve that makes it the cause instead of . What has that to do with bivalence?

-- ShaughanLavine - 12 Mar 2008