I’m not really a verificationist (see here for the beginnings of my account of content), but I have a lot of general sympathy for verificationism, and I think the verificationists were certainly onto something both true and important, even if they didn’t get the formulation quite right.
I am not a metaphysical anti-realist, though, and it’s always bugged me quite a bit that realism and empiricism (especially verificationism) are contrasted. Which they are, regularly, by realists like Boyd, and even by empiricists like Van Fraassen (who isn’t actually an anti-realist, but believes that the truth of scientific laws isn’t something we should be concerned with).
In case you’re not quite familiar with these terms (at least, as they’re used in a philosophical context): A verificationist is someone who believes that the content of synthetic sentences (those that make actual statements about the world, rather than merely following from the definitions of their terms) is their verification conditions: The hypothetical situations under which we could verify their truth. Such a statement, according to verificationists, is true if, and only if, the universe is such that we could, at least hypothetically, verify it (perhaps we can’t actually verify it, because, say, it’s occurring on a distant planet, but we could verify it if we were in a different situation).
A realist (in the most typical sense; more specifically a person like this is called a “scientific realist”) is someone who believes that statements of science make claims that are true (at least, that are supposed to be true–it’s not that they believe scientists can’t make mistakes) about real objects (or at least objects that are supposed to be real–though again, realism doesn’t mean you think scientists can never be wrong about what exists). In particular, this includes unobservable objects such as electrons.
Now, before I even talk about how a verificationist can be a realist, I want to give a quick-and-dirty reality check that shows that verificationism can’t possibly imply anti-realism. It’s fairly similar to what Nathan Salmon said about the baptismal theory of names and essentialism (which I discuss here). Verificationism, you see, is a statement purely about language: what sentences mean, and under what circumstances they are true. Anti-realism is a statement about ontology: what does and does not exist. And there’s really no way that a statement about language can, by itself, imply things about what does and doesn’t exist. As Salmon said (in a different context), you can’t pull a metaphysical rabbit out of a linguistic hat.
So, how can a verificationist be a realist? It’s actually more of a mystery how a verificationist could avoid being a realist. Let’s return to the example above: Do verificationists like, say, Moritz Schlick (the verificationist’s verificationist) think that unobservable entities (like electrons) exist?
Of course they do. Look at the sentence, “Electrons exist.” Under what circumstances, to a verificationist, is this true? Under the circumstances in which we could (theoretically) verify its truth, in a completed science. At the very least, an approximation to this verification has been acheived: Scientists have done a large number of experiments and proven (to the extent anything can ever truly be proven scientifically) that there are electrons. Van Fraassen would call this empirical adequacy: The existence of electrons is our best explanation for the various phenomena we observe. He made a distinction between empirical adequacy and truth, but the most important thing about verificationism is that it doesn’t–truth simply is empirical adequacy.
So to a verificationist, because “Electrons exist” is empirically adequate, it’s true. And by simple disquotation, it follows that electrons exist.
This sort of thing applies to all the claims of realism–a verificationist should believe, pretty much, that all entities scientists talk about exist (modulo the fact that, of course, current science might simply be wrong, in fact almost certainly is wrong about at least some things). If those things are the best explanation for observable phenomena, then their existence has, by that very fact, been verified. That’s what “verified” means. So to a verificationist, they exist.
This seems so simple that surely there must be some positive arguments out there that weigh on the other side, and, in fact, there are. I’ll address some of them, once I manage to dig out my Boyd, in another post.
2 Comments
just to make sure i’m not missing something – embodied in that last move (all entities that scientists talk about exist) is that someone who models scientific theories linguistically, say, would need to include sentences like “electrons exist” as part of the model (or at least make it plain that those are implied by all and any statements that use nouns to describe objects). Is that right?
Does this mean that causation also has to be real?
Well, yes, that’s right, but it’s putting it in a slightly less realist-sounding way. An important part of the verificationist belief system is that people who say, “OK, I guess modeling science linguistically requires saying ‘electrons exist,’ but do electrons *really* exist?” are talking gibberish. Having existence claims validatable through scientific methods, for a verificationist, is just what existence *is*.
And yes, it means causation has to be real, at least in the simple sense (and simple senses are really the only senses verificationists admit) that sometimes events cause other events. I mean, we might be wrong about this–it might turn out that everything in the universe just happens for no reason whatsoever–but that would be an *empirical* finding (and honestly, at this point, a highly unlikely one).
Post a Comment