Brain-in-a-vat art by Pete Mandik, and not by some dumb AI.

Consider the claim “all talk of brain representations is erroneous.” Such a claim is made by someone floating the idea that maybe there’s no such thing as representations in the brain, that so-called neural representations don’t exist.

Such a claim is presumably itself about something—it’s about brain representations, and what it’sclaiming about brain representations is that they don’t exist. Whatever BRAIN representations would be, one thing they would be is a kind of representation. And note that the claim under consideration is itself a representation, since it’s about something. What else does being a representation require besides having aboutness? Absolutely nothing. Of course, a claim is a linguistic representation, and it exists as written text or spoken speech. Text and speech are extra-neural things—they exist out there, outside of the brain. But what is going on in the brain of the speaker or writer before they write or say “brain representations don’t exist”? What is going on in the brain of the reader or listener after they have received the message “brain representations don’t exist”? One good answer would be “nothing at all”. Just kidding. That would obviously be a terrible answer. Obviously something in the brain is happening in both sorts of occasions.

Now let’s consider some slightly different occasions. Suppose someone is about to say “brain representations don’t exist” but they wait, because a loud emergency vehicle is going by and they wouldn’t be heard if they spoke just then. While they’re waiting, what’s going on in their brain? It’s gotta be something, right? It can’t be nothing. Might it be something with aboutness?

Maybe what’s going on in their brain before they write or speak is itself a representation of, that is, a representation about what they are going to write or speak. And not just a representation about the utterance considered as sound—since presumably a parrot can squawk out something that parrotingly resembles spoken English—but a representation that represents what the utterance itself represents. This is what makes a human speaker’s utterance an expression of a thought that brain representations don’t exist. Further, its absence from the parrot’s little bird brain is what makes the similar sound issuing from the little bird’s beak NOT the expression of a thought that brain representations don’t exist.

Of course, a neural-representation eliminativist can hold without contradiction that while spoken and written text has enough representational status to allow them to assert the non-existence of neural representations, no event happening anywhere in any brain can achieve such status. Unfortunately, in science, just having non-contradictory views isn’t good enough. You need stuff like evidence and explanations. What evidence could there be that linguistic representations have aboutness that couldn’t similarly apply to neural representations having aboutness? What explanation could there be for how linguistic representations get their aboutness that wouldn’t similarly apply to neural representations?

Great questions! I’m glad I asked them. With my brain!

And now, some comics: