Dennett's Where Am I Split a Person Across a Brain Vat and a Radio Link

In 1978 Daniel Dennett published a thought experiment in Brainstorms: Philosophical Essays on Mind and Psychology (Bradford Books, MIT Press) under the title “Where Am I?”. The narrator’s brain is removed from his skull, kept alive in a vat, and connected to the body by radio links to the nerve endings. The body keeps walking around under the narrator’s name. The question in the title is asked by the person who can no longer answer it by pointing at himself.

The split and what it exposes

Dennett sets the experiment up so that the ordinary answer fails. The narrator is told that he is the one who will remain awake while the operation happens, and then discovers that “remaining awake” has been done by the body while the brain did the going under. He then discovers that the radio link means he can be located with the brain and with the body at once, and that the two come apart.

The story gives the three pieces separate names, after the parts of Hamlet. Yorick is the brain in the vat. Hamlet is the original body. Fortinbras is a second body, built later and connected to the same brain after the first body is lost. The naming is doing argumentative work. Yorick is the skull that held the thought, and Hamlet is the man who walked around thinking it. Once they are separated there is no fact about which one is the narrator, and Dennett makes the narrator notice this and keep asking.

The value of the setup is that it removes both of the usual answers at once. You are not your brain, because the brain is somewhere else and you are still able to speak through the body. You are not your body, because the body is replaceable and Yorick persists. What is left is a set of functional relations and no further fact to appeal to.

The connection to gradual replacement

Dennett’s case is a sudden split rather than a gradual one, and the difference is the reason the two are usually treated as separate problems. In the gradual case there is never a moment when two things are both claiming to be the person. In Dennett’s case there is a moment when two things are plainly both claiming it, and the claim of one of them, the body, turns out to have been made all along by something that was never consulted.

The site’s treatment of the Moravec transfer sets out the three assumptions that gradual replacement rests on. Dennett’s case is the sharpest available test of the third of them. If continuous transition preserves the same subject, then there is no answer to “where am I” while the transition is running. Dennett’s narrator refuses to accept that and keeps asking, which is the point.

There is a second connection that is less obvious. Dennett is an illusionist about consciousness and he denies that there is a place where it happens. On that view the question “where am I” is malformed from the start, and the discomfort the story produces is the discomfort of asking for a thing that is not there to find. The analysis of the Multiple Drafts Model sets out that position in full.

Comparison to The Consciousness AI

Dennett’s case is a test of a design claim this project makes. The project treats the organisation of a system as what carries mind, rather than the material. If that is right, then relocating the processing while preserving the organisation should preserve the person, and the radio link should be enough to keep one subject across two locations.

Nothing in the codebase demonstrates that. The project’s Global Workspace layer and Affective Core are functional organisation and they can in principle be described independently of the hardware they run on. Whether that description carries a subject is an open question here in exactly the way it is open in Dennett’s story. His narrator’s refusal to accept “you are wherever the processing is” is a design objection and it should be read as one.

The interface detail in the story is also relevant and easier to settle. The radio links are a separable channel between the processing and the world. That is the same structural claim as Bernal’s proposal of undifferentiated nerve connections that can be given new end organs, worked out in the analysis of the 1929 case for removing the body. The channel is a design component. The subject is the unresolved part.

What the experiment does not establish

It is a thought experiment and Dennett gives no mechanism for the radio link and no evidence that a brain in a vat would report what his narrator reports. The naming does argumentative work but the argument is a question rather than a proof.

The story also does not separate copying from continuing. Dennett’s Fortinbras is a second body on the same brain, so it is a relocation rather than a copy. The case where two brains are produced and both claim to be the person is a different case and Dennett leaves it alone. That is the case Parfit builds in Reasons and Persons, and the analysis of Parfit’s branch-line case sets out why he thinks the question is the wrong one to ask.

What survives is the question in the title and the fact that it cannot be answered by pointing. Dennett made a philosopher ask where he is and then refuse to accept the two available answers. Twenty two years later Hans Moravec turned the same refusal into an engineering procedure.

Where Am I? by Daniel C. Dennett appears in Brainstorms: Philosophical Essays on Mind and Psychology, Bradford Books and MIT Press, 1978. It was reprinted in Part IV of The Mind’s I (Basic Books, 1981). The story names its three pieces Yorick, Hamlet and Fortinbras after the parts of Hamlet.