Functional Theories of Consciousness: Conflict vs Cooperation

[benedictlentil]

Julian Jaynes does a decent job describing consciousness functionally as a mental metaphor of one's self as an opaque container that contains one's feelings, thoughts, and intentions, so they're immediately available to you but not necessarily to others. Carruthers's notion of consciousness as the information globally broadcast within a multi-module mind seems also useful but distinct. We might want to think of Jaynes's notion as specifically denoting self-consciousness. I recently realized that I know two very different stories for how self-consciousness can arise.

Cooperative Theory of Self-Consciousness

Bands of animals often want to pass information between them, e.g. distress calls or a simple "here I am" call that allows members of a band to locate themselves relative to others. (Pygmy Marmosets learn distinct calls for their particular band.) As the information-passing system allows for increasing detail and precision, it eventually becomes capable of reference and recursion, i.e. universal grammar.

An universal language might start out with no idea of the self-as-describer, but only of the external phenomena described. But one of the things that an animal node in an information-processing network trying to keep itself together and alive is going to want to refer to is the nodes in the information processing network. And once you can talk about that, you're going to want to on occasion be able to refer to "this node, the one generating this message" with a short message length. This is enough to generate indexicals (I, you, they), and an idea of the self with privileged access to some intermediate results of local computation.

Conflict Theory of Self-Consciousness.

When one group conquers and enslaves another, the conquered would do well to pretend not to be the enemies of their oppressors, which requires keeping two distinct sets of books - what your thoughts, feelings, and intentions would seem like from the outside, vs what they "really" are. This requires and thus selects for subjective self-consciousness.

[ambimorph]

As the information-passing system allows for increasing detail and precision, it eventually becomes capable of reference and recursion, i.e. universal grammar.

I think this entails a proverbial mathematical miracle step.

Recursion seems to require not just reference, which can be between an utterance and a concept, but symbols, by which I mean the ability to refer to words with other words. This has superlinear scalability problems that force the partitioning of words into semantic and syntactic components.

I have tried to defend Terrance Deacon's thesis on this here: http://subsymbol.org/posts/2022/08/28/s ... rence.html

ETA: of course this has no bearing on the distinction made in the original post. Just eager morning thoughts.

[benedictlentil]

Hmm that puts me in mind of the Greek words έπος and λόγος, which both eventually mean “word” but seem to originally have been distinct ideas - the former to a phoneme cluster or discrete orally produced sound, and the latter to a physical cluster or perhaps bundle, related to its verb form which means “gather.” And then there’s ίδέα or έίδος, which was an image or “look” before it was an idea. AP David plausibly argues in “The Feminine Abstract and the Neuter Essential” (unfortunately I can’t find it online, though I found a review summarizing it here) that Socrates was talking with people who had the idea of words, speeches, and arguments, but perhaps not of essential ideas as the generators of patterns of reference.