Philosophy and semantics

So my philosophy 102A class “Reality and Metaphysics” has so far been pretty lackluster, but tonight I was actually “on the edge of seat” for about 45 minutes. Proof there really is a god who takes pity on the insanely bored? No – proof that even the most mundane, on the verge-of-misery-and-torture class can have a decent session.

Anyway, the topic was “consciousness” and after talking about Functionalism and Identity Theory, we came to Connectionism. To paraphrase, Connectionism basically suggests that the brain works like a massive computer application, made up of millions(?) of functions and operations (chemical, electrical, etc.) That work in concert to create thoughts, emotions, ideas, etc etc.

It makes perfect sense to me, possibly more so because I’m an “aspiring” Comp Sci. student.

So the Professor brought up this guy John Searle. He came up with a pretty fascinating thought experiment called the “Chinese Room.” Basically (paraphrasing again) it goes something like this:

A man is trapped in a room with lots of small balls that have chinese pictographs (characters) printed on them, along with an instruction manual. There is a hole in the wall that more balls enter through, and another hole in the wall that he sends balls out through. His instructions are to take the balls entering through the “input” hole, read the appropriate instruction from the manual, and send a corresponding ball out through the “output” hole, basically acting as a human sorting program.

Searle’s argument is that no matter how many pictograms the guy processes, he’s never going to learn Chinese.. and that syntax in no way implies semantics. He also relates this to computer AI, by saying today’s computers couldn’t figure out Chinese under these conditions either.

I’ll really have to look into some arguments against Searle’s “Chinese Room” but I was just sitting in class wondering: If the guy is really interested in trying to figure out the language, couldn’t he look for patterns in the process? Then he could start writing he own rulebook/dictionary possibly.

I think Searle is/was probably right about the limitations of today’s computers though, because I don’t know much about etymology or neural networks, but if I had to guess, I’d say that to simulate a brain with computer technology, you’d need:

relational databases with the ability to instantly reference any given object related to the current “thought”,

functions that were aware of each other, and constantly watching and evaluating each other’s operations/expressions down to the use of operator.

the ability to index data not only by it’s characteristics, but by possible contexts it could be used in.

Basically, I picture all of these things put together like an orchestra playing together which produces a song where the harmonics of each of the instruments mesh with each other to create a cohesive, purposeful sound that could never be achieved by any of it’s individual players/instruments. An intelligent thought, basically.

So yeah, that’s just a sci-fi analogy, but even that I think would be difficult at best to setup, nowadays. But then again, I read that they have simulation programs that can realistically model the hydraulic systems of an entire aircraft carrier, so maybe we’re not far off from something like that.