h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

I wonder just what Searle is trying to defend.

Wittgenstein pointed out the difficulty of describing thought. He said, If I ask two people a difficult (say mathematical) question, and one gives the answer effortlessly, while the other strides up and down wrinkling his forehead, pounding his fist into his palm, saying "don't tell me . . . " and generally working up a sweat (I paraphrase) which one is better at thinking?

Wittgenstein's description of "thinking that" is more like "behaving as if". The thing itself (thought) is not observable, only the results.

Turing was a pupil of Wittgenstein's for a while but I don't think they saw eye to eye; a pity, they could have stimulated each other. Maybe they did even on their slight acquaintance.

And the Wittgenstein test? "If a lion could speak, we couldn't understand him". Our language-behaviour is learnt by being a person in normal human relationships with other people. A machine can learn a lot of linguistic manipulations, but it is unlikely to learn enough to mask its lack of childhood memories, musical tastes and the rest, from an inquisitive questioner.

And yet I have no trouble saying machines think. If I log on as my son, the computer thinks I'm him. Why is it a big deal? Machines can't be human, but that's a different question.

And the Searle test is fatally flawed; he plays down the wonderful pages that show him how to answer questions in Chinese. *Any* questions? What sort of magical page is that? Another Wittgenstein quote: "The result of an imaginary experiment is not the result of an experiment."

Searle is trying to attack behaviourism, and claim that even if a machine and a person are functionally identical, it's still reasonable to claim that the person is intelligent and self-aware, and the machine is not.

I wasn't aware that Wittgenstein was a behaviourist. I do agree with him that language is conditioned on experience - you can often tell a lot about people, from their choice of words, sentence length, use of cliches, and so forth. No doubt many intelligent machines will similarly have a pronounced "accent".

>I wasn't aware that Wittgenstein was a behaviourist.

He certainly wasn't! You don't have to be one to disagree with Searle. There aren't many behaviourists left, though I have met one. His faith is touching.

See my entry A1024156 "Ludwig Wittgenstein" if you are interested.

I am largely in agreement with Steven Pinker, who attacks the behaviourist idea (the brain's infinite malleability) in "The Blank Slate". Pinker is not on Searle's side either.

Oh yes, I see your quote from him isn't as strong as I first read.

Perhaps I am misusing jargon. I always understood behaviourism to be the belief that something is intelligent if and only if it displays intelligent behaviour. Why does one require the idea of an infinitely malleable brain to believe that?

Behaviourism is something else; the theory that "mind" and such noumenous words have no usable meaning; one can only meaningfully discuss (a) things and (b) observed behaviour. Google for B. F. Skinner.

I don't see the distinction between this correct definition of behaviourism, and my incorrect mis-definition of it. You are saying that, under behaviourism, one can meaningfully discuss intelligent behaviour, but cannot meaningfully discuss intelligence qua intelligence. My understanding of behaviourism was that it defined intelligence as intelligent behaviour.

Ignoring semantics, these seem identical to me. I tried googling, but didn't find much helpful: Skinner seems to have had views beyond mere behaviourism (that all behaviour is conditioned, for example), which are now discredited, but I don't see that discredits behaviourism.

What am I missing?

Well maybe I am misreading. But to me the word "Behaviourism" means the denial of all things mental.

Wait, . . . "the behavorists tried to explain learning without referring to mental processes . . . learning is seen largely as a passive process in that there is no explicit treatment of/ interest in mental processes. The learner merely responds to the "demands" of the environment"."

In any case it would be highly misleading to describe Wittgenstein as a behaviourist. Even if in his early days he said that we couldn't meaningfully discuss the mind, he certainly didn't deny it.

But by your definition . . . well, it's very interesting, I see what you mean! (I think)

Perhaps it's like the difference between atheism and agnosticism?

No, that one's easy. Atheism is "There are no gods", agnosticism is "There is no knowing".

It may appear a similar distinction -- Behavioursm says "There are no minds" and Wittgenstein says "There is no talking about them". But Wittgenstein based his mystical stance on the fact that mind is undeniable.