Storobia Archive

Searle's Chinese Room

The Chinese Room - sometimes known as the Chinese Box - is a classic AI thought experiment proposed by John Searle in 1980. It has been the subject of much debate ever since and - like most ideas in philosophy - has produced more questions than answers.

What Is The Chinese Room?

Discussion of the Chinese Room often gets into detailed technical discussion of AI and philosophy, however the principle is simple. Searle's concept was in essence this: imagine that an English speaking man who knows nothing about Chinese is sitting in a room. He has with him a complete set of Chinese dictionaries, books of grammar and lists of rules for translating between English and Chinese. Chinese sentences are passed to the man under the door, he looks up the symbols in his books, follows his rules and writes answers which he passes back under the door.

Clearly the man himself does not understand the Chinese language, he's simply following rules. Therefore would a computer following the same rules be said to understand Chinese?

The next step is obvious: imagine we scale the experiment up so that instead of the "room" simulating a Chinese interpreter it simulates every facet of a particular Chinese person. Of a particular mind. Could the room be said to "be" intelligent in the same way as a mind?

Searle's position is that the Chinese Room does not "understand" Chinese since it only knows the syntax and has no comprehension of the semantics. It it simply manipulating the symbol in a mindless fashion. For Searle this is a refutation of the strong AI premise that a sufficiently well programmed computer could said to "be" an intelligent mind.

The main refutation of Searle's position is to argue that the Chinese Room is a system and that the system as a whole understands Chinese even though the man at the centre does not - he is just one part of the system. The "semantics" of Chinese are encoded system wide in the interaction between the dictionaries, books of grammar and rules.

There are, of course, refutations to the refutation. And refutations to the refutations to the refutations... and so on ad nauseum. Which is the main reason the Chinese Room remains so interesting. Of itself it doesn't "prove" anything, however it does provide a scenario on which to sharpen our definitions of concepts such as "understanding" and "intelligence".

Arguably the whole debate is irrelevant. The more interesting topic - at least to me - is not artificial intelligence but artificial consciousness. Even those who believe that the Chinese Room does understand Chinese are unlikely to classify the system as conscious.

This is an old page archive from Storobia. Please read the site terms of use