The following argument originated by John Searle in Minds, Brains, and Science: the 1984 Reith Lectures, Penguin Publishers, London, 1989, page 417. The idea is reviewed by Jack Copeland as Chapter 6 of his book. The following is a minor rewording.

Set Up: Consider the following scenario to accomplish language translation from Chinese to English:

• An individual (called Sam in the literature) works in an enclosed room and has no knowledge of the Chinese language.
• Sam receives Chinese text through an input device.
• Sam then receives questions, also in Chinese, and asked to respond in Chinese.
• Sam is given an extensive collection of rules to convert the symbols he reads to any needed answers.

Searle's Argument: Searle argues that rules can dictate syntax, but not semantics. That is, the rules can indicate the form of input symbols, but not the meaning. Thus, Sam may be able to answer the questions (if the rules are good enough), but Sam has no idea what any of this means; Sam has no understanding.

Shifting to the realm of computers, Searle argues that computers may be able to manipulate symbols, but they will never know what they are doing. Thus, computers cannot be intelligent.

Copeland's Response: Copeland distinguishes between the role of Sam and the role of "Miss Wong Soo Ling", "who speaks to us — or rather writes to us — in perfect Mandarin Chinese." From the standpoint of the symbol manipulator Sam, there is no understanding. However, from the Miss Ling's perspective, Sam is providing enough information to allow Miss Ling to construct a satisfactory answer. Miss Ling has understanding, even though Sam does not.

Walker's Notes: Within a person, the role(s) of Sam are played by neurons. Information comes into the brain via the senses, the neurons react, and a response is formulated.

Although we recognize the question-answer format as a common mechanism to determine understanding, the individual neurons have no understanding; neurons just respond to their environment. Any understanding comes from the collection of neurons coming together, not from the wisdom of the pieces. Asking the neurons (or Sam) about understanding seeks wisdom at the wrong place.

created 8 January 2007 last revised 9 January 2007

previous   next