Responses To The Chinese Room
| Philosophy : Philosophy Articles The Chinese Wall is an interesting thought experiment designed to show that the computational view of the mind is incorrect or that alone it is not enough. Briefly, you feed Chinese questions into a room and get intelligent answers in Chinese out. However inside is an English speaker simply following a set of rules from an input/output manual and copying the output for each input. The conclusion Searle draws is that the English speaker in the Chinese room does not understand Chinese, despite being seen by those outside the room to know Chinese due to the quality of the answers given. The most common response and perhaps the best response to the Chinese room is the so-called systems response and variants thereof. The system response says that whilst the English speaker in the room (the CPU equivalent) doesn't understand Chinese, the system as a whole does (e.g. everything that together makes up the 'room'). The actual validity of the assumption that it is possible to output valid Chinese in the way a person would is actually up for grabs and can suggest that the actual room itself is not possible - in other words the Chinese room experiment is not actually possible and suggests that the pure input/output model in itself is wrong. For instance, imagine the sentence: wh?t is the s?nten?e Now, you can read that sentence. But imagine that you have a similar sentence fed into the Chinese room with a letter substituted for a character that you just made up, so won't be in any of the inputs for the system. Remember that as a non Chinese speaker you work purely from the syntax. If there is no matching input for the Chinese room which there cannot be as you made up the symbol, it would not be able to answer the question - hence it doesn't have understanding of Chinese as a Chinese speaker would be able to answer this. In other words, separating the syntax from the semantics makes a difference as to whether a system can 'understand' because it is through recourse to semantics that we help solve syntactic puzzles. The two are inextricably linked in understanding and a system that does not connect the two processes like the Chinese room will not be able to answer these questions. Similarly, if a question is stated ambiguously, then again human speakers can nearly always work out what the questioner meant from the context. However, if the Chinese room is set an ambigious question, would it really respond in the way a human would? The bottom line: when fed particular questions, the Chinese Room could not display human like understanding, and therefore the system as a whole does not understand Chinese, based on its responses.
 Questions about philosophy:
|
|
Tweet this!
