--- In Wittrs@xxxxxxxxxxxxxxx, Gordon Swobe <wittrsamr@...> wrote: > At the beginning of his argument, Joe clearly defines the meanings > of the marks in plain English. I obviously anticipated that kind of response, though I had expected it to come from Joe. The idea behind AI, is that you could have a database that clearly defines the meanings of all of the marks used. And then you could manipulate those meanings by manipulating the marks. If you connect the results appropriately to external behavior, you should have a system that responds appropriately to meaningful input. The understanding would be in the system as a whole, as demonstrated by these meaningful responses. The reason I am skeptical of AI, is that I am skeptical about the possibility of having a database that clearly defines the meanings to be used. For the case in point, though you say that "Joe clearly defines the meanings", in fact the definitions are not at all clear - something that Stuart has already pointed out <http://groups.yahoo.com/group/Wittrs/message/5439> . It has long puzzled me, that philosophers so often make bogus logical arguments. They take an intuitive but unconvincing argument, symbolize it, and present it as a formal symbolic argument. Perhaps they believe that symbolizing the argument makes it more convincing. However, I see the effect as mainly one of obfuscation. Regards, Neil