[ExI] Meaningless Symbols.

[Gordon Swobe]

--- On Fri, 1/15/10, John Clark <jonkc at bellsouth.net> wrote:

>> print "Hello World"
> 
> The title of this thread is "Meaningless Symbols", if "print" was one of 
> those to the computer then it would not do exactly the same thing each
> time it encountered that symbol, instead it would do some
> arbitrary thing. Apparently the computer ascribed meaning to
> at least one of those "meaningless symbols".

You've stumbled close to the truth there, John.

As I pointed out in my original message, "print" counts as a syntactic rule. Although it stretches the definition of "understanding" I can for the sake of argument agree that s/h systems mechanically understand syntax. They cannot however get semantics from their so-called understanding of syntax. Nor can humans get semantics from syntax, for that matter, and humans really do understand syntax.

I mentioned also that the classic one-line "Hello World" program does not differ in any important philosophical way from the most sophisticated possible program. Someone made some scornful and ignorant comment about that. 

Let us say that we have a sophisticated program that behaves in every way like a human such that it passes the Turing test. We then add to that program the line 'print "Hello World"' (or perhaps 'speak "Hello World"') such that the command will execute at an appropriate time still consistent with passing the Turing test. That advanced program will not understand the meaning of "Hello World" any more than does the one line program running alone. 

S/H systems can do more than follow syntactic rules for crunching words and symbols. They have no way to attach meanings to the symbols or to understand those meanings. Those semantic functions belong to the humans who program and operate them.

-gts