[ExI] Meaningless Symbols.

[Stathis Papaioannou]

2010/1/15 Gordon Swobe <gts_2000 at yahoo.com>:

>> It doesn't help to say that some physical activity
>> happens in neurons which produces the understanding, not
>> because you haven't given the details of the physical activity, but
>> because you haven't explained how, in general terms, it is possible for
>> the physical activity in a brain to pull off that trick but not
>> the physical activity in a computer.
>
> But I have. You just don't believe me or understand me or both.

You've said that formal programs can't produce understanding but
physical activity can produce understanding. Computers not only run
formal programs, they also do physical activity. You have a hunch that
the sort of physical activity in computers is incapable of producing
understanding. But a hunch is not good enough in a philosophical
argument. To make your case you have to show that it is *impossible*
for the physical activity in a computer to support understanding. For
example, you would have to show that running a program actually
prevents understanding. That would mean that electric circuits
arranged in a complex and disorganised fashion that could not be seen
as implementing a program could potentially have understanding but not
if the same components were organised to form a computer. Is that
right?

>> Even if it's true that computers only do syntax and syntax can't
>> produce meaning (it isn't, since logically there is nowhere else for
>> meaning to come from)
>
> I think that last thought of yours needs some work. :)
>
> You say "logically there is nowhere else for meaning to come from", but *logically* nothing can get semantics from knowing rules of syntax, or vocabulary from knowing rules of grammar.

It's true that given an unknown string of symbols it's impossible,
even in principle, to work out their meaning even though you may be
able to work out a syntax. However, you can ground the symbols by
associating them with symbols you already know, a syntactic operation.
And ultimately the symbols you already know are grounded by
associating them with sense data, another syntactic operation. So
syntax is both necessary and sufficient for semantics. How else can
any entity, human or computer, possibly derive the meaning of
something other than through a process like this?

And my original point: even if you still believe meaning must come
from the physical processes inside a brain, why can't it also come
from the physical processes inside a computer?


-- 
Stathis Papaioannou