Re: ORGASM - Operant Reactionary Graduate Academic Scholastic Mentality

Michael Olea <[EMAIL PROTECTED]> wrote in message news:<[EMAIL PROTECTED]>...
> The relationship I see is that all 3 involve attempts to deal pragmaticaly
> with "meaning".  All 3 draw on concepts of information - mutual information,
> entropy, etc; and use many of the same tools - a family of metrics
> (variations of "edit distance") or equivalent similarity measures, hidden
> markov models, probabilistic grammars, support vector machines, cluster
> analysis, novelty detection...  I don't know that anyone has tried to model
> spike trains using grammars or that that would be a fruitful avenue of
> study, but the "stimulus reconstruction" techniques of Rieke et. al.
> ("SPIKES: Exploring the Neural Code") are a form of translation, an attempt
> to measure how well the "meaning" of a sequence of spiks has been
> "understood".  And all 3 involve attempts to automate processing of ever
> larger "corpora" - to figure out what it all means, Mr. Natural.  That's
> what I'm talking about.

Thank you for elaborating on the relationship between these problems.
What still troubles me is the relation of these to Quine's conception
of Cognitive Synonymy. I wonder, is it even well defined in any sense
of the word? A definition from the view point of Quine may well be in
order (Here an interesting and well informed take on Quine:
http://www.ifs.csic.es/sorites/Issue_13/tomassi.htm )

In other words, the common property of all three is that they use
formal techniques that deal with information. I wonder if there could
be any other scientific way of approaching such problems, and whether
these have anything to do with what any philosopher of science ever
said, leave alone the analysis of Quine?

I would appreciate further explanation on the philosophical import of,
say, using a mathematical means such as a support vector machine to
classify text documents with high accuracy.

In my opinion, trying to abandon the hammer that I am used to, that
means nothing except that the _operation_ of a support vector machine
may be similar to the working of human mind. It points out to the
possibility that categorical distinction amounts to no more than
developing such models over data. Furthermore, since there is an
effective learning algorithm that our minds may be employing
finitistic means of achieving this end.

Is there anything else to say?

Regards,

--
Eray Ozkural