🚀 go-pugleaf

pnm@goanna.cs.rmit.oz.au (Paul Menon) writes:

>In article <1991Jan3.205417.24369@agate.berkeley.edu>, bks@alfa.berkeley.edu (Brad Sherman) writes:

>> Forgive my extreme ignorance, but could someone give a thumbnail sketch
>> of "connectionism."  It seems not to exist in my dictionary, nor can
>> I find it in the title of any periodical in the UCB on-line database
>> (unless it is also known as "Connection Science").

>  Connectionism, I believe, is the term applied to the theory and application
>of neural-networks.  It may be far more general than that, eg, to include
>ultra-large networks of computers/entities.

Connectionism is a philosophy which has replaced machine functionalism as the
latest "in" philosophy.  Instead of believing in the mind/frog/whatever as a
collection of turing machines which operate on their internal tapes and
environmental inputs, connectionists believe that responses and intelligent 
behaviour are immergent properties which can be duplicated if only we can 
get enough connections and which because of their massive numbers can never
be truely understood, only worshi.. (I mean) studied.  This is not to say
everyone who likes neural nets and lots of connections ascribes to these 
beliefs, but then only the purists believe the brain is a true turing machine 
anyway.  The trick is to use the tools and avoid the dogma.

ejf

-- 
--
Erik J Fretheim
efrethei@afit.af.mil			AFIT/ENA Box 4151 (ATTN: CPT FRETHEIM)
(513)255-5276 AVN785-5276 		WPAFB, OH 45431  USA