mikes@tekecs.TEK.COM (Michael Sellers) (11/13/86)
[Someone recently asked why all this cognitive science stuff kept appearing
in net.cog-eng and net.ai . Where else *should* it go? Do we have enough
readership to justify a net.cogsci (it'd be nice)?]
I recently asked about cognitive science graduate programs; specifically,
why did you join the one you're in? I've gotten a number of responses
(thanks! keep that e-mail coming), and now I'd like to alter the question
a bit.
It seems that there are very few schools with anything like a bona fide
cognitive science (or cognitive anything) program at the grad level. There
are some schools that have some sort of mish-mash of AI & psych or linguistics
shoehorned into being a cogsci program (this is not meant to be an attack).
It may be that the field is still too young for there to be people to teach
it reliably, so a full-blown program is not yet possible.
My question, then, is in two related parts: first, what do you consider
appropriate and adequate preparation at the undergraduate level for a
graduate program in cognitive science (i.e., how much computer science, how
much math, how much bio or psych, etc); and second, what do you consider
would be a viable program at the graduate level to warrent an advanced degree
(MS or PhD) in cognitive science?
My main goal here is to start some discussion about how we should proceed
in this area in general; and more specifically so I can get some other ideas
about *my* possible program. In addition, it might be useful to approach the
question of "what is/should be cognitive science?" from this perspective.
Rather than arguing about whether neurology is more important than linguistics,
maybe we can settle on how much of each is a good blend.
E-mail to me is fine (I love the stuff), but I'd like to see a discussion
of this get going, so posting is probably good too.
--
Mike Sellers
UUCP: {...your spinal column here...}!tektronix!tekecs!mikes
"You learn lessons from no one so well as from yourself"spf@bonnie.ATT.COM (11/17/86)
>From clyde!rutgers!lll-crg!mordor!sri-spam!sri-unix!hplabs!decwrl!decvax!tektronix!orca!tekecs!mikes Mon Nov 17 10:05:58 EST 1986 > >and second, what do you consider >would be a viable program at the graduate level to warrent an advanced degree >(MS or PhD) in cognitive science? > Mike Sellers Well, I can tell you what I did (it must be right, no?). With a B.A. in Environmental Physics (no, I'm not sure what that is), I first took an M.S. in Computer Sciences, which gave me a solid foundation in algorithms, formal languages, and a little AI. At the time I was doing some industrial research with a cognitive psychologist (ex-EE), and came to appreciate the need for FORMAL training in experimental psychology. "Tech" types (myself included) tend to believe they can step into the psychology arena, derive a partial differential equation describing brain function, and solve it to know who people work. Of course, this is ridiculus; VERY little is known about how the brain (and it's owner) works. And experimenting with humans is not like doing physics experiments -- the sample rarely behaves the same way two days (or hours) in a row, and no two samples are alike. Besides that, the basic science of psychology is today where physics was in Newton's time (that's not a flame -- in fact, that's one reason why it interests me!). So, as a result of the above observations, I went back and got a second M.S. in Applied Psychology (experimental psychophysics, cognition, perception). Actually, I'm just finishing that M.S. now, and am considering grabbing a PhD (if my sheep farm doesn't intrigue me more). I think the duel M.S. has served me well; if I had taken HALF the courses in each of the two disciplines (CS and psych), as would have been the case with a single, combined masters, I don't think I'd have an adequate understanding of either. I don't know what the ideal PhD program for me would be, but I suspect it will consist of AI-type courses (pattern recognition, inference systems, etc), and advanced psych courses in cognition and perception, and possibly linguistics. Of course, this relates to my research interest in sensory data representation (auditory and visual) and other areas of "natural" intelligence enhancement. I'd enjoy seeing some discussion about "cognitive science" programs and curricula. Steve *** How can they know it's time for them to go? -- Sandy Denny