Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site duke.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!bellcore!decvax!mcnc!duke!crm
From: crm@duke.UUCP (Charlie Martin)
Newsgroups: net.cse
Subject: Re: Degrees, grades...
Message-ID: <7010@duke.UUCP>
Date: Fri, 14-Mar-86 09:48:44 EST
Article-I.D.: duke.7010
Posted: Fri Mar 14 09:48:44 1986
Date-Received: Sun, 16-Mar-86 00:45:47 EST
References: <4514@kestrel.ARPA> <3407@nsc.UUCP> <4588@kestrel.ARPA> <6987@duke.UUCP> <55@gilbbs.UUCP>
Reply-To: crm@duke.UUCP (Charlie Martin)
Distribution: net
Organization: Duke University
Lines: 66
Summary: 

In article <55@gilbbs.UUCP> mc68020@gilbbs.UUCP (Tom Keller) writes:
>
>   Computer Sciences, on the other hand, is a highly rigorous, mathematically
>oriented field.  Science is perhaps not a completely accurate description, but
>it's damned close.  These people don't *NEED* to know about the day-to-day
>nitpicky crap that the programmers do.  Many computer scientists go hours, days,
>weeks, even MONTHS without ever writing a line of code!  (*gasp*)

My goodness.  Poor people.  They spend all that time doing computer
science, doing all that intensive mathematical labor, and hardly ever
write a line of code.

Sorry, but you are completely off the wall here.  All that theory, all
the science or mathematics that is accumulated in CS has one and only
one purpose or reason to exist: it allows us to explain and understand
the processes involved in programming.  It has only one subject:
programs.  No matter how abstract, no matter whether you are a
front-line programmer or a theory person using dynamic logic to build a
model of concurrency, you are still occupied in understanding and
describing programs.  THERE SIMPLY IS NOTHING ELSE THERE.

And just as we expect physicists to understand the language of physics
-- i.e. lots of scuzzy differential equations, wave equations, etc. --
we simply should and must expect computer scientists to understand the
language of computer science: programming.

Would you expect a chemist not to understand how to put together a lab
setup?  A biologist not to understand statistical experiment design?  A
philosopher not to know how to write an argument?  Then why do you
expect that a Computer Scientist need know nothing about the practical
aspects of his (or her) field?

I admit it is often true -- but I'm an idealist.

Furthermore, many or most graduates from computer science programs will
take their BS or BA and leave, go out into the real world and make a
living -- just as many chemists will go out and become lab chemists.  It
behooves us as educators to see to it that these people are prepared to
do just that, just as it behooves a chemistry department to see to it
that their graduates know how to work and work effectively in a lab.
There will still be OJT, of course: but a chemistry graduate who learned
all the theoretical parts of chemistry without learning about life in
the lab would have been cheated.

Chemistry departments realize this, of course: how many chem departments
let someone graduate without several heavy lab courses?  Computer
science departments should do the same.

>   But the insistence of schools to lump programmers and computer scientists 
>together results in many problems.  The real issues here, it seems to me, are
>not whether a CSci degree has any value, but whether *PROGRAMMERS* need one.
>Whether we should try to combine the educational requirements for applied
>engineering (programming) and theoretical physics (Computer Science).
>
>   Would you consider a degree in theoretical physics to be terribly useful to
>someone wishing to design bridges?  Of course not!
NO, but it sure helps someone who is interested in VLSI.  Tools of the
trade ands all, you know.

Even a theoretical physicist needs to know about building experiments.


-- 

			Charlie Martin
			(...mcnc!duke!crm)