[net.misc] Computer Science or Computer Prescience?

otto (06/15/82)

I have often wondered about the "Science" appendage to fields whose
names by themselves would not appear to be scientific.  Are these new
fields in fact sciences? or does the term simply connote a more
methodical way of dealing with the topic at hand than would normally be
the case?

What is a science?  I adopt the position that something is a science if
it follows the steps outlined by Thomas Kuhn in his *The Structure of
Scientific Revolutions*, to wit, that there is an overriding paradigm
that those in the field agree defines the legitimate problems and
methods of solution of that field.  Changes in that paradigm can come
about (these are the scientific revolutions mentioned in the title of
his work), but at all times outside these moments of revolutionary
adjustment there is but *one* paradigm defining the field.

Before a field becomes a science it is in what is called its
"preparadigm" period, marked by numerous competing attempts to
understand the basic phenomena slated for study, with no clear unifying
principles tying the whole area together.  The period of the Alchemists
was just such a time for chemistry.  It was not until basic unifying
principles were discovered and promulgated by such as Dalton, Lavoisier,
etc., that chemistry finally acquired a paradigm and became a science.

Based on Kuhn's ideas, I think it clear that the fields of Computer
Science and Psychology are neither one sciences: both are in their
preparadigm phases.  In fact, I would be tempted to say that fields that
append "science" to their names, such as Military Science, Peace Science,
Computer Science, etc., are actually opting for *preparadigmhood*.  They
are claiming that, just as happened with the Alchemists, by gaining 
enough experience in the given field and noting enough events, that at some
point *in the future* someone will be able to formulate unifying
principles that will give the field its first paradigm and qualify it
as a science.

Thus, for me, the use of the term "science" in the title of a field of
study tells me more about the researchers hopes for its future than
about the current state of their knowledge about it.

George Otto
Bell Labs, Indian Hill
----------------------

steveg (06/16/82)

I'm not sure you could call hardware design a pre-paradigm science.
It has become pretty much a normal science, where people who violate
the Von-Neuman rules getting badly bruised. Look at the ammount
of pain Intel has with the 432, which could conceivably start the
next revolution.

AI, though, is your classic pre-paradigm science. I can't yet imagine
AI holding on to one paradigm for more than 3 years. Remember, it
wasn't that long ago that hill-climing, and perceptrons were in
vogue (they still are among a lot of the "yet to die off" generation).

Have the Kuhnians in the audiance read: "Night Thoughts of a Classical
Physicist", I think most would enjoy it.

				- Steven Gutfreund
				  University of Massachusetts,
				  L.O.A DEC Corporate Research