Path: utzoo!attcan!uunet!timbuk!cs.umn.edu!msi.umn.edu!src.honeywell.com!sol.ctr.columbia.edu!zaphod.mps.ohio-state.edu!wuarchive!emory!gatech!ncsuvx!news
From: fostel@eos.ncsu.edu (Gary Fostel)
Newsgroups: comp.ai.philosophy
Subject: Re: Emergent Properties
Keywords: chaos, science, prediction
Message-ID: <1990Oct26.214354.11063@ncsuvx.ncsu.edu>
Date: 26 Oct 90 21:43:54 GMT
References: <1990Oct12.214636.7945@ncsuvx.ncsu.edu> <30@tdatirv.UUCP> <1990Oct19.201604.7280@ncsuvx.ncsu.edu> <3369@aipna.ed.ac.uk>
Sender: news@ncsuvx.ncsu.edu (USENET News System)
Reply-To: fostel@eos.ncsu.edu (Gary Fostel)
Organization: North Carolina State University
Lines: 76

I said:

    The purpose of science is (I thought) to search for descriptions
    of the world, to be offered in terms of defined quantities, and
    then to test those descriptions for adequacy.

And Chris Malcom, at Edinbourough, replied:
  
    That is a simplified rational reconstruction of science, and is a
    suitable slogan for keeping teenage wannabee scientists in some kind of
    order. Since it is also communicatively efficient, scientists, just as
    do mathematicians, endeavour to present their results in accordance with
    this paradigm; and, just as with mathematicians, it bears little
    resemblance to the way they actually work.

Perhaps I am still a teenage wannabe scientist.  I am rather familar with
the work habits of "scientists" -- the good and the bad.  I had thought I
was talking about an issue of how one ought to proceeed, not how people
often do, for a variety of oft pressing pragmatic reasons.  There was a
time when scientists studied the philosphy of science as part of their
training; this is less common these days. It shows. 

Malcom continued:

    For example, this news group has recently been full of complaints and
    suggestions from people who -- motivated by precisely the view of
    science you present here -- would like all AI research to stop until
    someone has found a good definition of intelligence.

The fact that someone may misapply a basic principle of the scientific
method is not an invalidation of the principle.  Perhaps that is how these
people drew this conclusion.  I saw only a few remnants of rubble of that 
debate but if these "people" really said what Malcom thinks they said, they
clearly misapplied the principle.  I wonder if they think they said what
Malcom thinks they said.  If AI people are "scientists" than there may well
be some serious methodological weaknesses in current work -- weaknesses 
that are perhaps strengths if one relables the method.  Bad technique in 
science may be excellent philosophy, mathematics, or, esp. engineering.
A few words, brutally quoted out of context from Malcom's note, 
help make my point: "constructing", "building", "inventor". These are
terms drawn from engineering, not science.  

There are sciences, especially biological, in which enourmous effort is
put into "building a system" in order to use it to study a problem. However,
once the system is built, it is subjected to rigourous examination,  
usually called "characterizing" the system so that the REAL work of science
can then be done: performing well controlled experiments in a well understood
setting to prove or disprove theories about the elements involved in that
system.  The situation in AI ... and other branches of Computer "Science"
as well ... is quite different, as Malcom confirms when he says:
 
    It is not uncommon
    for a good research prototype to be the subject of heated debate over
    precisely what it exemplifies for a number of years. The final
    interpretation is sometimes -- even in the eyes of the inventor -- quite
    different from the original intention. One of the distinguishing
    features of a good research prototype is just this long-term
    fruitfulness. 

The "long term fruitfullness" of the "research prototype" comes from the
"heated debate" not from the specific well controlled experiments performed
using this prototype.  The difference is quite clear.  This is not to say
that everyone doing AI should pack it up, but it does suggest that if AI
comes under criticism for not being a proper scientific discipline, AI
people should not deny that such is the case.  If someone suggests that 
more rigourous scientific methods might provide long term benifits to the
field, this should not be refuted by claiming that such methods are already
being applied.  That just makes it look as if the practisioners do not
understand what it is they are being advised to consider.  Rather, it might
be the case that the scientific method simply can not be applied or would
not be fruitful, and that can be the basis for defending the status quo in AI.

But, of course, it might also be the case that those methods would help.  
  
----GaryFostel----                        Department of Computer Science
                                          North Carolina State University