3.250 computers and questions (53)
Willard McCarty (MCCARTY@VM.EPAS.UTORONTO.CA)
Sun, 16 Jul 89 20:24:38 EDT
Humanist Discussion Group, Vol. 3, No. 250. Sunday, 16 Jul 1989.
Date: SUN 16 JUL 1989 15:11:00 CDT
From: Jim McSwain <F0A8@USOUTHAL>
Subject: reply to Gilmore (Chronicle of Higher Education article)
Well done to Matthew Gilmore (LIBRSPE@GWUVM) for his comments on the
CHE article of 12 July 1989. I think the statement "extensive evidence
. . will lessen the weight given to exquisitely crafted explanations
based not on large numbers of cases, but on the force of argument"
refers back to an earlier statement that computers are "absolutely
trustworthy . . ."; i.e. they can seach an enormous datebase and not
leave out anything which human fatigue or carelessness might leave
untouched. I add the caution, however, that quite obviously a database
is only as "trustworthy" as the humans who compiled it, and they may
leave out data deemed irrelevant or emphemeral. Computers do
repetitive tasks quite well, the drudgery of looking through
stacks of records and texts for key words, dates, etc., but they
cannot anticipate the relevancy of previously unimagined data based
on new insights into the nature of the problems being pursued, which
is for me a tentative definition of one type of human thinking which
AI and microprocessors cannot as yet perform. Perhaps that is because
contrary to contemporary sentiments the mind is more than just brain
. . As for the original quote, I call to mind Keith Thomas's book,
RELIGION AND THE DECLINE OF MAGIC (1971?). In a critical review
Margaret Bowker objected that citing examples and counter-examples of
"cunning" men did not establish the extent of their influence among
the English masses. Large numbers of cases found through computer
searches still require human thought and imagination before they
make sense, and computers cannot do that for you. One must consider
for example what percentage surviving records were of the original
number of records, so that even apparently large numbers of cases
taken from existing records still may not tell one the significance
of the event or idea in the time under investigation. I welcome
massive databases and the equipment to use them, but human skill
and thought remain indispensable ingredients in the scholarly process.
What I am enthusiastic about in this article is the remark that
"computerized data bases . . . open up a new range of questions that
can be asked that would hitherto have been unthinkable . . ." Again
I see this as part of the humanistic task of ordering experience,
expanded in the case of computers and databases simply because our
practical range of speculation, examination and proof has been
enlarged relative to the fixed limitations of human energy, vision
and integration of vast amounts of data. Breath-taking stuff for
the lowly . . . JMcSwain