Truth, kids and Socrates.2
John McCreery (JLM@TWICS.COM)
Mon, 22 Apr 1996 19:19:51 +0900
Dwight Read writes,
"Note that McCreery uses the qualifiers: "I suspect", "I speculate".
His suspicions and speculatations may very well be insightful,
but they are not a reference to concepts established through
anthropology as a science. My question is still unanswered."
He is, of course, quite right. To keep the pot boiling, let me offer
the observation that cultural anthropology, as currently practiced,
is not a science. It may, however, be part of a larger scientific
enterprise, the study of human behavior.
I think I have mentioned before a favorite diagram in a book on
methodology that I read many years ago. The diagram is a four-
cell table defined by 1 and n steps on two dimensions: the
number of questions a researcher asks and the number of subjects
she asks them of. The (1,1) cell is research that consists of a single
question asked of a single subject, e.g., do you beat your wife? The
answer may be informative but does not, by itself, support
generalization. The (n, n) cell describes an ideal universe in
which, if n is large, large numbers of questions are asked of large
numbers of subjects. If the questions are properly designed and
the subjects properly sampled, a multivariate analysis will reveal
large numbers of connections. Doing this type of research is,
however, impractical for all except large, wealthy organizations.
In practice, researchers must normally opt for, what are at the
extremes, (1,n) or (n,1) strategies. The former design is ideal for
hypothesis testing, once a question is very well defined. The latter
is ideal for exploratory research that may generate great numbers
of insights, but is not, in itself sufficient to validate them.
Fieldwork, as practiced by most cultural anthropologists, is an
(n,1) strategy par excellence, immersion in a single case of which
we ask as many different questions as possible as we build our
files of fieldnotes. When we come to write up our observations
and sit sifting through our data, we may find many interesting
things to talk and speculate about. We are not in a position to
"prove" anything. Our role is, at best, to challenge received
wisdom in hopes that some of our (1,n) colleagues--in sociology,
psychology, political science, etc., find our ideas worth testing.
There are also, of course, a few anthropologists who do comparative
research and may be receptive to the insights we offer.
April 22, 1996