John L.McCreery (jlm@TANUKI.TWICS.CO.JP)
Fri, 7 Jan 1994 14:13:54 JST

First, a warm welcome back to Bob Graber and to everyone else
involved in this conversation which seems to be spreading from
postmodernism to epistemology in general. Best wishes and a Happy
New Year to all. To Bob G., I'd like to say thank you for his kind words
vis-a-vis "The Two Cultures." I'd also like to say that when he writes,

"I think, however, that McCreery gives too great a role
to experimentation in his conception of science. In my view,
commitment to evidence and reason is all that is crucial; science
consists more in a general epistemology (and its products) than in a
specific methodology--even when that "specific" methodology is as
diverse and far-reaching as experimentation. From this perspective,
experimentation is simply the ideal method for gathering evidence,
because it ensures that the evidence obtained will bear directly on one's

I agree with most of what he says. Where I take mild issue is where he
says that I give "too great a role to experimentation" in my conception
of science. What I meant to be saying is that experimental method is
the prototype, the intellectual center of gravity, around which
discussions of "science" are organized. Statistical methods (and more
recently, computer simulations as well) are attempts to approximate
the rigor of experimental method in situations where experiment is
impossible. These methods also impose conditions that anthropological
research rarely satisfies; nowhere is this more true that in interpretive
studies where diverse facts are assembled to create an overall picture of
what is typically a complex situation. Our problem is that we don't
have standards for what counts as "evidence and reason" in this kind of
study comparable to those we appeal to in evaluating experiments,
statistics or simulations. I would like to explore what they might be.

As starting points, let me suggest that anthropologists are not alone in
having to deal with assertions involving "diverse facts...overall
picture...complex situations." Lawyers deal with them daily, and there
an enormous literature on what should count as "evidence and reason"
in courtrooms. Doctors, too, must deal with complex cases, and there is
some literature (I don't know how voluminous) on the special problems
of clinical research. Historians have dealt with these problems for
centuries under the rubric of historiography. (Hermeneutics began as a
special branch dedicated to assessing the truth claims of scripture.) To
leap from "Even the rules of experimental method don't guarantee
truth" to "There are no rules worth thinking about" seems, to
understate the case, decidedly premature.

Looking forward to hearing from you.
John McCreery (JLM@Twics.co.jp)