Return-Path: <socgrad-relay@UCSD.EDU>
	sendmail 5.67/UCSD-2.2-sun
	Fri, 19 Feb 93 16:33:44 -0800 for /usr/lib/sendmail -oc -odq -oQ/var/spool/lqueue -oi -fsocgrad-relay socgrad-list
Date:         Thu, 18 Feb 93 18:01:33 EDT
From: Jon EPstein <JEPSTEIN@Kentvm.Kent.edu>
Subject:      Doesn't taste very good and all but boy is it full
To: socgrad@ucsd.edu
Statistics and mathamatical models are only good for testing theories which
spring from a positivistic philosphical orientation, for example
the work of B. Cohen is a good overview of this orientation.
However, much sociology is not appropriate to mathamatical tests for example
Goffmans work.  See    Reid 1990 (?) for a good discussion of this topic.
Also much of S.I., with the exception of Structural SI ala Stryker,
which it should be  pointed out violates the core assumptions of SI which
is kind of a problem, much of symbolic interactionism is not testable
with statistics.
In response to SREID1: Denzin is an outstanding symbolic interactionist
who has done a admirable job of addressing the contributions of Postmodernism
to SI.  His book Interpretive Biography is quite good as is The recovering
Alcoholic.  IMHO Denzin is much truer to the tenets of SI than are Stryker,
Burke et al. Also the "Chicago/Iowa" split is basically just a text book
thing that has little basis in reality (much like a lot of sociology especially
that which assumes that the social resembles a bell shaped curve. Reality is
not normally distributed)