The story goes like this. I have begun a project exploring the social networks that link the winners in one of Japan's largest advertising contests. I turn for help to Jim Ennis, until recently chair of the department of sociology and anthropology at Tufts University, a social network analyst, and briefly, when he was about to graduate and I was a green-as-a-gourd new professor just hired by Middlebury College. We go back a long way. While suggesting things that I might want to read, Jim suggests the work of U. of Chicago sociologist Andrew Abbott. That is how I find myself reading Abbott's _Time Matters: On Theory and Method_, a book of whose existence I might well have gone to my grave without ever knowing—and a book that impresses me greatly. Why? Because Abbott is a sociologist well-versed in quantitative methods and mathematical modeling who has also taken on board postmodern critiques of positivistic epistemology. His grasp of narrative and interpretive analysis is every bit as good as his grasp of math and statistics, a rare phenomenon, indeed. _Time Matters_ includes an autobiographical introduction and a series of papers written in the 1980s and 1990s. So far I have read one, "Transcending General Linear Reality," which analyzes the conceptual weaknesses common in quantitative research that assumes that social facts can be modeled as systems of linear equations. That assumption is often made primarily because without it the math involved in statistical analysis becomes very complicated, indeed. But it only takes a little thought to realize that life mostly doesn't work that day. (A thought that among other things--TANGENT WARNING!--directs my eye to the copy of Checkland and Scholes, _Soft Systems Methodology in Action_ sitting on the bookshelf next to me). Now I'm reading chapter two "Seven Types of Ambiguity," and, yes, the title is a tribute to William Empson's famous work of literary criticism. But Abbott is not the sort of social scientist to be happy with noticing ambiguity and throw up his hands in despair, throwing out the baby of formal analysis with the bath of simplistic models. He writes, "In this paper I analyze 'positivistically' a phenomenon normally assumed to forbid the possibility of positivism: the multiple and seemingly incommensurable meanins assigned to human events. I discover in current positivist research a source of untapped information about those multiple meanings. This discovery does not deny the normal critique of positivism, but rather suggests that the antipositivist concept of ambiguity can be used to fold the seeming flatness of positivism into a complex and subtle terrain. "For some readers, bothering with the research produced by sociological positivism and its cousins in political science and economics may seem a waste of time. For them, its philosophical presuppositions are ruined. Its correspondence theory of truth is broken. Its 'causality' has proved mere reification. Yet there are two problems with accepting the philosophical critique of positivism. First, even though positivist social science has been shown to be 'in principle impossible,' the vast majority of social science effort (and funding) is in fact spent doing it. Such research is often highly consequential, whether it be the market studies that shape consumeer demand or the census figures that determine political districts. "The second problem involves the motivations of those who deny the efficacy of positivism. Proclamations against positivism often mask an arbitrary unwillingness to think formally about the social world. One asserts that the world is constructed of ambiguous networks of meaning, argues for the complexity of interpretations and representations, and then simply assumes that formal discussion of the ensuing complexity is impossible. But this is obviously untrue. Many people have thought formally about ambiguity, representation, and interpretation. Nothing in those phenomena militates against thinking in a rigorous, even disciplined fashion, as we see in the work of Empson, Barthes, and many others." In an already overlong posting, allow me one example of what Abbott is talking about. He has noted that sociological studies often relate several measures to a single concept, so that, for instance, religiosity may be measured simultaneously by church attendance, daily prayer, personal statements of belief, etc. Then, he goes on, "But consider the reverse situation, the possibility that a given indicator is attached to more than one concept. In such a case, one measurable thing 'means' several conceptual things at once. Years in school, for example, 'means' education, in the sense that we presume (somewhat against our own experience as teachers) that time spent in school results in more or less monotonic increase in education. But years in school also 'means' exposure to popular culture or experience with bureaucracy. Years in school even 'means' reduced time available for criminal activity, because those in school are literally off the street. In fact, in many individual cases years in school is probably a more accurate indicator of time available for criminal activity than it is of degree of education. This simple type of multiple meaning is Empson's first type of ambiguity, a situation where one fact means several things." And there are six more to go. I'm hooked. Highly recommended. -- John McCreery The Word Works, Ltd., Yokohama, JAPAN Tel. +81-45-314-9324 http://www.wordworks.jp/ ------------------------------------------------------------------ To change your Lit-Ideas settings (subscribe/unsub, vacation on/off, digest on/off), visit www.andreas.com/faq-lit-ideas.html