Let me start to reply to the important points Naoki is making by saying
where I generally agree with what he said in his last posting.
That CA and ethnometh are not micro-social analysis, theory, etc. Certainly
they are not meant to be such within their own frame of reference. Yes,
they do try to challenge traditional macro-sociology (concerning which, see
below). I believe that they ADD an important perspective, and that their
critique of the traditional views is partly valid, but I do not see them as
a viable substitute. It is perhaps partly a clash of incommensurable
paradigms. Different questions are being asked, phenomena are being
construed differently. The general approach that one might call
phenomenological social analysis, to be more general about the aims than
just what CA or US or British ethnometh have historically made of them, is
perhaps a valid general alternative -- and Naoki is right that there is no
reason why they have to be limited to micro-analysis, or why their analyses
have to be micro. In fact I would say that no analysis can ever in fact be
micro; we cannot analyze anything without making use of semiotic tools that
link us as analysts into larger social networks with longer biographical
and historical trajectories.
But I do not often see this explicitly in CA and ethnometh work. I do not
see the explicit tracing of the larger networks through which the analysts
do their analyses. I do not see the analyst put back into the analysis in
the way that another kind of phenomenological social analysis in feminist
theory or queer theory or some kinds of postmodern critical theory actually
do (and are roundly criticized for in many quarters). I also do not see the
work of connecting the analysis of this occasion to the analysis of that
occasion out through the social networks far enough that the kinds of
issues which traditional social theory wants to address can be encompassed.
_Implicitly_ I grant that all this is necessarily happening in the
analysis; but this work does not present itself as mere DATA for other
people to analyze!
There are good reasons why phenom-soc research does not extend itself far
enough in these ways: it takes a very long time, more or less an extended
career of analyzing the pieces and then tracing out the connections. Very
little in our institutional structures for research is conducive to this
sort of longterm work. It is very much a problem of SCALE, and the linking
of scales. And by scale here I do NOT mean the notional divisions of
micro-, meso-, macro-, but the extents of times and places and participants
and settings in the work of research and social participation itself.
Naoki also notes that traditional sociology is a sort of mirror image of
cognitivism and that they really fit together very well within the
modernist project. Here I completely agree. Latour's discussion of this in
_We Have Never Been Modern_ seems pretty sound to me, though it is brief
and does not specifically address some of our concerns here. European
Modernism divided nature from society, body from mind in parallel ways. It
created a conceptual as well as political opposition between individualist
perspectives and communitarian perspectives. Mike Cole has often reviewed
here and in his _Cultural Psychology_ how the study of mind was split away
from the study of culture and society. I think it goes without saying on
xmca that one needs to see meaning-making and identity issues,
traditionally the province of pysch, in sociocultural and historical
perspectives. What is the mirror reflex of this, however? In
phenomenological social analysis, it is, I think, that we must see
large-scale and collective social-historical phenomena in relation to local
meaning-making practices and the positions and trajectories of how
identities are experienced and constructed. The ideals of phenomenoligical
social analysis are ones I tend to agree with in large measure, but I am
concerned that its practices do not live up to these ideals.
While not particularly phenomenological in inspiration, so far as I know,
Latour's version of a 'flat' or everywhere-local analysis has tried to link
together analyses of distant and distinct settings in ways that reveal
something of the larger-scale concerns of traditional sociology, though
perhaps less successfully so for historical time-scales (not studies in
period, but across periods), and with not as much precision in the analysis
of particular events and texts as various sorts of discourse or semiotic
analysis could provide.
I do believe ultimately that matters of practical method are more critical
now than issues of theory or methodology, and I believe that Mike Cole has
often gently tried to pull us round to this emphasis. I think that many of
us have quite a good sense of what we would like to accomplish in social
analysis, and that our goals are quite theoretically sophisticated. But our
methods are still quite primitive and inadequate, and most seriously we do
not very often effectively integrate methods in ways that link analyses on
different time scales, or even across distant space scales. I would say
that one of the best pieces of social analysis I have read in the last few
decades is Bourdieu's _Distinction_, but that work has very serious
problems and weaknesses precisely when it tries to substitute
cross-sectional data for longitudinal data (he talks about trajectories and
habitus-development, but has no longitudinal studies of cohorts), and when
it tries to integrate interview data with survey data, or data of what
people say they like with data about what people actually choose.
I will have a closer look at Suchman's chap 4 as Naoki recommends (and BTW
I am not a fan of Giddens except as a master synthesizer and reviewer of
the literature; I much prefer Bourdieu as a theorist), but I would very
much like to hear from people on xmca of (recent, say since 1980) exemplars
of method, where multiple methods of data generation and analysis show ways
to link local analyses across distant but linked sites in networks of
distinct (in type) but closely interdependent social activities, and/or
across time periods on a biographical or historical scale. I know a few
candidates, all flawed in one way or another, but I will wait to hear
others nominations.
JAY.
PS. I apologize if my casually written phrase 'necessarily distorting and
incomplete' referring to single methods may have seemed to some people to
imply a faith in the possibility objectively true accounts of phenomena
across scales. I lost this faith as a child. The mirror metaphor misleads.
Perhaps I should have written 'necessarily biased and incomplete', but I
would still not imply one can ever make a complete account or an unbiased
one. My ideal is merely a less incomplete and more multiply biased view,
preferably in ways relevant to my or our interests and projects, and in
ways that sit uncomfortably with one another, stimulating dialogue,
dialectic, dissatisfaction, and impetus to add yet more and more different
perspectives to the accounting.
---------------------------
JAY L. LEMKE
CITY UNIVERSITY OF NEW YORK
JLLBC who-is-at CUNYVM.CUNY.EDU
---------------------------