Hi,
The question that I raised indeed refers to analytical
units. The interesting thing is that I read Barker’s
study many years ago. I actually liked his neat units
of analysis, and done a study (on a very small scale)
using his concepts. But then, (I believe not only
because of Lave’s critique, but also because of the
Zeitgeist), I changed my position and forgot about the
paper and everything. The funny thing is that when I
asked the question I did not have in mind Barker’s
piece, but was probably looking for Barker’s kind of
elegant and neat system for marking the boundaries of
context.
The idea of the nested system is probably best known
through Bronfenbrenner’s work, but in a way, it seems
to be inherent to any kind of contextual approach to
human behavior: everything is always embedded in
something, it is always a part of some larger
category. Moreover in a paradoxical way, a context (an
activity system) is always more than it is, because it
also includes other contexts, mainly through subjects
who simultaneously operate in multiple activity
systems but also bring their experience from other
activities (as Yrjo wrote about it and Phil reminded
us).
I however have two problems with this:
1. In an empirical study we normally focus on a rather
limited number of levels and more “tangible” units. In
this case we can do two things. First we can take the
perspective of the subjects and arrive at units that
seem to be relevant to them (e.g., for small children:
the level of classroom, but not the level of the
municipality) in which case the boundaries are shaped
by the subjects' experience. The second: we may
certainly also assume that other levels are relevant
even if they are not present in one's experience and
thus we would naturally aspire to examine those levels
too. But then what are our criteria for choosing the
level of our analysis? Within AT, I believe, this
issue is treated by focusing on dyscoordinations and
contradictions within the activity system/s. Yet here
the issue of granularity and the issue of the
perspective emerge again. Do we, for example, focus on
dysfunctions of the educational system country-wide,
or on a particular school or on the problems in the
teacher's handling of the particular reading group?
Moreover, do we focus on the subjects' experience of
difficulties or we look for other indicators (grades,
the rate of production) etc.? I would assume that the
answer is more pragmatic and circumstantial (the
length of the paper, the resources and access to other
levels, our personal experience and knowledge of the
system etc.) Somewhat differently Rogoff (I think Cole
too) wrote about the units for analysis combining both
perspectives and different levels of analysis. In this
sense, I am not sure if AT could incorporate Barker’s
work, but no doubt it is worth of examining.
2. The second problem deals with the nature of the
relations between the levels. The nested system by its
nature (inclusive categories) suggests hierarchic
relations, implying that upper levels shape the lower
ones in which case I would suggest we loose the
incredible diversity and heterogeneity of activity
systems. I am aware that probably most people take the
notion of the nested system in a rather flexible way,
but perhaps we need to have another analytic
categories which would enable us to talk about
different relations between contexts.
Sonja
This archive was generated by hypermail 2b29 : Sun Apr 01 2001 - 01:01:22 PST