I treat AT boundaries primarily as a 'unit of analysis' question for several reasons. first, I draw heavily upon Yrjö's work and my interpretation of his formulation with nested systems is that it provides one way to think of the 'specialization' of the activity a system into interdependent (yet analytically separable) activity. 'subject producing' activity is one example of the separation (but neither isolated nor reduced) of a town or city's functions for schooling. (perhaps a bit of emic?) Second, there are parallels between activity systems and what Barker or Sarason have called "settings", but with important differences. Nevertheless, we can draw upon their insights. With Barker, he defined a behavior setting pretty much as the behavior-and-environment, and there are also components called 'synomporphs'. In some ways that i have not examined in detail, Barker's setting structure consisting of interdependent synomporphs is similar to nested activity systems. I think Barkers ways of determining and quantifying interdependence could use some looking at to inform the definitions of activity systems and their boundaries.
So a gymnasium is not a synomorph without the basketball game that goes one there. With Barker, the unit of synomorph pretty much has to do with examining what interaction goes on, and, in my own words, the relative density of interactions within a setting to those between it and other settings. In one way, the analytic (and provisional) isolation can be supported by what structures, either natural or artifactual (technoloigical), appear in the environment, and act to shape the density of interactions -- so for example, the walls of a gym pretty much constrain the basketball game going on there. Percentage of 'Shared space' is a particular way Barker uses to quantify the distinction. Similarly for a town, it may be separable for the purpose of analysis if it is geographically far from other towns. In a second way, and related to the first, the density of interactions may be shaped by the social structures in the setting -- perhaps what are "rules" in Yrjö's formulation. Back to the gymnasium these are the literally the rules for the game, and school rules for athletic events. For the town, this is in part the town government, and its *relative autonomy* from other town systems. There is a bit of ethnography necessary -- having to do with composing a unit that is not unrelated to the distinctions people in the study make. But of course a town (in the US) is embedded in a county that one could also (especially for economic reasons) take as a unit of analysis, and there are interactions with a state government that provides funding, imposes policy, etc. and similarly with a federal government.
Third, at a metatheoretical level I also consider the unit for being able to compose a study, not only that I can get my head around, but I also maintain that the boundaries must also be within what is acceptable for the research community i.e. how much can be included in a research paper or book chapter or book. We cannot just look at the study alone as what forms delimitations on it -- the study per se is a separable (but not a reduction or isolation) component of research activity, and it is shaped by the 'rules' of the research system as well. So for research studies to progress, it may be necessary to change the rules, i.e. develop new methodologies. In psychology, activity theory itself, and similarly Bronfenbrenners ecological psychology, has done this. Editors/reviewers sometimes recognize this contradiction between the research system and its future forms and help towards resolution -- Barbara White for example has a paper that takes up one whole issue of (to the best of my recollection) the journal cognition and instruction.
With my own aera paper, I take as separate systems such social-and-technological structures as the office of superintendant of public instruction (a state office), two towns and their school systems, and a regional support institution ESD 1113. I have not paid much attention yet to clearly defining these separably as 'activity systems' given the time available to get something written before aera. But it is my 'hunch', that I can support these distinctions, drawing in part upon Barker and Engestrom. With these distinctions made by 'fiat' to be the units of systems in the study, I have drawn 'boundaries' around them, and hence this allows me to think of the ways these boundaries become permeable to interactions through them. This is a methodological move that supports analysis of the study. The hunch is based in part of my own knowledge of american systems of government and schooling, and in part of my knowledge of activity sytems and Barkers way of making units.
Bronfenbrenners treatment could also be brought into the discussion, but i have just run out of time. This is a rather audacious posting, and this may be due in part to some situational stress which brings out this side of my personality.
bb
]
>This question, posed locally, seems to me very challenging and I would be
>interested in xmca help on the issue.
>mike
>-----
>
>Hi everybody, I am struging with the following
>question:
>If an activity (system)is a context of action, how one
>identifies its boundaries? Yrjo talks about "crossing
>boundaries" in his paper about policontextuality,
>referring to his examples (a worker moves to another
>production unit). I am not sure, I understood this,
>but I feel this is an important question for AT...
>Maybe, the issue of boundaries is not relevant at all,
>but then, how can we talk about 'crossing boundaries'
>Please talk back, I am desparate for other voices..
>Sonja
This archive was generated by hypermail 2b29 : Sun Apr 01 2001 - 01:01:21 PST