'Boundary objects' should I think be limited to material entities that
function in socially related contexts (networks of practices, activities).
Their function, as Leigh makes clear, is to help us think about how
different activities or networks of practices are actually articulated with
one another. And one way in which this happens is because the same material
objects (including, I think, human bodies/organisms) participate in both
(or all), but they are not quite functionally 'the same' in their different
native habitats.
The AT 'object', as Yrjo points out, is less limited to material entities
and serves rather the function of helping us understand how _one_ activity
or system of interrelated practices produces its wholeness or integrity --
how it is in fact one system, one activity (because all the actions in the
service of that activity share an orientation to a common
'object/objective', either presupposed or constituted by the
activity-in-progress itself).
These are rather different theoretical tools designed to help us solve
rather different conceptual problems.
Of course we can try to make some connections between them. We can ask, for
instance, whether boundary objects tend to correspond to mediational means
in AT models, rather than to primary AT 'objects'? not means in general
(i.e. not say language as such) but specific means (say a particular
document or type of document). We can ask when it happens that a community
recognizes two different spheres of activity as being parts of some larger,
over-arching activity, and if, in defining the unifying relationship, some
reference needs to be made to 'boundary objects' that participate in both?
My general sense, however, is that AT 'objects' are formulated at a higher
level of abstraction than are 'boundary objects'. They are more like types,
while the latter are more like things. In between lie classes of boundary
objects (like text-types, genres, categories of tools) and perhaps special
kinds of activity in which the 'object' is simply the deployment of some
mediational means.
If we are going to find a deeper unity of AT and ANT, I don't think we will
succeed by trying to fuzz out or blur the sense of these two kinds of
objects until they seem to include many of the same terms. Instead, we
ought to ask of each the basic question of the other:
ANT to AT: What are the fundamental means by which different activities
come to be socially and functionally interdependent?
AT to ANT: What constitutes the fundamental unity of all the practices
belonging to the same (sub-)network?
I believe that AT gives a fundamentally _historical_ (or
historical-developmental) answer to ANT's question, and that it is the
historical, or more generally the multiple-time scale problem that AT can
help ANT solve. ANT wants to agree with AT about the role and specificity
of history, but it is a little short on theoretical resources for doing so.
It is much stronger at crossing domains in a single era of time than at
crossing eras of time even in a single domain. AT, on the other hand, is
quite good at tracing out the history of a particular activity and the
evolution of its object(s), but I do not know of major studies that show
the history of the linkages between very different activities and their
objects. Are there some?
ANT would not, I think, accept the 'orientation to a common object-ive'
answer of AT to the second question. It would regard this as too
essentializing, as coming too close to a reification of particular
activities (even recognizing their historical contingency), and would
instead look into the very minute details of the practices that constitute
the activity, seeing how they interconnect (via 'internal' boundary objects
and otherwise), and also seeing how they diverge by connecting to other
practices that that are not usually considered part of the 'activity'. It
would then try to consider how people _draw_ the boundary around an
activity, by discourse practices for instance, and might also consider
quantitative criteria for strength or frequency of interdependence. (In my
own version of such a theory, there would be criteria to construct complex
hierarchies of different levels of organization of subnetworks of networks
in supernetworks, etc. as well as the interconnections on each level. ANT's
standard versions are a little more 'flat' than my own.)
more tomorrow -- JAY.
---------------------------
JAY L. LEMKE
CITY UNIVERSITY OF NEW YORK
JLLBC who-is-at CUNYVM.CUNY.EDU
---------------------------