Having now had a chance to look more at Gordon's paper, as well as to
respond a bit off-list to Steve (whose constructive response to my
constructively intended comments is a great example of how positive
critical dialogue can be around here ...), I wanted to add a few more
thoughts on the basic issue that Gordon poses.
He begins from the interesting perspective that subject-object relations in
dialogic activity are more akin to subject-subject relations, that is that
in sign-mediated processes, esp. face to face dialogue, we are acting on
one another (or with inner speech, say, on ourselves) as a special aspect
of the object-world. The discourse we produce then functions as a
mediational means for this effective (cf. perlocutionary) action.
But matters are not so simple, as Gordon develops the analysis of his
episode-in-focus. Sometimes the discourse really is the object (too), eg
when we negotiate how best to say something, or how to write it out. And
the discourse has to be expanded semiotically beyond just the words to
include the gestures and the actions-with-artifacts. Now those artifacts
can be mediational means in action directed toward another subject, or they
can be primary objects of action, when we work to "improve" them.
This suggests to me that, somewhat as Gordon proposes, both words and
things, both things-as-signs and things-as-material-interactants (cf.
actants), usefully bridge for us between a focus on (non-subject)-oriented
activity and subject-oriented activity. We can instrumentalize our dialogue
with others when we construe it as contributing to getting something done
on a non-subject object (making the land yacht go, the wheel turn). And we
can semioticize our action-on-an-object to turn it into a topic of talk, or
a part of "text" production (i.e. the things and actions as signs used
communicatively), to enact some agenda of building or maintaining an
interpersonal relationship, or communicating an idea.
Putting both signs and tools inside the activity triangles of
subject-object-means shows us that no phenomenon is inherently either a
sign or a tool; it depends on how we are using it at the moment. Likewise
no actant is inherently a subject or an object; it depends on how it is
construed within the activity at some juncture. All these vertices of the
triangles are ROLES (in the formal sense, i.e. "arguments" of a system of
relations, which is what Arne I think was well aware of). Something
occupies that role because we put it there, not because its nature compels
it to be in one role or another.
This perspective makes a good link to ANT, actant-network theory, an issue
we have often discussed here in the past. An actant is a participant in
discourse that functions semiotically (in Greimas' sense) because of its
role in some (for Greimas, narrated) activity. Roles can be filled by
persons, by things, by signs, by texts, by quoted speech, by reified
processes, etc. Latour makes of this a "symmetry" between persons-as-actors
and things-as-actors, calling both actants-in-networks. This causes a lot
of confusion if we insist on reading it within the traditional modernist
distinction between humanistic stances towards persons-as-subjects or
agents and the disparaged scientific stances towards persons as objects of
study. Some people recoil from equating people and other things because
they fear dehumanization, others because they can't take seriously the
apparent implication that things have human-like agency.
But a sophisticated reading of ANT shows that what Latour is doing is
redefining the notion of agency, moving it away from the sovereign
individual-as-actor, and re-integrating it into a larger social-technical
network: what we can do is a function of what everything else around us is
doing and how we are connected to the rest. This is no different for
nonhuman actants. Once we insert everything into a network, once we shift
ontology so that we don't believe it makes sense to talk about either
people or things in isolation, not part of any network; once some part of a
network becomes out minimal unit of analysis ... then the symmetry of
subjects and objects does not seem so strange.
I think this is exactly parallel to what Gordon, and Arne, are doing in
looking at how signs and tools, subjects and objects, are roles within
activity systems (call them activity networks ...) and the occupants of
these roles can shift role as activity unfolds. In fact, I think the most
interesting implication of Gordon's analysis is that activity-with-dialogue
unfolds precisely BECAUSE of such shifts. This in turns makes it credible
that we could classify the different ways in which dialogue can function as
activity and within activity by the configuration of roles and especially
by the sequence of transformation of roles. This gives us in fact a new
descriptive framework that is more dynamic than the static triangles of yore.
It's not quite phenomenologically dynamic. It's a sequence of snapshots and
a recipe of transformations of roles (a transition network model, states
and rules for going from state to state), but this is a rather common first
approximation to dynamical modelling in many fields. Keep your eye on the
land-yacht: now it's an improvable object, now its a symbolic token, now
it's a tool for transport, now it's the topic of a discourse, now it's a
gift, now it's a missile, now it's commodifiable property, now it's
evidence in an argument. Now it's an agent that makes us do something, now
its an agent that leads us to say something, now it's a model of something
else. The engine of dynamics is running here. Making the world now
enables/affords making the world a bit differently next. We can't ever
quite know in advance what we might decide to do once we see how our last
turn has turned out. It always means and does more than we intended in
doing it. To us as well as to others.
I've ended here with a rather individualistic language to make the point.
It should be corrected back to the less familiar language of the network
.... interactions, events, role-shifts, world-effect happen in the
changing, dynamical network. We can take up a stance within the network,
and we can try to imagine a chunk of the network larger than ourselves
(though always needing some viewpoint from which to see).
Suppose we now add one more feature: how it feels to participate in these
dynamic activity networks. We still don't have a handle on all the relevant
factors that co-determine (over- and under-determine) which transformations
of role happen at each stage. I don't believe there is any complete set of
determiners, from which all futures are predictable. But certainly we can
trace out the contributions of various co-determiners, and one of those has
got to be how we feel in some state of the network-now, and how we've been
feeling across some dynamics of the network recently (and longer term).
This affective aspect should also be symmetrizable within a network
ontology ... but that's a tall order.
JAY.
---------------------------
JAY L. LEMKE
Educational Studies
University of Michigan
610 East University
Ann Arbor, MI 48109 USA
http://www-personal.umich.edu/~jaylemke
---------------------------
This archive was generated by hypermail 2b29 : Fri Nov 01 2002 - 01:00:07 PST