some thoughts on boundary objects

Leigh Star (s-star1 who-is-at uiuc.edu)
Wed, 29 Oct 1997 13:27:11 -0700

Dear Friends, I too have been (with Jay, Susan Newman, Jean Lave, Naoki
Ueno, Carol Berkenkotter, Yrjo Engestrom and many other xmcaers) at the
Sociology of Science meetings. So forgive my not coming in sooner to this
very valuable discussion.

When I first wrote about boundary objects in 1989, the purpose was to
answer a specific set of questions: how do social worlds (or communities
of practice) intersect? How do people cooperate over a period of time when
they are sharing an object, but have different defintions of it, and
different uses for it? How does the simultaneous "sameness" and
"difference" of the object play out materially, and in specific work
situations?

Since then I have many times been asked "what is NOT a boundary object"?
The questions usually comes from what Yrjo quite rightly terms theoretical
bloating. There are two parts of the contribution of the boundary object
notion, both equally important. The first is the ambiguity of objects,
that is how they may take on different attributes in different contexts of
use. The second is the way in which they meet specific constraints, needs,
or features of the worlds in which they are used. Because I was looking at
problem-solving, I emphasized knowledge or informational needs. I looked
at issues such as differing time horizons between social worlds, different
units of analysis, different approaches to abstract vs. concrete.

I've always thought that a concept like boundary object should be evaluated
not for its essential qualities, but for the circumstances in which it is
most analytically useful. For me, it is most useful at a middle level of
scale, to explain arrangements made between communities over a reasonably
long period of time. That is, it is not particularly useful to call an
ephemeral event (as you have been fruitfully discussing here), a
once-spoken sentence, or a very large phenomenon (like the Holocaust or
world peace) a boundary object. Each has some of the features, but perhaps
may contribute more confusion that it's worth.

The conceptual background to some of these issues is in the social worlds
literature. I can provide references for anyone interested.

The most difficult thing for me to sort out, and what I've been working on
with respect to boundary object in the last few years, are two sets of
questions:

1. What is the relationship of boundary objects to standards and formal
classification systems?

2. How can we conceive of people's memberships and the "memberships" of
objects within social worlds, and what is the relationship between the two.

Geof Bowker and I have been working on a book which will be done in January
that I think -- I hope -- has some glimmers about these difficult
questions. It's called How Classifications Work, and will be published by
MIT Press. I would be happy to share parts of it on the MCA web page, or
to post brief exerpts to xmca if folks are interested.

Thanks for prompting me to write this down!

Leigh

********

Susan Leigh Star
Graduate School of Library and Information Science
501 East Daniel St. University of Illnois
Champaign, IL 61820 USA s-star1 who-is-at uiuc.edu