More on the definition of "boundary object"

The very first episode of Oddly Influenced was about Star & Griesemer's idea of "boundary objects." That's proven to be a fairly hard concept to pin down, and I did at best an average job. Fortunately, I'm rereading Étienne Wenger's book Communities of Practice to prepare for episode 21, and he adds some useful words.

A difficulty with the boundary object idea is that it's tied up with "meaning," not the most self-explanatory word. In my episode, I said things like "Animal collectors and university trustees attach different meanings to the Musem of Vertebrate Zoology." Fine, sure. But what do I mean by "meaning"?

Wenger has a what I think is a rather idiosyncratic idea of "meaning," but it works well for this purpose. He thinks an experience is meaningful if it either naturally fits a pattern we recognize or, if it doesn't naturally fit, we can make it fit. ("This is way more formal and stuffy and high-class than the usual lunch with the team, but it's close enough.")

A pattern is composed of two things, participation and reification.

Participation is about verbs. It's about the things we do when having the experience. When eating lunch, there's ordering the food (usually at the counter, this time from a waiter), there's the conversation. And not just any conversation: for some teams, "work lunch" includes talk about work, and for others it most definitely does not.

Note that participation is also about people. To Wenger, an experience completely divorced from other people doesn't count as meaningful. (Note: it's actually pretty hard to have a purely personal experience. When I used to give talks, I practiced them a lot, pacing alone in front of a laptop on our kitchen counter. But I was doing that with an audience in mind. Wenger says "Even drastic isolation – in solitary confinement, monastic seclusion, or writing – is given meaning through social participation." Monastic seclusion only makes sense against the background of other forms of living a religious life.)

Reification is about nouns. Wenger: "I will use the concept of reification very generally to refer to the process of giving form to our experience by producing objects that congeal this experience into thingness." From many varied experiences of eating around noon, we have formed a pattern called "lunch" and pushed it out into the world as something that affects practice. Lunch isn't really a thing, it's an abstraction of different experiences. But that's not how the Walgreen's pharmacist who gets a lunch break from 12:30 to 1:30 sees it. And it's not how the person who wrote her lunch break on the schedule sees it.

We humans are super-good at believing abstractions are concrete objects-in-the-world, as real as the rock you stub your toe on.

The point of reifications is that people organize practices around them. Taking a bump of cocaine is not part of a normal lunch, so the Walgreen's pharmacist doesn't get to do that, at least not overtly. Moreover, the existence of "lunch" helps people distinguish between what practices belong at work and what don't. (And the existence of "email" and "Slack" have served to change our understanding of the distinction since the last century.)

So: meaning is formed when the practices and reifications used (or experienced) in some sort of time-delimited experience are recognizable as fitting a pattern. (Of course, it all gets recursive: recognizing an experience, and fitting it to a pattern, is itself a practice, one that will change over time.)


To Wenger, boundary objects are reifications that facilitate participation by people who don't share patterns. A university trustee has no practices that involve skinned corpses in a drawer. An amateur collector has spent no time navigating the politics between east-coast and west-coast universities. But each can help the other.

Wenger puts it like this: "Boundary objects both connect and disconnect. They enable coordination, but they can do so without actually creating a bridge between the perspectives and the meanings of various constituencies."

A way to think about it is that different "social worlds" have adopted a single named reification but attached to it very different practices and ways of participation. Here, from Wenger, are different ways different people can view the same thing:

Modularity: each perspective can attend to one specific portion of the boundary object (e.g., a newspaper is a heterogeneous collection of articles that has something for each reader).
Abstraction: all perspectives are served at once by deletion of features that are specific to each perspective (e.g., a map abstracts from the terrain only certain features such as distance and elevation).
Accommodation: the boundary object lends itself to various activities (e.g., the office building can accommodate the various practices of its tenants, its caretakers, its owner, and so forth).
Standardization: the information contained in a boundary object is in a prespecified form so that each constituency knows how to deal with it locally (for example, a questionnaire that specifies how to provide some information by answering certain questions).

The key point is: "when a boundary object serves multiple constituences, each has only partial control over the interpretation of that object." For example, someone who sends an email cannot dictate how everyone will read it. The way people read it depends on their history: the connotations of words in their particular community, various kinds of email etiquette, what the proposal in the email suggests for practices (probably different for animal collectors than for trustees – note that the writer will have at best sketchy understanding of how trustees will interpret a proposal if the writer is not herself a trustee).


TLDR: .... say... ever notice how adding a "TLDR" section is a practice for particular communities? That helps to define them? Anyway...

  1. Boundary objects are reifications. Thus they are a focus of attention around which people can build patterns of practice; that is, meaning.
  2. As reifications – that is, nouns – they can accommodate lots of different practices, ones not shared between "social worlds" or communities (in the way that the trustee doesn't share "propose agenda items [relating to the museum]" with "baiting a trap [to catch a specimen for the museum]").
  3. The different communities will still have to work out how to work together. Often, some broker will have to operate in both of the worlds to translate from one to the other. (This is the role Grinnell played in the Museum of Vertebrate Zoology.)
  4. Working together will be difficult, but it will be easier than if the two communities had no reifications to coordinate around. By condensing a zillion practices into a single noun, a lot of unnecessary confusion and argument and misunderstanding can be avoided, leaving the focus on necessary confusion and argument and misunderstanding.

Hope this is useful.