Wednesday, September 22, 2004

Boundary Object

Susan Leigh Star and James Griesemer coined the term “boundary object” in a 1989 article ostensibly about the early days of the Berkley Museum of Vertebrate Zoology. Since then the article has been widely cited in a number of different fields such as computer science, management, and sociology. The idea that certain objects mediate communication between various individuals or groups is inherently quite interesting but hardly novel; exchanging object of significance is an inherently human trait. Star and Griesemer added a new twist in their definition of boundary objects:

“…those scientific objects which both inhabit several intersecting social worlds and satisfy the informational requirements of each. Boundary objects are both plastic enough to adapt to local needs and constraints of the several parties employing them, yet robust enough to maintain a common identity across sites. They are weakly structured in common use, and become strongly structured in individual-site use. They may be abstract or concrete. They have different meanings in different social worlds but their structure is common enough to more than one world to make them recognizable means of translation. The creation and management of boundary objects is key in developing and maintaining coherence across intersecting social worlds.” (Star & Griesemer, 1999 [1989], 508)

This passage is perhaps the most oft quoted description within my personal library of related journal articles. Star herself, however, amended her definition ten years later:

“Boundary objects are those objects that both inhabit several communities of practice and satisfy the informational requirements of each of them. Boundary objects are thus both plastic enough to adapt to local needs and constraints of the several parties employing them, yet robust enough to maintain a common identity across sites. They are weakly structured in common use and become strongly structured in individual-site use. These objects may be abstract or concrete... Such objects have different meanings in different social worlds but their structure is common enough to more than one world to make them recognizable, a means of translation. The creation and management of boundary objects is a key process in developing and maintaining coherence across intersecting communities.” (Bowker & Star, 1999, 297)

Boundary objects were initially formulated in response to Latour, Callon, and Law’s concept of interessement where entrepreneurs gradually recruit allies to stabilize their ideas or inventions. The notion of interessement almost assumes that these allies are total converts to the new innovations. In their study of a museum, Star and Greisemer recognized that although allies were required for the director to successfully complete the mission of the museum, the various actors involved were far from united or converted. To study the surrounding California countryside, the director had to coordinate and communicate with a variety of different individuals and organizations: his patron, the university administration, his professional staff, amateur collectors, naturalists, hunters and trappers, farmers, and any number of random others.

To a zoologist, a sample of flora or fauna only has relevance if it comes with complete documentation. Even though each of the various actors who acted as collectors, the director was able to coordinate them through the use of standardized forms that each would collect. It should be noted that the importance of these forms differed for each individual. To a professional scientist, accurately completing the form was a crucial part in the ritual of empiricism; to an amateur, a properly completed form could be read as a symbol of competence; to trappers, however, a completed form was required before they could collect a bounty for a particular sample.

In the closing paragraphs of their paper, Star and Griesemer articulate four different types of boundary objects (pp. 518-519). Unfortunately their descriptions are rather brief and I suspect that much of the subsequent work on boundary objects has been an effort to assign various documents and artefacts to the types articulated by the authors.

The first type of boundary object is a repository: “ordered ‘piles’ of objects which are indexed in a standardized fashion. Repositories are built to deal with problems of heterogeneity caused by differences in units of analysis.” (518) Libraries are, obviously, repositories. But so are many of the standard knowledge artefacts that we depend on such as encyclopedia, thesauri, and dictionaries. Similarly, Latour’s concept of the “centre of calculation” could be considered a repository. Repositories, however, introduce notions of scale. Certain materials may become artefacts or tools that may or may not involve markings or inscriptions. While a paper based encyclopedia could quite clearly be considered a boundary object, could an entire library? How about an entire government institution? At some point this concept of the repository boundary object must break down due.

The second type of boundary object is an ideal type: “an object such as a diagram, atlas, or other description which in fact does not accurately describe the details of any one locality or thing. It is abstracted from all domains and may be fairly vague. However, it is adaptable to a local site precisely because it is vague; it serves as a means of communication and cooperating symbolically—a ‘good enough’ road map for all.” (518) A typical example of an ideal type is the species. Of course, standardizing an ideal type takes a lot of socio-historical work. Even the construction of various species have become increasingly suspect (cf. Bowker, 1998).

The third type is a coincident boundary. “These are common objects which have the same boundaries but different internal contents. They arise in the presence of different menas of aggregating data and when work is distributed over a large-scale geographic area. The result of such an object is that work in different sites and with different perspectives can be autonomous. However, cooperating parties can share a common referent. The advantage is the resolution of different goals.” (518) Star and Griesemer start to lose me a bit with this concept. Coincident boundaries appear to be relevant only for geographic distinctions and may be difficult to operationalize in other settings. Do coincident boundaries exist in manufacturing settings? What are they? Different geographies? Different departments within the organizations? Do all of the parties involved have to agree on these boundaries and use them in their daily practices? I just don’t know and Star and Griesemer don’t give us any other hints.

The final type of boundary object is the standardized form. Star and Griesemer invoke the concept of Latour’s “immutable mobile” to describe standardized form. Their example from the museum is quite literally a standard printed form that everyone was provided to describe their samples. The standardized form is perhaps the most common example of the boundary object in both the literature and in common practice. JoAnne Yates, for example, describes the development of and importance of printed forms in the development of modern business practices (Yates, 1989). With standardized forms, however, it’s difficult to see how their usage could possibly be strongly structured in local practice or fulfill the information needs of a particular community. How, for example, does the timecard punched by manual labourers strongly structured? I would suggest that for the labourer—just as for Star and Griesemer’s trapper—the timecard is not an accounting tool but rather a means of justifying the pay they receive at the end of the week and perhaps a means of both establishing credibility within an organization (by putting in a lot of hours) or of rebelling against the organization in a manner reminiscent of Autonomist Marxists such as Tronti.

Michael Briers and Wai Fong Chua (2001) have added a fifth type of boundary object: the visionary object. “These are conceptual objects that have high levels of legitimacy within a particular community. They can evoke similar emotive and affective responses form a wide spectrum of people; possessing a sacred quality that makes it difficult for a ‘rational’ person to be against them.” (242) Briers and Chua effectively illustrate the role of visionary objects in a corporate setting where concepts such as “world best practice” play a very important role in aligning a disparate network of organizational actors.

Additional articulation of boundary objects has been provided by Roth and McGinn (1998) and Fujimara (1992). Roth and McGinn note the distinction between face-to-face boundary objects and distant boundary objects. Face-to-face objects mediate direct communication between individuals and are available to both parties. With distant boundary objects, however, there is an inherent power relation between the individuals that may introduce bias into the interpretation of the object.

Fujimara directly addresses one of the limitations of boundary objects. She contends that since boundary objects are value neutral and serve the interests of various parties, they cannot possible serve the purpose of recruitment or interessement:

“Although boundary objects promote collection action and coherence of information from different sites because they are more easily reconstructed (re-represented) in different local situations to fit local needs, they are equally disadvantageous for establishing the… ‘stabilization’ of allies behind ‘facts’. That is, while boundary objects can promote translation for the purpose of winning allies, they can also allow others to resist translation and to construct other facts. They have a wider margin of negotiation.” (174-175)

Fujimara introduces the concept of a “standardized object”—a kind of über-boundary object that contains sufficient granularity to allow value judgements and recruitment:

“A package differs from boundary objects in that it is used by researchers to define a conceptual and technical work space which is less abstract, less ill-structured, less ambiguous, and less amorphous. It is a grey box which combines several boundary objects… in ways which further restrict and define each. Thus, using a package allows for a greater degree of fact stabilization than using boundary objects.” (169-170)

A standardized package may contain a collection of both abstract boundary objects such as concepts and ideal types as well as very specific boundary objects such as standardized tests and forms so that other actors can both accept and continue a particular line of research. The combination of various boundary objects both supports other actors in accepting the theories presented within the package and limits discursive readings of the package. Of course, the exact form that various packages or boundary objects take still requires some articulation.


Bowker, G. C. (1998). The game of the name: nomenclatural instability in the history of botanical informatics. Paper presented at the Conference on the History and Heritage of Science Information Systems.
Bowker, G. C., & Star, S. L. (1999). Sorting things out: classification and its consequences. Cambridge, MA: MIT Press.
Briers, M., & Chua, W. F. (2001). The role of actor-networks and boundary objects in management accounting change: a field study of an implementation of activity-based costing. Accounting Organizations and Society, 26(3), 237-269.
Fujimara, J. H. (1992). Crafting science: Standardized packages, boundary objects, and "translation". In A. Pickering (Ed.), Science as practice and culture (pp. 168-211). Chicago; London: University of Chicago Press.
Roth, W. M., & McGinn, M. K. (1998). Inscriptions: Toward a theory of representing as social practice. Review of Educational Research, 68(1), 35-59.
Star, S. L., & Griesemer, J. R. (1999 [1989]). Institutional ecology, "translations" and boundary objects: Amateurs and professionals in Berkeley's Museum of Vertebrate Zoology, 1907-39. In M. Biagioli (Ed.), The science studies reader (pp. 503-524). New York; London: Routledge.
Yates, J. (1989). Control through communication : the rise of system in American management. Baltimore: Johns Hopkins University Press.

Tuesday, September 21, 2004


I’ve been avoiding this bit of writing for quite some time. It has been dogging me like a hangnail; or a tiny metal filing sliver in the tip of a finger; or like that tongue attracting flap of skin hanging off the top of your mouth after an encounter with a too-hot pizza. Well, you get the idea.

After separating markings from their underlying media, we must try to understand the nature of these markings and how and why they are used. While Ittelson provides some insight on how these markings come into being, he doesn’t provide any illumination on the role that these markings fulfill within a social environment. We need Latour.

Latour reduces the entire life-world of a natural sciences lab to a “cascade of inscriptions.” During his ethnographic observation of science in the making he observed how scientists and technicians would labour over various apparatus and create inscriptions: they filled out forms, punched in numbers, and took observations:

“It seems that whenever technicians are not actually handling complicated pieces of apparatus, they are filling in blank sheets with long lists of figures; when they are not writing on pieces of paper, they spend considerable time writing numbers on the sides of tubes, or pencilling large numbers on the fur of rats. Sometimes they use coloured papertape to mark beakers or to index different row on the glossy surface of a surgical table. The result of this strange mania for inscription is the proliferation of files, documents, and dictionaries.” (Latour & Woolgar, 1979, 48)

Furthermore, the lab was filled with “inscription devices”: machines that—to an uninformed observer—produced nothing but pieces of paper:

“By contrast with the expense and bulk of this apparatus, the end product is no more than a curve, a diagram, or a table of figures written on a frail sheet of paper. Is this document, however, which is scrutinized by participants for its ‘significance’ and which is used as ‘evidence’ in part of an argument or in an article. Thus, the main upshot of the prolonged series of transformations is a document which, as will become clear, is a crucial resource in the construction of a ‘substance’.” (Latour & Woolgar, 1979, 50)

All of this paper combines to form a scientific journal article—the ultimate product of scientific inquiry. Of key importance in Latour’s work is the idea that after an inscription has been produced, all of the intervening steps are forgotten. In the case of a journal article, the months of labour in the lab gets reduces to a few lines of boilerplate text in the methods section. This labour becomes invisible and therefore rhetorically unassailable. Similarly, in the case of inscription devices, the assumptions and work practices of the engineers who created the device becomes black-boxed and invisible. In Latour’s own words:

“One important feature of the use of inscription devices in the laboratory is that once the end product, an inscription, is available, all the intermediary steps which made its production possible are forgotten.” (Latour & Woolgar, 1979, 63)

This process of stripping off the various modalities in texts is a process of fact creation: “A fact is nothing but a statement with no modality… and no trace of authorship.” (Latour & Woolgar, 1979, 82)

Latour discusses how inscriptions are used to reify rhetoric to the status of fact. He also discusses how inscriptions are ironized through the use of rhetoric by “recruiting” various dissenting views from within the realm of science. In Latour’s world, the cascades of inscriptions produce various elite forms of inscription (e.g., journal articles) that go out into the world and do battle with other inscriptions. The winning inscription becomes canonical and continues without modality or trace of authorship (perhaps until a paradigm shift creates new modalities). These winning inscriptions inherently become ready-to-hand for the continuation of scientific studies. Latour does not, however, investigate the ramifications of a potential technological breakdown where these theories become present-to-hand (i.e., paradigm shift) or where the very apparatus used to create the inscriptions similarly fall into question.

In Latour’s later work, he tackled the process by which these inscriptions travel throughout the world and are used to create new knowledge. According to Latour, inscriptions (through some mysterious homing instinct) tend to accumulate within particular locations that Latour refers to as “centres of calculation”. Centres of calculation are essential for “acting at a distance”:

“…how to act at a distance on unfamiliar events, places and people? Answer: by somehow bring home these events, places and people. How can this be achieved, since they are distinct? By inventing means that (a) render them mobile so that they can be brought back; (b) keep them stable so that they can be moved back and forth without additional distortion, corruption or decay, and (c) are combinable so that whatever stuff they are made of, they can be cumulated, aggregated or shuffled like a pack of cards.” (Latour, 1987, 223)

In the pantheon of inscriptions that are both created by and stored within centres of calculation, Latour ranks the equation as the pre-eminent example of the genre (particular Reynolds’s equation for fluid mechanics). Equations are inherently very mobile and very combinable. Through the use of equations, centres of collection are transformed to centres of calculation.

While Latour’s ideas are interesting, he addresses only a very narrow epistemic window represented by scientists and engineers. Given the highly local nature of work practices (“local all the way down” according to Susan Leigh Star (1996)), Latour’s reliance on a single epistemic culture becomes problematic for our understanding of inscriptions as evidenced by his example of Reynold’s equation. While important and widely understood by engineer’s of all stripes, Reynold’s equation may be completely meaningless to other professionals who have to interact with engineers such as production planners or natural biologists. To a biologist, for example, Reynold’s equation is not an immutable mobile powering a particular centre of calculation but rather an impenetrable and meaningless morass of equations, tables, and nomographs. To non-engineers, even the concept of a nomograph can be challenging (see Ferguson, 1992). In this case, the biologist’s background and training introduces an entire set of conceptualizations and rationalizations that introduce modalities as the biologist and engineer converse. So how do we overcome this epistemic gap? Latour doesn’t give us an answer but Susan Leigh Star does: the boundary object.


Ferguson, E. S. (1992). Engineering and the mind's eye. Cambridge, Mass.: MIT Press.
Latour, B. (1987). Science in action : how to follow scientists and engineers through society. Philadelphia, PA: Open University Press.
Latour, B., & Woolgar, S. (1979). Laboratory Life : The Construction of Scientific Facts. Thousand Oaks: CA: SAGE.
Star, S. L. (1996). Working together: symbolic interactionism, activity theory, and information systems. In Y. Engestrom & D. Middleton (Eds.), Cognition and communication at work (pp. 296-318). Cambridge ; New York, NY, USA: Cambridge University Press.