The idea of boundary objects certainly has many applications and has been expounded upon by many disciplines, including computer science. However, it is impossible to discuss a boundary object without considering the broader context in which it is embedded. In Sorting Things Out, Bowker and Star emphasize the socially constructed and historically specific nature of classification systems, many of which have been internalized to the point of becoming transparent. That is to say, we understand as inherent and natural a system of classification that is actually the product of a very specific sociohistorical context (Bowker and Star, 1999). Boundary objects must be considered with the same eye towards the context containing them. In fact, when Star and Griesemer (1989) introduce boundary objects, they do so within the very particular context of the Berkeley Museum for Vertebrate Zoology. The artifacts that act as boundary objects in that context (specimens, maps, forms, etc.) are specific to the collaboration needs of the communities of practice that are brought together by their ties to the museum.

Reflecting on the evolution of concept, Star (2010), emphasizes the importance of this contextualization. While the fact that boundary objects can be flexibly interpreted is oft-cited, Star argues that less attention has been paid to the specific contexts in which boundary objects come into being. She stresses the fact that boundary objects are created when members of different communities must collaborate, that is they have a specific, common “information and work requirement” but differing interests (ibid, p. 602). Boundary objects allow actors to collaborate without necessarily coming to a consensus. For this reason, it is important to take a process-oriented approach to boundary objects: a defining feature of boundary objects is their ability to “tack back and forth” between being specific and abstract (ibid). A boundary object can be interpreted differently from the perspective of each community of practice that interacts with it. It is also transformed depending on the level of granularity with which it is examined: the loosely-structured object used in common by multiple groups takes on much more concrete features when examined from the specific perspective of one group. However, thinking critically about the context in which a boundary object is embedded, in addition to the ways that it is interpreted, is a much more productive line of inquiry (ibid, p. 613). Perhaps one of the difficulties in effectively engineering boundary objects is the shift in focus away from the processes that bring about boundary objects to the objects themselves.

Comment on this page