Do Categories Have Politics?

Suchman L. (1993) Do Categories Have Politics? The language/action perspective reconsidered. In: de Michelis G., Simone C., Schmidt K. (eds) Proceedings of the Third European Conference on Computer-Supported Cooperative Work 13–17 September 1993, Milan, Italy ECSCW ’93. Springer, Dordrecht.


This is a work of classic CSCW, focusing on the collaborative work of the late 80s and early 90s. Suchman examines system design for communicative practice, particularly Winograd et al.'s language/action perspective. The thesis is as follows: "This paper is an attempt to contribute to a critical re-examination of the place of coordination technologies in CSCW research and development, in particular that class of technologies that seeks to provide canonical frameworks for the representation and control of everyday communicative practices" (pg. 178).

Suchman argues that the adoption of speech act theory in the foundation of system design, by encoding speaker intentions into specific categories, " carries with it an agenda of discipline and control over organization members' actions" (pg. 178). At the time of writing, Suchman points out that speech act theory had become the dominant framework in CSCW.

Speech Act Theory

Suchman defines the basic premises of speech act theory and why it seems to appealing to computer research. The two relevant premises are: (1) language is a form of action; (2) "a science of language/ action requires a formal system of categorization" (pg. 179). The original theory by Austin and Wittgenstein argued it was impossible to theorize language as separate from use, but it had since become the belief that "a theory of language constitutes a theory of action" (pg. 179). She argues that this perspective, that dealing with language is dealing with action, has led to a lack of organizational changes. Some critics have pointed out the ways systems designed to project expected language sequences coerce individuals into using specific interactions. Systems predicated on speech act theory attempt to make intended acts clear, given the presumption that there is either speaker intent in being clear and grounded, or that others will benefit from a clear classification of speech. In the case of speech action systems, like Winograd et al.'s THE COORDINATOR, "the machine becomes the instructor, the monitor of one's actions, keeping track of temporal relations and warning of potential breakdowns" (pg. 181).

Categorization as Discipline

"Categorization has been taken up not just as a resource for analysts but as part of their topic or subject matter; that is, as a fundamental device by which all members of any society constitute their social order. With this move has come a rich corpus of theorizing and of empirical study about just how they do so" (pg. 181). She asks: " What is it about speech act theory that makes it so attractive as a way for practitioners of science and systems design to come to grips with organizational communications? Why do computer scientists go about making up all these typologies of interaction?" (pg. 182).

In describing how language is used to categorize, Suchman writes: " That is to say, systems of categorization are ordering devices, used to organize the persons, settings, events or activities by whom they are employed or to which they refer" (pg. 182). She argues that categorization systems "can be taken up as a resource in the development of more elaborated and formalised systems of social control" (pg. 182). She leans on Foucault's Discipline and Punish, particularly the focus of disciplining specific scientific disciplines as a development to problems associated with administrative power. Breakdowns of communication then become addressed through technological solutionism of a computer science discipline focused on providing order enforced through technology. In THE COORDINATOR, the "basic conversion for action" transforms colloquial expressions into a formalized, quantifiable system of categories reliant on people's willingness to reformulate their intentions into categorical actions. Conversation then becomes closed and can no longer be characterized in any other terms. Winograd et al. visualize the system as making "universal" and "simple" linguistic distinctions, which Suchman critizes as conflating the category's seeming simplicity with a simplicity of the language itself.

In analyzing the words of Winograd et al. about THE COORDINATOR, Suchman points out that the system was explicitly designed to maintain current status quo social orders, not challenge them or provide space for social action. She argues that the language/action perspective believes language must be brought "into compliance with a particular conventional order" (pg. 186). Computer scientists are now not only system designers, but designers of organizations. Like Winner, Suchman seems concerned with how technologies change work practice: "organization members are subjected to ever more elaborated systems of recordkeeping, measurement and accountability.5 Instead of the emancipating alternative that Winograd and Flores would seek, they seem to offer yet another technology designed to create order out of "nature" by, as Haraway would put it, 'policing her unruly embodiments'" (pg. 187). Speech act theory, Suchman argues, offers a model of speech promising a "universal basis for the design of technologies of accountability" which parties using that system are expected to be accountable to (pg. 188). Though she does not state it outright, it seems the answer to the title's question is yes, categories have politics.