The Activity of people is complex and modeling that activity is tricky.
Robinson's chapter takes the view that models, procedures, and scenarios are "tin ducks" that keep coming back to haunt analysts of computerization. Many analysts, he notes, treat procedures as actions whose outcomes are specified in advance, analogous to the technical notion of algorithms: procedures that are guaranteed to terminate and return an answer.
However, for Robinson, procedures do not necessarily exist independently of inquiries that people make about them. Managing the outcome of a procedure also means managing the procedure through its life, and managing the meaning or interpretation of the outcome. There is in principle no "guaranteed outcome." **Useful human action is locally situated**, whereas procedures are not – they are decontextualized systems of instructions.
Formal procedures can be useful as a "way in" to a situation for beginners, and they cover participants themselves against blame: A participant can say, "we followed the procedure." Procedures can serve as the core objects in a "dialogue over a boundary," namely, a dialogue among those inside and those outside a work situation. Procedures are not algorithms for producing work (because they are contingent and out of context), but instead are representations for consumption by outsiders.
In empirical studies of the information-management activities of people in relation to libraries (not reported here), Neumann and Bishop reported that people fall into two classes: Novices in a field often search for and retrieve information (e.g., from libraries), whereas more expert users "have been at their work long enough to have build up the social networks and power to be able to avoid libraries and command personal forces in seeking information" (personal communication, Bishop, July 9, 1996). That is, the formal information search and retrieval tools and procedures of libraries are often bypassed by experts, who replace them with ongoing social arrangements through which information flows naturally and continuously, without explicit "search" or "retrieval" activities. Formal tool-supported search and retrieval processes are, in effect, replaced by ongoing dynamic networked social processes for experts.
~
BOWKER, Geoffrey C., STAR, Susan Leigh, WILLIAM, Turner and LES, Gasser (eds.), 1997. Social Science, Technical Systems, and Cooperative Work: Beyond the Great Divide. . New York: Psychology Press. ISBN 978-1-315-80584-9, p. 122