Commentary 12 A closer look at Winogradg examples reveals much. Rethinking Documentation and Interface: Reflections on Categorical Approaches Dennis Wixon MS-A-1148 Microsoft Corporation 5600 148th Street NE Redmond, WA 98052 denniswi@microsofl.com ! found Dr. Winograd's paper thought provoking and at times frustrating. Frustrating because I would have liked to ask him about the implications of his examples. Consequently, I'd like to draw out some of the implications I saw in Dr. Wingrad's examples. I agree with Dr. Winograd that "every linguistic act is conversation" and my intention here is to begin a broader conversation about the nature of documentation and interfaces. On seeing the table contrasting interaction with documentation, my first question was what is the dimension or category that underlies each of the rows? Here are my suggestions: Starting with a classic reference manual, the author makes few assumptions about the goals of the audience; the communication is oriented toward completeness (e.g., here are all the parameters of the command). Similarly, the author of a manual makes few assumptions about the context of use, since the object is described completely so this information can be applied in any context. The manual is permanent and offers its own
/lp/association-for-computing-machinery/rethinking-documentation-and-interface-reflections-on-categorical-fkV5S20wQo