Abstract
The customary approach to classifying multiple audiences for written discourse is to recognize primary, secondary, and immediate audiences, and, in some cases, gatekeeping audiences. Based on findings from an ethnographic case study of engineering authors in an R&D setting, this article suggests that authors should also attend to watchdog audiences as they write. A watchdog audience pays close attention to the written transaction between the author and the primary audience. Authors must direct their discourse toward the primary audience, but they must also keep the motives and purposes of the watchdog audience in mind as they write and revise. The watchdog audience in my case study, while it had no direct leverage or other organizational power over the authors, still influenced the authors extensively as they revised their text. Evidence indicates that, beyond the apparent and traditional sources of power, there are more contextual, hidden, socially mediated power relationships equally capable of shaping written discourse.