Individuals sometimes feel compelled to confront a rule‐violator. Because the goal of a confrontation is to stop the objectionable action, the violator may feel that his or her autonomy is being threatened and may resist complying. To reduce the likelihood of negative outcomes, confronters are advised to engage in discourse that makes them appear face‐sensitive. However, we argue that the authority of a speaker and the type of directive (imperative or suggestion) that is spoken interact so as to affect the degree of face‐sensitivity attributed to a confronter. We conducted an experiment to test this notion. Consistent with our position, authorities are perceived as more sensitive when expressing suggestions and are attributed coercive potential regardless of the directive enacted Peers, however, are attributed greater coercive potential when communicating imperatives, while face‐sensitivity is unaffected by the type of directive. Implications for confrontation are discussed.

Social systems devise rules for member conduct and often specify punitive action for nonconformity. However, confronting and signaling the intent to punish a rule violator may be an inherently face‐threatening and volatile situation. As such, in this paper we seek to add to the research aimed at minimizing the negative effects of confrontation. We conducted an experiment to examine the impact of linguistic cues and coercive potential on message categorization and on receiver perceptions of threat and face‐sensitivity. Results suggest that threats might be considered a special class of warnings, distinguishable by a speaker‐based locus of punishment Locus of punishment did not, however, impact perceptions of having been warned. These findings thus call into question the assumed parallelism between researcher conceptualizations of threats and warnings and those of typical language‐users. Additionally, targets reported feeling less threatened and perceived more face‐sensitivity, in cases when the speaker was not the source of punishment. Perceptions of threat were decreased when disclaimers were employed and where the message originated from a peer rather than an authority. Power of speech had an impact in ambiguous situations. Implications for researchers and practitioners are discussed.