Edmundson, R. J. (1992). Evidence for native speaker notions of apologizing and accepting apologies in American English. Unpublished doctoral dissertation, Indiana University, Bloomington, IN.

Attempts to determine (1) how semantic formulae are interpreted by native speakers [so there is the intention and the interpretation], (2) the cues subjects use to interpret the sincerity of an apology and whether it was accepted, and (3) the rules needed to account for variety in interpretations of semantic formulae. Points out that studies looking at strategies in realizing a speech act are problematic in that often one element can be classified in more than one category. 161 native speakers of English (from 8 Intro to Language classes at Indiana U.) were asked to view one of two videos containing six apologies within the discourse of several popular TV programs and to answer several questions concerning each apology. There were some general patterns of interpretation but much variation in the responses. Subjects used mostly prosodic cues to judge sincerity of an apology. Females relied on lexical cues to judge the acceptability of the apology, while the males (1/3 the sample) relied on lexical, paralinguistic, and prosodic cues equally. The researcher found two interpretations of what accepting an apology meant. Some thought it meant acknowledging the offense and forgiving the offender. Others thought it meant that the social balance was fine (either because the social balance was restored or there was never anything wrong in the first place). Edmundson concluded that classification by semantic formula was completely unreliable as subjects might classify a single semantic formula into two or three different categories — i.e., "the choice to justify, explain, or excuse oneself while apologizing is a risky one which may have serious consequences for the social repair at hand" (98). Appropriateness of an apology was rated according to sincerity.