People identify apologies as unique types of actions as compared with kin-related moves, which remedy troubles or offenses, such as excuses and justifications (Goffman, 1971; Owen, 1983; Olshtain & Cohen, 1983; Sbisa, 1999). A feature of these apologies is the speaker's acknowledgment of personal responsibility for having caused trouble or offense (i.e., the "guilt") (Goffman, 1971) and a demand to the recipient to be absolved, that is, to either have the offense considered as nonharmful or so minimal it makes the apology irrelevant (Robinson, 2004). It appears that the two common formats for apologizing are "sorry" and "I'm sorry." According to Robinson (2004), who collects them into a single category ("sorry-based" units), they both overtly do the work of accepting "moral responsibility for offensive behaviour and initiate the process of negotiating absolution" (Robinson, 2004, p. 292). Heritage and Raymond (see EJ1088323) distinguish between the two formats of the "sorry" component, highlighting how local offenses (available to both speakers, indigenous to the interaction) are usually supported by "sorry," and more distal and face-threatening virtual offenses are more often supported by "I'm sorry." Elaborating Robinson's argument, and following the discussion opened by Heritage and Raymond, this article extends the analysis of the "sorry"-initiated and "I'm sorry"-initiated apologies. Particularly, the authors investigate to what extent the adoption of one (or the other) of the two formats by the apologizer indexes a different degree of moral responsibility for the offense and to what extent the differences--if any--in the use and sequential management of the two formats relate to how absolution is pursued and accomplished.