This paper argues that a cross-linguistic study of the lexical semantics of terms used to apologise in English and French can shed light on the conventionalisation of (mock) politeness. Despite the long historical interdependence between the two languages, the linguistic forms most often used to say sorry in Present-Day English and French derive from different etymons. S'excuser and désolé((e)s) are most frequently used in French while ‘sorry’ is more commonly used in English. Both languages draw more rarely on ‘pardon’, and ‘regret’. The paper explores the semantico-pragmatics of apology and aims to identify norms and variations in the expression of apology intra- and interlinguistically. Tokens of SORRY, APOLOGI/ES/SE/ZE, DÉSOLÉ, EXCUSE, PARDON and REGRET were identified and analysed in a range of (mainly) spoken corpora of American, British and Canadian English and Canadian and French French. Distributional patterns highlight differences in the rates and pragmalinguistics of apologies across (different varieties of) English and French but, crucially, also across different genres. While SORRY and EXCUSE have acquired the same illocutionary meaning in English and French, respectively, denoting APOLOGY, ironic uses of apologetic IFIDs are very rare in the data, arguably not denotative, and rely on a pragmatic mismatch to do mock politeness.