I turn now towards the fictional status of the text under consideration. I focus in particular on the part played by the narrator in author/reader interaction, and I suggest that any analysis of a fictional conversation should include a consideration of these figures as active participants.
Politeness theory has, of course, been used previously to explore literary texts (Ermida 2006, Leech 1992, Short 1989, Simpson 1989b). Ermida and Simpson demonstrate how Orwell and Ionesco respectively use excessive politeness from a socially more powerful speaker to serve some other purpose, Ermida following Chilton’s lead in suggesting that politeness has its more sinister side (Ermida 2006: 844, Chilton 1990: 205). Leech, Short and Simpson apply the framework to dramatic texts, which consist largely of dialogue, and Short argues that a pragmatic reading is useful because a ‘sensitive reader’ will in any case ‘[treat] the text as a series of communicative acts’ (1989: 143). Saki’s dialogue-heavy text resembles a playscript in many respects, with the narratorial comments being akin to stage directions, but I argue here that the role of the narrator is closer to that of another actor or speaker than a hidden director.
It is the narrator who depicts Tarrington’s increasing discomfiture through speech-act verbs (Leech, 1983: 198-228) which emphasise the difference between Clovis’ cool demeanour and Tarrington’s gradual loss of composure. Tarrington is assigned speech-act verbs which highlight his increasing frustration and his progress from off-record to on-record utterances (‘said’, ’resumed’, ’began again’, ’broke in’, ’exclaimed’, ’persisted’); it is clear also that Clovis is controlling the situation by denying Tarrington his conversational turn in ¶9 (‘broke in Mr Tarrington’, my emphasis). Clovis, by contrast, is given five instances of ‘said’, followed by one instance of ‘asked’ when Tarrington baldly announces the fact of their previous acquaintance: this change from ‘said’ to ‘asked’ functions as a marker for Clovis’ tactical shift and as an indication that the exchange has reached a crisis point.
The choice of quoting clauses is one way in which the narrator mediates the text, but what is far more important is the way in which the narrator makes the reader complicit in Clovis’ scheme. Once the aunt has fled, the narrator moves the reader much closer to Clovis and we witness the scene from his perspective. Moreover, we are witnessing events from Clovis’ new assumed perspective, that of being ignorant of Tarrington’s identity. This is marked in the several instances of elegant variation pointing to Tarrington: we know his name, but he is referred to in the text as ‘an affably disposed gentleman’, ‘the object scrutinized’, ’the newcomer’, ‘the candidate for recognition’. In this way, the severity of Tarrington’s FTA is both exaggerated and, crucially, enacted for the reader. The narrator is a co-conspirator with Clovis, and the adopted position of ignorance suddenly imposed upon the text invites the reader also to participate in this conspiracy – all to the exclusion of Tarrington. The narrator converses with the reader because the reader ‘hears’ both the conversation of the characters and the narrator’s interpretation of events. In this case, the narrator privileges one character’s wants over another by several means: in the positioning of the narrative point of view, by emphasising the enormity of the FTA inflicted on Clovis’ negative face, and by ridiculing Tarrington through reference to his persistence in speech-act verbs and accompanying adverbials.
There is not the space here to discuss any useful insights that may arise from an application of Sperber and Wilson’s relevance theory to complement this analysis, nor is there room to explore Kopytko’s suggestion that ‘it is the hearer who assigns politeness to any utterance within the situation in which it was heard’ (1995: 488), but such investigations would almost certainly prove fruitful.