Reflection: Cross-cultural variation in directness

The idea that speech acts vary in their degree of directness has had a huge impact on scholarship. British English culture is usually assumed to favor greater indirectness. Is this just a stereotype? Blum-Kulka and House (1989: 134) report the distribution of the major categories of directness across the languages of the CCSARP project. British English speakers use the most direct category least, so the stereotype seems partly true. But all speakers tend to avoid hints. This might not seem surprising, as conducting conversations in hints raises the risk of misunderstandings and creates a lot of work for hearers. However, it is worth remembering that hints have the advantage of leaving more room for negotiation of meaning across sequences of utterances by participants, a phenomenon which is not necessarily fully tapped into through the discourse completion tests utilized in the CCSARP project.

Other things being equal, one might expect some correlation between these directness strategies and politeness. However, Blum-Kulka (1987) found that informants considered conventionally indirect strategies more polite, arguing that they represented a trade-off between the indirectness required to be polite and not overloading the target with inferential work. Indeed, as Blum-Kulka et al. (1989a: 24) point out, we need to test “the possibility that notions of politeness are culturally relativized, namely, that similar choices of directness levels, for example, carry culturally differentiated meanings for members of different cultures”. This is an important point. For example, according to Wierzbicka ([1991]2003: 36), today in Polish the “flat imperative, which in English cultural tradition can be felt to be more offensive than swearing, in Polish constitutes one of the milder, softer options in issuing directives”. Stronger options include the use of impersonal syntactic constructions with the infinitive. Imperatives, in contrast, assume a second person addressee (some take this to be part of the semantics of imperatives). The explanation, Wierzbicka (2003: 37) suggests, is that “in Anglo-Saxon culture, distance is a positive cultural value, associated with respect for the autonomy of the individual. By contrast, in Polish culture it is associated with hostility and alienation”. However, we would go further in suggesting that, even within Anglo-English cultures, levels of directness in making requests are not necessarily used or perceived in the same way by all members across different contexts.

As for how indirect speech acts work, Searle’s account (1975) is focused on the inferencing required to bridge the gap between the direct speech act and the indirect speech act, and suggests the use of a framework such as Grice’s conversational implicature (1975), coupled with shared background information. Thus, the claim is that the utterance can you pass the salt would first be taken as a rogative speech act (a question), and this direct speech act would flout the conversational maxims. This, coupled with background knowledge (e.g. the bland food, the position of the salt relative to the speaker), would then lead to the computation of the indirect speech act of request. In this account, conventional indirect speech acts have a systematic relationship with the direct speech act’s felicity conditions. Thus, can you pass the salt orientates to the preparatory condition of the speaker having the ability to perform the act denoted in the request. Searle (1979: 73–74) spells out in detail the supposed inferential steps that we are supposed to take when processing the request can you pass the salt. However, even to arrive at Step 1, “X has asked me the question as to whether I have the ability to pass the salt”, the interpreter has already undertaken some inferencing, as s/he has inferred that the utterance counts as a “question” in this context and that they are the target of it. So, in fact, even the first step is not the literal step – the literal interpretation of form – it is supposed to be.

Given that Searle’s account emphasizes the interpretation of speech acts rather more than their linguistics, it makes sense to assess it in terms of psychological validity. Three particular and related questions can be addressed:

1. Are two speech acts entertained by the comprehender, first the direct one and then the indirect one?

2. In order to arrive at the indirect speech act, is the kind of Gricean inferencing outlined by Searle the only route?

3. Is it the case that orientation to felicity conditions is the basis for conventional indirect speech acts?

Answers to the first question are somewhat mixed. Clark and Lucy (1975), for example, seem to find evidence that the literal meaning is computed first and then the indirect meaning, but Gibbs (1983) found the contrary for certain types of indirect requests.

Nevertheless, Holtgraves (1998: 80) states that in the psycholinguistic literature “there does seem to be an emerging consensus that the literal reading of potentially indirect remarks need not be activated to comprehend a speaker’s intended meaning” (note that the wording here allows for occasions when they are activated). If the literal, direct meanings are not always activated, then already we have a challenge to the issue behind question 2 – that indirect speech acts involve two-stage Gricean inferencing, as there would be no deviation from the literal meaning of the utterance to account for. Some forms of indirect speech act, particularly requests, have developed conventional meanings which short-circuit, through associative inferencing, the two-stage processing model implied by the Gricean framework (see discussion of short-circuited implicatures). An indication of this is that, in cases like can you pass the salt, the word “please” is frequently added; pre-verbal “please” creates grammatical difficulties if the utterance is taken as a question. More specifically, Holtgraves (1994) provides experimental evidence to suggest that processing the literal meaning and then the implied, as Grice suggests, is not required for conventional indirect requests (e.g. can you pass the salt), but is sometimes required for non-conventional indirect requests (e.g. this could do with a little salt). As for the final question, Holtgraves (2005) conducted a production experiment in order to assess the ways in which people perform “implicit performatives”, including indirect requests. He found good support for Searle’s proposal that indirect speech acts are performed by referencing the relevant felicity condition.1