Relevance theory

Relevance is a crucial notion in the Gricean programme, and Grice had called (1989: 86) for it to be clarified. Working in a tradition inspired by Grice, Sperber and Wilson (1987, 1995, 2002) have developed an important theory of pragmatics in which a notion of relevance supplants all the other factors Grice considered. We noted in 4.2 that Grice’s theory depends on an intentional-inferential view of communication. For Sperber and Wilson, in contrast, language-use is an ostensive-inferential process: the speaker ostensively provides the hearer with evidence of their meaning in the form of words and, combined with the context, this linguistic evidence enables the hearer to infer the speaker’s meaning. But the hearer does not do this by entertaining assumptions about the speaker’s intentions and using them to explain infringements of a set of conversational maxims, as in Gricean approaches. Sperber and Wilson argue against any account of utterance interpretation which assumes that hearers decide what meaning or implication the speaker could have intended their words to have and assume that it was this meaning that the speaker intended to achieve. The range of possible meanings that a speaker may intend to convey is, they say, far too great to make this sort of procedure possible (2002: 10–11): the speaker could conceivably have been intending to communicate anything. Furthermore, they complain that Grice’s conversational maxims do not sufficiently narrow down the range of possible interpretations a hearer may reach for an utterance:

There may be a whole variety of interpretations that would meet whatever standards of truthfulness, informativeness, relevance and clarity that have been proposed or envisaged so far. The theory needs improving at a fundamental level before it can be fruitfully applied to particular cases. (Sperber and Wilson 1995: 37)

Rather, Sperber and Wilson claim that there is a single overarching cognitive principle, the Principle of Relevance, which determines the way in which hearers interpret – and speakers intend – utterances.

Relevance, in Sperber and Wilson’s definition (see e.g. 2002: 14), is a potential property of utterances and other phenomena (external events, thoughts, memories) which provide input to cognitive processes:

The relevance of an input for an individual at a given time is a positive function of the cognitive benefits he would gain from processing it, and a negative function of the processing effort needed to achieve these benefits. (Sperber and Wilson 2002: 14)

We can think of relevance as a kind of ‘cognitive nutrition’ (Breheny 2002: 181): a maximally relevant utterance is one that provides the desired information (for instance by answering a question, confirming a hypothesis, or correcting a mistake: Sperber and Noveck 2004: 5), and which does this in the way easiest to understand. Thus, if you ask me what my first name is and I reply ‘Nick’, my answer is maximally relevant: there is no other way I could have conveyed this information more efficiently. If, on the other hand, I reply ‘My first name is an abbreviation of the name derived from the Greek expression “victory of the people”’, then my answer will be considerably less relevant (though still true) because it will cost you considerable effort to interpret.

Every utterance, for Sperber and Wilson, ‘communicates a presumption of its own optimal relevance’ (1995: 158). This is the communicative principle of relevance. What it means is that a speaker implies the relevance of their words by the very act of speaking: in saying something to a hearer, a speaker implies that the utterance is the most relevant that they could have produced under the circumstances, and that it is at least relevant enough to warrant the hearer’s attention. Sperber and Wilson posit that there is a universal cognitive tendency to maximize relevance: human beings seek information from which they will benefit, and this tendency makes it possible, at least to some extent, to predict and manipulate the mental states of others, including those that result in the production of utterances. Thus, speakers will try to produce optimally relevant utterances, in the knowledge that this is what will be most useful for hearers, and hearers assume that the speaker’s utterance is the most relevant possible, given the speaker’s circumstances.

The principles which govern hearers’ interpretations of utterances are described by a twostep comprehension procedure (Sperber and Wilson 2002: 18):

(a) Follow a path of least effort in computing cognitive effects. In particular, test interpretative hypotheses (disambiguations, reference resolutions, implicatures, etc.) in order of accessibility.

(b) Stop when your expectations of relevance are satisfied.

 The most relevant utterance is the easiest to understand. Since the speaker is expected to make her utterance as relevant as possible, the hearer is justified in following the path of least effort by considering the most accessible (obvious) interpretation of the speaker’s words first. ‘The hearer is also justified’, Sperber and Wilson continue, ‘in stopping at the first interpretation that satisfies his expectations of relevance because, if the speaker has succeeded in producing an utterance that satisfies the presumption of relevance it conveys, there should never be more than one such interpretation’ (Sperber and Wilson 2002: 19). Let’s see an example of how the com prehension procedure accounts for the contextual interpretation of an utterance. Consider the following dialogue (taken from Sperber and Wilson 2002: 19):

Sperber and Wilson’s explanation of the processing steps involved here is worth quoting in full:

Peter’s mentally represented concept of a soldier includes many attributes (e.g. patriotism, sense of duty, discipline) which are all activated to some extent by Mary’s use of the word ‘soldier’. However, they are not all activated to the same degree. Certain attributes also receive some activation from the context (and in particular from Peter’s immediately pre ceding allusions to trust, doing as one is told, and defending interests), and these become the most accessible ones. These differences in the accessibility of the various attributes of ‘soldier’ create corresponding differences in the accessibility of various possible implications of Mary’s utterance, as shown in (4):

Following the relevance-theoretic comprehension procedure, Peter considers these implications in order of accessibility, arrives at an interpretation which satisfies his expectations of relevance at (4d), and stops there. He does not even consider further possible implications such as (4e)–(4g), let alone evaluate and reject them. (Sperber and Wilson 2002: 19–20)

Whereas on a Gricean account – as for many other pragmatic accounts – recognition of the literal meaning of the utterance is the basis of its sub sequent interpretation, for Sperber and Wilson this literal meaning is not even considered: Peter simply considers what are, for him at the time, the most relevant possible meanings of Mary’s statement, among which the literal one does not even figure. It is the Principle of Relevance which therefore allows the appropriate piece of factual (encyclopaedic) information about the concept SOLDIER to be selected as the relevant meaning of soldier in this instance.

Note that Sperber and Wilson’s argument depends on the implications of John is a soldier being ranked in the way shown in (4). Their justification for this ranking is that (4a)–(4d) are more accessible than (4e)–(4g) since they are already activated by the preceding context. They acknowledge, however, that the question of which implications are more accessible in a given context is an empirical one, and that their account must be tested experimentally. As they put it (1995: 138–139), ‘it is not enough to point out that information may be carried over from one conceptual process to the next; one would like to know which information is kept in a short term memory store, which is transferred to encyclopaedic memory, which is simply erased’, adding that ‘[h]ere we have neither formal arguments nor empirical evidence for any particular set of hypotheses’. On a traditional picture of linguistic meaning, the literal meaning is always the basis of interpretation: soldier means (something like) ‘member of the military’, and any account of its actual use must take this original meaning into account. We might, for example, see this as an example of a metaphor: the utterance of soldier immediately evokes the literal meaning ‘member of the military’, and the real-world knowledge that members of the military are associated with patriotism, team spirit, sense of duty and discipline is what allows Peter to recover Mary’s metaphorically intended meaning. The confirmation of the relevance theory account therefore depends on an independently confirmed ranking of the accessibility of the different pieces of information associated with an expression.

There is, however, some experimental evidence consistent with Sperber and Wilson’s claim that the ‘literal’ meaning of soldier may not be operative here: Gibbs (2002), for example, surveys a certain amount of evidence in favour of the ‘direct access’ view of interpretation. On this view, speakers go directly to the intended meaning without the literal meaning functioning as a preliminary step (for more discussion and supporting conclusions, see Glucksberg 2004). This evidence has considerable implications for our whole view of literal and non-literal meanings. For the moment, we can simply summarize Sperber and Wilson’s theory by observing that the postulation of a single cognitive principle – Relevance – as the central consideration in utterance interpretation is simultaneously a more comprehensive and a less verifiable theory of communicative behaviour than the Gricean approach which it is intended to supersede. As we have seen, Grice’s rather detailed list of maxims allows for the possibility of specific empirical disconfirmation of individual maxims (barring methodological problems like those discussed in 4.5 above). But Sperber and Wilson’s single generalized principle of relevance is much harder to pin down in the details of the empirical description of language use, whether cross-linguistic or psycho-linguistic. If relevance is assessed against considerations of cognitive effort and gain, how can we measure these without some specific insight into the details of cognitive functioning? How, equally, can we rank the possible interpretations of an utterance with respect to their abstract accessibility without direct investigation of the psychology of the speaker concerned?

Sperber and Wilson are far from the only theorists whose work raises this kind of question, and the newly emerging fi eld of Experimental Pragmatics (Noveck and Sperber 2004) promises to supply at least some answers. In particular, experiments reported in Van der Henst and Sperber (2004) lend support to the idea that principles of relevance are, in fact, operative in certain domains of psychology. There is still a long way to go, however. As Van der Henst and Sperber themselves put it (2004: 169), ‘it would take many more successful experiments involving a variety of aspects of cognition and communication to come anywhere near a compelling experimental corroboration of relevance theory itself’.

On a more general level, the basic presuppositions of theories of meaning like those discussed in this chapter have been the subject of consider able critique. Mey, for example, criticizes Relevance Theory for its treatment of speaker and hearer as ‘spontaneous individual[s] consciously working on unique problems, rather than . . . social agent[s] working within preexisting conventions with resources available to [them] of which [they] cannot be aware’ (1988: 294). This criticism could equally well be advanced against any of the approaches to communication in the tradition of Grice. In their focus on rational interactions between autonomous individuals, such theories can be charged with ignoring both the subconscious and the social determinants of linguistic behaviour. These traditions largely ignore the ways in which what we say and what we take others to mean are constrained by a whole range of factors well beyond the horizon of our individual conscious intentions, including social expectations and unconscious motivations, to name only the two most important. Do we, in fact, always speak with an intention? Could we always say what it is? Do we always react to others’ utterances by trying to determine either their intention, or the specific proposition that is most relevant? On the conscious level, the answer to these questions is clearly ‘no’. On the subconscious one, the answer is much less clear. Many linguists, however, would feel that in spite of these concerns, the inferential picture of communication is the best approach we currently have to the question of the relation between meaning and use. In spite of the problems confronting Gricean and related approaches, something like their view is assumed by most investigators of natural language.

اخر الاخبار

اخبار العتبة العباسية المقدسة

قسم الشؤون الفكرية يصدر عددين جديدين من نشرتي الكفيل والخميس

متحف الكفيل يستعرض حاملَ شموعٍ يعود إلى العصر القاجاري

ذوو شهداء مدرسة ميناب الإيرانية يتشرفون بزيارة مرقد أبي الفضل العباس (عليه السلام)

قسم الصناعات ينجز 75‎%‎ من أعمال إنشاء سور حديدي في شركة عين الحياة

المزيد

الآخبار الصحية

السجائر الإلكترونية والإقلاع عن التدخين.. نتائج متباينة تثير الجدل العلمي

مصر تحقق إنجازاً دولياً جديداً في منظمة الصحة العالمية

دراسة: السباحة فعّالة أكثر من الجري في الحفاظ على صحة القلب

عامل خطر في بداية الحمل يؤخر تطور الكلام والمهارات الحركية لدى الرضع

المزيد

الآخبار التكنلوجية

أحلامك ليست عشوائية.. إليك ما يفعله دماغك أثناء نومك

روسيا.. العلماء يضاعفون الاستقرار الحراري للألواح الشمسية ثلاث مرات

علماء يكشفون سر سرعة الدلافين في السباحة

سلمي أو دموي.. هكذا تنتقل السلطة في مجتمع الجرذان

المزيد

مواضيع ذات صلة

Are the maxims universal?

Questions about implicatures

Infringing the maxims

Gricean maxims and the Cooperative Principle

Conventional and conversational implicature

Discrepancies between truth-functional meaning and utterance meaning

Interpersonal context: implicature

Interpersonal context: speakers intention and hearers inference

Consequences of the illocutionary perspective

Locutionary, illocutionary and perlocutionary acts

Interpersonal context: illocutionary force and speech acts

Problems with the dictionary encyclopaedia distinction