Ruis de Mendoza Ibáñez: Implicatures

Francisco José Ruiz de Mendoza
Universidad de La Rioja
Implicatures, explicatures and conceptual mappings
LAUD Essen (LAUD A462). Related works.

This paper attempts to explore the relationship between some crucial aspects of the relevance-theoretic treatment of inference, more specifically the distinction between implicatures and explicatures (cf. Sperber & Wilson, 1986) and the Cognitive Linguistics approach to metaphor and metonymy as ways of understanding and reasoning about the world (cf. Lakoff & Johnson, 1980; Lakoff & Turner, 1989; Lakoff, 1987, 1993).

In Relevance Theory, metaphor and metonymy are treated as cases of interpretive (as opposed to descriptive or truth-conditional) use of language which involve the production of implicitly communicated assumptions or implicatures. In this metaphor and metonymy are no different from hyperbole, litotes and other figures of speech. While this account is innovative, there are two important weaknesses in it: (i) it overlooks the relevance of the mental mechanisms which underlie both metaphor and metonymy; (ii) it assumes that metaphor and metonymy interpretation is a matter of deriving strong and weak implicatures, thus ignoring the fact that the reasoning process for these two phenomena is not of the same kind as for the others. Such explanatory weaknesses have consequences. Thus, the relevance-theoretic account fails to note that the connection between metaphor and metonymy is stronger than that between, for example, metaphor and hyperbole. This special connection has been highlighted by Cognitive Linguistics, where metaphor and metonymy are described as conceptual mappings. However, cognitive linguists have not placed much emphasis on the communicative effects of such mappings.

In order to make both the pragmatic and the cognitive approaches compatible, we shall examine the relevance-theoretic implicature/explicature distinction in connection to metaphor and metonymy. Relevance theorists usually argue that explicatures are derived from the blueprint provided by the logical form of an utterance. This is achieved by reference fixation, disambiguation, and enrichment (see Carston, 1988; Recanati, 1989). These procedures are based on the development or expansion of an initial assumption schema. Against this view, we shall contend that sometimes explicatures are also obtained either through downgrading, as with scalar concepts like frequency and quantity, or through the modification of an assumption schema by means of a conceptual mapping, either metaphoric or metonymic. Finally, we shall explore the role of metaphor and metonymy in the production of explicatures. This will be done in connection with Fauconnier and Turner's (1995) theory of conceptual projection, within which the implicature/explicature distinction may find its place in Cognitive Linguistics.

Sperber & Wilson (1986) have criticized those pragmatists who, following Gricean postulates, have adopted as a working principle the view that any aspect of utterance interpretation which falls outside the domain of disambiguation and reference assignment is an implicature. Instead, Sperber & Wilson have made the interesting claim that some of the cases of what has been regularly treated as implicatures are in fact cases of explicit meaning, which they call explicatures. For them, an assumption is explicit if it is a development of the logical form encoded by an utterance (Sperber & Wilson, 1986: 182). A logical form, in turn, is "a well-formed formula, a structured set of constituents, which undergoes formal logical operations determined by its structure" (ibid. p. 72). When a logical form is semantically complete -and therefore capable of being true or false- it becomes a proposition. Incomplete logical forms are stored in conceptual memory as assumption schemas which may be completed on the basis of contextual information. Since, for Sperber & Wilson, completing an assumption schema -which has a logical form- in order to obtain a proposition -which also has a (more developed) logical form- is an inferential activity (i.e. it exceeds mere decoding), it follows that for them studying the way the logical form of an utterance is developed into its explicature is a matter of pragmatics[i].

Sperber & Wilson, together with other relevance theorists (e.g. Carston, 1988, Blakemore, 1992) have defended that there are three processes involved in getting from an assumption schema to a full proposition: disambiguation, fixation of reference and enrichment. While disambiguation and reference assignment are familiar linguistic phenomena, the notion of enrichment is entirely new. Consider the following example by Carston (1988/1991: 39):

By mere linguistic decoding and fixation of reference we obtain the information that the park is at some distance from where the speaker lives. However, this remark is but a truism in the sense that it is obvious that there must be some distance between the park and the speaker's home. In order for the utterance to be relevant, the expression "some distance", which is manifestly vague, has to be enriched to mean "further away (from where I live) than you think". As Carston (1988) has observed, when we deal with enrichment, the richer explicated proposition entails what is literally said.

In order to work out an implicature, on the other hand, the hearer needs to supply some implicit information which allows him or her to construct a reasoning formula of the condition-consequence type. Consider the following example, from Blakemore (1992: 58):

For B's response to be relevant, A needs to have access to the (implicit) assumption that one's comfort while on holidays may typically be affected by insects (rather than hidden microphones) and an excess of people. As a consequence, we reason that the speaker did not enjoy his holiday. This information is an implicature since it has its own distinct propositional form which functions independently of the explicated information as the conclusion of an argument.

Both Sperber & Wilson (1986) and Blakemore (1992) interpret the lack of literalness of metaphor and other 'tropes' as a matter of producing implicatures. For example, according to Blakemore (1992: 163), the metaphor

These are the more central implicatures. Other weaker ones would have to do with the nature of the neighbour's unfriendliness, together with her behaviour and appearance. It is these weaker implicatures that justify the speaker's not using a non-metaphorical utterance like My neighbour is fierce and unfriendly. Metaphor is thus seen as a way of optimizing relevance, which Sperber & Wilson understand as achieving the adequate balance between processing cost and meaning effects.

Recanati (1989) has made a distinction between two types of enrichment, which he calls saturation and strengthening. Saturation occurs when a sentence has to be filled with information from the context, as in He's not good enough. Strengthening applies in cases like "some distance" above where the richer specification entails the rather more vague explicit one. It is evident that strengthening is not matter of filling in missing information but of making explicit information match contextual requirements by upgrading some scalar concept in it.

Imagine (5) is a complaint where the speaker does not really mean that the hearer makes noise on every possible occasion but only very frequently, perhaps more than can be borne by the speaker. "Always" is used to reinforce the idea that the speaker is just as annoyed as if the hearer never stopped making noise. Also, take (6) as a rather loose way of indicating that the speaker is fascinated by John's ability as a driver, although he couldn't possibly say whether John is the best driver in the world. Now, according to the standard relevance theoretic treatment of hyperbole (cf. Sperber & Wilson, 1986: 235), examples (5) and (6) above would be considered exclusively as a matter of deriving implicatures such as (7) and (8) for (5) and (6) respectively:

(b) The fact that the hearer often makes too much noise bothers the speaker a lot

But it is obvious that there is a qualitative difference between these implications. Thus, the assumptions in (7a) and (8a) are somehow entailed by (5) and (6) respectively. Not so, however, the ones in (7b) and (8b) which, as example (2) above, take the form of implicated conclusions derived with the help of implicated premises like the following:

If the view we are putting forward is correct, when we are dealing with scales like quantity, degree and frequency it is possible to have not only strengthening but also downgrading as a form of deriving explicatures. The reasoning process when interpretation involves upgrading and downgrading -or when saturation is carried out- is one of adapting the logical form of an utterance to match contextual requirements provided that the relationship between what is said and what is implied is one of entailment. On the other hand, implicatures are the result of the hearer engaging himself in a reasoning process of the condition-consequence type, which makes the implied assumption relatively independent of what is said. Thus, (7b) and (8b) neither entail nor are entailed by (5) and (6).

With this in mind, let us go back to the case of metaphor (3) above and its purported implicatures (4a) and (4b). Since the expression is conventional, at least its most central interpretation is fairly straightforward, and is independent of the context. However, let us think of (3) in a context in which the speaker is using this utterance as a remark about his neighbour's latest demonstration of her fiery temper by using very aggressive gestures and bodily movements. In this situation, the uttering of (3) will probably lead the hearer to explore a wider range of implications than those to which he has normal conventional access (i.e. his neighbour being fierce and unfriendly). Thus, the hearer may understand that the neighbour's behaviour troubles the speaker, that he finds her manners particularly unpleasant, and that he is even afraid of her. If this is so, it is these implications that actually qualify as implicatures, since they need intervening propositions which are derived from the context. On the other hand, (4a) and (4b) would be closer to explicatures since they are accessed on the basis of the information explicitly provided by the expression.

This approach to the development of explicatures seems to pose a problem for the applicability of Carston's entailment test. As is well known, entailment tests have usually been used by semanticists to discover sense relations -such as hyponymy- and meaning components in componential analysis and truth-value semantics. Generally speaking, X is said to entail Y when if X is true, Y is true, and if Y is false, X is false (cf. Leech, 1981: 74, 134-136). For example, John is bachelor entails John is unmarried, and She smelt a rose entails She smelt a flower, since it is possible to say "if John is a bachelor, John is unmarried" and "if she smelt a rose, then she smelt a flower". The test holds for examples of enrichment like the one proposed by Carston. Thus, the proposition that "the park is some distance from where I live" seems to be entailed by "the park is further away from where I live than you think" since if the latter is true it follows that the former is also true. However, it does not seem to be applicable in the case of metaphor where it is not possible to say that the figurative expression is true. But the test is not a defining criterion for all types of explicature. In fact, it only deals with the limited range of cases which have so far been analyzed as explicatures by relevance theorists and perhaps not even with all of these. Thus, while it seems to apply well in cases of enrichment by strengthening, it is unclear how it would work when enrichment takes place by saturation. This is the case of the relationship between propositions like "he's not good enough" and "he's not good enough for the job", where one could hardly say that the latter entails the former since the former is an incomplete proposition.

Moreover, it must be noted that there is in fact some sort of entailment relationship involved between what is said in (3) and the implications in (4a) and (4b) since the latter follow from the former by virtue of the conventional meaning of the metaphor. Thus, if a woman is called a dragon, it follows that she is fierce and unfriendly. This form of entailment is essentially the same as the one used by structural semanticists since hyponymic relations (which depend on the way we establish hierarchies) and meaning components (which depend on how we conceptualize reality) are also determined by convention. The difference lies in that the entailment relationship in the case of metaphor cannot be assessed in terms of truth values but only in terms of the internal coherence of the cognitive model involved, i.e. a metaphorical model according to which a person's behaviour is understood in terms of animal behaviour (cf. the discussion of this sort of metaphor in Lakoff & Turner, 1989: 195-198).

In cases of more creative metaphors there would be neither logical nor conventional entailment, but rather what we may call entailment by analogy. For example, in

quoted by Lakoff & Turner (1989: 90), the general shape of an hourglass is mapped onto the general shape of a woman's figure. We may reason that if her waist is an hourglass, she has a thin waist in a certain proportion to the rest of her body, the same proportion that the central part of an hourglass keeps with rest of its sections.

Note, in contrast, that the entailment test is not possible in any of the three forms which we have distinguished when we are dealing with implicatures. Thus, for example (2), it is not possible to say that since the beaches were crowded it follows that speaker B did not enjoy his holidays, simply because we may have a situation in which the speaker does love crowded beaches. This observation combines with our previous discussion on the nature of implicatures to strongly suggest that the division between implicatures and explicatures cricually depends on the type of mental operation involved: while for implicatures, there is a reasoning process of the condition-consequence type where the speaker needs to supply (a set of) relevant premises which derive from the context, for explicatures there is a process of making the initial assumption schema provided by the utterance compatible with contextual requirements by means of adaptations which involve a number of steady procedures. Fixation of reference, disambiguation, saturation, strengthening, downgrading, and conceptual mappings would count as such procedures. In this respect, we must be careful to note that entailment is not a mental operation but a relation which may hold -in different forms- between an assumption schema and each of the explicatures derived from it.

There is one last issue to deal with in this section. Our claim is that (4a) and (4b) above are explicatures for (3). In the canonical treatment of relevance theory, on the other hand, the only possible explicatures for (3) would involve fixation of reference -as in (11) below- or, in the case of what Sperber & Wilson call "higher-level" explicatures (see Blakemore, 1992: Ch.6; Wilson & Sperber, 1993), the specification of a generic speech-act description, as in (12) (cf. also Sperber & Wilson 1986: 225, in this respect):

There is a fundamental misconception in this treatment of explicatures, though. Consider the following example, drawn from Wilson & Sperber (1993: 5):

According to Sperber & Wilson, Mary's utterance in (13b) may have -in the appropriate context- at least the following explicatures of which (14b-d) are higher-level explicatures:

The interpretation of this exchange is not particularly problematic since the higher-level explicatures embed a proposition which has its own truth conditions. Now, imagine that the hearer of (3) above has reasons to think that the speaker both believes and regrets what he is actually trying to communicate about his neighbour's behaviour. The hearer might derive the following additional higher-level explicatures:

But, strangely enough, in order to derive such explicatures the hearer would need to work out the implicatures beforehand. That this is not the case may be observed from a careful examination of other examples. First, imagine with respect to B's assertion in (2) above that the beaches were not very crowded as a matter of fact, at least from A's point of view. In this context, speaker A may derive the explicature in (16):

The implicature that B did not enjoy his holiday is worked out on the basis of this explicature and can be in no way previous to it.

Now imagine that sentence (1) above is produced by the speaker as a warning to the hearer that Central Park, where the hearer wants to go, is not within walking distance of his house, with the further implication that since the hearer has a a car he had better drive there. It is evident that this implication is an implicature since it may not be derived solely on the basis of what is said and that the truistic expression "some distance" is to be developed before this implicature may be derived. Furthermore, if we develop the higher-level explicature of (1) as a warning (i.e. the intention to make the hearer aware that there is a potential problem), this will embed the explicature but not the implicature, as is evident from :

(a) The speaker is warning the hearer that Central Park is further away from where he lives than the hearer thinks.

These observations tend to support the status of implications (4a) and (4b) as explicatures rather than implicatures. They also support the view that what is said in a metaphorical expression is simply a linguistic cue to the production of a set of explicatures by means of a conceptual mapping but not an explicature in itself.

In Cognitive Linguistics metaphor has been studied as a conventional conceptual mapping from one conceptual domain (called the source) to another conceptual domain (called the target). The source usually allows us to understand and reason about the target in terms of some of the relevant aspects of its conceptual structure (see Lakoff, 1993, for details). Thus, when faced with an expression like the one in (3) what we do is reject the literal interpretation and find some (culturally attributed) characteristics of dragons which apply to the speaker's neighbour's behaviour. Since the literal interpretation is never entertained, it may not be either an explicature or a source for explicatures. This is in keeping with what we do to interpret truistic utterances like (1), where what is literally said cannot be an explicature either. In (1) the expression "some distance" needs to be developed into "a longer distance than you expected" by means of enrichment; likewise, in (3) the expression "a dragon" needs to be converted into "someone fierce and unfriendly" by means of a conceptual mapping.

A metaphor is a mapping across discrete conceptual domains. If the mapping is carried out within a single domain, we will have a case of metonymy (see Lakoff & Johnson, 1980: Ch. 8). Consider:

In (18) it was not Napoleon but the army under his command that was defeated. In (19) it is not the fiction character Superman, but Christopher Reeve, the actor, who broke his back. The metonymy in (18) is a way of avoiding the use of a longer, heavier to process -and perhaps rather more vague- definite description. It is also a way of emphasizing Napoleon's more prominent role in the defeat. The metonymy in (19) would normally be used to avoid a long paraphrase like "Christopher Reeve, the actor who played Superman's role" in order to achieve successful reference. However, in spite of these similarities, the two metonymies are different in two respects. First, in (18) we have a metonymy where the target domain (the army) is a subdomain of the source domain (Napoleon), while in (19) it is the source (Superman) which is a subdomain of the target (Christopher Reeve, the actor) (see Ruiz de Mendoza, 1998, for details on the relevance of this distinction). Second, this difference between the two types of domain-subdomain relationship has communicative consequences. Thus, metonymies like the one in (18) allow the speaker to focus on the source domain as the most relevant one while avoiding an uneconomical description like "the army commanded by Napoleon". This mechanism is often useful when the speaker is unable to find the expression that actually designates the intended referent, as in:

where by using the expression "the White House" the speaker evades the problem of accurately describing the type of White House officials actually referred to.

Our discussion suggests that metonymic mappings are also used as ways of developing the explicit meaning of an expression. However, the explicatures produced by metonymic means are different from those produced by metaphor since in metonymic mappings only a single conceptual domain is involved. As a result, the hearer may either focus his attention on one aspect of the domain (a form of conceptual reduction), as in (18) and (20), or on the whole domain as prompted by a selected portion of it (which involves conceptual expansion), as in (19). This has consequences for the entailment test, which is not operative in any of its forms (i.e. logical, conventional, analogical) when we deal with metonymy. Consider the strangeness of reasonings like "if Napoleon lost at Waterloo, then Napoleon's army lost at Waterloo". However, since the propositional form obtained after carrying out the metonymic operation is not the result of the application of a reasoning formula of the kind which is used to derive implicatures, we are allowed to consider metonymy as a device for generating explicatures, the entailment test being only -as we remarked before- a partial test in assigning implicature or explicature status to an assumption.

Our treatment of the division between explicature and implicature has obviously benefited from the understanding of metaphor and metonymy not only as communicative phenomena, but also as cognitive operations whose exact nature pre-determines their communicative potential. In this section, we shall explore how the different interaction possibilities between metaphor and metonymy generate explicatures which are then available for application to specific situations with the subsequent production of implicatures. In this connection, we shall contend that the implicature/explicature division finds its place easily within the theory of mental spaces, as proposed by Turner and Fauconnier (1995) (see also Fauconnier & Turner, 1994, 1996).

In it the contenders are seen as figuratively getting ready for energetic action by girding their loins. In fact, the act of a person girding his loins metonymically stands for a more general situation in which the person is preparing to do something difficult or dangerous. We may diagram the metaphor-metonymy interaction of the figurative expression gird up one's loins as follows:

It may be observed that the most central explicature for (21) is but the final target domain of the mapping complex:

(22) Both sides were making preparations to get ready for a strenuous or otherwise difficult activity.

If we have contextual details about the activity we may derive the corresponding implicatures. For example, imagine that we are dealing with a debate between two contending teams which are determined to fight relentlessly. The following might be plausible implicatures:

These implicatures are derived with the help of information which is related to but independent of the information explicated by the expression.

Now, let us consider Turner and Fauconnier's theory of mental spaces. A mental space is a conceptual packet which draws partial structure from a conceptual domain. In metaphoric and metonymic mappings, the source and target domains of Lakoff's model would be but two input mental spaces which would combine their structure into a richer space, called the blended space. The correlation between the source and target input spaces would be licensed by another space which contains generic structure derived from source and target. For example, in the metaphor He dived right into the problem, the source has a diver and a body of water (e.g. the sea, a lake, a river) which needed exploring; in the target there is a person and a problem which required inspection. Since in both spaces, we have a doer, a doing, and an affected entity, the mapping becomes possible. The blend, in turn, would have such implications as that the person is very meticulous when solving problems, that the problem is so complicated that it requires careful examination, that the person is not afraid of dealing with difficulties, and so on. If we apply this theory to the interpretation of (21) above, it becomes evident that the explicature in (22) constitutes the generic space, while the implicatures in (23) are part of the blend. So, the role of the interaction in (21) is simply to develop the (set of) explicatures which will be used for further inferential activity when set in correlation with the context.

In this figurative expression the mind is seen as text, which can be read and understood; the person's thoughts, which are part of the mind, become readable. The interaction may take the following form:

Here the function of the metonymy is not to help produce the relevant set of explicatures which may allow us to construct a generic space. This is achieved solely on the basis of the metaphoric mapping (in both source and target there is a doer, a doing, and an affected entity). But the metonymy serves to bring into focus the part of the metonymic source (the thoughts) which is relevant for the understanding of the metaphoric correspondence on which it operates. It is interesting to note that while the metonymy in (21) is one where the source is part of the target, in the metonymy in (24) it is the target that is a subdomain of the source. Therefore, it becomes possible to correlate the two types of domain-subdomain relationship with specific functions. Thus, outside an interaction, a source-in-target metonymy, like (19) above, is used to develop a conceptual domain, while a target-in-source metonymy, like (18) above, serves to focus on some relevant aspect of it. In an interaction, on the other hand, source-in-target metonymies serve the additional purpose of signalling the hearer where to find the central inference from the metaphoric mapping. In (21) the metonymic source brings our attention specially to the specific action of the person girding his loins (i.e. making preparations). In contrast, target-in-source metonymies only allow us to understand the nature of one of the metaphoric correspondences, as in (24) above, where by "mind" is meant "thoughts".

In this paper we have argued that a sound understanding of such cognitive processes as strengthening, downgrading and the mapping of concepts, casts some light on the distinction between what is said and what is implicated, a distinction which has communicative consequences. We have proposed that implicatures obey condition-consequence reasoning patterns, resulting in independent assumptions of relatively low strength if compared to explicatures. Explicatures, on the other hand, are the result of developing the blueprint provided by the linguistic expression by means of various cognitive operations, including metaphoric and metonymic mappings. Finally, we have examined the relationship between Fauconnier & Turner's theory of conceptual projection on the basis of four mental spaces, and the way metaphor and metonymy yield sets of explicatures and implicatures. The former provide the necessary information for the construction of generic spaces which will allow projection of the source and target inputs into the blended space, which, in our view, is characterized by containing implicated information.

Carston, R. 1988. "Implicature, Explicature, and Truth-Theoretic Semantics", in Kempson, R. M. (ed.) Mental Representations: The Interface Between Language and Reality, Cambridge: Cambridge University Press, 155-181. Reprinted in Davis, S. 1991. Pragmatics. A Reader. New York, Oxford: Oxford University Press, p. 33-51.

Fauconnier, Gilles & Turner, Mark, 1994. "Conceptual projection and middle spaces". UCSD: Department of Cognitive Science Technical Report 9401. San Diego. [Internet document available from http://cogsci.ucsd.edu and from http://www.wam.umd.edu/~mturn].

Fauconnier, Gilles & Turner, Mark, 1996. "Blending as a central process of grammar". In A. Goldberg, ed., Conceptual Structure, Discourse and Language, 113-130. Stanford, CA: CSLI Publications.

Lakoff, G. & Johnson, M. 1980. Metaphors We Live By. Chicago & London: The University of Chicago Press.

Lakoff, G. & Turner, M. 1989. More than Cool Reason. A Field Guide to Poetic Metaphor. Chicago & London: The University of Chicago Press.

Lakoff, G. 1987. Women, Fire, and Dangerous Things: What Categories Reveal About the Mind. Chicago: University of Chicago Press.

Lakoff, G. 1993. "The contemporary theory of metaphor". In: A. Ortony, ed., Metaphor and Thought, 2nd. ed., 202-251. Cambridge University Press.

Leech, G. 1981. Semantics. The Study of Meaning. 2nd. ed. Harmondsworth: Penguin.

Recanati, F. 1989. "The Pragmatics of What Is Said", Mind and Language 4, 294-328. Reprinted in Davis, S. 1991. Pragmatics. A Reader. New York, Oxford: Oxford University Press, p. 97-120.

Ruiz de Mendoza, F. 1998. "The role of mappings and domains in understanding metonymy", in Barcelona, A. (ed.) 1998. Metonymy and metaphor at the crossroads, Volume in the Topics in English Linguistics Series, Mouton de Gruyter; volume in preparation.

Sperber, D. & Wilson, D. 1986. Relevance: Communication and Cognition. Oxford: Basil Blackwell. (2nd. ed., 1995).

Turner, M. & Fauconnier, G. 1995. "Conceptual integration and formal expression". Metaphor and Symbolic Activity 10: 183-204.

Wilson, D. & Sperber, D. 1993. "Linguistic form and relevance". Lingua 90: 1-25.

[i].- The relevance of the explicature/implicature division for linguistics is probably evidenced from the fact that the linguistic devices which impose constraints on implicatures (e.g. discourse connectives) are different from those which impose constraints on explicatures (e.g pronouns) (see Sperber & Wilson, 1993: 19-23).