Irony and the Use-Mention Distinction
Contents
1. 1978 iscourse A survey of recent i Descriptive Poetics and the Theory of Literature 3 249 087 Se ee Quintilian 1966 Institutio Oratori PA EA oria H E Butler Trans Loeb Classical Library Lon indirect speech Founda Eds 1976 An Integrated Theory of Lin ds Syntax and plicature In T Bever J J Katz and D T d Theory of Linguistic Abilities New York Crowell The psychological basis of this rhetorical theor ahaa ce y is discussed in Sperber 1975a and 318 Dan Sperber and Deirdre Wilson Sperber D 1975a Rethinking Symbolism Cambridge Cambridge ae Press 3perber D 1975b Rudiments de rh torique cognitive Po tique RHES y Sperber D and Wilson D eee Aea of Rhetorica eory F 3 Stylistics London Penguin 6 f wien Se ace D 1979 and forthcoming Oh Grice s theory of PNT French version in Communications 30 80 94 English version to appear in P Ed Conversation Speech and Discourse London Croom Helm n ae e e mrap me j A about meaning of 226 269 Accomplishments 67 Achievements 67 Additivity 70 Adverbials causal 112 duration 72 74 frame 72 73 frequency 72 74 manner 104 perfective 102 point time 73 time 72 75 112 Adversative relation 111 112 Affirmative proposition 269 Affirmative sentences 3 6 7 192 Akmajian A 18n 58 Allwood J 2 24 57 almost meaning of 257 270 Anaphoric nonanaphoric 226 231 and2. Sperber D amp D Wilson Irony and the Use Mention Distinction Readers are reminded that copyright subsists in this extract and the work from which it was taken Except as provided for by the terms of a rightsholder s license or copyright law no further copying storage or distribution is permitted without the consent of the copyright holder The author or authors of the Literary Work or Works contained within the Licensed Material is or are the author s and may have moral rights in the work The Licensee shall not cause of permit the distortion mutilation or other modification of or other derogatory treatment of the work which would be prejudicial to the honour of the author Reprinted from Radical Pragmatics Cole P ed Wilson D amp D Sperber On verbal irony pp 295 318 1981 with permission from Elsevier This is a digital version of copyright material made under licence from the rightsholder and its accuracy cannot be guaranteed Please refer to the original published edition Licensed for use at University College London for the Pragmatic Theory Online Course ISBN 0121796604 Permission granted December 2003 294 Ivan A Sag P and Zimmer K 1976 On the semantics of compounds and genitives in English Sixth California Linguistics Association Conference Proceedings Lewis D 1972 General Semantics In D Davidson and G Harman Eds Semantics of Natural Language Dordrecht Reidel Pp 1
3. has to do with the fact that ironical utterances often seem to be aimed at a particular target or victim 1 According to the traditional analysis irony and parody involve cuite different production and interpretation processes Irony involves change of meaning whereas parody involves imitation There is no necessary re lation between the two and any similarities that exist must result from similarities in the attitudes of ironist and parodist If irony is a type of mention however it is easy to account for the similarities and differences between irony and parody and for the fact that intermediate cases exist Both irony and parody are types of mention Irony involves mention of propositions parody involves mention of linguistic expressions In other words parody is related to direct discourse as irony is to free indirect dis course 2 Within the traditional framework the existence of an ironical tone of voice is rather puzzling Why not also a metaphorical tone of voice a synecdochical tone of voice and so on When irony is seen as a type of mention the ironical tone of voice falls quite naturally into place It is merely one of the variety of tones doubtful approving con temptuous and so on that the speaker may use to indicate his attitude to the utterance or opinion mentioned 3 It is well known that ironical utterances often involve a switch in style or register For example it is quite common to
4. in this case the set of possible interpre tations becomes to all intents and purposes nonenumerable And if this is so it is hard to see how one could even set about giving an account of disambiguation which is not we repeat a rare and marginal phenome non but a basic factor in the interpretation of every utterance Thus the notion of figurative meaning whatever its value for the anal ysis of figures of speech becomes a real source of difficulty as soon as we look at other aspects of the interpretation of utterances The question is whether these difficulties are caused by the complexity of the data which cannot be ignored or whether they result from some inadequacy in the concepts being used to analyze them Obviously a speaker may sometimes intend to convey something other than one of the literal senses of his utterance When he wants to convey something IN ADDITION TO one of the literal senses the notion of conver sational implicature is relevant This presents no problem for a theory of disambiguation on the contrary it has a role to play in such a theory If figurative meaning could be analyzed in terms of conversational implica ture as Grice has proposed disambiguation would be fairly straightfor ward However in the case of figurative language the speaker normally intends to convey something INSTEAD OF one of the literal senses of his utterance the implicature has to be seen as SUBSTITUTING FOR the literal sense The
5. and understood and to express the hearer s immediate reaction to it Apart from this instant echoing of a preceding utterance there are also cases of echoic mention that are less directly related to what has gone before for instance where the proposition mentioned is not the one just uttered but what the hearer takes to be one of its pragmatic implications FART MRI oh Se es IREAS sere om geen 2 ENE OT UCC TSN RENN o Irony and the Use Mention Distinction 307 31 a Pm a reasonable man b Whereas I m not is what you re implying There are cases where what is echoed is not an immediately preceding utterance but one that occurred some time ago 32 It absolutely poured I know it was going to rain you told me so I should listen to you more often There are cases of echoing where the sources are very distant indeed 33 Jack elbowed Bill and Bill punched him on the nose He should have turned the other cheek as it says in the Bible Maybe that would have been the best thing to do There are also what one might call anticipatory echoes 34 You re going to do something silly You re free to do what you want you ll tell me Maybe so But you still ought to listen to me Such cases of echoic mention are extremely common in ordinary con versation and considerably more varied than we have time to show here In each case the speaker s choice of words his tone doubtful question ing scornful co
6. distant is none theless clear Even when there is no prior utterance some vague echoing is still involved One normally sets off for a walk in the hope or expecta tion of good weather What lovely weather may simply echo these earlier high hopes In all these cases the remark in 1 is interpreted along exactly the same lines There is no question of a move from one figure of speech to another or one type of irony to another with quite different interpreta tion processes being involved the only move is from obvious cases of echoic mention to much vaguer and duller varieties of the same thing It seems to be raining Suppose someone had originally made this re mark just as the rain was starting By repeating it in the middle of a down pour the speaker of 2 shows how laughable it was in retrospect to be in any doubt about whether it was really raining Even when there is no prior utterance 2 would have a similar effect By pretending a degree of hesi tancy which is completely inappropriate in the circumstances it conjures up a picture of a quite ludicrous degree of inattention or failure to react It should be obvious without further contextualisation that 3 I m glad we didn t bother to bring an umbrella and 4 Did you remember to water the flowers are naturally interpreted as ironical echoes of advice on the one hand and obsession on the other which are both totally irrele vant in the circumstances Whether the advice wa
7. example it is often possible to process an utterance quite satisfactorily without consciously noticing that it is an utterance of a particular logical type closely related to quotation When the mention does not involve reported speech proper it is less easily iden tifiable still and it would not be too much of a surprise to come across whole classes of implicit mention of propositions that have so far been overlooked or misinterpreted Ironical utterances are a case in point 5 IRONY AS ECHOIC MENTION Consider the following exchanges 27 a I ve got a toothache b Oh you ve got a toothache Open your mouth and let s have a look 28 a Where can I buy pretzels at this time of night b Where can you buy pretzels At this time of night At Barney s of course 29 a I m tired a b You re tired And what do you think I am 30 a Doolittle Listen here Governor You and me is men of the world ain t we b Higgins Oh Men of the world are we You d better go Mrs Pearce G B Shaw Pygmalion In these examples the propositions used in a are implicitly mentioned in the responses in b These cases of mention are clearly not reported speech in the sense that they are not intended to inform anyone of the content of a preceding utterance such an intention would be pointless since the utterance has only just occurred Rather they are meant to in dicate that the preceding utterance has been heard
8. idea that an implicature could actually contradict the literal sense of an utterance as it would in the case of irony does not square with Grice s central claim that implicatures act as premises in an argument designed to establish that the speaker has observed the maxims of conver sation in saying what he said It follows that the interpretation of ironical utterances cannot be reduced to the search for conversational implicatures without grossly distorting the notion of implicature itself Grice does not succeed in integrating figurative interpretations into his overall pragmatic theory For more detailed discussion of Grice s treatment of figurative language see Wilson and Sperber forthcoming a eS a a ea ta Ni ts eh Td Re a ee ae ES Re i a 300 Dan Sperber and Deirdre Wilson This being so if the substitution theory of irony is correct some notion of figurative meaning seems called for and the problems it gives rise to seem to be forced on us by the data themselves However if it were possi ble to give just as good an account of the data but without any appeal to the notion of figurative meaning and using only independently motivated concepts literal sense implicature efc it is clear that this would be preferable to the traditional account An approach along these lines has already been suggested for the cases of metaphor synecdoche and meton ymy in Sperber 1975b in this chapter we attempt to extend the an
9. made in the case of irony Finally it fails to show that the conversational implicatures involved in irony are of the same type as the more standard cases of conversational implicature to which they are supposed to be assimilated For these rea sons Grice s purely pragmatic account of irony also fails In this chapter we offer an account of irony that goes some way toward solving the problem raised by both traditional semantic and pragmatic ap proaches In particular it explains why ironical utterances are made and why they occasionally but not always implicate the opposite of what they literally say Unlike the traditional theory it makes no reference to the notion of figurative meaning Unlike both the traditional theory and Grice s account it involves no substitution mechanism whether semantic or pragmatic Unlike Grice s theory it assumes that there is a necessary though not sufficient semantic condition for an utterance to be ironical Furthermore the crucial fact that ironical utterances convey not only propositions which can be accounted for in terms of meaning and impli Irony and the Use Mention Distinction 297 cature but also vaguer suggestions of images and attitudes finds a natu ral description in the framework we propose 2 SOME METHODOLOGICAL PRELIMINARIES There are a number of obvious similarities between linguistics and the study of rhetoric Rhetorical judgments like linguistic judgment
10. proper interpretation of an ironical utter ance is assumed to consist solely of conversational implicatures logically derived according to pragmatic patterns of inference If what we have been arguing in this chapter is correct then Grice s and other similar ap proaches to irony and more generally to figures of speech are too radi cal in one respect and not radicai enough in another In the first place ironical utterances do have one essential semantic property They are cases of mention and are thus semantically distin guishable from cases where the same proposition is used in order to make an assertion ask a question and so on As has been seen this semantic distinction is crucial to the explanation of how ironical utterances are in terpreted and indeed why they exist Without this distinction the echoic character of irony will be overlooked and it will thus be impossible to make the correct prediction that where no echoing is discernible an utter ance however false uninformative or irrelevant will never be ironical In this respect then a purely logical pragmatic approach to irony is too radi cal In a second respect though a logical pragmatic approach to irony re mains too close to a semantic one In both cases it is assumed that the interpretation of an ironical utterance consists solely of propositions whether entailments or implicatures intended by the speaker and recov erable by the hearer
11. show that one s ut terance is ironical by changing to a more formal or pompous style 7 It is widely accepted though not entirely uncontroversial that free indirect speech may be used to ironic effect see McHale 1978 p 275 276 and references therein What is lack ing in the extensive literature on this subject is any attempt on the one hand to explain why this connection should exist and on the other hand to construct a unified theoretical ac count of irony around it In this article we hope to have shown that such an attempt is worth making 312 Dan Sperber and Deirdre Wilson 41 That s done it you ve broken the vase I hope you re satisfied my lady There is nothing in the traditional account of irony that would lead one to expect such shifts However they can be quite easily explained on the assumption that irony involves echoic mention of a real or imagined utter ance or opinion In 41 the speaker is echoing the sort of deferential re mark that he implies the hearer is expecting As in free indirect dis course the implicit mention of a Proposition sometimes involves mention of an expression 4 From the point of view of the traditional theory there is a strange asymmetry in the uses of irony One is much more likely to say How clever to imply How stupid or How graceful to imply How clumsy than the other way round This connection of irony with implications of failure to reach a certain stand
12. the point of the indirect approach What difference is there between saying What lovely weather ironically and What awful weather literally On both standard semantic and pragmatic accounts there is nothing to choose between these two remarks and therefore no particular reason for ever being ironical This is clearly not right Moreover the data do not entirely support the standard description of ironical utterances The only clear intuition is that the speaker of 1 does not mean what he has literally said and lets it be understood that what he has literally said is the opposite of what he really believes Obviously his real beliefs can be deduced from this but we cannot necessarily conclude that his main in tention or even a subsidiary intention was to get these beliefs across He might instead have been trying to express an opinion not about the weather but about the content of 1 itself to indicate for example that it had been ridiculous to hope that the weather would be lovely It seems to be raining This clearly does not express the opposite of what the speaker thinks it just expresses Less than what he thinks Whereas 1 was odd because the speaker did not believe what he said 2 is odd because its truth is so patently obvious Although it might have been relevant or informative as the first few drops of rain were falling in the middle of a downpour it could never be seriously made except by A similar point is
13. 123 186 Antisubee divisibility 70 Approximating expressions 268 Aristotle 67 Aspect 75 79 perfective 112 117 120 123 124 125 126 Assertions 131 131n 133 139 140 155 Assumed familiarity 233 239 243 245 248 250 252 253 Assymmetrical expression 264 Atemporality of states 71 Atlas J D 2 3 11 24 24n 25 34 48 52n 55n 57 58 Austen J 307 308 Austin J L 85 85n 99 175 180 B Bahl K C 118 119 27 Bailey T G 103 110 t11 27 Banfield A 304n 3 7 Bar Hillel Y 41n 58 Barwise J 284n before relation 69 70 72 75 Bennett M 71 72 80 Bhatia T K 112 115 116 124 127 Bivalent language 22 Boer S A 57 58
14. 69 218 Montague R 1968 Pragmatics In R Klibansky Ed Contemporary Philosophy A Sur vey Florence La Nuova Italia Editrice Reprinted in Montague 1974 Montague R 1970 Pragmatics and intensional logic Synthese 22 68 94 Reprinted in Montague 1974 Montague R 1973 On the proper treatment O Hintikka er al Eds Approaches to Natura printed in Montague 1974 Montague R 1974 Formal Philosophy New Haven Yale University Press Nunberg G 1977 The Pragmatics of Reference Unpublished doctoral dissertation C U N Y Graduate Center New York Reprinted by Indiana University Linguistics Club 1978 Nunberg G 1979 The Non Uniqueness of Seman Philosophy 3 2 143 184 Scott D 1970 Advice on modal lo Logic Dordrecht Reidel Stern J 1979 Metaphor as d metaphors Unpublished doctoral dissert Kay f quantification in Ordinary English In J Language Dordrecht Reidel Re tic Solutions Polysemy Linguistics and gic In K Lambert Ed Philosophical Problems in emonstrative a formal semantics for demonstratives and ation Columbia University 4 Irony and the Use Mention Distinction Dan Sperber and Deirdre Wilson 1 INTRODUCTION Pict neta utterance is traditionally analyzed as literally saying one m gt e meaning the opposite Thus the ironical remark ee ovely weather would have the figurative meaning What awful Ee es a fe so se saa semant
15. Now it has long been recognized that the under standing of figures of speech and of irony in particular has nonproposi tional nondeductive aspects An ironical utterance carries suggestions of attitude and sometimes as in 49 of images which cannot be made entirely explicit in propositional form In this respect a logical pragmatic model does not provide a better description let alone a better explana tion than a semantic model On the other hand our analysis of irony as a case of echoic mention crucially involves the evocation of an attitude that of the speaker to the proposition mentioned This attitude may imply a number of propositions but it is not reducible to a set of propositions Our analysis thus suggests that a logical pragmatic theory dealing with the Irony and the Use Mention Distinction 317 interpretation of utterances as an inferential process must be supple mented by what could be called a rhetorical pragmatic or rhetorical theory dealing with evocation To conclude The value of current pragmatic theory as inspired b Grice s work lies mainly in the fact that it relieves semantics of a numb 4 of problems for which it can provide a more general and AE ee treatment However in the case of figurative utterances the move fro semantics to logical pragmatics merely creates a number of new b lems without providing solutions to many of the problems raised by tk traditional semantic appro
16. S Irony and the Use Mention Distinction 301 was his primary intention How would it be recognized Ironical utter ances are not always distinguishable by intonation from their literal coun terparts When there is no distinctive intonation it is clear that the choice between literal and ironical interpretation must be based on information external to the utterance contextual knowledge and other background assumptions rather than the form or content of the utterance itself Where such external information is lacking for example where 1 is said in the course of a long distance telephone call then the utterance would certainly be taken as literal In other words knowing the speaker s beliefs about the weather is a precondition for rather than a consequence of recognizing that his utterance was ironical The standard approach to irony which claims that the main point of an ironical utterance is to con vey the opposite of what is said would thus make every ironical utterance uninformative both on the level of what is said and on the level of what is implicated The speaker would be intending to communicate a certain be lief but in the absence of any special intonation his intention would only be recognized by someone who already knew that he held that belief We have already mentioned another problem with this account If the speaker of 1 meant to indicate that he thought the weather was awful why not say so directly What is
17. ach Taking irony as an example D to show that given an adequate semantic analysis of ironical AANE as echoic mentions the problems with both the traditional semantic 4 count and Grice s pragmatic account dissolve away A number of ae lems still remain what we are suggesting is that logical pragmatics ie in turn be relieved of these problems which can be given a more general w explanatory treatment within the rhetori i cal component of tended pragmatic theory P ARAE REFERENCES Banfield A 1973 Narrative st ld yle and the grammar of di tions of Language 10 1 39 oer Bever T Katz J J and Langendoen D T guistic Abilities New York Crowell Si Clee Fae and Semantics 9 Pragmatics New York Academic Press ay organ J d i ON s 1975 Syntax and Semantics 3 Speech Acts New York Cutl sayi i u ee se On saying what you mean without meaning what you say In Papers from as n egional Meeting of the Chicago Linguistics Society Depart i Be guistics University of Chicago Pp 117 127 l Gi ae pe P 1975 Logic and conversation In P Cole and J Morgan E PT ane 3 Speech Acts New York Academic Press Pp 41 58 ice H P 1978 Further notes on logi ion In P Cole gic and conversation In P 7 Semantics 9 Pragmatics New York Academic Press Pp T SASE A arnish R M 1976 Logical form and im Langendoen Eds An Integrate Pp 313 391 McHale B 1978 Free indirect di B
18. alysis to irony The problem will be taken up in more detail in Sperber and Wilson forthcoming 3 SOME BASIC DATA Consider the utterances in 1 and 2 repeated here for convenience and 3 8 exchanged between two people caught in a downpour in circumstances that are otherwise normal 1 What lovely weather 2 It seems to be raining 3 I m glad we didn t bother to bring an umbrella 4 Did you remember to water the flowers 5 What awful weather 6 It seems to be thundering 7 l m sorry we didn t bother to bring an umbrella 8 Did you remember to bring in the washing There are two obvious ways of grouping these examples First there are close syntactic and lexical parallels between 1 and 5 2 and 6 3 and 7 and 4 and 8 Second in a less straightforwardly definable way 1 4 have something in common which distinguishes them from 5 8 Consider 1 4 in turn a What lovely weather In the circumstances described it is inconceiv able that the speaker meant to get across the literal meaning of his utter ance In fact it is certain that he believes the opposite of what he has said However it is not so obvious that it was this belief that he primarily in tended to get across as would be claimed by both the semantic and prag matic accounts of irony referred to above In the first place suppose this oe ee ee ne d o ed a TAE RE t ATH IY A TR aR EAA
19. ard has often been noted There is no expla nation for it in terms of the traditional process of meaning inversion which should be able to work just as well in one direction as in the other However on our account there is a straightforward explanation Stan dards or rules of behavior are culturally defined commonly known and frequently invoked they are thus always available for echoic mention On the other hand critical judgements are particular to a given individual or occasion and are thus only occasionally available for mention Hence it is always possible to say ironically of a failure That was a great success since it is normal to hope for the success of a given course of action How ever to say of a success That was a failure without the irony falling flat the speaker must be able to refer back to prior doubts or fears which he can then echo ironically In the face of an imperfect reality it is always possible to make ironical mention of the norm In the face of a perfect reality there must be past doubts or fears to echo if the mention of a criti cal judgment is to count as ironical 5 The claim that ironical utterances are aimed at a particular target or victim is based ona variety of intuitions sometimes clear cut sometimes less so Within the traditional framework there are two quite separate processes that might account for this aspect of irony On the one hand every utterance whose literal sense would Carry over tone
20. d because like 2 it is so obviously irrelevant in the circumstances The speaker is not interested in the an swer he is much more likely to have asked the question precisely to high light its irrelevance and the pointlessness of asking or answering it in the circumstances If we also suppose that the hearer is fanatical about keep ing his flowers watered 3 will have the further implication that the ques tion is USUALLY pointless and that the hearer s obsession is ridiculous Thus what the speaker actually communicates is not question 4 itself but an attitude to it and to the state of mind that might give rise to it Although incomplete and imprecise these observations about 1 4 do at least make clear what these utterances have in common that distin guishes them from 5 8 Someone who utters 1 4 cannot but disso ciate himself from the content of his utterance either because it is clearly false as in 1 and 3 or because it is clearly irrelevant as in 2 and 4 The only way to understand him is to assume that he is expressing a belief ABOUT his utterance rather than BY MEANS OF it What 1 4 express is an attitude of the speaker to his utterance whereas what 5 8 express is an attitude of the speaker to what his utterance is about the weather the rain and the appropriate steps for dealing with them This distinction between two basic types of utterance is entirely missed by both standard semantic and prag
21. die s name was mentioned What am I to do with such a woman as this 25 The master began to understand and to share the intense disgust which the archdeacon always expressed when Mrs Proudie s name was mentioned What was he to do with such a woman as this he asked himself 26 The master began to understand and to share the intense disgust which the archdeacon always expressed when Mrs Proudie s name was mentioned What was he to do with such a woman as this Trollope Barchester Towers In formal language only the explicit mention of an expression is possi ble this is illustrated in 23 However implicit mention of an expression as in 24 explicit mention of a proposition as in 25 and implicit men tion of a proposition as in 26 are equally clear cases of mention from a logical point of view unless the concept of mention is arbitrarily re stricted to mention of expressions It is from a linguistic point o view that mention of a proposition is harder to identify than mention of an expres nar i nai aad e hat aa ainat aa itd a aariat ad Saat ea 306 Dan Sperber and Deirdre Wilson sion and implicit mention harder to identify than explicit mention hence the impression that there are mixed forms The most difficult cases to identify are those where a proposition is implicitly mentioned When the context gives no indication that the free indirect style is being used for reporting speech as in 26 for
22. er the analysis of irony as a type of mention does involve a quite central claim which has no equivalent in the traditional framework and which by itself provides a more satisfactory explanation for a much wider range of intuitions Within our framework an ironical remark will have as natural target the originators real or imagined of the utterances or opinions being echoed If the remark also carries critical overtones or if the hearer fails to detect the speaker s ironical intent the ironical effect may of course be reinforced but it may equally well be achieved when neither of these conditions is present In example 46 the victim is Jeremy because the utterance echoes an opinion of himself that he expects to hear In 44 the target is Billy be cause the utterance echoes an opinion imputed to him that he is nearly there In 42 the victims are all those who think or claim that Fitzgerald plays by the rules Fitzgerald himself and in certain circumstances the hearer too will be a victim in virtue of this In 48 it is the hearer who is the victim because the utterance echoes an opinion he is believed to hold In 1 if the weather forecast has predicted good weather it is this fore cast that is echoed in the remark What lovely weather on the other hand if no one in particular has actually made such a prediction our ac count correctly predicts that the irony is not aimed at any particular vic tim The analysis of
23. figurative senses rather than implicatures of 35 37 Then either the proposition that constitutes the figurative meaning must be understood as USED and the status of the utterance as echoic mention will disappear or it will be understood as MENTIONED and since it is not patently false inappro priate or irrelevant there will be no way of explaining the speaker s atti tude of mockery or disapproval Either way an account in terms of figu rative meaning will necessarily overlook a central and obvious aspect of the interpretation of the utterance The analysis we are proposing although it involves implicatures dif Irony and the Use Mention Distinction 309 fers from Grice s in at least two important respects Grice sees violation of the maxim of truthfulness as both a necessary and a sufficient condition for ironical interpretation When an utterance is patently false he argues the hearer interprets it as implicating the contradictory of what was liter ally said We have already mentioned one problem with this account Un like standard implicatures the implicature carried by an ironical utter ance must be substituted for rather than added to what was literally said because otherwise the total message conveyed would be a contradiction A more general problem is that violation of the maxim of truthfulness is in fact neither necessary nor sufficient for ironical interpretation It is not necessary because of the ex
24. ic theory designed to incorpo count would have to provide first iti uch a definition of figurati ae 1a urativ ee soa a mechanism for deriving the figurative ete sf an cas ans ee eas for explaining why figurative utterances peaker should prefer the ironical weather to its literal counter Meese part What awful weather whi i i t i which on this anal y means exactly the same thing It is because they provide no o to such questions that traditi i 1t10 fail nal semantic accounts of irony ultimately A sh i i Be Nee of one Pad appeared in French in Po tique 36 399 412 We iane Brockway R i silt niin Rishi y Robyn Carston and Julius Moravesik for a number of For intili i i example Quintilian defines irony in terms of the fact that we understand something which is the opposite of what i i intili CORDA at is actually said Quintilian IX H p 44 See also Turner 295 RADICAL PRAGMATICS Copyright 1981 by Academic Press Inc All rights of reproduction in any form reserved TERNE 1S marge 296 Dan Sperber and Deirdre Wilson At first sight Grice s 1975 1978 pragmatic approach to irony looks more promising than the traditional semantic approach Grice attempts to reanalyze the notion of figurative meaning in terms of his independently motivated category of conversational implicature Thus for Grice ironi cal utterances would conversationally implicate rather than figuratively mean the opposite
25. ird time that the ironical interpretation is really forced on us and from then on it increases in intensity In the traditional framework we would have to say that at the second or thire repetition the literal meaning is replaced by the opposite figurative meani 1g there can be no intermediate stage between literal and figurative interpretation In our framework Mark Antony has to be seen from the very first as MENTIONING the proposition that Brutus is an honorable man He first mentions it ina conciliatory tone of voice No doubt it is not his own most personal opinion but he is prepared to put it forward in a spirit of ap peasement echoing the sentiments of Brutus s supporters Then each time he repeats it he mentions it in the context of further facts which make it clear that he is dissociating himself from it more strongly each time The irony is first hinted at then strengthened then forced home Mark Antony carries his audience with him through a series of succes sively more hostile attitudes to a Proposition which itself remains un changed from start to finish At every stage the proposition is mentioned and not used This example brings us back to our preliminary remark that the concept of irony itself is open to reconsideration not just in its intension but also in its extension In classical terms an utterance either is ironical or it is not The picture we are suggesting is different Although an utterance ei ther is or i
26. irony as a type of mention thus makes it possible to pre dict which ironical utterances will have a particular victim and who that victim will be When the utterance or opinion echoed has no specific origi nator there will be no victim when there is a specific recognizable origi nator he will be the victim Thus when the speaker echoes himself the irony will be self directed when he echoes his hearer the result will be sarcasm In the traditional framework the ad hominem character of irony is a function of the propositional content of the utterance in our frame work it is a function of the ease with which some originator of the opinion echoed can be recognized The many cases where these two accounts make different predictions as in 48 should make it possible to choose between them 7 A FINAL EXAMPLE In Shakespeare s Julius Caesar Act 3 Scene 2 Mark Antony says six times that Brutus is an honorable man This is frequently cited as an ex ample of irony but on closer examination it raises a number of problems for the traditional theory The first time Mark Antony says Brutus is an RAO OO LRTI BAI om SEAM Mee heia a en Irony and the Use Mention Distinction 315 honorable man there is no perceptible irony the inclination is to take it as a propitiatory remark suitable to an occasion where Mark Antony is about to give Caesar s funeral oration with Brutus s permission It is not until he says it for the th
27. istence of ironical questions ironical under statements and ironical references to the inappropriateness or irrele vance of an utterance rather than to the fact that it is false Numerous illustrations have been given earlier in the chapter Furthermore as Grice himself points out 1978 patent falsehood or irrelevance is not a suffi cient condition for irony not every false or irrelevant utterance can be interpreted as ironical What is missing from Grice s account is precisely the fact that ironical utterances are cases of mention and that the proposi tions mentioned are ones that have been or might have been actually en tertained by someone On our analysis recognition of an ironical utterance as a case of men tion is crucial to its interpretation Once the hearer has recognized this and has seen the speaker s attitude to the proposition mentioned the im plicatures in 38 40 follow by standard reasoning processes They are typical cases of conversational implicature and not problematic in any way Our account of irony thus fits more naturally into Grice s overall framework than the account he himself proposes It might be suggested that there are two distinct types of irony echoic irony as illustrated earlier whose interpretation involves a rec ognition of its status as mention and standard irony whose interpreta tions involves a recovery of its figurative meaning The problem with this sugge
28. made in Harnish 1976 PO ee AR a E E IENE re ee ee ER ae Ne OOREEN eT 2 302 Dan Sperber and Deirdre Wilson someone with incredibly slow reactions The speaker is not trying to pre tend that he is such a person nor is he parodying anyone in particular What he is trying to do is to bring to mind just this exaggerated slowness of reaction which would itself be worth remarking on in the circum stances For an utterance to have this effect it must be obvious that the speaker is drawing attention to its content while at the same time disso citating himself from it What is important is the content of the utterance rather than what it is about Pm glad we didn t bother to bring an umbrella Like 1 this is a case where the utterance does not directly reflect the speaker s views where in fact he believes the opposite of what he says One could imagine 3 being used to echo an earlier remark made to or by the speaker or hearer before setting off Don t bother to take an umbrella or Let s not bother to take an umbrella for example By repeating this advice in the pouring rain the speaker of 3 underlines its futility Clearly someone who asks Did you remember to water the flowers cannot mean the opposite of what he says Indeed it is hard to see what would BE the opposite of 4 or of most other ironical questions This is a further argument against both standard semantic and pragmatic accounts of irony The question in 4 is od
29. matic accounts of irony On both standard accounts the ironical utterance of 1 is about the weather and is thus indistinguish able from the literal utterance What awful weather On our account there VAA an ann REE ne 5 n RET RN MRNA ty mE ae alam lace a eee ee Oe AATEC EAEN RIED RAEI Ain ee NOM RSE AN OET hs am v Irony and the Use Mention Distinction 303 isa crucial difference between the two utterances because one expresses an attitude to the content of an utterance whereas the other expresses an attitude to the weather 4 THE USE MENTION DISTINCTION The intuitive distinction we have just illustrated using 1 4 and 5 8 as examples is closely related to the distinction drawn in philosophy volves reference to what the expression refers t0 MENTION of an expres sion involves reference to the expression itself Thus marginal is used in 9 to refer to the doubtful grammatical status of certain examples 9 These examples are rare and marginal It is mentioned in 10a and 10b where reference is made to the word marginal itself 10 a Marginal is a technical term b Who had the nerve to call my examples marginal When the expression mentioned is a complete sentence it does not have the illocutionary force it would Standardly have in a context where it was used Thus the remark in Ila is uttered in 11b without actually being made the question in 12a is uttered in 12b without actuall
30. ntemptuous approving and so on and the immediate context all play a part in indicating his own attitude to the proposition mentioned In particular the speaker may echo a remark in sucha way as to suggest that he finds it untrue inappropriate or irrelevant 35 You take an eager interest in that gentleman s concerns said Darcy in a less tranquil tone and with a heightened colour Who that knows what his misfortunes have been can help feeling an interest in him His misfortunes repeated Darcy contemptuously yes his misfortunes have been great indeed Jane Austen Pride and Prejudice 36 Now just attend to me for a bit Mr Pitch or Witch or Stitch or whatever your name is My name is Pinch observed Tom Have the goodness to call me by it What You mustn t ever be called out of your name mustn t you cried Jonas Pauper prentices are looking up I think Ecod we manage em a little better in the city Charles Dickens Martin Chuzzlewit 308 Dan Sperber and Deirdre Wilson There are also cases where what is echoed is not a proposition expressed by an utterance but a thought imputed by the speaker to the hearer 37 Elinor looked at him with greater astonishment than ever She began to think he must be in liquor and with this im pression she immediately rose saying Mr Willoughby I advise you at present to return to Combe I am not at leisure t
31. o remain with you longer Whatever your business may be with me it will be better recollected and ex plained tomorrow I understand you he replied with an expressive smile and a voice perfectly calm yes L am very drunk A pint of porter with my cold beef at Marlborough was enough to over set me Jane Austen Sense and Sensibility We have presented examples 35 37 as cases of echoic mention we could equally well have presented them as cases of irony The utterances in question are patently ironical The speaker mentions a proposition in such a way as to make clear that he rejects it as ludicrously false inappro priate or irrelevant For the hearer understanding such an utterance involves both realizing that it is a case of mention rather than use and also recognizing the speaker s attitude to the proposition mentioned The whole interpretation depends on this double recognition Recovery of the implicatures 38 for 35 39 for 36 and 40 for 37 will follow automatically 38 He has not been the victim of misfortunes 39 You have no right to demand that I call you by your proper name 40 I am not drunk Not only it is unnecessary to appeal to the notion of figurative meaning in dealing with the interpretation of 35 37 and their implicatures 38 40 any account in terms of figurative meaning will actually be incom plete Suppose we treat 38 40 along traditional lines as
32. of grouping them which do not simply re flect conscious explicitly defined categories This is not because stereo typed responses or conscious categories are uninteresting but because they only provide insight into cultural peculiarities or idiosyncrasies A phere theory of rhetoric should be concerned with basic psychological aa mechanisms which remain invariant from culture to There is another closely related point The traditional study of rheto ric which dates back 2000 years offers a rich and subtle set of analytical 298 Dan Sperber and Deirdre Wilson concepts These concepts are interesting in themselves it is possible that some of them will have a necessary role to play in any future theory of rhetoric However it would be a mistake to prejudge the issue It should not be taken for granted that even the major rhetorical categories such as alliteration ellipsis hyperbole metaphor metonymy irony and so on correspond to genuine natural classes of facts playing clearly defined and distinguishable roles in speech production and perception It is possible that the whole idea of tropes and their classification is destined to go the same way as the notion of humors in medicine it is possible that verbal irony and its associated attitude have about as much claim on our atten tion as black bile and the atrabilious temperament The notion of irony is an abstract one based on a rather arbitrary range of examples which have themsel
33. of what they literally say What lovely weather would have no figurative meaning but would conversationally implicate that the weather was awful Grice s proposal would relieve semantic theory of the problems of defining figurative meaning and deriving the figurative mean ing of an utterance However these problems are not solved simply by transferring them from the semantic to the pragmatic domain It still has to be shown how the interpretation of ironical utterances can be success fully integrated into Grice s pragmatic framework In this chapter we shall argue that it can not and that existing pragmatic accounts of irony are as seriously defective as earlier semantic accounts Grice s departure from the traditional account of irony is not a radical one It is based on the same assumption the assumption that what the speaker of an ironical utterance intends to get across is the opposite of what he has literally said In fact the only disagreement between Grice and more traditional theorists is over whether the substitution mecha nisms involved are semantic or pragmatic Grice s account like the tradi tional one fails to explain why an ironical utterance should ever be pre ferred to its literal counterpart why someone should choose to say What lovely weather rather than the more transparent What awful weather As will be seen it also fails to make explicit exactly how the move from lit eral meaning to conversational implicature is
34. oing to rain 15 It is going to rain The reported speech framework is also too narrow because a number of clear cases of mention do not involve any report of speech even in the very loose usual acceptation Sentences 16 and 17 are examples 16 A yellow flag means stay away 17 Stay away is a grammatical sentence The opaque context framework is too broad because it covers cases of indirect speech such as 14 18 and 19 which clearly fall outside the scope of any notion of mention however extended 18 Oedipus wanted to marry Jocasta 19 Oedipus wanted to marry his mother The contexts in 18 and 19 are opaque since the substitution of the coreferential expressions Jocasta and his mother is not truth preserving However we would not want to say that these expressions are men tioned The notion of opaque context as usually understood is also too narrow because it does not account for cases such as 13b where Be quiet Be quiet is certainly mentioned but where no opaque context is involved see also 27 30 in what follows One could of course say that the null context is opaque under certain conditions but the conditions would have to be defined There is a real need for a comprehensive account of mention in natural language which would cover not only mention of an expression as in 11 3 but also mention of a proposition as in what is usually referred to as free indirect st
35. s are ulti mately based on intuition rhetoric like linguistics is a branch of cogni tive psychology It is well known that linguistic judgments may e fected by explicit teaching or conscious theorizing the same is true of rhetorical judgements only much more so because many rhetorical cate gories such as metaphor figurative or irony are part of everyda speech Because of this informant work in rhetoric must be a once with a certain amount of methodological caution i For example suppose we ask an informant whether 1 or 2 could be ironical when said by someone caught in a downpour 1 What lovely weather 2 It seems to be raining Anyone who has been taught the traditional definition of irony that ironi cal utterances say one thing and mean the opposite will naturally say that 1 but not 2 is ironical He will say this even though he may notice that both 1 and 2 could be said in the same wry tone of voice which a naive informant would precisely call ironical Given enough responses of this type we might well take the traditional definition of irony as being strongly confirmed however this would be a mistake since it is the defi sues aoe that is directly responsible for the judgments which con rm t The best way of avoiding these pitfalls is to ask questions that have no stereotyped response The ultimate goal is to find intuitive relationships among the data intuitive ways
36. s actually given or not whether the obsession was put into words or not does not affect the sta tus of the utterance as echoic mention but only its degree of pointedness What we are claiming is that all standard cases of irony and many that are nonstandard from the traditional point of view involve generally im plicit mention of a proposition These cases of mention are interpreted as echoing a remark or opinion that the speaker wants to characterize as lu dicrously inappropriate or irrelevant This account makes it possible to give a more detailed description of a much wider range of examples of irony than the traditional approach can handle In particular it provides a unified treatment of ironical antiphrasis and meiosis which are tradition ally regarded as two quite different things Moreover it makes no appeal PEPA R ent me Irony and the Use Mention Distinction 311 to the notion of figurative meaning nor to any other notion not fully justi fied on independent grounds 6 SOME FURTHER ASPECTS OF IRONY Our analysis sheds some light on a number of further problems with the treatment of irony We shall mention five of them here Four we shall deal with rather briefly the relation between irony and parody the ironical tone of voice the shifts in style or register that often occur in ironical utterances and the moralistic overtones that they sometimes have The last problem which we shall look at in more detail
37. s not a mention a mention may be more or less ironical with many intermediary and complex shades between Stereotypical cases of irony and other kinds of echoic mention 8 CONCLUSION As an undersized boy trips over his own feet while coming in last in the school sports one spectator turns to another and remarks 49 It s a bird it s a plane it s Superman This remark is clearly ironical Because it is an actual quotation it fits quite straightforwardly into our framework as a case of echoic mention however it poses considerable problems for both standard semantic and pragmatic accounts of irony Within these frameworks it would have to be analyzed as carrying the figurative meaning or conversational implica ture in 50 316 Dan Sperber and Deirdre Wilson 50 It s not a bird it s not a plane it s not Superman But if 49 figuratively means or conversationally implicates 50 which is literally truae it is hard to see why it would also be taken as a joke There is a further problem for Grice s approach since the implicature in 50 is completely uninformative in the context and would itself violate the maxims of conversation Our proposed analysis by virtue of its ability to handle 49 and similar examples proves both more explanatory and more general than either of the traditional alternatives Compared with traditional semantic approaches Grice s approach to irony is radically pragmatic The
38. s of approval will have a corresponding figurative sense with critical Overtones The intended victim on this account would be the object of the criticism For example if 42 has the figurative meaning 43 then Fitzgerald would be the victim of the irony in 42 42 Fitzgerald plays by the rules 43 Fitzgerald cheats Irony and the Use Mention Distinction 313 On the other hand the person to whom the ironical utterance was ap parently addressed may fail to detect its figurative meaning The immedi ate result will be that any third parties present who immediately detect the figurative meaning and are thus revealed as the true addressees of the irony will feel drawn into a conspiracy with the speaker at the expense of the person to whom the remark was overtly made For example if 44 has the figurative meaning 45 and if Billy fails to notice this then on this account Billy becomes the intended victim of the irony ia 44 44 Go on Billy you re nearly there 45 Go on Billy you re nowhere near These two processes may on occasion select the sar ie victim This would happen for example if 46 has the figurative m aning 47 and Jeremy fails to detect it 46 Go on Jeremy your story s really interesting 47 Don t go on Jeremy Your story s really boring Someone who restricts himself to examples of this last ty pe as happens rather too often would get the misleading impression tha the traditional
39. stion is that there is a whole range of intermediate cases between the clear cases of echoic irony and the standard cases see below If there were two totally distinct processes one based on mention and the other on figurative meaning each resulting in a different type of irony such in termediate cases should not exist It seems more accurate to say that all examples of irony are interpreted as echoic mentions but that there are echoic mentions of many different degrees and types Some are immediate echoes and others delayed some 6 See for example Cutler 1974 who distinguishes between spontaneous standard and provoked echoic irony 310 Dan Sperber and Deirdre Wilson have their source in actual utterances others in thoughts or opinions some have a real source others an imagined one some are traceable back to a particular individual whereas others have a vaguer origin When the echoic character of the utterance is not immediately obvious it is never theless suggested Within this framework we return to our original exam ples of irony Sentences 1 4 What lovely weather Suppose that as we were deciding to set off on our walk someone told us that the weather was going to be lovely It is quite clear that 1 is an ironical echo of this remark Or suppose we have spent a rainy winter talking about the walks we will have in the summer sun The echoic quality of 1 though its source is more
40. t utterances also con tain referential expressions which may have a wide range of possible re ferents even when the shared knowledge of speaker and hearer is taken into account It is thus quite typical for an utterance to have dozens or even hundreds of possible propositional interpretations However speaker and hearer are normally able to select a single one of these inter pretations without even realizing that they have made a choice It is gen erally agreed that this choice is a function of the context but to define the function as opposed to simply claiming that it exists is no easy task As long as we only have to choose among the literal senses of an utter sot anne ee eat nr tt NA nA A Irony and the Use Mention Distinction 299 ance the task is still an approachable one the set of possible interpreta tions remains finite and will be specifiable on the basis of a fairly re stricted range of semantic and referential variables One can think of several types of explicit procedures that could be used to eliminate all but one of the possible interpretations The difficulty lies not so much in con ceiving of such a procedure in principle as in choosing and justifying the right one On the other hand suppose that we have to take into account not only the literal senses of an utterance but also the whole range of figu rative senses that are loosely based on them via relations of resemblance contiguity inclusion or inversion
41. theory can provide a unified account of how an ironical utterance chooses its victim However anyone who looks at the differences between exam ples like 42 and 44 will immediately see that two quit different pro cesses are involved and that they are not necessarily rel ited at all Moreover it is easy to think of quite ordinary examp vs that are not accounted for in terms of either process Suppose the folk wing remark is made ironically to someone who dislikes classical music 48 Of course all classical music sounds the same On the one hand the figurative meaning of this remark has no critical con tent on the other hand in normal circumstances the hearer is unlikely to mistake the speaker s intentions In this case neither process will apply and there is no immediate explanation within the traditional framework for the clear intuition that the hearer of 48 is also its intended victim Within our framework it would be possible to define two processes that would correspond closely to those used in the traditional account Instead of figurative meanings there would be pragmatic implications or implica tures which might carry critical overtones instead of a failure to distin guish literal from figurative meanings there would be a failure to distin guish use from mention The framework we are proposing is thus at leas as explanatory in this respect as the traditional framework 314 Dan Sperber and Deirdre Wilson Howev
42. ves been rather inadequately described Because of this it seems to us to be a mistake to take IRONY itself as the object of investigation and to limit one s attention to its more standard cases There is a whole range of utterance types that can be more or less loosely called ironical The basic facts to be accounted for are the particu lar effects produced by particular utterances and the perceived similar ities among them We should be looking for psychological mechanisms that can account for these effects and their interrelationships When we have found some it might be interesting to make some comparisons be tween the resulting provisional conceptual scheme and the framework of classical rhetoric to see which notion of irony emerges if any but the existence of a unified category of irony should not be taken for granted Quite independently of the existence of irony there are already strong grounds for rejecting the notion of figurative meaning itself For example consider the treatment of disambiguation which is a major problem for any pragmatic theory Every hearer or reader almost instantaneously disambiguates each of the utterances he hears Even if we ignore figura tive meanings and consider only the literal senses of an utterance nar rowly defined almost every utterance is ambiguous In fact almost every utterance is multiply ambiguous with possible semantic interactions among its individual ambiguous constructions Mos
43. y being asked and the order in 13a is uttered in 13b without actually being given GI a What a shame b Don t just say What a shame do something 12 a What is irony b What is irony is the Wrong question 13 a Be quiet c Be quiet Be quiet And suppose I feel like talking This may be used as a test for distinguishing between use and mention of sentences The use mention distinction is a logical one In formal langua zes men tion is distinguished from use in a conventional way and there an be no question about whether a formula contains a mention nor abo t what is being mentioned In natural languages mentions take a wiety of forms some of which might seem to be intermediate cases falling some where between use and mention Moreover cases of mention in natural 304 Dan Sperber and Deirdre Wilson languages are not usually studied in their own right but only for the role they play within the frameworks of either reported speech or opaque contexts Both frameworks are inappropriate for the study of mention they are too broad in some respects and too narrow in others The reported speech framework is too broad because there are standard cases of indirect speech such as 14 where the indirectly reported propo sition 15 is part of a larger proposition and there is no demarcation line explicit or implicit to set it off as a case of mention 14 They say that it is g
44. yle gt In free indirect style an independent proposi tion is reported with an optional comment in a parenthetical phrase as in 20 5 For recent work on free indirect style see Banfield 1973 and McHale 1978 Tal LA MINCE oes 1 N amta aS SUA AN A OE RONEN AERC MORRO a o Irony and the Use Mention Distinction 3058 20 He will come at five he SANS It seems clear that 20 is equivalent in logical structure to 21 rather than 22 and is therefore a genuine case of mention 21 I will come at five he says 22 He says that he will come at five There are thus two properties that may be used to disting 1ish various types of mention in natural language looking at these one can see why it is sometimes felt that there are intermediate cases between the poles of pure use and pure mention On the one hand we can contrast explicit mention as in 23 and 25 with implicit mention as in 24 and 26 On the other hand we can contrast the two different types of obj ct that may be mentioned linguistic expressions as in 23 and 24 and p opositions as in 25 and 26 23 The master began to understand and to share the intense disgust which the archdeacon always expressed when Mr Proudie s name was mentioned What am I to do with such a woman as this he asked himself 24 The master began to understand and to share the intense disgust which the archdeacon always expressed when Mrs Prou
Download Pdf Manuals
Related Search