Speech Acts, Literary Tradition, and Intertextual Pragmatics more

28 The Intertextual Dimension of Discourse Works cited BAKHTIN, M. 1978. Esthetique et theorie du roman. Saint Amand: Gallimard. BERNARDEZ, Enrique. 1996. In «Pragmatics as Self-Regulation of Behaviour*, crossreference. BROWN, P. and S.C. LEVINSON. 1978. Politeness. Some Universals in Language Usage. Wiltshire: Cambridge University Press. GENETTE, G. 1972. Figures III. Seuil: Paris. . ENNION, V 1991. «Ishiguro». It's for Teachers. 5: pp. 48-49. FLO REN, C. 1993. «Recurrence in Ishiguro's Novels*. Actas del XV Congreso de AEDEAN, Asociacion espanola de estudios angloamericanos. Logrono 1991. 1993: 587-594. KRESS, G. 1985. Linguistic Processes in SocioculturalPractice. Hong Kong: Oxford Univer- sity Press. ISHIGURO, K. 1986. An Artist of the Floating World Suffolk: Faber and Faber. — 1988. The Remains of the Day. Suffolk: Faber and Faber. — 1995. The Unconsoled. London: Faber and Faber. HUTCHEON, L. 1985. A Theory of Parody. The Teachings of Twentieth-century Art Forms. Cambridge: Methuen HUTCHEON, L. 1989. The Politics of Postmodernism. Bury St Edmunds: Routledge. LAUSBERG, H. 1967. Manual de retorica literark. Madrid: Gredos. LEECH, G. 1983. Principles of Pragmatics Singapore: Longman. MANSON, G.M. 1989. -,«An Interview with Kazuo Ishiguro». Contemporary Literature XXX 3, University of Wisconsin. MEY, J.L. 1993. Pragmatics. An Introduction Oxford: Blackwell. PLETT, H. 1991. «Intertextualitites». In H.F. Plett (ed.). Intertextuality. Berlin: Walter de Guyter: 3-29. PRATT, M.L. 1977. Toward a Speech Act Theory of Literary Discourse. Indiana: Indiana University Press. RAJAN, T. 1991. «Intertextuality and the Subject of Reading/ Writing*. In Jay Claytib and E. Rothstein (ed.). Influence and Intertextuality in Literary*History. Madison: The University of Winsconcin Press. SEARLE, J.R. 1982. «Metaphor». In S. Davies (ed.) (1991) Pragmatics: A Reader. New York: Oxford University Press: 519-539. STALNAKER, R.C. 1972. «Pragmatic Presuppositions*. In S. Davies (ed.) (1991) Pragmatics: A Reader. New York: Oxford University Press: 471-482. SWIFT, G. «Kazuo Ishiguro*. Bomb Fall 1989: 22-23. TANNEN, D. 1989. Talking Voices, Repetition, Dialogue and Imagery in Conversational Discourse. Avon: Cambridge University Press. TRAUGOTT, E. «Generative Semantics and the Concept of Literary Discourse* Journal of Literary Semantics 2 (1973). Speech Acts, Literary Tradition, and Intertextual Pragmatics Jose Angel Garcia Landa Zaragoza University There is, at least potentially, a kinship between some central concepts of clas- sical literary criticism (such as tradition, genre, originality, allusion) and more recent ones developed by formalist/structuralist criticism (intertextuality, defamiliarization) and linguistic pragmatics (illocution, indirect speech act, pragmatic principles and maxims). Reflecting on the common ground shared by these notions may shed some light on the relationship between linguistics and literary theory. I will use as an example a short poem by Stephen Crane: I WALKED IN A DESERT. AND I CRIED: «AH, GOD, TAKE ME FROM THIS PLACE!* A VOICE SAID: «IT IS NO DESERT* I CRIED: «WELL, BUT— THE SAND, THE HEAT, THE VACANT HORIZON.* A VOICE SAID: «IT IS NO DESERT*' (The Black Riders, no. 42J Crane himself might be said to be telling us a story here, but (as noted by the mysterious voice) we should not seriously believe that the poet is walking in a desert. If we did so, we would be poor readers and worse linguists. I am grateful for the financial aid provided by the DGICYT (Programa Sectorial de Promocion General del Conocimiento, proyecto PS94-0057), which has allowed me to carry out this and other related proyects. 30 The Intertextual Dimension of Discourse It is fairly common for linguists and poets not to take each other seriously. Austin and Searle, the founders of speech act theory (SAT), hold that in order to analyse the meaning of a speech act, one must first make sure that the utterance was a serious one: for Austin, it does not count if we are speaking in jest or writing a poem (Austin, 1980: 9). And Searle differentiates the normal use of language in speaking of the real world from what he calls «parasitic» forms of discourse, such as fiction, theatre acting, etc. One concludes, maybe, that literature is a kind of linguistic jest, or at least a not very serious use of language. Sartre's dictum that man is serious when he takes himself for an object makes us wonder whether in taking language too seriously we might be taking it for an object as well: These notions of Austin and Searle's have of course been criticized by (other) literary theorists.1 It seems to me nevertheless that in spite of this inauspicious start, speech act theory may help to clarify the links between literature and linguistics. Defining the notions of literal speech acts, indirect speech acts and «non-serious» speech acts is an essential step in understanding the structure and the interpretation of speech activity, whether in literature or in other discourses. But first let us see in closer detail some basic concepts of pragmatics and their possible use in literary criticism and theory. An effective linguistic theory for the analysis of literary texts should fulfil at least two requirements which are not consistently satisfied by traditional grammars but are satisfied by linguistic pragmatics: -— The theory should contemplate the study of linguistic units larger than the sentence. — Its object of study should be the effective use of language, going beyond a description of language as an abstract system, and thus closing the gap between Saussure's langue and parole (language system and language use). For- instance, interpreting a metaphor requires that a distinction be drawn between meaning and use, which involves going beyond the scope of a simple systematic lexicon. In Ricceur's words, there are no metaphors in the dictionary.2 These new emphases, the textual and the contextual, converge spontaneously, since text formation is part and parcel of the actual use of language, or linguistic action.3 Thus, as is the case with any other use of speech, a literary work is not a sum of contextless sentences, but contextualized discourse, an effective instance of the use of language in a concrete situation.4 A study of sentence structures is clearly an / 1 See f. i. Derri'da (1988: 88ff); Petrey (1990: 22ff). 2 Ricoeur (1975), quoted in Schofer and Rice (1977: 135). 3 See Halliday (1970: 160-61). 4 SeeTodorov (1987: 32). For Benveniste, discourse is «toute enonciation supposant un locuteur et un auditeur, et chez le premier 1'intention d'influencer l'autre en quelque mafiiere» (1968: 242). On the Saussurean notion of discours as the syntagmatic aspect of language,' see S'egre (1985: 188ff), Hendricks (1973: 77ff). Speech Acts, Literary Tradition, and Intertextual Pragmatics 31 insufficient basis for a linguistic theory of literature: what is required is a textual and discursive linguistics. A text might be described as an atemporal structure of coexisting semiotic relationships. But a more fruitful approach to textual reality will give an account of the text in its temporal dimension, as a process, not an object. This is one meaning of the term «discourse»: a text considered as a temporal sequence of signs.5 - It is impossible to describe the form of a text without describing the function of that form, the use that is made of the structures. The early structuralist approaches to linguistics often neglected this aspect of language. The well-known divisions established by Saussure between language system {langue) and language use {parole),6 and by Chomsky between competence and performance, define the central task of linguistics as the study of the abstract language system, not the concrete use of language. Saussure, for instance, does not take into account the existence of regular forms in syntagms larger than the sentence. And, as we move towards hierarchically higher syntagms, it becomes harder and harder to determine the difference between langue and parole. Hence the contemporary development of a linguistics of parole which would sound paradoxical to a strict Saussurean, just as the idea of a theory of performance would be paradoxical for classic transformational grammar. Pragmalinguistics, then, tries to relate the study of the linguistic system to the study of linguistic process.7 The sentence is conceptualized nowadays as a structure subordinated to the text, which in its turn is a part of a contextualized semiotic process.8 Semiotics includes three types of studies. The first two have been the usual object of study of classical linguists: they are syntax, the shape of utterances, the relationships between signs, and semantics, which defines the meaning of linguistic forms, the relationships between signs and concepts. The third aspect of semiotics is pragmatics. The concern of pragmatics is the use of signs and sign systems in communication, the relationship between signs, referents and users. There have been in the recent past many structuralist theories which tried to describe the 5 See Lozano, Pena-Marin and Abril (1982: 33). 6 Saussure (1984: 27ff). 7. See Van Overbeke (1980: 396ff), Lozano, Pena-Marin and Abril (1982: 34ff). A historical analogue of the discourse analysts' reaction against structuralist linguistics (including generative and transformational linguistics) might be found in the eighteenth-century, in Condillac's reaction against the Cartesian grammatical tradition of Port-Royal (see Uitti, 1977: 74ff). 8 Witness Ingarden's early formulation: «The sentence-forming or duplicating operation (...) is in most instances only a relatively dependent phase of a much broader subjective operation, from which arise not only individual, out-of-context sentences but, instead, entire complexes of sentences or manifolds of connected sentences. When, for example, we conduct a proof or develop a scientific theory or simply narrate an account, we are attuned, usually from the very beginning, to the whole which we are to 'develop1 even before we have formed the individual sentences by which it will be 'developed'* (1973a: 103). See also Voloshinov (1986: 95). 32 The Intertextual Dimension of Discourse syntactic or the semantic deep structure of sentences and texts: from our present viewpoint, the deep structure of a text should be assigned a pragmatic description. That is, a comprehensive linguistic approach should contemplate discourse contextually, as it is used, rather than offering an abstractive syntactic or semantic description. The second meaning of the term «discourse», the one we wish to draw attention to, is the use of texts in a specific communicative situation. A linguistics of discourse is concerned hot only with the various structures of linguistic signs, but also with the modalities of linguistic enunciation and interpretation.9 Let us note in passing that the specificity of literary interpretation should be taken into account: for instance, the fact that the classics are usually asked to be as full of meaning as is possible without breach of coherence. If Crane's poem is to become a classic, it shall have to be analyzed and interpreted until it acquires a sufficient meaning. In the pragmatic approach sketched here the study of linguistic forms must be integrated with the study of discursive interaction: the analysis of signification needs to posit a systematic basis of intersubjective, standardized forms which nonetheless can be modified in the concrete communicative act.10 These forms include syntactic and semantic structures, modalities of enunciation or illocutions, models of linguistic action, linguistic and literary genres. The word used in a given context acquires a specific sense which must be deduced by the hearer through detective- like strategies, to use Biihleri expression.11 Let us consider, for instance, the meaning of the word «desert» in Crane's poem: its central meaning could be looked up in a dictionary, but its contextual sense must be interpreted and constructed, perhaps even negotiated. Likewise, we need to interpret who does the talking in the poem —is it Crane? Is it God that answers? Does the speaker answer himself? Any aspect of the effective use of language in a given situation, any element necessary forvthe study of language as text or discourse, may become the object of pragmatic study. These aspects and elements include the activity of the speaker and listener, the act of enunciation, kinds of utterance, context models, and the supralinguistic principles governing discursive interaction. The act of enunciation is reflected or inscribed in the text produced by the speaker, but only in a partial way. There is always a trace of the enunciation, as soon as we consider the text as an index 9 See Lozano, Pena-Marin ya Abril (1982: 35). 10 Biihler (1967: 115-23); Bach and Harnish (1979: 105); Lotman (1982: 71); Ducrot (1980: 5 5 Off). Cf. "Wittgenstein's notion of «language game» and his view of language use as an activity which is both regulated and rule-generating (1989: §§ 23, 117ff, 198ff). Compare Sbisa and Fabbri's description of how communicative intention is subject to negotiation (quoted in Lozano, Pena-Marin and Abril, 1982: 194)—a crucial insight which, however, may lead these authors to undervalue the hermeneutic step consisting in the hearers recognizing the speaker's intention. 11 Biihler (1967: 115-23). See also the distinction between the semantic and pragmatic perspectives on signification in 530ff. The significance of Biihler's work in the development of a comprehensive linguistic theory cannot be overestimated (cf. Van Overbeke, *J980: 416).' Speech Acts, Literary Tradition, and Intertextual Pragmatics 33 of the speech act .n However, in order to determine the meaning of a speech act, we need to be acquainted both with the text and the concrete circumstances of its enunciation, including the particular conventions which may apply to a given speech genre or a given period. Thus, textual linguistics ultimately converges with the general principles of hermeneutics.13 The theories of speech acts developed by philosophers and linguists (Biihler, Wittgenstein, Austin, Searle, etc.) try to systematize the principles of enunciation, and is an essential contribution to a general" linguistic pragmatics (including hermeneutics), a linguistic study of discourse.14 "Within the pragmatic level of analysis, language can be studied at different levels of abstraction. In Austin's terms, when we speak we are performing several simultaneous acts: locutionary acts (phonetic, phatic, rhetic acts), illocutionary and perlocutionary acts.15 The locutionary act consists in the speaker's addressing a hearer by uttering a linguistic form in a given context. For instance, the mysterious voice in Crane's poem utters the words «It is no desert.» In doing so, the voice conveys a set of semantically codified meanings, a proposition to the effect that the place is not a desert. By performing the locutionary act as part of a communicative exchange, the speaker performs an illocutionary act, a specific and socially recognizable communicative act, an intentional action. At this level of analysis, the enunciation has a pragmatic meaning or illocutionary force.16 For a speech act to be successfully 12 See Ducrot (1980: 534). 13 For an exposition of these principles, see Hirsch's works. It should be stressed in this connection that literary hermeneutics cannot limit itself to the study of a work's production, its original context, its historical or authorial meaning («meaning» in Hirsch). The study of «significance», the interpretive expansion of meaning, or present-day meaning, is a crucial step in the hermeneutic process. 14 The SAT approach is not, of course, radically new. Protagoras already distinguished four types of what he called «foundations of discourses* (pythmenas logon): question, request, answer and order. Other classifications may be found in Alcidamas and Anaximenes (Lopez Eire, 1980: 17ff). Aristotle himself offers a similar list of speech acts (order, request, explanation, threat, question, answer, etc.), adding that «the theory of such matters, however, belongs to Elocution and the professors of that art» and not to the theorists of Poetics {Poetics 1456b). Similar rudiments of pragmatics can be found through the centuries in many Logics, Pvhetorics and Grammars; they become gradually systematized in the 20th century (see Van Dijk, 1972: 24; Van Overbeke, 1980: 412; Chico Rico, 1988: 109ff). Before «speech acts» there came the «functions of languages or «uses of languge» (Malinowski, Brugmann, Sonnenschein, Steinthal, Bally, Richards, Biihler, Jakobson, Martinez Bonati, .Halliday, Castilla del Pino, etc.). These approaches should not be rejected in favour of SAT, since their aims and insights only partially overlap with the analytical approach of SAT. 15 Austin (1980: 95ff). Compare Searle (1980: 32); Van Dijk (1972: 318ff), 1980: 278ff); Schmidt (1977: 59ff); Pratt (1977: 80); Lyons (1977: 730); Lozano, Pena-Marin and Abril (1982: 188). An interesting foreshadowing of Austin's distinctions can be found in Ingarden (1973a: 107ff). 16 The conceptualization of language as a form of action is evidently prior to Austin's work. See, for instance, Ingarden (1973b: 382); Benveniste (1968: 265). 34 The Intertextual Dimension of Discourse performed, it must satisfy certain felicity conditions which vary from one act to another and serve to define them.17 In our poem, the voice, by saying that the place is not a desert, performs an illocutionary act, an assertion: it acquires a compromise with the truth of the propositional content. Moreover, by means of that assertion (a direct speech act) it answers indirectly the speaker's request, suggesting that it lacks an appropriate ground. Through the performance of illocutionary acts, language users perform yet another kind of speech acts: perlocutionary acts. Speakers exert influence on hearers, cause reactions on them (perlocutions or perlocutionary effects). In our example, the voice baffles the narrator/traveller. Maybe this was not its perlocutionary intent; we do not know. The speakers perlocutionary intention of causing such and such effect need not be manifest to the listener. Besides, the illocution may be successfully performed even if the speaker fails to achieve the perlocutionary purpose. Both traditional linguistics and structural linguistics derived from the work of Saussure or Bloomfield were concerned solely with the study of locutionary acts —in fact, semantics itself was often neglected as a non-systematizable aspect of language. That is, linguistics identified the study of language with the study of locutionary acts, relegating illocutionary acts to the status of language use, parole. According to Searle, an adequate study of speech acts would be a study of langue, not ofparole (1980: 27). This formulation is too rreat, and does not allow for the contextual flexibility and the constant evolution of the illocutionary level of language —in fact, the definition is couched in the very terms {langue and parole) which are radically questioned if we take the speech act approach seriously. Strawson observes that not all illocutionary acts are conventional in the same sense; there is a gamut between the poles of illocutionary conventionalization and merely locutionary communication.18 Searle's formulation, like the pair langue I parole, makes sense only within the framework of a sentence grammar; it is overriden by textual grammar. The study of the sentence as a locutionary act opens up only one level of signification: only the literal sense, the sense which is standardized and systematically codified, can be interpreted within this framework. The concern of semantics is the primary meaning of signs, not their signification in concrete situations.19 The product of a locutionary act is a proposition; that of the illocutionary act has to be a communicative move, an interpersonal action on the part of the speaker, a speech act proper.20 Illocutionary acts are socially 17 Searle (1980: 47, 65ff); Pratt (1977: 81); Lyons (1977: 733); Van Overbeke (1980: 458ff); Ducrot (1980: 519) A 18 Strawson (1964: 456-57). Compare Hirsch (1976: 67-71). 19 Austin and Searle, however, consider reference or the «rhetic act» as a component of locutionary meaning. 20 In Cicourel's SAT, communication is no longer a simple transaction„of meanings; it becomes an exchange of speech acts («Three Models of Discourse Analysis: The rol-e^ of social structures, Discourse Processes 3; quoted in Lozano, Pefia-Marin and Abril, 1982:*4l). Speech Acts, Literary Tradition, and Intertextual Pragmatics 35 codified, although not necessarily linguistically codified. Communication involves performing illocutionary acts, not just locutionary acts. The voice in our poem, for instance, does not formulate an abstract and universally valid proposition about the landscape: it informs, corrects and perhaps tells the narrator off. An illocutionary act can be said to be performed when the speaker makes the hearer recognize the speakers intention to make him/her recognize the illocutionary act being performed —that is, when the illocutionary force is correctly identified.21 This identification is an inferential process: it takes into consideration the locutionary act performed and also a tacit element in communication: mutual contextual beliefs (to use Bach and Harnish's term). Effective communication requires that both speaker and hearer believe that their interlocutor believes that both share these beliefs (a minimally coincident world-view, a similar language, common assumptions concerning the nature of the discourse event which is taking place, etc.). Mutual contextual beliefs consist of factual knowledge assumed to be shared by the speakers; to this knowledge we should add the norms of linguistic action which are also supposed to be shared.22 Of course, both the knowledge and the norms of discursive interaction which are assumed to constitute a common basis may eventually be questioned by one of the speakers, leading to a revision of the shared assumptions. The way common assumptions are negotiated is perhaps one of the key moves in Crane's poem: the narrator-walker defines himself as a victim; the voice does not accept that role for the narrator, arguing that it is based on incorrect presuppositions; the reader is left without a secure foothold. We have argued that the fulfilment of the illocutionary act consists in its being acknowledged as such, in its correct identification on the part of the hearer. This identification is not mechanically dependent on the (semantic) meaning of the locutionary act (Bach and Harnish, 1979: 10). Thence the possibility of both direct and indirect illocutionary acts. The voice in our poem says, as a direct assertion, that the place is not a desert; indirectly, it gives an answer to the walker's petition. A speaker may use knowledge shared with the hearer, and the hearer's inferential abilities, in order to perform a speech act by using (instrumentally and superficially) the performance of another speech act. Bach and Harnish define an indirect speech act as «an illocutionary act that is performed subordinately to another (usually literal) illocutionary act. It is indirect in the sense that success is tied to the success of the first act» (1979: 70). In general, there is a presumption of literalness of the 21 «So the performance of an illocutionary act involves the securing of uptakes (Austin 1980: 117). See also Strawson 1964, Searle (1980: 52); Lyons (1977: 733); Lozano, Pefia-Marin and Abril (1982: 194ff); W. Harris (1988: 84). Note that it is the illocutionary force that must be recognized, not a non-conventional perlocutionary intent, as held by Grice (see the refutation of Grice's position in Searle (1980: 5Iff). 22 For instance, Grice's Cooperative Principle and his maxims of conversational interaction (1989), or Bach and Harnish's Linguistic Presumption and Communicative Presumption (1979: 7passim). 36 The Intertextual Dimension of Discourse illocutionary act as long as the linguistic and contextual circumstances allow it.23 There is a relationship, albeit a flexible one, between locutionary and illocutionary acts: we cannot use in any circumstance any phrase in order to convey our meaning. Continual indirection may result in a standardization of the indirect illocutionary force. This is what happens in sentences such as «Can you pass me the salt?» which are directly interpreted as a request and not as a question.24 For our discussion of literature, however, we shall find it more interesting to focus on non- standardized indirect acts —those which require the receivers' inferential procedures and their involvement in the linguistic game of interpretation. If linguistics has always faced difficulties in dealing with literary speech, speech act theory is not an exception. Early versions of SAT, at least, afford only a limited perspective on discourse as a whole, since they rely on a sentential conception of grammar.25 The sentence is a useful abstraction as far as syntactic or semantic analysis is concerned, but the idea of pragmatics itself requires a more comprehensive level of analysis, the text,26 and the contextualized text, discourse.27 The relevant unit of analysis for the study of linguistic communication is not, there- fore, the text conceived as an abstract system of suprasentential relationships, but rather the production and understanding of the text in a given situation—discursive action.28 We shall need to define the precise application" of these notions to literature, so as not to re-vert to the naive intentionalist or expressivist positions which had been displaced by formal and structural approaches. Note, in passing, 23 Bach and Hamish (1979: 10). Compare Searle's contention: «E1 acto o actos de habla realizados al emitir una oracion son, en general, una funcion del significado de la oracion* (1980: 27). Bach and Hamish also show that even in those cases where the illocutionary force is explicitly stated as a part of the locutioanry meaning, the illocutionary level as such does not disappear: the speaker will have to recognize that the stated illocutionary force is correctly ascribed and must be interpreted literally. Besides, the semantics of the phrase would tell us only which type of illocutionary act is being performed: we would not be able to determine on a semantic basis alone that the act has been actually performed (1979: 204ff). 24 Searle (1979: 49ff); Bach and Harnish (1979: 192ff); Lozano, Pena-Marin and Abril (1982: 220ff). 25 According to some commentators, Austin's theory lends itself to a wider interpretation (Lozano, Pena-Marin and Abril, 1982: 173); but Searle is quite explicit about the sentential basis of his theory: «La unidad de la comunicacion lingtustica no es, como se ha supuesto generalmente, el si'mbolo, palabra, oracion, ni tan siquiera la instancia del si'mbolo, palabra u oracion, sino mas bien la produccion o emision del si'mbolo, palabra u oracion al realizar el acto de habla» (Searle, 1980: 26). 26 See Z. Harrik(1970); Dolezel (1971:. 95); Halliday (1970: 334); Petofi and Garcia Berrio (1978:245). 27 See Uitti (1971: 112), Ohmann (1971a: 245), D. Sperber («Rudiments de rhetorique cognitive)*, cit. por Lozano, Pena-Marin y Abril, 1982:. 37), Van Dijk (1972: 3, 1980: 32), Schmidt (1977: 51 ss), Lanser (1981: 71), Segre (1985: 377). Hjelmslev already treated certain^ semiotic phenomena, such as connotation, at discourse level (Martinet 1975: 177). , \ 28 Van Dijk (1972: 32 Iff). Speech Acts, Literary Tradition, and Intertextual Pragmatics 37 that SAT is also «structuralist» in the sense that it finds a system of regulated behaviour in an area (use of speech) previously relegated to individual, non-analyzed performance. Many applications of SAT to criticism and theory meet conceptual difficulties in situating such concepts as «literature», «narrative» or «fiction» in the framework of SAT. Since many speech act classifications rely on sentential, microstructural speech acts, most of them do not attempt to account for the infinite variety of discursive models.29 It is also common for the theory to be applied only to certain aspects of the discursive phenomenon (usually the production of discourse), while in fact its consequences should be taken into account also in the study of textual structure and of reception / interpretation. As I have pointed out, Austin and Searle have become.notorious among literary theorists because their models tend to brush aside fiction or poetry as a «non- serious» use of language.30 Their position is nonetheless understandable, because they were trying to theorise language at a given level of abstraction: the level of simple and primitive illocutionary acts. Literary works are complex and derived acts of speech. As we have been arguing, discursive action is heavily dependent on the relationship between the sender and the receiver and on the attitude of both towards the message. A detailed study of the. pragmatics of discourse requires a step beyond 29 We do not find a systematic approach to these varieties of speech in Ballmer and Brennenstuhl (1981). Their discursive types are not organized systematically. Next to «make rhymes», «write poetry», «produce (science) fiction* (!) are classified speech acts such as «draft a speech*, «keep a diary», «tell untruths», «prophesy». Clearly, a classification of speech acts needs relevant criteria to show the similarities and differences between these discourse acts as they are understood in ordinary human activity (see Lanser, 1981: 280, 289). Ballmer and Brennenstuhl classify narrative under the heading Utter. «narrate (a story)» stands next to «manifest», «mention», «say», «publish», «remark», etc. Neither does this classification account systematically for the difference between written and oral communication, which is no doubt a relevant trait for a classification of speech acts, besides being intuitively immediate. Of course, other schools have been dealing with these problems from their own perspective. Brugmann's (highly incomplete) classification of verbal activities, which includes eight categories, includes as a specific type the «statement about imagined reality* {Verschiedenheiten der Satzgestaltungnach Massgabe der seelischen Grundjunktionen, quoted by Jespersen, 1968: 301). Ingarden's analysis is already a detailed one. Other classifications of speech acts appear in Wittgenstein (1989: § 23), Austin, Searle, Habermas {Theorie der Gesellschaft oder Sozialtechnologie, quoted in Schmidt, 1977: 124ff), Schmidt (197:128ff), or Bach and Harnish. Wittgenstein is an exception: although he only deals with this issue in passing, and hardly in a systematic way, he includes literary fiction as one of his «language games». Austin (1980: 151 ff) dismisses such phenomena such as narrative, fiction, etc. In his classification of illocutions (1977), Searle ignores the poetical use of language (compare the contrast with Jakobson's account of the functions of language). Bach and Harnish seem to believe that the rules for communicative intercourse are suspended in narrative (see 1979: 90ff). Schmidt takes many more distinctions into account; in fact, he does not present a classification of speech acts but of «communicative activities*—for instance, apart from classifying illocutionary forces he distinguishes between «discourse type» (conversational, scientific, literary...) and «text type* (narrative, expositive, performative...). The articulation between these concepts, however, is not clear. 30 E.g. Searle, 1980: 65. 38 The Intertextual Dimension of Discourse linguistics: it is necessary to focus on a discipline or discourse activity which makes effective use of the texts in question. At an abstract linguistic level we can only define a set of nuclear and ideal speech act types—not actual speech acts at all, but rather instruments for the description of linguistic action.31 A study of actual communication requires going into the communicative and cultural context; the more complex and macroscopic the speech act under analysis, the more necessary is contextualization.32 If it is an analysis of the specificity of discourses that we are after, there is no sense in reducing this specificity to an abstract taxonomy of all possible types of speech act. "We do need the concept of a sequence of separate, well- codified, microscopic speech acts: the speaker uses the sentence as the basic fulcrum for the performance of these. But these sentential speech acts are instrumentalized at a textual/discursive level. At this level, we do not find just the simple speech acts analyzed and classified by Austin and Searle, but discourse acts or macro-speech acts.33 Discourse-level speech acts usually comprise a variety of micro-speech acts of different kinds guided by the discursive macrostructure which unifies and characterizes the global act. Types of global speech acts can be contemplated as specifications of or derivations from the types of primitive or microstructural acts. We could establish differences between levels of complexity, and thus differentiate basic or primitive discourse acts, such as «narrating», and derived discourse acts, such as «writing a poem, a novel».34 This distinction can be useful in defining the pragmatic nature of linguistic and literary genres, although it would be necessary to consider in each case the types of intentionality and shared conventions required for the act. The basic generic types of literature, such as narrative, drama and lyric, seem to be well established intermediate-level illocutions: although they are both complex and derived, they are recognizable almost universally and they have a well defined intentional structure. The pragmatic study of literature must account for literary- phenomena as a specific and socially codified use of language^. Literary pragmatics, in this sense, attempts the wholesale study of the texts literary aspects as such. The modalities of verbal fiction and narrative can be analysed as specific illocutionary 31 To conceive of communication as a sequence of these micro-speech acts, as Searle's analysis might suggest, will provide an excessively taxonomic and regimental picture of communication, as Leech rightly points out (1983: 23). 32 Compare Lyons (1977: 737). Note that what is not possible is to model macro-illocutions directly on Searle's illocutionary types other than figuratively (W. Harris, 1988: 47); this does not preclude, however, a convergence between an extended SAT model and genre theory. 33 Van Dijk (ls980: 325ff). Compare Pratt (1977: 85, 142); Chico Rico (1988: 215 passim), or, in a somewhat muddled way, Grice (1989: 362). According to Ohmann, in a novel, «behind the acts of stating is the all-encompassing illocutionary act of telling a story» (1971a: 247). Ohmann rightly points out that classic stylistics ignored the illocutionary level of discourse. 34 The structurally derived nature of literary discourse is suggested by the Russian Formalists and by Todorov (e.g., on narrative, Todorov, 1987: 38, 45). See also Genette (199.1: 4.8); Garcia Landa (1992:96). Speech Acts, Literary Tradition, and Intertextual Pragmatics 39 types in which the poetic function of language is dominant. This function was defined by Jakobson as an autotelic use of language: the main justification for the poetic message lies in the poetic experience itself, not in the social relationship between the speaker and the hearer or in direct reference to reality. For instance, Crane performs the speech act of writing a poem, an act which is «gratuitous» or offers itself as poetic in its primary structure, although it may pursue other aims (interactive, ideological, economic) through the non-dominant functions it also performs. The pragmatic perspective also helps us understand the structure of the text, a hierarchy of many types of pragmatic phenomena.35 A speech act inside a novel is, in its context, as real as a speech act in actual conversation, since the same interpretive principles are required for both. Therefore, the speech acts in literary works are not «etiolated» or «impaired and incomplete» as Austin or Ohmann (1971b) assert.36 In analyzing our poem we have had to take into account both the speech acts performed by the author (adopting a narrative persona) and the narrated speech acts (the conversation between the voice and the traveller). Finally, we must take into consideration the discourse act that we perform as readers or interpreters of the poem. The emphasis laid by most speech act theories on the illocutionary initiative of the speaker may blur the fact that the receiver, especially in the case of written literature, also performs an active role in determining the modalities of the effective use of speech and in deciding which speech act is being performed. This role cannot be encompassed by the mere notion of «uptake» or illocutionary completion as described by Austin. Reading is a communicative activity the initiative of which lies in the reader; and the traditional account of speech acts must be supplemented in the case of literature especially by a theory of the acts of reading and interpretation.37 "What of literature as a whole? Literature is one among many discursive situations defined contextually and culturally. Pratt (1977: 86) points out that literary works, just like any other communicative activity, are context-dependent, since literature itself consitutes a discursive context. Clearly, a «context» in this sense 35 Some definitions of literary pragmatics do not contemplate the pragmatic analysis of the communicative contexts signified by the text (e.g. Van Dijk, 1987: 191; Albaladejo, 1983: 191). 36 See Derrida 1988: 88; Genette 1991: 43-44. Genette, though, mistakenly suggests that in the case of first-person fiction there is no authorial speech act performed, only a narratorial one (1991: 45, 63). This is of course absurd; Pip may be the author of the text of Great Expectations as a memoir (real for him, fictional for us), but Dickens is the author of both his first-person and his third-person novels/speech acts. 37 Compare e.g. Voloshinov (1986: 86, 94); Garcia Landa (1992). Also W. Harris (1988: x), although Harris's concept of interpretation is too restricted to account for the variety of actual interpretive activities. Incidentally, there is a long critical tradition describing and prescribing reader responses which assumes that readers perform activities which go far beyond the mere description of reading as «passive consumption" which some contemporary theorists would have us see as the traditional view of reception. 40 The Intertextual Dimension of Discourse (as in any other, really) is nor so much a physical location as a set of conventions for the production and inrerpretation of discourse. Literary studies, of course, have always taken into account those specific conventions, even if their connexions with more general linguistic phenomena were not theorized. Consider, for instance, one of the concepts coined by Austin, the perlocutionary act. It is clear that rhetoricians and literary critics of all ages have discussed the finality of literature; they have described the emotions caused by literary works and developed theories of composition trying to account for the intellectual or ethical effects of works on the audience; they have also developed theories of evaluation. That is, literary theory has always devoted itself to the study of the specific perlocutionary effects of literature, and indeed of the pragmatic aspect of literature as a whole. The novelty of literary pragmatics lies in its attempt to work out the analogies between specifically literary conventions and more general verbal or semiotic phenomena—the search for common principles to systematize discursive action. "We could likewise seek the critical parallels between Grice's conversational maxims and the laws of the different literary genres: the maxims, like generic laws, can be flouted but act nonetheless as regulative principles (seeTodorov, 1987: 29). Leech's «politeness principle» might also be put to good use in the study of literary response and valuation: a bad poem or novel has always been considered to be an «impolite» act^of literature, a waste of time for the reading public and a loss of face for the author's public image. Comparative studies of this type can reveal much common ground between literary and other linguistic phenomena. We are assuming that literature is defined by the use of texts, not only by the kind of texts used; in Todorov's terms, we can provide a functional but not a structural definition of literature.38 In other words, there are many literary phenomena which can be given an explanation from the vantage point of pragmatics, and only some of them had been (partially) explored from a formalist viewpoint. Very sketchily, we can point out some relevant variables: — Artistic versus nonartistic discourses, and — Aesthetic versus nonaesthetic readings.39 ' — Fiction versus nonfiction. — Narrative versus descriptions, instructions, declarations, etc. — Written versus oral communication. A poem, that is, is different from an invoice in many respects: it is not just that it is (written and read as) literature, defined for instance as autotelic or as 38 Todorov (1987: 25). Also Genette (1991: 15). 39 On the difference between these two binomials,,see Genette (1991: 39-40, 148). Chambers' (1980: 403) attribution of literary illocution to the work (and not to the author—see Petrey 1990: 81) could be more adequately understood as an aesthetic use of the work—and thus, as the reader's illocutionary inititative. -(- Speech Acts, Literary Tradition, and Intertextual Pragmatics 41 intransitive reading matter.40 Crane's poem is, moreover, narrative and fictional in specifically literary ways. Narrativity and fictionality are discursive characteristics which are not exclusively literary; they acquire new funcrions in other contexts but their literary manifestations must be studied as literary conventions, traditions and forms produced by usage. Fictionality is closely linked to literariness because it favours and enhances the autotelic element in literary texts.41 Fiction is not, by the way, an anomalous use of language: it is the other side of referentiality. In order to be able to refer to reality, signs must be able to evoke it in its absence—that is, we must be able to construct a signified world representing reality on the basis of semiosis. There is always, then, a signified world emanating from discourse, and the «normal» informative use of the sign can therefore be reversed, in order to produce virtual effects of reference and a fictional world.42 Even if we hold that literature is not immediately practical and fiction is a scarcely serious speech act, we will have to admit that they are complex and interesting phenomena, a real laboratory of sense whose systematic study is still a challenge to linguistic pragmatics. It is difficult to define literature as a specific kind or set of speech acts. Some theorists try to define literature as a discursive context in which the conventions of literary genres would be the felicity conditions of the various speech acts performed.43 Even this weaker definition lays too much emphasis on authorial intention, perhaps unwittingly. Literature is not an illocution if by illocution we understand an intentional contract between author and reader.44' It shows, rather, that this concept of illocution defines an ideal case which cannot always account for the actual use of language. From the illocutionary viewpoint, literature might be defined at best as a set of illocutionary strategies renegotiated by the reader. There are, of course, clear-cut cases; as a matter of fact it is very likely that these account for the vast majority of «acts of literatures Authors do write «literature» within a given genre (drama, fiction, poetry) and are read according to the generic conventions they invoke. In this case, literary fictionality is an illocutionary intention taken up and fulfiled by the reader. But the notion of literature has always been notoriously difficult to define. The difficulty does not derive from any magical or evanescent quality which is impossible to formulate in logical terms; it comes rather from the variety of meanings of the word «literature» and the wide variety of 40 For these definitions, see Jakobson (1981) and Genette (1991). 41 See Genette (1991: 18-20). I must note in passing that Genette's equations of fictionality with Aristotelian mimesis and with constitutive literariness (1991: 17, 32) are, in my view, mistaken. 42 Compare the argument in Derrida (1988: 96). There does not seem to be any justification to link fiction specifically to the illocutionary type of «declarations», as suggested by Genette (1991: 50). 43 Pratt (1977: 86). 44 See Genette, 1991: 63; Petrey, 1990: 82. 42 The Intertextual Dimension of Discourse discourse activities which are grouped under the label of «literature»4,5—to say nothing of the internal quality of change and evolution inherent to the modern literary tradition. "When a poststructuralist critic reads Crane's poem, it is not with the same assumptions, intentions, or goals as those of an ideally innocent reader of the late nineteenth century, or one of the reviewers who found Crane's poems «extravagant» or «eccentric». All the same, all three are reading the poem as an illocutionary invitation to «literature»—which gives us an inkling of the vastness of application of the term and of how difficult it is to provide it with a univocal pragmatic characterization. The use of written texts in literature favours the diversification of the acts of reading. As Plato argued in Phaedrus, written works circulate everywhere and are read by all kinds of people, by the experts and by those who do not care about the work; writings do not know whom they should address and whom they should avoid. Plato seems to believe that these texts say the same to all readers. It would be more accurate to say, however, that although the words on the page are the same, the echo they find is different, and writing favours therefore a multiplication of significance. In this paper I use the term «significance» broadly in Hirsch's (1977) sense, but keeping in mind its etymological kinship with «signification», the labour effected by subjects through the play of semiosis. In Derrida's terms (1988: 56), the intentional meaning is deferred/differed there— sometimes with little ^deference to the author on the part of the readers. The consecration of a work as a canonic one, a recommended one, revered by a community and read by many readers with the expectation of aesthetic or ethical benefit further increases the significance of the work. Today we read Defoe as a writer of novels; Defoe, though, did not sign Robinson Crusoe with his own name, but with that of Robinson. In the early 18th century the genre was not consolidated, lacking the classical tradition of the novel inaugurated by Defoe himself among others. It would be therefore plausible to describe those supposed novels as memoirs, false (not fictional) for us and authentic for many (mistaken) readers among Defoe's contemporaries. If users/readers create the text, as Stanley Fish used to hold somewhat mischievously, one of their main strategies for doing so is by determining which act of communication and/or interpretation is taking place in the concrete reading situation. "We should nonetheless place some restrictions on readers' interpretative freedom by pointing out that works are read within an institutional context and a tradition: reading for the classroom and reading at a swimming-pool are two different literary ..games. Their rules do not depend solely on the individual reader, but on many conventions which help determine the kind of book which is most appropriate for each situation, or the kind/degree of attention and processing that 45 See Searle, 1979; Genette, 1991:11. Speech Acts, Literary Tradition, and Intertextual Pragmatics 43 is'required from the reader. Besides, reading a book for a classroom, the «obligatory» and «reverent» reading, usually presupposes earlier kinds of reading, either utilitarian or purely gratuitous; the very possibility of reading high literature as an institutional act derives from the historical possibility of either non-literary or «non-serious» previous readings—a possibility which need not be exist anymore in the case of some works (nobody reads The Pilgrim's Progress for one's edification or for one's entertainment any longer). The bewildering range of pragmatic constraints on literature makes it advisable for us to focus, for the remainder of this paper, on just one type of pragmatic relationship which happens to be unavoidable in literature. The definition of literature might incorporate the intertextual relationship between a work and the tradition it belongs to and which help define it. As I have pointed out, literature has often been a slippery object of study for linguists. This may be the result, in part, of the contextual and intertextual complexity involved in literary enunciation and reception. A classical text, for instance, is received in a given context (a literature course, for instance), but yet another type of context must be taken into account: literary traditions, a complex intertextual network. The informed reading of the text (we are considering only one among the possible uses of literature) involves the interpretation of a text not only as a dialogue with the reader, but also with the thematic or generic tradition it implicitly refers to. The specific nature of this «referring» is largely determined by the reader's activity, since the intertextual pointers may be unambiguous or otherwise. The critical reading of a text requires that a network of intertextual links be established between the text and many other texts. These links may consist of a study of the traditional motifs incorporated in a text, the narrative techniques, historical allusions, etc. Sometimes an author will consciously point out some intertextual connections of his work with the tradition. In this way, Pope could afford to perform complex intertextual manoeuvres, such as writing imitations of Horace in the style of Swift, (the whole subsumed by his own manner) and still be followed by his audience. Often, however, the intertextual signals, however deliberate, are not overt. And deliberate signals do not exhaust the' work's intertextuality, since many potential intertextual connections are possible— potential not in the sense of being intentionally buried for the benefit of the critic, but rather as a result of the fact that the significance of a text is not limited and calculable a priori. Many critics have held that literature functions essentially on the basis of the unsaid, of indirection and reading between the lines.46 Between the lines we find the unsaid, what is not yet written, which is nonetheless an implicit element of the literary process because the work of literature is an intertextual speech act. 46 For instance, Shklovski, Empson, Barthes, Bakhtin, Iser, Fish. 44 The Intertextual Dimension of Discourse The relationship between an innovative literary text and its traditional intertexts is proportionate to that between an indirect and a direct speech act.47 Analogously to the indirect speech act, intertextuality presupposes a complex processing strategy which includes among its premisses the literal speech act as well as a contextual hermeneutics. The result of interpretation, the speech act actually performed, cannot be understood in itself, in its literal sense. The receiver must construct its literariness, that is, the contrary of literalness. Readers of literary texts do not seek only to establish the dictionary meaning of words: they also construct the significance of the texts, finding their differential specificity or their hidden kinship; they try to make a text say something more than it would say if read in isolation, on its own. The interpretive strategies which presuppose the indirection of literature, or which establish it, are therefore essential to the reception of a work as a work of literature. The density of significance presupposed and demanded by the reader requires the activation of inferences, implicatures, underground comparisons, variations on canonical motifs, more or less deliberate allusion, conscious or unconscious manipulations of types and archetypes. This game of complexities finds its most elaborate manifestation in specialised literary criticism (psychoanalysis, semiotics, deconstruction...), but quite possibly no variety of literary reading is completely devoid of it. To read a work as literature is to refer it to a tradition, understood as an intertextual praxis of readings and writings, which multiplies the meaning of a text and eventually turns it into what it was trying to become: literature. To return to the intertextuality of Crane's poem: its desert is somewhat Biblical, as is the voice who addresses the speaker. The speaker has called out to God, but the poem does not say whether it is God that answers, whether the first voice answering is the same as the second, or whether the voice's mind-reading ability is proof of its divine or its authorial origin—a way.the poem thematizes its status as a «self- dialogue» of the poet's. The poem does seem to imply that what looks like a desert to someone need not do so to everyone, or that the aridity of the textual soil and the emptiness of the hotizon are also a function of the observing eye. Interesting parallels and differences could emerge, too, from a comparison between this poem and the trials of the Israelites in Exodus. Faith and providence have become problematic in Crane, God is more inscrutable. But the poem itself becomes less barren and more readable after an intertextual itinerary through the Bible. Many of Crane's poems are parables on nihilism, comparable to those of Nietzsche in his Zarathustra. We find in this mode a double level of indirection: the parable as such 47 Compare Chico Rico (1988: 214); Genette (1991: 55). It is clear, though, that these phenomena cannot be accounted for in Searle's or Bach and Harnish's models of SAT in their original formulation. These models rest, "as we have been arguing, on a sentential conception of grammar. Intertextuality comprises a whole new range of types of indirection at a textual-discursive level. The point of these types of speech act (to use a Searlean term) would consist precisely in establishing a connection between texts in order to increase significance. \ Speech Acts, Literary Tradition, and Intertextual Pragmatics 45 presupposes an underlying level of sense which invites interpretation; the deviant parable establishes an additional distance from the religious intertext with which it contrasts. The coincidence of both Crane and Nietzsche choosing the genre of the parable to oppose it to the Christian tradition is not only formal: it has ideological and parodic implications. But this is just one possible direction of analysis. Another relevant and equally interesting intertextual route might lead us towards Blake or Dickinson, or towards the translations of oriental poetry which encourage the use of free verse at the end of the nineteenth century. T. S. Eliot conceived of tradition as an empyrean of consecrated masterpieces («monuments») whose significance nonetheless changes, structurally avant la lettre, when a new masterpiece is added. More informally, Borges takes up this idea in an essay on Kafka and his predecessors. He argues that the tradition preceding an author, his predecessors, becomes visible (somewhat like a 3-D optical effect) with the appearance of a work which relates in an innovative way a set of previously unconnected elements in earlier works. Kafka makes us see Kafkian elements in earlier authors—his effect on the work of Hawthorne or Melville would be a good instance of this phenomenon. Perhaps we might go a bit further to argue that this intertextual network is constituted by Borges' act of reading as much as by Kafka's writing. We could finish by going back to Crane, yet another American («precursor of >) Kafka, to an additional piece full of Biblical and mythic echoes. I will use it as an allegory of the following hermeneutic quandary—that in understanding the structure, origin and nature of the literary object a dialectic is established between its objective nature and the interpretive act: A MAN SAW A BALL OF GOLD IN THE SKY; HE CLIMBED FOR IT. AND EVENTUALLY HE ACHIEVED IT- IT WAS CLAY. NOW THIS IS THE STRANGE PART: WHEN THE MAN WENT TO THE EARTH AND LOOKED AGAIN, LO, THERE WAS THE BALL OF GOLD. . NOW THIS IS THE STRANGE PART: IT WAS A BALL OF GOLD. AYE, BY THE HEAVENS, IT WAS A BALL OF GOLD. {The Black Riders, no. 35) 46 The Intertextual Dimension of Discourse The reader's interpretive act will help decide whether a poem is made of clay or gold, whether it shines because of its substance or because it is an illusion and a fiction; because we assume the right aesthetic distance, or because we have bothered to turn our gaze upon it more than once. Those who ask whether cool critical analysis is not a way of taking the lustre away from literature may also find an answer in this poem. But the poem needs to be asked. Works Cited ALBALADEJO MAYORDOMO, T. 1983. «La critica lingih'stica*. In Introduction a la critica literaria actual. Pedro Aullon de Haro (ed.). Madrid: Playor. 141-207. ARISTOTLE. 1954. Aristotle's Rhetoric and Poetics. Trans. W. Rhys Roberts {Rhetoric) and I. Bywater [Poetics). Introd. Friedrich Solmsen. New York: Modern Library. AUSTIN, J. L. 1980. How to Do Things with Words. 1962. Oxford: Oxford UP. BACH, K. and R. M. HARNISH. 1979. Linguistic Communication and Speech Acts. Cambridge (MA): MIT Press. BAKHTIN, M. 1981. The Dialogic Imagination. Ed. Michael Holquist. Trans. Caryl Emerson and Michael Holquist. Austin: U of Texas P. BALLMER, T. T, and W. BRENNENSTUHL. 1981. Speech Act Classification: A Study in the Lexical Analysis of English Speech Activity Verbs. Berlin: Springer. BALLY, Ch. 1977. El lenguajey la vida. Buenos Aires: Losada. BARTHES, R. 1971. «Litterature et significations. 1963. In Bardies, Essais critiques. Paris: Seuil. 258-76. BENVENISTE, E. 1968a. «De la subjectivite dans le langage». 1958. In Benveniste (1968c: 258-266). — 1968b. «Les Relations de temps dans le verbe francais*. 1959. In Benveniste (1968c: 237-258). — 1968c. Problemes de linguistique generale. Paris: Gallimard, 1968. BLOOMFIELD, L. J. 1933. Language. New York: Holt. BORGES, J.L. 1985. «Kafka y sus precursores*. In Borges, O-tras inquisiciones. 1960. Madrid: Alianza. 107-9. BUHLER, K. 1967. Teoria del lenguaje. 1934. Madrid: Revista de Occidente. CASTILLA DEL PINO, Carlos. 1983. «E1 psicoanalisis y el universo literario». In Introduction a la critica literaria actual. Ed. P. Aullon de Haro. Madrid: Playor. 251- 346. CHAMBERS, R. 1980. «Le Masque et le miroir: Vers une theorie relationnelle du theatre*. Etudes litteraires 13.3: 397-412. CHATMAN, S, (ed.). 1971. Literary Style: A Symposium. London: Oxford UP. GHICO RICO, F. 1988. Pragmdtica y construction literaria. Alicante: Universidad de Alicante. CHOMSKY, N. 1965. Aspects of the Theory of Syntax. Cambridge (MA): MIT Press. CRANE, Stephen. 1972. The Complete Poems of Stephen Crane. Ed. Joseph Katz. Ithaca: ■. Cornell UP. DERRIDA, J. 1988. Limited Inc. Evanston: Northwestern UP. DIJK, T.A. van. 1972. Some Aspects of Text Grammars: A Study in Theoretical Linguistics and Poetics. The Hague: Mouton. Speech Acts, Literary Tradition, and Intertextual Pragmatics 47 — 1980. Textoy contexto: semdnticaypragmdtica del discurso. 1977. Madrid: Catedra. — 1987. «La pragmatica de la comunicacion literaria*. In Mayoral (1987: 170-194). DOLEZEL, L. 1967. «The Typology of the Narrator: Point of View in Fiction*. In To Honor Roman Jakobson. The Hague: Mouton. 1: 541-52. — 1971. «Towards a Structural Theory of Content in Prose Fiction*. In Chatman (1971: 95-110). DUCROT, O. 1980- «Pragmatique linguistique II: Essai d'application: Mais—Les allusions a l'enonciation—Delocutifs, performatifs, discours indirect*.- In Parret et al. 487-575. ELIOT, T. S. 1951. «Tradition and the Individual Talent*. 1919. In Eliot, Selected Essays. 3rd ed. London: Faber. 13-22. EMPSON, W. 1984. Seven Types of Ambiguity. 1930. London: Hogarth. FISH, S.E. 1980. «Literature in the Reader: Affective Stylistics*. 1970. In Fish, Is There a Text in This Class? Cambridge (MA): Harvard UP. 21-67. GARCIA LANDA, J.A. «Speech Act Theory and the Concept of Intention in Literary Criticism*. Revista Canaria de Estudios Ingleses 24 (1992): 89-104. GENETTE, G. 1991. Fiction et diction. Paris: Seuil. GRICE, H.P. 1989a. «Logic and Conversation*. 1975. In Grice, Studies in the Way of Words. Cambridge (MA): Harvard UP. 22-40. — 1989b. Studies in the Way of Words. Cambridge (MA): Harvard UP. HALLIDAY, M.A.K. 1970. «Language Structure and Language Function*. In New Horizons on Linguistics. Ed. J. Lyons. Harmondsworth: Penguin, 1970. 140-165. — 1971. «Linguistic Function and Literary Style: An Inquiry into the Language of William Golding's The Inheritors». In Chatman, (1971: 330-365). HARRIS, W.V 1988. Interpretive Acts: In Search of Meaning. Oxford: Clarendon. HARRIS, Zelig S. 1970. «Discourse Analysis*. 1952. In Papers in Structural and Transformational Linguistics. Dordrech: Reidel: 313-48.. HENDRICKS, Wi.O. 1973. Essays on Semio linguistics and Verbal Art. The Hague: Mouton. HIRSCH, E. D., Jr. 1966. Validity in Interpretation. New Haven: Yale UP. — 1976. The Aims of Interpretation. Chicago: U of Chicago P. INGARDEN, R. 1973a. The Literary Work of Art: An Investigation on the Borderlines of Ontology, Logic, and Theory of Literature. 1931, 1965- Evanston: Northwestern UP. — 1973b. «The Functions of Language in the Theater*. 1958. In Ingarden, The Literary Work of Art. Evanston: Northwestern UP, 1973- 375-396. ISER, W. 1974. The Implied Reader: Patterns of Communication in Prose Fiction fiom Bunyan to Beckett. 1972. Baltimore: Johns Hopkins UP. JAKOBSON, R. 1981. Linguistica y poetica. I960. Trans. Ana Ma Gutierrez-Cabello. Madrid: Catedra. JESPERSEN, O. 1968. The Philosophy of Grammar. 1924. London: Allen and Unwin. LANSER, Susan Sniader. 1981. The Narrative Act-: Point of View in Prose Fiction. Princeton: Princeton UP. LEECH, G. 1983. Principles of Pragmatics. London: Longman. LOPEZ EIRE, A. 1980. Origenes de la poetica. Salamanca: Ediciones Universidad de Salamanca. LOTMAN, J.M. 1982. Estructura del texto artistico. 1970. Trans. Victoriano Imbert. 2a ed. Madrid: Istmo. LOZANO, J., C. PENA-MARIN and G. ABRIL. 1982. Andlisis del discurso: hacia una semiotica de la interaction textual. Madrid: Catedra. 48 The Intertextual Dimension of Discourse Speech Acts, Literary Tradition, and Intertextual Pragmatics 49 LYONS, J. 1977. Semantics. 2 vols. Cambridge: Cambridge UP. MALINOWSKI, B. 1930. «The Problem of Meaning in Primitive Languages». In C. K. O. and I. A. Richards, The Meaning of Meaning, 2nd ed. London: Routledge. 296-336. MARTINET, J. 1975. Clefs pour la semiologie. 1973. 3rd ed. Paris: Seghers. MARTINEZ BONATI, E 1972. La estructura de la obra literaria. I960. Barcelona: Seix Barral. MAYORAL, J. A. 1987. Pragmdtica de la comunicacidn literaria. Madrid: Arco-Libros. OHMANN, R. 1971a. «Speech, Action, and Style*. In Chatman, (1971: 241-254). — 1971b. «Speech Acts and the Definition of Literature*. Philosophy and Rhetoric A: 1-19. — 1972. «Speech, Literature, and the Space Between*. New Literary History 4.1: 47-63. PARRET, H., et al. 1980. Le Langage en contexte: Etudes philosophiques et linguistiques de pragmatique. Amsterdam: Benjamins. PETOFI, J.S., and A. Garcia Berrio. 1978. Lingiiistica del texto y critica literaria. Madrid: Alberto Corazon. PETREY, S. 1990. Speech Acts and Literary Theory. New York: Routledge. PLATO. 1989. Fedro, o de la belleza. Trans. M. Araujo. Introd. A. Rodriguez Huescar. Madrid: Aguilar. PRATT, MX. 1977. Toward a Speech Act Theory of Literary Discourse. Bloomington: Indiana UP. RICHARDS, I. A. 1964. Practical Criticism: A Study of Literary Judgment. 1929. London: Routledge. RICCEUR, P. 1975. LaMetaphore vive. Paris: Seuil. SAUSSURE, F. de. 1984. Curso de lingiiistica general Ed. Ch. Bally and A. Sechehaye. Barcelona: Planeta. SCHMIDT, S.J. 1977. Teoria del texto. 1973. Madrid: Catedra. SCHOFER, P. and D. RICE. 1977. «Metaphor, Metonymy and Synechdoque Revisited*. Semiotics 21.1-2: 121-149. SEARLE, J.R. 1979. Expression and Meaning: Studies in the Theory of Speech Acts. Cambridge: Cambridge UP. — 1980. Actos dehabla: Ensayo defilosofia dellenguaje. 1969. Trans. L.M. Valdes Villanueva. Madrid: Catedra. SEGRE, C. 1985. Principios de andlisis del texto literario. Trans. M. Pardo de Santayana. Barcelona: Critica. SHKLOVKSI, V. 1965. <Axt as Technique*. 1917. In Russian Formalist Criticism: Four Essays. Ed. Lee T. Lemon and Marion J. Reis. Lexington: U of Kentucky P: 3-24. SPERBER, D. 1975. «Rudiments de rhetorique cognitive*. Poetique 23: 389-415. STEINTHAL, H. 1961. Die Geschichte der Sprachwissenschaft bei den Griechen undRomern, mit besondere Rucksicht auf die Logik. 1863. Hildesheim: George Olms. ^STRAWSON, P. F. 1964. «Intention and Convention in Speech Acts*. Philosophical Review 73: 439-460. TODOROV, T. 1987. La Notion de litterature et autres essais. Paris: Seuil (Points): 27-46. UITTI, K.D. 1971. «Philology: Factualness and History*. In Chatman, (1971: 111-132). — 1977. Teoria Literaria y~Lingiiistica. 1969. Trans. R. Sarmiento Gonzalez. Madrid: Catedra. VAN OVERBEKE, M. 1980. «Pragmatique linguistiqu'e: I - Analyst de l'enonciation en linguistique moderne et contemporaine*. In Parret et al*(1980: 389-486). VOLOSHINOV, V. N. 1986. Marxism and the Philosophy of Language. 1929. Trans. L. Matejka and I. R. Titunik. Cambridge (MA): Harvard UP. WITTGENSTEIN, L. 1958. Philosophical Investigations. Trans. G. E. M. Anscombe. Oxford: Blackwell.