Deconstructive Intentions: On the Critique of the Hermeneutics of Understanding more

DECONSTRUCTIVE INTENTIONS: ON THE CRITIQUE OF THE HERMENEUTICS OF UNDERSTANDING Jose Angel Garcia Landa Univefsidad de Zaragoza Differences in interpretive assumptions, whether in literary hermeneutics or elsewhere, are usually the result of ideological differences, and therefore are not likely to be solved simply through rational debate. Nevertheless, a detailed discussion of the conceptual bases of the dispute must not be rejected out of hand, since it may clarify some issues and'help formulate a more comprehensive theory. A case in point would be the debate between a more traditional, historicist hermeneutics such as is articulated, for instance, by E. D. Hirsch (1967, 1976) and the post-structuralist versions of interpretation, such as deconstruction or reader- response criticism. Hirsch holds that there is an objectively determinable authorial meaning in the work, defined as the intention of the author as expressed in conventional linguistic forms. The aim of interpretation, according to Hirsch, is to retrieve that meaning through a process that is ideology-free. Thereafter, interpreta- tion is subservient to criticism, which is more ideological. Criticism seeks to evaluate works and determine their significance —not to be confused with meaning, since significance varies form age to age and from reader to reader. The position of post-structuralist critics on this issue is well known: they reject that meaning is objectively identifiable or recoverable: it is endlessly reworked and expanded in reading. These two critical positions cannot be reconciled on each other's terms, since they presuppose different objects of analysis and different roles for the analyst. In thispaperl will examinedeconstructiveandotherpost-structuralisfconceptions of meaning —above all, literary meaning— and point out what I think are some misconceptions in their analysis. I agree with the basic post-structuralist tenet that meaning is a function of discourse and is subject to ideological contention: Hirsch's conception of objectivist criticism cannot be sustained as such.1 Objectivism should not be defined as the attempt, to capture an essence, a historical or textual noumenon, a meaning in se; an objectivist theory does not try to discover Truth, but to translate different representations or theories into each other so that a greater understanding is achieved and two previously unrelated fields of inquiry can be seen to answer to a common logic. That is, a particular interpretation cannot be objective, but objectivist: the roleof truth is a purely ideal andregulativeone, and successive interpretive situations 1.-1 discuss this point further in comparing the critical positions of Hirsch and the New Critics ("Authorial Intention in Literary Hermeneutics: On Two American Theories"). will require new solutions, since the terms to be subsumed under the interpretation will be different too. Post-structuralist theories of interpretation could also benefit from this extended definition of objectivism in hermeneutics. The post-structuralist emphasis on the prol ifcralion of meaning and the productive role of reading can be given a more coherent formulation if some of the concepts of traditional hermeneutics are adapted to it—and, incidentally, post-structuralist theorists will be able to make these claims compatible with the possibility of communication, which is usually disregarded in their accounts. This new formulation would analyse meaning using the concepts of "intention" and "convention" in a sense close to the one they have in speech act theory. It will be convenient, therefore, to counter some post-structuralist critics' objections to speech act theory. I will concentrate on the notions of authorial meaning (as it relates to intention) and of significance, on the borderline between them and the conditions of their for- mulation. Both authorial meaning and significance are the object of theoretical debate —that is, their own status as interpretive concepts, and not only their application to specific texts, is the object of debate. In this they are not different from other critical concepts. Debates may lake place in two main ways: through the practical criticism of texts or in theoretical works. The latter need not be more deliberate that the former, though it usually is so. I will now examine some post-structuralist pronouncements on the role of intentions and conventions in interpretation, and try to show that some post- structuralist claims are based on a misguided concept of convention and intention. ]. Deconstruction and Interpretation: Jacques Derrida Dcconstruction has already been under attack from a variety of perspectives, ranging from Scarle to Fredric Jameson. The central issue of these polemics as far as we arc concerned here is whether dcconstruction has any implications for a theory of interpretation and for the possibility of communication. Attacking the conventions of interpretation has been denounccdas a self-destructive enterprise, and a contradic- tory one. Michael Fischer, for instance, opposes J. Hillis Miller's concept of interpretation as an activity which reaches no truth, or an endless wandering: "One wonders how Miller gets outside the prison house of language to see its walls" (1985, 53). Free play of meaning is an important axiom of deconstruction. But in Fischer's view, there can be no such thing as a free play of the interpreter: interpretive conventions must necessarily be constraining, because that is what allows them to be enabling (19.85, 118). Wendell Harris also warns against the confusion of deconstruction and interpretation. In his view, deconstructionists conflate what may be said of a word insofar as it is an item of langue with what may be said of it insofar as it is an instance of parole (1988, 25). The Saussurean theory of language has often been misrepre- sented by dcconstructivc critics (sometimes starling with Derrida himself), together with its implications for literary studies. For instance, the differential definition of language as a structure of wholly arbitrary oppositions, Of terms with a purely negative value, is presented by Derrida in his essay "Difference" (Derrida 1982) as the central 20 conception of Saussure's theory of language, but Saussure stresses just as much the positive terms which result from the negativity of linguistic difference: "Bien que le signifie et Ie signifianl soienl, chacun pris a part, purement differentiels et negatifs, leur combinaison est un fait positif' (Saussure 1949: 166). Moreover, the negativislic definition of the terms is only applicable to language as langue, to the study of the linguistic system, and not to the study of speech. Saussure knows that in parole the role of the context is all-important (Harris 1988, 8), and that parole rests on the relatively stable conventions of langue. Interpretation is concerned with the meaning of instances of parole, not with langue. The rclationshipof deconstruction to criticism at large has often been represented as doing away with the concept of authorial intention and celebrating the free activity of the critic, his creation without limits. An intellectual activity resulting from such assumptions is no doubt possible, but is perhaps not what we would want to define as "criticism." Hirsch sees Derrida as "the most fashionable of the theologians of cognitive atheism in the domain of literary theory" (1976,13). Possibly Derrida would feel quite pleased in being called a "cognitive atheist." But if Hirsch means thatDerrida's theory does not make or cannot make any difference between conscious (intentional) and unconscious meanings, or even between the intentionality of the author and that of the critic, he is wrong. This may be so in some of Derrida's followers. In Derrida's own version of deconstruction we find somewhat different assumptions, even if they exhibit their own kind of duplicity. Let us start with two statements on authorial intention by Derrida, found in works which are not primarily concerned with a critique of hermeneutics. The first discusses the relationship of authorial intention to the status of the critical comment as a signifying structure: Produirc cettc structure signifiante ne peut evidemment consister a rcproduire, par le rcdoublement efface et respeclueux du commentaire, le rapport conscient, volontaire, intentionnel, que l'ecrivain instituedans ses echanges avec l'hislorie a laquelle il appartient grace a l'element de la langue. Sans doutece moment du commentaire redoublant dpit-il avoir sa place dans la lecture critique. Faute de la reconnaitre et de respecter toules ses exigences classiques, ce qui n'est pas facile et requiert tous les instruments de la critique traditionnelle, la production critique risquerait de sc faire dans n'importe quel sense et s'autoriser a dire n'imporle quoi. Maiscelindispcnsablcgarde-foun'ajamaisfaitquepro/eger,iln'a jamais ouvert une lecture. (1967,227) Let mc only note that when Derrida speaks of the "garde-fou" or parapet of authorial meaning as being "indispensable," he means it. The second text is from Margins of Philosophy, and seems to be somewhat more adventurous in defining therole of authorial intention; The text —Hegel's for example— functions as a writing machine in which acertain number of typed and systematically enmeshedpropositions (one has to be able to recognize and isolate them) represent die "con- 21 scious intention" of the author as a reader of his "own" text, in the sense we speak today of a mechanical reader. Here, the lesson of the finite reader called a philosophical author is but one piece, occasionally and incidentally interesting, of the machine,2 It would be easy to turn Derrida's method against himself and show that this "incidentally interesting" element, this margin of his theory, plays a hidden regulative role in the analyses of more central or fascinating parts of the textual machine. After all, docs not a measure of their interest for the deconstructor derive form the fact that they are not the conscious intention of the author? Does not this deconstructi ve project depend on the previous identification of the construction? I will leave those as (rhetorical) questions, but there are other implications in Derrida's texts. They assume that the text has one (authorial, intentional) meaning, which is historically fixed. The textual machine would not be the same —maybe it would not work— without that component part. They assume that this meaning can be (must be?) identified, and that the analyst must be able to tell thedifferencebetweenauthorial meaning and other kinds of meaningorsignificance, between hisown activity and theauthor's. These assumptions are quite close to Hirsch's own, although Dcrrida does not stress the need to keep authorial meaning constantly in mind —presumably because his main concern is with another phase on the critical activity. In his textual analyses, Dcrrida often uses the conccptof authorial intention practical ly as well as theoretical ly, even as he subord inalcs this hcrmeneulic reading of the text to his own deconstructive reading.3 However, Derrida has sometimes implied that deconstruction has implications for the interpretation of authorial meaning, and has cast doubt on the validity of the central concepts of hermeneutics. In "SignatureEventContext" and "Limitedlncabc..." (in Dcrrida 1988a) he "deconstructs" the principles of Austin's theory of speech acts. According to Dcrrida, the basic notions of this theory, the notions of literal meaning, context, communication, intention, etc., are "logoceniric" constructs; they partake of a logic which rests on the notions of presence, being, speech, consciousness, purity, seriousness, originality, authenticity and ccntrality. This logic works with binary oppositions, in which one term is privileged at the expense of the other. The deep links between the non-serious, the impure, the marginal, the derived, itcrability, the pre- tended, writing, etc., cannot be understood in the terms of this hegemonic logic: "logoccntrism," Dcrrida argues, must be transcended. Derrida wants to belittle concep- tual and logical ("logocentric") thought, and to ppivilcge instead an alternative logic without concepts, a logic whose terms are undccidable and do not exclude their polar Other. Supposedly, such terms as "difference," "supplement," "pharmakon," "hy- men," and several others as deployed by Derrida would accomplish this feat: Derrida 2. - Dcrrida 1982, xi. Barihcs puts forward a comparable claim: "It is not that the Auihor may not 'come back' in the Text, in his text, but then he docs so as a 'guest'" (1977b, 161). 3. - See, for instance, Dcrrida 1988b, 199. 22 claims that they are neither words nor concepts; they open the play of logic and signification, rather than being caught in it. Unfortunately, while their status as words and concepts is evident from their presence in a linguistic text, their mercurial quality as non-words and non-concepts is merely asserted by Dcrrida (cf. Rorty 1989, 124 n. 6). Or rather, by means of a legerdemain, he transfers the pre-conceptual quality of the semiotic operations named by these terms to the name itself. If Derrida wanted to build an il-logic —well, he has succeeded. But this il-logic will not dismantle logic in the way claimed by Derrida. It may dismantle the straw logocentrism he constructs to challenge it afterwards. Derrida's characterization of what a concept is and how it works is itself seriously monolithic and essentialist. 'It is bound to be, so that his non-concepts may look alluring in comparison. Let us hear what he has to say on the logic of conceptual thought: To this oppositional logic, which is necessarilly, legitimately, a logic of "all or nothing" and without which the distinction and the limits of a concept would have no chance, I oppose nothing, least of all a logic of approximation [dpeuprei'l.asimpleempiricismofdiffcrenceindegree; rather I add a supplementary complication that calls for other concepts, for other thoughts beyond the concept and another form of "general theory", or rather another discourse, another "logic" that accounts for the impossibility of concluding such a "general theory." (1988a, 117) However, concepts are not defined by a simple logic of all or nothing (cf. Searle 1983b, 78: Harris 1988, 161). Or rather, this distinction does not require an equally exhaustive definition of concepts in all contexts. Once they are contextually defined, concepts follow the logic of "enough or not enough, here and now," rather than that of "all or nothing." Derrida obscures the circumstance that any logic and any use of a concept is always contexlualized —this, against his own claim that "context is always, and always has been, at work within the place, and not only around it" (1988a, 66). The context-sensitive nature of concepts affects Derrida's attack on speech act theory in at least two ways: - It invalidates his claim that a "logocentric" distinction can never be made pure, and is therefore questionable. One such polar distinction (for instance, between seriousness and irony) need only be made as pure as is necessary for the desired communicative effect to take place. If people interpret some utterances seriously and others ironically, it is sufficient proof that such distinctions are indeed made on sufficient (not exhaustive) reasons. - It brings out the fact that in this attack Derrida has ignored the contextual space defined by the theory itself. This is apparent, for instance, in his readiness to drive the discussion to metaphysical grounds which are beyond the concern of speakers and theorizers of speech acts alike (cf. Rorty 1989,132). - Derrida's version of analysis never addresses the social and cultural constraints of the interpretive context, even though they also might be said to be at work within the object and not only around it. Instead, Derridean analysis focuses on the object as a philosophical abstraction —in thecaseof speech acttheory ,on the"gcneral citationality" 23 of a sign, quite apart from the specific value assumed by the sign in a context. Petrcy has shown that Austin's theory has a greater potential for contextualized analysis than Dcrrida's: For Dcrrida, the conventions that matter apply to the units of every signifying form and thus inhere in the nature of the mark. Since conventions inherent in the mark are obviously trans-historical and universal, Derrida's conventions are independent of context whereas Austin's are coterminous with it.... Dcrrida's most sustained critique of speech-act theory is based on the nature of speech rather than on the acts it performs. (1990:139) My concern here, however, is not with the aim or validity of the whole project of deconstruction; only with its implications for the concept of interpretation I am discussing. That is, if there are any implications: Derrida's attitude is highly ambivalent in this respect. He seems to denounce the logocentric concepts as if they were the product of bad faith and led us to narrowness of thought, as if their legitimacy was invalidated by their metaphysical assumptions. For instance, he says that if the absence which is said to characterize written communication was shown to be a condition of all signs, the traditional conception of writing "would appear to be noncritical, ill-formed, or destined, rather, to insure the authority and the force of a certain historical discourse" (1988a, 7). The second half of the sentence suggests the cultural critique of false consciousness: surely nobody wants to be caught insuring "the authority and the force of a certain historical discourse"! Dcrrida surreptitiously suggests that the essential normativity of language, the conventionality of iterable signs (or signatures), of contextual relevance, of speech acts, is sustained by means of institutionalized violence: There is always a police and a tribunal ready to intervene each time that a rule... is invoked in a case involving signatures, events, or contexts... If the police is always waiting in the wings, it is because conventions are by essence violable and precarious, in themselves and by the fictionality that constitutes them, even before there has been any overt transgression, in the "first sense" of to pretend. (1988a__105) Just as fiction is a con vention, all institutions are conventions. This is a favourite deconstructivist claim, also made by Barbara Johnson and Stanley Fish. But the aim of this claim is not clear. Is it to assert that all institutions are fictional in the same sense that fiction is fictional? This would indeed be the direction pursued by this particular deconstruction of the legitimacy of hierarchies. It is hardly a description of the actual state of affairs, since the logocentrjc concepts denounced by Derrida are well alive and at work everywhere, including Derrida's own text, which is read as an essay and not as fiction. Is it, then, an injuction to subvert the present state of affairs and conflate fiction and nonfiction through our actions? And how? There is no clear answer in Derrida's text, 24 except the practice of deconstruction itself. Should we, then, always deconstruct, in lieu of interpreting? The answer would seem to be "yes." However, we also find disclaimers like the following: By no means do I draw the conclusion that there is no relative specifity of effects of consciousness, or of effects of speech (as opposed to writing in the traditional sense), that there is no performative effect, no effect of ordinary language, no effect of presence or of discursive event (speech act). It is simply that those effects do not exclude what is generally opposed to them, term by term; on the contrary, they presuppose it, in an asymmetrical way, as the general space of their possibility. (1988a, 19) This, however, need not be a concern for the interpreter who tries to retrieve the authorial meaning of a text. It is a metaphysical observation on hermeneutics, and not some new kind of interpretive principle. In the interview which serves as an afterword for Limited Inc, Derrida is even more explicit: I have never "put such concepts as truth, reference, and the stability of interpretive concepts radically into question" if "putting radically into question" means of contesting that there are and that there should be truth, reference, and stable concepts of interpretation. I have —but this is something entirely different— posed questions that I hope are radical concerning thepossibility of these things, of these values, of these norms, of this stability (which by essence is always provisional and finite). (1988a, 150) But this provisional stability does its work, something which is ignored in this account. Thisduplicityisatworkbehindevery definite slaiementinLzVra7&i/rtc.Derrida'S project is eminently ambiguous, since it both suggests and denies that it has implications for the theory of interpretation. He announces the break with the conception of communication as communication of consciousness, the disqualification ("or the limiting"!) of the concept of context, etc. (1988a, 8-9), only to let them in through the back door. Derrida constantly repeats the gesture of warning us against the police, only to tell us later that wC need not worry; they arc our friends, it was not a warning (1988a, 132). He was asking to be misunderstood or overemphasized, as he complains he has been, in creating the implied reader of his essays —and I do not think this escaped his attention, or that he is so clumsy a writer as not to be able to control the tone that will convey his meaning. Derrida speaks of Austin's theory as if it conceived of people as pretematurally self-conscious and hyper-legislatively-minded creatures: "In order for the context to be exhaustively determinable, in the sense required by Austin, conscious intention would at the very least have to be totally present and immediately transparent to itself and to others, since it is a determining center {foyer] of context" (1988a, 18). Austin never put forward such a preternatural concept of intention. As Dcrrida himself notes 25 in "Limited Inc abc," consciousness does not imply self-consciousness. Furthermore, a context docs not have to be exhaustively determined. On the contrary, it is assumed to be already determined unless an interlocutor asks for further specification, or behaves in a way that shows that his assumption about the context are different. In a characteristic gesture, Derrida accepts the notion of consciousness and normal circumstances, but giving them a marginal place in his theory: "those effects do not exclude what is generally opposed to them on the contrary, they presuppose it, in an asymmetrical way, as the general space of their possibility" (1988a, 19). Again, "they presuppose it" in a metaphysical consideration of language —not in a description of language use, like speech act theory. This move also equivocates, by introducing in a theory of language use a set of considerations which are quite irrelevant to it. We live among these "effects," and therefore we do not perceive them as effects, any more than we perceive our consciusness as an effect. They are inscribed in our mode of being as they are inscribed in our language. Derrida'scentralizationofthemarginhastheundesirableeffcctofmarginalizing the center. From the structural possibility of infelicitous speech acts, Derrida concludes that "the opposition success/ failure [echec] in illocution and in perloculion thus seems quite insuficientand extremely secondary [derivee]" (1988a, 15). They can only be said so by someone who can speak of "the teleological lure of consciousness" (1988a, 18) as if an account of communication which did not use theconcept of consciousness were possible. Normal circumstances, serious or felicitous speech acts work as normative principles precisely in that they do not presuppose their "other"—they are the unmarked case.4 Speakers assume that an utterance is literal unless they have a reason to think otherwise. A literal utterance is not interpreted through its contrary, or exhaustively compared to its context. The intrinsic, unmarked, normal context which is a part of the locutionary meaning of a sentence is enough to interpret it if it does not run against the extrinsic context of the utterance. On the contrary, to interpret a lie as a 1 ic we cannot help but set it against its other, the true statement it purported to be. To the extent that Derrida obscures this circumstance, he is not putting forward a theory of interpretation —he is noting the presence in theories of interpretation of the same logic of purity and impurity which is present in other semiotic areas, like the construction of sexuality or the cultural notions of East and West. This analysis does not have the same implications in all of these fields. A theory of culture without chauvinist cultural hierarchies is welcome, but as far as pragmatics is concerned I do not see any sense in a theory of communication which renounces thenotions of literal meaning of felicitous speech act as regulative elements used by the speakers. At least Derrida had not provided one. He wants to be both Fish and Hirsch. In ordinary speech, he allows the presence of intention or literal meaning as "derivative" effects; when he analyzes the source of the derivation, he can't help using the notions of presence, center and 4.- As to the debate on the status of fiction, Searlc and Derrida speak at cross-purposes because neither one is willing lo move into the reasonable middle-ground: that producing fiction (i.e., 'non-serious' speech acts) is itself a serious speech act A theory along theselines is put forward by Pratt (1977). 26 consciousness. The blind spot which opens the visibility of the text is its absent center.5 The principles of writing and supplcmentarity Derrida finds in Rousseau's writing are, it seems, central in these texts. They represent the mise en abyme of that lextuality: "Et nous verrons que 1'abTme n'estpas ici un accident..." (1967,233). Moreover, this blind spot is a (present) center in Derrida's critical commentary, since its identification is the organizing principle of his text. Deconstructive practice undermines deconstructive theory just as much as the theory undermines the practice. Derrida continually uses the language of essence and presence. Is deconstruction yet another discovery of centers? If not, what is the sense of "formulating," "producing laws" or essays with an "axial purpose," putting forward "ultimate justifi- cations," or constituting a "theory of structural necessity" as Derrida claims he is doing (1967,233)? By preserving the "vulgar concepts" (of writing, intention, etc) and setting them in a wider frame, Derrida's theory still has a thrust towards totalization.6 To the extent that a theory is an intentional construct and its purpose is the conscious recognition or establishment of relationships, Derrida's writing is a misguided experi- ment in mise en abyme —I suspect that his silence on this aspect of his writing is a wilful desire to have his text victimized by the same law it denounces, but his equivocations about the implications of the extended concept of writing make him play down the role of his own text as an intentional construct The text is unnecessarily victimized. Its effectivity is therefore doubly impaired, conceptually and rhetorically. As a conclusion, we may say that while Derrida's project aims at transcending hermeneutics, it has lo accept the "provisional" validity of the discipline in much the same lines as wc have been describing —a circumstance from which he is at pains to draw attention. He accepts a similar role of intcntionality in the interpretation of writing and of speech (1988a, 60), the conventional nature of linguistic intention in an account which isnotsodiffcrenlfrom Scarle'sown (Derrida 1988a, 129), the importance of rigor and a knowledge of tradition (1988a, 130—cf. Scholcs 1990), and the necessity of truth and fixed meaning as a regulative aim for the interpreter, one which is never wholly reachable for Derrida (1988a, 120,129). ' 2. Meaning in American Deconstruction So far I have been criticising the hermeneutics of radical suspicion implicit in some deconstructive; moves from the point of view of a hermeneutics of understanding, 5. - Derrida 1967,234. Or is it the absent center of the text as it is seen by the critic? This view, which would seem to use a more criLical version of presence, does not seem to be favored by Derrida. 6. - The counter-movement, which denounces the absurdity of this kind of enterprise, is fraught with false steps. For instance, in "Limited inc abc..." (1988a, 39) he speaks of "the interminable character" of speech act analysis, because the analysis itself must be a speech act. Derrida might equally well denounce the absurdity of a book on syntax on the basis that it must use syntactic constructions to describe syntax, or the absurdity of dictionaries because they, have lo explain words with other words. 27 a hermeneutics which starts from theassumption that understanding and communication are taking place all the time, and that it is one of the functions of criticism to explain those effects of meaning. The free play of meaning is being arrested all the time by the interpreters, both in everyday conversation and in literary interpretation, a fact which is difficult to account for if dcconstruction instead of understanding is taken lobe the basic interpretive manoeuvre. This is a hidden assumption in much of the work on dcconstruction done in the English-speaking academy. Other advocates of what Scholes calls "nihilist hermeneutics" are Paul de Man and Frank Kermode. According to de Man, the task of criticism is driven against the inherent duplicity of language, the impossibility of making the expression coincide with what is expressed: 'The interpretation of everyday language is a Sisyphean task, a task without end and without progress, for the other is always free to make what he wants differ form what he says he wants" (1983,11). Radical relativism is de Man's alternative to the fallacy of a finite and single interpretation (1983, 10). This is the most radically skeptic of de Man's misgivings about interpretation, because it bears on the use of language as a whole. However, it is made in the face of actual use of language, where interpretation is constantly taking place in a successful way (i.e., accomplishing its aims), and is not at all "unending" or Sisyphean —instead, it is most often immediate and unproblcmatic. As de Man pronounces the radical undecidability of language ex cathedra, without making any attempt at proving it or at contextualizing the validity of his assertions, his statement can only be adequately met in a Dr. Johnson- like way. I take it to be self-refuting, and therefore obviously false—unless deMan has felt free to make what he means differ from what he says he means! Some literary theorists have a confusing tendency to extend to any interpretive activity the conclusions of their own critical project, without taking into account that interpretation is an essentially contextualized activity, an examination of a text from a specific situation and with specific aims in mind. For instance, Frank Kermode's view of interpretive validity is (though only at times) equally bleak: "World and book, it may be, are hopelessly plural, endlessly disappointing: we stand alone before them, aware of their arbitrariness and impenetrability, knowing that they may be narratives only because of our impudent intervention, and susceptible of interpretation only by our hcrmeneutic tricks." (Kermode 1979, 145). Insomuch as this statement bears on linguistic and literary interpretation, its strategy is to obscure the fact that our "impudent intervention" and "hcrmeneutic tricks" are to a great extent pressuposed in the structure of the very intentional phenomena they analyze. These are therefore far from being arbitrary or impenetrable. An Urdu translation of Hamlet is impenetra- ble as far as I am concerned. I may find the English original ultimately impenetrable, but hardly in the same sense. Differences in degree are essential differences in hermeneutics. In spite of an occasional sweeping statement like the one quoted before, de Man is usually concerned with a level of interpretation which I think does not exclude the dctcrminacy of authorial intention. Rather, it presupposes some degree of detcrminacy. At limes, de Man is ready to recognize this relative determinacy, and also a regulative concept of objectivity; he affirms that a thrust towards totalization is necessary for criticism, even if it is endless (1983,31-32). Like Derrida's, de Man's 28 dcconstruction presupposes the interpretation of authorial meaning as a preliminary step in the critical activity. "The reader is given the elements to decipher the real plot hidden behind the.pscudo-plot, but the author himself remains deluded" (1983,1.04). That the authorial meaning is seen to be insufficient or deceitful, a "pseudo-plot", does not alter the fact of its presence. De Man's "reader" is in a privileged position to judge of the author's or the critic's blindness, to see both the intended statement and the inadvertent pattern which emerges behind it and which is for de Man much more interesting. So far, this sounds pretty reasonable. But: De Man shares with Derrida a puzzling tendency to erase the transformative power that his own critical activity exerts on the text, to ignore his own responsibility for the hidden patterns he finds behind the author's intention. After examining the critical work of Lukacs, Poulet, Blanchot and others, he concludes: "Critics' moments of greatest blindness with regard to their own critical assumptions are also the moments at which they achieve their greatest insight" (1983, 109). This beautiful paradox is grounded on an equivocation: it presupposes that the author is in some way responsible for anything we may find in his text. He did not intend that meaning but then he somehow intended it. But the insights which the critics' blindncss has made possible are de Man's own, and not theirs. Their own conscious insight counts as blindness for the critic. In the following passage on deconstructive reading, de Man's self-erasure is explicit and fully deliberate, although his account of intentionality is different: The reading is not "our" reading, since it uses only thelinguistic elements provided by the textitself...Thedeconstructionisnotsomethingwehave added to the text but it constituted the text in the first place. A literary text simultaneously asserts and denies the authority of its own rhetorical mode, and by reading the text as we did we were only trying to come closer to being as rigorous a reader as the author had to be in order to write the sentence in the first place. (1979,17) It seems that the author is no longer blind: he is a rigorous reader, and all works are deconstructive. Not for de Man: in themselves. There is aft illegitimate step in de Man's reasoning. If a text appears to contain tensions within its structure, such as an undermining of the metaphors it rests on, it does not follow that these tensions are apparent for the author or for just any reader. The interpreter who makes this claim has to invest an amount of comparison, logical reasoning and justification —critical work— in order to support it. Scmiotic phenomena do not exist in themselves. They always exist for somebody, and it is fair to say that Paul de Man creates the text's self- deconstruction as much as he identifies it. The same equivocation is at work at all critical moments where de Man uses the concept of intention. A third version runs thus: "The question as to whether the author himself is or is not blinded is to some extent irrelevant. It can only be asked heuristically, as a means to accede to the true question: whether his language is or is not blind to its own statement" (1983, 137). I own that the notion of how language can be blind or insightful in itself (as opposed to blind or insightful for a subject, an interpreter) completely escapes my understanding. If de Man's language is insightful in itself, that 29 does not solve my present problem, and his insight will not fare the better in my interpretation, for I take it to be wrong. De Man's account of meaning is not far from the idealist assumptions of some New Critics. This faith in the contents of "the text itself is certainly related to Wimsatt and Beardsley's invocation of the work as an autonomous, self-sufficient object without any necessary links to human intentionality (Wimsatt and Beardsley 1946). It is not surprising that de Man makes explicit the conflation of creative writing and criticism which was impending or subliminal in the New Critics: "Poetic writing is the most advanced and refined mode of deconstruction; it may differ from critical or discoursive writing in the economy of its articulation, but not in kind" (1979,17). Like Dcrrida's "absent center," de Man's conception of "blindness" or the self- deconstruction of texts is curiously essentialist: both seem to assume that a particular insight is in the text and constitutes its real structure even if nobody had perceived it before themselves. The meaning of the text is always already in it for de Man (1983,30). This claim does not apply merely to verbal meaning and the standardized illocutions: it includes the deconstructive reading of the text; for de Man, a text always deconstructs itself. In my view, it is Dcrrida's or de Man's perspective which creates the object they see. It does not exist "in the text" wailing for an interpreter at a given moment of its life and of the development of critical work. I do not think that the conclusions of these theorists of deconstruction arc simply wrong —they are often fascinating and challeng- ing. Nevertheless, they need to be set in a proper perspective where they do not set at nought their own basis and are no longer in conflict with a more comprehensive hermenculics. They must be rethoughton the basis of more relativist principles —I hope it no longer sounds puzzling that this increased relativism must also be an increased objectivism. Following de Man's ideas, Barbara Johnson holds the theory that (some?) works of literature and criticism already contain their own,dgconstruction. She has used this conception, for instance, in her critiques of readings by Dcrrida or Barthes. These deconstructions of deconstructions are curiously prone to invoking the conception of authorial intention at the crucial moment: de Man finds that Rousseau had deconstructed himself before Derrida followed suit with a somewhat simplified and sloppy job in Of Grammatology7; Johnson finds that Balzac is at least as subtle a deconstructor as the Barthes of SIZ (1990,11), or that a text by Lacan seems to have anticipated Dcrrida's criticism of it.8 In turn, the critics re-enact the false steps which they criticise in the authors they study. And when the self-deconsirucuon of a text forms a neat pattern, it is rewarded with the privilege of having been deliberate: "Whatever Derrida actually thinks he is doing here, his contradictory way of explaining it obeys the paradoxes of 1- De Man 1983,139. Dc Man makes the very undeconstructivist claim that "Rousseau's text has no blind spots" —I agree, with this qualification: there are no blind spots in a text until someone spots them. 8.- Johnson 1980,116. The texts in question are Lacan's "Seminar on 'The Purloined Letter"' and Dcrrida's'The Purveyor of Truth." 30 parergonal logic so perfectly that this self-subversion may have even be deliberate" (1980, 131). Johnson's concept of authorial intention is a misty one. Can the self- subversion be deliberate "whatever Derrida actually thinks"? I am afraid that the deconstructive critics, by paying an exclusive attention to the spaces between the lines, sometimes leave the more basic concepts unexamined. The (ale by Poe that Lacan, EtemdaandJohnsonarcwriungaboutoperiswimaStoicepigraph:"M^ acumine nimio" —"Nothing is more disagreeable to wisdom than too much cunning." The purloined letter was hanging in the middle of the mantelpiece for anyone to see. 3. Reader-Response Criticism and the Interpretation of Authorial Intention Reader-response criticism is not incompatible in principle with a historicist hermeneutics. However, some extreme critical positions among reader-response theorists deny any objectivity to the text, and reduce it to a function of the interpreter's strategies. This conception was originally sustained by Stanley Fish, and it has become widespread to a surprising degree. Fish's position on most critical issues has gone through substantial variations. His original project of an "affective slylistics" was extremely anti-intentionalisl: The formal units are always a function of the interpretive model one brings to bear; they are not "in" the text, and I would make the same argument for intentions. That is, intention is no more embodied "in" the text than are formal units; rather an intention, like a formal unit; is made when perceptual or interpretive closure is hazarded. (1976a, 447) ■ Paper has no intentions; so far so good. But Fish draws the wrong conclusions from this premise. Since the author's intention is a construction of the reader, he concludes that it is a fiction. Therefore, all kinds of formal features (since they, too, are "fictional") can be freely attributed to the author's intention. To- describe a reader's experience "is to describe his realization (in two senses) of an author's intention" (1976a, 475). This leads to the absurd conclusion of not making a difference between authorial attitude and the authorial attitude inferred by a reader. Fish disarmed his theory and declared that all inferences were equally valid and illegitimate. The reader and the writer of the text become indistinguishable, and the reader's activity constitutes the text.9 According to Fish, the formal features of the text do not have an objective reality: they are derived from the activity of the critic.10 Our definition of "objective" does not however invoke a jump into the noumenal. The formal features do have an objective existence. They are not derived from the activity 9. - These views arc echoed by many other critics, such as Horton (1979, viii) or Tompkins (1980, x). 10. -Fish 1976a, 476; cf. also Horton 1979, 22. 31 of the critic,11 but they are not brute facts either. Objectivity is the result of interpretation, not a device to foreclose it. In a more general way, the whole of Fish's enterprise has been directed at obscuring the difference between the metalinguistic and other uses of language. What a sentence means for Fish is, quite simply, what a sentence does to the reader, the play of expectations and frutrations during the reading process (1980b, 72). Langue, locu- tions and illocutions disappear as concepts in his theory, and leave their place to a mass of undifferentiated parole and perlocutions. In the wake of Derrida, Fish has denied the legitimacy of basic concepts such as "direct speech act" (Fish 1976b) or "literal meaning" (Fish 1978). Consequently, he cannot analyze the semiotic structure of the work: he has no tools left. This kind of enterprise has been sufficiently criticized.12 The extent of its in- tended bearing on hcrmcncutics is not even clear. According to Fish, "One wonders what implications [this theory] has for the practice of literary criticism. The answer is, none whatsoever" (1980a, 370). This theory is most useful in bringing out some aspects of the "naive" frst reading of a casual reader. Not that it will explain the whole of the reader's activity, since that would require the use of the metalinguistic constructs that Fish has rejected, and which are a part of any reader's experience. As a theory of interpretation and criticism, it is obviously null. And since it does not lay claim to validity (Fish 1976c, 195), we may as well leave the matter there. In any case, Fish has reneged many of his earlier claims, and in more recent work (1987) he has introduced a timid (and idealistic) version of holism. Reader-response critics are prone to indulging in Fish's same naive anti- theoretical assumptions. Horton, unlike Fish, attempts to draw a clear line between reading and interpretation, but believes nevertheless that the process of reading as such is immediately relevant as an interpretive alternative (1979,27). And, like Fish, she disregards the difference between the experiences of expectation and the surprises of the text which are ultimately rejected by the reader as accidents of the empirical reading process and those who contribute to his interpretation of the text and are interpreted as constituent parts of an idealized reading process. Horton opposes the "fun" of the reading process to the boring structures unearthed by structuralist criticism, without seeming to realize that these are not different interpretations of a text, but simply different kinds of focus on one phase or another of the interpretive process. She must use the concept of structure herself when she describes how a "spatial" pattern is built by the reader on the sequential activity of reading: moreover, she ignores the fact that the sequence of the work (which she opposes to this strucluralion) is itself the result of several layers of slructuration. But some post-structuralist critics are too ready to carry all their concepts in one bag. For instance, popular versions of T. S. Kuhn's theories (as voiced, for instance, by Fish) often-con Hate the status of the explanatory models of science wi th that of fiction. 11. -Sparshott 1976,104; Fowler 1976, 254. 12. - Cf. for instance Regis 1976, Bagwell 1983, Scholes 1984, Reeves 1986, 32 But the first are tools for making sense of reality, while fiction is a particular discursive genre understood with its own conventions, which require that both author and reader know that it is not true. Unless these conventions are at work in writing and reading, the text is not being understood as a fiction. However, the role of scientific models does not require that their users recognize them as provisional. Their explanatory power is the same if they are believed to be truth itself. Likewise, an author's attitude towards this problem is indifferent for practical purposes. It does not have to be recognized in order to use the model. Which means that scientific models are not, as far as their use is concerned, a form of fiction, but a form of factual discourse. In the context of scientific debate, they may be presented as hypothetical discourse, but still there is no identity with fiction. In the case of fiction, the author and the reader know that the fictive world they construct is possible only on the basis of the real world where they are. In scientific models, there is no such duplicity of worlds: it is the nature of the real world which is being debated. The hypothetical construct does not leave the world where the theory is being formulated unaffected, Instead, it coincides with this world, it constitutes it to some extent. The major claim of reader-response critics is to have dissolved the work into a variety of readings, to have made mean ing dependent on the reader as well as the author: "A literary work means one thing or another depending upon the ways in which we weave each particular detail into a coherent fabric of interpretation, and that, in.turn, depends upon which other details of the text we choose to bring to the fore with it" (Horton 1979, 56). "Meaning is no longer a property of the text but a product of the reader's activity" (Tomkins 1980, xvii). This claim jnvolves a deceptive sleight-of- hand: there are many different kinds of meaning in a text (f. i., graphic, locutionary, illocutionary, perlocutionary) and all are declared to be the product of the reader's activity. The difference in their degrees of sharabilily is ignored., "The position Fish argues in ['Literature in the Reader']," Tompkins claims, "does not deny that words have meanings, nor does it assert that the reader's response is untrammelled and free from textual constraints. The kinds of experience that li terature affords are regulated by the1 linguistic and literary competence of the individual reader" (1980, xvii). Fish is more subtle than that, I agree. He avoids those blatantly nonsensical claims but leads us to make the conclusions ourselves by obscuring a fundamental difference in the way the reader creates meaning, the difference between partially codified and wholly codified meaning. All semiotic maneuvers performed by the reader are then contemplated as equally creative and collaborative. It is not surprising that ignoring this difference leads Tompkins from asserting the total hegemony of the reader to the equally shocking claim of asserting the total hegeinony of the author: "The reader reacts to the words on the page in one way rather than another because he operates according to the same set of rules that the author used to generate them. The reader's experience, then, is the creation of the author; he enacts the author's will" (1980, xvii). This inflated conception of the authorial intention, which would seem to be so inimical to the principles of reader- response criticism, if often found in Fish's own early phase (1967; 1980). At times it is solved by going to the other extreme: the reader enacts the author's will, but then the author is only a construct of the reader's —and we have come.back full circle to the noumcnal conception of the text. 33 The theoretical thrust of Horton's enterprise is paralyzed when she evades an answer to the question whether wrong interpretations are possible: "it makes more sense to acknowledge and maybe even to celebrate the meaning-making capability of readers than it docs to try to deny its existence or hope to defeat it by reaching toward single, univocal interpretations of our text" (1979, 125), The attempt" to find an univocal meaning in texts is (and here Horton follows Fish) an instance of naive "positivism." We may observe at this point a shift from the mctacritical enterprise of this reader-response critic towards an evaluative standpoint, and this under the pretense of preserving an enlightened perspective over the petty univocal interpretations of critics. The celebration of multiple meanings will only do in some interpretive contexts. Many intellectual activities do notadmit it as a working principle. And a true mctacritical theory should explain why disagreements do occur and why they are thought to be important. Fish's variety of reader-response would submit the totalizing reading of a psychoanalytic cri tic ora Marxist critic toasetofrelativislic principles which isinimical to it, whilcprclcnding to accept iton itsown terms. The kind of reader-response criticism favoured by Fish or Horton is not a privileged, comprehensive perspective which embraces all other possible critical approaches, but a very definite critical activity on a level with them, with its own assumptions about the nature of a text and the aims of criticism. 4. Conclusion My basic objection to post-structuralist hermeneutics is that it conflates under the very general terms of "Fictions" and "conventions" many different types of "Fiction" and crucially different levels of "convention." The result is an impossibility to account for the possibil ity of communication between different sets of conventions, reconstruction and reader-response theory do not attempt to make the basic move of isolating the interpreter's consciousness from the author's, and in so doing their readings become blatantly ahistorical; all previous authors are transfigured into the post-structuralist stance. A theory of interpretation should preserve a regulative distinction between the historical authorial meaning, acccssibleatagcncral level through linguisticconventions, andothermeaningsof the text, which aredevclopcd by critical work in itsown and more speciFic sphere of activity. A theory of hermeneutics which preserves these regulative concepts will be more explanatory than those which does not recognise this distinction, because it accounts for more decisions actually made by critics. Critics (not least deconstructivc critics) often need to draw a line between the author's view of the text and their own. The distinction between authorial meaning and interpretation, however, should be thought as a regulative one. The theory thus becomes more inclusive and flexible. The hypotheses which do not recognise this distinction, or the purely regulative role of objective authorial meaning, cannot translate this theory into their own terms, while this hypothesis includes them and sets them in a wider context. Only a historicist hermeneutics offers a field wide enough to integrate (or increase the mutual translat- ability of) such disparate fields of theory as structuralist narratology, speech act theory, reader-response and reception theories, phcnomenological hermeneutics, Marxist 34 r ; literary theory and deconstruction —since these theories, whatever their explanatory power, are also historically' and culturally localized intellectual activities. . It will be observed that, according to the notion of objectivity defined above, the ; truth value of the previous assertions is strategical and situational —that is, they may be I true in the 1990s, in a specific context which I hope is not excessively narrow. I WORKS CITED ! . . . Bagwell, J. Timothy. 1983. "Who is afraid of Stanley Fish?" Poetics Today 4.1: 127- 134. Barthes, Roland. 1977b: "From Work to Text." 1971. In Barthes 1977c, 155-164. Barthes, Roland. 1977c. Image - Music - Text. New York: Hill and Wang. De Man, Paul. 1979. Allegories of Reading. New Haven: Yale UP. De Man, Paul.1983. Blindness and Insight: Essays in the Rhetoric of Contemporary Criticism. 1971. Minneapolis: U. of Minnesota P. Derrida, Jacques. 1967. De la gramrwtologie.Vms:Mm\i\i. Derrida, Jacques. 1982. Margins of Philosophy. Trans. Alan Bass. Chicago: U. of Chicago P. Trans, of Marges de la philosophic 1972. Derrida, Jacques. 1988a. Limited Inc. Evanslon: Northwestern UP. Derrida, Jacques. 1988b. "The Purveyor of Truth." In The Purloined Poe: Lacan, Derrida, and Psychoanalytic Reading. Ed. John P. Muller and William J. Richardson. Baltimore: Johns Hopkins UP. 173-212.' Fischer, Michael. 1985. Does Deconstruction Make Any Difference? Blpominglon; Indiana UP. Fish, Stanley E. 1967. Surprised by Sin: The Reader in Paradise Lost. New York: St. Martin's. Fish, Stanley E. 1976a. "Interpreting the Variorum." Critical Inquiry 2: 465-485. Fish, Stanley E. 1976b. "How to Do Things With. Austin and Searle; Speech Act Theory and Literary Criticism." MLN 91: 983-1025. Fish, Stanley E. 1976c. "Interpreting 'Interpreting the Variorum."' Critical Inquiry 3: 191-196. Fish, Stanley E. 1978. "Normal Circumstances, Literal Language, Direct Speech Acts, the Ordinary, the Everyday, the Obvious, What Goes Without Saying, and Other Special Cases." Critical Inquiry 4.4: 625-644. Fish, Stanley E. 1980a. Is There a Text in This Class? The Authority of Interpretive Communities. Cambridge (MA): Harvard UP. Fish, Stanley E. 1980b. "Literature in the Reader: Affective Stylistics." New Literary History 2.1 (1970): 123-162. Rpt. in Tompkins 1980, 70-100. Fish, Stanley E. 1987. "Change." South Atlantic Quarterly 86.4: 423-444. Fowler, Alastair. 1976. "Intention Floreat." In Newton-De Molina. 1976: 242-255. Garcia Landa, Jose Angel. 1991. "Authorial Intention in Literary Hermeneutics: On Two American Theories." Misceldnea 12 (Zaragoza, 1991): 61-92. Harris, Wendell V. 1988. Interpretive Acts: In Search of Meaning. Oxford: Clarendon. Hirsch, E.D., Jr. 1967. Validity in Interpretation. New Haven: Yale UP. Hirsch, E.D., Jr. 1976. The Aims of Interpretation. Chicago: U of Chicago P. Horton, Susan R. 1979. Interpreting Interpreting: Interpreting Dickens' Dombey. Baltimore: Johns Hopkins UP. v Johnson, Barbara. 1980. The Critical Difference. Baltimore: Johns Hopkins UP. Kcrmode, Frank. The Genesis of Secrecy: On the Interpretation of Narrative. Cambridge (MA): Harvard UP, 1979. Newton-De Molina, David. 1976. On Literary Intention. Edinburgh: Edinburgh UP. (Introduction, vii-xvii). Pratt, Mary Louise. 1977. Toward a Speech Act Theory of Literary Discourse. Bloomington: Indiana UP. Reeves, Charles Eric. 1986. "Literary conventions and the Noumenal Text: Stanley Fish's Egalitarian Poetics." Neophilologus 70.3: 334-340. Regis, Edward, Jr. 1976. "Literature by the reader The 'Affective' Theory of Stanley Fish." College English 38: 263-280. 36 Rorty, Richard. 1989. "From Ironist Theory to Private Allusions: Derrida." In Rorty, Contingency, Irony and Solidarity. Cambridge: Cambridge UP. 122-137. Saussurc, Ferdinand dc. 1949. Cours de linguistique genirale. Ed. Charles Bally, Albert Sechchayc and Albert Riedlinger. Paris: Payot. Scholes, Robert. 1984. "Who Cares About the Text?" Novel 17.2: 171-180. Scholes, Robert. 1990. "Interpretation: The Question of Protocols." In Scholes, Protocols of Reading, New Haven: Yale UP. 50-88. Searle, John R. 1983b. "The World Turned Upside Down." New YorkReview of Books 30.16 (27 Oct. 1983): 74-79. Sparshotl, F. E. 1976. "Criticism and Performance." 1967. Rpt. in Newton-De Molina 1976, 104-115. Tompkins, Jane P., ed. 1980. Reader-Response Criticism: From Formalism to Post- Structuralism. Baltimore: Johns Hopkins UP. Introduction, ix-xxvi. Wimsatt, W. K., and Monroe C. Beardsley. 1946. 'The Intentional Fallacy." Rpt. in Wimsatt, The Verbal Icon. New York: Noonday, 1958. 3-20. 37