The Pursuit of Signs: Semiotics, Literature, Deconstruction 
by Jonathan Culler.
Routledge, 256 pp., £7.95, July 1981, 0 7100 0757 4
Show More
Show More

If you teach or study literature in a university, the chances are you’ve spent at least some of your time recently arguing with colleagues about the uses and abuses of literary theory. Not only do structuralism, deconstruction and their offshoots draw the biggest audiences at professional conferences, but the quarrel over them has aroused the curiosity of mass journals like Newsweek. Amidst this swirl of publicity and controversy, Jonathan Culler has managed to remain calm long enough to write a reasoned defence of deconstruction, one from which both scholars and general readers will be able to learn a good deal. Culler is fairly termed an apologist for the post-structuralist critical modes typified by the work of Roland Barthes and Jacques Derrida, yet he writes without the supercilious tone of the former or the soporific word-play of the latter. Nor does Culler indulge in cheap vilification of opponents, suggesting that unwillingness to buy the whole post-structuralist bill of goods is a symptom of psychic repression or a police mentality. Even more refreshing, Culler doesn’t obscure the steps of his argument – and there is one, though this is a book of separate essays – by resorting to slogan-like clichés or great fogs of verbiage. Unlike critics who claim to take risks in their work while writing so evasively as to be protected from any criticism, Culler takes genuine risks in presenting deconstructionist insights as arguments rather than as dithyrambic assertions. He is thereby open, not only to the counter-arguments of opponents, but to the dismissal of allies, who will be able to accuse him of vulgarising the mysteries.

This last charge is easy to level at anybody who ventures to define deconstruction, the thing being so protean, so taken up itself with the equivocal nature of ‘representation’, that almost any attempt to represent it will by definition be a misrepresentation. Depending on which deconstructionist text one reads, deconstruction may be seen as a philosophical system, a critique of systems, an analysis of the conditions of system-making, a method, a critique of method, a mode of reading, a mode of literature in its own right, or as all or none of these things. Nor are the consequences of deconstruction very clear. Derrida writes that he is not ‘junking’ traditional metaphysical concepts but trying to ‘transform’ them, ‘to turn them against their presuppositions’. But when one asks to what end and from what standpoint the concepts are to be transformed, it’s hard to pin Derrida down to anything more than vaguely apocalyptic sentiments. Nevertheless, certain general tendencies of deconstruction can be summarised. Against the traditional view that philosophy and literature are founded in the nature of things or the spirit of man, deconstruction borrows from structuralism the argument that these disciplines, and ‘man himself’, are, in Culler’s words, ‘a product of discursive forces’. Deconstruction amends structuralism, however, by adding that these discursive forces are not fully under our control and cannot be fully charted by any science of language (hence the term ‘post-structuralism’, often used synonymously with ‘deconstruction’). ‘Skeptical of the possibility of mastering meaning within a comprehensive system or discipline,’ Culler writes, deconstruction ‘investigates what the most powerful and interesting texts have to tell us about signification and shows how they undo the logics of signification on which they rely’. Specifically, what these texts betray, under the scrutiny of the deconstructor, is the quixotic nature of trying to represent in language the ‘presence’ of the self and the world. Representation tries to pass itself off as a token of that presence, but the fact that we need a substitute only dramatises the inadequacy of the representation. The language we use to express ourselves and the world is finally what prevents our ever succeeding.

To prove his point, the deconstructor frequently exploits the element of figuration which, he claims, underlies even the most abstract language, but which is seen most obviously in literature. Thus, in a chapter on the apostrophe, one of several set-pieces of practical deconstruction, Culler maintains that this rhetorical figure simultaneously asserts the power of language to bend the world to human desires and concedes that the power rests on language alone. An apostrophe such as Blake’s ‘O rose, thou art sick’ figuratively closes the gap between the post-Enlightenment poet and the natural object he addresses. Yet ‘it figures this reconciliation as an act of will, as something to be accomplished poetically in the act of apostrophising; and apostrophic poems display in various ways awareness of the difficulties of what they purport to seek.’ In its self-undoing nature, however, the apostrophe is merely typical of all tropes and figures, and all language, for the deconstructor.

In another piece of sustained deconstruction, Culler examines some statements from M.H. Abrams’s The Mirror and the Lamp in order to argue that Abrams, one of the principal critics of deconstruction, has himself unwittingly practised deconstructive strategies on romantic literary theory. Abrams had pointed out that Coleridge introduced his favorite analogy for poetry, the growing plant, in order to avoid mechanical determinism, yet if we follow the logic of the figure, we see that one form of determinism has merely been substituted for another. Furthermore, according to Culler, Abrams’s book as a whole enforces a different thesis from the one Abrams intended. Whereas Abrams presents the romantics as abandoning the mirror (mimesis) for the lamp (self-expression) as the model of poetic activity, his material keeps suggesting that they did not succeed in getting rid of mirror-images. They illustrated what appears to be a kind of deconstructionist law: that in cases of figurative opposition, ‘the repressed term always inhabits its opposite.’ Culler’s proof of this in part rests on the point that ‘the logic of the figures themselves’ undoes the opposition: ‘a mirror is no use without light, and there is no point in illuminating a scene unless something will register or reflect what is there.’ In short, ‘a language come alive... is a language unable to control its own tropology.’

If all this sounds (and indeed, I think, is) a bit strained, one should note that Culler doesn’t offer these exercises as ‘interpretations’, and he laments the fact that deconstruction in the United States has been misunderstood and misappropriated as a method for generating more explications of literary works, of which we already have a superfluity. Derrida, Culler argues, does not interpret philosophical texts so much as ‘describe a general process through which texts undo the philosophical system to which they adhere by revealing its rhetorical nature’. Derrida is interested, not in the ‘meaning’ of a text, if by that term one refers to ‘a unifying purpose that would assign each part its appropriate role’, but rather in the general conditions of meaning which the text exemplifies. But ‘when deconstruction comes to America,’ Culler observes, a shift takes place whereby ‘the text does not just contain or perform a self-deconstruction but is about self-deconstruction, so that a deconstructive reading is an interpretation of the text, an analysis of what it says or means.’ Paul de Man and J. Hillis Miller, according to Culler, have turned deconstruction into a ‘methodological principle’ by which semantic ‘undecidability’ is, as Miller puts it, ‘always thematised in the text itself ...’ In other words, deconstructionist ideas are made into ‘privileged themes of literary works’, and deconstruction by this route is ‘tamed’ into a ‘version of interpretation’.

This point is part of Culler’s larger quarrel with what he calls ‘the widespread and unquestioning acceptance of the notion that the critic’s job is to interpret literary works,’ the idea that the touchstone of critical work of any kind is ‘whether it actually assists us in our understanding of particular works’. It’s not that Culler is opposed to critical interpretation as such, but that he thinks we ought to be paying more attention to the assumptions that underlie interpretation, ‘the conventions and operations of an institution, a mode of discourse’. Readers of Culler’s earlier Structuralist Poetics will recognise the argument: structuralism has shifted the perspective of criticism – rather than a criticism which discovers or assigns meanings, structuralist poetics aims to ‘define the conditions of meaning’. What the present book adds to the earlier position (or subtracts from it, depending on your viewpoint) is the argument that deconstructive insights should be included within semiotics in order to keep the whole operation, the ‘investigation of the ways and means of literary signification’, from taking itself too seriously, from setting up as a ‘master discipline’ claiming to be ‘a complete and non-contradictory science of signs’. I have doubts about the workability of this partnership of deconstruction and semiotics, which is rather like a rocky marriage in which the partners depend on one another but never cease quarrelling. But there is cogency in Culler’s argument that the object of criticism ought to be not ‘literary works themselves but their intelligibility: the ways in which they make sense, the ways in which readers make sense of them’.

Since Culler is the author of a first-rate interpretive study (Flaubert: The Uses of Uncertainty), one can’t explain his wish to demote interpretation as mere self-interest. Nor should one confuse his argument with celebrated attacks against interpretation during the Sixties, which were really attacks on thinking for curtailing our due allotment of sensory gratification. Culler is calling for more thought, not less, in response to the discontinuities of our cultural and educational situation. One pernicious effect of making interpretation the primary model of critical activity, Culler suggests, is to isolate literature from all contexts except the putatively ‘literary’ at the very moment when literature has already been isolated from the cultural environment. In a chapter on ‘Literary Theory in the Graduate Program’, Culler argues that semiotics can help construct the needed contexts by encouraging students to see literature ‘in its relations to other ways of writing about human experience, such as philosophy, sociology, anthropology and history’.

If that sounds like an affront to ‘traditional’ views of literature, one should consider whether critics such as Sidney, Dr Johnson and Arnold would have been likely to recognise any tradition with which they were familiar in the assembly-line industry which professional literary explication seems to have become. It’s only relatively recently that criticism has come to be expected to engage in elaborate exegesis of meanings, frequently to the neglect of the general moral and cultural concerns that had interested traditional men of letters.

Nor is it very helpful to object that what students really need is not the distractions of theory but immersion in the great texts themselves. ‘We often complain,’ says Culler, ‘that students have not read enough when they come to college, but the problem is not a quantitative one that would be solved by more assigned readings. The problem is structural, involving the marginal situation of literature within the students’ cultures.’ (And not just theirs, one should add.) Exposing students to the classics is surely important, but to imagine that such exposure is all that is needed fails to address the problem. Specifically, such a view ignores the ‘intertextual’ nature of meaning, the fact that meaning is not immanent in words and phrases but depends for its comprehension on the construction of the reader, which in turn depends on what the reader brings with him, on his knowledge of prior presuppositions and texts. (This point is spelled out in Culler’s chapter, ‘Presupposition and Intertextuality’.) Reading founders when this prior knowledge is no longer transmitted by the culture, at which point, it can be argued, theoretical awareness must intervene in order to help reconstruct what has ceased to be tacitly assumed. How this theoretical awareness is to be integrated into the graduate and undergraduate curriculum, of course, remains a problem.

Nevertheless, Culler’s analysis of the reading process strikes me as potentially useful to teachers. It also seems to me compatible with the premises of standard historical interpretation: but on this point Culler would evidently disagree. One senses that a further impulse underlying his reservations about interpretation is doubt that interpretation is a rational activity subject to principles of adjudication. One attraction of semiotics for Culler may be the metacritical perch it affords him above the vulgar strife of competing interpreters, each of whom naively fancies he is getting closer to a ‘correct’ reading of a corpus of work, or at least believes that such a goal is not mere self-delusion. ‘To make the goal of literary studies knowledge of the meaning of each individual literary work,’ Culler states, ‘involves the futile attempt to impose a particular standard and a single goal upon the activity of reading.’ One could object that knowledge of a work’s meaning might embrace many goals, and that we don’t merely ‘impose’ a standard by fiat but give reasons for our interpretations which are open to refutation. Culler, however, would retort that all such reasons depend on a standard which is finally culturally-determined. There is no position outside history and culture from which to claim, as the naive historical critic does, that one has recovered the ‘original meaning’ of an author.

Culler is no ordinary relativist, however: he attacks Norman Holland for the view that textual meaning is a function of the individual reader’s ‘identity theme’. For Culler, there are norms of interpretive correctness, but since these norms are institutionally-produced, they are changeable. This view would seem to align Culler with the most recent work of Stanley Fish (only Fish’s earlier views are taken up by Culler here), who argues influentially in Is there a text in this class? that interpreters have no recourse to ‘independent evidence’ in disputes over texts, since what counts as evidence is necessarily constituted for any reader by the interpretive strategies which produced his reading in the first place. What Fish takes to be the consequences of this argument is far from clear, but he seems to want to hold that it discredits the idea that an interpretation can be ‘true to the text’. For though ‘the text’ is ‘always set’ for any given community of readers, Fish says, it is set only by the standards of right reading which happen to be ‘in force’ in that community, and therefore ‘it is a text that can change.’ Culler argues to similar effect when he suggests that the standard assumption that a reading can be grounded in a prior text is valid only within a limited causal logic. From the standpoint of another logic, a deconstructive one, this priority is reversed, so that it’s readings that predetermine texts and not the other way around. Since for Culler these ‘two logics’ can never be reconciled by an overarching system – as Kant, for example, reconciled the antinomy of necessity and freedom – we are returned to the conclusion that meaning is permanently unstable.

The argument is ingenious, but it poses problems. If deconstruction and standard interpretation belong to different logics, how can one put the other in question? An unwanted implication of Culler’s two-logics scheme would appear to be that deconstruction is no threat to orthodox reading practices. This consequence would appear to follow, too, from his argument that deconstruction shouldn’t be confused with interpretation, that it deals with the conditions of signification and not with ‘what a text says or means’. Culler says that deconstruction shows how texts ‘undo the logics of signification on which they rely’, but if this undoing is not part of a text’s meaning, what precisely is undone? Presumably, what is undone is the desire for ultimate unity with the self and the world, that is the vain goal of all representation. Deconstruction provides elaborate proof that concepts can’t pretend to be the things they stand for. But if one never supposed that concepts made this pretence, one might feel that it is the deconstructors who are promoting the confusion, perpetuating superstitions in order to justify their campaign against them.

There is thus an ambiguity about the self-undoing effect and how we can tell it has occurred. Is the effect an affair of words, consciousness, or both? When Culler says that ‘apostrophic poems display in various ways awareness of the difficulties of what they purport to seek,’ his phrasing leaves unclear who or what possesses the ‘awareness’. (Abrams has made a similar point in a commentary on Culler’s treatment of the mirrors and lamps in Abrams’s book.) Culler’s assertion that ‘the logic of the figures themselves’ is self-undoing, that ‘language come alive ’ is ‘unable to control its own tropology’, attributes the effect to language. Yet when Culler restricts the effect to ‘post-Enlightenment poetry’, as he seemingly does at at least one point, he implicitly locates it in the awareness of a specific group of poets, not merely in their language. If the effect could be attributed to language alone, there would be no reason not to ascribe it to apostrophes written in any period. The ‘life’ of language would appear to be the ace-in-the-hole of deconstructive reading: if there is evidence that the writer of a text is aware of the falsity of his own claims, the deconstructor scores a point, but if this awareness can’t be found in the writer, all is not lost since it can then be attributed to the ‘desire’ inherent in language.

This view that language is ‘alive’ (alive and ill?) incongruously traps Culler in that superstition of linguistic immanentism which it was the great project of structuralism and modern linguistics to wipe away. Of course that’s the point: the return of repressed mystery haunts the would-be science of signs. The dilemma, however, is self-induced, for, as Culler is well aware in sober moments, it isn’t the logic of ‘figures themselves’ that determines meaning but the conventions of usage. One can attribute awareness or life to figures only by abstracting them from the communicative situation in which they occur. A friend of mine once ridiculed a trite phrase by serving drinks containing pebbles and stones to guests who’d asked for them ‘on the rocks’. His ploy illustrated, not that his guests’ language couldn’t control its own tropology, but merely that the same words can be used to perform very different kinds of speech act, so that we can determine what an expression means (what act is being performed) only by an inference from the total speech situation. As E.D. Hirsch has argued, words in themselves, at the level of the dictionary (the langue), represent a wide range of potential meanings, only a limited number of which are actualised in any particular speech-event (or parole). Even if one agrees with the structuralists that human beings are constituted by discursive systems, that the ‘self’ is merely a kind of text, the fact remains that until words are used by somebody on a particular occasion to perform some task, they can mean virtually anything or nothing. ‘On the rocks’ is just as legitimately read as a reference to stones (or shipwreck, or a failing marriage) as to ice cubes if the phrase is considered apart from a situation in which somebody is trying to order a drink. This dependence of meaning on context occasions no metaphysical crisis – unless, again, one mistakenly expects words to be inherently tied to what they signify.

There are problems, too, with Culler’s other argument for the instability of signification: that since interpretive norms are culture-bound, the ‘correct’ reading of a text may vary for different interpretive communities. It is certainly true that reading conventions come and go. Nobody today except the Moral Majority ‘reads’ events as direct messages from God, as American Puritans once did, and before Freud nobody read texts as symptoms of Oedipus complexes. The question remains, however, whether some common thread may survive such changes in reading conventions, whether, in fact, we could even claim to identify novel conventions or lapsed ones if we could not draw on a metaconvention that stayed intact. Do not basic conventions such as making a statement, asking a question, ordering, requesting, promising, boasting, and myriad others, recur in all languages, as they would seem to have to in order to justify any claim to recognise them across cultural and temporal gaps? (Chomsky’s ideas about innate mental structures may be relevant to this question.) In any event, until we have looked closely enough at the dynamics of convention change to be sure there are no such continuities, the culturally-produced nature of interpretive norms is no argument for claiming that an interpretation could be right for one community and wrong for another. It’s not inconceivable that a Nazi and a Zionist would find many of the same meanings in Mein Kampf, even though they would react differently to them. Of course, a new way of looking can make it possible to discover things heretofore unnoticed. The deconstructors’ special interest in self-undoing effects allows them sometimes to see features in texts which previous readers had ignored. But it doesn’t follow that these features were not in some sense there before anyone inferred them – which is to say, recognisable by extension of the older extant conventions.

Of course how you deal with the problem of the stability of interpretation may depend finally on what kind of stability you are looking for. Hirsch, Abrams, Wayne Booth and other opponents of deconstruction have argued that interpretation is inherently unstable because of the difficulty of guessing with certainty what any person may mean or have meant on any occasion, an uncertainty deriving from the fallibility of the inferences we have to make and the often partial evidence with which we make them. But given a spectrum of degrees, these critics argue, we can be more confident of some interpretations than of others, especially – in the case of historical research – if we’ve done some homework. To call an interpretation ‘unstable’, then, makes sense only if one specifies: unstable as compared with what, unstable for what purposes. For certain purposes, I know well enough what my druggist means, just as for certain purposes readers know well enough what novelists and poets mean, at least much of the time. By the same token, to charge an interpretation with being ethnocentric makes sense only with reference to an implied comparison with some other interpretation which might be less so. The big mistake is to begin by assuming that the goal of interpretation is to secure unconditionally ‘stable’ readings, ones which are immune to possible correction or refutation. Once this assumption is made, it becomes easy to see how disappointment at the predictable failure of readers to reach that goal could lead to the conclusion that meaning is beset with some absolute radical instability.

None of this, however, will satisfy those who suspect even the most modest defences of validity, objectivity and rational adjudication of harbouring a concealed ideology of idealism, essentialism and logocentrism. Culler says little about the political aspect of deconstruction, yet it’s difficult to ignore the political animus suggested by attacks on ‘mastery’, ‘privileged positions’, conceptual hierarchies, and so forth. The real question for Derrida is finally not whether we can understand our druggist, but whether the medium of exchange by which the understanding takes place is not encoded with important presuppositions about economy and power, presuppositions which might be exposed by dislocating language. Derrida tirelessly searches for ways to ‘transgress’ the master v. slave hierarchy that he finds ‘inscribed’ in the binary oppositions of discourse. He might point, for example, to the figurative implications of ‘adjudication’, which presuppose the domination of the judge over the judged party, and he might point to the support such encodings lend to the dominant social order. To speak, therefore, of resolving disputes by recourse to a metalanguage or shared reality is merely mystification, an attempt to hold oneself innocent of politics and history in order to disguise one’s complicity. It can be said for this view of things that it generates inquiries into the ways power informs language that might not otherwise be undertaken, though it shouldn’t be forgotten that such inquiries had been undertaken from quite different standpoints – by Orwell, for example – long before deconstruction had been heard from. Since the deconstructors question the notions of truth and common ground by which critics like Orwell measured the injustices built into language, it remains unclear how their displacings of past and present systems is supposed to lead to something better. Nietzsche, the tutelary spirit of deconstruction, didn’t have this problem because he didn’t believe social justice was either desirable or possible. As Richard Rorty has pointed out in these pages,* to be a Nietzschean and to pursue social justice, without inconsistency, is no mean feat. No wonder the concept of ‘inconsistency’ has to be deconstructed.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN

Please include name, address, and a telephone number.

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences