Vol. 4 No. 3 · 18 February 1982

Deciding Derrida

David Hoy on the work (and play) of the French philosopher

4584 words

Of the essays collected and excellently translated in Dissemination, the best example of Derrida’s own practice of the deconstructive criticism he fathered is ‘Plato’s Pharmacy’.* Here he pursues his question why the metaphysical tradition from Plato to the present subordinates writing to speech. Derrida is not claiming to reverse Plato and to subordinate speech acts to écriture, intentions to texts. His suggestion is rather that the attempt throughout the history of philosophy to think about the relations of language, truth and reality is continually biased by the misguided oppositions between writing and speech, signifier and signified, the metaphorical and the literal, presence and absence, sense and intellect, nature and culture, or even male and female. For Derrida these dichotomies are set up not rationally, but with an implicit preference for one side or the other. His procedure for showing the prior exclusion of the other side is to study not the logic but the rhetoric used in such cases as Plato’s attack on writing, especially the metaphors and myths in the Phaedrus.

In particular, Derrida draws our attention to Plato’s frequent presentation of writing as a drug, a pharmakon. Words can be drugs, and like a drug the one word pharmakon can be taken to mean either cure or poison. Hemlock, for instance, poisons Socrates, but Socrates employs both rhetoric and metaphysics to argue that he is actually being cured. Similarly, in the Phaedrus’s mythical account of the origins of writing, the god Theuth invents writing and wants to give it to humanity as a gift. Theuth presents it as a pharmakon, meaning a beneficial ‘recipe’ for both memory and wisdom. However, the king who has the authority to accept the pharmakon hears the word differently and is suspicious of the addictive drug with its more probable narcotic effects.

The inventor of écriture is thus accused of smuggling drugs. The Czechoslovak authorities who arrested Derrida this January on the same charge may have been repeating the classic mistake of metaphysics by taking language as reality. Although having risked imprisonment in the name of the free discussion of philosophy certainly makes Derrida a political hero, it still will not make him a philosophical hero like Socrates. In fact, his play with puns and multiple meanings strikes his critics as the opposite of truth-seeking, and more like the eristics satirised in Plato’s Euthydemus. The attacks in the Academy on Derrida and de-construction run from the disdainful to the vehement, often implying the need to arrest such illicit activity. Are the charges against his writing legitimate, even if those against his person and political action in Prague were not?

Geoffrey Hartman’s Criticism in the Wilderness and now Saving the Text represent an extensive and largely sympathetic rumination on Derrida’s textual practices and their influence. That Derrida has had considerable effect, whether for good or ill, on recent literary criticism in English cannot be denied, although there has been relatively little written on him by English-speaking philosophers (Richard Rorty being a notable exception). Many may think the peak of Derrida’s influence is past, but Hartman is still justified in saying that Derrida will not be forgotten, if only because he will not be forgiven.

Exactly what is so unforgettable is difficult to say from Hartman’s discussions, however, and this very difficulty may well be what motivated him to devote so much attention to Derrida’s Glas, a continuation of the practice of dissemination. We do not learn much about the philosophical content of Glas from Hartman, but he is suggesting that despite a long exposition of some remote parts of Hegel, the book is not a work of philosophy. Similarly, despite the attention in the other half of the book to Genet, it is not a piece of literary criticism either. Taking his cue from the idiosyncratic presentation of these two seemingly unrelated halves of the book in separate columns on each page, Hartman prefers to think of it as breaking with any identifiable genre and becoming a new art-form in its own right. His subtitle spatially represents Derrida as being undecidably in between literature and philosophy. Treating Glas as an art-form allows Hartman to dwell on its stylistic complexities – its puns, solecisms and typographical oddities as well as its lack of a beginning, end and signature. To ask about the point of Glas, or its argument, would then be a mistake.

Of course, there is a point, for style and content are not as disjunct as the columns on Derrida’s pages. Derrida’s notion of writing explodes the very idea of the Book, whether philosophical or literary. Traditional interpretive assumptions, such as that a book is or ought to form a totality or whole, with an underlying unity preferably provided by an author’s explicit intention, are said to be vestiges of the ‘metaphysics of presence’ or ‘logo-centrism’ that led Plato to think of truth as the presence of a formal reality in the language and mind of a speaker. Hartman has long been concerned with literary examples of the desire for unmediated vision, and now with Derrida’s Glas he has a case of writing no longer based either on ‘the wish to put ourselves in an unmediated relation to whatever “really” is, to know something absolutely’ or on ‘a desire to be defined totally’.

Critics of Derrida’s earlier works implied that if his attempt to sustain arguments against the notions of truth, reference and reality succeeded, these works themselves could not have been written, or at least that if Derrida really believed what he was saying, his subsequent books would be very odd. Glas seems to have been written simply to say: ‘Yes, very odd indeed – in fact, whether it is even a “book” is precisely the question.’ Derrida could then indicate Hartman’s many essays on it as evidence that even such a text will have its readers.

Perhaps it is simply a matter of acquired taste whether readers will like Glas. which Hartman calls ‘a scandalous literary pudding, a French trifle’. He admits there will be many who find Derrida’s style ‘exhibitionistic’ and ‘cheapened’. He does not claim that Derrida is unique, but acknowledges Derrida’s debt to the ‘heavier Heidegger’. Also, the use of puns and other devices he grants may be ‘aleatory and overburdening’, ‘at best witty and at worst trivial’. Despite Derrida’s call for play and his Nietzschean objections to seriousness, his analyses often do not dance, but are ‘elegantly humourless’ and ‘may tire us into antagonism’. Hartman implies that Derrida is himself conscious of these objections, and is writing a ‘deceptive prose, modeled in part on Mallarmé, that evokes a radical equivocation’. Rather than deal with the question whether Derrida’s ‘punsterish, catachrestic style’ is good or bad, Hartman settles for the thought that while there is no such thing as a good pun, ‘puns are the only thing beyond good and evil.’

Hartman is hereby engaging in the same ‘task of equivocation’ he recognises to be resisted by the Anglo-American tradition. Hartman appears to be reflecting on his own, equally punsterish style when he suggests that these likely objections are more serious for the interpreter who is influenced by the writer being interpreted: ‘For the very fullness of an interpreter’s style, if richly allusive, or speculative, or contaminated by the writer under discussion, leaves us with a hollow feeling.’ Yet Hartman is much better at Derrida’s stylistic game than Derrida himself, for Hartman is always witty and never tiresome. Hints could be explored further, suggestions argued and justified, but it is impossible to feel antagonism.

I hope, however, that Hartman overstates his case in claiming that while Derrida’s writing may seem extreme now, in another decade it could be the norm. His prediction is based on the assumption that there is a continuity between the earlier, more theoretical writings like Of Grammatology and the increasing oddities of the more recent works, such that Derrida’s impact is due to the earlier and later works equally. If, however, Derrida’s influence, like Heidegger’s, stems from the point of view (and not necessarily the style) of the earlier works like Grammatology rather than from the less theoretical writings like Glas, there is no reason to think subsequent literary critics could not state more clearly and put into practice better what Derrida started.

Although the three books Derrida published in 1967 (Of Grammatology, Speech and Phenomena and Writing and Difference) formulate what Hartman aptly labels the ‘endgame vision’, Hartman takes the three books Derrida published in 1972 (Dissemination, Positions and Marges de la Philosophie) and Glas (1974) as demonstrating how strongly his ‘will to write’ persists. Unfortunately, our will to read subsides. Derrida himself begins ‘Plato’s Pharmacy’ (published in Tel Quel in 1968) with an apparent allusion to his earlier writings: ‘To a considerable degree, we have already said all we meant to say ... Since we have already said everything, the reader must bear with us if we continue on awhile.’ In fact, there is so much indirectness in these later essays, and Derrida goes on for such a long while, that the vision is itself disseminated and dispersed. The weight of allusion is so great that Jonathan Culler remarks, in Structuralism and Since, that Dissemination is ‘Derrida’s most forbidding and difficult book’.

If the reader is familiar with Derrida’s earlier work, though, there are enough cues in Dissemination to see what is going on. First, there is a narrative line with Plato at the beginning and Mallarmé at the end. This is the story of metaphysics exhausting its own metaphors, degenerating into manipulatioin of an empty vocabulary. Second, the specific aspect of this history Derrida investigates is why the very concept of writing is so marginal to Plato but so central to Mallarmé. Third, in addition to the general problem of the reference of language to reality, Derrida also analyses their conceptions of aesthetic representation, suggesting a change in the notion of mimesis between Plato and Mallarmé. Finally, the link between the aesthetic and the logical representation of reality is seen in Plato’s use of rhetoric in drawing the metaphysical dualisms and deciding the values of the poles, much as the English translator must decide to render pharmakon as either remedy or poison.

Derrida’s Mallarméan thesis is that the metaphysician and the translator are in a similar predicament because they seem forced to choose one side or the other of the dyad when such oppositions are undecidable. The thought is that the translator is not inaccurate in choosing ‘remedy’ in one context and ‘poison’ in another as the proper translation of pharmakon, but that nevertheless the translation can ‘neither be accepted nor simply rejected’. The translation fails to capture the ambiguity of the Greek word, its very ability to be taken either way. Only a translation in which we could still hear the other possibility would do justice to the original text, and rarely if ever do different languages have corresponding words with the same ‘chain of significations’ or range of connotations.

This familiar point is not strong enough, however, to support Derrida’s conclusion that the failure of the translation to capture the ambiguity makes an ‘understanding of the context’ probably ‘impossible’ and the reference undecidable. Clearly the translator can understand the context well enough to choose the appropriate modern term accurately, even with some loss of ambiguity that can be corrected only with a successful paraphrase like Derrida’s own. Without further arguments for the indeterminacy of translation Derrida’s theory of textuality appears ungrounded, since what he takes to be a fundamental paradox is not really so difficult to understand.

Derrida’s intention is often more straight-forward than his language makes it seem. He wants to show that the very idea of authorial intention is less useful as an interpretive guide than the associated meanings with which the words in a text are vested. Since pharmakon is caught up in a chain of significations, it may connote more than the author may have or even could have intended (although in this case Plato seems to have been quite in control of the word). When dealing with linguistic relations, interpreters are to avoid all recourse to ‘crude tools’ like the conscious/unconscious, voluntary/involuntary distinction. Derrida himself seems to slip back into metaphysics or at least mystification, however, when he claims that by concentrating on the language of the text alone he can uncover the ‘deeply buried necessity’ and ‘powerful constraints’ that organise the internal structure of Plato’s text quite independently of Plato’s intention, and even undercutting that intention.

Deconstruction is thus the investigation not of authorial intention but of textual dissemination. ‘Textuality being constituted by differences and by differences from differences,’ writes Derrida characteristically, ‘it is by nature absolutely heterogeneous and is constantly composing with the forces that tend to annihilate it.’ The crux of disseminative interpretation is the notion of undecidability, which the Tel Quel group culls from Gödel’s proof. Unlike Julia Kristeva, though, Derrida at least admits that there is only some analogy – a rather forced one – between the undecidability of propositions in a formalised axiomatic system and the indefiniteness of rhetorical ambiguities and textual allusions in an author like Mallarmé (or Derrida himself). Derrida’s own propensity to take as undecidables single words that have opposed meanings seems much closer to Freud than to Gödel. Although Derrida does not refer to Freud’s essay on primal words in this connection, Freud was also interested in examples from ancient languages of single words with antithetical meanings (e.g. strong and weak, or inside and outside).

If undecidability is to be a technical term in interpretive practice, it is not to be confused with either indecisiveness or indeterminacy. Examples like pharmakon in Plato, ‘supplement’ in Rousseau, and ‘hymen’ in Mallarmé are not to be simply vague. The inability to decide is not to be the result of the interpreter’s indecisiveness, but of the necessity of seeing both senses in the text, even if the text blinds itself to one of them. Derrida claims these words ‘have a double, contradictory, undecidable value that always derives from their syntax’. Deconstruction is limited to exploring the supposedly infinite interplay of syntactical connections and freed from determining their reference and truth-value.

Like Nietzsche, Derrida tends to overstate the case and imply that the nature of textuality is such that there is in texts no reference and truth-value at all. The slogan in Grammatology is that ’il n’ y a pas de hors-texte’ (‘there is nothing outside the text’), and in Dissemination that there is no ‘extratext’, nothing before the text, no ‘pretext’ that is not already a text. To be consistent Derrida therefore writes a preface (‘Hors Livre’) that disclaims being a preface, and is in fact an entertaining discussion of famous prefaces in literature and philosophy. Here he modifies his earlier claim by saying that if there is no outside to a text, there is no inside either, the consequence being that there is no text-in-itself, and no purely intrinsic interpretation. The text will always affirm and refer to an outside, Derrida now admits, but that outside is nothing else than just another text. Textuality implies reference, not to an external reality, but rather to other texts, to intertextuality. ‘We have fallen into the condition of viewing all things as texts,’ says Hartman. Or, as Barbara Johnson puts it in her resourceful introduction, ‘nothing, indeed, can be said to be not a text.’

At that an earlier Dr Johnson might again have kicked a stone, and G.E. Moore would perhaps have challenged Derrida to hold up his own hand and say: ‘Here’s a text!’ Derrida, however, rightly wants to avoid the ideal ism-realism controversy. His point may be rather that philosophers like Plato converted statements in ordinary usage into what Derrida dubs ‘philosophemes’, philosophical terms of art formed into beliefs that only seem to be determinably true or false. The lack of translatability in Plato’s text is attributed further to the transformation of a non-philosopheme into a philosopheme, to ‘the problem of the very passage into philosophy’.

Derrida thus fits into a philosophical tradition as old, perhaps, as philosophy itself. Overcoming metaphysics and bringing philosophy to an end is a recognisable genre now practised by showing how past philosophemes deconstruct themselves. The philosophical text is found to ‘double’ itself: not simply bipartite but actually duplicitous, the text contains, even if only by excluding, an ‘other message’ that undercuts its cognitive claims. While Wittgenstein describes his similar aims as straightforwardly therapeutic, Derrida’s double-edged intents are more ambiguously pharmaceutic. Heidegger also practised a form of Abbau on Plato’s (and Carnap’s) Aufbau, but Derrida thinks Heidegger’s term Sein is yet another philosopheme that succumbs once again to the illusory belief in an ultimate ground, in a ‘transcendental signified’. So Derrida tries a move he attributes to Mallarmé. On Derrida’s account of mimesis in Plato and Mallarmé, Plato gives mimesis a metaphysical interpretation, ‘which implies that somewhere the being of something that is, is being imitated.’ Mallarmé, in contrast, adopts the surprising tactic of rejecting the metaphysical interpretation, but not mimesis. His art is to arrange words and sentences syntactically, apparently without regard for, but relying parasitically on, their semantic intelligibility, since the reader will inevitably construe meanings. Derrida describes Mallarmé’s poetry as this ‘mimicry imitating nothing’, ‘this speculum with no reality’, ‘this mirror of a mirror’, ‘a reference without a referent’. The same is probably to be said of Derrida’s own ‘philosophical’ remarks, such as ‘there is no simple reference.’

Given Derrida’s attempt to be for philosophy what Mallarmé is for poetry, he could not object to the judgment either that he was not a philosopher or that he had not shown there could be no philosophers. Furthermore, no one should look to him for a philosophical grounding of the practice of textual interpretation, since that would miss his point that there can be no philosophical grounding of anything. Barbara Johnson suggests that Derrida is creating a new, ‘non-binary logic’, and is generating ‘nothing less than a revolution in the very logic of meaning’ from his discovery of undecidables, terms that say ‘Neither/nor, that is, simultaneously either /or’.

Since a paraphrase can explain unparadoxically why and how such a term can be used unambiguously, it is stretching the point to imply that we need a new logic. His project is better construed as the less serious and more amusing one of discovering in classic places like Plato’s attack on writing the same sort of ‘kettle-logic’ Freud noticed in dreams: ‘1. The kettle I am returning to you is brand new. 2. The holes were already in it when you lent it to me. 3. You never lent me a kettle, anyway.’ Derrida formulates Plato’s various attitudes toward writing analogously: ‘1. Writing is rigorously exterior and inferior to living memory and speech, which are therefore undamaged by it. 2. Writing is harmful to them because it puts them to sleep and infects their very life which would otherwise remain intact. 3. Anyway, ... because living memory is finite, it already has holes in it before writing ever comes to leave its traces.’

It is not hard to imagine schematising Derrida’s own cheerful thinking in turn. Indeed, he should welcome the following playful sketch as long as it is offered with curative and not poisonous intent. 1. There is nothing outside the text, and the text is impervious to attempts to get out. 2. All attempts to stay within the text fail (since there is no inside, no unity to a text), and those thematic readings that look for a didactic message distort the text. 3. Since reference to reality is impossible, there aren’t really any texts anyway.

Dissemination could thus be a form of dreaming in which the existence of genuine practitioners of deconstruction could be doubted. Whereas literary critics like Hart-man may be concerned with Derrida’s solecisms, philosophers will suspect him of solipsism. Looking at the practice rather than the theory of disseminative reading, what does distinguish it from other forms of practical criticism? Hartman himself is not necessarily practising deconstruction as such in his reading of Glas. In fact, his interpretation is more of a reconstruction, for it pursues Derrida’s verbal plays and allusions in order to make sense of them and to generate a theory of literary writing and reading. He generalises Derrida’s style with the thought that ‘style may be a continued solecism,’ with no end or ‘closure’ to the stream of words. What holds for writing holds for reading as well: ‘The more pressure we put on a text, in order to interpret or decode it,’ says Hartman in Criticism in the Wilderness, ‘the more indeterminacy appears’; and in Saving the Text he reiterates that in penetrating a text, ‘what one comes to know is the unintelligibility ... of the literary work.’

Pressure on Hartman’s own text yields, however, a quite intelligible view about literature and the practice of reading it. He agrees with Derrida that intention is never simple, that ‘we don’t always know what we mean.’ His thesis that in writing ‘partial knowledge is the normal condition’ is less strong, though, than Derrida’s Mallarméan conception of literature and philosophy as play, ‘an adventurous excess of a writing that is no longer directed by any knowledge’. Hartman’s notion that literature both builds and at the same time demolishes an effet de réalité is compatible with Derrida’s critique of presence. Yet Hartman implies that Derrida’s theory does not allow for the pathos or power of literature. Purely semiotic analysis fails to capture the ability of words to wound, as he puts it in one chapter. Although Hartman prefers to think of reading as dissemination rather than proclamation, he wants to restore to literature, if not truth, at least ‘troth’: ‘To call a text literary is to trust that it will make sense eventually, even though its quality of reference may be complex, disturbed, unclear.’

This restoration ought to be appreciated by literary theorists, but Derrida himself would probably accuse it of retaining a logocentric desire for unmediated vision. Trusting that not only literary but also philosophical texts will make sense eventually is for Derrida characteristic, not of dissemination, but of what he names hermeneutics. Rather than bracket the referential function of language indefinitely, as dissemination does, the hermeneutical reader tries to make sense of the text by integrating its disparate elements in a unifying framework. In contrast, a disseminative reading is a potentially infinite proliferation of possible meanings.

Derrida’s own texts deliberately frustrate the hermeneutic reader, as if to underwrite the point that the choice between the playful, disseminative strategy and the integrative, interpretive one is an either/or. To break the grip of the vestigial metaphysical desire to comprehend totally and absolutely, Derrida turns into paradigms texts that seem unreadable to the interpreter searching for certainty. Through elaborate and myopic analyses of obscure, marginal material he delights in discomposing us and our normal contexts for understanding texts. He sometimes suggests that dissemination works by treating sentences, whole works, and even an author’s corpus, as if they were in quotation-marks, mentioned but not used, and thereby removed from any determinate context. Generalised to all texts, including Derrida’s own, dissemination is intended, not to complement trustful readings, but to show the failure of hermeneutics to comprehend the text and close the circle.

Those like myself who think there must be more to interpretation than pure dissemination can insist that deconstruction presupposes a prior construction of the text’s unity or sense. Even if the disseminative reading then succeeds in disrupting this understanding, the fact that there are difficult interpretive decisions in practice does not entail undecidability or the impossibility of understanding in principle. Although any particular context for making a text intelligible can be called into question, it does not follow that no context is justifiable. Giving up the ideal of determining the one and only appropriate context by appeal to a decisive criterion like the author’s intention should not result in making undecidability the new ideal of all writing.

Dissemination and hermeneutics need not be contrasted so extremely. They are more plausibly seen not as irreconcilable theories of meaning, but as practical interpretive strategies, as facets of any good reading. So regarded, disseminative practice ensures that the text’s complexity is not underestimated, while the hermeneutical sense-making activity keeps the dissemination from wandering off infinitely. The search for sense in the text will not bring the reading to a premature halt, for complexity and sense are rarely in equipoise. The reflective equilibrium aimed at in one reading is likely to become unbalanced in the next, and re-established only by other means. What would stop the will to read, especially in the case of philosophy, is the discovery of the text’s final undecidability or unintelligibility. If I understand Hartman correctly, he is right to think that reading will continue only so long as it trusts that the failure to make sense is its own and not the text’s.

Over and above this context of reading strategies, however, there is another issue that should not be conceded to Derrida without argument. His attempt to be the Mallarmé of philosophy is not the only way to avoid falling back into Platonism. Relinquishing a notion of knowledge as direct contact of mind with reality does not entail either giving up on knowledge or being left merely with a doctrine of linguistic immanence. Literature, philosophy and literary criticism itself are legitimate forms, among others, of knowledge and understanding. There is no reason to think that with plausible evidence their central statements could not be understood and decided.

Although a hermeneutical, sense-making reading of Derrida’s essays is indeed difficult, deciding where he stands is not impossible: he is an anti-realist and post – empiricist who is less interested in advancing a positive doctrine than in debunking the metaphysical strains he finds not only in Plato but also in moderns like de Saussure and Lévi-Strauss. He identifies more with figures like Mallarmé and Nietzsche who are like him in two respects: they break with traditional ways of writing, and they are often misinterpreted as idealists. Derrida’s critics in the Academy may grow impatient with his style, but they should see that his indirectness results from the difficulty of deconstructing without also constructing.

Without the possibility of understanding Derrida and deciding the merits of his statements, readers would in fact quickly lose interest in his persistent will to write, not poetry, but something approaching philosophy. He introduces terms like ‘differance’, ‘trace’ and ‘undecidability’, yet then with tongue-in-cheek consistency implies that what these quasi-philosophemes mean may in turn be undecidable. Should we look forward to an era where undecidability is the norm in philosophy? Derrida may think undecidability has always been the norm, but deliberately constructing a really undecidable philosophy is incoherent. The irony of his position is that if the genre and sense of his writing could not be understood and decided, we would have another alternative than to infer that he had created a genre-breaking art-form or a new style of reason. We could decide his texts were neither literature nor philosophy, nor anything else.

Send Letters To:

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


Please include name, address, and a telephone number.


Vol. 4 No. 6 · 1 April 1982

SIR: It may interest readers of the review of Dissemination by Professor David Hoy (LRB, 18 February) to know that another work of Derrida mentioned in his review, Positions, first published in France in 1972, is available in an English translation published by the Athlone Press in 1981.

Brian Southam
Managing Director, Athlone Press, London WC1

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