What today do we make of these issues, so central, so provocative, even galvanizing in Derrida’s early texts: significations and codes, speech and writing, citation and iterability? The temptation is to respond: not much. It’s been a long time since most literary critics or even theorists were moved by concerns pertaining to language and discourse. No doubt, this is with some right—these issues became “played out,” as some say, with seemingly little new or urgent left to be said about them, or so it seemed at the time.1

These topics appear worth revisiting now, however, in part since what remains of these questions, the fallback stances taken by many critics when pressed about them, tend to be some version of Derrida’s appraisals, especially those that Derrida set out in “Signature Event Context” (SEC).2 SEC, if nothing else, provides a set of handy takeaways, shorthand theses, upon which not only critics but Derrida himself subsequently relied.

In SEC, to put it in the parlance, Derrida’s early approach to matters textual and linguistic became “sedimented.” Sedimentation is a notion borrowed from Edmund Husserl. It designates a circumstance where a discourse and its subject matter no longer are engaged from the ground up, in respect both to the total chain of reasoning standing behind the conclusions pertaining to them, as well as the topic’s most foundational aspects and appearances. In physics, for example, no one now examines the requirement to describe motion in mathematical formalisms; that requirement has become sedimented.

In SEC, Derrida’s own thinking about language and discourse, I suggest, suffers the same fate. Situated apart from his reflections on the continental philosophical tradition, treating a philosophical school with which he is less familiar, and forced to present his results in a fashion unusually positive, the deepest roots of Derrida’s own thinking fall from view. The compression of more involved and nuanced arguments and the obscuring of the underlying matters they concern become still more acute when successors adopt Derrida’s conclusions.3

The impact of its sedimented views is not the sole reason to return to SEC. SEC gives expression to a conviction, almost an intuition, common among those who work in literary studies, concerning textuality. SEC’s neologisms have entered into the lexicon of many critics and theorists usually because they buttress and ramify a seemingly implacable practical belief pertaining to what texts are and how they should be approached.

A hoary but still relevant example from the beginning of Hegel’s Phenomenology, with which I will begin, lets this common, powerful conviction be made explicit, along with some of the questions it leaves unanswered.4 That discussion introduces my examination of Derrida’s own approach to writing and the text in part one of SEC. Derrida, to be clear, does not himself, of course, fail to reflect on what many critics merely assume; taking his earlier findings as results, his notions of iterability and “signifying form”—the latter in SEC almost entirely unexplicated—nevertheless encourage Derrida to skip some of these issues’ broader and deeper contexts. Austin’s own program, with its very different starting points for approaching speech and writing, helps highlight the partially unacknowledged ground on which Derrida’s conceptualizations move. At the same time, Derrida’s criticism of Austin in part two of SEC—namely, that Austin cannot account for literary and other sorts of supposedly non-serious speech—is also, to my mind, cogent. Hence, after first raising some questions about Derrida’s depictions of writing and textuality, in the second part of the present essay I lay out a new view of literature and literary criticism, spurred by Derrida’s concerns, which ultimately takes on board aspects of both his and Austin’s approaches.

1. Ideality and Iterability

In the first chapter of his Phenomenology of Spirit, Hegel writes down the statement “now is night,” offered in response to the question “what is now?” Hegel’s aim is to debunk the claim of “sense-certainty,” the first entry in the Phenomenology’s survey of candidates for absolute truth. To delegitimate sense-certainty’s insistence that truth is equivalent to sensuous immediacy, Hegel underscores the legibility of his inscription and what it states. “Now is night” speaks and lays claim to being true even when it is no longer night, when it is day. Hegel thereby exhibits the gap between what sense-certainty identifies as its truth-maker (what is now) and the expressions and concepts it employs.5

The ramifications of Hegel’s demonstration for his own thought cannot be further entered on here. “Now is night,” the appearance it yields, does, however, exemplify what still seems true about writing and language to many critics. What is written down, as Hegel points out, and they would concur, has the ability to stand untethered from all circumstances, apart from any “other-being [Anderssein],” as Hegel puts it. Once inscribed, the statement operates on its own, entirely independently of any context. Accordingly, language, meaning, and texts operate at a distinct remove from all “sensuous being”—technically here they are its negation—and from what in other contexts would be called the real. The inscription “now is night” floats free of all such reference points, while itself remaining “preserved [aufbewahrt],” as Hegel says.6

The question arises, however, of whether what is set down in writing genuinely is capable of such independence, whether written statements or anything else can exist and function without regard to a temporal and spatial situation, as in Hegel’s example.7 Are there, for instance, universals or concepts of some sort that find their way into and somehow underpin language and discourse? Do writing (and language) express, and may themselves even be, instances of the ideal? Ideal entities are presumed not to be spatio-temporally individuated, in contrast to those that are, the real.8 “2+2=4” (or, for that matter, “2+2=5”), on this view, is said to express the same, single, self-identical sense to all who come across it, regardless of circumstances. Its meaning is one, while its apprehensions are many. Were this not so, no understanding or discussion of what it says would be possible, it is maintained.9 Similarly with texts, were its expressions and what they express not ideal, we could only talk about similar texts, not the same ones—many Moby Dick’s, not a single one—and use only similar words, with no guarantee that their meaning is or can be shared.

Most critics, to be sure, do not place themselves in the idealist camp; they appeal to language, to signifiers, and to related reference points. Nevertheless, even these accounts invoke or imply the ideal to a much greater degree than is often believed. Appeals to linguistic structure, to words, to signs all conjure on the ideal.

Of course, the physical inscription, the fact of being set down in ink on paper appears to give a material basis for how statements can last and mean in the manner Hegel describes. Ink or pixels, however, cannot yield what we call a text, no less its meaning and signification, when taken in their own sensuous immediacy. This is because physical signifiers, or so-called tokens, are not able, on their own terms, to form identifiable classes that might permit signification to occur on their basis. Some actual signifier, a so-called token—one of the “the’s” printed on this page, for example—may indeed be something real, pixels or ink; accordingly, the hope would be that by recognizing this instance as like other physical instances, what each signifies (their self-same meaning or signified or concept) also can appear. Being one of a class, thanks to a physical likeness, a token could convey what is not simply physical, its meaning or sense, without any need to appeal to the ideal.

Nothing, however, warrants believing that a token conceived on a simply material or physical basis belongs to a language or to the same language (e.g., “four” is found in both French and English). Nor is it obvious that tokens so conceived necessarily belong to any language at all—otherwise an ant tracing a line in the sand might be referring to itself in English, perhaps beginning to perform the cogito. Moreover, being phonic and graphic and gestural, signifiers don’t share any genuine physical resemblances. This remains true, though less evident, even within a single signifying medium. Let Stalk Strine is the title of a guide to Australian pronunciation. Individual physical inscriptions and utterances comprise neither a recognizable nor a relevant class when taken in their own right.10 Accordingly, pixels, ink, and so forth cannot allow signification to come to pass; they do not permit conveying expressions independently of the ideal as many believe.11

Some of those who hold these views certainly think that their stances proceed from Derrida’s writings. His earliest accounts, however, are far subtler than the foregoing. For example, in Of Grammatology, when glossing Saussure’s claim that the signifier is a “psychic image,” Derrida insists, against Saussure, that the signifier be understood in terms of Husserl’s phenomenological epoché, which by definition involves an intentional component. Signifiers can never be studied by a “mundane science,” Derrida declares, breaking with all linguisticism and also from any straightforward or reductive materialism.12

Neither in Of Grammatology nor in SEC does Derrida endorse ideality, of course, not even in the guise of affirming language’s or the mundane signifier’s role, as do many of his followers. Derrida does not underwrite ideality; he does, however, deconstruct it. That or any deconstruction’s workings, moreover, entail that, while not subscribing to ideality, Derrida does not simply wholly reject it, as do I or many in the analytic tradition. Rather, deconstruction leaves a certain amount of running room or leeway for ideality and its achievements. Deconstruction as a practice, broadly, lets “philosophemes” such as ideality be affirmed within their own domain and limits, while arguing for these notions’ inclusion in still greater, more comprehensive wholes that contest the values, ends, and origins that purportedly give birth to and that guide these same conceptions—above all, the values of presence and self-presence. In part one of SEC, accordingly, Derrida clearly attempts to go ideality one better: he wishes to unbind its scope—to broaden the array of instances the repeatability it seemingly guarantees covers—while offering an alternative construal of ideality’s workings, beyond any appeal to meaning, to consciousness, or to context, thereby putting into doubt ideality’s philosophical framing.13

Derrida’s strategic retention of ideality shows up in SEC’s leading notion of iterability; it becomes further evident and undergoes a new twist at the end of part one of SEC in an instance of grafting that Derrida offers by treating an example from Husserl’s Logical Investigations. With the latter, the central problem of SEC, by mine and also by Derrida’s lights, emerges: the status of context (SEC, 2).

a) Iterability

Derrida’s overall strategy in respect to iterability falls out of his interrogation of a discussion of the origin of language and writing by the Abbé de Condillac. Condillac effectively establishes the still current understanding of “communication,” according to Derrida, precisely in terms of ideality. “Communication,” in Condillac’s wake, consists in “circulat(ing) a representation as an ideal content (meaning).” “Writing,” Derrida adds, is considered “a species of this general communication” (SEC, 6).

To arrive at iterability, then, Derrida undoes what Condillac has wrought: namely, the subordination of writing to communication more broadly understood, including this ideal component. Derrida denies writing’s determination as “a species of … general communication” organized around the ideal.14 Simultaneously, Derrida conceives of writing—and, with it, ideality or ideality’s effects—on new terms. Viewing writing as intrinsically tied to absence, Derrida frames a notion of iterability more encompassing than ideality, wherein writing stands as the genus of which Condillac’s communication emerges as species.

Derrida’s reversal (of species and genus) and reinscription (of writing) accords with a description of deconstruction that Derrida repeatedly offers and explicitly recurs to later in SEC.15 Here this operation pivots on writing’s purported ability to function in the radical absence of a receiver (and eventually of a sender). Writing entails a “certain absoluteness of absence” (“un certain absolu de l’absence”), as Derrida puts it, that Condillac ignores (SEC, 7; Marges, 374).

Derrida’s demonstration of writing’s absence’s absoluteness (a depiction that may conjure up Hegel’s “now is night”) initially depends on the notion of code. Following “the death of the receiver, or even of both partners, the mark left by one of them is still writing,” Derrida affirms, and then adds, only “to the extent that, organized by a code, even an unknown and nonlinguistic one, it is constituted in its identity as mark by its iterability” (SEC, 7).

Writing’s alliance with an absolute absence, its operation in the absence of a sender or receiver, thus corresponds to its being “organized by a code,” even if that code is also unknown. Neither the inscription’s physical disposition, nor those “causal” powers some attribute to Derridean écriture, at least at this moment, account for this capacity. Instead, the marks, writing, function thanks to their appurtenance to a code, a code which, Derrida will further tell us, though perhaps available, can never be “structurally secret” (SEC, 8).

Derrida’s argument at this moment is novel in the context of his own work, and it also may finally not be wholly persuasive, including in SEC’s context, as Derrida will soon distance himself from code. Code’s stamp on writing and iterability for Derrida, nevertheless, at this juncture gives writing and iterability an unparalleled scope or reach, one crafted to include, while undermining, ideality and its functioning.

More specifically, inscriptions and all relevant marks on the foregoing basis are conceived by Derrida as “communicable, transmittable, decipherable … for every possible user in general” (SEC, 8); similarly, these inscriptions are said to continue to operate “in the radical absence of every empirically determined receiver in general” (SEC, 8).16

Derrida’s somewhat unusual use of both “every” and “in general” attests to the breadth that Derrida assigns to code’s potential functioning. Ideality’s deconstruction requires such an affirmation, however. Iterability must outstrip the ideal, while accounting for all of ideality’s effects. Iterability must make possible ideality’s allowing the same understanding by all (“every possible user in general”), even as it contests the ideal as such, thanks to its commerce with an absolute and fundamental absence.

By my lights, however, Derrida at this moment grants too much to ideality to begin with, even as he aims to contest its significance and the terms of its functioning. Languages and codes are themselves, after all, all finite, empirical; they come into and go out of existence. Hence, communications employing languages, or depending on codes that have been invented (whether or not known), can never even in principle pertain to “every empirical receiver” and operate for “all possible users in general.”

Many codes—signals, for instance, or book codes—are not so decipherable and may thus remain permanently, or “structurally,” secret. For example, a spy may agree with their mole to turn a red light on in their house every Friday to signal a meeting on the following Tuesday. If spy and mole both die on an intervening Sunday, even if the red light happens to be on, there is no reason to suppose that their code is still recognizable as a code by anyone.

Similarly, a code may depend on an arrangement of books in a shop window; seeing it no one not apprised of the code will know it is an instance of a code at all. Still more gravely, witness the huge number of indigenous languages that have and are everywhere disappearing. These languages’ “decipherability” ceases in the absence of certain empirical receivers and users, their native speakers. Nor does their writing when similarly situated always betray its having once proceeded from a code or language. Whether certain inscriptions are or are not instances of what is sometimes called “asemic writing” can remain unknown, as can whether entire manuscripts are products of languages or code at all, as with the so-called “Voynich manuscript.”17

Derrida’s positioning in SEC, then, arguably initially gives too much fealty to the ideal—to Hegel’s “preservation” and the ability for inscriptions to be available for every empirical receiver and to all possible users in general. So proceeding, Derrida misprizes the temporal character of codes, languages, and discourse themselves.

Derrida’s affirmations concerning code’s scope become arguably still more questionable as his discussion unfolds. Derrida soon introduces the notion of “grafting,” which he defines as “a written syntagma[’s]” capacity to be “detached from the chain in which it is inserted” (SEC, 9). As subject to grafting, writing so understood, he tells us, is capable of breaking with “all possibility of communicating” (if not “all possibility of functioning”). On this basis, Derrida, accordingly, emphasizes what he calls a mark’s “force of rupture” and its “essential drift” (SEC, 9).

At this moment, however, marks and writing, now distanced from “any possibility of “communicating” seemingly are no longer decipherable by anyone, no less everyone. They have ceased being decipherable at all. To be sure, Derrida’s idea is that such grafting takes place after writing and the mark have been constituted thanks to reference to a code. Hence grafting requires now viewing code as “the possibility and impossibility of writing, of its essential iterability” (SEC, 9). Writing’s capacity to be repeated, its very identity, however, purportedly recurred to its being “organized by a code.” If that identity now turns out to be owed neither to its physical shape nor to any actual code (since the latter is also its condition of impossibility), nothing seemingly any longer lets the mark be a mark, writing be writing, rather than a random gash or an ant’s pathway. With code and its decipherability out of the picture, the grafted mark’s identity completely disappears as well.

Of course, this may in part be Derrida’s point, since marks and inscriptions otherwise would not function beyond meaning; they could not be affirmed to exceed “every horizon of semio-linguistic communication,” as Derrida will soon put it (SEC, 12). Code and iterability, accordingly, must operate with this duality—decipherable for all possible users while also at some point not being decipherable for anyone—to fulfill Derrida’s aim to encompass and undermine ideality at once.

b) Signifying form and context

The foregoing considerations, in the end, may have worried Derrida himself since he next dispenses with code. Developing further the possibility of “the graft” or “grafting,” Derrida confesses: “I prefer not to become too involved here with this concept of code which does not seem very reliable to me.” Instead, “we must be able to recognize the identity, roughly speaking, of a signifying form” (SEC, 10).

Apart from its recognizability through all “empirical variations,” as he puts it—again echoing the ideal—what signifying form may otherwise be remains in the dark in SEC, nor, clearly, does it cover non-linguistic cases as code did. And should signifying form itself somehow recur to languages or a language, the ideal, as found in linguistic structure, would not be avoidable, not to mention why we would need iterability at all, language now turning up as the latter’s ultimate reference point.

Nevertheless, despite the obscurity attendant upon signifying form, an important development follows from it in which a decisive new aspect of the mark’s functioning comes on the scene: its relation to context. On the basis of signifying form, Derrida pursues the notion of the graft as it relates to context, ultimately by way of an example taken from Husserl’s Logical Investigations. Context’s treatment raises rather different problems than those previously discussed.

More specifically, after affirming that the iterable “mark” can appear across an “infinity of contexts”—the appeal to a positive “infinity” being yet another trace of ideality’s erstwhile operation—Derrida now adds that this “does not imply that the mark is valid outside of a context, but on the contrary that there are only contexts without any center or absolute anchoring [ancrage]” (SEC, 12). No mark, Derrida now indicates, is valid outside of a context, no message or inscription functions apart from one. Earlier, however, seemingly to the contrary, Derrida, speaking of the mark’s “force of rupture,” declared that “no context can entirely enclose [the mark]” (SEC, 9). Hence Derrida in SEC seems to affirm both that the mark breaks with all context and that it is always found in one. From this perhaps also results the belief that the effects of marks are somehow causal.

Whether or not two contradictory characterizations of the mark on Derrida’s part appear in SEC, in the present context, his present account of context raises a still more fundamental issue. On this last construal, after all, Derrida’s stipulation that the mark lacks absolute anchorage is finally irrelevant and inconsequential. Once context is deemed necessarily at work in every instance, once it is granted that a communication—an utterance or inscription of some sort—always operates in some context, as Derrida has just affirmed, how can what might happen in some other context matter and what weight can talk of anchorage, absolute or not, have? What happens, or could happen elsewhere, on another occasion has no bearing once the importance of its context for a mark or writing is granted, once it is granted that no marks are ever valid “outside” some context. That a waitress, after her shift, might say “my dogs are tired” and so also might, on another occasion, the owner of some aging Weimaraners—these clearly have no effect, no force in respect to one another; something all the more true, if in some contexts some marks cease meaning anything at all, as Derrida also indicates.

Derrida’s own invocation of context in fact ultimately shows why context’s role cannot be maintained and the universality that ideality bequeaths to iterability sustained. If context is decisive—as it must be once physical features, codes, and unexplained, stable (verbal) meanings are set aside—then speaking of an unbounded, infinite repetition of some writing or mark seems both impossible (since what is being repeated can no longer be identified across these instances) or relevant (since within these series of valid contexts no one event bears on any of the others—except of course in yet another context which brings them together and makes the case for them being so related). Accordingly, by the present lights, Derrida’s appeal to radical contextualization must be embraced, while dispensing with what seems but a pure, empty repeatability, bordering on the ideal, a repeatability that stands at a distance from all operative writing and discourse.

The alternative here beginning to be proposed, as well as what Derrida himself concretely has in mind, can both be better grasped thanks to the example of grafting that Derrida in the light of the foregoing offers from Husserl. At issue in Husserl’s text is how specifically logical, and, for him, necessarily ideal, meanings, Bedeutungen, function in the propositions that they comprise. Derrida departs from Husserl’s treatment of such Bedeutungen by focusing on an instance of what Husserl calls “agrammaticality” and claiming that this instance is still subject to that repurposing that we have seen Derrida call grafting.18 For Husserl, by contrast, agrammatical employments entail a complete loss of sense such that their character as meaningful and their corresponding ideality and/or repeatability wholly cease to function. Husserl’s example of this kind of foreclosed and shut down proposition is “the green is or.” In this expression, a so-called syncategoramatic term, or logical operator, “or,” appears in the wrong grammatical slot. Thanks to this, for Husserl, “or” itself, as well as the rest of the proposition’s components, say nothing at all and necessarily fall short of the ideal’s, or any, mode of being (SEC, 12). They fail to express any thought in Gottlob Frege’s sense.19

Derrida insists, however, that even in this instance where all standard semantic capacity seemingly fails, and an expression and its signifying form on the face of it lack any plausible “vouloir-dire” (“meaning,” or “wanting to say”), the marks in question may still operate. For Derrida, Husserl’s example remains viable, albeit as agrammatical, strictly speaking meaning nothing at all.

More specifically, Derrida points out, first, that in French “the green is or” retains significance, owing to the ambiguity of “ou” “or,” which in French can also mean “where,” yielding “the green is where.” Derrida further claims that Husserl’s phrase may continue to function, though not mean, by being used as an example of agrammaticality (SEC, 12).

“The green is ‘or’” is Derrida’s sole example in SEC of grafting, the only instance he offers illustrating the repurposing of “the written syntagma.” Yet both of Derrida’s proposals, insofar as they work in, and need context, appear at odds with his own stipulations. For one, the phrase’s French gloss, “the green is where,” obviously depends on an accident owed to an existing historical language, namely French, thanks to which it is not agrammatical at all. Moreover, should Husserl’s phrase really not mean anything and still serve as an example of agrammaticality, it seems unclear even for Derrida himself, how far its status as such extends, Derrida’s point now being that context is always needed: does this example operate only for those aware of this debate; of Derrida’s contestation of Husserl; for all German speakers; or actually for everyone, somehow for all possible users in general?

In both cases, the fate of “the green is or” seems limited to the occasion of its use and is certainly not available across an unbounded repetition; it functions only in some circumstances and not others. A portion of Derrida’s concerns at this moment are well aimed, however. Husserl’s original distinction between contradictory and agrammatical “propositions” makes of the logos, the grammatical statement, the benchmark of all meaning and all discourse, something which is true more generally for Husserl. By contesting Husserl’s treatment of agrammaticality, Derrida, accordingly, wants to call into doubt the logos’s traditional authority, the statement’s claim to govern the rights of all modes of writing and speech.

Derrida’s own alternative, by aiming to comprehend the logos and its traditional capacities on different terms, while granting their workings and their authority within limits, unfortunately ends up landing in obscurity. Another route exists, however, for contesting the logos’s rights that also recognizes context’s force and (wholly) eschews the ideal.

To stay with Derrida’s example, an interior decorator might well be imagined, who, when poring over color samples, exclaims “the green is or” and adds “but I what I want is and! That yellow over there, that’s and, not or.” Husserl’s phrase may indeed be repurposed, grafted if one wishes, while both its use and what it says prove new, such that its reach or significance would not be confined to the logos, to Bedeutungen as traditionally understood. For “the green is or” to function in this way, however, requires context—it only works thanks to the above or a similar scenario, such that no break or force of rupture with all contexts nor a structural appurtenance to a positive infinity of them can be affirmed. Contexts can be envisioned, after all, in which Husserl’s phrase can’t be repeated or thought to operate. “The green is or” won’t be repeatable in response to someone asking “please pass the salt” nor, more than likely, at the beginning of a U.S. State of the Union address.20

Derrida himself, of course, doesn’t settle for this modest version of dispensing with all “absolute anchoring.” I am suggesting he does not because ideality (here and in all such instances) would simply cease to be relevant and lose its purported rights and thus would iterability as well. Only individual instances in individual contexts of use would matter, to be sorted out as they may—by readers and speakers, authors and auditors on a given occasion.

To be sure, context, the very concept of context, Derrida also insists in SEC, recurs to presence, to presence to a consciousness; substituting context for the framework of the ideal thus for Derrida finally accomplishes nothing. It is relatively easy to show, however, that context’s fealty to consciousness does not hold in Austin’s account nor, then, the present one. Austin portrays speaking—doing things with words—as a kind of action, subject to the same constraints and evaluations as are all the other acts we perform. All such acts, however, can and do, at least at times, function without regard to an individual’s conscious thoughts or beliefs. Austin, accordingly, explicitly rejects having recourse to “inward and spiritual” acts to determine these acts’ validity, specifically with respect to promises.21

Derrida’s commitment to outstripping the ideal, then, accounts for Derrida’s reluctance to countenance context in all its radicality and his claiming both that the mark entails rupture with context or an appurtenance to a positive infinity of them. Were the phrase and whether it is being repeated adjudicated solely on individual occasions of use, as is actually the case, ideality and iterability both would cease to matter; what is said or written in a given context alone would be available and it would suffice.

2. Quotation and Literature

Iterability, however, is not so easily dispensed with. Whatever their theoretical weaknesses, the iterable mark, and the corresponding possibility of unlimited grafting, seem singularly able to account for Derrida’s own interpretative practice and also for that of literary criticism and other scholarship in the humanities. Since iterability focuses on use, on discourse, it extends beyond the contribution of language and does not wholly forfeit a text’s connection to a moment of production—to an author and an intention in some guise—while also allowing the text to operate beyond them.22 Iterability, accordingly, seems to explain the unique combination of faithfulness and innovation characteristic of literary criticism and of textual interpretation more generally. It and related conceptions make evident both why attention is warranted to the context of a text’s production and to the author’s role, as well as to considerations that exceed these same parameters.

When it comes to accounting for literature and literary criticism, the ability of Austin, of so-called Ordinary Language Philosophy, and indeed of any context-based theory more generally to do the same work is less clear. To be sure, Stanley Cavell, Charles Altieri, Cora Diamond, Alice Crary, and numerous others have taken bites out of this very apple. Their success, however, has been scattershot: literature in these contexts is often viewed in an overly realist vein, and ignored or downplayed is that verbal experimentation and excessiveness that a notion like iterability appears designed to capture. Many, though not all, of these critics and thinkers, moreover, like Austin himself, grant a relative priority to non-literary employments of language.23 They affirm pre-existing modes of speech in more familiar contexts and with established ends, as found in primary senses, mother tongues, and so on. In this way, literature’s scope is narrowed, along with that of literary analysis and textual interpretation themselves.

To be clear, even in the present instance, something like the foregoing may hold—at the very least, literature’s subject matter will be something rather than nothing (though the former may include literature itself, as well as the aesthetic). Nevertheless, what is here being set out departs from the foregoing attempts in that it dispenses entirely with any jumping-off points or guardrails in prior customary or conventional modes of discourse. Thanks to this, literature and the literary text end up furnishing the primary instance and best mapping of discourse’s operation, thereby opening the door to literary writing and reading of essentially unprecedented sorts.

Concerns about literature’s standing in many context-oriented accounts obviously align with Derrida’s own stated worries in SEC’s second part, when he discusses Austin’s analysis of performatives. As is well-known, Derrida questions Austin’s failure to consider uses of performatives when “said by an actor on the stage, or if introduced in a poem, or spoken in soliloquy,” employments which Austin himself seemingly discounts under the headings of “sea-change,” “the ‘non-serious,’” “parasitism,” “etiolation,” “the non-ordinary” (SEC, 16). At this moment, Derrida, slightly tweaking his own previous focus, suggests that the broader category under which all of Austin’s exclusions fall is “quotation.” Austin’s talk of the non-serious, the etiolated, and so on effectively rejects, Derrida asserts, “the possibility for every performative utterance (and a priori every other utterance) to be ‘quoted’” (SEC, 16).

From Derrida’s account of quotation and the further possibilities that it raises, the present understanding of literature and text eventually emerges. Before turning there, note, however, that though Derrida condemns Austin’s exclusion from his analysis of literary and similar instances—he questions what he deems Austin’s drawing “a ditch” around some uses of language and omitting others when framing his theory of speech-acts (SEC, 17)—Derrida himself eventually grants validity to these same distinctions. In line with what was said at the outset about Derrida conceding the functioning of standard philosophical conceptualizations and self-evidences, ultimately within new limits and on a new (contestatory) basis, Derrida himself denies “that citationality in this case [of a performative used in everyday life]” finally “is of the same sort as in a theatrical play, a philosophical reference, or the recitation of a poem”; he indeed affirms, along with Austin, “a relative specificity, as Austin says, a ‘relative purity’ of performatives” (SEC, 18). Though “general iterability,” as conceived by Derrida, encompasses both classes, literary instances and everyday uses nevertheless answer to different “kinds of iteration” (SEC, 17).

In the present debate, then, both Austin and Derrida distinguish literary and related uses from more standard ones, and neither, as it happens, give an account of literary employments, at least in their present texts.24 In turn, the line that Derrida and Austin both draw becomes effaced here, ultimately thanks to some hints given by Austin at the end of How to Do Things with Words as well as certain suggestions made by Derrida. The latter’s focus on citation or quotation enables the unfolding of literary discourse’s features—features which ultimately pertain to every discourse, to all employments of language.

Potentially a fertile notion, citation, in Derrida’s own hands, suffers a problematic fate, one allied to context’s obscurities. Derrida in section I, it became clear, maintains the role of context, while also affirming the mark’s force of rupture or break with it. In part II, Derrida again backs off from context: he does so now by eliding the difference between quotations, or citations, which necessarily refer to (prior) contexts, and applications of codes, structures, or conventions.

Quotations, though repetitions, indeed further entail a concrete occasion on which the phrase in question has previously been used. (John said, “I hate that color,” when I showed him my new rug.) By contrast, employments of a code or convention are not usually deemed quotations or citations, precisely because they don’t imply a prior actual moment when the phrase was used; instead, these uses recur to a rule, procedure, or algorithm. Quotations are thus repetitions that invoke prior contexts, while conventions and codes and their instances don’t require such contexts.25

Derrida himself, however, while criticizing Austin’s omission of the citational, elides this distinction and conflates conventions with quotations or citations as such (SEC, 15). He reworks the notion of a coded utterance, now revived and again identified with the iterable, making it coincide with an actual quotation. “Could a performative utterance succeed,” Derrida asks, “if its formulation did not repeat a ‘coded’ or iterable utterance … if the formula I pronounce in order to open a meeting, launch a ship or a marriage were not identifiable as conforming with an iterable model, if it were not then identifiable in some way as a ‘citation?’” (SEC, 18).26

Having brought Austin’s exclusion of literary and other instances under the umbrella of citation or quotation, Derrida attempts to make the latter notion converge with his own iterability. If, however, this distinction is kept intact and no recourse to an unbounded and context-rupturing iterable mark is had, a working picture of literature and of all discourse can be achieved. Taken at its word, Derrida’s invocation of quotation indeed indicates how literary texts in their specificity operate, as well as how all other discourses do—literature finally, on the present view, serving as a better example of language’s operation than paradigms that focus on single-sentence expressions, whether constative or performative.

More concretely, quotation, citationality as occurring in and through contexts (and thus not entailing any rupture with context), in truth backstops all discourse including all understanding and production of (literary) texts—a fact that not only places literature on the same footing as other discourses but entails that literary discourses supply their joint best example.

Literature’s foregrounding, thanks to a radical embrace of (quotation in) context, to be sure, will also come with a cost. To make quotation, the repetition of a phrase used in context, central to literary and other discourse, requires affirming reference (to some possible worldly subject matter) as ineluctably intertwined with what texts and other expressions say. With ideality’s dismissal, and literature and its repetitions proceeding always in contexts, expressions and understanding not only refer to prior instances of what is said but also refer to some matters previously talked about, some subject or topic or aim, which, like the usages themselves, function both as precedented and not.

To see why this may be the case, it is worthwhile recurring briefly to work done by Mary Louise Pratt quite a while back, in her Toward A Speech Act Theory of Literary Discourse. Pratt there had already focused on citations in life and literature and insisted on breaking down the barrier between the two. To explain the workings of fictional narratives, Pratt more specifically contests the distinction between the illocutions employed to tell stories in everyday life and those found in fiction. In the case of the former, “[w]ithout the slightest hint of infelicity,” Pratt avers, “I can recount an anecdote I heard from someone else whose name I can’t remember to an audience I don’t know about events I didn’t witness that happened somewhere I have never been.”27

Citations in literature and in life share the same footing for Pratt; in both, speakers and authors may lack knowledge of the subjects to which they refer, while in both quotations can still function equally well, without “any infelicity,” as she puts it.

Yet even if quotation in ordinary circumstances and in literature operate identically for Pratt, for her, both also recur to non-citational instances and governing rules of the sort that she takes Austin to analyze, the ones that he (and also Derrida) recognize “in their relative purity.” In the absence of H. P. Grice’s “axioms” (as she calls them), as well as of Austin’s illocutionary rules, neither in literature nor in life could one know what is being talked about when speaking of marrying or promising and so forth—when using performatives and other illocutions more generally. Grice’s and Austin’s analyses thus establish these speech-acts’ original and primary contours—ones that quotations and citations assume, whether in literature or elsewhere.

Literature’s priority and originality as discourse for Pratt reaches a limit, then, in such (pragmatic) conventions, all of which refer back to standard, non-literary speech situations. Consequently, if literature is to be conceived as on entirely the same footing as other discourse, no less given an exemplary status, it is clearly necessary to go Pratt one better by insisting on the following. Everything, including our ordinary usage, must begin from quotation, with conventions, codes, and so on (pragmatic or not) playing no regulative or constitutive role. All verbal performances and their identification depend solely on prior encounters, some earlier commerce, with employments of what in a new context are taken to be the relevant phrases—hence forming one-time series, each with their own “historicity,” as I call it. Conventions of speech of a strong and controlling kind do not exist—from tokens and signifiers up to propositions themselves.28 All discourse instead relies on citation: uses of language with their various achievements deemed having actually previously occurred, even if not necessarily fully recalled as such—hence on a citationality without reserve, as it might be called, since this capacity explicitly lacks any other conditions, inputs, or guardrails.29

Alongside such citationality, as part and parcel of discourse’s historicity, it is also necessary, as already noted, to embrace what Austin’s work makes available: the role of context and worldly reference in all writing and speech—today sometimes referred to as occasionalism. Occasionalism, as attributed to Austin, is a conception that pertains to statements, propositions (though as we will see literature proves a still more apt case). Statements, truth claims, such as 2+2=4, for Husserl, Frege, and others, again, are taken to be necessarily ideal, non-spatio-temporally individuated meanings. Austin contests the ideal’s status by insisting that such claims have sense and can be adjudicated as true or false solely in contexts, in what he deemed “historic situations.” Austin thus would reject Husserl’s insistence, to which Derrida as we have seen also subscribes, that “the flat mountain is golden” or other such utterances can be seen to be false based on their meanings alone. Instead, whatever is being expressed, spoken or written, needs contexts of some sort to be legible in the first place. No stand-alone meanings or expressions exist. (In the above case of “the flat mountain,” one might, after all, be referring to a butte.)

Austin’s occasionalism and its consequences for literature and literary interpretation become clearest in Austin’s views of truth, of “is true,” both prior to HTDTWW and when he recurs to this topic at its end. Austin himself glosses “is true,” it should underscored, in terms of quotation. “Is true” designates (or controversially “describes”) the successful coordination in some given “historic situation” of an expression’s indexical, referential capacity with its descriptive potential. To say of some statement that it “is true” is to affirm that when some claim was once actually used, it melded description and reference appropriately. A statement’s being branded true thus always harbors a quotational or citational vector. Austin pushes back rather radically against the proposition and its framework by tying the notion of something being true to an occasion in this fashion.30

In turn, Austin’s initial mapping of “is true” directly sheds light on literature’s and literary criticism’s operation, as does also its subsequent fate in HTDTWW. Even in the simpler instance of single claims, Austin rejects identifying one part of what is said as indexical and referential and another portion as descriptive. Only the sentence or phrase in use as a whole, he insists, refers and describes.

This holistic requirement, however, is much more evident in literature than elsewhere. A text’s worldly points of captation, its primary references and themes, in literary interpretation, are always gauged through its descriptive aspects, by working through different descriptions it offers, while the reverse is also true: what is said in such passages requires an eye to what may be being talked about, to possible candidates for reference. The text’s referent and what it says about that subject matter (its descriptions) in literature are always parsed together; descriptions direct our eye toward possible reference, while what such descriptions actually say or intimate can only be determined along with that to which they might refer. Literary interpretation is a sorting and articulating of both reference and description.

In the case of literary works, this leaves open an extraordinarily wide range of topics and interests. One can read Henry James’s Turn of the Screw, for example, as telling us something about childhood, or narrative form, or late nineteenth-century England, and so on. It will, however, in its own fashion be understood to be saying something about something, and this will, to varying degrees, relate to the contexts of both its readers and James. One might well perform a “feminist” reading of Screw, largely but not wholly emphasizing the reader’s stance, or “queer” James, but one won’t be taking Screw as a commentary on Hittite burial practices or Schroedinger’s equations. Austin’s account makes clear how and why that is so.

In criticism, as here viewed, then, readers articulate a work’s or corpus’s subject matters, some of its themes and interests—pertaining to various features of human life, historical events, technological developments, and so on—while also unfolding what the author says and shows about them (including how they do so), a task that also can be informed by the precursors that the author’s text’s own citationality affords within these contexts.31 Criticism takes into account not only a given text and what happens in it and to what it refers, possibly along with other texts by the same author, but also texts prior to the present one and the circumstances and ideas that they might treat. Henry James’s The Princess Casamassima, for example, taken with an eye to its own workings, as well as to other texts by James, such as Portrait of a Lady, in part, clearly concerns promising, just like HTDTWW. Whether James is interested in promising in the first place and what he says and shows about this topic, as well as finally the critic’s own appraisal of James’s treatment and these subject matters, in turn, are necessarily approached together: is Casamassima ultimately canned Kant, insightful social commentary, neither, or both—the different possible takes in play, here only examples, by no means being mutually exclusive.

No single piece of criticism, of course, can account for all of a text’s relevant aspects; none can grapple with everything a given literary text may say and be about, given the open-ended citationality here evinced. Hence the understandable proliferation of criticism that surrounds most well-known literary texts and corpuses. This same interplay of precedence (of the texts and circumstances that provide a given text’s contexts) and novelty (in what it says and its descriptions), best instanced in literature, nevertheless does not foreclose discussions concerning whether some single, given interpretation is more or less successful or convincing. With a given reference in view, a reading of a novel or poem can turn out to be implausible or even incorrect.

Still more worrisome in the context of current literary studies than the possibility of declaring some interpretations wrong-headed may be the necessity to approach literature in terms of Austin’s chosen predicate at all: to regard literary works and corpuses as showing or expressing or saying something possibly “true.” This determination more than likely is not the one that today first leaps to mind when considering literature and literary discourse. Veridical concerns appear radically foreign to contemporary criticism, despite the fact that literature is more and more being read with an eye toward its referents—race, money, sexuality, income inequality, the environment, and so on—and it is so read, one assumes, with the aim of arriving at some sort of insight or takeaway, or illumination, that is, some truth or insight, in respect to these same themes, even if one pertaining solely to the depth and urgency of the problems they pose.

Turning to Austin’s final brush with truth and “is true” in HTDTWW, by way of conclusion, allows for elucidating truth’s workings as presently understood, rendering the present account of literary study’s work more complete. In the closing pages of HTDTWW, Austin takes a decisive step: he ceases to view questions of truth (assessments of worldly circumstances, of what is the case) and assessments of other sorts (pertaining to “value” broadly understood) as fundamentally distinct. Austin suggests, instead, that all discourse draws on both facts (truth) and values (other sorts of assessments), at this moment dissolving the distinction between constative and performatives (HTDTWW, 148).32 Performative utterances rely on and sometimes explicitly include assessments that would be traditionally assigned to claims and judgments—as, in the case of a bet, that a race is indeed taking place at two o’clock today or that a given horse is actually running in it. Claims and statements, in turn, necessarily imply and sometimes explicitly incorporate other sorts of evaluations: of what is interesting, appropriate, or in some other way relevant on some occasion. The assertion that “France is hexagonal,” Austin points out, possesses neither truth nor falsity apart from such an evaluation (HTDTWW, 142).

His own analysis, Austin concludes, pertains, then, not to constatives or performatives, but to what he calls “the total speech act in the total speech situation” (HTDTWW, 148). The act so characterized is the genuine unit of expression and the concrete object of his inquiry. Despite its repetition of the word “total,” this phrase does not at all amount to the totalitarian paean to consciousness that Derrida makes it out to be at the beginning of SEC part II (SEC, 14n6, 17). Perhaps ironically, given Derrida’s own taking this phrase out of context, Austin with it instead announces his most radical affirmation of context and his most mature position on “is true.” True/false, he avows, in contrast to his earlier characterization, does not refer to “anything simple at all”; these are “not names for relations [or] qualities” (HTDTWW, 144, 149).

This final, quite open-ended, situating of truth at which Austin arrives more fully illuminates how truth’s consideration operates in criticism and literature. A text’s and its author’s assessments (of what is desirable, required, abhorrent, and so on), along with the subject matters to which those assessments relate, and the worldly circumstances and affairs they invoke, are and must be, Austin indicates, confronted all at once, and this is why the critic too cannot help but engage with what is being said and whether and how it is insightful. Every instance of literature and every interpretation, that is, involves both what is meant by values and facts, and critics must wrestle, ultimately simultaneously, with the aptness of both sorts of determinations. The difference in literature’s case from most speech is that both of these dimensions are always explicitly on the table, often in complex, albeit concrete settings, scenarios, and narratives. To be sure, some critics, like some authors, will be more concerned with assessments that classically would be those of value—ultimately plumping for some looks at the world and critiquing others; other critics and authors will stress various registers of description. Yet in all these instances, these vectors are interwoven, such that a description of what is putatively the case always implies some interest and viewpoint; in turn, affirmations or critiques of some project, plan, or program, always imply some estimations of circumstances and what actually is the case. Literature and literary critical performances necessarily engage at once with aptnesses pertaining to value and to truth and this holds sway whether they concern the future, past, or present.

Criticism, on this view, then, never could be restricted to a discussion of the author’s own viewpoint, which does not mean that such intentions are simply irrelevant. For similar reasons, ascertaining what a text and author says often invites reflection on their positioning within greater social, cultural, or historical milieux.33 The genre in which a text functions, its style, as well as its expressed and implicit concerns may all open on to reflection on topics of wider scope.

None of what the text says and is about can be determined beginning from such considerations, however; texts are not functions of cultural, social, or historical “structures” or “logics.” Such suppositions erase the text itself, by subordinating it to new and again unexplained systems or structures, implicitly resurrecting the ideal. Appeals to structures or logics necessarily trade on otherwise unexplained ideal last instances, ones hailing now not from language but from society, history, the economy, and so on.34 Their invocation thus renders the text in question effectively equivalent to the common understanding of “now is night,” an inscription somehow without a writer, receiver or context—subordinating it to a new ideal instance and thereby depriving it of its status as an authored text.35

In sum, then, SEC’s legacy, on the present view, will be to have gone furthest, along with a few others, in engaging propositionality’s paradoxes and ideality’s seeming inevitabilities in the hope of arriving at what may lie beyond them. Doing so, Derrida’s piece invites a moment when ideality and propositionality, structures and constitutive conventions, have entirely departed from our thinking and the question of what other instances speak when we do ceases to be posed.36 Even when the interest in what or who else speaks abates, it will still remain true, however, that another speech, a different instance of writing and thought, always precedes our own—that all discourses and their topics remain related to earlier instances of speech, understandings, and encounters with things that they do not fully master—and thus that no discourse ever entirely has either a first or a last word.