How to do Things with Texts by M. H. Abrams

With the appearance of his The Mirror and the Lamp, a study of critical theory of the romantic period, M. H. Abrams became known as a lucid and thorough scholar of the thought of that age. His second major book,Natural Supernaturalism, was an impressive overview of romantic literature.Throughout his career Abrams has produced important essays, mainly on romantic poetry, but in his later work he has entered the

contemporary theoretical wars with essays that are openly critical of developments occurring around de-construction and the question of whether determinate meaning is possible. In a well-known essay “The Deconstructive Angel” (Critical Inquiry 3 [1977]), Abrams took as his target in particular the later deconstructive writings of /. Hillis Miller, who responded in the essay in this volume.

This selection of Abrams’s is a critique of the work of Jacques Derrida, Stanley Fish, and Harold Bloom. Critical of all three, he is nevertheless able to provide, in his characteristic way, a clear description of the positions they hold. Abrams recognizes their differences, but he sees one overarching similarity among them, and he does not like it. That is their common rejection of presumptions about the meaning of literary texts, indeed of all texts, that have been fairly commonly held by traditional humanists—that authors had something to say which they conveyed in such a way within a tradition of linguistic conventions as to make possible the assumption that their meaning could be construed by a reader. Abrams does not imply that new readings cannot reasonably arise. He holds that we read according to the linguistic strategy employed by the author of the work, and clearly he believes that in situations where a past text provides special dif­ficulties this strategy is theoretically recoverable by the work of humanistic scholarship.

Abrams’s principal works are The Mirror and the Lamp: Romantic Theory and the Critical Tradition (1953); Natural Supernaturalism: Tradition and Revo­lution in Romantic Literature (1971); and The Correspondent Breeze (1984), a collection of essays on romanticism. See Wayne Booth, “M. H. Abrams: Histo­rian as Critic, Critic as Pluralist,” Critical Inquiry 2 (Spring 1976).

***********

The Age of Criticism, which reached its zenith in the mid-decades of this century, has given way to the Age of Reading, and whereas the American new critics and European formalists of the Age of Criti­cism discovered the work-as-such, current literary theorists have discovered the reader-as-such. This reader, as everyone knows who has kept even cur­sorily in touch with the latest Paris fashions, is not the man he used to be. He is a wraith of his old self, stripped of everything human, as part of a system­atic dehumanizing of all aspects of the traditional view about how a work of literature comes into being, what it is, how it is read, and what it means.

For purpose of comparison, let me sketch the sa­lient and persistent features of the traditional, or humanistic paradigm of the writing and reading of literature. The writer is conceived, in Wordsworth’s terms, as “a man speaking to men.” Literature, in other words, is a transaction between a human au­thor and his human reader. By his command of lin­guistic and literary possibilities, the author actu­alizes and records in words what he undertakes to signify of human beings and actions and about matters of human concern, addressing himself to those readers who are competent to understand what he has written. The reader sets himself to make out what the author has designed and sig­nified, through putting into play a linguistic and lit­erary expertise that he shares with the author. By approximating what the author undertook to sig­nify the reader understands what the language of the work means.

In our Age of Reading, the first casualty in this literary transaction has been the author. To the noninitiate, it is bemusing to observe the compla­cency with which authors of recent books and essays announce their own demise. “It is about time,” says Michel Foucault, “that criticism and philosophy acknowledged the disappearance or the death of the author.”¹ “As institution,” according to Roland Barthes, “the author is dead: his civil status, his biographical person, have disappeared.”² The necrology extends to the human reader, and in­deed to man himself, who is reduced to an illusion engendered by the play of language, or as Foucault puts it, to “a simple fold in our knowledge,” des­tined to “disappear as soon as that knowledge has found a new form.”³ In these new writings about reading, accordingly, the author deliquesces into writing-as-such and the reader into reading-as- such, and what writing-as-such effects and reading- as-such engages is not a work of literature but a text, writing, ecriture.⁴ In its turn the text for­feits its status as a purposeful utterance about human beings and human concerns, and even its in­dividuality, becoming simply an episode in an all- encompassing textuality—dissolved, as Edward Said has remarked, into “the communal sea of lin- guicity.”⁵ Consonantly, the relations between au­thors which had traditionally been known as “influ­ence” are depersonalized into “intertextuality,” a reverberation between ownerless sequences of signs.

It might be expected that, evacuated of its hu­manity, reading-as-such would become an interplay of bloodless abstractions. Quite to the contrary. We find in French structuralist criticism and its Ameri­can analogues that reading is a perilous adven­ture—not of a soul among masterpieces,⁶ but of the unsouled reading-process as it engages with the text-as-such. Persistently this inhuman encounter is figured in a rhetoric of extremity, as tense with the awareness of risk and crisis; anguished by doubts about its very possibility; meeting everywhere in the “action du signifiant”⁷ with violence, disrup­tion, castration, mysterious disappearances, mur­der, self-destruction; or as overcome by vertigo as the ground falls away and leaves it suspended over an abyss of recessive meanings in a referential void. In this Gothic context of the horrors of reading it is a relief to come upon Roland Barthes’s The Plea- sure of the Text, with its seeming promise to revive the notion, as old as Aristotle and Horace, that the distinctive aim of a literary work is to give pleasure to its readers.⁸ But then we find in Barthes’s account that the pleasure is not in the artful management of the human agents, interactions, and passions sig­nified by the text, but in the engagement with the text-as-such, and that Barthes adapts the tradi­tional concept to current connoisseurs of textuality by a running conceit sustained by double entendres, in which textual pleasure is assimilated to sexual pleasure; the prime distinction is between the mere plaisir effected by a comfortably traditional text and the orgasmic rapture, jouissance, in the close encounter with a radical “modern” text which, by foiling the reader’s expectations, “brings to a crisis his relations with language.” It seems safe to pre­dict that the innocent reader, seduced by Barthes’s erotics of the text, who engages with a nouveau ro­man is in for a disappointment.

My concern, however, is with the strategy and the rhetorical tactics of structuralist criticism only as a background for considering three current writers who put forward radical new ways of reading texts. One, Jacques Derrida, is a French philosopher with an increasing following among American critics of literature; by pressing to an extreme the tendencies of structuralism, Derrida proposes a mode of read­ing which undermines not only the grounds of struc­turalism itself, but the possibility of understanding language as a medium of decidable meanings.⁵ The other two, Stanley Fish and Harold Bloom, are Americans who set their theories of reading in op­position to what they decry as the antihumanism of structuralist procedures.¹⁰ All three are erudite, for­midable, and influential innovators who found their strategies of reading on an insight into a neglected aspect of what enters into the interpretation of a text. These theorists differ, we shall see, in essen­tial respects, but they share important features which are distinctive of current radicalism in inter­pretation. In each, the theory doesn’t undertake simply to explain how we in fact read, but to propa­gate a new way of reading that subverts accepted in­terpretations and replaces them with unexpected alternatives. Each theory eventuates in a radical scepticism about our ability to achieve a correct in­ terpretation, proposing instead that reading should free itself from illusory linguistic constraints in order to become liberated, creative, producing the meanings that it makes rather than discovers. And all three theories are suicidal; for as the theorist is aware, his views are self-reflexive, in that his subver­sive process destroys the possibility that a reader can interpret correctly either the expression of his theory or the textual interpretations to which it is applied.

It is worth noting that such Newreading—by which I denote a principled procedure for replacing standard meanings by new meanings—is by no means recent, but had many precedents in Western hermeneutics. We find such a procedure, for ex­ample, in ancient Greek and Roman attempts to uncover the deep truths hidden within Homer’s sur­face myths and fictions, and to moralize the im­moral tales of Ovid; we find it also in the rein­terpretations of the Old Testament by writers of the New Testament, as well as by Jewish Kabbalists; we find a similar procedure in medieval and later exe- getes of the many-leveled allegorical meanings in the entire biblical canon. These old reinterpretive enterprises, however diverse, all manifest three pro­cedural moments, or aspects: (1) The interpreter in­dicates that he understands the standard, or ac­cepted meanings of a text or passage (called by biblical exegetes “the literal meaning”). (2) He replaces, or at least supplements, these standard meanings by new meanings. (3) He mediates be­tween these two systems of signification by setting up a transformational calculus which serves to con­vert the old meanings into his new meanings. We can, I think, discern a parallel procedure in our cur­rent Newreaders. In considering their proposals, I shall ask the following questions. What sort of things does each Newreader undertake to do with texts? By what transformational devices does he manage to do these things? And then there is the general question: What is there about the way lan­guage functions that enables a Newreader to ac­complish the surprising things he does with texts?

The Science of Nescience: Jacques Derrida

How is one to make entry into the theory of Jacques Derrida, the most elusive, equivocal, and studiously noncommittal of philosophical writers? I shall try to break through with a crashing generalization: As a philosopher of language, Derrida is an absolutist without absolutes.

Derrida proposes that both the Western use of language and philosophies of language are “logo- centric”; that they are logocentric because essen­tially “phonocentric” (that is, giving priority and privilege to speech over writing); and that language is thereby permeated, explicitly and implicitly, by what, in a phrase from Heidegger, he calls “the metaphysics of presence.” By “presence”—or in al­ternative terms, a “transcendental signified” or “ultimate referent”—he designates what I call an absolute; that is, a foundation outside the play of language itself which is immediately and simply present to us as something ultimate, terminal, self- certifying, and thus adequate to “center” the struc­ture of the linguistic system and to guarantee the determinate meaning of an utterance within that system. The positing of some form of presence, it is suggested, is the expression of a desire—which is the motivating desire of metaphysics—to establish a conceptual replacement for the certainty about language and meaning provided by the myth in Genesis of language as originated and guaranteed by a divine, hence absolute, authority, or else by the theological view that language is certified by the omnipresence of the Logos. In a remarkable series of readings of diverse texts, philosophical and literary, Derrida subtly uncovers the presupposi­tion that there is an absolute foundation for lan­guage, and displays the internal paradoxes and self- contradictions that are attendant upon such a presupposition. The quest for presence, then, is doomed to unsuccess, whether that supposed abso­lute is the presence of his meaning to the conscious­ness of the speaker at the instant of his utterance; or Platonic essences that underwrite the significations of verbal names; or a fixed and simple referent, “the thing itself,” in the world “outside of language”; or Heidegger’s “Being” as the ultimate ground of sig­nification and understanding. But having, in the critical aspect of his reading of texts, dismantled the traditional absolutes, Derrida remains committed to absolutism; for he shares the presupposition of the views he deconstructs that to be determinately understandable, language requires an absolute foun­dation, and that, since there is no such ground, there is no stop to the play of undecidable mean­ings: “The absence of a transcendental signified extends the realm and the play of signification to in­finity.”¹¹ In this aspect of his dealings with lan­guage, Derrida’s writings present variations on a Nietzschean theme: Absolutes, though necessary, are dead, therefore free play is permitted.

  It should be remarked, however, that the philoso­phy of language offers an alternative to the supposi­tion that language requires an absolute foundation in order to be determinately meaningful. This alter­native sets out from the observation that in practice language often works, that it gets its job done. We live a life in which we have assurance that we are able to mean what we say and know what we mean, and in which our auditors or readers show us by their verbal and actional responses whether or not they have understood us correctly. This alternative stance takes as its task not to explain away these workings of language, but to explain how it is that they happen, and in instances of failure, to inquire what it is that has gone wrong. A prominent recent exemplar of this stance is the Philosophical Inves­tigations of Ludwig Wittgenstein.¹² There are simi­larities between Wittgenstein’s views of language and Derrida’s, in the critical aspect of Derrida’s reading of philosophical texts. Like Derrida, for ex­ample, Wittgenstein insists that it is not possible to use language to get outside “the limits of lan­guage”; he holds that the concept that language di­rectly represents reality is simply “a picture that holds us captive”; he rejects the account of the meaning of an utterance in terms of the objects or processes to which its words refer, or as equivalent to the conscious state of the speaker of the utter­ance; and, in his own way, he too deconstructs the traditional absolutes, or “essences,” of Western metaphysics. He also rejects as futile the quest for an ultimate foundation for language. Philosophy, he says, “can in the end only describe” the “actual use of language,” for it “cannot give it any foundation”; in giving reasons for the working of language, “the spade turns” before we reach an ultimate reason. But Wittgenstein’s stance is that language is “a prac­tice” that occurs as part of a shared “form of life,” and that this practice works; as he puts it, “this game is played.” His Investigations are designed to get us to recognize when language works, and when it doesn’t—“when language is like an engine idling, not when it is doing work”—to get us to under­stand how the slippage occurred.

Derrida of course acknowledges that language works, or as he puts it, that it “functions”—that we constantly perform what we take to be successful speech acts and successful instances of oral commu­nication, and that a written text is lisible, “legible,” that is, strikes us as having determinably specific meanings. But he accounts for this working as no more than “the effects of ideality, of signification, of meaning and of reference”—effects which are en­gendered by the play of differences within language itself; he then proceeds to “deconstruct” these effects by undertaking to show that, since they lack a ground in presence, their specificity of meaning is only a simulation. Derrida’s procedure might be summarized as follows. He agrees that language works, then asks, “But is it possible that it really works?” He concludes that, lacking an ultimate ground, it is absolutely not possible that it works, hence that its working is only a seeming—that, in short, though texts may be legible, they are not in­telligible, or determinately significant.

Of each of the traditional terms and distinctions used to analyze the working of language—terms such as “communication,” “context,” “intention,” “meaning,” and oppositions such as speech-writing, literal-metaphorical, nonfictional-fictional—Der­rida requires not only that they be grounded in ab­solute presence, but also that they be certified by criteria of what he calls “ideal purity” and “ulti­mate rigor” if they are to be determinately used and understood. For example: in order to communicate “a determinate content, an identifiable meaning,” each of these words must signify a concept “that is unique, univocal, rigorously controllable,” and its contextual conditions of use must be “absolutely determinable” and “entirely certain”; while the ut­terance of a determinate speech act must be tied to “the pure singularity of the event.” Of course such analytic words cannot meet these criteria of ab­solute fixity, purity, and singularity, nor can any words, for it is an essential condition of a language that a finite set of words, manageable in accordance with a finite set of regularities, be capable of gener­ating an unlimited variety of utterances adaptable to an unlimited diversity of circumstances, pur­poses, and applications. But Derrida’s all-or-none principle admits of no alternative: failing to meet absolute criteria which language cannot satisfy without ceasing to do its work, all spoken and writ­ten utterances, though they may give the “effect” of determinate significance, are deconstructable into semantic indeterminacy.

Derrida describes his “general strategy of de­construction” as a mode of “double writing”: it first “inverts” the hierarchy of the terms in standard philosophical oppositions such as speech-writing, signifier-signified, then it “displaces” what was the lower term in the hierarchy (or a derivative from that term) “outside the oppositions in which it was held.” The latter move generates, in place of the standard terms used to analyze the workings of lan­guage, a set of new terms which, he says, are neither words nor concepts, neither signifiers nor signifieds. These invented pseudoterms, however, although “displaced” from their locus within the system of language, nonetheless are capable of producing “conceptual effects”; and these effects operate in two dimensions. On the one side, they account for the fact that texts are “legible,” yielding the effects of seemingly determinable meanings. On the other side, they serve as what I have called a set of trans­formers, which Derrida employs to “disseminate” these effects into their deconstructed alternatives.

            The chief transformer is differance¹³—Saussure’s key term “difference,”¹⁴ twice-born and re-spelled with an “a”—which conflates “difference” and “deferment.” In one aspect of its functioning, the “differences” among signs and among the condi­tions of their use explain how they generate their apparently specific significations; in its deconstruc­tive aspect, it points to the fact that, since these sig­nifications can never come to rest in an absolute presence, their specification is deferred from sub­stitute sign to substitute sign in a movement with­out end. Similarly with the other nonwords for non­entities with which Derrida replaces standard terms for dealing with language; in place of the spoken utterance or written text, the “general text” or “proto-writing”; in place of the word, “mark” or “grapheme”; in place of significance, “dissemina­tion” or a large number of other “nicknames” that Derrida resourcefully coins, or else adapts to his equivocal purpose from common usage. All in their double function account for the legibility of a text at the same time that they “open” the apparent clo- sure of the text “ett abyrne,” into the abyss of an endless regress of ever-promised, never-delivered meaning.

Derrida emphasizes that to deconstruct is not to destroy; that his task is to “dismantle the meta­physical and rhetorical structures” operative in a text “not in order to reject or discard them, but to reconstitute them in another way”; that he puts into question the “search for the signified not to annul it, but to understand it within a system to which such a reading is blind.” He can in fact be des­ignated as, on principle, a double-dealer in lan­guage, working ambidextrously with two semantic orders—the standard and the deconstructed. He writes essays and books, and engages in symposia and in debates, that put forward his deconstructive strategy and exemplify it by deconstructing the texts of other writers. In this deconstruction of logo- centric language he assumes the stance that this lan­guage works, that he can adequately understand what other speakers and writers mean, and that competent auditors and readers will adequately understand him. In this double process of constru­ing in order to deconstrue he perforce adopts words from the logocentric system; but he does so, he tells us, only “provisionally,” or sous rature, “under era­sure.” At times he reminds us of this pervasive pro­cedure by writing a key word but crossing it out, leaving it “legible” yet “effaced”—an ingenious doublespeak, adapted from Heidegger, that enables him to eat his words yet use them too.

Derrida’s double-dealing with texts is all-inclusive, for he is aware that his deconstructive reading is self-reflexive; that, although “exorbitant” in inten­tion, it cannot in fact escape the orbit of the lin­guistic system it deconstructs. “Operating neces­sarily from the inside,” as he says, “the enterprise of deconstruction always in a certain way falls prey to its own work.” The invented nonwords which serve as his instruments of deconstruction not only are borrowed from language, but are immediately re- appropriated into language in the process of their “iteration” (in Derrida’s double sense of being “re­peated” and therefore “other” than absolutely self­identical). And the deconstructive reading these in­struments effect, he says, is a “production,” but “does not leave the text.. . . And what we call pro­duction is necessarily a text, the system of a writing and a reading which we know is ordered by its own blind spot.” Even as they are put to work on a text, accordingly, the deconstructive instruments decon­struct themselves, as well as the deconstructed translation of the original text which Derrida, as deconstructor, has no option except to write down as still another deconstructible text.

Derrida’s critical lexicon, therefore, as Gayatri Spivak, his translator, has said, “is forever on the move.” In the consciously vain endeavor to find a point outside the logocentric system on which to plant his deconstructive lever, he leaps from neolo­gism to neologism, as each sinks beneath his feet en abyme. His deconstructive enterprise thus is a boot­strap operation, a deliberate exercise in ultimate futility, in a genre of writing he has almost single- handedly invented—the serious philosophy of the absurd. The most earnest and innovative passages in Derrida are those which, on the surface, seem at best playful and at worst embarrassingly arch—pas­sages which deploy grotesque puns, distorted words, false etymologies, genital analogues, and sexual jokes; which insist in our attending to the shapes of printed letters, play endless tricks with Derrida’s own name and with his written signature; or collocate wildly incongruous texts. In such pas­sages—extended to the length of a nonbook in his Glas—Derrida is the Zen master of Western phi­losophy, undertaking to shock us out of our habit­ual linguistic categories in order to show what can­not be told without reappropriation into those categories: what it is to experience a text not as conveying significance, but as simply a chain of marks vibrating with the free and incessant play of differance.

Occasionally, however, Derrida ventures the at­tempt to tell what can’t be told, that is, to make his deconstructive concepts, although “in intimate rela­tionship to the machine whose deconstruction they permit,” nonetheless “designate the crevice through which the yet unnameable glimmer beyond the clo­sure can be glimpsed.” This glimpse is of an apoca­lyptic new world which, he prophesies, will be effected by the total deconstruction of our logo­centric language-world—“the ineluctable world of the future which proclaims itself at present, beyond the closure of knowledge,” hence cannot be de­scribed but only “proclaimed, presented, as a sort of monstrosity.”

To realize the inclusiveness of the new world thus proclaimed, we need to keep in mind what Derrida calls “the axial proposition” in Of Grammatology, his basic theoretical work: II n’y a pas d’ hors-texte, “there is no outside-the-text.” Like all Derrida’s key assertions, this sentence is multiple in significance. In one aspect, it says we can’t get outside the writ­ten text we are reading—it is a closure in which both its seeming author and the people and objects to which the text seems to refer are merely “effects” engendered by the internal action of differance. In another aspect, it says that there is nothing in the world which is not itself a text, since we never expe­rience a “thing itself,” but only as it is interpreted. In this inclusive rendering, then, all the world’s a text, and men and women merely readers—except that the readers, according to Derrida, as “sub­jects,” “egos,” “cogitos,” are themselves effects which are engendered by an interpretation; so that in the process of undoing texts, we undo our textual selves. The apocalyptic glimpse, it would seem, is of a totally textual universe whose reading is a mode of intertextuality whereby a subject-vortex engages with an object-abyss in infinite regressions of de­ferred significations.

At the end of his essay “Structure, Sign and Play,”¹⁵ Derrida hazards his most sustained endeavor in the vain attempt to put names to “the as yet unname- able which cannot announce itself except. . . under the formless form, mute, infant, and terrifying, of monstrosity.” The annunciation is of “a world of signs without error, without truth, without origin, which is offered to an active interpretation,” in which one “plays without security” in a game of “absolute chance, surrendering oneself to genetic indeterminacy, the seminal chanciness of the trace.” Derrida suggests that we at least try to overcome our age-old nostalgia for security, with its hopeless dream of an absolute ground in “full presence, the reassuring foundation, the origin and end of the play,” and to assume instead toward this prophecy of deconstruction triumphant the nonchalance of the Ubermensch, “the Nietzschean affirmation, the joyous affirmation of the freeplay of the world.” If one cannot share the joy, one can at least acknowl­edge the vertigo effected by Derrida’s vision, yet take some reassurance in the thought that, even in a sign-world of absolute indeterminacy, it will pre­ sumably still be possible to achieve the “effect” of telling a hawk from a handsaw, or the “effect,” should the need arise, of identifying and warning a companion against an onrushing autobus.

Reading Between the Words: Stanley Fish

Of the deconstructive “interpretation of interpreta­tion” Derrida remarks that it “attempts to pass be­yond man and humanism.” Stanley Fish represents his theory of reading as a ringing defense against “the dehumanization of meaning” in the “formal­ism” of current linguistics and stylistics, as well as in structuralist criticism, which raises “the implied antihumanism of other formalist ideologies to a principle.” Such theory “is distinguished by what it does away with, and what it does away with are hu­man beings.” Fish himself undertakes to explain meaning by reference to “the specifically human activity of reading,” proposing as his humanistic “point of departure the interpretive activity (experi­ence) by virtue of which meanings occur.” His model for interpretation is that of a reader who confronts the marks on a page and generates mean­ings by his informed responses to it. In the tradi­tional humanistic view, it will be recalled, there is an author who records what he undertakes to sig­nify, as well as a reader who undertakes to under­stand what the author has signified. In terms of this paradigm, Fish’s rehumanization of reading is only a half-humanism, for it begins by diminishing, and ends by deleting, the part played by the author. In Fish’s later writings, we shall see, the reader be­comes the only begetter not only of the text’s mean­ings, but also of the author as the intentional pro­ducer of a meaningful text.¹⁶

Fish differs from other systematic Newreaders in that, instead of setting up a matrix of transform­ers—a set of revisionary terms—he proposes a “method” or “strategy” which is in fact a set of moves to be enacted by the reader in the process of construing a text. These moves are such as to yield meanings which are always surprising, and often antithetic to, what we have hitherto taken a text to mean. As the key to his method, he proposes that we replace our usual question while reading— “What does this sentence (or words, phrase, work) mean?”—by what he calls “the magic question,” namely: “What does this sentence do?” The result of this magic question, if persistently applied by readers, is that it “transforms minds.”

In all Fish’s expositions of his method, however, “the key word,” as he himself remarks, “is, of course, experience”; and what in fact works the transfor­mative magic is his major premise, express and im­plied, “Reading is an experience.” On the common assumption that the term “experience” can be predi­cated of any perception or process of which one is aware, this assertion seems self-evident, and inno­cent enough; it can, however, lead to dubious con­sequences when posed as the premise from which to draw philosophical conclusions. Take, for ex­ample, one of Fish’s favorite sources of sentences to demonstrate his method of reading, Walter Pater’s “Conclusion” to The Renaissance.'⁷ In one virtuoso paragraph, Pater begins by casually positing that the perception of all “external objects” is an “expe­rience,” then dissolves the experience of each object “into a group of impressions,” translates this into “the impression of the individual in his isolation,” and reduces it “to a single sharp impression” in a fleeting moment, bearing traces of “moments gone by”; to this, he asserts, “what is real in our life fines itself down.” From the premise that everything we perceive is our experience, Pater has taken us head­long down the metaphysical slope to his conclusion of a solipsism of the specious present—that one can validly assert reality only for one’s single sense- impression in a fugitive “Now!” The example should make us wary about the consequences for interpretation that Fish deduces from his premise that reading is an experience, and what he proposes as its immediate corollary—that “the meaning of an utterance ... is the experience—all of it.”

One conclusion that Fish draws from this claim that meaning is all of a reader’s experience (all the experience, as he qualifies it, of a “competent” or “informed” reader) is that, since the “response in­cludes everything” and is a “total meaning experi­ence,” you can’t make valid use of the traditional distinction between subject matter and style, “process and product (the how and the what)” in an ut­terance. Another and related conclusion is that you can’t distinguish, within the totality of a declarative sentence, what is being asserted. He excerpts, for example, from Pater’s “conclusion” to The Renais­sance: “That clear perpetual outline of face and limb is but an image of ours.” In standard stylistic analysis, he says, this is “a simple declarative of the form X is Y.” He then analyzes the experience of reading the sentence in accordance with the ques­tion, “What does it do?” and finds that “in fact it is not an assertion at all, although (the promise of) an assertion is one of its components. It is an experi­ence; it occurs; it does something; . . . [and] what it does is what it means.” Turn Fish’s method of read­ing back upon his own writing (I find nothing in the method to prevent our doing so) and we get the in­teresting result that his assertion about Pater’s sen­tence—“In fact it is not an assertion at all. . .”—is in fact not an assertion at all, but only an evolving experience effectuated in a reader.

I want to focus, however, on an important aspect of Fish’s strategy for transforming accepted mean­ings. He supplements his basic equation of meaning with the reader’s total response by proposing a start-stop-extrapolate method in reading:

The basis of the method is a consideration of the temporal flow of the reading experi­ence. ... In an utterance of any length, there is a point at which the reader has taken in only the first word, and then the second, and then the third, and so on, and the report of what happens to the reader is always a report of what has happened to that point. (The re­port includes the reader’s set toward future experiences, but not those experiences.)¹⁸

What happens at each stopping point, then, is that the reader makes sense of the word or words he has so far read, in large part by surmising what will come next. These surmises may, in the text’s sequel, turn out to have been right, but they will often turn out to have been wrong; if so, “the resulting mis­takes are part of the experience provided by the au­thor’s language, and therefore part of its meaning.” Thus “the notion of a mistake, at least as something to be avoided, disappears.” And the point at which “the reader hazards interpretive closure” is inde­pendent of the “formal units” (such as syntactical phrases or clauses) or “physical features” (such as punctuation or verse lines) in the text written by the author; the method in fact creates what the reader takes to be formal features of the text, “because my model demands (the word is not too strong) percep­tual closures and therefore locations at which they occur.” In reading the sentence from Pater’s Renais­sance, for example, Fish hazards brief perceptual closures after each of the four opening words: “That clear perpetual outline . . .”

It is apparent that by Fish’s start-stop strategy, a large part of a text’s meaning consists of the false surmises that the reader generates in the temporal gaps between the words; and this part, it turns out, constitutes many of Fish’s new readings. To cite one instance: Fish presents a three-line passage from Milton’s Lycidas which describes one consequence of Lycidas’s death:

The willows and the hazel copses green Shall now no more be seen,

Fanning their joyous leaves to thy soft lays.

Although, he tells us, it is “merely a coincidence” when a perceptual closure coincides with a formal unit or physical feature such as the end of a verse line, it happens in this instance that the reader’s process of making sense “will involve the assump­tion (and therefore the creation) of a completed as­sertion after the word ‘seen’” at the end of the sec­ond line; he will then hazard the interpretation that these trees, in sympathy with the death of Lycidas, “will wither and die (will no more be seen by any­one).” And though this interpretation will be un­done “in the act of reading the next line,” which re­verses it by going on to say that they “will in fact be seen, but they will not be seen by Lycidas,” the false surmise remains part of the text’s meaning.

I recall a new reading of the closing couplet of Lycidas which William York Tindall of Columbia proposed to me many years ago. Tindall suggested the following perceptual closures (I cite the first edi­tion of 1637):

At last he rose, and twitch’d. His mantle blew.

To morrow to fresh Woods, and Pastures new.¹⁹

Those who know Bill Tindall may suspect he was not wholly serious in this proposal. Yet according to Fish’s strategy, it is the way a first reader might haz­ard his perceptual closures. The thought that, even after subsequent correction, this misreading re­mains an element in the poem’s meaning is to me disquieting.

I have myself tried, by way of experiment, to read in accordance with Fish’s method. By stern self- discipline, I managed to read word by word and to impose frequent perceptual closures, resisting the compulsion to peek ahead in order to see how the phrases and clauses would work out in the total sentence. And instead of suspending judgment as to meaning until the semantic Gestalt was complete, I solicited my invention to anticipate possible mean­ings and actuated my will to fix on a single one of these possibilities. The result was indeed an evolv­ing sequence of false surmises. I found, however, that the places where I chose to stop rarely coin­cided with the stopping-places of Stanley Fish, and that my false surmises rarely matched his, especially in the startling degree to which they diverged from what actually followed in the text. What am I to conclude? A possible conjecture is that Fish himself has not always resisted the impulse to peek ahead; that in fact many of his novel readings are not pro­spective, but retrospective; that in local instances they are the result of a predisposition to generate surprising meanings between the words; and that in large-scale instances, when he presents a new read­ing of a total literary work, they are the result of a predisposition to generate a system of surprising meanings of a coherent sort.

In his earlier writings, despite some wavering as to what is implied by his use of the term “method,” Fish represented his analyses primarily as a descrip­tion of what competent readers in fact do; its aim was simply to make “available to analytic con­sciousness the strategies readers perform, indepen­dently of whether or not they are aware of having performed them.” In his recent theoretical writings, however, Fish asks us to take his method not as “descriptive” but “prescriptive”; its aim now is to per­suade us to give up reading in our customary way and instead to “read in a new or different way.” Fish’s current views are an extreme form of method­ological relativism, in which the initial choice of a method of reading is “arbitrary,” and the particular method that the reader elects creates the text and meanings that he mistakenly thinks he finds. “Inter­pretive strategies” are procedures “not for reading (in the conventional sense) but for writing texts, for constituting their properties and assigning their in­tentions.” “Formal units,” and even “the ‘facts’ of grammar,” are “always a function of the inter­pretive will one brings to bear; they are not ‘in’ the text.” It turns out, indeed, that there is nothing either inside or outside the text except what our elected strategy brings into being, for “everyone is continually executing interpretive strategies and in that act constituting texts, intentions, speakers, and authors.” Starting with the premise that the mean­ing is all of a reader’s experience of a text, we have plunged down the metaphysical slope to the conclu­sion that each reader’s optional strategy, by deter­mining his responsive experience, creates every­thing but the marks on the page, including the author whose intentional verbal acts, we had mis­takenly assumed, effectuate the text as meaningful discourse.

From this position Fish draws the consequence that, since all reading strategies are self-confirming, there is no “right reading” of any part of a text; there are only agreements among readers who be­long to an “interpretive community” which hap­pens to share the same strategy. And with his usual acumen, Fish acknowledges that the reading strat­egy he himself proposes is no less “arbitrary” in its adoption and therefore no less a “fiction” than al­ternative ways of reading; his justification for urg­ing it upon us is that it is “a superior fiction.” It is superior because it is “more coherent” in the rela­tion of its practice to its principles, and because “it is also creative.” Insistence on a “right reading” and “the real text” are the fictions of formalism, and as fictions they have the disadvantage of being confining. My fiction is liberating. It relieves me of the ob­ligation to be right (a standard that simply drops out) and demands only that I be inter­esting (a standard that can be met without any reference at all to an illusory objectivity). Rather than restoring or recovering texts, I am in the business of making texts and of teaching others to make them by adding to their repertoire of strategies.

In these claims Fish does his own critical practice less than justice. Many of his close readings of liter­ary texts effect in his readers a shock of recognition which is the sign that they are not merely interest­ing, but that they are right. In such readings, how­ever, he escapes his own theory and reads as other competent readers do, only more expertly than many of us; his orientation to the actual process of reading serves in these instances to sensitize him to nuances effected by the author’s choice and order of words that we have hitherto missed. And even when, in conformity with his stated strategy, Fish creates meanings by reading between the words, the new readings are often, as he claims, interesting. They are interesting because they are bravura critical per­formances by a learned, resourceful, and witty in­telligence, and not least, because the new readings never entirely depart from implicit reliance on the old way of reading texts.

1 remain unpersuaded, therefore, that the her­meneutic circle is inescapably, as Fish represents it, a vicious circle—a closed interplay between a reader’s arbitrary strategy and his interpretive find­ings. I persist in the assurance that a competent reader of Milton, for example, develops an exper­tise in reading his sentences in adequate accordance both with Milton’s linguistic usage and with the strategy of reading that Milton himself deployed, and assumed that his readers would deploy. This ex­pertise is not an arbitrary strategy—though it re­mains continuously open to correction and refine­ment—for it has a sufficient warrant in evidence that we tacitly accumulate in a lifetime of speaking, writing, and reading English, of reading English literature, of reading Milton’s contemporaries, and of reading Milton himself. Those who share this as­surance set themselves to read Milton’s text, not as pretext for a creative adventure in liberated inter­pretation, but in order to understand what it is that Milton meant, and meant us to understand. For our prepossession is that, no matter how interesting a critic’s created text of Milton may be, it will be less interesting than the text that Milton himself wrote for his fit readers though few.

The Scene of Literature: Harold Bloom

Harold Bloom’s theory of reading and writing lit­erature centers on the area that Derrida and the structuralists call “intertextuality.” Bloom, how­ever, employs the traditional term “influence,” and presents his theory in opposition against “the anti- humanistic plain dreariness of all those develop­ments in European criticism that have yet to dem­onstrate that they can aid in reading any one poem by any poet whatsoever.” “Poems,” he affirms, “are written by men”; and against “the partisans of writ­ing . . . like Derrida and Foucault who imply . . . that language by itself writes the poem and thinks,” he insists that only “the human writes, the human thinks.” Unlike Stanley Fish, then, Bloom restores the human writer as well as reader to an effective role in the literary transaction. But if Fish’s theory is a half-humanism, Bloom’s is all-too-human, for it screens out from both the writing and reading of “strong” literature all motives except self-concern and all compunction about giving free rein to one’s will to power:

. . . the living labyrinth of literature is built upon the ruin of every impulse most generous in us. So apparently it is and must be—we are wrong to have founded a humanism directly upon literature itself, and the phrase “hu­mane letters” is an oxymoron. . . . The strong imagination comes to its painful birth through savagery and misrepresentation.²⁰

Like many recent critics, Bloom posits a great di­vide in literary history and locates it in the seven­teenth century; his innovation is to account for this division as the change from the relative creative nonchalance of a Homer, Dante, or Shakespeare in “the giant age before the flood” to the acute anxiety of influence suffered by all but a very few poets since the Enlightenment. A modern, and therefore “belated,” poet awakens to his calling when irre­sistibly seized upon by one or more poems of a pre­cursor or father-poet, yet experiences that seizure as an intolerable incursion into his imaginative life- space. The response of the belated writer is to de­ fend himself against the parent-poem by distorting it drastically in the process of reading it; but he can­not escape the precursor, for he inevitably embodies its distorted form into his own attempt at an abso­lutely original poem.

Bloom’s theory, as he points out, is a revision for literary criticism of what Freud sardonically called “the Family Romance.” The relation of reader and poet to his parent-precursor, as in Freud’s Oedipal relationship, is ambivalent, compounded of love and hate; but in Bloom’s detailed descriptions of reading and writing, love enters only to weaken the result of the process, while the aspect of hate, jeal­ousy, and fear is alone given a systematic and cre­ative role to perform. This role is to deploy, with unconscious cunning, a set of defensive tactics, “the revisionary ratios,” which are in fact aggressive acts designed to “malform” the precursor in the attempt to disestablish its “priority” over the latecomer, both in time and in creative strength. “Every act of reading is . . . defensive, and as defense it makes of interpretation a necessary misprision. . . . Reading is therefore misprision—or misreading.” And since “every poem is a misinterpretation of a parent poem,” he concludes that “the meaning of a poem can only be another poem.” “There are no right readings”; the sole alternative is between “weak mis-readings and strong mis-readings.” A weak misreading attempts, although unavailingly, to get at what a text really means in itself; it is the product of an inhibiting timidity, or at best of an excess of “generosity” toward the parent-poet. A misreading is strong, hence creative and valuable, in proportion to the boldness with which the reader’s emotional compulsions are licensed to do violence to the text that he strives to overcome.

It is sometimes argued against Bloom’s theory that his claim, “all reading is misreading,” is in­coherent, on the ground that we cannot know that a text has been misread unless we know what it is to read it correctly. This argument overlooks an inter­esting feature of Bloom’s theory, that is, its quasi- Kantian frame of reference. At times Bloom’s idiom corresponds closely enough to Kant’s to qualify, in Bloom’s terms, as a “deliberate misprision” of Kant’s epistemology. Terms which recur on almost every page in which Bloom discusses misreading are “necessity,” “necessary,” “necessarily,” “must be.” Such terms are to be taken seriously; they signify an a priori necessity. In Bloom’s theory, that is, the compulsive revisionary ratios through which we ex­perience a poem correspond, in Kant’s philosophy, to the cognitive forms of space, time, and the cate­gories that the mind inescapably imposes on all its experience of the world. Consequently Bloom’s reader can only know the phenomenal poem consti­tuted by his own revisionary categories; he cannot possibly get outside these categories to know the noumenal Ding an sich, or what Bloom calls “the poem-in-itself” or the “poem-as-such.”

But Bloom’s aim, he says, is not simply to propose “another new poetics,” but to establish and convert us to “a newer and starker way of reading poems.” The product of this new way of reading is “an an­tithetical practical criticism, as opposed to all the primary criticisms now in vogue.”

Let us give up the failed enterprise of seeking to “understand” any single poem as an entity in itself. Let us pursue instead the quest of learning to read any poem as its poet’s delib­erate misinterpretation, as a poet, of a pre­cursor poem or of poetry in general.²¹

Bloom therefore, like Derrida and Fish, proposes a way of reading a text that will displace the mean­ings that “primary,” or traditional readers have hitherto found in it. As applied in his reading, Bloom’s revisionary ratios in effect function as an inventory of transformers for translating accepted meanings into new meanings; he conveniently pre­sents a one-page table of his transformers which he calls “The Map of Misprision.” And such is the vir­tuosity of these devices that they cannot fail to effect Bloom’s antithetic meanings; in his own re­peated assertion, “It must be so.”

In this analysis I deliberately enact the role which Bloom, in a phrase from Blake, calls “the Idiot Questioner,” whose presence as an aspect of his own mind Bloom recognizes but sternly represses. (In the present instance “the Idiot Questioner” can be translated as a stolid inquirer into the credentials of a critic’s interpretive procedures.) Pursuing such an inquiry, I note that Bloom, in his tetralogy of books on the theory and practice of antithetic criti­cism, sets up six revisionary ratios which he names “clinamen,” “tessera,” “kenosis,” and so on. He goes on to assimilate each of these ratios to a vari­ety of other reinterpretive devices—to a Freudian defense-mechanism; to a concept of the Hebrew Kabbalists; to one of the rhetorical tropes such as synecdoche, hyperbole, metaphor; and to a recur­rent type of poetic imagery. These amalgamated transformers are not only versatile enough to estab­lish each of Bloom’s new readings, but also anti­thetical enough to convert any possible counter­evidence into a confirmation of his own reading.

Take, for example, the Freudian mechanisms of defense—which Bloom calls “the clearest ana­logues I have found for the revisionary ratios”—as he applies them to interpret any poem as a distorted version of a precursor-poem. If the belated poem patently echoes the parent-poem, that counts as evi­dence for the new reading; although, Bloom as­serts, “only weak poems, or the weaker elements in strong poems, immediately echo precursor poems, or directly allude to them.” If the later poem doesn’t contain such “verbal reminders,” that counts too, on the basis of the mechanism of repression—the belated poet’s anxiety of influence has been strong enough to repress all reference to his predecessor. And if the belated poem differs radically from its proposed precursor, that counts even more de­cisively, on the basis of the mechanism of “reaction- formation”—the poet’s anxiety was so intense as to distort the precursor into its seeming opposite. This power of the negative to turn itself into a stronger positive manifests itself frequently in Bloom’s ap­plied criticism. For example, the opening verse paragraph of Tennyson’s Tithonus has traditionally been read as expressing the aged but immortal pro­tagonist’s longing for death. Bloom, however, reads it antithetically as a revision, or

swerve away from the naturalistic affirma­tions of Wordsworth and of Keats. What is absent in these opening lines is simply all of nature; what is present is the withered Ti­thonus. As Tennyson’s reaction-formation against his precursors’ stance, these lines are a rhetorical irony, denying what they de­sire, the divination of a poetic survival into strength.²²

Perhaps so; but it will be noted that the reaction- transformer charters the antithetic critic to speak without fear of contradiction, while stranding his Questioner in a no-win position.

Bloom’s theory, like that of other Newreaders, is self-referential, for he does not exempt his own in­terpretations from the assertion that all readings are misreadings. In his recent books on Yeats and Stevens, he often writes brilliant critiques that com­pel assent from a “primary” critic like myself. The extent of Bloom’s own claim for these readings, however, is that they are strong misreadings, in that they do violence to the texts they address, by virtue of his surrender to his need for autonomy and to his anxieties of the influence exerted on him by his critical precursors. And in lieu of any possible crite­rion of rightness, such readings can be valuable only to the degree that they are “creative or interesting misreadings.” By their strength, he says, such read­ings will provoke his critical successors to react by their own defensive misreadings, and so take their place within the unending accumulation of mis­readings of misreadings that constitute the history both of poetry and of criticism, at least since the Enlightenment.

While acknowledging that his theory “may ask to be judged, as argument,” Bloom also insists that “a theory of poetry must belong to poetry, must be poetry” and presents his work as “one reader’s criti­cal vision” bodied forth in “a severe poem.” Let me drop my role as Idiot Questioner of Bloom’s evi­dential procedures to read him in this alternative way, as a prose-poet who expresses a founding vi­sion of the Scene of Literature. In the main, this has been traditionally conceived as a republic of equals composed, in Wordsworth’s phrase, of “the mighty living and the mighty dead” whose poetry, as Shelley said, “is the record of the best and happiest mo­ments of the happiest and best minds.” In Bloom’s bleak re-vision, the Scene of Literature becomes the arena of a savage war for Lebensraum¹³ waged by the living poet against the oppressive and ever­present dead—a parricidal war, in which each new­comer, in his need to be self-begotten and self- sufficient, undertakes with unconscious cunning to mutilate, murder, and devour his poetic father. The poet’s prime compulsions are like those of the Freudian Id, which demands no less than every­thing at once and is incapable of recognizing any constraints on its satisfactions by moral compunc­tion, logical incompatibility, or empirical impos­sibility. And the poetic self remains forever fixed at the Oedipal stage of development; for Bloom ex­plicitly denies to the poet “as poet” the Freudian mechanism of sublimation, which allows for the substitution, in satisfying our primordial desires, of higher for lower goals and so makes possible the growth from the infantile stage of total self-concern to the mature recognition of reciprocity with other selves. The war of which each poem is a battle­ground is ultimately futile, not only because every poet is inescapably fathered by precursors but also because, according to Bloom, his will to priority over his precursors is, in deep psychic fact, a defense against acknowledging his own human mortality. The conflict, furthermore, is doomed to terminate in the death of poetry itself, for the population of strong poets will soon usurp so much of the avail­able living-space that even the illusion of creative originality will no longer be possible.

In Bloom’s own idiom of rhetorical tropes, one can say of his critical poem about poetry that it is a sustained synecdoche which puts a part for the whole. By this device, and by his subsidiary device of strong hyperbole, Bloom compels us to face up to aspects of the motivation to write and misread poems—self-assertiveness, lust for power and pre­cedence, malice, envy, revenge—which canonical critics have largely ignored. To those of us who yield ourselves to Bloom’s dark and powerful eloquence, the Scene of Literature will never look the same again; such a result is probably the most that any writer compelled by an antithetical vision can hope to achieve. But the part is not the whole. What Bloom’s point of vantage cannot take into account is the great diversity of motives for writing poetry, and in the products of that writing, the abundance of subject-matters, characters, genres, and styles, and the range of the passions expressed and repre­sented, from brutality and terror and anguish, in­deed, to gaiety, joy, and sometimes sheer fun. In sum, what Bloom’s tragic vision of the literary scene systematically omits is almost everything that has hitherto been recognized to constitute the realm of literature.

On Bloom’s critical premises, I am of course open to the retort that I have misread both his criticism and our heritage of literary texts. But knowing from experience Bloom’s geniality to his own critical pre­cursors, I am confident that he will attribute my misreading to an amiable weakness—to my fallacy, that is, of misplaced benevolence.

Newreading and Old Norms

I shall conclude by considering briefly my third question: What makes a text so vulnerable to the diverse things that Newreaders do with it? The chief reason is that our use and understanding of language is not a science but a practice. That is, what we call “knowing a language” is not a matter of knowing that or knowing why, but of knowing how, of having acquired a skill. We are born into a community of speakers and writers who have al­ready acquired this skill, and we in turn acquire it by interplay with these others, in which we learn how to say what we mean and how to understand what others have said by a continuous process of self-correction and refinement, based on what are often very subtle indications of when and in what way we have gone wrong.

The successful practice of language depends on our mastery of linguistic uniformities that we call conventions, or norms, or rules. Linguistic rules, however, differ radically from the rules of chess or of a card-game to which they are often compared. The rules which constitute these games are stipu­lated in an authoritative code to which we can refer in order to resolve disputes. The use and under­standing of language, on the other hand, depends on tacit consensual regularities which are multiplex and fluid; except in very gross ways, these regu­larities are uncodified, and probably uncodifiable. In our practice, therefore, we must rely not on rules, but on linguistic tact—a tact which is the emergent result of all our previous experience with speaking, hearing, writing, and reading the language.

Stanley Fish seems to me right in his claim that the linguistic meanings we find in a text are relative to the interpretive strategy we employ, and that agreement about meanings depends on membership in a community which shares an interpretive strat­egy. But if we set out not to create meanings, but to understand what the sequence of sentences in a lit­erary work mean, when we have no choice except to read according to the linguistic strategy the author of the work employed, and expected us to employ. We are capable of doing so, because an immense store of cumulative evidence provides assurance that the authors of literary texts belonged to the lin­guistic community into which we were later born, and so shared our skill, and the consensual regu­larities on which that skill depends, with some di­vergencies—which we have a variety of clues for de­tecting—which are the result both of the slow change of communal regularities in time and of the limited innovations which can be introduced by the individual author.

When a Newreader, on the basis of his contrived interpretive strategy, asserts that a passage means something radically different from what it has been taken to mean, or else that it means nothing in par­ticular, we lack codified criteria to which we can appeal against the new interpretation; in the last analysis, we can only appeal to our linguistic tact, as supported by the agreement of readers who share that tact. But such an appeal has no probative weight for a reader who has opted out of playing the game of language according to its constitutive regularities; nor is the application of our own in­herited practice verifiable by any proof outside its sustainedly coherent working. All we can do is to point out to the Newreader what he already knows—that he is playing a double game, introduc­ing his own interpretive strategy when reading someone else’s text, but tacitly relying on commu­nal norms when undertaking to communicate the methods and results of his interpretations to his own readers.

We can’t claim that the Newreader’s strategy doesn’t work, for each of these ways of doing things to texts indubitably works. Allowed his own prem­ises and conversion procedures, Derrida is able to deconstruct any text into a suspension of num­berless undecidable significations, Fish can make it the occasion for a creative adventure in false sur­mises, and Bloom can read it as a perverse distor­tion of any chosen precursor-text. These substitute strategies in fact have an advantage which is a prin­cipal cause of their appeal to students of literature. Our inherited strategy, although it has shown that it can persistently discover new meanings even in a classic text, must operate always under the con­straint of communal regularities of usage. Each new strategy, on the other hand, is a discovery procedure which guarantees new meanings. It thus provides freshness of sensation in reading old and familiar texts—at least until we learn to anticipate the lim­ited kind of new meanings it is capable of generat­ing; it also makes it easy for any critical follower to say new and exciting things about a literary work that has been again and again discussed. But we purchase this advantage at a cost, and ultimately the choice between a radical Newreading and the old way of reading is a matter of cultural cost- accounting. We gain a guaranteed novelty, of a kind that makes any text directly relevant to current in­terests and concerns. What we lose is access to the inexhaustible variety of literature as determinably meaningful texts by, for, and about human beings, as well as access to the enlightening things that have been written about such texts by the humanists and critics who were our precursors, from Aristotle to Lionel Trilling.