Electronic Reserve Text: Robert Scholes, "Canonicity and Textuality."

From Introduction to Scholarship in Modern Languages and Literatures, ed. Joseph Gibaldi

Nay here in these ages, such as they are, have we not two mere Poets, if not deified, yet we may say beatified? Shakespeare and Dante are Saints of Poetry: really, if we think of it, canonized, so that it is impiety to meddle with them. The unguided instinct of the world, working across all these perverse impediments, has arrived at such result. Dante and Shakespeare are a peculiar Two. They dwell apart, in a kind of royal solitude; none equal, none second to them: in the general feeling of the world, a certain transcendentalism, a glory as of complete perfection, invests these two. They are canonized, though no Pope or Cardinals took hand in doing it!
-Thomas Carlyle

Beaucoup trop d'heröisme encore dans nos langages; dans les meilleurs--je pense a celui de Bataille--éréthisme de certaines expressions et finalement une sorte d'heroisme insidieux. Le plaisir du texte (la jouissance du texte) est au contr':lire comme un effacement brusque de la valeur guerrière, une desquamation passagere des ergots de l'écrivain, un arret de "coeur" (du courage).
-Roland Barthes, Le plaisir du texte(1)

    For Carlyle, lecturing in 1840, the greatest poets were heroic figures, canonized saints of literature, whose names could readily sustain such adjectives as "royal" and such nouns as "transcendentalism," "glory," and "perfection" (85). Indeed, his lecture itself was called "The Hero as Poet." But for Barthes, writing in the early 1970s, the pleasure of the text emerges only when the writer's impulse toward heroism is in abeyance, when valor and courage are overcome.

A text is, he says, or should be, like a "flippant person who shows his bottom to the Political Father" (84). Nothing saintly or heroic about that. These two statements, I believe, reveal something of the depths beneath our present debate about canonicity and textuality--and something of what is at stake in this debate.

The debate itself is the occasion of the present essay. If the concepts of canonicity and textuality were not currently active in our critical discourse, there would have been no reason for a discussion of them to be included in this volume. It is important to note, then, that these concepts are not merely active in our discourse but active in an oppositional way. Despite some shared meanings, and implications in their etymological past, the two terms now stand in opposition (an opposition embodied in my epigraphs), as names (however crude) for

[139]

two different conceptions of our practice as scholars and teachers: the literary, structured according to the hierarchical concept of canon, and the textual, disseminated around the more egalitarian notion of text. I cannot pretend to impartiality in these debates. I am a textualist. But I shall try, nonetheless, to give a fair idea of what is at stake in this dispute and to avoid excesses of special pleading. Even so, the reader, as always, should be on guard.

Let us begin gently, judiciously, by considering the history of the words canon and text as they have moved through Western culture from ancient times, when they first appeared in Greek, to the present. My survey is partial, of course (perhaps in more than one way),' but I believe that a more ample and detailed study would produce histories much like those I recount. In ancient Greek we find the two words from which the modern English word canon (in its two spellings, canon and cannon) has descended: (kanna) 'reed'; and (kanon) 'straight rod, bar, ruler, reed (of a wind organ), rule, standard, model, severe critic, metrical scheme, astrological table, limit, boundary, assessment for taxation' (Liddell and Scott 2). Like canon, our word cane is also clearly a descendant of the ancient kanna, but its history has been simpler and more straightforward than that of its cognate. However, the second of the two Greek words, kanan, has from ancient times been the repository of a complex set of meanings, mainly acquired by metaphorical extensions of the properties of canes, which are hollow or tubular grasses'; some of which are regularly jointed (like bamboo), and some of which have flat outside coverings. The tubular channel characteristic of reeds or canes leads to the associations of the word canon with functions that involve forcing liquids or gases through a channel or pipe, while the regularity and relative rigidity of canes lead toward those meanings that involve measuring and controlling (ruling--in both senses of that word). And it is likely that the ready applicability of canes as a weapon of punishment (as in our verb to cane, or beat with a stick) supported those dimensions of the meaning of kanan that connote severity and the imposition of power.

In Latin we find the same sort of meanings for the word canon as were attached to the Greek kanan, with two significant additions, both appearing in later Latin. These two additions are due to historical developments that generated a need for new terms. On the one hand, the rise of the Roman Catholic Church as an institution required a Latin term that could distinguish the accepted or sacred writings from all others, so that "works admitted by the rule of canon" came themselves to be called canonical or, in short, the Canon. In this connection we also find a new verb, canonize--are, to canonize. On the other hand, with the importation of gun powder and the development of artillery, the tubular signification of the word led to its becoming the 1;lame, in late Latin, for large guns (Lewis and Short). A common theme, of course, in these extensions is power. It is worth noting here that when the Hebrews became the People of the Book, the word they adopted for their canonical texts was also a word that meant the Law: Torah. As Gerald Bruns argues, the establishment of the Torah as the written Law in Jewish history meant the victory of a priestly establishment

[140]

over the independent voices of the Prophets. In particular, once the Law was fixed in written form, the spoken words of Prophets could not make headway against it, leading to the replacement of prophecy by commentary on the now canonical Book in which the Law was embodied.

For our purposes, the significant point is the way that canon in Latin also combined the meaning of rule or law with the designation of a body of received texts. In its Christian signification, however, canon came to mean not only a body of received texts, essentially fixed by institutional fiat, but also a body of individuals raised to heaven by the perfection of their lives. In the latter signification, the canon referred to an open, not closed, system, with new saints always admissible by approved institutional procedures. This distinction is important because in current literary disputes over the canon, both models are invoked, one on behalf of a relatively fixed canon and the other on behalf of a relatively open one. In any case, our current thinking about canonicity cannot afford to ignore the grounding of the modern term in a history explicitly influenced by Christian institutions. As the epigraph from Carlyle indicates, the conscious use of religious terminology in literary matters is at least a century and a half old.

We must now backtrack a bit to note that the word canon also has a more purely secular pedigree going back to Alexandrian Greek, in which the word kanon was used by rhetoricians to refer to a body of superior texts: (hoi kanones) "were the works which the Alexandrian critics considered as the most perfect models of style and composition, equivalent to our modern term 'The Classics'" (Donnegan). Exactly how the interplay between the rhetorical and the religious uses of the notion of canon functioned two millennia ago is a matter well beyond the scope of this inquiry. What we most need to learn from the ancient significations of canon, however, is that they ranged in meaning all the way from a text possessing stylistic virtues that make it a proper model to a text that is a repository of the Law and the Truth, being the word of God. We should also remember that the word, as a transitive verb, referred to a process of inclusion among the saints.

In the vernacular languages, the meanings of canon found in late Latin are simply extended. In French, for instance, we can find the following in a modem dictionary: canon 'gun, barrel of a gun, cannon; cylinder, pipe, tube; leg (of trousers)'; and canon 'canon. Canon des ecritures, the sacred canon; école de droit canon, school of canon law' (Baker). The French is especially useful in reminding us that the word for gun and the word for the law and the sacred texts are simply branches of a single root rather than two totally different words. That in English we regularized separate spellings (cannon and canon) for the guns and the laws in the later eighteenth century has tended to obscure the common heritage of both these spellings in the ancient extensions of a word for reed or cane. In English the most relevant meanings of the word canon for our purposes are these:

canon 'a rule, law, or decree of the Church; a general rule, a fundamental. principle; the collection or list of the books of the Bible accepted by the Christian


[141]

Church as genuine and inspired; hence, any set of sacred books; a list of saints acknowledged and canonized by the Church' (OED).

The nature of the connection between the Christian canon and the literary canon is crucial to our understanding of the present disputes about canonization. This connection was made most forcibly and enduringly in English letters by Mathew Arnold, as Northrop Frye pointed out more than thirty years ago in an exemplary discussion of Arnold's touchstones that laid bare Arnold's motivation:

When we examine the touchstone technique in Arnold, however, certain doubts arise about his motivation. The line from The Tempest, "In the dark backward and abysm of time,'; would do very well as a touchstone line. One feels that the line "Yet a tailor might scratch her where'er she did itch" somehow would not do, though it is equally Shakespearean and equally essential to the same play. (An extreme form of the same kind of criticism would, of course, deny this and insist that the line had been interpolated by a vulgar hack.) Some principle is clearly at work here which is much more highly selective than a purely critical experience of the play would be.

Here we should pause to notice that Frye's notion of a "purely critical experience" conserves much of the Arnoldian project--which remains at the 'center of our present critical debates. We shall return to this point. But: first, let us continue with Frye's next paragraph:

Arnold's "high seriousness" evidently is closely connected with the view that epic and tragedy, because they deal with ruling class figures and require the high style of decorum, are the aristocrats of literary forms. All his Class One touchstones are from, or judged by the standards of, epic and tragedy. Hence his demotion of Chaucer and Bums to Class Two seems to be affected by a feeling that comedy and satire should be kept in their proper place, like the moral standards and the social classes which they symbolize. We begin to suspect that the literary value judgments are projections of social ones. Why does Arnold want to rank poets? He says that we increase our admiration for those who manage to stay in Class One after we have made it very hard for them to do so. This being clearly nonsense, we must look further. When we read "in poetry the distinction between excellent and inferior. ..is of paramount importance. ..because of the high destinies of poetry," we begin to get a clue. We see that Arnold is trying to create a new scriptural canon out of poetry to serve as a guide for those social principles which he wants culture to take over from religion. (21-22)

Like so much in that extraordinary book of Frye's, these crucial paragraphs opened the way to all our subsequent discussions and disputes about the literary canon. In particular, Frye made literary scholars and critics aware of two things that had been overlooked or concealed during the academic hegemony of the New Criticism. First, that' "literary value judgments are projections of social ones" (though he tried to reserve for himself a field of "purely critical experi-


[142]

ence"). And, second, that the Arnoldian tradition in criticism involved "trying to create a new scriptural canon out of poetry." The way we currently use the word canon in literary studies is very much the way we learned to use it from Northrop Frye. And it was also Frye who-when very few students of literature thought of calling their enterprise "literary theory"-told us that "the theory of literature is as primary a humanistic and liberal pursuit as its practice" (20). When Frye wrote, the word canon was used in literary studies mainly to refer to the body of texts that could be properly attributed to this or that author (a significance that is acknowledged in the Supplement to the OED). The MLA annual bibliographies are full of articles with titles like "The Shakespeare Canon" or "The Defoe Canon." Since Frye, however, and especially in the last decade, literary scholars have come to use the word as the name for a set of texts that constitute our cultural heritage and, as such, are the sources from which the academic curriculum in literature should be drawn. This situation is full of complexities and perplexities. We shall return to the problems of literary canonicity after further complicating matters by considering the cultural history of the word text.

This word has a history that is perhaps even more interesting than that of canon, in that it has been susceptible to a greater range of fluctuations in meaning-a process still very much alive. The variability (or duplicity) of the word is apparent even in its Greek beginnings:

(tik, tikos) Of or for childbirth, a medicine used for women lying in, a KOV (pharmakon).

TIKTOO (tikto) bring into the world, engender; of the father: beget; of the mother: bring forth; of the earth: bear, produce; metaphorical: generate, engender, produce.

    TEKTOOV (tekton) worker in wood, carpenter, joiner; generally: any craftsman or workman; metaphorically: maker, author.

TEXVll (techne) art, skill, cunning of hand; cunning in the bad sense: arts, wiles; an art or craft; a method, set of rules, or system of making or doing, whether in the useful arts or the fine arts; work of art, handiwork; treatise (on grammar or rhetoric).

(adapted from Liddell and Scott)

The single theme that runs through all these words and their meanings is that of creation. In this, the word tikto appears central, with its fundamental meaning of physical or natural production (of children and the fruits of the earth) and its metaphorical extension to cover all kinds of production. Around this central core of meaning some curious and interesting extensions play. First, tik, the pharmakon, or drug, used to make childbearing easier for women. And here, perhaps, we should note that the two opposed meanings of pharmakon are very similar to those of our modem English word drug, which refers to both harmful and beneficial kinds of ingestible substances. Jacques Derrida has made much of this in "Plato's Pharmacy," but the double meaning of the word was fully noted

[143]

regularly in Greek lexicons before Derrida' s influential essay. 'For our purposes it is important to note another play in the meanings of tik and tikto: on the one hand, the natural--begetting, engendering, and bringing forth--and, on the other hand, the artificial: the drug that must mitigate the "unnatural" pains of the natural process, or come to the aid of nature in this instance.

The word tekton extends the meanings of tikto in the direction of artifice or craft, by pointing first toward carpentry and other physical (though not natural) acts of making and .finally to mental creation, production, or authorship. One facet of this extension is that it tends to obliterate female production as it moves away from nature and toward art, craft, and authorship. In the earliest formulations the role of woman as child .bearer and the earth as feminine bearer of fruits (regularly portrayed as a goddess rather than a god) were dominant. But gradually the gender emphasis shifted. In Greek culture, carpentry-a male occupation-assumed a central position in the paradigmatic structure of this word and its meanings. In the word that named the maker's skill, techne, there was some room for female handicraft, but the general pattern of thought embodied in this language seems to have aligned women with nature as primitive producers and men with culture as producers of consciously constructed objects of daily use and art. The word techne itself was frequently used to refer to metal work, ship building and other trades associated with male workers. This word, like pharmakon, has its pejorative sense, too, referring to guile or cunning. Finally, and this is especially relevant to our concerns, the meanings of the word were extended to refer to the methods or systems of the developing verbal disciplines of grammar and rhetoric.

When we pick up the history of these terms as they appear in Latin, we find that the notion of joining as in carpentry, or constructing as in metalwork, has been replaced by weaving, as the guiding concept of textual fabrication. We can see also that the extension from material handicraft to verbal construction is reinstated and extended, taking on specific references to verbal composition or style (as opposed- to weaving). In Latin we find texo, texere 'weave; to join or fit together any thing; to plait, braid, interweave, interlace, intertwine; to construct, make, fabricate, build; to compose' and textum 'that which is woven, a web; that which is plaited, braided, fitted together, a plait, texture, fabric'; figurative: 'of literary composition, tissue, texture, style' (Lewis and Short). The meanings related to weaving. and woven fabric were to remain with these words and with many of their descendants (texture, textile, etc.), but the verbal extensions of meaning toward literary style and composition became more and more important in the history of the word textum itself. In particular, the masculine textus, which appears first in poetry and in post-Augustan prose, seems to have become the favored form for the verbal and stylistic meanings of the word; and in medieval Latin, in particular, we find the masculine textus carrying specific adaptations for Christian verbal functions, including, finally, specific reference to the New Testament as the Text: textus 'text, wording, contents of speech or writing; charter; Gospel book' ("Dedit rex Serenissimus Augustus quattuor

[144]

1. Acquaintance with "letters" or books; polite or humane learning; literary culture. Now rare and obsolescent. (The only sense in Johnson and in Todd 1818).

2. Literary work or production; the activity or profession of a man of letters; the realm of letters.

3. a. Literary production as a whole; the body of writings produced in a particular country or period, or in the world in general. Now also in a more restricted sense, applied to writing which has claim to consideration on the ground of beauty of form or emotional effect.

We should be attentive to a number of things about this word that has been so important to the enterprise of the Modern Language Association and to the individuals and institutions connected with it. First, we should note that in Samuel Johnson's Dictionary and as late as Henry John Todd's version of it (1818), literature referred to a learning or culture possessed by an individual rather than to a set of lettered objects or written texts. That is, it was customary to say that a person had literature rather than that a person read literature. Second, the word also came to name whatever texts were produced by a person (clearly thought of as male) who possessed "literature" or book learning. Third, the word, when it did refer to the written or printed texts themselves, included everything, the whole body of writings of a time or place--or of the "world." Fourth, in a process that emerges fully only at the end of the eighteenth century, literature came to mean belles lettres, as opposed to other kinds of writing.

In 1762, in what became an enormously influential textbook in American colleges, Lord Kames proposed a science of criticism. His Elements of Criticism was meant to apply to all of what he called the "fine arts" but was focused mainly on the arts of language as they are displayed in poetry, drama, and some prose. So far as I can tell, in over a thousand pages, he found no occasion to use the word literature. It is also true, of course, that for him the whole art of language lay in finding the proper means for the expression of human sentiments and passions. For him, the imagination was no more than one human faculty, and by no means was it the most important faculty with respect to the fine arts:

Such is the nature of man, that his powers and faculties are soon blunted by exercise. The returns of sleep, suspending all activity, are not alone sufficient to preserve him in vigor. During his waking hours, amusement by intervals is requisite to unbend his mind from serious occupation. The imagination, of all our faculties the most active, and not always at rest even in sleep, contributes more than any other cause to recruit the mind and restore its vigor, by amusing us with gay and ludicrous images; and when relaxation is necessary, such amusement is much relished. But there are other sources of amusement beside the imagination.

(1: 337; typography modernized)

What Kames meant by imagination was simply the "singular power of fabricating images independent of real objects" (3: 386). He has neither our notion of

[146]


literature nor our sense of imagination. The momentous linking of these two terms, which was necessary for our modem idea of a literary canon to be developed, is primarily if not exclusively the work of Romantic writers, especially the Jena Romantics and their followers (among whom we must still, with whatever mixture of emotions, consider ourselves). This Romantic notion of imagination as the primary quality of literature was, of course, given its special potency in English by Coleridge. Coleridge himself thought of literature in the pre-Romantic manner. When, in the Biographia Literaria, he sought to advise young men not to take up literature as a "trade" (129), his list of literary men who had other careers included Cicero, Xenophon, Thomas More, Richard Baxter (the Puritan divine), and Erasmus Darwin {the poet of evolution). In its fully Roman.. tic usage, however, literature took over the meanings formerly assigned to poetry alone, with the added insistence on the transcendental powers of imagination. The crucial distinction thus fell not between poetry and prose but between imaginative writing and writing that lacked this divine spark. This distinction was already nascent in Kames's separation of the fine arts from the useful arts, but the addition of imagination as the decisive criterion made the distinction more invidious. Imaginative writing-and Coleridge made this especially clear with his description of primary and secondary forms of imagination-connected certain texts directly to Divinity or Absolute Spirit. Materially useless but spiritually precious texts: that is what literature had come to mean when Mathew Arnold conceived the project of replacing dogma with literature-the project criticized by Frye in the "Polemical Introduction" to his Anatomy of Criticism. We must be aware, however, that this Amoldian project was extended by writers as diverse as T. S. Eliot and I. A. Richards, as well as by the New Critics teaching in American universities after World War II. And it was alive and well in the work of Northrop Frye himself, as Barbara Hermstein Smith vigorously demonstrates in an essay in an important collection entitled Canons (ed. von Hallberg), to which we shall return shortly.

By the end of the nineteenth century two simultaneous processes (or two facets of the same process) had led to the establishment of a literary caQon. One of these was the separate, superior status claimed for works of verbal imagination, which, thus empowered, constituted a literary canon. The other was the professionaliz-ation of teaching in the newly established (and in particular the American) universities and graduate schools. As the study of the modem literatures, especially English, replaced the classics and oratory at the center of the humanist curriculum at the end of the nineteenth century, authors such as John Locke and Francis Bacon, who had loomed large in our early college curricula, gave way to writers who were literary in the now accepted sense of that word. For a time, criticism struggled against philology for power within the new English departments, each with its own canon of proper texts for study. In this struggle philology, which was really an attempt to carry on classical studies without classical texts, was doomed because its canon was based on antiquity, privileging

[147]

texts in Old and Middle English, while criticism could select its canon on the basis of "pure literary merit."

Actually, of course, this selection required an institution to debate and ratify canonical choices, and, at the proper moment, the institution came into being. It was called the Modem Language Association. In the professionalization of literary studies, the canon supported the profession and the profession supported the canon. Likewise, the canon supported the literary, curriculum and the curriculum supported the canon. The curriculum, in literary studies, represented the point of application, where canonical choices were tested in the crucible of student response.. Works that proved highly teachable (like Shakespeare) remained central in the canon as well as in the curriculum. The revival of John Donne may have been begun by Herbert Grierson and T. S. Eliot, but it continued with such notable success because the New Critics found Donne perfectly suited to their pedagogy, their curriculum, and, hence, their canon. Authors who proved less amenable to critical exegesis (Oliver Goldsmith, for instance, who once bulked large in the American curriculum) were quietly allowed to drift out of that curriculum and, hence, out of the canon.

The most important thing about these processes is not that they went on but that they went on unnoticed. Until the last few decades they were seen as "natural"-or even as not occurring at all. What has happened to literary studies in those decades is a part of larger cultural happenings that can be described (and deplored, if you like) as the politicization of American life. Once upon a time we believed that if the best men (yes) were appointed 'to the bench, we would get the best judicial decisions. Now, we know that one set of appointments to the Supreme Court will give us one set of laws and another set of appointments will give us others. What is happening is part of the evolution of a democratic society. With respect to the literary canon, Frye's statement about Arnold's touchstones was a political bombshell: "We begin to suspect that the literary judgments are projections of social ones." Which is to say that the literary canon is a social, and therefore a political, object, the result of a political process, like so much else in our world.

Thus the battle lines were drawn, and the battle is still in progress. On the one hand are those who defend a universal standard of literary quality (among whom we may find Frye himself), and on the other are those who argue that standards are always relative, local, and political. There are militant universalists and there are laissez faire or pragmatic universalists, which complicates matters.

And the relativists are divided, also, into champions of different excluded groups, seeking canonical status for their own class of texts, and. anarchists or absolute, relativists, who would undo all canons and standards if they could. At this point I suppose I should run up my own flag, since I cannot pretend to neutrality on these matters. I do not see how anyone can teach without standards, but I cannot find any single standard for determining the worth of a text. I do not, that is, believe in literature either as a body of spiritually informed texts or as a

    [148]

universal standard of textual value. I have lost my faith (and, yes, I once had it) in literature as an institution. It is not my intention here, however, to preach literary atheism or to make my position central to this discussion. I mention it merely as a bias the reader may wish to discount, as I move on to what I take to be the best single focus of current canonical disputes, Robert von Hallberg's collection of essays, made under the auspices of Critical Inquiry in 1984.

We have already had occasion to note Barbara Herrnstein Smith's rich and complex argument, in von Hallberg's volume, against the transcendental or universal valorizing of 'our canonical texts-an argument I must summarize rather brutally here as suggesting that what supports canonical texts is not so much their own merits or relevance to our purposes as it is the way they have already been inscribed into an intertextual network of reference. That is, they are culturally important because they have been culturally important, a situation from which both the inevitability of change and its equally inevitable slowness may be inferred. Smith's essay has been positioned first in the volume, with the result that many of the others can be read as amplifications, qualifications, or counterstatements to it, though most of them were written quite independently. The most direct counterstatement is that made by Charles Altieri in an essay called "An Idea and Ideal of a Literary Canon."

Altieri argues that only an appeal to "a general high canon" can provide the "authority" we need to resist local and specific abuses of power. In making this case, he suggests that "if we want to measure up to a certain kind of judgment" we must turn to "those models from the past that have survived such judgments" (57, 55, 56). Drastically simplified, the argument is that we cannot have ideals such as justice unless we ground them in texts that have been judged ideal. Whether literary judgments and ethical or political justice have enough in common to support this necessary connection is a problem that the essay never quite resolves.

Altieri argues that works in the high literary canon are there because they have three qualities: the forceful and complex presentation of moral categories, semantic scope and intensity, and either technical innovation or wisdom and ethical significance. He further argues that the submissive study of such texts is necessary for us to develop our ability as readers and judges of our own culture. He does not quite complete the argument by concluding that study of the canonical texts is the only path to an ethical and effective life, but he certainly implies this.

It would be possible to criticize this essay by probing into its internal contradictions and terminological slippages, as exemplified by the lumping together of technical innovation and wisdom in the same category. I prefer, however, to make two, more general observations. One is that it leaves us wondering how writers and thinkers who had a very limited canon themselves ever became canonical. How did Homer, Aeschylus, and Plato, for instance, acquire the qualities that Altieri would attribute to them, since they had scarcely any access to the canon now held to be so indispensable. Socrates apparently knew his Homer but most of what he had learned he seems to have learned from

[149]

the Sophists. My second point is that I see little evidence that prolonged study of canonical literary texts has made professors of literature either wiser or more virtuous than anthropologists, say, or carpenters. A more complex counterposition to Altieri's, however, is to be found within the volume of Canons itself. I refer to John Guillory's discussion "The Ideology of Canon...Formation: T. S. Eliot and Cleanth Brooks."

Just as Altieri is not responding directly to Smith,' Guillory is responding not directly to Altieri but to the general position that Altieri represents. His discussion traces the path from Eliot's reconstruction of the canon of English literature in his early essays to the institutional underwriting of that very canon by the New Critics through such books as Cleanth Brooks's Modern Poetry and the Tradition (1939) and The Well- Wrought Urn (1947). Guillory shows how this process functioned as a subtle and more attractive alternative to Mathew Arnold's attempt to replace dogma with literature. He does this by reminding us that what Eliot's essays suggested and the New Critics instituted was the replacement of doxa with paradox. Under this regime, canonical texts were seen not as repositories of truth and beauty but as embodiments of a discourse so ambiguous that it could not be debased and applied to any practical or dogmatic end. The study and teaching of the new canon of specifically noncognitive texts would of necessity fall to those trained to show that they are canonical precisely because they resist reduction to doxa or dogma. Those who understood this, either as teachers or students, became members of what Guillory calls a "marginal elite," an elite based on a canon of texts that aspired neither to scientific nor didactic status but to a literary purity defined explicitly as the absence of such ambitions:

Nevertheless, literary culture has aspired to canonical consensus, an illusion reinforced by the cognitive silence of the literary work, the silencing of difference. Very simply, canonical authors are made to agree with one another; the ambiguity of literary language means nothing less than the univocity of the canon. I now want to examine this rule of canonical self, identity as it governs the institutional dissemination of literature.. Eliot's fantasy of orthodoxy passes into the university (both as an ideology of the marginal elite and as an instruction in the marginal relation of the poem to truth. (350)

To document his case, Guillory looks at Brooks's crucial treatment of Donne's poem "The Canonization"-a truly overdetermined choice by all con... cerned, including myself. Guillory sees Brooks as basing the poem's canonical status as poetry on its ability to offer and to inhabit a realm removed from and "above" the world of power and cognitive assertion:

[T]he ideological function of Brooks' reading concerns the demarcation of a spiritual realm between the crudities of power and the crudities of fact. The spiritual realm is defined by the audience the essay addresses: the auditors are conceived at a moment of apostolic succession, at just the moment of transition between Eliot and Brooks, as representative figures of literary culture. The incognito clergy is

[150]

relocated within a visible social structure: the pedagogical institution. The idealized reading of the lovers' withdrawal must be understood as symptomatic of the professional commitment to the preservation of value: just as the lovers institute love in their act of renunciation, so it is the marginality of value which is both deplored and established by the idealization of literature. (356)

As he has observed earlier in the essay, "in teaching the canon, we are not only investing a set of texts with authority; we are equally instituting the authority of the teaching profession" (351). Guillory's point is 'partly that we should strive for a certain critical distance in determining our own stake in maintaining a canon, but he also means to suggest a possible direction out of New Critical orthodoxy into a "state of heterodoxy where the doxa of literature is not a paralyzed allusion to a hidden god but a teaching that will enact discursively the struggle of difference" (359-60).

This is a brave conclusion to an elegant essay, but it seems to me dangerously close to simply replacing the New Critical canon with a new set of texts privileged by their heterodoxy or their enactment of the "struggle of difference." The problem, I believe, is that "difference" is itself a notion that has gained its privileged position in recent American theory partly because it allowed an easy transition from New Critical paradoxes. Believing, for instance, that the best texts are those "that deconstruct themselves" is just a step from equating paradox with literary value. It is a useful step, to be sure, but my own feeling is that something simpler is necessary: not texts that embody difference but just different texts. Perhaps this is what Guillory means, but 1 am wary of the tendency of American literary deconstruction to lead back to a canon more traditional than even that of the New Critics. Certainly the compatibility of a certain sort of deconstruction with traditional literary values is writ large in J. Hillis Miller's much quoted statement, "I believe in the established canon of English and American literature and in the validity of the concept of privileged texts. 1 think it is more important to read Spenser, Shakespeare, or Milton than to read Borges in translation, or even, to say the truth, to read Virginia Woolf" (qtd. by Froula in von Hallberg 152).

The issue of canonicity turns finally on the notion of literature itself, as, for instance, Arnold Krupat suggests in his discussion "Native American literature and the Canon," in which he makes the following point:

In our own time, the canon is established primarily by the professoriate, by teacher-critics who variously-passively or actively but for the most part--support the existing order. As Leslie Fiedler has remarked, "Literature is effectively what we teach in departments of English; or, conversely, what we teach in departments of English is literature" ([Fiedler and Baker] 73).

Roland Barthes has offered a similar observation. "The 'teaching of literature,' " Barthes said, "is for me almost tautological. Literature is what is taught, that's all" (Doubrovsky 170). What the pedagogical canon includes from the past and from current production generally

[151]

and substantially works to ratify the present and to legitimate an established hegemony. (310)

Krupat's position is necessary to his argument that by attending to .Native American works in the curriculum, we will also establish them in the canon. Others in the volume who would not deny the connection of the curriculum to the canon would see the mechanisms of canonization as being more complicated. Alan C. Golding, for instance, brings to our attention the way that over the past century and a half American poetry anthologies have played a vital role in shaping the canon, but he, too, notices. that over the decades curricular needs have become more influential even on the anthologies. Similarly,' Richard M. Ohmann, examining \ contemporary mechanisms of canonization, describes a complicated process but suggests that the greatest power lies with a class that stretches from the marginal elite of the universities to a less marginal elite group in the magazines and publishing houses.

In another important essay in Canons ("When Eve Reads Milton: Undoing the Canonical. Economy") Christine Froula, reminding us of how the canon has functioned as an instrument of domination, argues that the proper answer is both to add new textual. voices to our curricula and to read the old texts in a different way:" (19)

Few of us can free ourselves completely from the power ideologies inscribed in the idea of the canon and in many of its texts merely by not reading "canonical" texts, because we have been reading the partriarchal "archetext" all our lives. But we can, through strategies of rereading that expose the deeper structures of authority and through interplay with 'texts of a different stamp, pursue a kind of collective psychoanalysis, transforming "bogeys" that hide invisible power into investments both visible and alterable. In doing so, we approach traditional texts not as the mystifying (and self-limiting) "best" that has been thought and said in the world but as a visible past against which we can teach our students to imagine a different future. (171)
This volume on canons, which in turn points to other important discussions of the question, is certainly the place for later inquirers into canonical matters to begin. My own conclusion, however, after examining both this book and many of the texts cited therein, is that as long as we refrain from challenging the hegemony of literature itself, the essentially conservative and patriarchal processes of canonization will continue to function in much the same way. In a tendentious and vigorous essay in this volume, "The Making of the Modernist Canon," for instance, Hugh Kenner develops a literary standard that enables him to relegate Virginia Woolf and William Faulkner to secondary status as "provincials" and to largely ignore such writers as Gertrude Stein, Dorothy Richardson, Jean Rhys, and Djuna Barnes as beneath his notice. They are simply not high enough or modernist enough to be visible. As Guillory reminds us, and


[152]

as Altieri implies, our present canonical situation is what it is precisely because of the passage of literature- through modernism. What Kenner demontstrates, though it is not his explicit intention, is that Romanticism filtered through avant-gardism (which is the formula for modernism) yields a literary canon in which the oppressively absolutist and patriarchal tendencies of canonization are more visible than ever before.

Specifically, the modernist canon installed by Pound, Eliot and their followers in English departments in this country called for, on the one hand, an aggressively innovative approach to literary form and, on the other, a learned appropriation of mythology and poetical texts drawn from the ancient, medieval, and Renaissance canons. This modernist notion of literary excellence worked powerfully (and "naturally") against women who had no easy access to classical education and for whom the traditional verbal forms were in themselves experimental, in that they had never before been used to express the experience of women in a world of possibilities opening all too slowly-but opening nevertheless at the end of the nineteenth century. A novel like May Sinclair's Mary Olivier (1919), for instance, shows us both how painfully difficult and how profoundly radical it was for a provincial woman to adapt to her situation the bildungsroman form recently energized by D. H. Lawrence and James Joyce. The result is a novel that is powerful and important but will never match the canonical works of the modernist masters by their own literary criteria. And this is just one example of countless texts in which marginal voices have found expression in forms too humble for canonization or already discarded in the relentless modernist search for innovation.

My point is not that modernism itself was some sort of error but that it represented the culmination of a process of literary canonization begun by the Romantics--a process that is now unworkable because it has become too visible and because we have at last become aware of its social costs. In response to this situation I (and it must be obvious that these are not the conclusions of the MLA itself) would argue that we need to scrutinize critically and if possible undo the privilege we have so long granted to the notion of literature itself. This is why the opposition of text to work and of textuality to literature is so important.

As we have seen," the history of the word text and its cognates is not so different from that of canon. Both sets of signifiers passed through alliances with the significations offered by history; both took on Christian significance in the Middle Ages; and both have some specifically verbal significations in our own world. But where canon has persisted in its exclusionary and hierarchical functions, allowing only such qualifications as Alastair Fowler's potential, accessible, and selective canons (all literature, literature currently in print, and approved literature-discussed by Golding, in von Hallberg 279), text has acquired, especially at the hands of French theoreticians like Roland Barthes and Jacques Derrida, some new significations that are programmatically subversive of canonincal distinctions. The new meanings of text are usefully summarized in the introduction to Dominick LaCapra's Rethinking Intellectual History:


[153]


Text" derives from texere, to weave or compose, and in its expanded usage it designates a texture or network of relations interwoven with the problem of language. Its critical role is to problematize conventional distinctions and hierarchies, such as that which presents the text as a simple document or index ofa more basic, if not absolute, ground, reality, or context. Yet the use of the notion of the text (or of textuality) to investigate a relational network inevitably raises the specter of "textual imperialism" or "pantextualism." When the notion of the text is itself absolutized, one confronts the paralyzing and truly abstract sort of interpretative bind that the appeal to the notion of textuality was intended to avoid or at least to defer... (19)

This statement is helpful in two ways. It directs our attention to some primary features of current notions of textuality, and it warns us about the abuse of such notions. The important primary features are (1) the way in which textuality insists on the connection or "network" linking any particular bit of language to other bits and to the whole network, and (2) the way that this particular linkage supersedes or forestalls any limitation of the meaning of a particular textual object to some nontextual referent, author, or situation that could entirely regulate the flow of meanings evoked by that object. The warning LaCapra offers is also important. He reminds us that textuality itself is a metaphor that can be used and abused. One abuse is a denial of referentiality so absolute as to become a mere formalism, a problem addressed by Terry Eagleton in Literary Theory. Eagelton objects to the way that deconstructive critics of the Yale school have "colonized" history itself, viewing "famines, revolutions, soccer matches, and sherry trifle as yet more undecidable 'text' " (146). Fredric Jameson has tried to mediate between these positions in an important passage of The Political Unconscious, arguing that "history is not a text, not a narrative, master or otherwise, but that, as an absent cause, it is inaccessible to us except in textual form" (35). In Jameson's language, textuality refers to a collaboration between language and the human unconscious that always distances us from reality with; out ever replacing that reality.

This poststructuralist notion of textuality is based on the semiotic and deconstructive projects of Charles Sanders Peirce, Ferdinand de Saussure, and Jacques Derrida, in which human interaction with the world is understood as always mediated by signs that can be interpreted only by connecting them to other signs, without ever leading to some final resting place of interpretation that might be called Reality or Truth. This we may think of as the strong sense of the word textuality as it is used in contemporary literary theory. It is this sense to which Derrida referred in his famous statement about there being no outside to textuality: "il n'y a pas de hors-texte" (Of Grammatology 158), 'which means that we can "make sense" of things only by establishing our own connections within the network of textuality that enables our thinking and perceiving in the first place--as I have just done by interpreting Derrida's phrase. As LaCapra warns us, however, it is a mistake to take this metaphor of textuality literally a mistake that can only be made by ignoring the way that the idea of the "literal"

[154]

is ruled out by the metaphor itself. The function of this sense of textuality, then, resituate the reading or interpreting of texts in a more creative or, as Derrida , "exorbitant" mode. But what is a text?

Here, Roland Barthes is our liveliest guide--and within the metaphor of textuality, liveliness contends with reliability (some would say supersedes itnot I) for the most important attribute of guidance. In one of his most influential essays, "From Work to Text," Barthes uses the opposition named in his title as a way of situating his new criticism in opposition to the old. His method of accomplishing this at first seems to align Barthes's nouvelle critique with the American New Criticism as John Guillory described it. Barthes tells us

the Text is that which goes. to the limit of the rules of enunciation (rationality, readability, etc.). Nor is this a rhetorical idea, resorted to for some "heroic" effect: the Text tries to place itself very exactly behind the limit of the doxa (is not general opinion-constitutive of our democratic societies and powerfully aided by mass communications--defined by its limits, the energy with which it excludes, its censorship?). Taking the word literally, it may be said that the Text is always paradoxical. (Image 157-58)

What makes Barthes's formulation quite different and in certain respects opposed to American New Criticism, however, is his specific opposition of the text to work:

The difference is this: the work is a fragment of substance, occupying a part of the space of books (in a library for example), the Text is a methodological field. ... [T]he work can be seen (in bookshops, in catalogues, in exam syllabuses), the text is a process of demonstration. ..; the work can be held in the hand, the text is held in language, only exists in the movement of a discourse (or rather, it is Text for the very reason that it knows itself as text); the Text is not the decomposition of the work, it is the work that is the imaginary tail of the Text; or again, the Text is experienced only in an activity of production. It follows that the Text cannot stop (for example on a library shelf); its constitutive movement is that of cutting across (in particular, it can cut across the work, several works).

...The author is reputed the father and the owner of his work: literary science therefore teaches respect for .the manuscript and the author's declared intentions, while society asserts the legality of the relation of author to work. ... As for the Text, it reads without the inscription of the Father. Here again, the metaphor of the Text separates from that of the work: the latter refers to the image of an organism which grows by vital expansion, by "development" (a word which is significantly ambiguous, at once biological and rhetorical); the 'metaphor of the Text is that of the network; if the Text extends itself, it is as a result ofa combinatory systematic. ...Hence no vital "respect" is due to the Text: it can be broken. ..j it can be read without the guarantee of its father, the restitution of the intertext paradoxically abolishing any legacy. It is not that the Author may not "come back" in the Text, in his text, but he then does so as a "guest." ...He becomes, as it

    [155]


were, a paper-author: his life is no longer the origin of his fictions but a fiction contributing to his work: ..; it is the work of Proust, of Genet which allows their lives to be read as a text. (Image 157, 160-61)

In American New Criticism, the boundedness of the literary work, its organic unity, its status as a "verbal icon" supported the role of the literary critic (or teacher) as a quasi-priestly exegete, introducing outsiders into the hermetic mysteries of literature. The notion of textuality weakens the boundaries of the individual textual object and reduces the strength of its connection to an individual author or a specific situation, in order to emphasize the intertextuality of every such object and the freedom of the reader to establish connections among many texts at many levels. This notion also changes the critic or teacher from a figure of automatic authority to one reader among others, whose performance of the reading act will have to be its own justification. The criteria for judging such performances may well include (I would say, must include) such traditional interpretive virtues as learning, attention to detail, and intensity of thought, but they will now also include range, creativity, and even exorbitance along with the traditional virtues. With this extension of the reader's range comes also a new freedom to take with equal seriousness (and playfulness) texts outside the "selective canon" of literature and indeed outside "literature" itself.

Perhaps the simplest and most radical implication of the concept of textuality is that it breaks down the barriers between verbal objects and other kinds of signification. The word text is useful--and indeed necessary-if we are to discuss the common semiotic properties of pictures, films, plays, operas, jokes, graffiti, poems, songs, stories, speeches, advertisements, novels, essays, and other. ..other what? Well, other texts, of course. A text .is a cluster of signs or potentially signifying entities that can be connected by an act of reading to other such clusters.

A few years ago when a consortium of teaching organizations (the English Coalition) sought foundation support for a conference on the future of English studies from kindergarten to graduate school, one powerful foundation refused' to consider supporting the proposal until the word text in the proposal was replaced everywhere by the word literature. This, my friends, is a true story, and it suggests that these matters, which may seem like trivial questions of terminology that concern only scholars and teachers, really do have political and economic consequences. In this opposition, text is aligned with the extension of democratic social, economic, and political processes and canon with the maintenance or recovery of more hierarchical structures. At their extremes, these two positions may imply anarchy and absolutism. In the middle, where most of us work\ and struggle, they may only be Jeffersonianism and Federalism. In any case, we may be certain that concepts of canonicity and textuality are themselves imbedded in the larger processes of our social text.

One final word. While I still have some control over this collection of words--before, that is, they enter the web of textuality even further--I would


[156]

like to make a disclaimer. The notion of text deployed here is no panacea. It should be, at best, a stimulus to rethinking our enterprise. There is, I suppose, a possible curriculum of textuality that many writers on these matters could specify. And what, you may well ask, prevents that set of authors and works from becoming a canon just as exclusive and oppressive as the old one? I have two tentative answers to this very pertinent question. One is that, insofar as what we are considering is a set of theoretical writings, they are bound to be largely subsumed (Aufgehoben) by later theoretical writings. The other. is that to the extent that we have really made it legitimate to consider-and study in our courses-any kind of textual object from graffiti to The Making of Americans, we have gone beyond canonization, because a canon requires that there be much more outside of it than inside. Without a canon, of course, we shall have to live our academic lives on an ad hoc basis.- Individually, we may all be governed by habit and inertia more than we should', but perhaps we won't think that our hobby horses are cast in the mold of some Platonic Pegasus.3

Brown University

NOTES
l"Far too much heroism still in our languages; in the best-I think of Bataille'serethism [excessive irritability] of certain expressions and finally a sort of insidious herolsm. The pleasure of the text (the joy of the text) is on the contrary like an abrupt erasure of warlike valor, a momentary desquamation [scaling off] of the writer's spurs, a stoppage of 'heart' (of courage)" (Barthes, Plaisir 50; trans. mine).

2 In citing dictionaries and lexicons, I do not give page numbers because the words serve as their own locators. I also abridge and omit freely, in the interests of controlling what still seems like an ungainly amount of philological matter, though I think it is necessary to the discussion and may be useful beyond the immediate context.

3 In preparing this essay I have received any amount of useful advice from anonymous rea~ers and some very specific and extremely helpful criticism from John Murchek of the University of Florida. While teaching a course in canonicity, he located and drew my attention to the quotation from Carlyle that serves as the first epigraph. He also pointed to a number of weaknesses in an earlier draft, which I have done my best to remedy. For all this assistance I am extremely grateful.


WORKS CITED AND SUGGESTIONS FOR FURTHER READING

Altieri, Charles. "An Idea and Ideal of a Literary Canon." V on Hallberg 41-64.
Baker, Ernest A. Cassell's New French Dictionary. New York: Funk, 1930.
Barthes, Roland. Image, Music, Text. Trans. Steph~n Heath. New York: Hill, 1977.
.Le plaisir du texte. Paris: Seuil, 1973.


[157]


Brooks, Cleanth. Modern Poetry and the Tradition. Chapel Hill: U of North Carolina P,
1939.
.The Well...Wrought Urn: Studies in the Structur-e of Poetry. New York: Harcourt,
1947. Bruns, Gerald L. "Canon and Power in the Hebrew Scriptures." V on Hallberg 65-84. Carlyle, Thomas. On Heroes, Hero... Worship, and the Heroic in History. Centenary Edition.
London: Chapman, 1897.
Coleridge, Samuel Taylor. Biographia Literaria. Ed. George Watson. London: Dent, 1982.
'"
Derrida, Jacques." Of Grammatology. Trans. Gayatri C. Spivak. Baltimore: Johns Hopkins
,
UP, 1976.
."Plato's Pharmacy." Dissemination. Trans. Barbara Johnson. Chicago: U of
Chicago P, 1981.61-171.
Donnegan, James. Greek and English Lexicon. London: Bohn, 1896.
Doubrovsky, Serge. The New Criticism in France. Trans. Derek Coltman. Chjcago: U of
Chicago P, 1973.
Eagleton, Terry. Literary Theory: An Introduction. Oxford: Blackwell; Minneapolis: U of
Minnesota P, 1983.
Fiedler, Leslie A., and Houston A. Baker, Jr., eds. English Literature: Opening Up the
Canon. Baltimore: Johns Hopkins UP, 1981.
Froula, Christjne. "When Eve Reads Milton: Undoing the Canonical Economy." V on
Hallberg 149~ 76.
Frye, Northrop. Anatomy of Criticism: Four Essays. Princeton: Princeton UP, 1957. Golding, Alan C. "A History of American Poetry Anthologies." Von Hallberg 279-308. Guillory, John. "The Ideology of Canon...Formation; T. S. Eliot and Cleanth Brooks."
Von Hallberg 337-62.
J J
.Jameson, Fredric. Th.e Political Unconscious: Narrative as a Socially Symbolic Act. Ithaca:
Cornell UP, 1981.
Kames, Lord [Henry Home]. Elements of Criticism. 1762. 3 vols. New York: Johnson,
1970.
Kenner, Hugh., "The Making of the Modernist Canon." V on Hallberg 363-76. Krupat, Arnold. "Native American Literature arid the Canon." Von Hallberg 309-36. LaCapra, Dominick. Rethinking Intellectual History. Ithaca: Cornell UP, 1983.
Lewis, Charlton T., and Charles Short. A Latin Dictionary. Oxford: Clarendon-Oxford
UP, 1980. .
Liddell, Henry George, and Robert Scott. Greek...English Lexicon. Oxford: Clarendon-Ox...
ford UP, 1961. "
Niermeyer, Jan Frederik. Mediae Latinatis lexicon minus: A Medieval Latin...French/English
Dictionary. 2 vols. 1954. Leiden: Brill, 1976.
Ohmann, Richard M. "The Shaping ofa Canon: U.S. Fiction 1960-1975." Von Hallberg
377-401. / Oxford English Dictionary. New York: Oxford UP, 1971. Peirce, Charles Sanders. Philosophical Writings of Peirce. Ed. Justus Buchler. New York:
Dover, 1955.


[158]

McGraw, 1966.
Smith, Barbara Hermstein. "Contingencies of Value." Von Hallberg 5-40.
.Contingencies of Value: Alternative Perspectives for Critical Theory. Cambridge:
Harvard UP, 1988.
von Hallberg, Robert. Canons. Chicago: U of Chicago P, 1984.