SOURCE: “The Function of ‘Prufrock’ for Criticism,” in T. S. Eliot Annual, Vol. 1, edited by Shyamal Bagchee, The Macmillan Press Ltd., 1990, pp.155-96.

This is the second of two unforeseen essays on the most familiar English poem of the twentieth century. My original intention was simply to write a modest (and tractable) piece pointing to certain important qualities of ‘The Love Song of J. Alfred Prufrock’ I did not find mentioned in the extensive published criticism. It was frustrated by a gradual awareness of the extraordinary historical relationship Eliot's poem has both with (to use his word) the tradition behind it, and with English literature and literary criticism since its advent.

Remarkably, in 1911 a graduate student barely into his twenties, and working in virtual isolation, created the principal harbinger/archetype of the English poetry of Modernism; yet nothing appears ex nihilo, and Eliot himself was aware of a great deal behind ‘Prufrock’. During the half-century following its advent, the movement that evolved and prevailed granted Eliot's youthful poem special canonical status. In part, this was because it seemed perfectly suited to the objective literary criticism—and the related teaching method—that became dominant during Modernism; yet ‘Prufrock’ actually contravenes important assumptions of that criticism. Its relevance persists because it also vindicates the most basic assumption of that objective criticism against the radically opposed basic assumption of much criticism that has evolved during the past two decades. At issue is the assumed nature of a literary—to beg the basic question—work: is it an artifact that embodies meaning, as the modernists assumed? Or is it a sign-sequence that stimulates meaning, as many assume today?

The remarkable advent of the harbinger/archetype directs attention to both prior literary history and its historical moment. The spectrum of assumptions about literature the canonical modernist poem contravenes directs attention to criticism since its advent. This latter extraordinary historical relationship is the subject of the present essay.

Both decades of commentary, and recent emphasis on the opposition of any particular reader to attributed evidence, increasingly disclose cryptic elements in ‘Prufrock’; as a result, once confident and seemingly unassailable assertions have been transformed into questions. One has practical priority. ‘Prufrock’ is in Eliot's second ‘voice’ of poetry, ‘a tale told’ by the poet ‘speaking through a mask’.¹ But what precisely is told through the mask as happening—in the basic physical sense—in it? (Does Prufrock attend the party?; does he set out for it and change his mind en route, and if so, where?; does he set out?)

The specific question leads to three general ones which have been asked, more or less consecutively, during the past two decades, and readily answered in the negative. Most immediate (and earliest in time): must what is happening be determinable? More abstracted and theoretical: is something definite happening in ‘Prufrock’—is the meaning in the words that constitute a poem determinate? Finally, the ultimate and fundamental question: is there meaning in it at all? Does this or any poem have (immanent) meaning? ‘Work’ begs this basic question about literature so familiar today, because the term usually is understood to signify an entity made of language—a meaningful quasi-object.

Neither interests nor qualifications make me fit for theoretical debate. But this practical critic's present subject is a poem that implicates him in issues of critical theory—to which the poem also provides access. The contribution I hope to make to the appreciation of ‘Prufrock’ itself will emerge with the probing of the critical issues whose singular paradigm it is.

Before that dual process can be begun, my conception of relevant criticism must be declared, since it subsumes the process. The following pretends to be no more than that brief declaration. A number of distinguished studies of the criticism of the past half-century accomplish aspects of the description and evaluation I am not attempting. Of the many shorter ones concerned with North America's principal modernist criticism, René Wellek's recent ‘The New Criticism: Pro and Contra’, in The Attack on Literature (Chapel Hill: University of North Carolina Press, 1982) is noteworthy for its authority—both knowledge and familiarity—with respect to its subject. The essay relates the New Criticism succinctly to both its antecedents and its successors, while providing a brief, respectful but probing appraisal of it.² In her introduction to the collection, Reader-Response Criticism: From Formalism to Post-Structuralism (Baltimore: Johns Hopkins University Press, 1980), Jane P. Tompkins traces the evolution out of modernist criticism of that emphasis in the newer criticism. As a final example, and a complement to these, the major part of Frank Lentricchia's After the New Criticism (Chicago: University of Chicago Press, 1980), ‘A Critical Thematics: 1957-77’, is an ‘historical … study [of] the past two decades of critical theory in the United States’ (p. xii).

I. ON POETRY AND CRITICS

From Aristotle's challenge twenty-two hundred years ago to his teacher's characterisation of literature in the Republic, until the waning of Modernism the day before yesterday, critics’ conception of a literary composition remained constant in a major respect. Implicit in the Poetics is the assumption that a philosopher can talk about plays and epics as Aristotle does because as verbal compositions they are artifacts. Like any other substance, the composition of words has an essence or entelechy, part of which is its meaning. In different phases of Western culture critics characterised the language of poetry differently, considering it mainly either propositional, or evocative—even to the extent of being, in Dante's phrase, ‘beyond sense’—or idiolectic (and its creator respectively sage, or mystic, or magician). But the critics agreed implicitly with Aristotle that the meaning of that language, though it may be elusive, even ineffable, is both there—immanent in the structure of words—and, because it is immanent, determinate.

Eventually, the romantic conception of organic form precluded a separable content/meaning. Then in the present century, ‘the meaning of meaning’ was scrutinised, and the difference between an abstract account of meaning and its embodied reality was emphasised. By mid-century, a theoretical foundation for the previously implicit assumption that meaning is immanent in literature had been articulated in the American ‘New Criticism’.

The appellation was inept even before it became anachronistic; ‘Formalist-Cognitive Criticism’ may be cumbersome, but it designates what seem to me the linked fundamental doctrines of that criticism. My understanding of that fundamental dyad can be stated briefly. A poem is an autonomous structure of words existing in the cognition of that meaningful structure.³ An object not physically but cognitively, an apprehended formal entity, a system, the poem is a functioning ‘heterocosm’ whose lineaments and adornments properly monopolise the critic's attention. And since it not merely imparts but is (among the other things it is) meaning, meaning in it undoubtedly is both immanent and determinate. This conception of literature (‘poetry’ usually signifying more than verse composition), articulated in Formalist-Cognitive criticism, became essentially the prevailing one during the period ‘Prufrock’ helped to inaugurate.⁴ But, as we shall see, ‘Prufrock’ does not fully confirm it.

Aristotle's mimetic artifact—Work about the Universe (to cite two of the four ‘Co-ordinates of Art Criticism’ in the famous conceptual model presented in the first chapter of M. H. Abrams's study of romantic criticism, The Mirror and The Lamp [New York: W. W. Norton, 1953])—effectively displaced Plato's poem as expression/evocation of the appearance of an illusion. Throughout literary history thereafter, critics generally emphasised Audience and then Artist, until with Modernism those two of Abrams's ‘coordinates’ were jointly challenged and by mid-century eclipsed by the objective one. And now, as the century wanes, most major varieties of critical theory constitute their parabolic joint reinstatement.

Abstractly, the change from the habitual modernist reference to a ‘work’, to current use of the French/linguistic word ‘text’, is the substitution of a neutral term for a question-begging one. In fact (that is to say, historically), that substitution has proved to be the vehicle for the revolutionary challenge to the traditional Aristotelian myth that meaning in literature has objective (real) status, by the regenerated Platonic myth that it is subjective (ideal). They are alternative myths—one or the other fully believed by most critics, neither one apparently provable. The issue seems part of the Zeitgeist: it invokes the general, epistemological question of the relation of knower to what is known, and has analogues not only in the alternative ‘God's truth’ and ‘hocus-pocus’ views of the structure of language, but also in the alternative conceptions of mathematics as ‘discovered’ (‘God-given mathematics’) and ‘invented’.⁵ Although neither Aristotle nor Plato often is mentioned in these current debates, opposed reference to ‘realism’ and either ‘conceptualism’ or ‘idealism’ usually occurs.

My purpose is not to participate in the current debate about criticism, but to show how ‘Prufrock’ arbitrates the alternative conceptions of meaning and literature which are at its centre. About two decades ago, ‘A new turn in poetic taste’ occurred, in the words of René Wellek's recent essay, bringing with it ‘attempts to dismiss T. S. Eliot both as poet and critic and to reduce the role of all modernism [that] imply a rejection of the New Criticism also … ’ (p. 101). By that time, some critics in France and North America had not just denied that meaning in literature is positively determinable, but rejected the belief that any ‘sign-sequence’ has determinate meaning. Their doctrine derived, of course, from the attitude toward language and meaning propounded in the newly revived structural linguistics of the Swiss pioneer, Ferdinand de Saussure.

From the structuralist principle that social phenomena are understood not by the specific content of their elements or instances, but in their relations, it follows that ‘language is a form and not a substance’. And Saussure specified the relations constituting that ‘form’ in his doctrine that every verbal structure is a sequence of signs having two components, the signifying set of sounds or shapes and the mental events they signify—signifiant and signifié—and having only those two. Although the user of language normally is being referential, this ‘lexicon of signification’ (Roland Barthes) frees the reading of the ‘sign sequence’ from immediate connection beyond the shapes or sounds and the mind experiencing them. The two-part signs are autonomous structures—they rely on no (necessarily external) referent. Hence, literary ‘texts’ are hermetically sealed off from referential meaning not because of any special status as ‘aesthetic’ artifacts, as modernists claimed, but a priori: language may embody referential meaning, but as apprehended it does not refer; it stimulates mental events (based on prior experience). Simultaneously, Saussure's doctrine raises the experiencing mind to correlative status with the experienced language in a stimulus-response model. And therefore it makes the free-floating, subjectively apprehended, meaning indeterminate.

The structuralist conception of the relationship of language and meaning in literature developed into ‘post-structuralist’ conceptions as, with historical appropriateness, its current sequels are called. And the foundation was Saussure's doctrine that language is apprehended as a structure of hermetic two-part signs. Setting aside that all reality is known as the mental experience of it, the mental act which is Saussure's latter (signifié) must be conceptual—‘idea’—since it is not referential. And with their affection for imagery and tropes, modernists might object that Saussure's doctrine is refuted by the way imagery actually works. That I would share their objection is relevant, in part because the structuralist misrepresentation of the way imagery works corresponds to the modernist misrepresentation of the way allusion works; and with the eloquent example of ‘Prufrock’, I shall object to that below.

Most discussions of imagery involve its role in metaphor. The extensive attention of philosophers, linguists, literary critics and psychologists has emphasised the creation of metaphor (Aristotle first addressed this), the process of understanding it, and its truth-value. My concern is limited to a different question: What makes metaphor effective? The basis for my objection to the structuralist view of language (and the analogous modernist view of allusion) is the way imagery seems to work to make metaphor so.

A recent contribution by Paul Ricoeur to ‘the somewhat boundless field of metaphor theory’ characterises the image as accomplishing ‘the picturing function [sic] of metaphorical meaning’ in ‘a pictorial or iconic moment’. The moment is not strictly cognitive but more broadly psychological, as his title suggests: ‘The Metaphorical Process as Cognition, Imagination and Feeling’. And by the last word he does not mean an emotional reaction: ‘Feelings … are interiorized thoughts’ which ‘accompany and complete the work of imagination as schematizing a synthetic operation …’⁶ Ricoeur is concerned with making and understanding metaphors, not with the source of their effectiveness. But he insists that their images are apprehended in a ‘synthetic operation’ of ‘cognition, imagination and feeling’. I shall not presume to an equally authoritative assertion, but attest instead that, for me (in me), an image works to create its effectiveness, and so works to endow a metaphor employing it with that effectiveness, because my apprehension of the image is just such an inclusive—synthetic—activity.

When I apprehend an image, it does not simply signify. Instead, it functions in me precisely imagistically: it transcends the hermetic complex of signifier-signified to invoke—as experience, immediately and directly—my vicarious equivalent of the bit of experienced reality it names. Not just ‘cognition’ occurs, but also something like ‘imagination and feeling’. This seems to be so for other readers, too. Hence, the power to function as image of any image in literature is coextensive with the experience—in fact or imagination—the image names. An image that is universally effective is so not because its words are universally understood, but because they invoke a community of experience.

On the first page of a classic work of modernist criticism that also influenced greatly the historical development being recapitulated here, Seven Types of Ambiguity, William Empson discusses his variation of a common specimen sentence, ‘The brown cat sat on the red mat’. Are one's experienced instances of brownness-and-catness/readness-and-matness merely coalesced in mental acts? How can an image function effectively in a metaphor (for example, ‘as it set, the sun laid a red mat on the distant field’), if the reality-as-object-of-sensory-experience that the image names is not invoked? It is actual red-matness that transforms the equation in language into an experience of language. Indeed, when Othello says, without imagery, ‘But yet the pity of it, Iago! O Iago, the pity of it, Iago!’ is the effective meaning of the repeated phrase merely what the words signify? Or is the human experience relevant to the represented life of the play invoked to supply what mental act would never discover in the simple phrase of four words and five syllables?

Whatever the merits of my testimony, rejection of the hermetic view of language is now decidedly ancien régime. The objective criticism of Modernism has been largely superseded, in a development dialectically implemented by certain of that criticism's own doctrines. The history runs through Structuralism to Post-Structuralism, first promoting the doctrine of indeterminable meaning in a poem or story, then rejecting determinate (‘metaphysical’ in a term of Deconstruction) meaning, then denying that language has any immanent meaning at all, and so enabling (in the idiom of Phenomenology) ‘The production of the meaning of literary texts’.⁷

The revolution has crested. Thus, in Is There a Text in This Class? (Cambridge: Harvard University Press, 1980), Stanley Fish declared:

No longer is the critic the humble servant of texts whose glories exist independently … it is what he does that brings texts into being … (p. 368)

Fish's discussion of one's ‘beliefs’ about a text (‘I spend most of my time … fielding questions that sound disconcertingly like objections from my former self’) constitutes a witty, humane and intelligent endeavour to justify ‘a persuasion model’ in which ‘critical activity is constitutive of its object’—an endeavour to justify the contemporary myth that meaning is not immanent in the text—against its venerable contrary myth, which maintains that ‘critical activity is controlled’ by the work (pp. 364-5). Although by no means uniquely, ‘Prufrock’ nevertheless tangibly discredits this currently popular denial of immanent meaning in poetry.

II. IS THERE A WORK IN THIS TEXT?

A representation as the subject's language of consciousness itself, ‘Prufrock’ anticipates the most sensational and famous instance of this special narrative method the modernist focus on psychology engendered: the final chapter of Ulysses. In both cases, the ‘stream of consciousness’ portrays an inner conflict between alternative attitudes; and in both cases, the conflict is crucial because its issue is fundamental to the character's destiny. Finally, as in most persuasive uses of the method, in both cases the process of the character's consciousness is made associational.

Hence, the ‘I’ says ‘Let us go … when’, then ‘Let us go, through’ what, then ‘Let us go and make our visit’, to ‘the room’ with ‘the women’ inside and ‘the yellow fog’ outside. Committed to the character's associational process by the method, an author can endow his or her portrayal with coherence and teleology only by digging, discretely, a channel for the character's stream of consciousness only by exploiting controlling devices not relinquished in the method.

Some devices implementing control are compatible with the character's process of consciousness because they function as indistinguishable from it. The two paragraph breaks between visit proposal and description of fog are hiatuses in Prufrock's thought; the satiric, soon repeated, jingly, go / Michelangelo couplet also is Prufrock's mental play. But some devices remain compatible with the character's process although distinctly the author's product.

The difference can be illustrated in Joyce's chapter. The eight typographical units in which he disposed Molly Bloom's ‘soliloquy’, whose relations to each other impose form and direction on her paratactic reflections, also are elements of her process and so constitute an indistinguishable controlling device. But the point in her reflections at which the chapter begins and the point at which it ends signify the author's selection from his character's ostensible stream of thought, and together complete the significant form that makes Molly's stream of consciousness function as the conclusion of Joyce's novel. She is unaware of that selection, so the distinguishing of his product from her process is not a naive attempt to separate form from content; it is the recognition that a telic design informs a discourse totally innocent of such design.

The young Eliot was resourceful in informing Prufrock's process with controlling devices both indistinguishable and clearly distinct from it. And the telic design that makes the poem (his product) out of its discourse (his character's process as a composition of language) also illuminates the critical dispute about language and meaning in literature. It indicates that ‘Prufrock’ is no mere signsequence stimulating mental acts that create different meanings in different individuals or at different times, but a quasi-object constituted by its immanent meaning. Two theoretical challenges to this assertion can be anticipated.

Like the separation of form from content, the pure Cartesian separation of apprehending subject from the object apprehended is naive: as Georges Poulet observed, every thought is a thought about something. On the other hand, one is not obliged to grant that nothing definite can be said about meaning associated with a literary work, so long as no one has successfully disproven that a community of discourse to some extent constrains, and directs, the apprehension of language. The issue is strictly whether Eliot's composition of words is a stimulating sign-sequence or a meaningful quasi-object—the precise nature of the subject-object relationship when a poem is read.

The second challenge is that one cannot assume ‘Prufrock’ to be ‘out there’ as more than a stimulus of black marks on white paper, when drawing evidence from it to contest the currently popular denial that any more substantial entity can exist. And no such a priori assumptions have been relied on. Instead, the modernist doctrine will be shown to provide a more adequate account of Eliot's poem—to accommodate more of its characteristics—than the current alternative doctrine. And the phrase ‘its characteristics’ will be shown to refer, in some cases, to elements as substantial as the margins on this page.

The initial controlling device with which Eliot informs Prufrock's process is distinct from that process because it has not yet begun. Whether they embody their meaning or merely stimulate it, the words ‘The Love Song of J. Alfred Prufrock’ introduce Eliot's poem. Like the title Joyce gave the novel whose last chapter it anticipates, the title Eliot gave ‘Prufrock’ is a key to significant relations in it. Two equal-length incongruous phrases mediated by a brief genitive preposition, the title not only announces Prufrock's immobilising predicament, but also indicates the three principal forms of the (indistinguishable) controlling device Eliot uses for emphasising Prufrock's inability (like a patient etherised) to overcome his predicament. The title embodies frustrated expectation, binary opposition, and chiasmus. These are the three principal forms of Eliot's controlling device; the first two are stylistic tropes of vacillation and equivocation, the third, of return or inertia.

Instances of them in the poem come readily to mind (the two lines beginning ‘In a minute’ combine all three); but it is the ubiquity of this stylistic complex that identifies it as a controlling device, Eliot's product. For example, in the first verse paragraph: ‘Let us go’—not now but ‘when’; you—and (the opposition is yet to emerge) I; the conventional romantic reference to the evening sky—and the completing simile; the streets which do not invite going—but repel it; half-deserted—those streets mutter; they do not lead—they follow; like an argument—tedious but insidious; following—they lead; not us—but you; and the question the ‘you’ is told not to ask—is put.

The stylistic pattern does not demonstrate the deconstructionist proposition that all sign-sequences subvert their meaning. It is Eliot's design for making his poem express formally the crucial meaning the poem portrays—that the process of Prufrock's consciousness is a doomed psychomachia.

Two other major controlling devices enabling Eliot to shape his product in Prufrock's process are prosody in the poem and, most important for the question of the relationship of language and meaning in literature, his formal arrangement of his character's stream of consciousness.

Couplets and variants on couplets—triplets, one couplet or two in an envelope rhyme—end the great majority of the lines. People do not speak (or think) in rhymed couplets, but the mere presence of a prosodic pattern in a poem could be more than convention. The first three paragraphs of ‘Prufrock’ assert that its prosody functions positively as the index of a controlling presence.

In the first paragraph, ‘an overwhelming question’ is followed by three dots, a conventional mark for interruption Eliot uses elsewhere in ‘Prufrock’ and in other early poems. All the other lines form couplets except the third, which presents the unexpected and shocking simile. The second paragraph is the jingly couplet. The third breaks the poem's pattern of couplet rhyme temporarily, but substitutes a metrical pattern for it. The arch proto-imagist cleverness of its fog-cat (apparently Eliot's earliest cat poetry) can be attributed to Prufrock; but Eliot informs Prufrock's extended metaphor with a pattern of iambic lines. The first two lines combine in opposition iambic heptameter, which is inherently unstable, and the established tight regularity of couplet rhyme. The latter five lines alternate iambic pentameter and hexameter. And to reinforce his controlling presence, again he made the third line irregular.

Prosodic pattern functions as a controlling device most emphatically at the place in ‘Prufrock’ where Eliot temporarily suspends it; and there, as will be shown, it functions in combination with his ordering of the parts of the poem.

His most overt means of ordering the parts is refrain. Refrain in the poem is indistinguishable from his character's process, as when Prufrock repeats ‘Let us go’ three times in the first paragraph; yet Eliot makes it a controlling device. When the couplet which serves as the second paragraph recurs as the fifth, the paragraphs bracketing it have the repeated phrase ‘there will be time’ and numerous other repetitions of the word ‘time’. This verbal refrain reinforces their common subject, to link the fourth and sixth paragraphs. And the refrain couplet between them, which also occupies the central position in the first three paragraphs, thereby creates a grouping of paragraphs four through six and another (retroactively) of one through three. The controlling device remains indistinguishable. Even the refrain couplet central to both sets of three paragraphs is precipitated in Prufrock's process by his thought first of making ‘our visit’, then of taking ‘a toast and tea’.

Not individual verses, but verse paragraphs, are the significant formal elements of ‘Prufrock’. The next three paragraphs are unified by the phrase ‘I have known’ and ‘know’. And the last nine paragraphs in the poem are grouped like the first nine in three sets of three, although this is achieved less by refrain than by other means. (Refrain phrases are used only in the latter two paragraphs of the fourth set, and not at all in the fifth and last sets.) The last set of three concerns the mermaids, and is introduced by the line ‘I have heard the mermaids singing, each to each’. The fifth set, which concludes with that line, is Prufrock's appraisal of his present condition and future prospects.

That fifth set is all, appropriately, in the indicative mood (mode). The verb forms that record the process of Prufrock's ‘love song’ evolve in a pattern conforming to and so reinforcing Eliot's disposition of most of the poem into two clusters, each with three sets of three paragraphs.

The ambiguous ‘Let us go then’ aside, the verbs in the first three paragraphs all are present indicative. Those in the second set of three are future indicative, with general statements and the question ‘Do I dare?’ in the historical present. The third set combines the present perfect indicative—in which Prufrock reviews his chronic situation—with the more malleable subjunctive mood, in which he poses the question ‘should I?’ The fourth set—the first of the second cluster—repeats the combination of the third, except that the subjunctive indicates conditional contrary-to-fact (‘would it have been worth … / If one … / Should’), not possibility. The fifth set, like the second, is future indicative with general statements—and the significant mutation, in its final paragraph, of the repeated question in the corresponding final paragraph of the second set, ‘Do I dare to eat a peach?’—in the historical present. Following the denial of a better prospect (‘I do not think that they will sing to me’), the last set reasserts in the present perfect indicative the constancy of Prufrock's predicament.⁸

If Eliot's strict ordering by way of mutually reinforcing patterns of prosody, refrain and verb form were obtrusive, his poem would be as pedantic and dull as this recapitulation. He made the ordering devices indistinguishable and assimilate their working into Prufrock's process. His reasons for the whole enterprise are indicated by the bit of the poem he excluded from the pair of clusters of three subsets of three paragraphs: the tenth and eleventh paragraphs, set between the clusters. It is in those two central paragraphs that he makes prosodic pattern function most by suspending it (he avoids refrain as well).

That they embody a definitive development in Prufrock's story, in two stages having a familiar pattern, is indicated by the respective tenses of the two paragraphs. The first follows the set of three paragraphs reiterating the subjunctive ‘should I’; and it comprises a single future indicative question: ‘Shall I say …?’ The second returns to the subjunctive. And here ‘should’ expresses neither the possibility expressed in the set of three paragraphs preceding it, nor the conditional contrariness to fact of the set following it, but the psychological bridge from aspiration to rationalisation: complaint. It specifies a preferable alternative, the desolate nature of which, conveyed at the centre of the poem as much by ‘scuttling’ and ‘silent’ as by the crab image itself, eliminates doubt either about the futility of his ‘love song’ or about his destiny.

The third and fourth sets of paragraphs are grammatically alike, and so are the second and second-from-last sets. This minor symmetry, and the major symmetry that disposes the 20 paragraphs in two clusters of three sets of three with one plus one between them at the centre, reinforce the repeated instances of verbal and phonetic chiasmus in the poem to express stasis, inertia, futility. The chiastic pattern even extends to certain themes; for example, evening in the first set, and aging in the last; his physical appearance in the second, and in the fifth.⁹

But the important point for the controversy about language and meaning in literature is Eliot's undeniable ordering of Prufrock's process; and his most explicit ordering device is most significant for that controversy.

A series of five dots appears at each of three points in the poem: after the first cluster of nine paragraphs; before the second cluster (bracketing the tenth and eleventh paragraphs); and after the first set of three paragraphs in the second cluster.¹⁰ The crucial (in more than one sense) bracketed paragraphs present the decisive development in Prufrock's process, and portend his destiny. The third series of dots separates his contrary-to-fact conditional rationalising of inertial from ‘No! I am not Prince Hamlet’, which initiates the set of paragraphs projecting his future life and ending with the line introducing the mermaids. In other words, two series of five dots mark off the climax or crisis of the action of Prufrock's psychomachia, which rises in the first part of the poem and falls in the symmetrical second part: frustrated expectation, binary opposition, chiasmus. And the third series of five dots designates the conclusion of the poem, comprising the last two sets of paragraphs.

Most of the individual details in ‘Prufrock’ that Eliot combined as controlling devices are indistinguishable from his character's process. However, the three series of dots are distinct poet's product, not inside but between the units of language: they do not simultaneously belong to Prufrock's thought-discourse. And they not only shape the discourse, but shape it meaningfully. Especially the two series of dots at the centre enact a meaning for the poem containing Prufrock's thought-discourse which is independent of the meaning in (or those meanings signified by) that discourse. Furthermore, the meaningful shaping is not itself verbal: Prufrock's discourse has been subsumed in a physical composition. Overtly enacted by the three series of dots, Eliot's physical composition incorporates his elaborate arrangement of verse paragraphs.

There is an analogy with picture poems; but those do not clarify the question whether meaning is immanent in a ‘work’ or just stimulated by a ‘text’, because they employ visual mimesis of an actual shape, achieve a kind of pictorial paranomasia. (The case with concrete poems is more involved.) The meaningful shape of ‘Prufrock’ is geometrical, not pictorial: it conveys the poem's own anatomy, rather than an external object's morphology. Therefore, the shape of the poem enacts—articulates—meaning for the language of the poem.

The basic critical views about form in a work of literature correspond to those about the relationship of language and meaning. That the form in a work is conceptual, an abstraction from the composition of words, is agreed. But the current view is that this abstraction is a model created by each reader, each model of the work's form an instance of the ‘process by which men give meaning to things’; and the modernist (objective) retort to this idealist view is that a reader's ‘model is not the structure, for the structure is always in the object, latent as it were but only if latent is not opposed to real’.¹¹

In ‘Prufrock’, the form is not wholly latent but partly manifest, and that part of it could scarcely be more real. Its reality is attested by patterns of physical arrangement too complex to be accidental, any more than justified margins are accidental. Within his character's stream of consciousness, which constitutes the language of the poem—and so indistinguishable from that language which is innocent of its presence—Eliot created the geometrical configuration described. (As though to assert this creation, in one instance he made a paragraph of a single verse.) He added to his indistinguishable articulation of the parts of his poem the distinct demarcations created by the three series of dots. The form of three threes as nine, one and one as two, and three as nine, with the final two sets of paragraphs the conclusion, must be conceptualised; it is ‘latent as it were’—‘but only if latent is not opposed to real’. (The paragraphs can be counted, for example.) In the same way, the ‘silently sounded’ (Roman Jakobson) prosody—as well as its absence at the centre of the poem—is real sounds; so also are the instances of phonetic chiasmus: an oscilloscope could record them.

These agencies of formal articulation are ‘in the object’—immanent in ‘Prufrock’: where else can they be? And they do not convey Prufrock's meaning but enact fundamental components of the meaning of Eliot's poem. In other words, the meaning in ‘Prufrock’ is not a matter of the language of the poem alone, but partly the product of physical properties creating a geometrical context for that language. This augmentation of the meaning of the ‘sign-sequence’ or ‘text’ of Prufrock's language itself, which is partly outside that language, to which Prufrock the speaker of the language is not privy, is the crucial evidence that a literary composition probably is a work embodying meaning. For although the augmentation of meaning could be merely the poet's ordering the reader's mental acts, it is more likely that the meaning in the speaker's ‘text’ shares the ontological status of the elements of the poem which augment it: the physical context Eliot created for it. There is no experience of cognitive disjunction: it is as if meaning is just as immanent in the text of Prufrock's thought-discourse as it is in the formal context Eliot created in making his poem out of his character's text. One can never prove that the meaning of Prufrock's language does not result solely from a reader's mental act, for the alternative myths about meaning and language abide because they are impervious to final proof. But the relationship between Prufrock's text and the partly physical context it manifestly does have indicates—and the physical elements of Eliot's meaningful context prove with respect to themselves—that the Aristotelian myth more adequately accounts for the reality of this work of literature.

One result of the conception central to Modernism of art-works as—and of the pressure artists felt to create—quasi-objects, is that Eliot's canonical poem and harbinger of English modernist poetry discredits the current doctrine that denies immanent meaning to language, comports with the objective conception of language and meaning in literature implicit in the metaphor for a poem a prominent Formalist-Cognitive critic took from Donne's ‘Canonisation’ and made famous among the last critical generation.

III. A WELL WROUGHT URN?

It is an historical irony as well as a tribute to the potter's artistry that Formalist-Cognitive and other modernist critics overlooked some of Eliot's prodigious formal working in ‘Prufrock’. But the irony is much less important than the doubt ‘Prufrock’ raises about two related implications of Donne's metaphor (or Wimsatt's own ‘verbal icon’) which suited those critics: the ‘cognitive’ implication that, although unparaphraseable, and fully knowable only by the avowedly fictional Ideal Reader, meaning in a work is not only immanent but also definite; and the ‘formalist’ implication that, when its intrinsic nature as literature is respected, meaning in a work has a hermetically isolated status—a poem is an ‘aesthetic’ entity, an idiolect. This section invokes ‘Prufrock’ to probe the first, the fourth and last sections, to probe both of these two central modernist critical doctrines.

Cleanth Brooks himself (and his collaborator, Robert Penn Warren) helped create the canonical distinction of ‘Prufrock’. The first two extended critical studies of the poem both appeared in 1938; one was in the first edition of their classic textbook—probably more famous, and more influential on literary culture in America, than any other since McGuffey's elementary-school Eclectic Readers.¹² About understanding ‘Prufrock’, Understanding Poetry declares: ‘The poem has some complication, of course, but its primary difficulty for the reader is the apparent lack of logical transitions’ (p. 595). The revised commentary of the 1950 and 1961 editions points out that ‘the events are not as fully indicated in Eliot's poem as in Tennyson's’ dramatic monologue ‘Ulysses’, but says a careful reading will ‘permit us to realize the implications of the whole poem’ (pp. 443-4; p. 390). In a more recent festschrift for W. K. Wimsatt, Brooks wrote: ‘One can make out in “The Love Song” a vague narrative …¹³ And the subject of the festschrift had already written in 1952, the year before Abrams described objective criticism as the new ‘reigning mode’, that:

In Prufrock it is nearly possible, tantalizingly plausible, to suppose a basic story of a little man approaching a tea party at which there is a woman to whom he might … propose marriage, or to whom he stands, rather, in such a casual relation that his very thoughts of proposal are almost hallucinatory.¹⁴

The extent of the ‘uncertainties’ Wimsatt labelled four sentences later is revealed in the total subverting of his initial ‘plausible’ supposition by the alternative he introduced quietly with ‘or’ and ‘rather’.

Formalist-Cognitive critics acknowledged the ineffable dimension in poetry, insisted that its meaning cannot be paraphrased adequately, and augmented our understanding of the richness of plurisignation. But ‘Prufrock’ confronts what Abrams called ‘the objective point of view’ with—precisely—uncertainty. The doctrines of both structuralist and post-structuralist criticism make the reader's uncertainty a consequence of the interaction of language and experience, and welcome it. To modernist criticism uncertainty is far more disquieting, for the doctrine that meaning is definite dictates that uncertainty be attributed only to inadequate reading, or to a defective work. ‘Prufrock’ shows that both the recent and the earlier doctrines simplify the reality of poetry because both are right. It discloses three sources of uncertainty: reading competence, ‘Prufrock’ itself, and the critic's ‘bias’ (Barthes)—the existential conditions of a particular reading. Meaning may be immanent in a poem, but even if it is so, it is not necessarily all stable, or even definite.

In his recent book on Eliot, Piers Gray asks: ‘What is the nature of the very opening lines …?’, and declares ‘the poem immediately opens before us several levels of [sic] uncertainty’.¹⁵ Its first four words illustrate the case. Does ‘Let us go’ mean ‘[you] permit us to go’, or is it imperative? In either case, who is the auditor, ‘you’? Does ‘then’ relate to ‘When’ at the beginning of the second line, signifying that evening has not yet come, which would be stylistically clumsy? Or does it mean ‘consequently’, in which case ‘when’ may or may not mean ‘now that’? In either case, do ‘you and I’ go in the evening?

Who is ‘you’?, and Do ‘you and I’ go?, have been answered by many critics in many ways over the past half-century. But they cannot properly be separated, nor isolated from the meaning of ‘then’ and that of ‘When’. If it is not yet evening, the visit simply cannot take place in the poem; a later reference to ‘the afternoon, the evening’ specifies that it can take place in the poem, and so indicates that ‘then’ means ‘consequently’. That being so, one is returned to the first question. For whether or not the visit takes place is more or less significant depending on whether or not You can command I—whether ‘Let us go’ is consenting, petitioning or even commanding—which depends on who the You is. For example, the I may be announcing his own resolve, as a consequence of consideration with the You (before the poem begins), of the proposal by the You to make the visit (‘Let us go then’). If that is so, and yet they do not go, the poem constitutes a subversion of the apparent resolution that is its starting point.

The reader's irresolution about the meaning of the opening words mirrors Prufrock's irresolution throughout the poem, and may be a strategy of Eliot's to stimulate negative capability respecting Prufrock's predicament. And the interdependence of one's alternative inferences about those first words is another warrant (if one is needed) of the poem's coherence. But the very range of critics’ answers to the questions Who is You?, and Do You and I go?, documents the uncertainty.

In the ‘Words Set Free’ section of The Pound Era (Berkeley: University of California Press, 1971), Hugh Kenner asks of the opening line, ‘what meaning do we attach to it, for instance to you?’; He mentions the three most common alternative answers to the question: You ‘is possibly the reader …; or is possibly some other part of … Prufrock; or is possibly … Dante or some analogue of Dante's’. Then he dismisses the question as anachronistic (‘Georgian’); for the modernist line ‘now causes no difficulty’, its ‘words set free’ to embody not meaning but ‘effect’ (quoting from Yeats's 1900 essay ‘The Symbolism of Poetry’) ‘too subtle for the intellect’ (pp. 130-1). The argument bridges modernist and structuralist approaches: uncertain meaning is no problem since poetry is idiolectic and (citing Mallarmé) ‘made not of ideas but of words … the poem can convey’ effects, not meanings.

Many passages of poetry and even some whole poems are accessible only in terms of effect (and so refractory against formal analysis); in those instances concern for uncertainty about substantive meaning is frivolous. But that is not the universal case, even in symboliste and modernist poetry. Critics continue to try to identify You because the identity has an instrumental function in Eliot's poem.

In the same year, 1950, an extended ‘Reading’ announced simply ‘Prufrock is speaking to us. We are the “you”’, and the second edition of Understanding Poetry both declared You to be ‘the generalized reader’ and proposed that the poem ‘in the end, is not about poor Prufrock’: like Guido da Montefeltro in the epigraph, Prufrock ‘speaks to the “you” of the poem—the reader—only because he takes the reader to be damned too’, afflicted by a general disease’ in modern Western society.¹⁶ These two studies may have been the first to identify the You in print as the reader. Using the epigraph more elaborately than Brooks and Warren, a subsequent critic makes the same identification, but draws the opposite inference from theirs. He explains ‘then’ as a transitional word, whose function is to link Prufrock's self-revelation rhetorically to the epigraph; and as Guido is thus pointedly made to correspond to Prufrock, so Guido's auditor, the poet Dante, corresponds to the reader. However, Guido's presumption about Dante is erroneous, which signifies that Prufrock's reader is not ‘damned too’ in a poem of comment on modern society but only, like Dante, ‘visiting.’¹⁷ The more direct and strictly logical connection has now been made: the You whom Prufrock addresses who ‘functions … to elicit confidences … and write the poem—bring back the story’, ‘the Dante-figure of the poem’, has recently been identified as the analogous poet—Eliot himself. In support, Eliot's statement in a letter that ‘the “you” in THE LOVE SONG is merely some friend or companion’ is quoted, and this instance of the usual evasive politeness of Ol’ Possum interpreted as signifying that You is himself.¹⁸

Of course, the conclusion of most critics of Eliot's poetry, beginning with F. O. Mathiessen in his book published in 1935, is the least ingenious one—that the You is neither the reader nor the poet, but a constant inner ‘companion’ of Prufrock's, one able to ‘hear’ his consciousness.¹⁹ The logic of the poem's narrative method indicates this, and a wealth of supporting detail makes the uncertainty about it derive from the first of the three sources mentioned—reader competence. For example, the reader can be neither ‘here, beside Prufrock on that particular evening, nor preparing a face, and receiving a plate, at tea. As for Eliot's being Prufrock's Dante, although ‘then’ can be seen as stylistically sequential to the epigraph, the meaning ‘Because you cannot repeat what I say, then let us go’; (and go where, if forever in hell?) is not a logical sequence. Indeed, if Prufrock were addressing Eliot in the mistaken belief that the poet could not reveal what he confided, and the poet betrayed his confidence, it would be a cruel joke on a pathetic figure—and a morbid joke as well, for after all Prufrock is nothing but Eliot's own words.

Eliot does not have to be addressed by Prufrock for his analogy with Dante's portrayal of a character in hell revealing himself to operate. Prufrock's auditor will never repeat what he ‘says’, for it is his silent thoughts. A symptom of his problem is that he does not—would not, cannot—say to another person what he is thinking. Like Guido he is in hell and assumes that his confidences will be kept because they are imparted there; but his assumption is more sound than Guido's, because in his private hell he speaks only to You who perforce cannot but keep his confidences. The critical consensus that, in Mathiessen's phrase, Prufrock is engaged in a ‘debate with himself’, is the reasonable inference about a thought process which is a response. Piers Gray writes:

If one speaks to oneself … you has been given a role, in a sense, independent from I. And one is therefore … both subject and object. The French for this state is dêdoublement …

Eliot himself uses the term in describing Laforgue's irony as the instrument ‘to express a dédoublement of the personality against which the subject struggles.’²⁰

In the poem I—its manifest Prufrock—presents his response to You about a possible action involving also (perforce) You; presumably You desires it (‘Let us go then’) or it would not be at issue; I aborts it; only You meets faces, and a single face is prepared (for You: one says ‘How do I look?’ but thinks ‘How does me look?’) on those occasions; You and I both may be talked about; I speaks of his physical attributes, of daring to eat a peach, of determining—and of the satisfaction he takes in—his dress; both ‘have lingered’ by the ‘sea-girls’, but only I has seen them—it is one's I who monitors the optic nerve. Affixing to I and You names from Cartesian philosophy, or Freudian or Ego psychology, is unnecessary and would itself engender uncertainty. The poem discloses who I and You are, and we all know both from personal experience, though the I in most of us is less timorous and forlorn. In ‘“O Where Are You Going?” Said Reader to Rider’, Auden presents both sides of an anti-Prufrock's debate; it is not a dédoublement, for only the I is involved; and in contrast to Prufrock ‘he’ (rider, farer and hearer) frees himself from ‘them’ (reader, fearer and ‘horror’, the I's paralysing inhibitions), to precisely ‘go’. The identity of neither antagonist, in neither poem, need be uncertain.

Do Prufrock's You and I also go? The answers of critics who have addressed the question directly are a mirror of uncertainty. Prufrock goes to the party and then to the beach; attends then leaves the party; attends the party; ‘has mounted’ but then descends the stairs; goes to the outside of the house; and is walking to the party but ‘is not there yet as we hear him speaking’.²¹ Grover Smith boldly proposed the significant alternative to all these various inferences, in T. S. Eliot's Poetry and Plays: ‘the action [is] limited to the interplay of impressions, including memories, in Prufrock's mind’ (p. 16). Others treat the question diplomatically: ‘it is impossible to say which … elements are externally “there” and which are the disjecta membra of a disordered consciousness’; ‘As he enters the room, or as it enters his consciousness’.²² Or they concur in Smith's view, with diffidence: ‘the streets and fog to be traversed (or, more likely, contemplated)’; ‘One of the puzzles of the poem is … whether Prufrock ever leaves his room. It appears that he does not … ’²³

Most of the various inferences of the first group of critics occurred earlier than the proposals that Prufrock never stirs. As this fact suggests, while the poem cannot be absolved of all responsibility for the prevailing uncertainty, it is more definite than the multifarious criticism indicates.

The most purposeful (richest) alternative is that Prufrock does not stir: the poem begins when I expresses to You a willingness to ‘go then’, and constitutes a subversion of the apparent resolution that is its starting point. This alternative is most consonant with its narrative strategy, and with the elaborate patterns indicating equivocation, vacillation and inertia. It also seems to be promoted by a careful use (again) of verbs. For example, when the coming and going of the talking women is first mentioned, the account is made a general one by the immediately subsequent description in the past tense of the fog ‘curled’ about the locale of the party: what is described is not being witnessed, but remembered. Almost every detail of experience in the poem—involving the streets, the party, the beach, and so on—is carefully made general, proleptic or conditional: ‘streets that’, ‘there will be’, ‘They will say’, ‘I have known’, ‘I have seen’, ‘when I am’, ‘would it have’, ‘I shall wear’, and so on; hence, most are in the future, the present perfect and the subjunctive.

There seems to be three exceptions to this pattern of general, proleptic and conditional reference, in addition to the initial ‘Let us go’—three instances in the poem in which specific present circumstances are mentioned. The first is the couplet ‘Is it perfume from a dress / That makes me so digress?’; the second is the passage in which ‘the afternoon, the evening, sleeps … / … / … or it malingers / Stretched on the floor, here beside you and me’; and the third is the passage in which Eliot deftly characterises his poem by way of Prufrock's simile: exasperated by sunsets, dooryards, streets, other details made familiar by the poem, and ‘this’ itself, Prufrock exclaims: ‘It is impossible to say just what I mean! / But as if a magic lantern threw the nerves in patterns on a screen: / Would it have been worth while / … ’

Perhaps anachronistically for a poem written in 1911, some critics see a ‘cinematic-type structure’ in ‘Prufrock’, the magic lantern ‘presenting fragmentary shots’ in place of Prufrock's saying just what he means.²⁴ But while the nerves are his, the ‘lantern’ making the elegant patterns of them and the ‘screen’ on which they are focused have been shown to be the poet's. Prufrock's simile cannot be invoked as warrant to characterise Eliot's poem as indefinite. However, the passage does not contradict, and even suggests, that Prufrock is still at home.

The second of the three passages in which Prufrock mentions specific present circumstances provides additional evidence that he never stirs. It follows directly the determination, in the poem's two central paragraphs, of the outcome of his whole deliberation. The afternoon-evening is personified as either tired or malingering ‘here beside you and me’. To characterise evening—a relentless and constantly changing natural phenomenon—as tired or malingering, is nonsense unless the epithets apply as well to ‘you and me’, who are ‘here beside’ the evening ‘stretched on the floor’ of the room You and I have not yet left (although the party has already largely been described). And the afternoon has begun to be evening—time to go—Prufrock acknowledges. So he is not tired; he precisely malingers. His decision having just been made in the two central paragraphs, he can be expected to continue to—as he himself puts it—malinger.

The first of the three passages is the one element in the poem not wholly consonant with a reasonably certain answer to the question ‘Do You and I go?’ It is not perfume that causes Prufrock's digression. The poem gives the cause—arms. But to explain the question in his thought is difficult unless he is in the vicinity of dresses and perfume. Possibly, the perfume too is recollection from a familiar experience, but the words say something else.

This detail is not the only source of gratuitous uncertainty in the poem. For example, considering ‘all’ those bare arms that actually have hair, is the ‘one’ Prufrock mentions a specific person he wishes to court? Presumably, but by no means certainly. However, that uncertainty is a pointless minor ambiguity, not a significant inconsistency in the portrayed situation. The inconsistency of ‘Is it perfume from a dress’ is a positive flaw in a supposed ‘heterocosm’ which otherwise has no evidence that Prufrock stirs from his room. It makes uncertain Eliot's otherwise coherent and powerful portrayal of Prufrock as epitomising his predicament by his immobility, and so totally subverting the apparent resolution with which he begins.²⁵

Unlike inadequate reading, these inconsistencies and gratuitous ambiguities are uncertainties whose source is the work itself. The third source of uncertainty contravenes the Formalist-Cognitive doctrine, and general modernist assumption, that unless flawed, meaning in a poem is definite, that—at least ideally—there can be an Ideal Reader. The locus is the mimetic dimension of Eliot's poem—specifically, attitudes Prufrock is expressing. The source itself seems to be the human nature readers share with Eliot's character.

This third source of uncertainty must be distinguished from gratuitous uncertainty and pointless ambiguity about Prufrock's attitudes. For example, when he asks if he dares to eat a peach, is he pointedly citing something trivial to mock himself by its contrast with the symmetrically opposed portentous question ‘Do I dare?’? Or is his speaking of a peach purposive—perhaps metaphorically, or as a symbol? It may even be because of some current health fad; for his question is bracketed by his references to current youthful fashions of hair and trouser-bottoms. Hence, although self-mockery is not unusual for Prufrock, or inappropriate at this point, the context does not indicate whether or not his question is ironic—and it ought to do so.

A good example of the kind of uncertainty in ‘Prufrock’ that is not attributable to a failing either of reader, or of poem, is Prufrock's proto-imagist extended metaphor of the fog as a cat. At one extreme of a range of interpretations. A. D. Moody says of it, ‘the nightmarish and the seductive become confused. The cat is familiar, yet vaguely terrifying … ’ (p. 33). Hugh Kenner's Invisible Poet describes the ‘inimical clouds of yellow fog’ as part of Eliot's ‘incantation’ of ‘a hell’ (p. 10). Grover Smith's chapter on Eliot's pessimistic early poems is entitled ‘The Yellow Fog’; and he believes that an opposition between the house of the tea-party, and the fog around it, helps convey ‘Prufrock’'s sense of impotent inferiority or isolation’ (p. 18). Elizabeth Drew selects Smith's first alternative: ‘mingled self-pity and self-disgust, are … brought home to us through the images of the tortuous streets and the fog-cat’ (p. 35). Brooks and Warren specify the other, isolation—but it is ‘of the drawing room’; and they find the passage signifying the ‘relaxed, aimless quality of Prufrock's world’ as well (p. 391). J. Hillis Miller sees the fog as ‘express[ing] Prufrock's wish that he too could curl once about the house and fall asleep’ (p. 139). Morgan and Wohlstetter consider the cat ‘a sexual symbol’, with the fog expressing Prufrock's ‘desire’ as well as his ‘indecision’ (p. 33). George Williamson's similar interpretation is slightly more positive; the ‘mental state’ it ‘reflect[s]’ is unalloyed ‘desire which ends in inertia. If the cat image suggests sex, it also suggests the greater desire of inactivity’ (p. 60). And a note about the fog-cat by Arthur E. Waterman in The Explicator (XVII [June 1959], no. 67 [n. pag.]), calls ‘this beautiful image’ Prufrock's first ‘allusion to’ the idyll he contrasts to his situation: ‘it [the fog] is capable of natural freedom within the filth of the city … ’.

What is Prufrock's attitude toward the fog he likens to a cat? Whatever his uncertain attitude may be, his extended metaphor expressing it functions strictly to develop the portrayal of him; and all the quoted interpretations are consistent with the character portrayed elsewhere in the poem. Hence, there is neither a problem of gratuitous uncertainty, as with the references to perfume, ‘one’, and a peach, nor one of inadequate reading.

Precisely because none of the proposed answers to the question misread or ignore elements in the poem, and yet they differ so widely, they must derive from sources within the respective readers quoted that are not strictly readerly—must be, in the most precise sense, subjective. The third kind of uncertainty, indicating how far from real reading is the Ideal Reader, belies the modernist assumption that unless flawed, meaning in a poem is definite. Not all instances of this kind of uncertainty are so congenial as Prufrock's fog-cat.

It has been observed often that ‘Prufrock’ is filled with references to parts of the body. Actually, with the exception of the back of the fog-cat, those references are all to the two most expressive parts: the head (muzzle, tongue, hair, chin, face, eyes, head) and—in one instance ‘arms and legs’—the arms (hands, claws, fingers, arms). Recognising the exquisite interplay of related images of head and arms is one of the rewards the poem offers. The scant hair on Prufrock's head, the hair on the woman's arms, and the metaphorical hair of the waves combed by Prufrock's mermaids connect, and together augment, Eliot's portrait. Even more elaborately, the ‘restless nights in … hotels’ relates to the ‘lonely men … leaning out of windows’ on their arms whom Prufrock thinks of invoking for the purpose of a pathetically inept suit: an appeal for rescue from loneliness. He contemplates making it at the climax of the poem, in the one single instant he contemplates doing anything; and the leaning-men image of that instant relates to the stimulus to and reaction from the instant—respectively the female arms, and the crustacean claws he ‘should have been’—purposefully ‘Scuttling’, in contrast to his irresolute inertia. Finally, all relate to the fingers that ‘smooth’ his malingering immediately after the climactic section in which the instant occurs. Considering the centrality in the poem of this complex of images, the question ‘hands’ drop ‘on your plate’ may well be the ‘overwhelming’ question.

But is it? Are the two complexes of images actually brought together at the end, when Prufrock—no crab but ‘human’—‘drowns’? Does he drown because he lingers by the mermaids of his imagination, who are combing the hair of the waves? Is the secondary meaning of overwhelm—to submerge, engulf—invoked? Respecting these subordinate but tangible potentialities of meaning, is there a definite text in this class?

The issue is focused by the ‘Arms that are braceleted and white and bare / (But in the lamplight, downed with light brown hair!)’ The exclamation mark, the persistence in his consciousness of the arms, and the immediate occurrence of the climax:

Arms that lie along a table, or wrap about a shawl.
          And should I then presume?
          And how should I begin?
.....Shall I say …

indicate two things. The first is that his ‘digression’ is not caused by the perfume of the terse couplet between these two passages, but by that which is in parentheses: his image of real human female arms. And the second is that the image directly precedes and so precipitates his feeble and momentary resolution to act and save himself.

The resolution seems to me salutary in ‘The Love Song of J. Alfred Prufrock’. And Prufrock's ‘digression’ (characteristically) into what should be his proper state of mind, is prompted by the impulses of his You, responding to the image of the real arms—the healthy impulses which have compelled I to engage in his whole psychomachia that is the poem. I read ‘downed with light brown hair’ and the exclamation mark as indicating a favourable response to the physical reality Prufrock's image embodies. (It may not be irrelevant that I consider such a response understandable and healthy.) Consequently, I read the prompt frustration of his inclination to act as an index of the hopelessness of his predicament.

However, my account of the climax of the poem and its motivation depends on my inference that Prufrock is attracted by the real woman's human arms. Perhaps, instead, they repel him, and his repulsion is what causes his prompt abandonment of any attempt to save himself. Uncertainty about Prufrock's attitude toward those arms is far more critical than the similar situation with his fog-cat. Does the image motivate, or does it in fact defeat, his brief inclination to act? The difference is significant for our understanding of him and of the poem.

The ambiguous remark in A. D. Moody's recent book, ‘And what a complex of feeling there is in “But in the lamplight, downed with light brown hair!”’ (p. 34), may be his acknowledgement of both the problem and the contrast between my sort of reading and the exactly contrary one. That one is put succinctly by Brooks and Warren, who quote the line, and then ask:

Is this a mere observation, or does it indicate something about Prufrock? The fact that the observation of the ‘real’ arms is put in contrast with the ‘romantic’ arms, modifies the attraction: against the attraction there is a hint of revulsion, a hint of neurotic repudiation of the real, the physical. (p. 392)

Of course, the ‘something’ it indicates ‘about Prufrock’ can be ‘the attraction’, not ‘revulsion’. The dimensions of the problem are revealed in the context in which J. Hillis Miller quotes the line and the one preceding: ‘Eliot's early poetry is dominated by disgust for the body. The protagonists … recall James's heroes in their fastidious distaste for “Arms …”’ (p. 184).

Theories about Eliot's ‘emotive imagery’ are no help here. Does one generalise about Eliot's early poetry to learn how, at the age of 22, he would have regarded not a moral issue, but a natural and fairly general fact of human anatomy? Ought the poem to have specified, more than by the conjunction of elements it presents, and the familiarity with its subject's mind it allows, what response Prufrock would have had to that particular image? Is the passage gratuitously ambiguous? Perhaps the answer is Yes in both cases. But both the weakness of the evidence in favour of those conclusions, and the specific nature of the image at issue, suggest that contrary inferences about Prufrock's response to his image—about this aspect of Eliot's mimesis of a human—derive not from uncertainty in the poem, but from a source closer to home in each reader. And if that is so, is one's own possibly unformulated ‘aesthetic’ (in the most radical sense) response to down on women's arms not relevant to one's inference about Prufrock's response to arms ‘Downed with … hair!’? This last question is partly answered by the fact that the normal procedure of a reader unaware of the contrary inference from the parenthetical line is to draw his or her own with certainty, rather than to apprehend ambiguity in the passage. This procedure is very different from the reader's in analogous cases not involving an image that, in a sense, reads the reader. For example, although we know Prufrock, we do not know how sound is his suspicion that the ‘voices dying with a dying fall’ he recalls would do so because they had been talking about him; but we are aware we do not know—and it is not material.

Meaning may be immanent in the work of literature, but the act of cognition is a particular human's act, and ‘Prufrock’ provides evidence that in certain instances the human particularity and not the structure of words is—as many recent critics say of all instances—definitive.

IV. GHOSTLIER DEMARCATIONS

When in ‘Ulysses, Order, and Myth’ Eliot praised Ulysses for pioneering ‘the mythical method’, he was referring to its reliance on systematic allusion to the Odyssey as, in Joyce's phrase, a ‘way of working’: for Eliot the efficacious ‘method’ (‘a step toward making the modern world possible for art’) consisted not in employing mythic material, but in ‘manipulating a continuous parallel between contemporaneity and antiquity’.²⁶ Hence, even though his own recently-completed poem relied on the general myth of a wasteland and a quest for the redeeming Grail, Joyce's use of a particular antique literary work could be Eliot's exemplar. Enabling felicities the modernists valued, such as historical juxtaposition, tonal enrichment and parsimony, allusions understandably were used to do work of fundamental thematic and structural kinds in a great deal of modernist literature. Correspondingly, modernist critical practice was alert to the functioning of allusions in literature of all periods.

But the literary device was given little theoretical attention: the autonomous status of a work was not considered compromised by the allusions in it. For example, Formalist-Cognitive critics likened allusions to direct references, such as proper nouns and cited book titles; and all such special references were treated as simply analogous to the references to their lexical meanings made by conventional words and phrases. The general view was that a literary work employs a fact of a particular culture, its language, some of which is lexical, and some of which is referential to persons, places, concepts, myths and works of literature in the culture. Readers ‘need to understand the language of the poem including the ideas and allusions’.²⁷ And the special information readers require for the needed understanding of references in a work compromises the intrinsic status of those references—the autonomy of the work—no more than the special information required for understanding obsolete meanings or dialect words.

However, although the term allusion can be used loosely to designate any reference, its strict meaning of tacit as distinct from explicit reference not only is more useful, but also reveals the flaw in ignoring the distinction. To treat as language a ghostly quasireference, an inferred presence whose very existence by definition is uncertain, is to simplify its relationship to other elements of the work; and, no less than the uncertainties in ‘Prufrock’ attributable to readers’ human subjectivity, some of its allusions controvert the doctrine of definite meaning. But there is a more important consequence. To treat allusions as language is to obscure what an allusion does, when the manifest presence of the ghost in a work overcomes the uncertainty as to whether it exists. To treat the allusions in ‘Prufrock’ as language-in-an-autonomous-artifact is to obscure the way the literary device actually works in the poem.

In other words, ‘Prufrock’ exemplifies that the Formalist-Cognitive conception of allusion as a species of referential ‘language’ in an autonomous artifact obscures two significant characteristics of the device. Allusions in the poem controvert both the ‘cognitive’ doctrine that meaning in a literary work is definite, and the ‘formalist’ doctrine that a literary work is an autonomous artifact. For the two significant characteristics are that allusions are tacit, and they are conjunctive.

Formalist-Cognitive critics did identify one problem with allusion: the failure of some other critics to distinguish relevant special information from that which is irrelevant. An aspect of Intentionalism addressed in ‘The Intentional Fallacy’ is ‘to suppose that we do not know what a poet means unless we have traced him in his reading’; for ‘There is a difference between internal and external evidence for the meaning of a poem’ (Verbal Icon, pp. 14, 10). In contrast to external ‘author psychology’, however, ‘allusiveness’ in a poem is cited as an instance of internal evidence for meaning, a kind of language that functions as part of the poem, and so does not violate its integrity (pp. 14-15).

The last two pages of the revised and expanded commentary on ‘Prufrock’, in the second (1950) edition of Understanding Poetry, are devoted to the ‘question … raised by the literary allusions’ in it: ‘has the poet a right to expect this knowledge …?’ (p. 442). The answer is that the allusions in it are precisely both language, and functional: ‘the reading of any poetry requires some preparation’ (p. 444), and ‘the critics and scholars are there to help us. Then we can try to see if the allusions … are really functional …’ (p. 443). Although in the next edition and its final form (1960), the commentary was shortened by a number of excisions, an addition was made to the end of its last paragraph, the whole of which concerns literary allusions generally. Its new conclusion comprises a pair of sentences with two subjects. One is the then still-current calumny that Formalist-Cognitive critics—who worked to recast the history and revise the canon of English poetry—were antihistorical; the other is the status of allusion as ‘language of the poem’. ‘A poem does not exist in isolation. It exists in history’, Brooks and Warren declare, and, to ‘understand and appreciate’ it, one must know about both ‘the world that brought it to being and the world to which it refers’ (p. 399).

The conception of reference (including allusion) in literature as a kind of language seems more or less adequate for explicit references. For example, no important distortion of ‘Prufrock’ occurs when ‘Michelangelo’ or ‘the eternal Footman’ is conceived of as ‘internal’ and definite, merely requiring identifying information analogous to the lexical information required by archaic words. This is true as well of explicit references that are literary, such as ‘Prince Hamlet’ and the dual reference in ‘I am Lazarus, come from the dead’ (to the brother of Martha and Mary in John 11, and the beggar of the parable in Luke 16). It can be argued perhaps that since there are two, one of the Lazarus references must be tacit; but characters in specific works are designated, explicitly, as distinct from historical Dances or Jews, or even from the Amlothi of the Prose Edda or Saxo's Amleth: the language is definite, so to speak.

But tacit literary references are a different matter. The allusion in ‘Prufrock’ to another character originating in the Gospels reveals one inadequacy of the Formalist-Cognitive conception:

Though I have seen my head (grown slightly
          bald) brought in upon a platter,
I am no prophet …

The readiest ‘meaning’ to be given this ‘language’ would identify the account of the death of John the Baptist in Matthew 14 and Mark 6. However, an alternative meaning proposed for the ostensible language designates ‘the hero of Wilde's Salomé’ (‘Jokanaan’); and a third proposed identification is ‘Laforgue's John the Baptist’ (in Salomé, one of the Moralités légendaires).²⁸ Unlike the two accounts of different characters ‘come from the dead’ named Lazarus, the natures of the three very different works portraying John's death are such that all cannot be ‘meant’ by the ‘language’ of the passage in ‘Prufrock’.

According to the eminently practical criterion invoked by both ‘The Intentional Fallacy’ and Understanding Poetry, the narrative in the Gospels can be set aside as irrelevant (not functional). Comparing himself to the prophet, Prufrock speaks of his head as brought in on a platter—that is, into the room with ‘tea and cakes and ices’ in which ‘one’ sits. In the Gospels, Salomé is a young girl who has no relationship with or attitude toward the prophet, and asks Herod for his head at her mother's bidding. Wilde's ‘tragedy in one act’, in contrast, is an erotic action of heroic stoicism, unrequited love and revenge. Also, much is made of Jokanaan's thick black hair; and the seductive and implacable agent of his beheading laughs at his head when it is brought ‘in’ to Herod's party. The analogies and contrasts with the situation of the balding, self-conscious Prufrock, afraid to invoke, at a party, the banal power of his own temptress, are both apparent and richly functional.

In fact, both of the then-recent works about the death of John are suitable (‘functional’) objects of literary allusion in ‘Prufrock’. Laforgue's Salomé is relevant to Prufrock's situation and sense of himself. Furthermore, the general affinities Eliot's poem has with the characteristic tone, diction and versification of Laforgue's poetry constitute ‘internal evidence’ of the relevance of a poetic oeuvre Eliot often declared influenced him during that period; critics ‘have traced him in this reading’ of that oeuvre. However, Laforgue's Salomé simply is incompatible with Wilde's; its decadence the very opposite of sensual—enervated and nihilistic, with both principals rejecting life as trivial (‘nous désséchons de fringales supraterrestres’).

The problem with the conception of allusion as language which this element of ‘Prufrock’ illuminates is a variant of the problem raised by some of the true language in the poem: indeterminacy. There is absolutely no basis on which either Laforgue's Salomé or Wilde's Salomé, or both, can be positively identified as alluded to. The criterion of functionality weighs against the simultaneous allusion to both, and seems to indicate Wilde's play; but the process of determination is in the realm of degree, not kind. Identification originates in nothing more positive than a subjective assessment. Allusion being by definition tacit, the evidence for a suspected allusion always is circumstantial. And the weight of evidence assessed in reaching a judgement (the degree of likelihood determining attribution) is the cumulative testimony of such things as similarity of details in the two works, and number of similar details which all have relevance / value for the alluding work. A suspected allusion exists because, like Thoreau's trout in milk, it is present, and it is present because it functions. But how many of the three—and which—apparent allusions to works about the death of John the Baptist are (is) real? Whether or not a suspected allusion is functioning can itself be a difficult question.

The conclusion of ‘The Intentional Fallacy’ raises the question of indeterminate allusion in ‘Prufrock’:

… the line: ‘I have heard the mermaids singing, each to each,’ … bears a certain resemblance to a line in a Song by John Donne, ‘Teach me to heare Mermaides singing,’ so that … the critical question arises: Is Eliot's line an allusion to Donne's? (Verbal Icon, 17-18)

The concern of Wimsatt and Beardsley is to distinguish the intentionalist procedure in coping with the question—whereby ‘the critic writes to Eliot and asks … if he had Donne in mind’—from ‘the true and objective way of criticism’: ‘the way of poetic analysis and exegesis, which inquires whether it makes any sense … ’ They ignore the cognitive implications of allusion as indeterminate ‘language’; but they are fully aware of the problem:

This method of inquiry may lead to the conclusion that the given resemblance between Eliot and Donne is without significance and is better not thought of, or the method may have the disadvantage of providing no certain conclusion. (p. 18)

The circumstantial evidence inclines Wimsatt and Beardsley to the not-certain verdict that ‘the given resemblance’ to Donne's ‘Song’ fails to function in ‘Prufrock’—‘makes [no] sense’—so no allusion exists.

However, in his recent book Piers Gray cites the ‘winde’ that ‘Serves to advance an honest minde’ in Donne's poem, associates it with the wind blowing the waters in the ‘Prufrock’ passage, and eloquently makes the unusual case that the conclusion of Eliot's poem is positive, portraying ‘the transcendence of the lyric over the ironic’, on the grounds precisely that Prufrock has heard mermaids singing, signifying that his ‘honest’ mind, his imagination, will comfort him (pp. 74-9).

Wimsatt and Beardsley mention ‘a line of a sonnet of Gèrard de Nerval’ as evidence against Donne's ‘Song’ (p. 18). Presumably the line is one in ‘El Desdichado’: ‘J’ai rêvê dans la grotte où nage la sirèné. (It is quoted by Arthur Symons in The Symbolist Movement in Literature, the book that led the undergraduate Eliot to the French poets.) Gray invokes the same line, declares ‘We have lingered in the chambers of the sea’ to be ‘a rendition of’ it, and uses the Nerval to advance his interpretation based on the Donne poem (p. 79).

What ‘Prufrock’ exemplifies in this case is that even the arbitrary but practical modernist criterion of functionality (‘makes sense’, is ‘really functional’) can be not merely simply but doubly unreliable. It always involves a judgement incapable of proof; in addition, it can be logically and hermeneutically faulty. Explication itself can be determined by a prior, and so premature, higher-order activity of interpretation.

In this case, two different putative allusions contribute to the formulation of alternative explications of the ending of ‘Prufrock’—each of which view of the ending, by circular reasoning, is then used to demonstrate the functionality and hence the existence of the appropriate allusion. Arguing in his almost equally recent book the more familiar view that the end of ‘Prufrock’ is negative, A. D. Moody finds a persuasive allusion to a poem in which a merman at the shore calls without success to his mortal wife to leave civilisation and rejoin him in the sea, and which has a song, a reference to combing, and other similarities; but his bald assertion, ‘The allusion to Arnold's “The Foresaken Merman” is obvious’, actually emphasises the indeterminate status of his proposal (pp. 34-6). Allusion to Arnold's poem, direct or ironic, certainly would be incongruous with the positive ending Piers Gray describes, in which Prufrock has achieved ‘an imaginative triumph’, his mermaids either ‘an image of happiness’ or ‘an antecedent type of the female figure who is later to represent spiritual guidance’; it even would be incongruous with a less positive ending, in which the ‘drowned’ Prufrock has been made to ‘enter into real life’, much as baptism is ‘the first stage of true life or salvation’.²⁹ Allusion to Arnold's poem would be congruous with and enrich the negative ending A. D. Moody and most critics find, in which Prufrock's lingering in the sea by mermaids is terminated by human voices (perhaps his own two voices) waking ‘Us’, causing something like death.

But do the voices drown him, or do the mermaids? Mermaids traditionally drown human beings. However, his mermaids ignore him. Mermaids do not really exist, so ‘waking’ is accepting reality; is that as onerous as Prufrock himself believes? The many difficulties presented by the ending of the poem are relevant here because of its possible allusions. The Nerval line could be alluded to in either case. But a positive ending to the poem is major circumstantial evidence that the ending also is alluding to the Donne poem; and a negative one is major circumstantial evidence that, instead, it also is alluding to the Arnold. Simultaneously, the contrary potential allusions reciprocally help to effect the contrary potential endings.

Even when the circumstantial evidence by which an allusion is identified does not include a controversial interpretation (to which it reciprocally contributes), the identification remains—by definition—indeterminate. Indeterminate status is shared by all identified allusions in a work, while indeterminate meaning is only an exceptional circumstance in the case of lexical language, and even in the case of explicit reference. This fact alone creates a fundamental doubt about the Formalist-Cognitive conception of allusion as language.

That conception, expressive of the objective critics’ insistence on the autonomy of a work of literature, paradoxically withholds objective status from another literary work that is precisely the object of an allusion. Allusions not only are tacit; they also are conjunctive. If indeed the work alluded to were merely the meaning of language, it would constitute relevant literary history absorbed into the autonomous alluding work. As the object of an allusion, it is a discreet entity with a relationship to the alluding work.

The denial of separate status to the object of allusive ‘language’ is one manifestation of a general inadequacy in the Formalist-Cognitive hypostatising of the autonomy of a work. The inadequacy was identified early by the ‘neo-Aristotelian’ critics, who argued that Formalist-Cognitive critics are not objective with sufficient rigour, consider poetry ‘one of many modes of discourse’ rather than ‘a special class of made objects’, and so fail to make generic distinctions between ‘kinds of wholes’; hence the concept of ‘the intentional fallacy’ fails to accommodate some implications of the artist's sense of genre.³⁰

By definition, the ineluctable context of any individual work that is not literally sui generis includes ‘kinds of wholes’, whether they are part of its creator's former experience and operative awareness, or not; that context includes as well conventions, motifs, types of character and action, persistent imagery, all the elements of the entity literature that are exemplified in the work. The ‘neo-Aristotelian’ Critics and Criticism, Ancient and Modern was published the year before The Mirror and the Lamp; and four years before that, in 1948, Northrop Frye published ‘The Function of Criticism at the Present Time’, which he transformed into the ‘Polemical Introduction’ of The Anatomy of Criticism (Princeton: Princeton University Press, 1957). A principal subject of polemic in both essay and introduction was the failure of criticism to evolve a ‘conceptual framework’. The ‘organizing or containing forms’ of that framework would derive from ‘the assumption of total coherence’ in literature enabled by the axiom that ‘literature is not a piled aggregate of “works”, but an order of words’ (Anatomy, 16-17).

The ‘Co-ordinates of Art Criticism’ in The Mirror and The Lamp are restricted to one plane. A complete or three-dimensional conceptual model would situate behind Work, in a different order of relationship to it than Universe, Artist and Audience, if not the whole ‘verbal universe’ of literature as Frye maintains, certainly those Works that form the more limited context within literature of its genre, its imagery, and other characteristics.

Although in practice they appreciated the function for a particular work of its context within literature, the Formalist-Cognitive critics denied the importance of external, and so contextual, relationships to a work theoretically. They accommodated relevant literary history by absorbing its manifestations into the work. They were not joined in that denial by the creator of ‘Prufrock’, whose criticism most of them admired greatly. The title of Frye's essay alludes to that of Eliot's essay, which Eliot took from the essay by Arnold. In the revised version of his essay, Frye wrote:

Mr. Eliot's essay The Function of Criticism begins by laying down the principle that the existing monuments of literature form an ideal order among themselves, and are not simply collections of the writings of individuals. (Anatomy, 18)

On the second page of Selected Essays, beginning the earliest critical essay he chose to reprint, ‘Tradition and the Individual Talent’ (1919), Eliot made his famous statement that a writer works out of a ‘historical sense’ that all ‘the literature of Europe from Homer and … of his own country has a simultaneous existence and composes a simultaneous order.³¹ And in the next paragraph occurs the passage he chose to quote verbatim six years later, to begin the essay cited by Frye, in which he declares that the ‘complete meaning’ of any artist involves ‘his relation to’ antecedents: ‘The existing monuments form an ideal order among themselves … ’

A distinction must be made between Frye's postulation of a significant literary context for every literary work (or the structuralist concept of ‘utterances’ in literary works as ‘permutation of texts’ that Julia Kristeva named intertextualité;) on the one hand, and on the other, the neo-Aristotelian concern for a writer's awareness of literary provenance.³² The former relation of the individual work is to the ‘verbal universe’ of literature, and it is inevitable. The latter is to a specific literary context, and it is telic; it is to this that a writer's use of allusion is analogous. Eliot does not commit himself to either alternative; but the conclusion of ‘Ulysses, Order, and Myth’, published the same year as ‘The Function of Criticism’ (1923), declares that the value for modernist writers, of ‘the mythical method’ of fundamental reliance on allusion, derives from those writers’ strong sense of the relation between their contemporary culture and the past. And it is plain that, for the creator of ‘Prufrock’, the conception of allusion as language, with the object of that allusion absorbed in the alluding work, denies ‘the relation’ between his poem and its context in the ‘tradition’—in the ‘order’ of literature.

Since allusion is tacit, its indeterminacy is unavoidable, with the potential consequences already shown exemplified in ‘Prufrock’. But that is the minor significance of the inadequate Formalist-Cognitive conception of allusion as language which is contained in an autonomous work. The major significance of that conception is that it obscures the way all allusions function.

Being tacit, the allusion in ‘Prufrock’ to the circumstance of the prophet's death in the sensual Salomé; of Wilde (or—here judged less likely—in the nihilistic one of Laforgue) exists only because it functions. But in functioning, it does more than merely exist. It makes into a context the work alluded to, so that the nature of that work determines the attributes and the quality—supplies the value—of the functioning allusion. In other words, the nature of Wilde's Salomé; is invoked by the allusion; and so, if the allusion to Salomé; in ‘Prufrock’ is to accomplish its full effect, Wilde's play must have been experienced by the reader. Even the simple and explicit reference ‘No! I am not Prince Hamlet’ functions more eloquently for one who has experienced the play than for one given the relevant information.

An allusion (be it even a pun) is a kind of analogical trope. It is like the instrumental part (Richards’ ‘vehicle’) of a metaphor in that both are tacit, and both rely on relevance (functionality) to assert presence. And as the irrelevant qualities in the instrumental part of a metaphor are excluded from the analogy (such as the soft texture of red mats in the metaphor for a setting sun), so are irrelevant writings about the death of the prophet John the Baptist (such as the Gospel accounts) ‘better not thought of’, as ‘The Intentional Fallacy’ says of Donne's ‘Song’.

The currently popular doctrine that the apprehending of language is a non-referential ‘mental act’, it has been pointed out, denies the prior experience on which an image such as ‘red mat’ depends, when those two words are functioning effectively as the instrumental part of a metaphor. Correspondingly, the Formalist-Cognitive view of allusion as ‘internal’ language denies the prior experience (of Wilde's Salomé) on which an allusion to Salomé depends, when that allusion is functioning effectively in ‘Prufrock’.

‘Prufrock’ probably demonstrates most eloquently that literary allusion functions by invoking the literary object alluded to—by making that work a context in literature for the alluding work—with an allusion whose circumstantial evidence is generally considered incontrovertible. Even the title of his poem is implicated in the ironic relationship Eliot seems to have created between it and the most famous courtship poem in English, ‘To His Coy Mistress’. Two elements in ‘Prufrock’ effect the allusion most overtly. One is Prufrock's question ‘would it have been worth while’ to have ended ‘some talk of you and me’ in order ‘To have squeezed the universe into a ball / To roll it toward some overwhelming question’. The other is his repeated insistence that ‘we’, in contrast to Marvell's couple, have ‘enough’ time. The two elements are linked when ‘Do I dare / Disturb the universe’ occurs in the context of ‘there will be time’, and the first paragraph associates making ‘our visit’ with confronting the ‘overwhelming question’. Prufrock does not dare disturb his ordinary universe of ‘tea’, ‘porcelain’ and ‘talk of you and me’, by squeezing all of it into the ball Marvell's lover proposes for breaking ‘Through the iron gates of life’—the ball Prufrock would use on the question that overwhelms him. The analogy suggests that to put his question would be to break through the iron gates of his life; and the poem seems to indicate that. His overwhelming question corresponds to the question the very putting of which constitutes the confident exhortation by Marvell's lover that is Marvell's poem. And if he dared shape his ball, he believes, it would only roll ‘towards’ his question. In other words, the analogical relationship to ‘Prufrock’ of the seventeenth-century poem is profound, involving Prufrock's character, predicament and whole deliberation about his predicament; and the analogical relationship cannot fully do its work for these fundamental elements of Eliot's poem unless Marvell's poem itself—the order of words that creates the lover's wit, self-assertiveness, patronisation, sarcasm, erotic intensity and triumphant confidence—is part of the reader's experience. For the allusion to it invokes it as experienced reality.

Although Eliot belittles the ‘syllogistic relation’ of the ‘three strophes’ of Marvell's poem in ‘Andrew Marvell’ (Selected Essays, 254), that ‘relation’ enters into the ironic functioning of the allusion. For the three elements in Marvell's logical sequence of If (‘Had we’), But, and So (‘Now therefore’) employ respectively the subjunctive mood, the indicative mood and the imperative mood; and the body of ‘Prufrock’ (to the set of dots before the six-paragraph conclusion), which corresponds to the lover's exhortation, exactly reverses the sequence: So (‘then’), the sham conclusion in the imperative mood with which Prufrock begins, is followed promptly by But, employing various forms of the indicative (‘the women come and go’, ‘there will be time’, ‘I have known’), after which he progresses to his rationalising If (‘would it have been worth it’) in the subjunctive. The information that the logical sequence in the object of Eliot's allusion is reversed in his own poem could be imparted, like nuances of meaning in an archaic word. But here too, the eloquent and devastating irony Eliot has created cannot be fully experienced by a reader of ‘Prufrock’ unless a prior experience of ‘To His Coy Mistress’ is available for ‘Prufrock’ to invoke.

V. CONCLUSION

The title of Paul de Man's Blindness and Insight designates his thesis that a critic's insights controvert his or her (inevitably partially blind) general formulations about literature. Frank Lentricchia expresses my hope as well when he invokes that sense of ‘blindness’ in the Preface to After the New Criticism:

One of the lessons that I hope I've learned from reading contemporary criticism and philosophy is that no one is in a worse position to judge the blindnesses of a particular point of view than the one who subscribes to it … (pp. xi-xii)

Few would deny that the half-century, from the ascendancy of the literary movement that identified ‘Prufrock’ as one of its seminal works in English, to today, has produced one of the richest bodies of literary criticism in history. One aspect of the historical significance of Eliot's poem is its relevance for both the criticism belonging to its own period, and the challenges by critics of the past two decades. The insights produced by recent criticism augment the bountiful legacy to modernist criticism, to enrich us all; but there were inevitable blindnesses in general formulations during the half-century. Some fostered critical doctrines about the fundamental nature of literature that have important limitations.

Those doctrines concern the relationship of the meaning in a literary work to its nature as a composition of words; the extent to which its meaning, if immanent, also is determinate; and the extent to which its status is autonomous. ‘Prufrock’ eloquently exemplifies their limitations. That it could serve this purpose is additional evidence of its historical place; that it could do so without being diminished is additional evidence of its perpetual power.

Notes

1. ‘The Three Voices of Poetry’, On Poetry and Poets (New York: Farrar, Straus & Cudahy [Noonday], 1961) pp. 105, 104.

2. Most studies of the New Criticism seem to be essentially polemical. Among exceptions, Murray Krieger's The New Apologists for Poetry (1956; rpt. Bloomington: Indiana-Midland, 1963) probably deserves pride of place. Gerald Graff's book, Poetic Statement and Critical Dogma (Evanston, Ill.: Northwestern University Press, 1970), is less broadly-ranging and sympathetic. Professors Wellek and Graff debate the former's essay in Critical Inquiry 5 (1979) pp. 569-79. The last chapter of Frank Kermode's Romantic Image (New York: Macmillan, 1957) warrants mention for its relevant historical account of ‘the Symbolist conception of the work of art as aesthetic monad’ (p. 157) in modernist criticism.

3. ‘The poem conceived as a thing in between the poet and the audience is of course an abstraction. The poem is an act. … But if we are to lay hold of the poetic act to comprehend and evaluate it, and if it is to pass current as a critical object, it must be hypostatized.’ W. K. Wimsatt, The Verbal Icon (Louisville: University of Kentucky Press, 1954) p. xvii. The Chicago ‘Neo-Aristotelian’ critics’ emphasis on ‘wholes’ expresses the connection between Aristotle and the modernist critics.

4. For example, Wimsatt defines ‘poetry’ as ‘literature in its most intensive instances’ (Verbal Icon, p. xv).

5. See, for instance: review by Fred W. Householder, Jr. of Zellig S. Harris, Methods in Structural Linguistics (Chicago: University of Chicago Press, 1951), International Journal of American Linguistics, 18 (1952) pp. 260-8, p. 260; André Martinet, ‘Structure and Language’, in Structuralism, ed. Jacques Ehrmann (New York: Anchor Books, 1970) p. 7; and Allan Calder, ‘Constructive Mathematics’, Scientific American, 241 (October 1979) pp. 146-71 (quoted phrases on p. 146).

6. In Critical Inquiry, 5 (1978) pp. 143-59. The quotations are on pp. 143, 144, 145 and 156 respectively. Professor Ricoeur's distinguished book La métaphore vive is translated as The Rule of Metaphor (Toronto: University of Toronto Press, 1977).

7. Quoted from Wolfgang Iser, ‘The Reading Process: A Phenomenological Approach’, in Reader-Response Criticism, pp. 50-69, p. 68. It is reprinted from his The Implied Reader: Patterns of Communication in Prose Fiction From Bunyan to Beckett (Baltimore: Johns Hopkins University Press, 1974). The evolution in ‘reader-response’ criticism Professor Tompkins traces, in the course of introducing the selections in her volume, corresponds to the general history given here. For example: ‘Michael Riffaterre shares with [Walker] Gibson and [Gerald] Prince the assumption that literary meaning resides in the language of the text, but he attacks the idea that meaning exists independently of the reader's relation to it’ (p. xiii); ‘What Wolfgang Iser sees when he examines the same process is … a reader actively participating in the production of textual meaning. … But he does not grant the reader autonomy … from textual constraints’ (p. xv); ‘The next event in the drama … is that … the reader's activity is declared to be identical with [sic] the text. … Stanley Fish [is] the first critic to propose this theory of reading. … Meaning, according to Fish, is … an experience one has in the course of reading’ (p. xvi).

8. In T. S. Eliot: The Pattern in the Carpet (Berkeley: University of California Press, 1975), Elisabeth Schneider discusses the relationship between ‘the changing moods and tenses of verbs’ and ‘The rise and fall of the merest possibility of action’ (pp. 25-6).

9. R. G. Peterson, ‘Concentric Structure and “The Love Song of J. Alfred Prufrock”’, T. S. Eliot Review 3 (1976) pp. 25-8, adduces Eliot's roots in old traditions of number and symmetry’ (p. 28), and proposes ‘the repetition in reverse order of nine thematic groupings of obviously related images’ (p. 26), around the couplet about perfume from a dress ‘at the exact numerical center of the poem (11. 65-66)’ (p. 25). Although some of Peterson's formulation seem Procrustean to me, enough of the images and themes in the poem are symmetrically disposed to augment their other chiastic patterns. In ‘Critical Calculations:Measure and Symmetry in Literature’, PMLA 91 (1976) pp. 367-75, Peterson discusses ‘numerological and symmetrical patterns’ throughout Western literature, and recent critical attention to them.

10. He uses this typographical device again in three new poems in Poems (1920): ‘A Cooking Egg’, ‘Whispers of Immortality’, and ‘Mr. Eliot's Sunday Morning Service.’

11. Roland Barthes, ‘The Structuralist Activity’ (from Essais Critiques), quoted from European Literary Theory and Practice: From Existential Phenomenology to Structuralism, ed. Vernon W. Gras (New York: Dell, 1973) p. 161; Martinet, p. 4. An instance of Kenneth Burke's attraction for recent critics is his modulating these opposed views by defining form in literature, in the aptly-named ‘Psychology and Form’, as ‘the fulfillment of expectation’.

12. All four editions (1938, 1950, 1961, 1974) were published in New York by Henry Holt and Co. The commentary originally emphasised irony in the poem. It was reprinted, extensively revised and expanded, in 1950, and slightly revised and cut, in 1960; both poem and commentary were removed from the fourth edition. Unless otherwise specified, references are to the third (1960) edition. The other study is: Roberta Morgan and Albert Wohlstetter, ‘Observations on “Prufrock”’, The Harvard Advocate, 125, No. 3 (December 1938) pp. 27-30, 33-40.

13. ‘T. S. Eliot as a “Modernist” Poet’, pp. 353-77, p. 365, in Literary Theory and Structure: Essays in Honor of W. K. Wimsatt, ed. Frank Brady, John Palmer and Martin Price (New Haven: Yale University Press, 1973).

14. W. K. Wimsatt, ‘Prufrock and Maud: From Plot to Symbol’, Yale French Studies, No. 9 (Spring 1952) pp. 84-92, p. 92. It is reprinted in Hateful Contraries, pp. 201-12.

15. Piers Gray, T. S. Eliot's Intellectual and Poetic Development: 1909-1922 (Sussex: Harvester Press, 1982) p. 56.

16. Morris Weitz, ‘A “Reading” of Eliot's “The Love Song of J. Alfred Prufrock”’, The Philosophy of the Arts (Cambridge, Mass.: Harvard University Press, 1950) pp. 93-107, p. 95; Understanding Poetry, 2nd ed., pp. 434, 440. For the persistence of this view, see Burton Raffel, T. S. Eliot (New York: Ungar, 1982) p. 26.

17. Frederick W. Locke, ‘Dante and T. S. Eliot's Prufrock’, Modern Language Notes, 78 (1963) pp. 51-9.

18. Philip R. Headings, T. S. Eliot, Revised Edition (Boston: Twayne Publishers, 1982) pp. 24-5. The interpretation is not in the original (1964) edition.

19. F. O. Mathiessen, The Achievement of T. S. Eliot, 3rd ed. (1958); rpt. New York: Galaxy-Oxford University Press, 1959) p. 53. See also Elizabeth Drew, T. S. Eliot: The Design of His Poetry (New York: Scribner, 1949) p. 34; George Williamson, A Reader's Guide to T. S. Eliot: A Poem-By-Poem Analysis (1953; rpt. New York: Noonday, 1957) pp. 59-60; Grover Smith, T. S. Eliot's Poetry and Plays: A Study in Sources and Meanings (1956; rpt. Chicago: Phoenix-University of Chicago Press, 1961) p. 16; Robert Langbaum, The Poetry of Experience: The Dramatic Monologue in Modern Literary Tradition (1957; rpt. New York: Norton, 1963) p. 190; and Professor Kenner's book on Eliot, The Invisible Poet: T. S. Eliot (1959; rpt. New York: Harbinger-Harcourt, Brace and World, 1969) p. 6.

20. Gray, p. 68; on pp. 67-74, Eliot's conception of ‘double selves’ is related to the thought of Bergson and the contemporaneous French psychologist, Pierre Janet.

21. See respectively Understanding Poetry, pp. 392, 395 (2nd ed., pp. 436, 439); Langbaum, p. 190; Morgan and Wohlstetter, p. 27, and Williamson, p. 63; Leonard Unger, T. S. Eliot: Moments and Patterns (Minneapolis: University of Minnesota Press, 1966) pp. 164, 19; Alwyn Berland, ‘Some Techniques of Fiction in Poetry’, Essays in Criticism 4 (1954) pp. 380-5, 381-4; and Kenner, The Invisible Poet, p. 24.

22. Bernard Bergonzi, T. S. Eliot (New York: Macmillan-Collier, 1972) p. 16; and A. D. Moody, Thomas Sterns Eliot: Poet (Cambridge: Cambridge University Press, 1979) p. 33.

23. Headings, p. 21; and J. Hillis Miller, Poets of Reality: Six Twentieth-Century Writers (1965; rpt. New York: Atheneum, 1969) p. 139.

24. Gertrude Patterson, T. S. Eliot: poems in the making (Manchester: Manchester University Press, 1971) p. 110; Morgan and Wohlstetter, p. 27. See also Unger, T. S. Eliot, pp. 19-20.

25. For an even more apparent narrative inconsistency in The Waste Land, involving the use of quotation marks in ‘The Burial of the Dead’, see Stanley Sultan, Ulysses, The Waste Land and Modernism: A Jubilee Study (Port Washington, New York: Kennikat, 1977) pp. ‘63-4.

26. Originally published in The Dial, 75 (November 1923), pp. 480-3.

27. Cleanth Brooks, ‘The Problem of Belief and the Problem of Cognition’, Appendix 2 (pp. 226-38) in The Well Wrought Urn: Studies in the Structure of Poetry (London: Dobson, 1949) p. 227. See also the comment on ‘our habitual knowledge of the language’, in “The Intentional Fallacy’, Verbal Icon, p. 10.

28. The phrases are quoted from Williamson, p. 64 and Moody, p. 35. See also Morgan and Wohlstetter, pp. 27, 36-7, and Understanding Poetry, p. 393, for the Wilde identification.

29. Quoted from, respectively: Gray, p. 79; Unger, T. S. Eliot, p. 33; Weitz, pp. 102, 106.

30. See R. S. Crane, ‘Introduction’, Critics and Criticism, Ancient and Modern (Chicago: University of Chicago Press, 1952) pp. 13-17; the quoted phrases are on p. 13. Crane's thesis is that Aristotle was supplanted by a ‘Hellenistic-Roman Romantic tradition’; in that tradition, the Formalist-Cognitive critics lack ‘sufficient theoretical bases for’ considering ‘the peculiar natures of the artistic wholes’ writers ‘were engaged in constructing’ out of their ‘commitment to certain kinds of poetic structures and effects rather than others’ (p. 15).

31. Selected Essays (New York: Harcourt, 1932) p. 4. According to Valerie Eliot's prefatory note, the two 1917 essays in To Criticize the Critic were reprinted in response to requests.

32. For intertextualité, see, for instance, Julia Kristeva, Desire in Language: A Semiotic Approach to Literature and Art (ed. Roudiez), (New York: Columbia University Press, 1980) pp. 36-8.

    (I am indebted to Bernard Kaplan, Professor of Psychology at Clark University, for his helpful criticism.)