SOURCE: "Longinus at Colonus: The Grounding of Sublimity," in The Reach of Criticism: Method and Perception in Literary Theory, Yale University Press, 1983, pp. 47-80.

[Below, Fry uses Sophocles's Oedipus as a touchstone to compare Longinus and Aristotle. He concludes that the former discards fundamental distinctions—e.g., language and spirit—that are fundamental and problematic in the Poetics of the latter.]

The capacity to be able to act theoretically is defined for us by the fact that in attending to something it is possible to forget one's own purposes.… Theoria is a true sharing, not something active, but something passive (pathos), namely being totally involved in and carried away by what one sees [Hans-Georg Gadamer, Truth and Method]

In undertaking to show the relevance of Longinus to the concerns of criticism at the present time, it may be useful to begin by considering opinions of his treatise that are recorded by two modern theorists of criticism. W. K. Wimsatt thinks that On the Sublime is incoherent in every way, that Longinus is incapable of distinguishing clearly among author, text, and audience and incapable likewise of distinguishing between such pairs of terms as nature and art or thought and language. The result of these confusions is, according to Wimsatt, that Longinus cannot sufficiently distinguish even between his own five "causes" of the sublime, the two that are inborn (the power of forming great thoughts and the ability to feel passion) and the three that can be learned (use of figures, use of diction, and word order). It is imprecision of this sort that augurs poorly, in Wimsatt and Brooks's Short History, for the future of author-oriented criticism.¹ As far as Longinus is concerned, according to Wimsatt, a verbal work is only an accidental spark of contact between two souls, like a piece of loose wiring.

Elder Olson agrees with this last view, in effect, but he feels that the spark is a happy occasion. As he puts it, the Longinian sublime is "the communication of nobility."² Apart from this point, Olson's view of Longinus is almost entirely in disagreement with Wimsatt's. He tries to make Longinus a kind of Aristotle, with the result that in his hands On the Sublime becomes so coherent that even the contents of the massive lacunae can be inferred.³ Olson neatly determines that the "great thoughts" are the province of the author, the strong passions belong to the audience, and the three rhetorical categories refer the work to its own composition, to the author's choice of words, and to the affective quality of the words as sounds, respectively.

I think it inadvisable to attempt the rehabilitation of an obviously intuitive writer by stressing the well-ordered complexity of his ideas. Olson saw something he liked in Longinus but had to distort his author's way of proceeding enormously in order to bring it out. Wimsatt saw nothing he liked, but he saw the way Longinus's mind worked very clearly. He saw that Longinus is always talking about the same thing no matter how various the headings he devises may seem. I should like to make use of Olson's sympathy and Wimsatt's insight in order to show that the tendency of Longinus to slide from category to category and to leave large areas of overlap between his terms is in fact a highly desirable approach to theory, and one that is preferable to the formalism of Aristotle.

As theorists, both Olson and Wimsatt are primarily concerned, in their different ways, with structure: Olson to establish generic structures within which works can be identified, Wimsatt to define structure as an intrinsic quality of the works themselves.⁴ This common concern brings both of them, but especially Olson, closer to Aristotle than to Longinus. Longinus takes a Chicagoan view of the accidental or intermediary function of language, but he takes a decidedly anti-Chicagoan, agnostic view of the conceptual structures, if there are any, that exist apart from words. As Wimsatt points out unsympathetically, text and soul seem to be analogous in Longinus;⁵ neither has any definite shape or identity because each tends so easily to merge with other texts and other souls. Here too Longinus can be shown, however, to have taken a wise course. For him the question that has plagued Aristotle and his descendants never arises, the question as to whether structures are extrinsic or intrinsic to embodied works. The main question for Longinus is whether the knowledge that passes through discourse is finally to be understood in structural terms at all.

I do not wish to speculate, here or later, about the degree to which the argument of Longinus as I will interpret it is self-conscious, or cognizant of the implications that I will find in it. The Peri Hupsous is a series of fragments written by an unknown author; I will not deny that I take advantage of the vagueness that both characterizes and surrounds this work in order to build a poetics out of its hints and obliquities that can preside over the rest of this book. I will say, though, that I have paid more attention to the continuity of detail in the text than any commentator before me, and that I have not ignored or deliberately slanted any of it. To summarize what follows, then: I will concentrate on Wimsatt's observation that the difference between text and soul is very slight in Longinus. His hupsos, which means "height," or, in context, "elevated language," must be understood both as elevation and as language, difficult as it is to do so. His thinking can be viewed in two different but closely related ways. Either it is a theory of the interaction of consciousness with a phenomental world that is perceived almost as another consciousness, or else it is a theory of the fluidity with which utterances can move from one consciousness to another. The quality that gets transmitted can be understood either as language or as spirit but it cannot be divided into language and spirit, language and thought, or language and reference. Once these distinctions are set aside, it will become apparent that it is much the same, though never wholly the same, whether one speaks of nature or art, author or audience; or whether one speaks—to return to the distinction that emerged at the end of the last chapter—of unconsciously or consciously imposed form.⁶ Thus I will show Longinus to have accommodated, without by any means cathartically "resolving," the difficulty concerning the place and function of form that disturbs the argument of the Poetics.

I

In comparing Longinus and Aristotle it would be most appropriate, perhaps, to align the rhetoric handbook called Peri Hupsous with Aristotle's Rhetoric, especially because in the Rhetoric Aristotle allows more vividness and energy to the oratorical performer (1413b-14a) than he allows to the players in the Poetics. Speechmaking in the Rhetoric requires pathos and spectacle before all things, whereas dramatic imitation uses these aids as sparingly as possible. However, although in some ways the Rhetoric resembles the treatise of Longinus more closely than the Poetics does, the Peri Hupsous itself can be as readily considered a poetics as a rhetoric. Longinus maintains no distinction between modes of utterance, at least not in a systematic way. At the outset he speaks of his topic as a special quality that belongs to "the very greatest poets and prose writers."⁷ On two occasions later he does discriminate between poetry and prose (15.8, 30.2), but at these moments of greater specificity he has evidently remembered suddenly that he is supposed to be a rhetor—which would imply, of course, that oratory must be a special discipline.⁸ If it be granted, though, that the Peri Hupsous is both a rhetoric and a poetics, in this respect being yet another reflection of Longinus's failure, or unwillingness, to make distinctions, then the Poetics will not appear to be an inappropriate text for comparison after all.

Longinus intermittently discusses tragedy. He twice mentions Oedipus the King, once apparently to concur with Aristotle that it is supreme among tragedies (33.5) and once again to quote it, to quote as an instance of the sublime a central passage of the sort that Aristotle never considers:

   Weddings, weddings,
You bred me and again released my seed,
Made fathers, brothers, children, blood of kin,
Brides, wives, mothers—all
The deeds most horrid ever seen in men.
[23.3; trans. Russell]

Longinus cites this passage as an example of troping singular objects with plural endings. That is what it is; it stresses the multiplicity of horrors that befall Oedipus as if to remind Aristotle that misfortunes rarely come singly: they are not "one." There is nothing in Longinus about tragic structure,⁹ but still, on the strength of this passage alone, he inspires confidence in his ability to get at what is tragic in general, irrespective of what may be tragic in genre. In Aristotle, art, which is a movement external to the thing moved, traces a curve of order; in Longinus, as his plunge into the midst of the Oedipus would signify, "disorder goes with emotion, which is a disturbance and movement of the mind" (20.2; ataxia de to pathos, epei phora psuches kai sunkinesis estin). If the mind moves at all, if it transports or is transported, its movement is disordered and traces a disorder.

Longinus is not quite the rhapsodist that his cooler readers have made him out to be but he certainly lacks the Peripatetic neutrality of voice. His own consciousness of this difference is plain in his one citation of Aristotle, whom he makes to say, with Theophrastus, "that there are ways of softening bold metaphors—namely by saying 'as if, … Apology, they say, is a remedy for audacity." "I accept this doctrine," Longinus continues, "but …" (32.3).¹⁰ No one can either be or give the impression of being inspired who thus says "as if." Si vis meflere: he who expresses emotion must not just move but be, if not always witlessly possessed, at least greatly moved in his turn. Thus in moments of emotion the writer has "no chance to delay," as the onrush of what he says will have "outstripped its creator" (27.2, 1). Although Longinus is capable of sarcasm concerning the affectation of madness (see 15.8), this is not because madness or a state like it is extravagant but rather because its affectation in the wrong place is not extravagant. It is one form, "the pseudo-bacchanalian" form, of the false sublime (3.5). To be transported we must first have our feet on the ground; we cannot be duly and unflaggingly inspired, after the perfunctory suggestion of Aristotle, "by a strain of madness" (ch. 17).

Although he willingly turns to the tragic literature of the Attic period for his examples of the sublime, Longinus appears to have little use for "the trappings of the stage" (7.1).¹¹ His distaste for the excitements of performance seems even more extreme than Aristotle's and may at first confirm one's fear that the "sublime" (as Boileau translated hupsos, following Latin translations) will prove to be nothing but a fancy word for sublimation. Here, ostensibly, is the closest agreement between Longinus and the Poetics, to the effect, namely, that characterization, or the mere portrayal of "manners" (ethe), is inferior to, is indeed inessential to, the highest kind of expression. Narratives of manners, says Longinus, are the pastimes of old age: In the Odyssey the dotard Homer returns to the mere imitation of character which, as Aristotle had said, is also natural to childhood. (Whatever one may think of this judgment of the Odyssey, one should notice that at least it safeguards Longinus from emulating Aristotle's attempt to isolate a single course of action in that poem.)

For Longinus, as for Plato, greatness is steadfast, perhaps even inflexible; it eludes the many-talented Hyperides, for example, and appears instead in Demosthenes, who "has no sense of character" (34.3). Character in Longinus cannot be as freely varied as it is even in Aristotle. Slaves and women are permitted to speak by Aristotle as long as they speak in character, whereas in Longinus the drunkards of Herodotus (one who could arguably be released, as a historian, from the precept of poetic heightening) are censured for having spoken at all, even though, or rather because, they spoke in character (4.6-7). Longinus evidently feels, with Plato, that an ignoble taste for ventriloquism estranges the soul from the singleness of purpose it should cultivate. Even though Longinus easily violates this viewpoint in both taste and practice, it remains the most rigid, least thoughtfully integrated, and, I would say, least original feature of his thinking. That this view of character is what unfortunately leads to the facile indictment of his contemporaries in the last chapter is a disquieting aspect of Longinus's approach to art. It authorizes the Superman at certain moments; but in general, as I will try to show later, the sublime threatens more than it consoles autocracy of all kinds.

Not only is Longinus uninterested in the symmetries of dialogue and conflict that are made available by the imitation of character, but he also disdains, again in company with Plato, the group of emotions that carries drama forward. "Emotions, such as pity, grief, and fear" (8.2), he says, are "divorced from sublimity and [have] a low effect." These are just the emotions that determine the tragic structure in Aristotle. Again, though, this judgment is probably connected with Longinus's dislike of the ventriloquistic part of dramatic composition. His "fear" (phobos, as in Aristotle) refers only to cowardice, for example the cowardice, one may conjecture, of the suppliant Lykaon in the Iliad, which draws forth the famous response of Achilles: "Come friend, face your death, you too" (book 21; trans. Fitzgerald). But the fear involved in this exchange is not only dramatic; it is not only Lykaon's fear. There is fear for any listener in the reply of Achilles, and fear even for Achilles himself, perhaps, in having come to fathom the apathy of his own clearsightedness.¹² In a nondramatic setting, where emotions like fear can become ontologically charged rather than merely pragmatic, they could probably be restored, with the approval of Longinus, to the circle of emotions "without number" (22.1) that do readily supplement the sublime. But in rejecting pity and fear Longinus still does tellingly reverse the judgment of Aristotle. "In ordinary life," writes Longinus, "nothing is truly great which it is great to despise: wealth, honour, reputation, absolute power" (7.1). These are the "trappings" of status upon which the interest of the drama, considered as a representation of society, must solely depend. For a king who is not allowed to have developed a real personality, these possessions make up the whole of what is lost in the moment of tragic reversal.

As Longinus almost certainly did not know the Poetics, it is remarkable how precisely he inverted Aristotle's values.¹³ But to what end? Would it not seem again that Longinus, even more than Aristotle, is bent on doing away with the last contingencies of human life in criticism? I would say decidedly not. It is most important to qualify and delimit his distaste for the material things that have enslaved his sottish contemporaries and that the dramatization of despotism relies upon. Longinus describes the domestic atmosphere of the Odyssey, to take up a case in point, as a "realistic" one (9.15); that is, it is made up of biologoumena, the stuff of daily life that every theory of criticism until 1800 or thereabouts may be said to have banished from the precincts of all but the lowest genres. Now, this stuff cannot be what is missing in Longinus's earlier complaint that in the Odyssey "the mythical element [muthikon] predominates over the realistic [praktikon]" (9.14). Admittedly, it is not the details of reality but the Aristotelian "probable" that is in question here, a factor that I took to be purely intellectual in the last chapter. But in the case of Longinus there is, in the praktikos, a bias toward actual experience.

It is difficult to agree with Longinus that myth outweighs reality in the Odyssey. Instead we would want to say that the poem quite amazingly holds the mythical and the realistic in balance. The all-too-human gullibility of Polyphemus and the fireside knitting of Calypso are "natural" (for that is perhaps a better word than "realistic") whereas the massacre of the female domestics, on the other hand, is savage on a scale that is plausible only in myth. Perhaps, though, for Longinus, this interdependence of the real and the fantastic is just the problem. Homer's realism, he may feel, is not in good faith because it is not made interesting for itself or its determination of the action but only for its meretricious connection with the mood of folktale. If this is what Longinus means, it will not redeem his judgment of the Odyssey, but it will provide us with some reassurance that, for him, reality is indeed the basis of the sublime. And this basis is crucial, as I shall argue everywhere in the present book; without it the sublime is merely the quaint remnant from the rhetoric of ahistorical aestheticism that its critics suppose it to be. Longinus does not care for moonshine; he dislikes the fabulous in Homer because one cannot discover it either in "real life" (3.2) or in any religion that truly inspires awe, like that of "the lawgiver of the Jews"—"no ordinary man" (9.9), says Longinus, but certainly, he must suppose, a real one.

II

The sublime is "grounded," then. There is a necessary connection between sublimity and the earth, or more necessary in any case than the connection between art and nature in Aristotle. The Longinian sublime is also more closely connected with the earth than the Kantian sublime, which is the opposite of the beautiful precisely in not being given but only implied as an absence by the natural world. What makes this connection possible in Longinus is the absence of radical dualism from his thinking. The effect of his penchant for sliding categories is to promote mergers.¹⁴ There are mergers between genres, as we have seen, and there is also the merger of gods and men. This occurs in a passage that is mildly critical of Homer for having promoted the merger himself, but I think that too much has been made of Longinus's Platonic piety in this instance.¹⁵ Homer exalts even while he demeans: In the Iliad he has made "the men of the Trojan war gods, and the gods men" (6.7). It is quite a different matter to have blended superstition and folkways in the Odyssey; in the Iliad the stakes of existence, and its conditions, are fought for at the height of human potential, which is figured forth in the attacks of Diomedes and Achilles on the gods and in the appearance of the unarmed Achilles, under the aegis of Athena, that fells twelve Trojans with heart failure. This height (hupsos) of human strength merges in turn with the voice of the poet, whose Iliad was composed "at the height of his powers" and at a "consistent level of elevation" (9.13).

It is this last sort of fusion, in which the person inspired takes on the qualities of what inspired him, that is most often noted both by sympathetic and by hostile readers of Longinus. The commentator in turn is inspired. The sublime, understood as the "echo of a noble mind" (9.2), is transmitted from the text to the voice of its author to the voice of the commentator, who can stand, as a result, in the place of the author: "Filled with joy and pride, we come to believe we have created what we have only heard" (7.2). As it is effected rhetorically by Longinus, this identification may remind one of the fallacy of imitative form, with the important proviso that it is not formal. Thanks to the looseness of his categories, Longinus can subordinate the structure even of his own insights to the continuousness of experience. He does not deliberately constitute himself, or so he would imply, as one who is inspired—as the great sublime he draws. We are conscious of his rhetoric not as the imposition of a pattern but as a movement;¹⁶ first there is the interjection into his own text of some fragment, which is often broken up still further by misquotation; then Longinus reminds us of the force of heroism—or divinity—that is mirrored in authorship; and finally there comes an onnish of commentary, tumbling out, confused, in every way the record of experience, not of reflection (see 9.6-7, 10.3). The sublime in the theory and practice of Longinus is not infinite, though it may intimate boundlessness; it is always an experience in time, and thus restores to reading what formalization, which is necessarily spatial, has removed.

I have made no proper distinction, so far, between intention and accomplishment. I have spoken of "false" madness as though sincere madness were easy—and desirable—to single out, and then again I have spoken of sublimity and "rhetoric" in the style of Longinus as though there were no difference between them, or as though the difference did not matter. It does matter, but perhaps only when it is considered as success or failure in the representation of perceptiveness. All the least futile discussions of sincerity, from Johnson on Lycidas to Richards on The Chinese Classics to Lionel Trilling on Jane Austen, have touched only lightly on the question of hypocrisy and stressed, instead, a correlation that can be demonstrated, at least in part, between the occasion of an utterance and its manner. Thus in Longinus insincere madness is "untimely … emotion where none is in place" (3.5). If madness is appropriate to an occasion, however, one is then free to affect it in all sanity. If the real Erinyes should attack an orator where he stands as though he were Orestes, then even if he is calm at heart he may resort hypocritically to the figure called phantasia, or "visualization," which is said to betoken madness.

There is no necessary connection, therefore, between sincerity—considered as spontaneity—and the sublime. Even the noble Demosthenes could scarcely stand trial for sincerity in Longinus's reading of his rhetorical questions: "Emotion carries us away more easily when it seems to be generated by the occasion rather than deliberately assumed by the speaker.… The figure of question and answer arrests the hearer and cheats him into believing that all the points made were raised and are being put into words on the spur of the moment" (18.2). On the other hand, however—and this is a crucial distinction—at the affective end of the sublime experience sincerity is essential. For the reader or commentator it is not enough even to be persuaded, "for persuasion on the whole is something we can control" (1.4). If we are free to disregard the sublime, it is not the sublime. With respect not to one's reading but to one's reader, however, as the chain adds links, one is again freed, having taken the place of the author, to dissemble enthusiasm if need be.¹⁷

The very complexity of this issue, which is scarcely allowed to arise in the Poetics, must show that in Longinus there is no easy way to subsume art in nature or nature in art. There remains evidence in his own text, instead, of the conflict between the two which is just the conflict between design and compulsion that I discussed earlier. One reason why no clear relation can be established between art and nature is that for Longinus they are not easy to distinguish. He vacillates in his treatment of them but his vacillation is rigorous, I think, rather than weak-minded. For him the issue is at bottom a moral one, touching as it does on the question whether "we can develop our nature to some degree of greatness" (1.1)—the question, in other words, whether we can improve our nature by art. As a teacher justifying his own existence, he must part company with the widespread opinion, dating from Pindar, that greatness is solely "a natural product" (2.1), but he plainly feels that he must also stop short of the Sophists' notion that there are rules for all things given that all things are unknowable and must therefore be devised by artifice as need arises. The former view is typically that of the poets and the latter that of the rhetors, and Longinus, as usual, takes his stand between them.

Although his eventual assignment of two natural and three artificial causes to the sublime suggests that nature and art are fully separable, the discussion that leads up to this list is more complicated and—apparently—more confused. Of the "three points" (2.2) he makes to refute the contention that genius is artless, the second anticipates the later headings and the third is mostly cliche, but the first is significant: "Though nature is on the whole a law unto herself in matters of emotion and elevation, she is not a random force and does not work altogether without method." If this is so, nature is partly art and can receive from art, in that case, only a supplement of itself. The last sentence of the next paragraph (2.3) again undermines the standard contrast, this time not by merger but by dialectic: "The very fact that some things in literature depend on nature alone can itself be learned only from art." Whether by "art" here is meant criticism, or the reflective judgment, or the recognition of having failed to achieve nature by artificial means, is not wholly clear. What is clear, though, is that the sentence is reversible: "Art" and "nature" could change places, with "nature" now meaning "experience."¹⁸ This possible reversal would point to an exactly complementary moment in the learning process, and would suggest, in its very exactness, that nature and art are not casual aids to each other but two facets of an indivisible dynamic. The exemplary Demosthenes later illustrates this dyad more than once, showing, for example, that "sobriety is needed even under the influence of inspiration" (16.4).

A more interesting later outgrowth of this interaction of categories occurs near the end of Longinus's by and large conventional contrast between genius and mediocrity: "We may say that accuracy is admired in art and grandeur in nature, and it is by nature that man is endowed with the power of speech" (36.3). Thus, language belongs to man's natural course of development and is not an art implanted in man by fiat. The notion that man differs from the animals in possessing inborn art, the chief sign of which is his conversion of random sounds into the articulate sounds of speech, was perhaps the most crucial presupposition of Aristotle's formalism in the Poetics. Longinus's view is neither that art complements nature nor that it is nature but simply that it comes to us along with the rest of our inheritance. If this is so, art cannot be an activity we perform but must be instead an activity that takes place in us—like nature. Longinus cannot quite say, then, with the formalist tradition from Aristotle to the present, that art, "in cooperation with the conscious will" (Coleridge), defends the fortress of the self against the siege of nature. He appears to suggest, rather, that both art and nature come as strangers to hold their combat in a remote corner of what at first we may not even recognize as the self: "Do you not admire the way in which [Sappho] brings everything together—mind and body, hearing and tongue, eyes and skin? She seems to have lost them all, and to be looking for them as though they were external to her" (10.3). This difficult response to the so-called "Ode to Anactoria" is drawn forth in particular from Sappho's "My tongue is broken, a subtle fire runs under my skin."¹⁹

III

The "broken tongue" of Sappho could be an emblem of the Longinian sublime. Her alienation from herself resembles the sudden switch to apostrophe whereby Demosthenes "divides a single thought between two persons in his passion" (27.3) and thus parodies the course of the sublime from transmitter to receiver: "this shameless monster, who—you vile wretch!" Longinus says that Demosthenes is touched by genius because, unlike Hyperides, he "lacks fluency" (34.3). But it is not only this loose application of Sappho's trope that merits attention. There is also the more precise metaphor in which the broken tongue represents the lapse into incoherence, the disarticulation of syntax or "semiotic discontinuity," as Thomas Weiskel has called it, that is caused by certain figures of speech.²⁰Polyptoton, the term that designates all repetitions of a word in different inflections, is also the general term that includes Demosthenes' change of person, and it is one of the figures that Longinus stresses most.

The sublime figures and tropes, which I shall now survey, are not quite sublime in themselves any more than art is quite nature, although, as we have seen, neither of these paired sets of terms constitutes a clear dichotomy. Just as Sappho's broken tongue is not itself her speech—neither her allegedly inarticulate stammerings nor the well-formed "sapphics" that record them—so in turn speech itself is not the sublime.²¹ If the sublime were indeed a property of words themselves, it could be quantified with respect to their combinations. It might prove, for example, to be some "principle of equivalence" like that of Jakobson and would in any case certainly appear as a function of structure. In other words it would be part of the matrix of continuous, orderly composition by which, according to Longinus, the orator effects "persuasion." But "persuasion" in Longinus is just the opposite of "transport," which is caused by the sublime (1.4). Persuasion is an affect that we are free to resist. It arises from the "ability to order and arrange material" and takes its effect, such as it is, "when we see the whole context" (1.4). The sublime, on the other hand, is what we cannot resist. It is not surprising that we can stand aloof from persuasion. It is more surprising that we can be persuaded at all, except by the most translucent of styles, because carefully organized discourse forms a latticework that seems to resist us; it proclaims its autonomy far more than its authority. In recognition of these characteristics, the New Criticism affirms the autonomy of the work of art. The sublime, on the other hand, as Longinus's descriptive terms for it will show, is that which forces its way through an opening it has widened in the latticework of persuasion. It has no other way of appearing; as Kant also showed, its occasion is natural, but it cannot appear in or through nature without a prior appeal to a formulation of the mind—an inner discourse—that arises in response to nature. Thus it is not a quality of words, but it does depend on their close proximity.

Just so in criticism, I would venture to add, the sublime will not appear in formal discussion as such, but it cannot appear without the context of a formal discussion. The sublime is "what is left to the imagination"; it is not the words in the text or a paraphrase of them but it is still prompted by them. Where the text has left words out or transposed or delayed them, the imagination must supply them. I shall try to demonstrate that this familiar exercise of multiple choice has all the attributes that Longinus and his successors call sublime. Far from cloaking itself in a nimbus of ineffability, the sublime is closely related to interpretation itself. "It is only through inevitable omissions," writes Wolfgang Iser, "that a story gains its dynamism."²² These omissions are the reader's share. The sublime (to continue to call it that for the time being) is potential in any interpretation that does not suppose that it can either leave a text as it found it or else exhaust it by assigning it a form. The sublime that is closest to interpretation responds to a text by augmenting it. What Longinus calls auxesis, or "amplification," is not the sublime, he says, because, like persuasion, it is determined finally by quantity; but it is still close to the sublime, close enough to call for a fine discrimination. "Sublimity depends on elevation, whereas amplification involves extension; sublimity exists often in a single thought, amplification cannot exist without a certain … superfluity" (12.1).

Amplification, about which there will be more to say in another place, is not a "broken" figure. One of the few exceptions to the criterion of fragmentation in Longinus's survey of figures, it remains close to the sublime because it overloads "persuasive" composition as much as the sublime disrupts it. A more typical figure that generates the sublime is asyndeton, which presents words or phrases "without connection" (19.1). There is nothing deliberative or slow, nothing spondaic, as it were, in the wide spacing of this figure; rather it rushes by in order to get beyond an impasse which is passed along, in the process, to the reader's effort of reconstruction: "Disconnected and yet hurried phrases convey the impression of an agitation which obstructs the reader and drives him on. Such is the effect of Homer's asyndeta" (ibid.). The passage Longinus has quoted just before this observation marks a crisis of speech in the Odyssey which is also a crisis for the status of speech as a means of persuasion. The breathless Eurylochus returns to tell Odysseus that Circe has turned their comrades into "squealing," inarticulate swine, "though [their] minds were still unchanged" (book 10; trans. Fitzgerald). Panic-stricken as he is, Eurylochus is unable to imitate in his asyndetic speech the well-knit construction of what he saw: "We went as you told us, noble Odysseus, up the woods, / We saw a beautiful palace built in the glades" (ibid.; trans. Russell, 19.2). As befits its ambiguity for men, Circe's palace is both persuasive and sublime, a contexture in a "glade" or gap; but in its outward structure it is like the figure that constrasts with asyndeton in Longinus, polysyndeton, which creates "smoothness by conjunctions" between phrases (21.2) and thus undermines the "harsh character of the emotion" (21.1) one would find in the same phrases written without conjunctions.

Another figure conducive to the sublime is hyperbaton, which is an arrangement of words that differs from the "normal sequence" (22.1). This figure creates gaps not in syntax but in expectation, whether by delay or by prematurity. From the standpoint of the author, however, hyperbaton is fully unified, not in the arrangement of its parts but with respect to the ground of thought or feeling from which it arises. More than any other figure, hyperbaton implies spontaneity and brings the problematic terms art and nature into their closest possible conjunction: "Hyperbaton is the means by which … imitation approaches the effect of nature" (22.1). Again, ars est celare artem. Yet, working under this concealment, the art of hyperbaton actually disrupts the nature it pretends to resemble: Thucydides and Demosthenes "show ingenuity in separating by transpositions even things which are by nature completely unified and indivisible" (22.3). Thus it is not the figure of speech itself that the sublime thunderbolt gaps (as one gaps a sparkplug) but rather the natural ground of the figure. Hyperbaton opens abysses, vacancies. If we recall the full play of the word arthra in Aristotle, which according to usage means both the presence and the absence of jointure, we may also grasp the significance for Longinus not only of Sappho's broken tongue but also of the antifeminist trope, quoted from Plato's Timaeus, that labels "the seat of the desires 'the woman's quarters"' (32.5). The sublime thus can be further understood as a condition of desire, an intimation of presence transmitted not only through a figural rift but also through a "cleft in the ground" (13.2; the phrase appears in an important passage that will occupy us below).

The topics of Longinus's quotations nearly always reflect, in some way, the rhetorical devices they are chosen to illustrate.²³ This is to some extent a consequence of the "representative" form that Pope was inspired—partly by Longinus—to illustrate in his Essay on Criticism. But in the case of Longinus the pattern of representative form is so continuous that it seems, irrespective of purpose, to be a necessity of expression. Evidently Longinus is drawn as much to a certain view of nature as to a certain group of tropes and figures. As hyperbaton reveals, nature itself is noble, impassioned, and broken. We noticed before that Longinus, unlike Aristotle, stresses the plurality of Oedipus's misfortunes; he assumes that the heart of tragic experience is the moment of dismemberment, not resolution, and he expects that that moment will be represented in language that risks inarticulateness. The language that prepares for tragedy's bloody sowing will likewise be scattered abroad; hyperbaton, asyndeton, and polyptoton are all present in the passage about careening flight that Longinus quotes from Euripides' lost Phaethon:

"Steer towards the seven Pleiads."
The boy listened so far, then seized the reins,
Whipped up the winged team, and let them
 go.
To heaven's expanse they flew.
His father rode behind on Sirius,
Giving the boy advice: "That's your way,
  there:
Turn here, turn there."

Equally characteristic are the passages that describe in nature the wound they illustrate in syntax. Under the heading of "ordinary words" and how to use them, Longinus indulges in an outburst of rapid-fire quotation: "'Cleomenes in his madness cut his own flesh into little pieces with a knife till he sliced himself to death.' 'Pythes continued fighting on the ship until he was cut into joints [i.e., steaks]"' (31.2). The effect of slicing is just the same in hyperbaton as Demosthenes uses it: "Now, for our affairs are on the razor's edge, men of lonia …" (22.1).

Nothing could have a more overtly rhetorical effect than the violence in these "ordinary words." At present among critics there is a tendency to discredit the discipline of "rhetoric" as it has been traditionally practiced because it is so difficult to imagine a "pure" employment of language that would not be rhetorical.²⁴ If to some degree nature is revealed as art by the ubiquity of rhetoric, it is equally true, however, that in the absence of a zero-degree or full transparency in language, art disappears back into nature simply for the reason that the province of art has no discernible boundaries. One wonders whether Longinus, whose technical categories so readily dissolve, might not agree with this recent view. Is there an Ordinary Language (idiotismos; 31.1) whose idioms are not cut to pieces by use? What is the "normal sequence" that the artifice of hyperbaton disrupts?

IV

Although for the most part Longinus's observations about the need of art to conceal art are commonplace, there is one moment at which he seems to become fully attentive to Horace's maxim. In preparing to assert, inter alia, that hyperbaton approaches nature most nearly of all the figures, Longinus declares that "figures are natural allies of sublimity" (17.1). "Allies," here, is a precise metaphor. Longinus is about to describe with surprising, almost digressive, amplitude the adversary relation between the orator and the prince or judge who listens to him. What the orator must avoid, says Longinus, is the danger of making the figural gaps, or "fallacies," of his oration look like subterfuges; he must not "raise the suspicion of a trap, a deep design." For Longinus, again, the irregularity of figures reflects the violence in nature, and now he seems to perceive that with a provenance thus shocking figuration itself will appear to be an evil that must be concealed. And certainly, for all his complaisance, the orator is indeed an aggressor.²⁵ Except in epideictic oratory, which Longinus does not discuss, the orator is always trying to prevail over opposition while frequently pretending not to be hostile. If his eloquence fails, the violence of nature, of his nature, and of the nature of his speech, will have a colorful outlet indeed: "Such a person [as "tyrants, kings, governors …"] immediately becomes angry if he is led astray like a foolish child by some orator's figures. He takes the fallacy as indicating contempt for himself. He becomes like a wild animal" (italics mine).

Here a complication arises. Although there is violence in all figures, evidently it is not the studied figures but the ones that are most wildly irregular that will seem the most natural and hence irritate the hearer as little as possible. The inkhorn techniques, on the other hand, the ones that are conducive to the unity of composition and are therefore not sublime, bring on a violent reaction—the reaction, say, of Wordsworth to "poetic diction" or of Whitman to "the beauty disease" ("Poetry To-day in America"). The force of figures that are themselves broken by violence goes unnoticed but does so only because it has overwhelmed and disarmed opposition without the knowledge of the opponent. "Amazement and wonder exert invincible power and force and get the better of every hearer" (1.4). The sublime stuns the hearer, spends its aggression in so doing and becomes, once it has entered the hearer, the appreciative condition known as "transport" (ecstasis). This process is parallel to, and I think more generally useful in criticism than, the process of catharsis in Aristotle.

The Longinian catharsis is a conversion of power into light, and in this Longinus differs from later theorists of the sublime. The difference between hupsos and the sublime of Edmund Burke is not so great as it is sometimes said to be,²⁶ but there is, in tendency at least, one important distinction to be stressed. Burke's sublime wears many guises but in general it inclines toward mist and obscurity, whereas the sublime of Longinus begins in darkness but then bursts forth into clarity and illumination: "'God said'—now what?—'Let there be light,' and there was light' (9.9). The illustration that comes next makes the pattern clear while showing what the triumph of light in the sublime is like. The "sheer brilliance" (17.2) of the sublime speaker may be a cruel light but it is also honest; without sublimation it shows even death for what it is: "Darkness falls suddenly. Thickest night blinds the Greek army. Ajax is bewildered. 'O Father Zeus,' he cries,"

'Deliver the sons of the Achacans out of the
 mist,
Make the sky clear, and let us see;
In the light—kill us.'

It is as if, through the premature expiation of having asked to be killed by his "Father," Ajax is hoping that in coming to light prior to its manifestation as force, the oedipal situation need never develop.²⁷ The wished-for moment will be violent, certainly, but at least the suppliant will have deferred to authority, just as the orator must defer to the ruler who will become "a wild animal" if he suspects that he is being mocked by the evasions of figure. In each of these three instances the coming-to-light is a purification of aggression: of divine omnipotence heretofore without an outlet, of blind rage in battle, and of rhetorical guile, respectively. It is a blazing forth that is apt to be followed by a general diffusion of force once pressure has been released by the cleft in the ground. Power becomes light in an allegory of acculturation.

Longinus's famous list of natural analogies for the sublime includes the brilliant light that issues from "the craters of Etna, whose eruptions bring up rocks and whole hills out of the depths, and sometimes pour forth rivers of earth-born, spontaneous fire" (35.4). It also includes the force of the ocean, which elsewhere has the might of a god, as when Poseidon causes Hades to quake in the underworld "for fear the earth-shaker … might break through the ground" (9.6). In every analogy Longinus can think of, some force breaks through a natural barrier, just as the sublime breaks through the contexture of figures. But what Poseidon threatens in this last instance is in fact such a powerful rupture that even Longinus wishes to channel it. The threat of "the whole universe overthrown and broken up" makes Homer's battle of the gods "blasphemous and indecent unless it is interpreted allegorically" (9.7). "Much better," says Longinus in the next paragraph, are the passages in which the gods, like the orator or like his judge bemused in a state of transport, are no longer hostile. Even then, in an episode Longinus prefers to the battle of the gods, the image of the gap appears once more, this time constructively, with the coming of Poseidon: "The sea parted in joy, and the horses flew onward" (9.8). The effect of these two passages about Poseidon when they are read in sequence exactly parallels the progress of the sublime from the broken ground of chaos toward the benign channeling of its force. Longinus's next two examples, to confirm the pattern, are the one given from Moses, in which "now what?" makes a gap in the quotation, and the one from Homer's speech of Ajax, who longs for the incisiveness—"kill us"—of the light that pierces darkness.

In keeping with this same pattern, the effect of Demosthenes' oratory is at first violent and then less so. Longinus quotes the passage beginning "The aggressor would do many things" (20.1), and then shows how Demosthenes responds in kind: "The orator is doing here exactly what the bully does—hitting the jury in the mind with blow after blow" (20.2). By the end of the paragraph, which concerns the mixture of a regular, recurrent figure, anaphora, with an irregular one, asyndeton, Demosthenes has somewhat diffused his force: "His order becomes disorderly, and his disorder in turn acquires a certain order" (20.3). Demosthenes, who has "divine gifts, it is almost blasphemous to call them human" (34.4), always subjects the hearer to an "abrupt sublimity" (12.4) which has the effect of being "catapulted out" (21.2) from the gap or opening we have been keeping in view. The effect he has on his audience is twice compared with thunder and lightning (12.4, 34.4). This simile as Longinus uses it is not at all novel in ancient literature, even with reference to Demosthenes, but in Longinus it becomes a motif and recalls the crucial preliminary definition of the sublime: "Sublimity … tears everything up [panta diaphorese] like a whirlwind [1.4; Prickard: "lightning flash"; Roberts: "scatters everything before it like a thunderbolt"]." When Aristotle expresses his fear that if one part of a tragedy is moved from its place "the whole will be disjointed," he uses the word diapheresthai. Once again, then, Longinus can be seen to have planted his unstable compound in the very midst of the Aristotelian order.

V

It is not clear whether Longinus's "good friend" the philospher in the final chapter is an opponent putting a wrongheaded argument or simply Longinus himself varying his delivery, good orator that he is, by putting one of two equally valid arguments in another voice. I incline to the second possibility, in part because to suppose that the two speakers are one will bring into consideration a special effect of the Longinian sublime, namely, the identification of the listener with the speaker, which I shall now discuss from various standpoints. Whatever the diagnostic merits of the friend's case (which will be found in Tacitus and many others), the rhetoric of his closing comment provides the essential background for the sublime: "I understand that the cages in which dwarfs or Pygmies are kept not only prevent the growth of the prisoners but cripple them.… One might describe all slavery, even the most justified, as a cage for the soul, a universal prison" (44.5).

Although this speaker does retain a conception of "justified" slavery (meaning only justified imprisonment, perhaps), we can still infer from his remarks a state of culture in which, despite the overthrow of democracy by despotism, the institution of slavery could be subject to criticism as it never really was in the thought of Plato or Aristotle. For them, each person, even if he is not officially a slave, must understand his station in the social and moral order to be fixed within justifiable restraints. Especially in Plato, this narrowness (adherence to one vocation, for example) is good for the soul and best suits its natural, uncorrupted wish for conformance with the Good. Hence Plato cannot share the rather complex opinion without which the idea of the sublime cannot be entertained, the opinion of Longinus that the soul by nature craves freedom whether its constraint is justified or not. Supposing that there is merit in this opinion, either man is an entity divided against itself or else his soul is a visitant from without, longing for escape from an oppressive host. Either of these views makes a sublime object of man himself: no longer a whole or a unity, man is divided by a fault through which what is sensed as alien and senses him as alien comes and goes.

There is unquestionably an element of the uncanny in the sublime, a quality which has been convincingly evoked by Freud as the sudden appearance of oneself as another.²⁸ It is by his sense of the uncanny, arguably, that Longinus's ascription of sublimity to speeches favoring the oppression of aliens is dictated. In light of these remarks, here is a long passage in which it is hard to decide, as one reads, who is the slave, who the governor, and who the liberator, or to decide which of these is the host, at the present moment, of the sublime in transit:

It is when it [visualization] is closely involved with factual arguments that it enslaves the hearer as well as persuading him. "Suppose you heard a shout this very moment outside the court, and someone said that the prison had been broken open and the prisoners had escaped—no one, young or old, would be so casual as not to give what help he could. And if someone then came forward and said 'This is the man who let them out,' our friend would never get a hearing; it would be the end of him." There is a similar instance in Hyperides' defense of himself when he was on trial for the proposal to liberate the slaves which he put forward after the defeat. "It was not the proposer," he said, "who drew up this decree: it was the battle of Chaeronea." (15.9-10)

The first emphasis falls upon fact—ironically, perhaps, because the facts will be very difficult to pin down. It is visualization" (phantasia) that "enslaves" the hearer, that much is clear, but we do not know whether this device works for or against the facts. Neither do we know, as yet, why the sublime is said to be an enslavement of the hearer that cooperates with rational persuasion, when hitherto in the Peri Hupsous transport and persuasion have been treated as alternatives that are practically synonymous with rebellion and conformity, respectively.

All these questions are answered, but in a very surprising way, in a later passage: "We are diverted from the demonstration [of fact] to the astonishment caused by the visualization which by its very brilliance conceals the factual aspect. This is a natural reaction: when two things are joined together, the stronger attracts to itself the force of the weaker." In short, might makes right, but only when might is subversive. Whatever the opinion of Demosthenes concerning slavery and whatever sympathy his idolator Longinus may have for it, the force of the sublime as Longinus records it covertly transfers power from the oppressor to the oppressed. The "enslaved" hearer, in the first place, is the citizen who has custody of the prisoners; or, if he is a first-century reader of Demosthenes, he is a slave by virtue of the despotism he lives under. (Longinus's interlocutor Terentianus was possibly a Roman senator in prospect,²⁹ but he would nonetheless merely have been one of the emperor's "flatterers in the grand manner" [44.3].) Furthermore, in the first quotation, a voice from "outside the court" penetrates it with tidings of escaping prisoners, and the prison in turn is said to have been "broken open." Liberator, prisoners, messenger: these in turn burst through some barrier and thereby prove themselves, and not the possessors of temporal power, to have carried out the movement of the sublime exactly.

The situation of Hyperides, who is not elsewhere considered to be sublime (see 34.1-4), is equally complex. He was once a liberator and is now on trial for his temerity, an outcome which in the previous citation Demosthenes declares to be impossible (so great is his respect for lawless force). Thus enslaved, Hyperides would appear to have transmitted his power to the slaves he released, and yet he is now speaking with sublime effect on his own behalf because the power of yet another has been transmitted to him: It was not me, he says, but the disruption caused by the battle which did this thing.

All the confusion in this long passage arises, then, from the unexpected discovery on the part of each speaker in turn, including Longinus, of the self in the other. If perhaps in some cases it is the other way around, if it is the self that seems to have been invaded, that is because the sense of demoniacal possession Freud speaks of in this context is sure to appear where there are political passions. It is the rule of "transport" in the sublime, as this passage also shows repeatedly, that to have power one must be enslaved, possessed by another. As an instance of this intimation of Hegel's master-slave dialectic in operation, one may cite the subversion of Edmund Burke's politics by his aesthetics. In his Reflections on the Revolution in France, the established throne and all the other "fixed forms" of government, especially the pitiable queen, exactly conform with Burke's condescending notion of the Beautiful in the Enquiry, whereas the mob in full cry on 6 October is much closer to his notion of the awesome Sublime. Cattle are nothing much, he says in the Enquiry, but the wild ass in Job who knows no master is sublime.³⁰ What then, in the Reflections, of the "thousands of great cattle reposed beneath the shadow of the British oak," who commendably "chew the cud and are silent"?³¹

By "visualization," Longinus means that in phantasy, as hearers, we subvert our own identities. We conjure up presences that usurp our places. The hearer neither sees Demosthenes nor remains conscious of his own freedom but sees instead the hitherto unknown liberator of the slaves and becomes himself a slave to what he sees. He discovers himself to be Oedipus as surely as the Aristotelian audience does not. But in this discovery there is strength. If the sublime is a possession that is distinguished by a coming-to-light, and thus casts out the uncanny almost in the instant of evoking it, its effect may be "amazement" (12.5) and clarity all at once, an epiphanic moment of presence that quickly becomes self-presence. When what is absent stands before one, even takes possession of one's consciousness, first there is a recoil, as from an invader, and then there is a surge of empathy, as in the "momentary checking of the vital powers and … consequent stronger outflow of them" that accompanies the sublime in Kant's Critique of Judgment.³²

Common and nowadays somewhat ineffective rhetorical devices for this making present in Longinus are the sudden switch from past to present tense in narrative (25) and the sudden apostrophe to the hearer (26.1-2). We can still be affected, though, by the skillful use of a third device, conjuration. Demosthenes' oath, "By those who risked their lives at Marathon" (16.2), is remarkably glossed by Longinus: "He was suddenly inspired to give voice to the oath by the heroes of Greece." Demosthenes does not call up the past, then; past voices call through him, thus enacting the course of the sublime from speaker to speaker. The finest event of this kind in Longinus is "the appearance of Achilles to the departing fleet over his tomb" (15.7). The heroic past, which is in a sense the whole text of the Iliad and the other cyclic poems, stands above a gap in the ground whence it has come, still alive with, or revived by, the charge for posterity it carries. The past in the person of Achilles haunts the puny survivors whose victory has been won and who are now launched, a "departing fleet," upon the historic decline Longinus laments in his final chapter. "Greatness of mind wanes," he says there (44.8)—unless it should return to possess the living.

VI

The sublime is chthonic, "earth-born" like the volcano, yet it is also divine—or else it is the human euhemeristically exalted. It appears at the horizons of perceptual experience that are left out of the Poetics. Hence Longinus's touchstones frequently bring extreme conceptions of matter and spirit together. He quotes, for example, the Homeric account of the rebellious giants, who themselves are heavy and dull and make heavy work of their climb into the sky: "Ossa on Olympus they sought to heap; and on Ossa / Pelion with its shaking forest"—shaking like the beard of a giant—"to make a path to heaven—" (8.2). In their lumpishness these giants must use the arithmetical auxesis rather than the true hupsos or megethos (greatness) to serve their purpose. At the same time, however, their instinct cannot be taken lightly if it enables them to forge a "path" toward their goal: "And they would have finished their work …" This coincidence of the animal and the divine is repeated, not so much in the next Homeric passage itself—"So long is the stride of the gods' thundering horses,"—as in Longinus's response to it: "If the horses of the gods took two more strides like that, they would find there was not enough room in the world" (9.5).

The oxymoron of measureless substance is exemplified for Longinus by the ocean. He himself takes its measure, thinking perhaps of a strait or bay that becomes a lake at low tide, in the course of assigning an inferior place to the Odyssey: "We see greatness on the ebb," he says, glancing forward again to the waning of greatness in his last chapter: "It is as though the Ocean were withdrawing into itself and flowing quietly in its own bed" (9.13). When a vast, possibly boundless expanse becomes a self-contained structure, it stops flowing out of itself. It no longer influences in the way the ghost of Achilles, representing Homer, can influence, not to say inundate, a departing fleet. There is no "outpouring" (ibid.) from the Odyssey that can affect us. In a passage that wholly anticipates Kant, Longinus identifies the sublime as something illimitable in the mind, something that must overflow and appear in contrast with the finite objects in nature: "The universe therefore is not wide enough for the range of human speculation and intellect. Our thoughts often travel beyond the boundaries of our surroundings. If anyone wants to know what we were born for, let him look … above all [at] the Ocean" (35.3-4). Just as boundlessness and measure are closely interdependent, so also, it will be recalled, art and nature, words and thought, are nonidentical word pairs that are nonetheless inseparable. Freedom, to repeat the essential homily concerning the sublime, needs its obverse in confinement. We could not identify the "abundant, uncontrolled flood" of Archilochus if it were not a "bursting forth of the divine spirit" from "under the rule of law" (33.5).

The sublime, again, is neither a great mind, precisely, nor the words of a great mind but rather "the echo of a noble mind" (9.2).³³ If we are to take this turn of phrase seriously, the sublime must then be the reverberation, or resonance, of words. That characterization may bring it a shade too close to language, however, too close to uttered language, that is, and for this reason Longinus's ensuing sentences seem to counteract the sense of the word echo: "This is why a mere idea, without verbal expression, is sometimes admired for its nobility—just as Ajax's silence in the Vision of the Dead is grand and indeed more sublime than any words could have been."³⁴

The distinction between silence and the echo in this passage is as crucial as it is subtle. Again Longinus has turned to a crisis of communication in the Odyssey. An attempt to open a path between heroic ghosts and present needs, the crisis begins when Odysseus invites the sublime by making a wound in the earth. "With my drawn blade/I spaded up the votive pit," he recalls, all for the purpose of speaking with Tiresias, "the prince of those with gift of speech" (book 11). All the women and men who are allowed to approach after Tiresias has spoken his fill are echoes of the seer, reverse echoes, as it were, which confirm the authenticity of his prophecies with the accuracy of their recollections. The eloquent silence of Ajax, which is balanced as a motif by the vagueness of Tiresias's forecast beyond the moment of planting the oar, is made possible by the eloquence of Tiresias on all prior topics, without the exhaustion of which Ajax could not have approached the pit.³⁵ Furthermore, the silence of Ajax is itself pregnant with speech, as even the snubbed Odysseus realizes: "Who knows if in that darkness he might still / have spoken, and I answered?" The silence of Ajax, in short, is wholly dependent on words—on Odysseus's present narrative, on the speech Odysseus made to Ajax rehearsing the old grievance, and on our own feeling that even in that moment of growing darkness, which recalls Ajax's darkling petition for death quoted earlier by Longinus, the suicide could have said much.³⁶

The sublime echo thus wavers between the aftersound of words and their negation. It is as difficult to locate precisely as the ghost of Ajax disappearing into Erebus. The "voice of the dead" in such moments is not really a voice; rather it is the speaker's memory short-circuited so that it seems as though the dead were now saying what the speaker has always known them to have said when alive. Even if the "dead" should in fact be living persons, the effect of astonishment brought about in the speaker by means of visualization, brokenness of figure, and his accession to "divine gifts" makes the speaker seem merely a mouthpiece for what is remote from his ordinary sphere. "'The dead writers are remote from us,"' wrote Eliot, "'because we know so much more than they did.' Precisely, and they are that which we know."³⁷ In the experience of the sublime they are not remote, however, but so vividly present that they take possession of consciousness.

The controversy that surrounds the theory of influence today has in part to do with the question of whether influence occurs in the form of words or thought, echo or silence. The early Humanists thought that it is the words of the past that either intimidate or educate us. Edward Young thought that the spirit of the past can liberate our own genius if we disregard the words. Harold Bloom imagines a powerful presence from the past, at once word and spirit, which will dominate the new speaker unless he contrives to sap its strength. Longinus can be helpful in ascertaining how far the idea of influence can be carried beyond the discussion of buried allusion, but more importantly he can show how far beyond this point it needs to be carried. If nonverbal thought is necessarily at least "colored" by reading, that coloring is not washable. It must be seen as a dye: "The choice of correct and magnificant words … makes … a kind of lustre bloom upon our words as upon beautiful statues … it is indeed true that beautiful words are the light that illuminates thought" (30.1-2). Allusion, then, which can be defined as words taken from their place in history, colors the whole mind. This is a far subtler view than the notion of words as the "dress" of thought, but it is also more troublesome when it comes to assessing originality. When it is difficult to discriminate between words and thought, it becomes more difficult to know how great a portion of what is spoken belongs properly to the speaker.

In that case more than ever, the speaker must fight for his share of credit as an originator or "author."³⁸ This is the point at which he ceases to be the "hearer" and assumes, in his turn, the role of orator. Whatever it is that reaches the hearer, and perhaps echo will have proved as good a word as any for it, he is at first stunned but then he reacts, and in so doing he passes into the speaker's role.³⁹ The aggression of the original speaker is met with counteraggression on the part of the listener. One such listener is Longinus, who cannot praise his literary ancestor Plato without some word of qualification. There is a measure of tension in his reverence for Plato from which his reverence for Demosthenes, for example, is wholly free. On several occasions Longinus taxes Plato with abuses of style. The "otherwise divine Plato" (4.6) too often exhibits the vice of frigidity, and he is also given to such needless and dull embellishments as periphrasis (29.1) and surplus metaphors (32.2). In these belittling gestures, which have the effect of pointing to something petty in Plato's nature, the reader will recognize one reflex of the anxiety of influence in the theory of Harold Bloom. It is hard not to conclude that Demosthenes is praised more unreservedly than Plato is because he is less important. Demosthenes commanded only one register of the sublime, and his style does not exert any direct influence on the sinuous and chameleonic style of Longinus himself. By far the more powerful influence upon Longinus is Plato, whose style and whose thinking—his concept of nobility, his dislike of the theater—figure with great prominence in the Peri Hupsous.

There is an additional charge to be whispered against Plato. With Xenophon, says Longinus, Plato was "trained in Socrates' school" (4.4); or again, late in the passage toward which the present analysis is tending, Plato "diverted to himself countless rills from the Homeric spring" (13.3). In short, Plato can be charged with unoriginality as well as pettiness. Longinus is aware that his readiness to find fault is a fault—"I myself cited not a few mistakes in Homer and other great writers" (33.4)—but he goes on to argue, notoriously from the standpoint of his detractors, that the faults he finds in others are sure proofs of their genius; and in that case his own fault would prove that he himself is at least not insipid. Even in this gesture of restitution toward genius, however, there is a trace of malice. Take Caecilius, he continues, author of the treatise on the sublime against which he has pitched his present essay: consider what Caecilius says against Plato. Apparently Longinus is about to defend Plato. But then, with a devious stratagem common both in oratory and in neoclassical wit, Longinus faithfully repeats the indictment made by Caecilius: "Loving Lysias more deeply than he loves himself'—possibly the talk of philia here is meant to recall the Phaedrus—"[Caecilius] yet hates Plato with an even greater intensity.… In preferring Lysias to Plato he thinks he is preferring a faultless and pure writer to one who makes many mistakes" (32.8). Plato is ever so faintly damned, then, both by contrast with Lysias—for Longinus will not glorify mistakes until the next chapter—and also by comparison with him. In repeating the invidious comparison of Caecilius, Longinus himself causes Plato to appear merely as a competitor, and an unsuccessful one at that, for superiority in phrasemaking.

Longinus's attitude toward Plato is not in itself terribly significant. I have lingered over it to show further that captiousness of some kind belongs properly to the dynamics of the sublime (a point to which I shall return), and to suggest that this inevitably critical moment does raise an epistemological question: Is the sublime firmly enough situated in the nature of things to survive criticism? It is the first premise of Longinus that the sublime is objectively verifiable. What he says to establish this point has a Johnsonian ring: "Reckon those things which please everybody all the time as genuinely and finely sublime" (7.4). Here Longinus differs from Kant, whom he anticipates in so many other respects, and rejoins the eighteenth-century "empiricist" writers, from Addison to Burke and Kames, who found the sublime indisputably to be present in certain works of art and natural objects.

At least, this is partly so. Coleridge, a Kantian whose Biographia attacks the premises of empiricism, still struggles whenever possible, more clearly so in other writings, to establish an objective ground of judgment,⁴⁰ and Longinus, no matter what he says, must undertake the same struggle. Although, again, nature plays a vital role in the production of the sublime for Longinus, it is not clear that nature, which "is not a random force and does not work altogether without method," can be very safely distinguished from art. It is not easy to decide whether the sublime passes from object to subject as a volcanic echo or from subject to subject as an oratorical echo. Coleridge, in the well-known anecdote of the lady he heard calling a waterfall "pretty," himself knew, absolutely, that it was "sublime." But the anecdote at least partly bears witness against itself because Coleridge's knowledge, absolute or otherwise, would not have been possible without information given by reading and by the oral tradition of the educated.⁴¹

The Longinian sublime cannot be located in words or things with confidence for similar reasons. His conviction that the sublime pleases all at all times must be undermined, furthermore, by the prevalence of the instinct he notices in himself and his predecessors, like Caecilius, for faultfinding. If the sublime is the most exalting experience one can have that is not purely religious, and if it must therefore be spontaneously desired by the soul (we can reasonably expect that Longinus would want to take a Platonic view of the question), then why do we act so aggressively toward it? Why do we decry its pretenders and even express doubts about the illustrations of it that we bring forward ourselves? It may be replied that vigilance is needed to protect the fane of the sublime from false priests who go about "poking the thyrsus in at the wrong place" (3.5; my adaptation of the term parenthyrsus). Yet it is troubling that such vigilance leaves so little intact. The Odyssey, the battle of the gods in the Iliad, Plato—all but the safely peripheral Demosthenes, peripheral both in style and in politics, are quite badly mauled. As Longinus himself would say, they are "dissected": "We have to ask ourselves whether any particular example does not give a show of grandeur which, for all its accidental trappings, will, when dissected, prove vain and hollow" (7.1).

It appears from this metaphor that the aggression of the hearer is an imitative aggression, corresponding as it does to the tearing apart that marks the appearance of the sublime itself. Although it takes place in a different mood, perhaps, the rending of the sublime still anticipates interpretation, the dissection that Wordsworth called murder. The listener's response to the sublime resists its assault. At first the listener is possessed, and then he bestirs himself to expel the alien voice, which ceases to possess him in the process and becomes his property instead: "It is our nature to be elevated and exalted by true sublimity, Filled with joy and pride, we come to believe we have created what we only heard" (7.2).⁴²

After these last observations have been made, the dramatization of mimesis that follows will almost interpret itself:

Many are possessed by a spirit not their own. It is like what we are told of the Pythia at Delphi: she is in contact with the tripod near the cleft in the ground which (they say) exhales a divine vapour, and she is thereupon made pregnant by the supernatural power and prophesies as one inspired. Similarly, the genius of the ancients acts as a kind of oracular cavern, and effluences flow from it into the minds of their imitators.… Plato could not have put such a brilliant finish on his philosophical doctrines or so often risen to poetical subjects and poetical language, if he had not tried, and tried wholeheartedly, to compete for the prize against Homer.… As Hesiod says, "this strife is good for men[.]" Truly it is a noble contest and prize of honour, and one well worth winning, in which to be defeated by one's elders is itself no disgrace.

We can apply this to ourselves.

(13.2-14.1)

Possession, which is divine, comes from a cleft in the ground that represents the natural origin of the sublime. The priestess is then the "author" of the sublime, and we who receive her prophecies are hearers-turned-authors or hearers-turned-interpreters. After the Delphic comparison Longinus changes the subject a little, revealing from that point forward the aggression in the relationship between speaker and listener.

VII

Nearly all commentators suppose that there is a drastic narrowing of the Aristotelian mimesis in Longinus and other late Greek and Latin writers. Verbal imitation has replaced ideal representation, symptomizing the overthrow of true philosophy by the "Alexandrian" logomachy. But in the case of Longinus this judgment is unduly harsh.⁴³ In his treatise the spirit is not suppressed, leaving only the gaunt body of philosophy in view; rather spirit and letter are nearly interchangeable, and each very adequately embraces the mimesis of Aristotle. The common distinction we make between mimesis, alluding to Plato and Aristotle, and imitatio, alluding chiefly to Quintilian, Longinus, and the rise of Humanism, is a good deal more radical than it should be. Pope's lines on Virgil are perhaps closest to the equipoise of Longinus: "When t'examine every part he came, / Nature and Homer were, he found, the same" (Essay on Criticism, I, 134-35). The "nature" that Pope has in mind is the phusis, or "rules," of Aristotle, but the "nature" of Longinus, had he written this couplet, would have been more dynamic.

One way in which Longinus does maintain a slight functional separation between nature and art is in speaking, as he routinely does, of "mere" words. These are what the false sublime is made of: it looks sublime, but in the long run, for "the man of sense and literary experience …, it fails to dispose his mind to greatness or to leave him with more to reflect upon than was contained in the mere words" (7.3). Only the implications of the word more can save Longinus from saying, here, that the sublime itself is merely a matter of words. It is not solely because this conclusion would trivialize the sublime that Longinus is fortunate to have evaded it; he also escapes having to admit that as a "text-book" (1.1) his treatise must be a failure. And indeed it is a failure in that genre. Because by his own demonstration there is no trope or figure that is not subject to abuse, then truly, though Longinus has argued to the contrary (2.1), the sublime cannot be taught as a rhetoric. This he does later admit in an unguarded moment: "Evils often come from the same source as blessings; and so, since beauty of style, sublimity, and charm all conduce to successful writing, they are also causes and principles not only of success but of failure" (5).

If the sublime is not a rhetoric, and I think Longinus knows it is not, it cannot be identified with any structure. This article of belief must be kept in mind in turning to a number of passages that seem to undermine it by identifying the sublime with order and resolution. The first of these describes the Iliad during Longinus's contrast of that poem with the Odyssey. The Iliad, he says, has a "consistent level of elevation which never admit[s] of any falling off (9.13). What is perplexing about this statement is that elsewhere Longinus has insisted that the sublime appears in disruptive fragments of less brilliant surrounding compositions. If "persuasion" is achieved by observing the Horatian decorum (see 1.4), the "transport" of the sublime must be a breach of decorum. There are several ways of resolving this anomaly. First, it could be said that a whole text, seamless in itself, would become sublime if it disrupted other texts. The reader of the Iliad, especially if he or she is a poet, might find that as a whole it overwhelms his or her own inspiration. It could also be said that for the Iliad to have unified the story of Troy is a feat that surpasses ordinary powers of composition. Thus the composition of the Iliad is an extreme instance of making the plural singular, which will always result, as Longinus says elsewhere, in "surprise" (24.2).

A third rationale for Longinus's suddenly holistic phrasing, briefly stated, might be this: When he speaks of sustained elevation he is not actually touching upon questions of structure or even of unity. He is speaking, rather, of a device that more closely resembles the blows of Demosthenes' bully, the device, namely, of repetition. This is the most satisfactory explanation for Longinus's praise of the Iliad because it can also be applied to his other totalizing figures. This I shall do in a moment, having first resisted the temptation to gloss over the anomalous passages prematurely. For they are anomalous. Recalling Aristotle, for example, Longinus claims that periodic structure in a sentence can be sublime: "The beauty of the body depends on the way the limbs are joined together, each one when severed from the others having nothing remarkable about it, but the whole forming a perfect unity" (40.1). He pursues this train of thought pertinaciously: "Similarly, great thoughts which lack connexion are themselves wasted and waste the total sublime effect" (sundiaphorei kai to hupsos). Because it is not evident how more than one great thought can appear in, or even be discussed with respect to, the structure of a single sentence, Longinus's notion of the "total sublime effect" continues, in this context, to be unintelligible. Perhaps he is affirming what is crucial, again, in any theory of the sublime that has subtlety: the principle that sublime moments need a formal context of some sort and cannot take their course in a vacuum.

If this is so, it remains only to characterize the interstices of the "great thoughts." It is not clear from this passage alone whether sublimities should come at us in a steady stream, creating, e.g., the Iliad's "consistent level of elevation," or whether they should be indirectly but no less surely connected by decent intervals of decorum. Longinus himself begins to resolve this question, and to show how the sublime can indeed appear in formal composition, a few sentences later in this same discussion of periodicity. Having quoted a passage from a lost play of Euripides about Dirce being haled up and down by a bull, a passage that is as agitated as its subject ("it writhed and twisted round"), Longinus remarks that "the word-harmony is not hurried and does not run smoothly" (40.4). He means that clusters like perix helixas and petran drun slow one down in pronouncing them. For this reason "the words are propped up by one another and rest on the intervals between them; set wide apart like that, they give the impression of solid strength." Like Antony bestriding the world, they are colossal. It is in this manner, then, that Longinus can praise unifying effects in style without contradicting himself. Apparently neither sustained elevation nor periodicity is to be considered smooth or even fluent. The sublime elements of composition are like the stone slabs of an entryway before the invention of the keystone; up to a point, the strength of the arch is proportionate to the width of its base. Thus even in passages in which a sustained effect is aimed for, the composition should be, as Longinus puts it in the next section, "rough at the joins" (41.3; skleroteta episundedemena).

Longinus more than once resorts to architectural metaphors—rather surprisingly, it has been pointed out, considering that the Greeks did not rank architecture among the fine arts.⁴⁴ Having in general insisted on the need for a "cleft" through which the sublime can pass, Longinus at one point seems to change his mind. In a discussion of gathering details for a sublime effect, he says that good writers "have taken only the very best pieces [i.e., stones], polished them up and fitted them together. They have inserted nothing inflated, undignified or pedantic. Such things ruin the whole effect, because they produce, as it were, gaps and crevices" (10.7). But again there is really no contradiction. An architecture in which wedges of foreign matter (wood, perhaps, or roughcast) leave actual "chinks and crannies" in a wall would be much cruder than the rough architecture of the sublime needs to be or can be.⁴⁵ It would be ludicrous architecture, like the wall with its chink that separates Pyramus and Thisby in A Midsummer Night's Dream. The opening of the sublime must be well-knit to a certain degree in order to resist what will tear through it.⁴⁶

After this review it can be asserted with some confidence that Longinus's fifth "source" of sublimity, which is "dignified and elevated word-arrangement" (8.1; sunthesis), is not out of keeping with the other four, at least three of which stress division and passion rather than composition and composure. But the exact nature of the linguistic "joins" under this fifth heading remains difficult to determine. Much depends, in any theory of form, on whether connecting links are thought to be interstices, like the cushion of air that keeps colliding bodies from touching in the physical world, or joints, like helical or chain structures. Whichever of these models may seem more accurate, there can be no question of a third, of the blending or ooze that is implied in the Goethean concept of "organic form." There is nothing liquid or even porous about those linguistic signs which are not themselves the sublime but remain necessarily its vehicle. A sign cannot merge with its neighbors without losing the syntagmatic features—the differences—that make it intelligible. This observation may be an instance of what the German hermeneuticians condemn as Modernisierung, but it still provides a convenient way of loosening the strong grip that the word sunthesis appears to exert upon the disruptive potential of the sublime. Even the evidence of the Peri Hupsous itself makes it clear, in any case, that sunthesis is not synthesis. Longinus's sunthesis is actually a somewhat limited technical term meaning the order in which words are written down; it is quite closely equivalent, that is, to the modern word syntagmatics.

Although the semantic possibilities of imitative form are not to be overlooked in studying word arrangement (one thinks of the poetic line in Virgil), the primary concern of Longinus in this regard is with rhythm.⁴⁷ This being so, the word composition that is used in most translations of the section on sunthesis is especially misleading. Rhythm may be syncopated but it cannot form an are or return on itself. It is essentially repetitious, "the ticking of a watch made softer," as Yeats put it,⁴⁸ and thus has very little to do with the techne of the classical formalists. Longinus himself tips the balance between variation and repetition clearly toward the latter by devoting a section (20) to the sublime effects of asyndeton combined with anaphora. He stresses the need for variety "to save the sentence from monotony and a stationary effect" (20.2), but much more significantly he recommends that anaphora be used with asyndeton, which is percussive in effect, rather than, e.g., with hyperbaton, which would involve an unexpected deviance from norms of recurrence. At this point, furthermore, he expatiates on the bullying of Demosthenes, "hitting the jury in the mind with blow after blow" (20.3). It may be observed in addition concerning this point that Longinus later contrasts the "wonderful spell of harmony" cast by the "varied sounds of the lyre"—the mild spell of "persuasion," that is, or of the beautiful—with the Phrygian pulsation of the flute, which inspires one mood, though by no means the only one, of the sublime (39.2-3).

In short, repetition is more important than variety if one's purpose is to stun the auditor. Again we come to an aspect of the sublime that we may prefer to wish away. Mesmerism, demagoguery, Madison Avenue, ecstatic dancing, and the "meditation" on a name or a sacred mantra—all undertake to "enslave" the hearer, to bring about the identification of an audience with some transcendently authoritative voice by means of repetition.⁴⁹ But Longinus is able to show that repetition has more acceptable uses. A mature taste in the discrimination of sublimities, for example, comes from experiences that take place "many times over" (7.3). From this standpoint taste can be defined, very plausibly, as a superior force of habit, an informed, at least partly voluntary self-hypnosis. Repetition also provides a way, finally, of revealing what the sublime is and what it is not. Partly owing to the element of sameness in repetition, the sublime is one—and largely indivisible—but no version of the sublime is ever unified, anywhere in the Peri Hupsous, as an increment of structure or of harmony.

VIII

Some difference remains, however, between the repetitious sublime and the sublimity that takes effect with a single blow. Verbal composition is mainly a process of accretion, "building up" (39.3), a process that is equivalent to what Longinus describes, on the figural level of composition, as auxesis, or amplification. There are enough passages assigning repetition rather than singleness to the sublime to warrant the suggestion that Longinus anticipates Kant's distinction between the "dynamic" and the "mathematical" sublime,⁵⁰ which latter heading denotes the effect of being staggered by the sheer volume of something: too much information, too many anxieties, mountains too high, and so on. The addition of this category completes my discussion of the rhetoric, the conditions, and the modes of the Longinian sublime. Using the mathematical sublime as a point of departure, I shall now discuss some of the implications of the sublime, with an eye toward showing why I think it is preferable to the identification of form as an end in view for contemporary criticism.

The mathematical sublime has been analyzed by Thomas Weiskel, and more recently by Neil Hertz, as a frustrating buildup or sedimentation that finally causes the sensation of "blockage," the "checking of vital powers" noted by Kant. Hertz admires Weiskel's thesis but argues that Weiskel identifies the blocking agent in Kant—the Reason (Vernunft), which implicitly rebukes the Imagination (Einbildungskraft) for failing to reduce the overload to proportion—too hastily and arbitrarily with the Father in the "family romance" of Freud.⁵¹ Weiskel cuts his way through the blockage of his own knotty problem, in other words, simply in order to have got through it. Passed on from commentator to commentator, each time with a new qualification, the mathematical sublime renews itself according to the pattern described earlier: possession, resistance, response. Hertz's solution—not to cut through the blockage but to learn to live with it until its familiar presence ceases to frustrate—is designed to put an end to the game, but that may not be possible. It is too likely that one's orientation in the chain of audition along which the sublime is passed is, as Weiskel says, irreducibly oedipal. Any forfeit of that orientation would then simply become a promise that one will not mind being overwhelmed.

There is another topic in Freud, however, that brings one closer to the way it feels to cope with the mathematical sublime from day to day. This topic is the "mastery of repetition" that Freud supposes to be a constructive displacement of the death instinct.⁵² At least one phase of the sublime, the initial phase that all the eighteenth-century theorists agreed in calling awe-struck or frightened, is loaded with intimations of mortality. Freud described the favorite topics of the obsessional neurotic as those "upon which all mankind are uncertain,"⁵³ including the afterlife and the reliability of memory. The neurotic's repeated efforts to break through to conclusions about these topics, efforts which transfer to waking life the unpleasant doing and undoing that everyone performs during a feverish doze, are futile attempts at auxesis. Replete with its "great thoughts," on the other hand, the sublime can become a working through that makes the neurotic compulsion purposeful. The sublime reveals its power of sublimation even in what it does to the signals of repetition itself: a dripping faucet becomes the sound of waves, a metronome softens into meter, and the dullest monotony becomes "The murmurous haunt of flies on summer eves."

The issue of repetition dominates even Longinus's passing remarks about metaphor. A structurally oriented poetics is likely to make metaphor, whether Aristotle's "transference of an alien name" or Jakobson's "poetic function," the microcosm of its largest concerns. Both in the Poetics and in the Rhetoric Aristotle showed how metaphor is the most important trope in the repertory of the artist because the "eye for resemblances" that it requires is what is also needed to connect the beginning with the end of a composition. On the other hand, Longinus, whose poetics is disruptive, evinces no interest in the structural properties of metaphor.⁵⁴ What does interest him is the traditional question among the rhetors concerning the relation between metaphor and repetition:⁵⁵ How often should the trope be repeated in elaborating a single topos? The answer Longinus gives is that there should be no limit as long as the result of any repetition is the pathos, or mesmerism, that erases one's memory of repetition and "never allow[s] the hearer to count the metaphors because he too shares the speaker's enthusiasm" (32.4). It is not metaphors themselves, then, but their repetition that achieves the sense of identity—empathetic rather than semantic identity—that can be managed only by the "transference of alien names" from the standpoint of structural analysis. Because the temporality of tropes and figures is unrelieved in Longinus, the vertical structure even of metaphor collapses into sequence.⁵⁶ He says nothing about the synonymy and identity that are enforced by radical metaphor in formalist thinking.

Repetition in the mathematical sublime can function as a homeopathic cure and thus resembles one widely sanctioned interpretation of catharsis in Aristotle. Like the frenzied music mentioned as a katharsis in the Politics (1341b-42a) and as a form of repetition by Longinus in his section on rhythm, repetition in excess precipitates the hearer toward a state of calm. Such is the outcome, Longinus thinks, of the onslaught of metaphors describing the "bodily tabernacle" in the Timaeus: "Finally," writes Longinus, at once quoting the passage and describing the effect of it on the reader, "when the end is at hand, the soul's 'ship's cables' are 'loosed,' and she herself 'set free."' This extended metaphor seems to invite death, and in that respect it can be compared with both homeopathic medicine (more poison) and homeopathic catharsis (more suffering). But the therapy of repetition does not depend on the suspension of life that is required by the unified structure of catharsis. Rather it comes about in keeping with the ordinary footfall of time, which is metric, rhythmic, and anaphoric. The dynamic sublime, a single thunderbolt, has somewhat different purposes, but in the mathematical sublime there is reassurance in knowing that "Hectors and Sarpedons came forth" and kept on doing so (23.3), and there is likewise a sense of being "renewed" by the Ciceronian discourse that is "repeatedly fed with fresh fuel" (12.5).

IX

I have tried to show in my analysis of the Poetics that the crisis—the "recognition scene"—of a unified structure exposes the function of the double, or projected self, at the crossroads of choice; sexual or otherwise, this choice is always equivalent to the "ambiguity" of the New Critics. The crisis that comes with the sublime, on the other hand, exposes the function of the double in the prefiguration of death; confronting the uncanny, the listener tries to wrest authority from what seems to be a former self. Just as Addison and Burke associate the Beautiful, which is soft, smooth, round, and well-formed with sex, generation, and the plenitude of earthly existence, so they and most other theorists incline to associate the sublime with darkness, solitude, the unknown, and the "checking of vital powers." The sublime seems always to have been viewed as a trial confrontation with death. Whereas the theme of the Beautiful is the destiny of others as it appears manifest in their forms (their shapeliness in life, their roles in drama), the theme of the sublime is the destiny of ourselves, which we confront in the act of trying to win an authentic self from the forms that stand in our way. When the contrast between these old rivals for aesthetic attention is put thus provocatively, awarding all sanity, vital health, and humanity to the Beautiful—which is the flesh of formalism—then anyone who is not merely morbid and still prefers the sublime as an objective for interpretation has a lot of explaining to do. The worst that can be said about the sublime is now said: In bringing the death instinct rather than the pleasure principle close to the surface of aesthetic experience, it is necessary for the sublime to risk the irresponsible and exhibitionistic courtship of danger. To this effect Longinus praises a Homeric passage about mariners in a storm: "He has in effect stamped the special character of the danger on the diction: 'they are being carried away from under death"' (10.7). However, the immediate danger attendant upon the sublime is not death, but ridicule. With the slightest overemphasis the sublime becomes the ridiculous: ham-handed, humorless, and provincial. This risk may be unavoidable unless the sublime is transformed into another, quieter quality, one that I think Longinus himself has anticipated.

The false sublime gets puffed up and loses control. The danger of visionary speech (the dynamic sublime) is Icarean, the risk of a great fall, as when "the writer's soul … shares the danger" with Phaethon (15.4) or when the listener must share "the speaker's peril" during the suspension of meaning in hyperbaton (22.4). The danger of interpretive speech (the mathematical sublime) is Daedalian, the temptation "to go too far" in the proliferation of tropes (32.7), to lose the sublime in confusion rather than achieving it by sustained attention. The mathematical sublime must turn out to have been pregnant and not dropsical (3.3). All its amplifications may be nothing but "puffy and false tumours" (3.4). These dangers, which once more show the close proximity of words and nature, have to do in general with bodily malfunction and monstrosity.⁵⁷ Expressions may be "incomplete and abortive" (14.3), constipated, as when the "bowels" of the too-literary mariners of Aristeas "heave in pain" (10.4), or dwarfed, hardened beneath the surface like chancres, by the pettiness of conception that dwarfs reality: vultures that are tombs for men (3.2), a book that compresses the conquest of Asia (4.2), and the forced synecdoche crossed with homonymity (in the word kore) that traps a maiden in an eyeball (4.4). The failure of the mathematical sublime, then, which is the failure of interpretation, results in monsters, grotesqueries, and misbirths of the study.⁵⁸

Dangers of visionary speech, however, are not so easily dispelled, nor are they quite so clearly deserving as targets for satire. A transitional instance between the two kinds of sublimity, an instance of euphemism, runs a more complex risk: "The goddess struck the Scythians who plundered the temple with a feminine disease" (28.4), by which is meant, presumably, impotence.⁵⁹ We could classify this misfortune as a bodily disorder that is related merely to the collapse of the mathematical sublime, for there are certainly as many Scythians as Hectors and Sarpedons. There is a key difference, however: for a prior offense, a rape, the Scythians are now cast as victims of authority instead of hapless authors. The danger now becomes ethical, in other words, and returns us to the issue of enslavement. Originary power endangers audiences just insofar as it realizes itself. If Icarus falls, no one notices; that, I take it, is what Brueghel meant and Auden emphasized:

           the ploughman may
Have heard the splash, the forsaken cry,
But for him it was not an important failure.

But if Phaethon, having risen against the order of things and having succeeded in driving the sun, should survive, or transmit his blinding success as a legacy, we would then find ourselves at the mercy of a petulant and spoiled youth. "Persuasion" is a democratic, tolerant discourse that leaves the hearer free to choose whether to be persuaded. It may be a novel, a book of self-help, a primer in economics, a discourse on Method; in the marketplace the ploughman listens attentively and casts his vote. "Transport," on the other hand, as we have seen, whether it drives the hearer across the sky or drives him mad, casts him into slavery, robbing him even of his proper self.

It is "natural," as Longinus says, that "'when two things are joined together, the stronger attracts to itself the force of the weaker" (15.11). The sublime seems too readily to belong in Plato's utopia; the integrity of the unenslaved individual counts for very little in either case. We cannot pretend that the concluding remarks of the Peri Hupsous as we have it are not in the main the sort of diatribe against the herd instinct that betrays a longing for tyranny. Longinus offers himself the chance to blame the disappearance of the sublime on the defeat of democracy but he refuses the offer—and is right, by his own lights, to do so. Always excepting Demosthenes, the oratory of "persuasion" that best suits the forum of democracy is not in any sense sublime. It is not certain, in fact, whether even persuasion is dependent on one political climate more than another: "In our age there are minds which are strikingly persuasive and practical, shrewd, versatile, and well endowed with the ability to write agreeably" (44.1). The more insidious enemy of the sublime, materialism, likewise flourishes regardless of most political changes; "most," that is, however, because the possibility of founding an austere tyranny, a reign of philosopherkings or perhaps a theocracy, is the greatest temptation of those who are capable of "great thoughts."⁶⁰ Hence, again, there appears to be a bond of sympathy between tyranny and the sublime.

I think, however, that this sympathy exists in appearance only. In the course of its attack on materialism the argument of Longinus is almost undermined by a reversal of the values attached to certain metaphors. This reversal is inevitable and ultimately robs the austere tyrant of his authority, leading to the democratization of the sublime that I wish to propose. In Longinus's last chapter the tropes associated with materialism gradually begin to imply that materialism is an epidemic so far-flung that it must itself be accounted sublime. The grosser passions fight an "unlimited war," wealth and the lust for it are "measureless and uncontrolled," and we are "slaves" to the love of pleasure (44.6-7). The rout of the sublime being sublime, then, Longinus is almost prepared to acquiesce in things as they are, to let the inundation come in whatever form: "Perhaps people like us are better as subjects than given our freedom. Greed would flood the world in woe, if it were really released and let out of its cage, to prey on its neighbors" (44.10). The Longinian sublime is certainly a tyrant but it is not always, as we might have supposed, a celestial dictator, a sky-god or Platonic Houyhnhnm; it is not surprising, when one considers the intimacy between mind and nature that is revealed elsewhere in the text, that the tyrant is sometimes our own instinctual life. A Yahoo tyrant, at least nominally the oppressed rather than the oppressor, it is demagogic, revolting, universally in charge, so much so that it must be softened and coaxed into epicurean channels, sublimated like our "private parts," which nature "concealed as well as she could": "and as Xenophon says, [she] made the channels of those organs as remote as possible, so as not to spoil the beauty of the creature as a whole" (43.5). The term beauty—kallos—is no doubt a vague one here but it may serve to remind us that if the "creature" in question were sublime, or conducive to the sublime, all its gaps might be visible. It should be clear, in any case, that the sublime extends the lower as well as the upper reaches of the Poetics.

X

The bridge between the sublime, thus revealed in its role as a negative force, and the beautiful, a tender-hearted experience that could not in itself entail the possibility of being critical, is the mathematical sublime, or sublime of interpretation. The most elementary temptation of any antiformalist interpreter is to identify with all the darkest visionaries and to find dark values exclusively, with Melville in The Encantadas, on the only side of the tortoise he or she can see. But the interpretive sublime is on neither side. It exists as an alternative to alternation and reveals the lack of opposition in false dichotomies. The discovery of hidden meanings is no more adequate as an exclusive end of criticism than the summary of overt ones.⁶¹ Derivative myth-criticism, "Freudian interpretation," and most of the merely rancorous modes of demystification oversimplify in this way. At the end of the last chapter I suggested that any criticism necessarily goes astray in some measure because it can never estimate the distance between the architectonic purpose and the instinctual purposiveness of its object—or of its own subjectivity. The "effect" of literature, the sublime or uncanny effect which sets it apart (but not categorically apart) from the relatively featureless discourse of practical exchange, does not appear squarely within either form or the negation of form but in the signs of estrangement, the gaps, between formal design and the form of its attempted negation. Neither of these forms completely occludes or represses the other; each harbors an unsubdued strength that deforms the other in ways that cannot be scientifically reduced any more than psychology can fully discriminate or arbitrate between will and instinct in the mind.

So the sublime is not identical with the unconscious, and its appearance is not properly to be described as sublimation, although that plays its role. Undoubtedly we have grown too familiar with our fictions about the fictions of the unconscious, with all their pathos and economy—so familiar, indeed, that the whole psycho-analytic romance may turn out to have sublimated whatever it is that is still wholly unconscious. The sufficiency of the oedipal explanation is undergoing attack from every direction—humanist, feminist, Marxist, deconstructionist—because too obviously it cuts the Gordian knot of interpretation. As Hertz argues, the oedipal explanation breaks through the mathematical sublime arbitrarily. I am not quite sure; in my view the oedipal explanation seems arbitrary because it over-determines the language of desire, desire which is not for one thing, arguably, but for the sheer sensation of presence. But the oedipal explanation is still, if I may so put it, a valid allegory. Roughly the same objection can be made, followed by a roughly comparable defense, concerning the closely related explanations called "bourgeois ideology," langue, and the "precursor"—concerning any explanation, in short, that seems unequivocally to insist that poets are written, not writers.

Antithetical factors are not interpretive ends in themselves any more than the factors conducive to formal unity were before them. Systematically approached, antithetical factors will soon constitute an antithetical formalism. There is undoubted value in a science of this sort, which will certainly find ways of making whole dimensions of art visible for the first time, but I do not see how this or any other quickly exhausted way of processing literature—for that is what it is—can be heralded as the future task of criticism and recommended specifically as an antidote to interpretation.⁶² It is under the banner of science, and not as a hindrance to science, that the interpretation industry has come to monopolize criticism in the last forty years. Interpretation has supplanted both judgment and taxonomy in scholarship; this triumph has been a recent one, not because interpretation had never been done before but because for the most part until this century it had seemed a simple job that could be done silently. It may be noted that the interpretation industry has always dominated theology, and always will dominate it, because where it really matters interpretation has never seemed easy. Interpretation is inevitable; it is thought itself, and it will seem exhaustible, not when it is supplanted by science, but only when it is itself mistaken for science.

The New Critics used to ask the philological historians how they could confidently classify ideas, terms, and texts without reading them attentively, without a due concern for nuance and irony. It seemed a good and fair question, and the answer came as too much interpretation—to the point of "blockage"—with that sort of concern. But ridicule and self-exhaustion alike have now done their work, and very few more new books with titles like "Ordered Flux" are likely to appear. The new blockage is more likely to consist of general rhetorics, revisionary histories, and narratologies, entitling us to ask, in behalf of interpretation, the same old question: How can the new anatomies proceed with confidence if they do not read attentively, openly, and to some extent unsystematically? The science of the greatest modern anatomist, Northrop Frye, has passed inevitably into eclipse, but rival sciences alone are churlish enough to scorn the grace of his interpretations.

The function of the sublime is to keep interpretation from closure. It persists between forms, as close to one as to another; it should be remarked in this regard that the most sophisticated operative terms of formalism, from Aristotle's "recognition"—duly interpreted—to Skhlovsky's "defamiliarization" and Sigurd Burckhardt's "disturbing element,"⁶³ point clearly toward what I mean by the sublime. A valuable instance of the necessary interaction between forms and the sublime appears in Longinus's reading of Sappho's "ode" to Anactoria. Self-convicted of fragmentation, her tongue "broken" by passion, Sappho "brings everything together," as Longinus says (10.3) with commanding metrical skill. If this were all, the topic of interpretation would still be "ordered flux," but Longinus has merely set interpretation in motion by suggesting an equipoise of motive. One would have to continue—to point out, for example, the consecrating excitement of love-anguish in contrast with an encroaching ennui that is intimated in the very composure and elegance of these sapphics, thus questioning the presupposition we have as readers that form is comfort. One would have to show, always inconclusively, how passion and control infect and finally disfigure each other in the course of their endless give-and-take of values.

The effect of a well-conducted interpretation along these lines, drawn out more and more finely in a dialectic that never quite repeats itself, is the effect of the mathematical sublime. But we still have not faced the imputation of tyranny in our victimage by the victim Sappho. Rumors of tyranny have followed the concept of the sublime ever since they were first prompted by the typical metaphors of Longinus himself. Perhaps it can be shown at least that they have been greatly exaggerated. Tyranny exists without legitimate authority but it still requires that some authority, however specious, be invoked for it. The sublime by contrast is unauthorized, as we have seen, a phantom of possession that is always in retreat from one site to another. Although the sublime never falsifies its nature (quod semper, quod ubique, quod ab omnibus) and thus evades the charge of fickleness leveled by Plato against both poets and democracies, yet at the same time it never regiments response; it always surprises and elevates, as Longinus says. Its tonic effect is not subject to conventionalization and it also lacks the monotonous insistence of obsession. Not at all necessarily suggestive of spectacular gloom but of the happy absence, rather, of inhibition, it is best seen, on reflection, as a grace beyond the reach of art.

In an article called "'A Grace Beyond the Reach of Art,"' which is meant to supplement his book on the sublime, Samuel Holt Monk demonstrates that Pope's phrase is less directly indebted to Longinus than to a flourishing and longstanding concept of "grace" per se.⁶⁴Charis, venustas, gratia, je ne sais quoi, sprezzatura.—roughly synonymous with all these terms, grace was very nearly equivalent in seventeenth-century poetics to sublimity in the century succeeding Boileau's translation of Longinus (1674).⁶⁵ (Hazlitt's gusto may offer itself as a nineteenth-century equivalent, while indeterminacy and free-play are somewhat unsatisfactory candidates for our own.) The very real difference between the two terms was that grace in the aesthetic milieu of the Restoration suggested nothing gothic, ponderous, or frightening. It encompassed the baroque fillip, the shining of a countenance, the unexpected twist of a period, or the simplicity of an epigram. Its effect, says Monk, was "sudden and surprising," like that of the sublime, and in its time it was, like the sublime, "a repository for the irregular and irrational elements in art,"⁶⁶ but apart from these parallels it was a very different quality, a carelessness in elegance (like Pyrrha's hair in Horace) rather than a breach of decorum.

Monk says ("Grace Beyond the Reach of Art," p. 134) that Lysias was the stock example of grace in literature, an example taken by seventeenth-century writers from Longinus's probable contemporary Demetrius (see On Style, 3, 128). When one considers what Longinus says about Lysias, one comes to suspect that he may be extending the terms hupsos and megethos to incorporate the charis discussed at length by Demetrius (127-42) and also in passing by Dionysius of Halicarnassus. It is important not only that Longinus compares Plato with Lysias while pretending not to, showing thereby that grace, whatever Plato may lack, is certainly what Demosthenes lacks. What is more telling is the fact that the concept of grace would rationalize the presence in Longinus's canon of Sappho's ode, a poem that many scholars, most notably Saintsbury, have refused to call sublime.⁶⁷ When Demetrius affirms that "one cannot sufficiently admire [the charm: epicharitos] in the divine Sappho,"⁶⁸ Longinus would certainly applaud the precision of his sentiment. All things considered, he would not demur unduly, I think, if one were to introduce grace into the domain of the sublime.

A modern essay that is splendidly Longinian, provided that one is willing to extend the meaning of hupsos in this manner, is "Language as Gesture" by R. P. Blackmur, whose examples, especially those taken from the plastic arts, have a quality resembling "grace." Blackmur reminds one of Longinus both in his virtuoso, paronomasiac style and also in his way of choosing and then sympathetically fusing his commentary with his examples. He concludes "Language as Gesture" with a tribute to Shakespearean gesture that could be a translation of Longinus: the power of implication in Shakespeare "must overwhelm us even though we realize as we consent to it, that we have made it ourselves."⁶⁹

XI

I suggested [elsewhere] that the tradition deriving from Aristotle has tended to constitute itself as the only tradition. In recommending a different perspective I have tried (a victim, perhaps, of some obscure law of recoil) to recover nearly the whole field for Longinus, including much to which he never explicitly laid claim: a theory of representation, a rigorous understanding of the interplay among nature, feeling, and language, and a grasp of aesthetic qualities beyond the so-called sublime—the Burkean trombone—extending even to amplification, charm, and delicacy. For the moment I have not pressed these conclusions into a theory of interpretation, but in rough terms the orientation of such a theory will be obvious. Even though the theory itself should remain as evasive as the sublime, in practice it amounts to this: whenever possible, the interpreter should delay the semantic closure that is urged upon him by his or her own will to form and that of the text. The enterprise of interpretation is best honored by those who do not agree to go no further, to get no closer, to honor tenuous symmetries. Having arrived at a liminal understanding of form (both as the determination of consciousness and as unconscious determination⁷⁰) by feeling along its edges like someone who is blind, the interpreter begins, at that point of exhaustion, to interpret.

The Sophoclean Oedipus disregarded by Aristotle but nonetheless present in the Poetics realizes that, as Longinus says, his misfortunes are plural. Because of his resolution to become a scapegoat he sacrifices the "external trappings" of his governorship—as helmsman and tyrannos—to the interior form of his family romance, scattering the symbols of his far-seeing guardianship on the ground in order to make the character beneath his role visible for the first time. In Longinus all of this seems a preliminary matter; it is a model for the structure of disruption that tells us very little in itself, "catharsis" notwithstanding, about the way the experience reaches the spectator. A fuller and subtler moment in the Oedipus story is commemorated by Longinus, just before he cites the apparition of Achilles over his tomb, as "Sophocles' account of Oedipus dying and giving himself burial to the accompaniment of a sign from heaven" (15.7).

In the messenger's speech recording this event, there are "clefts" in the sky, in the earth, and in knowledge, all of which regions at this moment, especially as Oedipus views them, are nearly indistinguishable. Although each region is sundered, there is still no point of division between them:

Then very quickly we saw him do reverence
To earth and to the powers of air,
With one address to both.
        But in what manner
Oedipus perished, no one of mortal men
Could tell but Theseus. It was not lightning,
Bearing its fire from God, that took him off;
No hurricane was blowing.
But some attendant from the train of Heaven
Came for him; or else the underworld
Opened in love the unlit door of earth.⁷¹

Longinus recognizes his affinity with a seer who can bring mind and nature so close together that one sudden obeisance will do for both. A sweep as unconfined as this cannot be imagined without some loss of categorical nicety. Theseus, like Horatio and so many others to come, survives to interpret the experience, reconstructing in his own person the monarchical form of Oedipus while passing on the family form to Hippolytus. Like Longinus, Theseus will aspire to the mathematical sublime; he will be concerned with the quality of life in society after the dynamic sublime has been buried. Leaving the sacred ground, he carries the vision of Oedipus into a community that is preserved, in all its necessary forms, by taking that vision to heart in order not to relive it. The power of Oedipus is now explanatory, descriptive, and perhaps lacking in magnificance, but in being recalled by the voice of Theseus it is made to return from its antisocial isolation. To follow and confirm that return is the whole art, or nature, of interpretation.

Notes

¹ For these arguments, see William K. Wimsatt and Cleanth Brooks, Literary Criticism: A Short History, 2 vols. (Chicago: Univ. of Chicago Press, 1978), 1, 97-111.

² Olson, "The Argument of Longinus' On the Sublime," in Critics and Criticism: Ancient and Modern, ed. R. S. Crane (Chicago: Univ. of Chicago Press, 1954), p. 259.

³ See "ibid.," p. 233. Should there be any doubt of Aristotle's influence on Olson's allegedly "pluralistic" interpretation, Olson's assertion that the sublime "is a kind of mean" between vices of style ("ibid.," p. 242) will indicate what I mean.

⁴ For the differences between these views as assessed by the participants, see R. S. Crane, "The Critical Monism of Cleanth Brooks," in Critics and Criticism, ed. Crane, pp. 83-107; Wimsatt, "The Chicago Critics: The Fallacy of Neoclassic Species," in The Verbal Icon: Studies in the Meaning of Poetry (Lexington: Univ. of Kentucky Press, 1967), pp. 41-65; Elder Olson, "The Dialectical Foundations of Critical Pluralism," TQ 9 (1966), 202-30. Under the tutelage of Richard McKeon, the Chicagoans tended to read Aristotle as though the distinction between language and concept, or referent, were the keystone of his system. See especially McKeon, "Aristotle's Conception of Language and the Arts of Language," Critics and Criticism, pp. 173-231.

⁵Literary Criticism, 1, 101.

⁶ Thomas Weiskel points to "the confusion of nature and art, author and work, which will become the trademark of the Longinian or affective sublime" (The Romantic Sublime: Studies in the Structure and Psychology of Transcendence [Baltimore: Johns Hopkins Univ. Press, 1976], p. 12).

⁷"Longinus" on Sublimity, trans. D. A. Russell (Oxford: Clarendon Press, 1965), to be cited by section number and marginal number, 1.3. For translations I have also consulted Longinus on the Sublime, trans. A. 0. Prickard (Oxford: Clarendon Press, 1906), and On the Sublime, trans. W. Rhys Roberts (Cambridge: Cambridge Univ. Press, 1899). 1 have used the Greek edition and commentary of D. A. Russell, "Longinus" on the Sublime (Oxford: Clarendon Press, 1964). Although the identity of the author is unknown, I follow the nearly universal convention of calling him "Longinus." It is interesting to note that a leading Longinus scholar, G. M. A. Grube, has returned to the belief, or at least leans toward it, that our author was indeed the third-century Cassius Longinus celebrated by Boileau and Gibbon (The Greek and Roman Critics [London: Methuen, 1965], p. 341).

⁸ The most extreme instance of debunking based on this error is Walter Allen, Jr., "The Terentianus of the Peri Hupsous," American Journal of Philology 62 (1941), 51-64.

⁹ Neil Hertz describes the "oedipal moment" itself, insofar as Oedipus's recognition can be identified with resolution, as the "sublime of conflict and structure" ("The Notion of Blockage in the Literature of the Sublime," in Psychoanalysis and the Question of the Text, ed. Geoffrey H. Hartman [Baltimore: Johns Hopkins Univ. Press, 1978], p. 76).

¹⁰ The fact that Longinus almost certainly found this remark in Quintilian (Institutio Oratoria, 8, 3, 37) makes his assignment of it to the Peripatetics yet more significant.

¹¹ So Roberts. Russell settles for "external trappings" even though the word is prostragodoumenon.

¹² Grube has also argued (Greek and Roman Critics, p. 344) that Longinus must have in mind the emotions of characters, not audiences. Aristotle himself says, however (Rhetoric 1408a), that pity, grief, and fear are "low emotions" in oratory. On this point see also Allen Tate, "Longinus and the 'New Criticism,"' in The Man of Letters in the Modern World (New York: Meridian, 1955), p. 188.

¹³ I mean only the structural values of the Poetics. In the Nicomachean Ethics (1124a) there is a portrait of the "magnanimous man" which closely anticipates the sublime individual in Longinus.

¹⁴ I do not think that Longinus's distinctions are completely empty. I would agree, for instance, with Russell's excellent summary ("Longinus" on the Sublime, p. 126) of the difference between figures (schemata) and tropoi, which is, to put it negatively, that if a figure fails the result is a solecism whereas if a trope fails the result is a barbarism. Deconstruction is not the only current school of thought, in any case, which professes a disregard for generic and other such distinctions. See the persuasive article by John Bayley, "Against a New Formalism," Crit Q 10 (1968), 60-71.

¹⁵ See, e.g., Russell, "Longinus" on the Sublime, p. 91, and G. M. A. Grube, "Notes on the Peri Hupsous," American Journal of Philology 78 (1957), pp. 365-66.

¹⁶ Olson's "Argument" notwithstanding, I would agree with Neil Hertz that one cannot keep a fixed pattern or structure of the text clearly in mind ("Lecture de Longin," Poetique 4 [1973], p. 292).

¹⁷ Readers familiar with Jacques Lacan's "Seminar on 'The Purloined Letter'" (YES 48 [1972], 38-72) will recognize in this chain of dissimulation and enthrallment, which is so much more complex than the Horation Si vis me flere, the ring of pursuit as Lacan understands it in Poe's story—and in the psychoanalytic transference.

¹⁸ On this point see Tate, "Longinus and the 'New Criticism,"' p. 183.

¹⁹ Hertz's reading of Longinus's response to Sappho differs from mine in stressing compositional qualities; I do not agree with his assertion that "La doctrine de l'unite organique a rarement et presentee avec autant de ferveur" ("Lecture de Longin," p. 295).

²⁰Romantic Sublime, p. 17. Weiskel feels on the whole, however, with Hertz, that Longinus is committed to "organic continuity" (ibid., p. 21). Wimsatt is nearly alone among the commentators in having remarked that Longinus's figures "tend to have to do with abnormalities of syntax and peculiarities of structure" (Literary Criticism, 1, 103).

²¹ Allen Tate is the most extreme proponent of the notion that the sublime is a quality of words ("Longinus and the 'New Criticism,"' p. 177). See also Elizabeth Nitchie, "Longinus and the Theory of Poetic Imitation in the Seventeenth and Eighteenth Centuries," SP 23 (1935), p. 586. Boileau in the "Preface" to his 1674 translation of Longinus insisted persuasively that the sublime is not wholly a question of style (The Continental Model: Selected French Critical Essays of the Seventeenth Century, ed. Scott El ledge and Donald Schier [Ithaca: Cornell Univ. Press, 1970], p. 272).

²² Iser, The Implied Reader (Baltimore: Johns Hopkins Univ. Press, 1974), p. 280.

²³ See Hertz, "Lecture de Longin," p. 292.

²⁴ See Geoffrey Hartman on the impossible ideal of "purity" in language in Criticism in the Wilderness: The Study of Literature Today (New Haven: Yale Univ. Press, 1980), esp. pp. 115-57.

²⁵ As Weiskel says (Romantic Sublime, p. 5), the sublime "is always cloaked in metaphors of aggression."

²⁶ D. A. Russell is especially concerned to deny the similarities between Burke and Longinus. See, e.g., the introduction to his translation, "Longinus" on Sublimity, p. xvi.

²⁷ Hertz ("Lecture de Longin," p. 299) discusses the oedipal situation that appears in this and many more of Longinus's quotations, including the passage on Phaethon's flight cited above.

²⁸ Freud, "The 'Uncanny,"' in On Creativity and the Unconscious, ed. Benjamin Nelson (New York: Harper Torchbooks, 1958), p. 156n. I return to this issue in chapter 5 (p. 175).

²⁹ Allen, "Terentianus of the Peri Hupsous," pp. 52-53.

³⁰ Burke, A Philosophical Enquiry into the Origin of our Ideas of the Sublime and the Beautiful, ed. James T. Boulton (London: Routledge & Kegan Paul, 1958), p. 66.

³¹ Burke, Reflections on the Revolution in France (London: J. M. Dent, 1960), p. 82.

³² Kant, Critique of Judgment, trans. J. H. Bernard (New York: Hafner, 1972), p. 83.

³³ Perhaps the only commentator who approaches this distinction carefully is George Saintsbury in his eccentric but interesting essay on Longinus in A History of Criticism and Literary Taste, 3 vols. (Edinburgh: William Blackwood and Sons, 1900), I, 162.

³⁴ Russell cautions us that grammatically the silence of Ajax is only an analogy and not an example ("Longinus" on Sublimity, p. 9n.), but I am not sure it matters which way we take it.

³⁵ On the elaborateness of the frame composition in Homer, which encodes information in specular pairs of the kind I have stressed in this passage, see Cedric Whitman, Homer and the Heroic Tradition (Cambridge, Mass.: Harvard Univ. Press, 1958), esp. p. 294-84.

³⁶ A. C. Bradley describes nocturnal silence as "a peace … that may make the face of death sublime" ("The Sublime," in Oxford Lectures on Poetry [London: Macmillan & Co., 1955], p. 49).

³⁷ Eliot, "Tradition and the Individual Talent," in The Sacred Wood (London: Methuen, 1964), p. 52.

³⁸ Not all hearers do so, clearly, but the response which does not in itself become an "influence" becomes, however idiosyncratic and interesting, a terminal mutation. Every significant reader is in some sense also a writer. See the text of Longinus at 7.4.

³⁹ Harold Bloom's theory of influence presides over these next few pages in roughly the form of development, lacking the tropes and defenses, that appeared in The Anxiety of Influence (London: Oxford Univ. Press, 1975). Also in this context Weiskel writes admirably as follows (Romantic Sublime, p. 32): "To consider the problem of originality is to find the two kinds of sublimation, poet's and reader's, compounded or superimposed."

⁴⁰ I refer especially to the second of three newspaper essays that are usually entitled "On Genial Criticism": see "On the Principles of Sound Criticism: Essay Second," in Miscellanies Aesthetic and Literary, ed. T. Ashe (London: George Bell, 1885), pp. 10-14.

⁴¹ There are several versions of this anecdote. All of them are conveniently reviewed by C. D. Thorpe, "Coleridge on the Sublime," in Wordsworth and Coleridge, ed. E. L. Griggs (Princeton: Princeton Univ. Press, 1939), pp. 193-94. See also Bradley, "The Sublime," p. 37.

⁴² Quintilian's well-known anticipation of this passage (Institutio Oratoria, 8, 2, 21) is a sarcastic joke at the expense of those who pride themselves on deciphering obscure passages. The relevance of this joke for anyone listening, say, to the priestess at Delphi, should be clear.

⁴³ As Grube writes (Greek and Roman Critics, p. 347), "Clearly Longinus uses mimesis in the broadest, not the restricted rhetorical sense."

⁴⁴ See Prickard, Longinus on the Sublime, p. xvi.

⁴⁵ On the complexities of this passage, see Russell, "Longinus" on the Sublime, p. 106. Some scholars have suggested emending psugmata (gaps) to psegmata (dust, chippings).

⁴⁶ Hertz argues ("Notion of Blockage," p. 70) that a theoretical concern with "blockage" arises only after the decline of Longinus's influence. I think that this concern is already present in Longinus and can be found in certain phrasings of Addison (especially Spectator 412) more apparently than in any author in the decades just preceding Kant.

⁴⁷ Wimsatt (Literary Criticism, 1, 109) goes so far as to equate sunthesis with rhythm.

⁴⁸ Yeats, "The Symbolism of Poetry," in Essays and Introductions (New York: MacMillan, 1961), p. 159.

⁴⁹ As I understand it, the meaning of Kierkegaard's allegory, Repetition: An Essay in Experimental Psychology, is that repetition, for the reasons here outlined, is a means of finding grace. It is, in any case, as Valery explained, an enemy of our rational wish to process art as information (see "The Idea of Art," in Aesthetics, ed. Harold Osborne [Oxford: Oxford Univ. Press, 1978], p. 28).

⁵⁰ Kant, Critique of Judgment, pp. 86-106. The mathematical sublime is what Thomas Weiskel calls the "hermeneutical sublime," anticipating what I have said earlier in the present essay about the sublime in interpretation (see Romantic Sublime, esp. p. 28).

⁵¹ Hertz, "Notion of Blockage," p. 74. See also p. 76 and Hertz, "Lecture de Longin," p. 304. It seems to me that in order to project an adequate dynamic into the structure of the Kantian sublime, Weiskel has had to draw imagination and reason dangerously close to their English meanings, so that imagination swells into the glorious faculty of Wordsworth, significantly an "unfathered vapour from the abyss," and reason shrinks from the radiant proportions Kant awards it to the old patriarchal bogey that the Romantics attributed to the Augustans.

⁵² Freud, Beyond the Pleasure Principle, trans. James Strachey (New York: Norton, 1961), pp. 8-13, 30.

⁵³ Freud, "Notes upon a Case of Obsessional Neurosis" ("The Rat Man"), in Three Case Histories, ed. Philip Rieff (New York: Collier Books, 1976), p. 88.

⁵⁴ It has been speculated that there is a lost section on metaphor; that may be, but I would prefer, of course, to think not. See T. R. Henn, Longinus and English Criticism (Cambridge: Cambridge Univ. Press, 1934), p. 65.

⁵⁵ Jakobson would see metaphor as a basis for repetition in that the "poetic function" is an imposition of equivalence on signs ("Linguistics and Poetics," in The Structuralists from Marx to Levi-Strauss, ed. Richard DeGeorge and Fernande DeGeorge [Garden City, N.Y.: Doubleday Anchor, 1972], p. 95).

⁵⁶ The definitive discussion of the absence of simultaneity from the functioning of tropes is that of Paul de Man, "The Rhetoric of Temporality," in Interpretation, ed. Charles Singleton (Baltimore: Johns Hopkins Univ. Press, 1964), pp. 173-209.

⁵⁷ What all the faults have in common, says Grube, is "swellings" ("Notes on the Peri Hupsous," p. 364).

⁵⁸ The use of the topoi of malformation in satire has been treated theoretically by Michael Seidel, Satiric Inheritance, Rabelais to Sterne (Princeton: Princeton Univ. Press, 1979), esp. pp. 3-59.

⁵⁹ I would not have guessed it, frankly, but see the conjecture of Russell, "Longinus" on the Sublime, p. 148.

⁶⁰ For a modern discussion of the politics of sublimity, see Iris Murdoch, "The Sublime and the Beautiful Revisited," YR 49 (1959-60), 247-71. For the fullest discussion of Longinus's closing remarks, see Charles P. Segal, "Hupsos and the Problem of Cultural Decline," Harvard Studies in Classical Philology 64 (1959), 121 ff.

⁶¹ Errors of this sort are well discussed by Martin Price, "Form and Discontent," NLH 4 (1972-73), 383.

⁶² See Jonathan Culler, "Beyond Interpretation: The Prospects of Contemporary Criticism," CL 28 (1976), 244-56.

⁶³ See Skhlovsky, "Art and Technique," in Russian Formalist Criticism: Four Essays, ed. Lee T. Lemon and Marion J. Reis (Lincoln: Univ. of Nebraska Press, 1965), pp. 3-57, and Burckhardt, Shakespearean Meanings (Princeton: Princeton Univ. Press, 1968), 285-313. Weiskel (Romantic Sublime, p. 19) speaks well in this context of Wordsworth's "great program of defamiliarization."

⁶⁴ Monk, "' A Grac e Beyon d the Reac h of Art," 'JHI 5 (1944), 131-50.

⁶⁵ Or somewhat more than a century. Although Burke ignored Longinus and Hugh Blair abused him, his influence is still reflected, according to Monk, as late as 1787 (The Sublime: A Study of Critical Theories in XVIII-Century England [Ann Arbor: Univ. of Michigan Press, 1960], p. 25).

⁶⁶ Monk, "A Grace Beyond the Reach of Art," pp. 132, 150.

⁶⁷ Saintsbury, History of Criticism, I, 154.

⁶⁸Demetrius on Style, ed. W. Rhys Roberts (Cambridge: Cambridge Univ. Press, 1902), p. 131.

⁶⁹ Blackmur, "Language as Gesture," in Language as Gesture: Essays in Poetry (Westport, Ct.: Greenwood Press, 1974), p. 24.

⁷⁰ I do not know of a fully developed argument that anticipates what I have been suggesting about the two "forms" and their relations; M. H. Abrams has written interestingly, however, in discussing John Keble, of a "conflict of motives" between composition and repression (Mirror and the Lamp [New York: Norton, 1958], p. 146).

⁷¹Oedipus at Colonus, trans. Robert Fitzgerald, Sophocles I, ed. David Grene (Chicago: Univ. of Chicago Press, 1970), p. 150.