SOURCE: A review of The Poetics of Indeterminacy, in Criticism: A Quarterly for Literature and the Arts, Vol. XXIV, No. 1, Winter, 1982, pp. 64–70.

[In the following review, Steiner criticizes what she considers to be specious arguments and inaccurate semiotic analysis in The Poetics of Indeterminacy.]

It is hard not to admire the courage of Marjorie Perloff's work. She sets out to do nothing less than recast the modernist canon, writing with evident pleasure of poets disdained for their incoherence and exclusively cerebral appeal. Drawn forth from obscurity and isolation, these sports of art become themselves a fecund species, a line fully as productive as the Romanticist-Symbolist dynasty to which they are contrasted. Moreover, once identified, this “Other Tradition” begins to encroach on its High Modernist opposite, claiming Pound, Williams, and Beckett as its own, the heirs or coevals of the likes of Rimbaud, Stein, Apollinaire, Ashbery, Cage, and Antin. Anyone compelled by such writers will feel grateful for Perloff's book, which promises to release us from the need to apologize to the guardians of the Great Tradition for our preoccupation with Minimal Minors.

And yet, though one might applaud Perloff's intent and eagerly await the change in literary values her book calls for, The Poetics of Indeterminacy is not the vehicle to effect this change. Its argument is problematic in almost every respect, threatening a critical indeterminacy that its author does not anticipate. Still, even this indeterminacy is interesting, and in fact instructive about the nature of criticism itself. And so, with sincere respect for Marjorie Perloff's learning, sophistication, and independence of mind, I would like to take issue with her book.

What is an indeterminate text? The answer that Perloff provides varies as the book proceeds. No reference to Heisenberg appears, and Derrida is consigned to a single footnote (of which, more later). Instead, we have Todorov's notion of “undecidability” defined initially on a pragmatic basis: because of the violation of the normal (!) relation between signifiant and signifié found in Symbolism, there are no controls on the associations that arise with texts of the Other Tradition. Thus, “it becomes impossible to decide which of these associations are relevant and which are not. This is the ‘undecidability’ of the text” (pp. 17–18). To illustrate, Perloff juxtaposes texts of one tradition to those of another, finding the High Modernist ones determinate and the undecidable ones indeterminate.

The manifest danger of such a technique, the critic's admission that she is at an interpretive loss, is not just that it is embarrassing, not just that the perverse reader will inevitably find such texts perfectly clear, but that the very status of criticism would require redefinition if texts were allowed to remain opaque, to have this “ambiguity of literalism.” The more normal critical response is not to speak of that whereof one can make no sense. For example, discussing an Ashbery poem similar to one treated by Perloff, Richard Howard quips: “I should say that was beyond critical dispute, or should be, simply because it is largely inaccessible to critical procedure. Fortunately (for my enterprise) not all of Ashbery's work … resists analysis or even interest so successfully.”¹ But Perloff not only wants to allow for opaque poems; she wants to talk about their opacity. And the effect of such talk is to make either the text become intelligible or the critic appear obtuse.

In the first case, time after time Perloff's analyses of indeterminate works culminate in summaries that seem perfectly determinate:

[Stein's “Edith Sitwell”] explores the nature of concord and discord, sameness and difference between two friends.

[p. 80]

[And though] poem after poem in [Williams's] Spring and All is characterized by … Cubist mobility and indeterminacy [p. 129]. Spring and All enacts the difficult process whereby this “hell” is “lit” by flashes of the “dark woman,” the Kora who is waiting to be discovered. … Out of the “messy” and unwieldy prose, out of the disorder of language, the bland crowds and “patches of standing water,” “dazed spring approaches.”

[p. 137]

The critical act is unfortunately one of patching and mending, of reconciling words to systems of value, and thus whenever Perloff lets her guard slip she collapses the opposition between determinate and indeterminate art that she is out to establish.

When she keeps her guard up, on the other hand, she is just as likely to strike the reader as obtuse, willfully blind to a pattern of meaning presented. For me, the most glaring example of this blindness is the treatment of Stein's “Melanctha,” one of the most relentlessly plotted and coherently characterized stories in all of literature. Faced with Stein's picture of the self-defeating, contradictory heroine and her contagious effect on her lover, Perloff is stymied: “Melanctha is submissive but wild, graceful but self-destructive, soothing but always getting into trouble, intelligent but never able to get what she wants. A similar indeterminacy is found in the characterization of [her lover] Jeff Campbell” (p. 93).

The idea that realist character is normally without inconsistency is just one of many simplistic tests for indeterminacy applied to literary works. One even suspects at times that Perloff might be talking down to us, enlisting the aid of what she takes to be the naive realist in order to establish the identity of indeterminate art. For example, she assumes that a text that does not directly illustrate its title is indeterminate: “Rimbaud evokes ‘cities’ [in “Villes”] that are, from the start, impossible to locate in ‘real’ space. For although the poem unfolds a metonymic network of urban images … these references to a possible city are consistently canceled out by images of wild nature” (p. 50). The logical response to such a state of affairs would be to decide that the poem was not about cities in “real” space, but about something else, and that the title was to be taken figuratively. But the first breach between text and title is enough to establish its indeterminacy for Perloff, regardless of the potential significance in such a breach. The corollary of this attitude is Perloff's irritating habit of comparing poems on the basis of their titles. Thus, we find “city poems” or “lake poems” or “box poems” ranged against each other like the control group and the subject of an experiment, as if to say, “When Eliot writes a poem about a city, he writes a poem about a city—but Rimbaud—now there's an indeterminate writer.”

The naïveté of Perloff's stance is sometimes extremely jarring. She calls our attention to a line from Beckett's How It Is: “only one thing to do go back or at least only another thrash round where I lie.” Then she comments: “Supplying the missing [syntactic] links is … not the reader's main problem; the real puzzle is semantic. Why is the ‘only thing to do’ to ‘go back’? Why is the inertia of ‘thrash[ing] round where I lie’ ‘at least’ the ‘only other’ thing to do? There is no way of deciding” (p. 232). There is none, indeed, if one fails to consider existentialist philosophy, absurdist literature, and most of the mainstream of twentieth-century culture.

This contextual innocence on Perloff's part is all the more surprising in light of the sophisticated critical concepts that she marshals on behalf of her point of view. The problem here unfortunately is that she often invokes them incorrectly. One repeated error is the confusion of the reference to specific existent objects with reference in general—in semiotic terms, denotation versus designation. For example, “Unlike, say, Gertrude Stein or, for that matter, Rimbaud, [Pound] does not call into question the relationship of signifier to signified. We can readily identify the fresco ‘at Capoquadri … over the doorway,’ Francis Thompson's then modern poem, ‘The Hound of Heaven,’ or T. E. Lawrence's photographs of ‘rock temples in Arabia Petra.’ … But these illusionistic, literal images are consistently ‘interfering’ with one another, so as to remind us that the world of the poem is not, after all, the real world” (p. 196). The relation of signifier to signified does not remain intact just because we can identify the works of art mentioned; that is, denotation does not imply straightforward designation. Moreover, when has one ever assumed that the world of a poem was the “real world”? Even when an artist musters up every gesture and convention of realism, the relation of work to world is always problematic. This careless use of semiotic terms renders Perloff's position tenuous at best.

So much for the pragmatic definition of textual indeterminacy: the presence in a work of conflicting, undecidable interpretations. The possibility that this is more a condition of the reader than the text does not occur to Perloff, but I find it hard to avoid. We are constantly gaining insight into texts that previously seemed indeterminate or incomprehensible, by growing as readers. Moreover, a text's intelligibility and determinacy are also a function of critical and aesthetic history. In the early years of this century, Mallarmé and Eliot were hardly the determinate retrogrades that Perloff paints, nor will her Other Tradition be able to maintain its otherness under the onslaught of critical interpretation—an onslaught, I should add, of which The Poetics of Indeterminacy is a part. Yet Perloff clearly holds that indeterminacy is a property of texts, not readers, and of post-Rimbaldian and only post-Rimbaldian texts at that. “I am aware that here I take issue with Derridean theory. ‘Indeterminacy,’ as I use that term in this book, is taken to be the quality of particular art works in a particular period of history rather than as the central characteristic of all texts at all times” (p. 17, note). The idea that indeterminacy is a property at all seems contradictory, given the pragmatic nature of the term's initial use. Moreover, it is a pity that this interesting dispute with Derrida should run its course in a footnote.

Is it really possible to declare without a blush that The Waste Land is determinate whereas the “poetry of Rimbaud and his heirs” defies determinate interpretation? The idea that allows Perloff to do so is the sloppy notion of aesthetic semiosis that was touched on earlier: in the mainstream poem from Romanticism to Symbolism to High Modernism, however difficult the meaning may be to decode, “the relationship of the word to its referents, of signifier to signified, remains essentially intact,” whereas this relationship is undermined in the Other Tradition (pp. 17–18). In the Symbolist line, the way to meaning is difficult but possible; in indeterminate art there is no meaning but the surface (pp. 27–28). In the first, words have “specific connotations”; in the second they have, rather, a “compositional value” (p. 23). In the first, metaphor is the predominant semantic mode: “Mr. Eugenides is related, along the axis of metaphor or substitution, to all the other sinister charlatans in the poem, just as every other woman in The Waste Land is a version of ‘Belladonna, the Lady of the Rocks / the Lady of Situations’” (p. 16). In indeterminate poetry, however, metonymy prevails: “in Kora, drinking tea may be either good or bad depending on what has just happened or is about to happen. When the poet finds himself at nightfall alone at the inn without the desired woman, he naturally concludes: ‘what poor tea it was.’ The axis of contiguity thus replaces the axis of substitution” (p. 119). And, we might add, Pop semiotics thus replaces a thorough examination of the subject. Perhaps this imprecise semiotics is the single most disheartening feature of the book: that theoretical concepts as precise and powerful as these can be so imprecisely and inertly used, that conclusions as suggestive and accurate, I think, as those Perloff intuits could be justified by such fallacious reasoning.

To be specific, what does it mean to say that the signifier-signified relation remains intact? Sometimes Perloff means that terms denote rather than merely designate, as we saw earlier; sometimes that they appear in grammatical sentences; sometimes that they are metaphoric rather than metonymic; sometimes that they are not excessively repeated; sometimes that they are concrete or at least do not appear in indefinite, long sentences. Clearly the terms “signifier” and “signified” are themselves indeterminate here, and the dismissal of fundamental issues such as Derridean slippage begins to look either sinister or unforgivably careless. No matter how troublesome Derrida and the other theoreticians of semiotics, structuralism, and post-structuralism may be, no matter how quickly one wants to get to the poetry at hand, the category of indeterminacy and the placement of certain writers within it will remain utterly meaningless if the issues are not laid out consistently and logically.

Just to give an idea of how a little Jakobson can be a dangerous thing, we might pursue Perloff's use of “metonymy.” She claims that Stein rejected realism, producing in Tender Buttons purely metonymic texts. Aside from the fact that she is wrong about Stein's understanding of realism, Jakobson specifically identifies metonymy as a technique of realism.² Moreover, Jakobson elsewhere³ reveals how slippery the term “realism” is, establishing meanings for it that would include any work to be found in the Other Tradition. Indeed most of these indeterminate writers would probably justify their unorthodoxy in the name of realism.

The comparison of indeterminate poetry to Cubist art has the same amateurish quality. Referring continually to one idea by Gombrich—that illusion is suspended in the presence of two conflicting interpretive possibilities—Perloff feels at liberty to pursue the analogy to the visual arts anywhere the inspiration of the moment leads. Thus, one ends up with completely famous comparisons: “Just as the ‘Cubist’ painter recognizes that, in Apollinaire's words, ‘You may paint with whatever material you please, with pipes, postage stamps, postcards or playing cards …,’ so the verbal artist like Gertrude Stein takes words and unlinks them ‘from their former relationships in the sentence’” (pp. 114–15). What that “so” means here is a real mystery to me. But more importantly, the idea lifted out of Gombrich's extensive and complicated writings is terribly misleading. Though it is true that in Cubism conflicting interpretations interfere with illusionism, this conflict is not the only reason for the defeat of illusionism in Cubism. Moreover, it does not necessarily eliminate illusionism outside of Cubism. One thinks of Dali's or Archimbaldo's double images, which are miracles of illusionism. Visual indeterminacy is thus flattened into a single trait, and then it is shifted bodily over to literature. “Image,” however, does not mean the same thing with reference to the two arts; the presence of two—even of two conflicting—interpretations does not necessarily interfere with the “illusionism” of a literary work (whatever that is). As with the semiotic terms, the art-historical component of Perloff's argument is very weak.⁴

But even acknowledging all this, one feels that in some sense Perloff is right. There are some texts produced during the past hundred years that take particularly troublesome liberties with language. The critic faced with this fact might be led to ask why writers have taken to writing this way, in other words, what the cultural function and value of such art might be. It is just here, however, that Perloff is most irritatingly silent. She insists that some works are simply not open to the kind of interpretive action that critics are so prone to undertake, and then she stops. “The meaning of ‘A Substance in a Cushion,’ like that of the title Tender Buttons, remains latent, impossible to translate into something else [shades of ‘The Heresy of Paraphrase’?—no, that indeterminacy belongs, according to its propounder, to all art]. And indeed the important thing is not to establish a fixed meaning for any one item here …, but to see how carefully Gertrude Stein has structured the whole sequence” (p. 107). We know she was careful, apparently, because it is not easy to create such verbal indeterminacy, and a more careless hand (that of Edith Sitwell is adduced as an example) would have slipped into mere ambiguity. But when Perloff takes the time to describe this structuring, she resorts to the most primitive—and often incorrect—formal analysis. (The treatment of accent on p. 317 is inconsistent with any linguistic or poetic theory that I know; the discovery of consonance on p. 127 equates the /z/ in “ladies” with the /s/ in “socks.”) And we are finally left to wonder why Perloff values indeterminacy at all. I know why I value it, and you no doubt have your reasons, but Perloff's case is minimal: such art reacts against an outdated tradition (a question-begging justification), it is interesting (how? why?), it is hard to produce (isn't all art?), and so on. It is not that one would want her to give in and dig out The Meaning of the Poem and weigh it in The Scales of Contemporary Values. But one is left with a feeling of blank mystery and dogmatic prohibition: “don't try to act like a critic and come to a determinate reading of this text; that would be pure conservative wilfulness, an imposition on what is to be valued as a stream or a concrete shard or any other numinous but indeterminate object.” No serious (or humorous) critic can be satisfied with such a demand for passive assent.

Thus, The Poetics of Indeterminacy seems to be only a first stage in the critical reception of the difficult poets it treats. Still, as such, it is extremely useful in delineating the issues in that reception: the model of the visual arts, the concepts of modernism and postmodernism, the influence of romanticism, the importance of continental, especially French, influences, and the need for (an accurate) semiotics for understanding the complex semantics of such art. Perloff's wonderful sensitivity to French nuance, her command of modern poetry and its criticism, and her style and authority cannot but evoke our admiration. At the same time, they should not blind us to the attitude toward intellectualism implicit in the faults of this book. In an earlier study of Frank O'Hara Perloff commented: “Throughout this book I have tried to keep in mind O'Hara's own strictures on literary criticism, so charmingly put forward in the little poem, ‘The Critic.’ … I hope that if O'Hara were alive today, he would not consider me ‘the assassin of [his] orchards.’ I have tried, on the contrary, to respect his wish: ‘Do not / frighten me more than you / have to! I must live forever.’”⁵ It may very well be, however, that the orchard cannot bloom unless it is first assassinated, nor immortality come to a poet whom criticism has not affrighted. This fearsomeness is not merely a critic's power to say yes or no, but the unleashing on a text of the full force of his or her knowledge and self-awareness. Anything less cheats the text, protects it where it should need no protection, and thus inevitably enfeebles it. To understand the difficult texts that Perloff considers we need a poetics of indeterminacy and not a religion of it.

Notes

1. Richard Howard, Alone with America, enlarged ed. (New York: Atheneum, 1980), p. 53.

2. Roman Jakobson and Morris Halle, Fundamentals of Language (The Hague: Mouton, 1971), p. 90–96.

3. Roman Jakobson, “On Realism,” in Readings in Russian Poetics: Formalist and Structuralist Views, ed. Ladislav Matejka and Kristyna Pomorska (Cambridge, Mass.: M.I.T. Press, 1971).

4. I should mention, too, some problems of fact and concept. Perloff claims that Rimbaud is the great source of the indeterminate tradition. Yet Williams frequently denies any French influence in his literature and Stein was directly influenced by the determinate (in Perloff's view) Flaubert. Further, Perloff offers the following potential models for Stein's portraiture: Edgar Lee Masters's Spoon River Anthology (1915), “Yeats's mythologizing portraits of Maud Gonne in The Green Helmet (1910), Pound's Browningesque ‘Portrait d’une Femme’ (1912), or Eliot's ‘Prufrock’ and ‘Portrait of a Lady’” (both 1910–1911). She then goes on to say that these were not models that Stein used. But not only were they not—they could not have been. Stein began her portraiture in 1908, and had been developing the theory behind it ever since her days as a psychology student before the turn of the century. Influence is such an ambiguous concept that it hardly seems worth pressing this point, but this chronological imprecision is troublesome. Similarly, on p. 111, Perloff writes that “Picasso's painting was considered to be the meeting-ground of these different schools, ranging as it does from the neo-Romanticism of the Blue Period to the severities of Analytic Cubism to Surrealist fantasy. What all these painters [mentioned in Apollinaire's Les Peintres cubistes] had in common—and this is Apollinaire's point about ‘l’esprit nouveau’—was a rejection of an art that is primarily representational.” But Apollinaire's view was enunciated in 1911 when neither Picasso nor anyone else had ventured into “Surrealist fantasy.” This imprecision continues in Perloff's failure to distinguish analytic from synthetic Cubism, her treating the two as consistent or identical. And her claim that Williams's symbolism in Paterson was a mode that he was uncomfortable with, his earlier indeterminacy being instead his native element, is belied by his vehemently symbolist history, In the American Grain, published as early as 1925.

5. Marjorie Perloff, Frank O'Hara: Poet Among Painters (New York: George Braziller, 1977), pp. xiii–xiv.