Melville's chaotic style and the use of generative models: an essay in method - Herman Melville

"Ourselves are Fate." So ends chapter 75, "'Sink, Burn, and Destroy' - Printed Admiralty orders in time of war," of Melville's White-Jacket (321). This strangely constructed sentence does not seem to have troubled any of the editors, typesetters, or others involved in preparing the original editions and copies on which the Northwestern-Newberry edition of Melville's works is based, nor does it seem possible for a hyphen to have been dropped. On this sentence, moreover, the Northwestern-Newberry editorial appendices are silent.

Readers of Melville know, of course, that silences are often dizzyingly vocal, even multivoiced. And so this one. The strangeness of the construction may become more apparent if we replace plural with singular forms: "Myself is fate [I am fate?]. Herself is fate. Yourself is [you are?] fate." Melville was far from being a careful proofreader of his own work; perhaps he simply meant to write "Our selves [our personalities] create our fate" - a fairly straightforward rendition of the "character is destiny" motif. On the other hand, the original might be taken as an elliptical construction: "Our selves are the result of the operation of fate," or more succinctly, "Our selves are fated," or with a twist, "We ourselves are fated." This reading fits the context established by the preceding paragraph but not the one in which "Ourselves are Fate" occurs, which insists that "in our own hearts we fashion our own gods" (320). Nor does the immediately preceding sentence provide much help: "In two senses, we are precisely what we worship" (321) - the "two senses" are left unspecified and cannot be inferred from the immediate context.

"Ourselves are Fate" is a small example of what I shall refer to as Melville's chaotic style: strings of words that are punctuated as sentences but are flawed either semantically or syntactically or both, that command no single parsing, and that occupy positions of rhetorical emphasis. This style, which can be found in other significant passages but has not been considered by Melville specialists, is my main theme, and it can be usefully discussed with the help of the generative model of deterministic chaos. I will describe two ways in which to apply this model, one mathematical, the working out of which is beyond my scope here, and the other metaphorical, using several features of deterministic chaos to generate new observations about Melville's problem sentences. Both applications follow accepted procedures for using models to understand a problematic domain (in this case, Melville's style); hence they differ from the way in which scholars in the humanities and human sciences have used chaos theory in the past few years. It is because of this difference, I will argue, that the promise of chaos theory as both an analytical tool and a generative model has not yet been fulfilled.

THE CHAOTIC STYLE: EXAMPLES

Punctuated sentences that are both flawed and indeterminate appear in passages frequently cited in discussions of Melville. (By "punctuated sentence" I mean any string of words that begins with an upper-case letter and ends with a mark of end punctuation, whether or not that string forms a grammatically complete sentence. Hereafter when I use "sentence" alone I mean "punctuated sentence.") The three final paragraphs of Moby-Dick's infamous chapter 42, "The Whiteness of the Whale," for instance, contain a number of such sentences (194-95). For the sake of readability, I have numbered sentences and commented on each individually.

1. "Though neither [the colt nor Ishmael] knows where lie the nameless things of which the mystic sign gives forth such hints; yet with me, as with the colt, somewhere those things must exist." The use of "with" implies both "for" and "within" - a straightforward case of double meaning, which, however, may cause a reader to strain just a bit.

2. "Though in many of its aspects this visible world seems formed in love, the invisible spheres were formed in fright." The grammatical parallelism emphasizes a contrast: one visible world apparently formed one way but an unspecified number of invisible spheres definitely formed another. As with the previous sentence, a repeated preposition is used in an odd way: what does it mean to be "formed in" either love or flight? And whose fright?

3. "Is it that by its [whiteness's] indefiniteness it shadows forth the heartless voids and immensities of the universe, and thus stabs us from behind with the thought of annihilation, when beholding the white depths of the milky way?" The secondary senses of metaphors here seem not just mutually contradictory but jumbled: "indefiniteness shadowing forth voids" implies something physically intangible yet visible that is external to the perceiver, but "stabbing us from behind" suggests something tangible and hidden from view, and "stabbing with thought" suggests something intangible. How do all these stabbings happen when we view the Milky Way, which is visible "out there" rather than hidden behind us? In fact Melville is probably attempting to have Ishmael render the felt sense of an inexplicable psychological phenomenon, a point I will discuss later.

4. "Or is it, that as in essence whiteness is not so much a color as the visible absence of color, and at the same time the concrete of all colors; is it for these reasons that there is such a dumb blankness, full of meaning, in a wide landscape of snows - a colorless, all-color of atheism from which we shrink?" The sense of this sentence is strained by the paradoxical statement that whiteness is both an absence and something concrete and by the assertion that a void can be meaningful. What actually causes the breakdown of sense is that the sentence's question-and-answer structure implies that what it asserts (there is a blankness, this is the all-color of atheism, and we do abhor it) arises from the opening question about the essence of whiteness, but the two categorical opposites (absence and presence, blank and meaningful) cannot be resolved into a meaningful and unambiguous proposition or set of propositions.

While critics who have used these sentences to argue about the meaning of Moby-Dick agree on the most general level about the gist of the sentences, they offer widely divergent views about what specific propositions the passage may be advancing. Nor has anyone considered what effect the passage is likely to have on readers who attempt to analyze its structure. As with the sentence "Ourselves are Fate," these sentences, as well as the ones considered below, apparently did not pose a problem to the editors of the Northwestern-Newberry edition.

Like "The Whiteness of the Whale," chapters 14, 33, and 44 of The Confidence-Man are frequently cited by Melville specialists and usually interpreted as Melville's commentary on the problematic relationship between fiction and experience and the expectations of readers regarding that relationship. Elizabeth S. Foster describes the peculiar quality of the language in The Confidence-Man as "double-writing" (xix) and R. W. B. Lewis as "self-erasing" (272). Lewis's label provides a good starting point because it describes how the sentences move forward grammatically but seem to nullify what has gone before, thus carrying with them no unambiguous propositional content. The chapter headings themselves exhibit a variation on this characteristic:

Chapter 14: Worth the consideration of those to whom it may prove worth considering (69)

Chapter 33: Which may pass for whatever it may prove to be worth (182)

Chapter 44: In which the last three words of the last chapter are made the text of discourse, which will be sure of receiving more or less attention from those readers who do not skip it (238)

In these headings, self-referentiality combines with an ironic double-voiced discourse (created by the addresses to "those to whom" and "those readers") to call into question any assertion that takes the chapters themselves as straightforward expressions of Melville's views on fiction. An even clearer warning comes near the end of chapter 44: "In the endeavor to show, if possible, the impropriety of the phrase, Quite an original, as applied by the barber's friends, we have, at unawares, been led into a dissertation bordering upon the prosy, perhaps upon the smoky" (239). References to smoke carry metaphoric weight in this novel, linked as they are to devilish behavior of an ambiguously benevolent sort or to benevolent behavior of an ambiguously devilish sort or to ambiguous behavior of a devilishly benevolent sort. Smoke is thus as problematic a phenomenon as whiteness, and for the same reasons: the word smoke carries a complex mixture of malevolent and benevolent associations, and actual smoke interferes with perception.

The following smoky sentences, from chapter 44, should be worth the consideration of those to whom they are worth considering:

Furthermore, if we consider, what is popularly held to entitle characters in fiction to being deemed original, is but something personal - confined to itself. The character sheds not its characteristics on its surroundings, whereas, the original character, essentially such, is like a revolving Drummond light, raying away from itself all round it - everything is lit by it, everything starts up to it (mark how it is with Hamlet), so that, in certain minds, there follows upon the adequate conception of such a character, an effect, in its way, akin to that which in Genesis attends upon the beginning of things. (239)

Because "consider" typically is followed by a noun-phrase complement, at first glance the initial sentence may seem to be incomplete. The usage, however, may be regarded as an acceptable although unusual formula in which "consider" is to be read as a synonym for a verb phrase that includes a complement. (When my adolescent daughters ask what's for dessert, I might respond, "Well, I'll have to consider," meaning "I'll think about it." But if I don't give exactly the right intonation to consider, I confuse them, language literalists that they are when something like dessert is on the table for consideration.) Were this the only unusual expression, the passage might be comprehensible. But the second sentence continues with odd usages: the unnecessary modifier "essentially such," the confusing shift from "itself" to "it." The verb "ray" is particularly strange. None of the definitions given in the OED includes "to illuminate," a sense that fits if we take the verb as transitive and "all" as its object, whereas the phrase "raying away from itself" is more similar to an intransitive than to a transitive structure.

The sense alters after the mid-sentence dash: the clause "everything starts up to it" identifies not intrinsic traits of the original character but the effect of such a character, first on the rest of the work containing it and then on "certain minds." "Everything starts up to it" commands no single paraphrase, and additional vagueness is added with the uncertainty about who is to be included in that group of "certain minds": does the phrase refer to readers' minds or the author's mind, or is it part of a statement about human beings in general? In the rest of the sentence, the qualifications ("adequate" and "in its way") compound this uncertainty. In sum, this passage from The Confidence-Man, like the sentences on whiteness from Moby-Dick and "Ourselves are Fate" from White-Jacket, can be read many different ways, but no single interpretation can tie together all of the passage's sentences. Any given reading is likely to be endorsed only on the basis of its critical ideology: Bakhtin-influenced critics, for instance, have recently rediscovered the multivoiced quality Foster named forty years ago.

MELVILLE'S STYLE AND THE CRITICS

Explanations of Melville's difficult prose style fall into four broad and somewhat overlapping categories: biographical, literary-cultural, social, and metaphysical. A brief survey of these categories reveals two interesting consistencies. First, the critics who argue most persuasively recognize that Melville's style has a rhetorical effect that can be connected in some way to his interest in what is inexplicable, ineffable about human experience. Second, none of the critics admits that some sentences are intransigent in ways that would earn a freshman English student a lengthy tutorial session.

Criticism that falls into the biographical category is the least problematic but also the least interesting: its theme is that Melville's uncertain mental health probably had something to do with his style. Barbara Smith-Lemeunier notes that shortly after completing The Confidence-Man, Melville suffered a nervous breakdown, that actually could have surfaced occasionally while he was working on the book, darkening its mood and consistently leading him to challenge readers' conventional expectations (24); the garbled sentences might likewise be the result of a sick mind. Similarly, the all-day sessions in which Melville composed a sizable portion of Moby-Dick could have contributed to mental strain as well as to the eye strain that made him uncomfortable reading anything not in large type. Nor was he careful in revising his work. When considered as aesthetic objects, the novels present a different kind of challenge as a result of Melville's tendency to reconceive them halfway through but not to reconcile the earlier parts with the later ones. (The classic study of this problem in Moby-Dick is by George R. Stewart; see also Brian Higgins and Hershel Parker on Pierre.) We also know that it took Melville some time to master the methods not only of Shakespeare and Milton but of prose stylists like Thomas Browne who impressed him so much. For instance, Brian Foley, in his excellent discussion, notes that in Mardi Melville was still basically imitating the surface features of Browne's style, but that by 1850, when he began Moby-Dick, he had mastered its essence. This explanation, however, does not account for flawed sentences in Moby-Dick, Pierre, and The Confidence-Man; by the time he was writing these later works, Melville presumably had mastered his craft.

None of these biographical approaches is satisfactory. Melville, no Henry James, was remarkably reticent about how he composed and revised: whether in crafting paragraphs and chapters he had either actual or ideal readers in mind, whether he derived pleasure from giving a nice turn to a sentence, or if he even thought about such things when he was reading rather than writing. Melville kept no writer's log, nor does the record include any worksheets that show him practicing the styles of Shakespeare, Milton, and Browne; his imitations of their styles take place within his novels, and we simply do not know how conscious or planned these imitations were. Thus very little can be proven about Melville's conscious intentions at the level of the sentence. Most of what we know comes from allusive, cryptic, or ambiguous pieces, such as "Hawthorne and His Mosses" (Melville's famous 1850 review of Mosses from an Old Manse), and from comments embedded in his fictional writings, a context that complicates their interpretation. For this discussion, the most important point to be made about Melville's composing process is an inference that can be drawn from the fact that, rather than being evenly distributed, the problem sentences in his writings tend to be grouped together in places where the ideas are already complex. Because we know that Melville loved to grapple with difficult ideas, it seems reasonable to assume that these are the locations he would have scrutinized most closely, unlike setting or character descriptions that he may have tossed off quickly and proofread (if at all) on the fly. Thus we might expect fewer flawed sentences when he gives the impression of delving deeply into thought. This phenomenon of grouping, to which I will return, stands as one of the chief reasons for regarding Melville as a writer with a chaotic style.

Because Melville was so closely associated with the literary culture of his era, the dynamics of his participation in, struggle against, and withdrawal from that culture must have influenced his prose style. Criticism that falls into the literary-cultural category places Melville's style either within the range of language available to him (Brodhead, Porter) or within the context of the rhetorical effects he may have been aiming to achieve (Smith-Lemeunier, Putz, Bryant, Buell). Such work is important, and the pieces by Porter, Putz, and Lyons are essential reading for anyone who hopes to understand the many vectors that informed Melville's stylistic choices. In the best and most recent of these studies, Paul Lyons demonstrates that "Melville's interaction with sources" is informed by a method that results in, for instance, "the surface disorderliness of Ishmael's languages" (445). This method is based on an assumption Lyons attributes to Melville that readers respond sensually and associatively, rather than always logically, to stylistic allusions, to what Lyons aptly terms "metastylistic" moments. Melville thus "anticipates Bakhtin's sense that 'each sphere in which language is used develops its own relatively stable'" generic types (455). Lyons does not, however, look within sentences at faulty syntax or incompatible semantic features. His analysis implies that Melville expected readers not to attempt to make sense of each sentence or to identify cohesive links between sentences but to operate sympathetically, one might say, attending mainly to generic and tonal echoes. But Lyons's assumptions must remain speculative unless someone uncovers a trove of lost letters in which Melville presents his view on the craft of writing. Unable to be certain how Melville approached the crafting of sentence-level rhetoric, the stylistic critic must consider all of the data, including those sentences, which I term intransigent, that we may wish Melville had never written.

The same limitation is present in David Reynolds's massive study, Beneath the American Renaissance. In one section, Reynolds discusses how what he terms "the carnivalization of American language" manifested itself in the "irrational style" of the urban humorists, who "possessed a genre awareness" and thus recognized that "popular sensational novels were typically alinear and centrifugal, with successive chapters often being so inconsequential that time-space unities were broken" (480). He summarizes a burlesque of advice on writing "the Diabolic style," which emphasizes, in the words of the burlesque itself, "discontinuity and incongruity" (481): "Some urban humorists so exaggerated the irrational style that they produced new intensifications of what I have termed the American Subversive Style, characterized by extreme linguistic fragmentation and bizarre juxtapositions of disparate images" as well as by "free language play" (481-82). Because Melville had so fully assimilated "all popular humorous idioms of the day" (549) as well as "the energetic ironies and paradoxes of American working-class literature" (288), according to Reynolds he was quite likely to write a style as mixed as White-Jacket's patchwork coat. But the examples Reynolds gives of the popular literature contemporary with Melville do not show syntactic and semantic flaws of the sort under discussion here.

While he was writing a place for himself in the literary world of his time, Melville was also responding to social and political realities in relation to which his style is occasionally discussed, for instance by William Wasserstrom and Smith-Lemeunier. Most thorough and convincing are those pieces that treat his style as an attempt to present in a disguised fashion realities that could not be named: for instance, incest in Pierre (Gray 123) and homosexuality in Redburn and White-Jacket (Justus 53). However, this disguise typically came in the form of complex punning, double and triple meanings, and so forth, rather than in semantic or syntactic intransigence. Only in Pierre can one find passages in which an unnameable reality corresponds with intransigent sentences; as I will show below, what is being disguised is probably not the fact of incestuous desire in itself but the psychological turmoil that such desire stimulates.

Before the recent rush of literary-cultural explanations, the most common approach to Melville's style was to connect it to his metaphysical concerns. Most of these explanations agree that Melville intended to render the experience of what Ishmael, in chapter 42, refers to as the "mystical and well nigh ineffable" (188). Because such an experience is by definition beyond language, he sought a style that would not try to encompass but, as Robert Zoellner puts it (borrowing a word from Ishmael), to "subtilize" the phenomena (164-65). Such a style is built on inconsistencies and disruptions (Beaver 38, Brodhead 4-5), emphasizes processes and passions (Matthiessen 130-31, Brodtkorb 26-27, 147), and places intuitive over verbal understanding (Kawin 8). What Millicent Bell observes about Moby-Dick is equally true of the individual sentences making up this novel and those in the other novels as well: "Melville's masterpiece does not boil down in our try-pot, and if it states anything it is that experience is ultimately irreducible to thought - but [the book] is itself reducible neither to this meaning nor to any other" (31).

Like the other categories of criticism, metaphysical explanations do not really account for the syntactic and semantic flaws in Melville's writing, but they do suggest an interesting possibility: the problematic sentences manifest a rhetorical intention of working on readers at the level of preverbal, intuitive, passional response. This may sound like a curious suggestion to be making about a category of explanations I've termed metaphysical, because "metaphysics" is usually taken as having to do with the most encompassing questions of human existence, questions that are approached analytically, rationally. Yet as Roger Fowler points out, individual sentences do reflect in their structure the ways human beings make sense of the world; our intuitive knowledge of this fact is what leads us to infer an author's or a character's world-view from stylistic details (16-17). Hence it makes sense to assume that Melville's most general purpose (or the purpose we attach to the voice behind these texts, the voice we call "Melville") is reflected in the smallest unit of these texts that carries the Melville stamp: the punctuated sentence.

Despite their variety, critical studies of Melville's style display a strong continuity. From the 1940s and 1950s all the way through to this decade, critics have shown an ongoing interest in the emphasis of Melville's style on process, multiplicity, and indefiniteness and in its use of figurative and rhythmical language. Not only professional but general readers of Melville's novels sense that certain phenomena, such as the "ineffable" and the "chaotic," are essential to Melville even though they are nearly impossible to describe. Parts of Melville's work bear a strong resemblance to that "boggy, soggy, squitchy picture" Ishmael viewed in the Spouter-Inn, a picture with "a sort of indefinite, half-attained, unimaginable sublimity about it that fairly froze you to it" (Moby-Dick 12).

Although all four categories of criticism yield fairly traditional explanations of Melville's style, together they also suggest some points about his intransigent sentences. The biographical approach offers the possibility that if Melville worked the way many other writers work, then he occasionally had a "felt sense" of what he wanted to say if not the precise words with which to say it. (Sondra Perl discusses the role of the "felt sense" in the writing process and Alice Brand the related importance of emotions in driving intention.) As a participant in a society and a literary culture that still held romantic notions about the importance of inspiration, he probably felt authorized occasionally to suggest, rather than to spell out, by means of ungrammatical or semantically strained punctuated sentences, just as Whitman felt authorized to mix freely together formal and colloquial levels of diction. Some recent novelists have attempted to explore beyond the boundaries of language and rational thought, developing linguistic means for rendering the ineffable. (For discussions on the work of such novelists, see Kawin, Guetti, and Saldivar). Melville can be read this way as well, with the assumption that his styles were his means of "subtilizing" such inexplicable psychological phenomena as the fascinating terror evoked by whiteness, the terrible fascination of incest, and even the conflicted desire of readers for artful artlessness.

A writer who operates (either intentionally or intuitively) on the level of "felt sense," suggestion, and "subtilizing" will probably demonstrate two of the types of difficulty described by George Steiner: the tactical and the ontological. Tactical difficulties such as sentences that cannot be coherently parsed, display what might be termed "anti-grammar"; such difficulties often create in readers "a distinct sense in which we know and do not know, at the same time" (40). Ontological difficulties result from a violation of "the contract of ultimate or preponderant intelligibility between poet and reader, between text and meaning" (40). They lead to "blank questions about the nature of human speech, about the status of significance, about the necessity and purpose of the construct which we have ... come to perceive as a poem" (41). Melville's intransigent sentences display anti-grammar, for instance, when they play fast and loose with prepositions or mix transitive and intransitive senses of verbs, and they violate the contract between writer and reader by using rhetorical devices that assert a self-erasing or tautological explanation.

THE CHAOTIC STYLE: A GENERATIVE MODEL

I now turn to the task of demonstrating how Melville's intransigent sentences can actually constitute a style I am terming "chaotic." Neither traditional nor current techniques of literary analysis have recognized that the style of these sentences raises a problem of interpretation, and thus a radically new approach is needed. I propose to study Melville's sentences through the theory of deterministic chaos by combining the concepts of "generative metaphor" (drawn from Donald A. Schon) and "model" (as defined by Max Black). According to Black, a theoretical model will be compelling only to the extent that it generates inferences about the problematic domain that can be "independently checked against known or predictable data in the primary domain" (Black 231). In other words, a model must draw on scientific rather than on general knowledge. The model as Black defines it, however, assigns no role to intuition, an important element of Schon's generative metaphor that enables a researcher to move from a hunch to a new theory. A generative metaphor begins with the researcher's pre-analytic awareness of a similarity between a baffling phenomenon and one that is well understood; it generates elements of a theory about the problem area, but it does not yield a simple one-to-one correspondence between the primary and the problem domains (Schon 254-79). Like the model, the researcher arrives at a generative metaphor by drawing on a domain that she understands on an analytical, if not strictly scientific, level.

My term "generative model" captures important aspects of both Black's and Schon's theories. A generative model may be discovered on a hunch, but once discovered it establishes fairly explicit correspondences between the primary and the problematic domains. The most important work of a generative model is, like that of Black's theoretical model, to make predictions and then suggest ways for checking them. It differs from the theoretical model in that the primary domain suggests ways of checking the predictions within the problematic domain. My generative model takes deterministic chaos as the primary domain and the intransigent sections of Melville's prose as the problematic domain; hence my title, "Melville's Chaotic Style." In taking such an approach, I am not denying Richard Freadman and Seamus Miller's assertion that "literature is in important respects theory resistant" (205); if we are to have a theory of literature or a theory of style worthy of the designation theory, however, we must possess some means of making predictions on the basis of the theory as well as the possibility of checking these predictions against new information.

The relevant features of the primary domain - that is, of deterministic chaos - an be illustrated with reference to turbulence, which may be defined as the motion of a fluid in which local velocities and pressures fluctuate randomly (my own definition, dredged up from engineering courses years ago). One of the most intractable problems of fluid dynamics is that of describing what happens when a smooth-flowing fluid becomes turbulent. Fluid that is slow-moving, in the manner of light crude oil as it flows through a pipeline, will look smooth. Turbulence develops when the fluid reaches a constriction in the pipe, forcing it to move much faster.

One of the interesting things about turbulence is that the above definition, which may seem, as we used to say in Physics I, "intuitively obvious," is actually incorrect. Computer modeling and sophisticated experimental techniques have recently revealed that the turbulent motion of fluid is not random but exhibits what is called "deterministic chaos": although the motion of any molecule or group of molecules within the flow cannot be predicted, and although any two adjacent molecules can after an infinitesimal time be arbitrarily far apart, the motions will all display certain patterns. (These patterns, technically termed "strange attractors," are not visible to the naked eye and are extremely difficult for the mind's eye of all but a few gifted mathematicians to visualize; they are present not in three-dimensional space but in a mathematically constructed "phase space.") Thus the motion of fluid is not wholly random; its pattern is both "deterministic," that is, governed on the large scale by the laws of fluid dynamics, and "chaotic," that is, locally unpredictable. Any system displaying deterministic chaos will also be characterized by "sensitive dependence on initial conditions." When a fluid is flowing smoothly, a small change in the conditions of flow will cause a correspondingly small change in the structure of the flow and hence in the relative positions of any two molecules. But when the fluid approaches the onset of turbulence or actually reaches a turbulent state, a small change in the conditions of flow can be magnified exponentially. Such magnification is known as the "butterfly effect" after the famous paper by Edward Lorenz titled "Predictability: Does the Flap of a Butterfly's Wings in Brazil Set Off a Tornado in Texas?" (Gleick 322, note to page 20).

Although at first it may seem counterintuitive, the concept of deterministic chaos may be found in countless natural phenomena, from a dripping faucet to species populations. Thus we should not be surprised to find strong hints of deterministic chaos in linguistic systems. We saw above how intransigent sentences are distributed unevenly throughout Melville's works: the sentences tend to cluster at the ends of chapters just as the motion of the oil molecules becomes jumbled where the pipe narrows. A second similarity between the primary domain (deterministic chaos) and the problematic domain (Melville's sentences) is the breakdown of well-formedness. Molecules flowing smoothly through a pipe can arrange themselves in a large number of orderly ways, but when the flow becomes turbulent, the positions of the molecules can no longer be predicted. In much the same way, any set of rules (a syntax) operating on a finite lexicon can produce an infinite number of well-formed sentences as Noam Chomsky pointed out long ago; when a sentence breaks down, however, there can be no way of telling what word (or even what kind of word) will come next.

A model becomes generative when it suggests a new line of thinking. For instance, can these clusters of sentences be shown to contain strange attractors? To answer this question would require a combination of linguistic and mathematical analysis such as I have as yet only sketched out: it would involve treating passages of Melville's prose as systems evolving through time and then applying the mathematical technique of logistic mapping to aspects of those systems that might reveal patterns similar to those discovered by Robert Shaw in the famous dripping-faucet studies (Ian Stewart 188-91). (One might look, for example, at the time-evolution of the topic-comment structure; a second possibility is the time-evolution of the topic at the deep-structure level. Like the dripping faucet, these can be treated as one-dimensional phenomena.) This line of thinking is where my approach radically departs from both traditional and current literary studies as well as from the use to which chaos theory has been put by literary and cultural theorists. (I will return to this point in my conclusion.) The generative model actually suggests a research program that could lead to new knowledge about Melville's style, and new knowledge (as opposed to new approaches) is a rarity in literary studies.

My main interest in the remainder of this essay, however, is to regard the model of deterministic chaos metaphorically more than mathematically, using it to generate a question of a different sort than those raised by biographical, metaphysical, and other approaches to Melville's style. The most important such question is this: What might be the "initial conditions" that result in clusters of intransigent sentences? The answer requires acknowledging that if "initial conditions" include the readers experiencing the sentences - that is, if a rhetorical perspective is taken - then the whole system changes since some readers will not even notice their intransigence. I have demonstrated that the passages from Moby-Dick and The Confidence-Man cited above contain intransigent sentences, but their effect on normal semantic and syntactic relationships is not so great as to make the passages unreadable. In fact, if critics' use of these passages is any indication, the sentences pose more problems for the general reader than they do for the Melville specialist. In his description of three types of reading strategies - perception, comprehension, and interpretation - George Dillon emphasizes that interpretation governs the other two types, especially "when a text is grammatically irregular, obscure, or elliptical: the processes of perception and comprehension do not get enough data, or enough consistent data, to select a reading, so one imposes a propositional structure or contextual frame which supports one's sense of what the passage should be saying" (xxii; my emphasis). Melville specialists, unless they set out to find sentence-level problems, read either to confirm or to apply what they know; because they have interpretive strategies solidly in place, they will not experience the intransigent passages as such. Encountering "Ourselves are Fate," for example, which clearly contains "tactical difficulties" of the sort Steiner describes, the specialist has an extensive system of relationships that aids in the interpretation of the key term "Fate" and in the identification of the "two senses" in which the word can be understood - senses, I repeat, that the passage does not specify.

The general reader lacks the background to impose a propositional structure on this passage and thus may encounter the full force of the flawed syntax. But this reader also desires to understand and therefore may make a preconscious choice to interpret the sentence roughly as the specialist does: "Oh, sure, we make our own fate, and the fate we make reflects our character." For the specialist as well as the general reader, this interpretation is "what the passage should be saying" based on the local context and on the role we in the twentieth century typically assign to "fate." Once a passage is interpreted, discrepancies between this interpretation and what the passage can be proved to mean cease to exist. A reader's drive to interpret controls this rhetorical system in exactly the same way that the person who manages the pipeline slows down the pumps when a turbulent condition is imminent. For readers who have strong defenses against tactical difficulties, "sensitive dependence" on such difficulties (or on other initial conditions) simply cannot be present.

On the other hand, when readers are taken out of the system and when the syntactic flow of Melville's sentences is correlated with the flow of topics from chapter to chapter and from paragraph to paragraph, a "sensitive dependence" does seem to exist between sentence-level chaos and topics that participate in "the ineffable." Most discussions of Melville's style sooner or later address his attempt to render the experience of the ineffable, which has been variously described as emphasizing intuitive rather than verbal understanding (Kawin 8), stressing processes (Matthiessen 130-31) and passions (Brodtkorb 26-27, 147), foregrounding the irreducibility of experience to thought (Bell 31). These discussions, however, tend to concentrate on phenomena external to the characters that have elicited verbal responses from the characters, rather than on the stylistic traits of those responses. Such a concentration does not fit with Melville's steady focus, in his mature works, on the inescapably subjective nature of perception. "With me," Ishmael writes, "those things must exist" (Moby-Dick 195). "With" here is another of Melville's baffling prepositions, but two plausible readings, substituting "within" and "for," emphasize subjectivity. The point of "The Whiteness of the Whale" is not whiteness in itself but the experiencing of whiteness, the psychological and physiological responses of humans (or of horses, as in the chapter's final six paragraphs) in its presence. Ishmael implies at the beginning of the chapter that these phenomena can be described: the "thought, or rather vague, nameless horror concerning" Moby Dick is "so mystical and well nigh ineffable" that he "almost despair[s] of putting it in a comprehensible form" (287). The qualifiers "well nigh" and "almost" imply that the horror or thought only approaches ineffability and that his apprehension about succeeding at his task only approaches despair. Yet if the chapter is read as the unrevised record of Ishmael's thoughts as they occurred to him rather than as a step-by-step analysis of whiteness, then its rhetorical climax justifies his despair. Ishmael began this chapter believing he might be able to put "in a comprehensible form" his reaction to the white whale. However, the initial condition - his choice of this topic for the chapter - quickly destabilized first his sentence-to-sentence cohesion and then the internal structure of his sentences to the point that the chapter's final sentences exhibit not only tactical but ontological difficulties, calling into question language's capacity to signify. Put simply, it is Ishmael's verbal encounter with the idea of Moby Dick that culminates in intransigent, chaotic sentences; at best, these sentences dramatize for a reader the powerlessness of language in the face of ineffable experiences.

The same line of reasoning applies to the end of book 7 of Pierre: the sentence-level chaos results not from the abstract topic of the passage - the fascination of incest - but from the specific experience of one human being with this topic. Pierre has just had an erotically charged but nonphysical encounter with Isabel, his half sister:

Now Pierre began to see mysteries interpierced with mysteries, and mysteries eluding mysteries; and began to seem to see the mere imaginariness of the so supposed solidest principle of human association. Fate had done this thing for them. Fate had separated the brother and the sister, till to each other they somehow seemed so not at all. Sisters shrink not from their brother's kisses. And Pierre felt that never, never would he be able to embrace Isabel with the mere brotherly embrace; while the thought of any other caress, which took hold of any domesticness, was entirely vacant from his uncontaminated soul, for it had never consciously intruded there.

Therefore, forever unsistered for him by the stroke of Fate, and apparently forever, and twice removed from the remotest possibility of that love which had drawn him to his Lucy; yet still the object of the ardentest and deepest emotions of his soul; therefore, to him, Isabel wholly soared out of the realms of mortalness, and for him became transfigured in the highest heaven of uncorrupted Love. (142)

I have shown that at the deep-structure level "the actual agent in all these associations, identifications, and mystifications" is not Fate but Pierre, even though on the surface the agent seems to be Fate (Kearns 48). Except that Pierre's thoughts are being rendered by an omniscient narrator, this passage is similar to Ishmael's meditation on whiteness both in the progression of its thoughts and in the totally baffling strings of words in its climactic paragraphs. ("Somehow seemed so not at all" is probably the best example of such a string.) Similarly, the passage from chapter 44 of The Confidence-Man does not result from some overwhelming confusion about the concept of originality even though it does foreground that issue. The voice of this passage is quite restrained, even distant; the sense of urgency is much less than it was in the passages from Pierre and Moby-Dick. But the problem raised by this passage and, indeed, by all three of the self-reflexive chapters in The Confidence-Man is just as knotty: how do notions of "the real" transfer between fiction and life? The Confidence-Man, Pierre, and Moby-Dick each suggests that people can simultaneously accept mutually exclusive ideas about topics that carry a great deal of emotional weight (whiteness is both sacred and profane, characters should be both realistic and idealized). But these ideas are not amenable to any dialogic balance; in Melville's work their tendency to destabilize language is so strong that the result is intransigent sentences.

CONCLUSIONS AND PROSPECTS

Typically we think that the ineffable for Melville had something to do with metaphysics, but by the time he was writing his mature works he had a well-thought-out system for speculating on and dramatizing metaphysical questions: White Jacket's fall from the masthead, Ishmael's solitary two-day float, Ahab's pursuit, and other famous events all illustrate this system. I would like to suggest that from mid-1849 (when he was working on White-Jacket), and perhaps even earlier, Melville's goal was no longer to explain the ineffable but to dramatize its psychological effects. He probably came to understand, either consciously or by means of his writer's intuition, that it is impossible to control entirely a reader's response to any passage and therefore he had to do something other than provide well-formed descriptions of the ineffable if he hoped to confront at least his first-time reader with the reality of these subjective psychological phenomena. Hence his tendency to combine an analytical voice (rather than narrative or expository) with a psychological phenomenon that cannot be circumscribed by language. For those readers who approach the style on the level of words and sentences and do not leap immediately to interpretation, this combination generates a "felt sense" that lies outside the boundary of rational thought but is embodied within the text in passages revealing the chaotic style.

This explanation of the chaotic style does not require a belief that Melville consciously crafted some sentences to be intransigent but only that he wanted to portray limited human beings struggling to articulate the ineffable. The chaotic style itself, as a hypothesis, suggests several research projects besides the logistic mapping project already mentioned. For example, I have not studied Mardi for intransigent sentences, but if they could be found in the context of an analytical voice expressing a sense of the ineffable, my claim that Melville's works reveal a chaotic style would be strengthened. It would also be interesting to analyze how readers paraphrase the chaotic passages; if the paraphrases tend to contain concepts to which the label "ineffable" can be attached (even if those concepts are embedded in a variety of propositions), this result would support my assertion that such passages communicate a felt sense of something not amenable to containment by language. Finally, other passages in which an analytical voice expresses a sense of the ineffable could be examined; if any such passages did not demonstrate the chaotic style, my hypothesis would need to be radically revised or rejected outright. (Based on years of reading and teaching Melville, I do not expect this to happen.)

The kind of analysis I am suggesting could also rehabilitate chaos theory for literary scholarship. My search through the MLA Bibliography turned up five articles using chaos theory in 1987, two in 1990, three in 1991, and one in 1992, as well as two books and a collection of essays in 1991 and a dissertation in 1992. Probably not coincidentally, I found nothing before 1987, the year in which James Gleick's best-selling Chaos appeared. The sudden loss of interest in chaos theory can be explained by the highly metaphorical quality of the scholarship that originally made use of it. This scholarship relied on typical categories of literary analysis, such as figurative language, ambiguity, and repetition, as well as on typical critical moves, such as deconstruction's search for conflicting codes embedded in texts. For example, N. Katherine Hayles used concepts from chaos theory such as self-similarity and iteration to view culture as reflecting a destabilization of the order-disorder opposition and to suggest that Derrida's wordplay demonstrates the inescapability of indeterminacy in writing ("Chaos as Orderly Disorder" 313-14). Alexander Argyros similarly argued that narrative may serve a culture as a "flexible and turbulent laboratory in which to invent new knowledge" ("Narrative and Chaos" 670). He also offered a suggestion I find fascinating, that the human experience of self as well as "[t]he bulk of human ideas ... can be represented by chaotic attractors" (Blessed Rage 291,296).

Although such connections are wonderfully thought-provoking, they really are nothing more than alchemy or metaphor masquerading as theory. While it is interesting to note that a culture's structures may be similar at different levels, or, as chaos theory would put it, may display self-similarity, such connections have not been shown to lead to any research program that could verify or disprove their basis in theory. In the collection of essays edited by Hayles, she titled the first section "Chaos: More than Metaphor" and praised one piece for going "beyond a metaphoric connection to assert a deeper congruence" between chaotic systems and metafiction ("Complex Dynamics" 21). I admire and respect Hayles's work, which embodies a commitment to interdisciplinary inquiry, one of the strengths of current literary studies; however, her claim for a deeper congruence, without any support from a qualitatively different realm of inquiry, such as mathematics or even stylistics, looks suspiciously like the trace from a myth of presence. The same is true of Hayles's repeated use of the term "isomorphic" to describe what is no more than a similarity. Ian Stewart's caution is worth keeping in mind: not everything that looks chaotic necessarily has "an underlying deterministic cause," and hence, "[i]f you want to use chaos, get a mathematician who understands it on the team" (327-28).

I have a hunch that Melville intuitively grasped the concept of deterministic chaos. In Moby-Dick, early in his chapter titled "Cetology," he wrote, "the classification of the constituents of a chaos, nothing less is here essayed" (228-29). Moreover, in The Confidence-Man he wrote that just as an orbit can have only one planet, a "work of invention" can have only one original character: "Two would conflict to chaos" (239). Why? Was he thinking of the intractable three-body problem of celestial mechanics (two planets plus the governing star), which displays all the characteristics of deterministic chaos? Perhaps Melville felt it to be true that the movements of two wholly original characters under the influence of a central governing mind could never be predicted, although, like the "constituents of a chaos," those movements could still be classified - probably not in the normal sense, but apprehended by the human brain nonetheless at some level that goes beyond the linguistic. Melville's chaotic style may be thought of as exemplifying a point made by Steven Pinker about expository writing: it "requires language to express far more complex trains of thought than it was biologically designed to do" (401).

Every sentence can mean more than it says as long as it exists in a context to which some set of interpretive procedures can be applied. I have shown that Melville's syntactically flawed sentences can "mean" in ways that have little to do with what they say, and that this meaning may result from characteristics of language and cognition that have only begun to be understood. Keeping this possibility viable is one reason for rehabilitating deterministic chaos as a theory-generative model in the human sciences and humanities: the model may shed some light on the noncognitive realm of understanding. It may also help us understand how sentence style helps shape a reader's experience of a text, especially when that experience yields a felt sense that cannot be reduced to propositions about meaning. More generally, it may produce insight into how human beings find meaning in any chunk of language that resists linear analysis, from Emily Dickinson's enigmatic two-line poem number 1127, "Soft as the massacre of Suns / By Evening's Sabres slain," to Lucky's monologue in Waiting for Godot.

Works Cited

Argyros, Alexander. A Blessed Rage for Order: Deconstruction, Evolution, and Chaos. Ann Arbor: U of Michigan P, 1991.

-----. "Narrative and Chaos." New Literary History 23 (1992): 659-73.

Beaver, Harold. "Mardi: A Sum of Inconsistencies." Herman Melville: Reassessments. Ed. A. Robert Lee. Totowa: Barnes & Noble, 1984. 28-40.

Bell, Millicent. "The Indeterminate Moby-Dick." Approaches to Teaching Melville's Moby-Dick. Ed. Martin Bickman. New York: MLA, 1985.23-34.

Black, Max. Models and Metaphors: Studies in Language and Philosophy. Ithaca: Cornell UP, 1962.

Brand, Alice. "The Why of Cognition: Emotion and the Writing Process." College Composition and Communication 38 (1987). Rpt. Rhetoric and Composition: A Sourcebook for Teachers and Writers. 3rd ed. Ed. Richard L. Graves. Portsmouth: Boynton/Cook, 1990. 152-59.

Brodhead, Richard H. "Trying All Things: An Introduction to Moby-Dick." New Essays on Melville's Moby-Dick. Ed. Brodhead. Cambridge: Cambridge UP, 1986. 1-21.

Brodtkorb, Paul. Ishmael's White Word: A Phenomenological Reading of Moby Dick. New Haven: Yale UP, 1965.

Bryant, John. "Melville's Comic Debate: Geniality and the Aesthetics of Repose." American Literature 55 (1983): 150-70.

Buell, Lawrence. "Moby-Dick as Sacred Text." New Essays on Moby-Dick. Ed. Richard H. Brodhead. Cambridge: Cambridge UP, 1986. 53-72.

Dillon, George. Language Processing and the Reading of Literature. Bloomington: Indiana UP, 1978.

Feidelson, Charles. Symbolism and American Literature. Chicago: U of Chicago P, 1953.

Foley, Brian. "Herman Melville and the Example of Sir Thomas Browne." Modern Philology 81 (1983-84): 265-77.

Foster, Elizabeth S. Introduction. The Confidence-Man: His Masquerade. New York: Hendricks House, 1954. xiii-xcv.

Fowler, Roger. Linguistics and the Novel. London: Methuen, 1977.

Freadman, Richard, and Seumas Miller. Re-thinking Theory: A Critique of Contemporary Theory and an Alternative Account. Cambridge: Cambridge UP, 1992.

Gleick, James. Chaos: Making a New Science. New York: Viking, 1987.

Gray, Richard. "'All's o'er, and ye know him not': A Reading of Pierre." Herman Melville: Reassessments. Ed. A. Robert Lee. Totowa: Barnes & Noble, 1984. 116-34.

Guetti, James. The Limits of Metaphor: A Study of Melville, Conrad, and Faulkner. Ithaca: Cornell UP, 1967.

Hayles, N. Katherine. "Chaos as Orderly Disorder: Shifting Ground in Contemporary Literature and Science." New Literary History 20 (1988-89): 305-22.

-----. "Complex Dynamics in Literature and Science." Chaos and Order: Complex Dynamics in Literature and Science. Ed. Hayles. Chicago: U of Chicago P, 1991. 1-33.

Higgins, Brian, and Hershel Parker. "Reading Pierre." A Companion to Melville Studies. Ed. John Bryant. Westport: Greenwood, 1986. 211-39.

Justus, James H. "Redburn and White-Jacket: Society and Sexuality in the Narrators of 1849." Herman Melville: Reassessments. Ed. A. Robert Lee. Totowa: Barnes & Noble, 1984. 41-67.

Kawin, Bruce. The Mind of the Novel: Reflexive Fiction and the Ineffable. Princeton: Princeton UP, 1982.

Kearns, Michael. "Interpreting Intentional Incoherence: Towards a Disambiguation of Melville's Pierre; or, the Ambiguities." Bulletin of the Midwest MLA 16 (1983): 34-47.

Lewis, R. W. B. Afterword. The Confidence-Man. New York: Signet, 1964. 261-76.

Lyons, Paul. "Melville and His Precursors: Style as Metastyle and Allusion." American Literature 62 (1990): 445-63.

Matthiessen, F. O. American Renaissance: Art and Expression in the Age of Emerson and Whitman. London: Oxford UP, 1941.

Melville, Herman. The Confidence-Man: His Masquerade. 1857. Evanston: Northwestern UP, 1984.

-----. Moby-Dick. 1851. Evanston: Northwestern UP, 1988.

-----. Pierre; or, the Ambiguities. 1852. Evanston: Northwestern UP, 1971.

-----. White-Jacket or The World in a Man-of-War. 1850. Evanston: Northwestern UP, 1970.

Perl, Sondra. "Understanding Composition." The Writer'sMind: Writing as a Mode of Thinking. Ed. Janice N. Hays, et al. Urbana: NCTE, 1983.43-51.

Pinker, Steven. The Language Instinct. New York: William Morrow, 1994.

Porter, Carolyn. "Call Me Ishmael, or How to Make Double-Talk Speak." New Essays on Moby-Dick. Ed. Richard H. Brodhead. Cambridge: Cambridge UP, 1986. 73-108.

Putz, Manfred. "The Narrator as Audience: Ishmael as Reader and Critic in Moby-Dick." Studies in the Novel 19 (1987): 160-74.

Reynolds, David. Beneath the American Renaissance: The Subversive Imagination in the Age of Emerson and Melville. New York: Knopf, 1988.

Saldivar, Ramon. Figural Language in the Novel: The Flowers of Speech from Cervantes to Joyce. Princeton: Princeton UP, 1984.

Schon, Donald A. "Generative Metaphor: A Perspective on Problem-Setting in Social Policy." Metaphor and Thought. Ed. Andrew Ortony. Cambridge: Cambridge UP, 1979. 254-83.

Smith-Lemeunier, Barbara. "Coconut Shy on the Mississippi: Melville's Style in The Confidence-Man." Hagiographie et Iconoclastie: Modeles Americains. Ed. Serge Ricard. Aix-en-Provence: U de Provence, 1984. 7-29.

Steiner, George. "On Difficulty." On Difficulty, and Other Essays. New York: Oxford UP, 1978.

Stewart, George R. "The Two Moby-Dicks." American Literature 25 (1954): 41748.

Stewart, Ian. Does God Play Dice? The Mathematics of Chaos. London: Blackwell, 1989.

Wasserstrom, William. "Melville the Mannerist: Form in the Short Fiction." Herman Melville: Reassessments. Ed. A. Robert Lee. Totowa: Barnes & Noble, 1984. 135-56.

Zoellner, Robert. The Salt-Sea Mastodon: A Reading of Moby-Dick. Berkeley: U of California P, 1973.

Michael Kearns is an associate professor of English at the University of Texas of the Permian Basin and the author of recent and forthcoming articles on Henry James, Nathaniel Hawthorne, Herman Melville, and cognitivist approaches to narrative. His Metaphors of Mind in Fiction and Psychology (University of Kentucky Press, 1987) won the 1986 Midwest MLA Book Award. His current project is a book on rhetorical narratology.

COPYRIGHT 1996 Northern Illinois University
COPYRIGHT 2004 Gale Group