Chapter 2:

The Socio/Cognitive Context

Reader, Text, Reader-Text

It is not enough to have established the distinction between literary and instrumental language; we must contextualize it as well. The theoretical propositions argued in the first chapter are only interesting at this point; to judge their ultimate import they must be examined within two specific contexts: the context of the reader, and the context containing the reader, the socio-historical context. The former will determine the effect and reception of literary and instrumental language on/by the reader; the latter will determine their effect and reception on/by the culture at large.

It has been suggested that instrumental language functions in the mind of its user as a tool, something put to a particular purpose without consciousness of its nature or operation. Certainly this is the way most of us use language on an everyday basis; indeed, to remain ever-conscious of how one's language functions would likely result in awkward, unnatural speech. We also do not expect our audiences to question the meanings of our utterances or analyze our language in any detail, unless we cue them to do so or the context suggests it. In this sense, instrumental language in and of itself has no significant effect on its user or audience. Similarly, a reader of an instrumental text will receive the text's monosemic message without becoming aware of the text as text.

The user of instrumental language, we might then note, requires no special training; oral literacy is the only essential for this type of communication. Within the field of linguistics, this capacity would be described as a user's linguistic competence and performance. Concerning literature, however, there are several critics and theorists who argue that the reading of literary texts demands specific skills; they would posit what Jonathan Culler describes as a literary competence. In this approach, literature is not defined according to textual properties, but according to a theory of discourse. As Culler has it, "To read a text as literature is not to make one's mind a tabula rasa and approach it without preconceptions; one must bring to it an implicit understanding of the operation of literary discourse which tells one what to look for" (SP 115-6). He suggests, as an example, that a reader lacking this competence would not know what to make of a poem; such a person would be routed by the unusual typographic organization, among other things. Gerard Steen, in establishing the grounds for his own empirical study, further distinguishes between a "professional" reader and a "private" one, noting that the former seeks out, even creates, ambiguity and metaphor where the latter might read instrumentally (30). Culler and Steen both seem to agree, in principle, that the reader must bring some machinery to the text to process it other than instrumentally. Steen, however, openly admits that he does not distinguish between poetic and conventional metaphor; this makes his work essentially incompatible with this study.

This discourse-oriented approach is not wholly convincing. It is of course granted that familiarity with significant works, theories, and forms will aid one in interpreting literary texts within institutionalized discourse; indeed, literary competence is a genuine set of skills and knowledge. For example, if one reads John Berryman's "Dream Song 310" and comes upon the lines

Henry, monstrous bug, laid himself down

on the machine in the penal colony

without a single regret. (13-15)

without knowing Kafka's stories "In the Penal Colony" and "The Metamorphosis," several allusions will be missed; some of the poem's resonance will be lost. The concept of allusion is itself principally literary, and familiarity with it is likely a part of competence. However, this is a very different thing from being unable to read the poem altogether, or resolve some meaning out of it. Of course, even maximal literary competence will not much aid a reader in processing the metaphor

(2) Tom is a thumb.

Knowing this is a metaphor has little to do with understanding it. In this example, as with all literary language, the presence of polysemy is the principal challenge to the reader, and this polysemy presents itself regardless of one's training; all readers must and will delineate connections between "Tom" and "thumb." Few, if any, will stop at the first meaning arrived at with the belief that the "right" meaning has been discovered; it is far more likely that a reader will go on to evaluate other possible meanings. Literary competence may aid in distinguishing more or less significant meanings, and the practice of literary interpretation will provide structure upon which to array these meanings, but these acquired skills are not necessary for literary language to affect the reader in ways that instrumental language does not.

In defining literary competence, Culler makes an incorrect observation that, when righted, defines the experiential field in which literary language has its greatest effect on the reader. Discussing Empson's 7 Types, he concludes, ". . . the most obvious feature of literary competence is the intent at totality of the interpretive process" (SP 126). Culler presents the idea of textual totality, which is essential to the New Criticism, as unique to the interpretation of literature. This is not so. The intent at totality is common to all readers reading all texts, almost without exception. It is, in fact, essential to all communication. For example, a reader of a magazine article automatically assumes that its contents are all interrelated and, when combined, function as a whole. Similarly, the critic who attempts to demonstrate the wholeness of a poem, envisioning it as an independent object, more or less unconsciously does the same with an essay he might read about that same poem. Drawing from the mundane, we expect in conversation that each dialogue itself is of a piece, something that, if studied, would reveal its interconnections, yield some kind of framework. In every case, totality is assumed and the text is processed based upon this assumption. It is, in a sense, an ever-present belief in authorial intention.^[1]

This intuitive assessment is realized in a communication model recently developed by Dan Sperber and Dierdre Wilson, one described by and studied under the heading "Relevance Theory." This theoretical model maintains that we cognitively attempt to maximize the relevance of communicative inputs and outputs (261). That is, we process information according to a belief that it will be meaningful to us, that it will somehow "relate" to us; we further assume that communicative acts directed at us (whether they actually are or we merely perceive them to be) have a purpose and merit reception. As they describe it, "To communicate is to claim an individual's attention: hence to communicate is to imply that the information communicated is relevant" (vii). Given the number of potential sensory inputs we constantly receive, all of which are messages of a sort, we are forced to restrict our attention to those which seem uniquely significant.

This concept is, at its extreme, a paranoia of the sort most strikingly demonstrated by conspiracy theories. Take, for example, a man who, in the course of a day, observes fifteen different people wearing black hats appear in his vicinity at fifteen different times. It is quite possible that he will construe these unusual though potentially unconnected events as evidence of a plot against him. In reality, the black hats may be delegates attending some convention, and their distinguishing attire may be symbolically related to their purpose. Yet the poor would-be target has every reason to believe that this ostensibly coordinated communication is in some way connected to him. Another case of maximization of relevance exists in the mindset of the medieval period, where because of the guarantee of authorial intention (God's), all potential signs are read as pointing to the grand scheme of the universe. Thus the medievals created such works as bestiaries, which explained what animals "meant" according to this scheme.

In conjunction with the assumptions of totality and relevance, we may add that readers further expect some kind of closure, as Eco and Barthes have observed. Whether the text is a poem, an essay, or a conversation, a conclusion is generally arrived at, even if it is a provisional or temporary one. It is through these three communicative expectations, then, held for all texts and all discourses, that the difference between literary and instrumental language is most immediately discerned and felt. It is at this level, foremost, that the reader is affected by literary language and unaffected by instrumental. Instrumental language, the tool of speech, of the unconscious mind, naturally lends itself to the satisfaction of totality, relevance, and closure; we use it to communicate to these ends because we know our audience expects that; we know our audience expects that because we ourselves do.^[2] The limited informational capacity of instrumental communications ensures that an audience will not read irrelevant meanings out of the message; nor will the message carry a suspended conclusion, for to do so would require the indecision created by competing meanings. Totality, of course, will readily be created by the audience out of the relative paucity of meanings. Indeed, if the ideal of an instrumental text is one meaning, then totality is inherent to the instrument.

Literary language, in contrast, denies totality, relevance, and closure. The surplus of meaning makes all three logistically problematic, if not impossible. In terms of information theory, the sheer improbability and unexpectedness of such a maximally polysemic message leaves its recipient unable to process it, at least according to familiar methods. This is what Barthes means when he says that the poetic (the polysemic) excludes men:

It initiates a discourse full of gaps and full of lights, filled with absences and

overnourishing signs, without foresight or stability of intention, and thereby

so opposed to the social function of language that merely to have recourse

to a discontinuous speech is to open the door to all that stands above

Nature. (WDZ 48-9)

In contrast, instrumental language (his "classical language") "postulates the possibility of dialogue, it establishes a universe in which men are not alone. . ." (49). This distinction between the polysemy of literary language and the monosemy of instrumental language is transhistorical, in this sense. Human beings are essentially totalizing machines. The effect of literary language on a reader, then, is the result of a violation of totality, relevance, and closure. Whereas instrumental language fulfills these communicative expectations and is therefore processed without question, literary language calls attention to itself. As a vehicle for a presumed intended meaning, it is suspect. This cognitive break, this awareness of the medium has the potential to change the reader. Keeping Culler's literary competence in mind, it is not so much that the reader must have skills with which to process the text, but that the text processes the otherwise unaware reader.

When a reader is faced with an inconclusive (literary) text, he is immediately faced with the possibility that language may "fail," at least in the traditional, instrumental sense; and at the point that an instrumental reading fails, the reader is forced to make a decision. If the text is determined to be at fault, and totality, relevance, and closure are then similarly determined to be unreachable, the reading will be abandoned. In the case of Lot 49, for example, such a reader would conclude that the author somehow could not communicate what he "meant." To this reader, the ambiguity of the Trystero, among many other things, is due to a problematic message, "noise," as it were. As Peter Stoicheff observes, writing of metafiction and polysemy, such texts are "ostensibly meaningless for the reader of mimetic fiction" (CaO 88). Mimetic fiction, we may understand, is principally instrumental; it demonstrates a faith in the representativeness of words, their ability to accurately reconstruct reality.

Most readers, however, refuse to deny totality, relevance, and closure, essential as they are to human communication. This choice, this communicative auto-da-fé, will lead them to assume that the misunderstanding is their own, and that the message, if examined more thoroughly, will yield its proper meaning. This is, again, essentially the New Criticism: using binaries and other tools, a trained reader will analyze each element of the poem-message, and subsequently categorize and order them to fit an interpretive whole. The layman, of course, possessing no such tools, will make homelier connections to create not an interpretation, but a relevant, total meaning of the message.

The effect of a literary text on a reader who chooses to abandon it is limited to that identified above: a forced awareness of the potential for communication failure. In this case, the failure is likely to be perceived as unusual, its source inarticulable. Given a reader who continues processing that text, however, the potential effect is great. This reader will begin his reading anew, now attempting to identify and pursue the ever-increasing number of competing meanings in the assumption that this process will eventually reveal the totality, relevance, and closure that has thus far eluded him. As he discovers more meanings, the number of connections he attempts to make among them also grows, exponentially. During this process he will constantly revise his "total" reading to incorporate the increasing number of discontinuities. In the case of a complex text, it is highly probable that some of these perceived meanings are not authorial constructs; indeed, as the reader becomes more sensitized to the play of polysemy, the number of these unintended polysemies is likely to increase. In turn, as the meanings pile up, the reader will begin to doubt authorial intentionality and the possibility of totality; in other words, he will begin to realize the improbability of the text's author having planned all of the meanings he has discovered, and this will naturally lead to the conclusion that there is no predetermined yet undiscovered scheme that will contain them. When these events occur, and the few initial polysemies that first alerted the reader give way to a suspicion of all language as polysemic, a perceptual phase transition has occurred. It is at this point that one is most aware of the immanence and indeterminacy of the Word.

Yet the reader cannot completely discount the possibility of totality; indeed, to approach a text without such a faith in totality is absurd, and counter to the nature of human communication. This is why a randomly generated text, as previously demonstrated, may be closer to an informational/polysemic ideal, yet our awareness of the absence of totality or authorial intent makes us uninterested in processing it. Ideally, of course, one might simply enjoy the free play of meanings, sans totality, and make the connections because one can. This is the game quality of literary texts, and as Theall observes, it is an open game, without chess-like boundaries. He values it, appropriately, because such games "lead to a liberating possibility" (202). In this case, one is freed from the constraints of instrumental language use and the concepts encoded in it.

The likelihood of the perceptual phase transition occurring is determined partly by the degree to which a text is polysemic and partly by the reader's own cognitive bent. Certainly there are some texts, and Finnegans Wake comes to mind here, that cross a threshold level of polysemy so that the phase transition is in effect forced upon the reader, the only alternative being the abandonment of the text as incomprehensible.^[3] Ideally, once this phase transition takes place, it will hold for other texts as well. That is, the reader's awareness of the potential atotality of texts, of the artificial nature of closure, and of the polysemy (both intended and not) lurking in all language will alter his reading process. This is consonant with Olsen's notion that postmodern fiction creates, through its destruction of old assumptions, "the possibility for construction of a new set of assumptions about language and experience" (114). This is literary language's capacity to transform the reader, give him new lenses through which to observe the operation of texts, to become more aware of the nature of the communicative act. This new "vision" is comparable to Steen's notion that a professional interpreter will always discover more polysemies than a non-professional; the difference here is that it has nothing to do with profession or the assumption of a role. Once a person becomes aware of language in this manner, only being lulled into instrumental unconsciousness can create a return to "normal" reading. We might say literary critics of this era are permanently enlightened, much as their medieval counterparts, who searched all reality for meanings, clues, to their Book.

Another effect of literary language, no less important, and indeed more immediately productive rather than destructive, is the creation of new meanings out of the noise of a complex text. Intuitively, it is clear that the presence of multiple meanings will encourage the reader (again, in the movement towards totality) to try to make connections between these meanings, just as a strong metaphor will cause its reader to connect its S and P. Encountering for the first time

(11) Paradise is an Erector Set

one is likely to take away from the statement a new meaning of Paradise, of Erector Sets, or both. These meanings, whether sacred, sacrilegious, or obscene, may then be used by the reader to alter his world-concept, or alter that of others. It is further worth noting that these meanings may not have been intended or predicted by the metaphor's author. Ideas, then, that did not exist prior to a reading of (11) now have the power to shift perceptions, generate new thought. This is what Ricoeur is addressing when he writes of metaphorical connections and their "power always to push the frontiers of non-sense further back" (95). Example (11) is also not necessarily the ideal, because it is an intentional construct; the variety and range of new meanings produced in the free play of a highly complex text are, due to their unexpectedness, even more likely to break new conceptual ground.

Concerning the generation of meaning out of noise, information theory again provides another perspective on the operation of literary language. Based upon the work of Henri Atlan, who first addressed the issue as a biological problem, contemporary theorists recognize noise as having a twofold character. As William Paulson relates it, the first quality of noise Atlan termed "destructive ambiguity," which is the capacity for noise to distort or damage the intended meaning of a message (CaO 40). Thus, given the three statements

(11) Paradise is an Erector Set

(12) The religious heaven is a child's toy

(13) Man may create his own ideal state

one may articulate (11) with the expectation that one's audience will interpret it to mean (12), yet it may be the message represented by (13) that is actually communicated. In this case, the polysemy of the metaphor, its noise/ambiguity, allows for the "destruction" of (12). Of course, nothing has really been "destroyed," in the conventional sense; meaning (12) remains within (11), only it is not arrived at by the hypothetical audience in this example. The reception of (13) instead of (12) is an example of the second quality of noise, which Atlan terms "autonomy-ambiguity." In the sense that the received meaning is not equivalent to the intended meaning, the former is "autonomous" from the latter. It is a new, independent meaning, not a copy of the original. Thus a polysemic text has the potential to generate unintended meanings, and new, autonomous texts. The biological origin of Atlan's work provides a useful parallel example, one suggested by Hayles. She cites the development of desirable, adaptive traits in an organism's offspring through genetic mutation as a function of biological noise (CB 56). In this scenario, the duplication of the genetic "message" encoded in an organism's DNA is affected in some way so that the DNA it passes to its offspring is not an exact copy of the original. This new, altered code may unexpectedly create an organism with traits superior to its parent. One may extrapolate from this biological example and arrive at the notion that a polysemic text could generate, unexpectedly, a meaning superior to its intended. There is also a transgressive quality to this; concerning the "Erector Set" metaphor, an author may use it with the intention of communicating a positive message about self-determination (13), but some members of his audience may read it as a jab at religion (12) and arrive at a completely different conclusion.

In this sense, the creation of new meaning has much in common with literary language's resistance to totalization: both deny the closure of language; both inject new energy into the language system by denying this closure. If the Sapir-Whorf hypothesis accurately represents the relationship between thought and language, then we further the boundaries of our expressible existence through literary language. What complex literature ultimately teaches us, then, is that the road to totality is a treacherous one, survived only by the blind. That "making sense" is an act, an unconscious tendency to artificially create closure. That everything is taken for granted. The free play of meaning resists totality, the ultimate goal of the mind.

In practice the reader is rarely left so alone; even the ideal text is only partly composed of literary language. The instrumental language in a text provides some definition, a hint at the possibility of totality, and is itself a kind of sub-totality. Again, the fifty-percent ideal encompasses all that is desirable in communication: partly fulfilled expectations combined with the radical potentialities of new information.

Reader-Text, History, Reader-Text-History

The impact that literary language can have on an individual reader and his linguistic knowledge is, as observed, very significant; yet history has not much favoured this language. Jonathan Culler has asserted that the distinction between "poetic" and "practical" language varies according to historical period (PoS 54). While admitting that certain rhetorical figures and styles have probably crossed over from one category to the other at various times, in Western culture the distinction has actually not varied much. In fact, the break between polysemy and monosemy has long existed, with the latter heavily privileged over the former. The cultural weight of Greek and Latin rationalism is an example of this privileging, which Umberto Eco observes is the origin of the modern sciences ("Interpretation" 29). The essential material of this rationalism is causality and linearity of argument, whose monosemic quality is readily evident: cause-effect relationships require a linear understanding, an inherent temporality; these cannot embrace contradiction or ambiguity. Donald Levine observes that this privileging only increased as the sciences made impressive advances in the 17th and 18th centuries (1). Nor is this appeal unique to science, which has always depended upon concreteness; Ted Cohen, citing Hobbes and Locke, asserts that "There has been a very strong line in Western philosophy, especially in that strain running from British empiricism through Vienna positivism, which has denied to metaphors and their study any philosophical seriousness whatsoever" (1).

Polysemy, on the other hand, has principally been home to the arts (sometimes) and mysticism (to this day). Indeed, Donald Levine goes so far as to identify ambiguity as a "handmaid to mysticism" (30). This mysticism opposes scientific rationalism as a world-view. Whereas with rationalism a thing is proved true if it can be explained, with mysticism something is true only if it is inexplicable, beyond human knowledge. The place of polysemy in this is to embody the coexistence of opposites, to represent the uncertainty of meaning, to deny the totality of human understanding. Mysticism must not be confused with religion, either, which is typically highly structured through ritual, pattern, and hierarchy. Though the Bible itself is an exceedingly polysemic text, and for many religions the source of their beliefs, the function of the ecclesiastical body is to limit and legitimate interpretations of that text. For example, the Catholic father Augustine, in his De Doctrina Christiana, does in fact glory in the "gift" of textual ambiguity, but simultaneously affirms the classic four-fold method of interpretation (literal, allegorical, moral, anagogical) (Eco Sem 149). He opens texts to polysemic readings while limiting the number that are acceptable. This highlights the danger of literary language to institutions; the same texts that may affirm a hegemony may be used to deny or subvert it.

Without straying too far, the iconoclasm of mysticism is parallel to the resistance of literary language to totalizing readings, and to the awareness of language's inherent polysemy that it induces. However, mysticism's connection to polysemy has greater significance; its traditional marginality within Western culture is comparable to the marginality of polysemy in that same culture. In unusual company, science and religion are thus the dominant powers, relegating art and mysticism to the role of subverter or nonessential minority. In fairness, Eco has noted that there has always been something of a dialogue between the two, and that the "qualitative" nature of the latter has often influenced the "quantitative" interests of the former ("Interpretation" 34). Yet polysemy has gained little ground because of this. Indeed, in recent centuries, certain Western societies have even more markedly resisted polysemic language; America is notably among them, as Levine points out.^[4] In America, the scientific support for monosemic language has been heavily endorsed by a general cultural pragmatics traceable to the Puritans, the nation's progenitors. The Puritan ascetic mode of living, exemplified by the characteristic plain-style sermons and preference for sincere, direct speech over courtly grace, had no place for polysemy (Levine 2).^[5] Sacvan Bercovitch describes the clerical approach to Biblical exegesis: ". . . literalism precludes personal interpretation. It serves as a wall of flame to secure the pristine Word against any snare of the intellect, all flights of the imagination" (111). Nor did the demands of frontier life, which oriented the early settlers towards dominating nature rather than indulging in it, much encourage anything other than the most direct and clear modes of communication (Levine 31). This practical orientation never disappeared from the society that it engendered, and the industrialization that followed, leading the country's rise to economic superiority, continued this tradition of heavily favoring the totalizing and the absolute.

Appropriate to capitalized nations such as the United States, there is an economics of information, and it too privileges monosemy over polysemy; one example of this has already appeared in the discussion of information theory. Recalling the information theorists' definition of multiple meaning as noise, we find that the use of the pejorative term is consistent with the economics of science, particularly engineering, in which a surplus of meaning costs productivity: time spent sorting relevant meanings from irrelevant. In the so-called "private sector," the economic ideal in business is to give customers precisely what they ask for, neither more nor less; this necessitates carefully controlling output, which in turn requires carefully controlled communication. If the output is itself information, then it is imperative that the recipient does not have to spend additional time decoding a complex message, or doubt its intended meaning. To that end, the presence of additional information not ordered and paid for, so to speak, is also giving away something for nothing.^[6] It is this quality of polysemic texts that resists capitalization; their unaccountability makes them literally unaccountable in the business sense. Considering the contemporary post-industrial information era that some say we are upon, or well into, it is less than surprising to consider issues of multiple meaning as having an economic role. Of course, as Norbert Wiener wryly observed in 1954, "The fate of information in the typically American world is to become something which can be bought or sold" (113).

Barthes' "gourmandism" (as Hayles describes it) on the other hand, is opposed to this monosemic economy, as it focuses on the possibility of unlimited rereadings of a single polysemic text (CB 188). Indeed, rereading alone, as he observes, runs counter to the contemporary consumer/throwaway culture in which reuse is viewed dubiously; the notable exception to this is recycling, but in this case consumers have only to separate their trash, and not view its reconstitution into fresh product; they see only the end result, sold as new (S/Z 15-16). Rereading, as Barthes amusingly suggests, is "tolerated only in certain marginal categories of readers (children, old people, and professors)" (16). The economics of rereading, especially as connected to a surplus of meaning, provides for a new appreciation of a classic thought experiment: if one is going to spend the rest of one's life on a desert island, what single book would one take? The commonest answers, of course, are 1) a volume containing the whole of Shakespeare, and 2) the Bible. Now, notwithstanding the heavy cultural weight carried by the former and the religious dimension of the latter, both texts are quite probably chosen because, even in the average reader's eyes, they are eminently rereadable. This rereadability is only one index of their highly polysemic nature. This example also articulates the manufacturer's worst nightmare: a consumer who is satisfied with a single good.

Hayles points out that within the discipline of literary studies there is yet another economics at play; only in this system the values are, in part, inverted (CB 189). She begins her exploration of this economy with the stellar rise of deconstruction, a movement that for some lauds polysemy, the play of meanings, the absence of closure. For centuries previous, in various forms, a canon of accepted literature for study has existed. One may trace this back to the medieval period, in which the limited number of classical texts available established what was arguably the first English canon outside of the Bible, itself the source of "canonicity." A canon is necessarily limiting, and to some this is a positive quality, an established hierarchy. Yet it also restricts the range of literary studies; the fear that one may have nothing new to write about Shakespeare is (or was) legitimate, and if one has nothing to say about Shakespeare, or some other canonized author, one is going to have a difficult time getting a job. This does not even take into consideration the frustration of participating in a dialogue in which one's role is reduced to the repetition of previous thought. Postmodern theory, then, opened up the canon, even when it did not radically alter it. For even without the addition of works written by the Other, the standard canonical works became fecund once again when postmodern/deconstructive readings (including Barthes' own S/Z) of these texts challenged old long-established interpretations. Again, this is comparable to the fecundity of medieval "commentaries" which, though principally interpretations, also permitted the introduction of new materials via already accepted canonical texts.

The Cultural Contradictions of Complex Literature

Polysemy's relationship with the literary arts is by no means agreed upon nor desired by all, especially when it results in highly complex texts; in this sense, only mysticism has never rejected literary language. In the United States particularly, there remains a preference for "realistic" poetry and prose, which tends to be more instrumental in nature. With the rise of modernism at the turn of the century, however, texts generally moved in the direction of becoming significantly more polysemic than many of their predecessors; one has only to compare the American Realists to James Joyce, T. S. Eliot, and Djuna Barnes to see this taking place. Among the notable exceptions to this rule, of course, are the proletarian writers such as Upton Sinclair and John Steinbeck, who for the purposes of propaganda had no wish to open up their texts to multiple interpretations, nor overly challenge potential readers. These modern, polysemic texts were not received quietly by the American literary public; one dramatic example of this is the so-called Bollingen controversy, which erupted in 1948 over the awarding of the Library of Congress prize to Ezra Pound for his Pisan Cantos. The initial furor concerned Pound's political activities; at the time of the award he was being held on charges of treason, having been arrested for broadcasting pro-fascist propaganda from Italy during World War II. The focus of public debate encompassed, however, heated attacks on the supposed "unintelligibility" of the poem, as Archibald MacLeish dramatizes in Poetry and Opinion. This charge of unintelligibility shared company with additional indictments of obscurantism and elitism, all of which were (and often still are) leveled at the moderns. Without oversimplifying, one may read a majority of the objections as implicit rejections of polysemy, a hallmark of many modernist works. There has been a similar public (as well as critical) reaction in recent decades to deconstruction. Both the works it is accused of furthering, like Pynchon's and Joyce's, and its own theoretical texts are often charged with being "unintelligible" or self-defeating (though this is in fact part of the game). As one of its principal tools is polysemy, it is not too much to read the negative reaction to deconstruction (as well as postmodernism, which, according to most definitions, also embraces non-linear/polysemic texts) as another negative reaction to polysemy itself.

Given that literary language resists institutionalization, there is a certain logic to the repudiation of polysemy; it threatens to disrupt the status quo, leaving nothing material in its place. It replaces the dialectical with the idiolectical. Even within the bounds of literary study, it offers no solace or support in its extreme manifestations; Finnegans Wake is a fine example of this. As Margot C. Norris has observed in a review of Wake criticism through the early seventies,

The attempt to assess the teleology of Joyce's Finnegans Wake has always

presented critics with a dilemma: the choice between a radical and a

conservative interpretation of the book. A radical interpretation would

maintain that Finnegans Wake subverts not only the literary status quo, but

the most cherished intellectual preconceptions of Western culture as well. .

. . (206)

In contrast, she asserts that the majority of Wake criticism conforms to the conservative interpretation, which reads the work as having "fixed points of reference in the manner of the traditional novel." Thus the text which, as Eco describes it, was "made to liquidate grammars and dictionaries" threatened its own discourse with liquidation, resulting in thirty years of criticism battling its own extinction (Sem 25). This case demonstrates, ironically, the power of literary language to shatter conventions, forge new ground.

There is a still greater spectre haunting complex texts, and that is the charge of elitism; it sometimes manifests itself in terms of of the work being "undemocratic." The latter has a comic half-truth to it; as literary language is anti-social (a la Barthes), it is certainly undemocratic. The challenges, however, stem from an opposition to the difficulties inherent in reading literary texts. It seems that readers are rebuffed by texts that refuse to nicely concede totality, relevance, and closure; they perhaps assume that texts that so immediately deny them are therefore asserting some kind of knowing superiority. Even some otherwise-enlightened critics conclude the same; Lance Olsen laments that while postmodernity was supposed to have democratized art, it "simply replaced one sort of elitism with another," in this case with techniques that make the work "inaccessible" (120). He cites Gravity's Rainbow and Mulligan Stew as counterparts to the "elite modernist" The Waste Land and Ulysses. Without addressing these particular works, it must be asserted that the charge of elitism does not altogether hold. Though it is true that some literary texts require greater literary competence to process adequately, polysemy alone requires no additional literacy skills. A highly polysemic text is as accessible to the public as a maximally monosemic one; the only difference between the two is the effort demanded by the former. Recalling the random and highly polysemic "statement"

(8) Square even wish events the

and providing two others,

(14) I would like to attend the events in the square

(15) For a long while Abel considered a Paracelsian receipt for his

madness^[7]

it is immediately clear as to the relative accessibility of polysemic language. Statement (14) is highly monosemic, and requires only contextualization to identify the "events" and the "square" in which they will take place; even without the context, the meaning is self-evident. Also fairly monosemic is (15), except in this case Biblical literacy is necessary to resolve any meaning out of the statement. In contrast, (8) is highly polysemic and requires considerable effort to work out the relationships between the words, to make connections and establish meanings; however, it does not require any more preparation to interpret it than (14) does. Thus a reader that can process (8) and (14) may be incapable of understanding (15), which requires an additional competency.

It is beyond the scope of this study to address in any meaningful fashion the causes of these particular misperceptions/prejudices; one might speculate that their origins are in a confidence in the present institutions and a desire not to question them; in an anti-intellectualism, especially virulent in the United States; in the pragmatics necessitated by the quotidian on one hand, and by the economic on the other. It is, however, within bounds to assert that the marginalization of polysemy and the high privileging of monosemy is unhealthy for the language system as a whole. Whether we agree with Gregory Bateson, who coldly observes that "mere purposive rationality . . . is necessarily pathogenic and destructive of life," or with Philip Wheelwright, who concludes that the metaphor is key to preventing a fall "from the ambiguous grace of being human into the unsignative security of the reacting mechanism," it is certain that literary language has a powerful role to play in human communication (146, 123). The effect it can have on a reader should not be limited to those willing to cross some sociocultural boundary.

It is worth observing that the rejection of complex literature does not permeate all of Western culture; as Susan Sontag notes, this is much more an Anglo-American trait. She describes French literary culture at mid-century, contemporaneous with the Pound controversy:

Such "difficult" literary tendencies as Symbolism and Surrealism, and in

particular the line of post-novel prose narrative from the Surrealist fictions

to those of Borges, Beckett, and Robbe-Grillet, are taken to occupy the

central position in contemporary letters -- while most novels in traditional

"realistic" forms (such as continue to this day to be critical successes in

England and America) are regarded as essentially uninteresting, barely

noteworthy products of a retarded or reactionary consciousness. (ix)

This perspective provides an interesting counterpoint to the efforts the French have made in recent years to insulate their language from borrowings and outside influences.

Literary Language and Contemporary Literature

Given the significant role that complex literature has to play within the language system, it is unfortunate that it occupies a marginal place in mass culture; the popularity of genre fiction, which is principally monosemic, has always overshadowed literary works. Even "literature," as the term is generally taken, has shifted its attention away from postmodern fictions to conservative, neorealist works in the last decade. This is the heyday of Bobbie Ann Mason, Frederick Barthleme, and Ann Beattie, whose works, however progressive their vision, are essentially realistic. A great deal of this marginalization of polysemic fiction is due to a common notion, certainly dating back to the Victorians, that literature is principally a medium of entertainment; long gone is the classical ideal that texts should both "delight" and "teach." Thus we have the phrases "a good book" and "beach reading," as well as the innumerable adaptations to the screen, that most entertainment-driven medium. Complex texts, on the other hand, are generally not "entertaining" in the same fashion as realistic, monosemic texts are, if only because they require much more effort to process. Nor do they lend themselves to the beach or the screen, given their polystructural or astructural nature. Indeed, some of the most highly complex texts put such a demand on the reader that time becomes a factor; however overstated the case may be, the demands made by (corporate) society of the individual have increased in the last few decades.

Contemporary reader-response theory emphasizes the role of the reader in creating meaning; however, it does not often acknowledge that the potentially creative act of reading is work, something that may not appeal to the average reader. An instrumental text, using language as a tool, communicates much more directly than a polysemic text, and popular literature, traced back to its origins in orality, is the direct communication of stories. Laura Miller, writing in the middlebrow New York Times Book Review, takes the practitioners of hyperfiction to task and asserts that ". . . story is fiction's trump card. People who read for nothing else will read for plot . . ." (43). She continues, "Readers like me stubbornly resist hyperfiction's efforts to free them from what [Robert] Coover calls 'domination by the author.'" The readers she describes find pleasure in submitting to the ideology of the text; their pleasure is not Barthes'.

What may be needed, then, is a new perception of literature in conjunction with the new definition of literature proposed here; an acceptance, perhaps, that the reading of complex texts is an important cognitive and social function; a conclusion that maybe, just maybe, Pynchon is Good For You.