I

In chapter 1 of this study I outlined some criteria by which I anticipated that my discussion of syntactic stylistics would, explicitly or implicitly, be judged. As an analytical description of a certain set of data, I suggested, the model I advanced would have to be clearly articulated and unambiguous in its claims and predictions. I would have to make every effort to apply it directly to relevant literary examples. Finally, I should also seek to examine its implications outside the field of stylistics as such. The first two of these three announced goals I would claim already to have attained. Beginning in chapter 2, I sketched, sometimes in fine detail but more often with broad strokes, my theory of syntactic stylistics, a theory that I believe can clarify and enrich debate in both critical and linguistic studies. In the process I also undertook a variety of textual analyses illustrating my proposals. It remains now only to review that theoretical model as a whole and to assess its applicability to disciplines less integrally connected with the academic study of literary works.

As Figure 23 indicates, I claim that stylistic generalizations fall into three broad classes. Technical commentary allies itself closely with contemporary linguistics, borrowing both its terminology and its analytical principles from the most widely accepted current versions of syntactic theory. To linguistics it contributes in its turn formal accounts of significant nonstandard literary “dialects.” As just one of the three subdisciplines embraced by syntactic stylistics, technical analysis stands out as an undeniably scientific pursuit; the stylist’s technical claims represent precise hypotheses with empirically testable consequences. (As we shall shortly see, the same cannot be said so confidently, if at all, for the other branches of the discipline, nor, therefore, for syntactic stylistics as a whole.)

FIG. 23

Orthodox syntactic theory includes no aesthetic component. Any stylistic judgments, therefore, that draw on aesthetic concepts must by definition belong in some category other than the technical. Perceptual stylistic analysis, as I have called it, arises from just such an application of general aesthetic principles to linguistic material. To the extent that the study of aesthetics itself can be systematized, perceptual stylistics will approach technical stylistics in its rigor and theoretical precision. Its raw materials, however, will never overlap with those designated for technical description, since notions such as “balance,” “antithesis,” and “delay” have no role to play in that aspect of language captured by traditional syntactic theories.

Interpretive stylistics differs fundamentally from both technical and perceptual analysis. In this subfield, the stylist draws on either of the other two subfields (or on them both) to provide one side of a delicate balance, as he links syntactic form, however tenuously, with what he understands of poetic meaning. It is ironic, of course, that this treacherous domain of interpretation should represent both the goal of most stylists’ work that carries the greatest popular appeal and that area in which they venture furthest from their roots, furthest from the relatively “safe” territory of technical description. Yet in this respect stylists share their temerity with critics of many other persuasions—biographers, historians, and psychoanalysts, for example, all of whom eventually find themselves forced to lay aside their respective specializations and to make unambiguous critical claims about a specific literary text. To the extent that stylistics rests at least one leg of its argument on linguistic bedrock, on fundamental claims about a central aspect of every text’s form—the common core of meaning contained in its language—to that extent at least the interpretive stylist enjoys as good a chance as anyone of promoting for himself and for a wider audience his critical insights into a given work.

As I have summarized it here, and as I developed it throughout the first four chapters of this study, my theory of syntactic stylistics is just that: an analytical description of an intellectual discipline. I have attempted, that is, to examine what stylists do and to discover how they do it. Keeping in mind the distinctions I have drawn among types of stylistic statement should make it easier for readers to determine whether particular stylists have compatible or incompatible views on a particular passage, where those views diverge, and what kinds of evidence might decide between them. Conversely, my model should also help to isolate “nonarguments” in stylistics, debates in which the parties involved mistakenly believe themselves to be at odds while their analyses actually differ so radically in assumptions and materials that rational discussion is effectively impossible. Finally, I hope that my enumeration of areas of legitimate stylistic comment may help to prompt textual analyses where none were considered before, analyses which may lead us to either extend or restrict the specific taxonomy that I have proposed. All of these potential developments I see as proper and indeed desirable consequences of my proposals.

At a number of points earlier in this study, though, I was careful to point with concern to other possible interpretations or applications of my theory which I viewed as illegitimate. Among the most potentially damaging, I would number, first, the assumption that this theory might be either a theory of reading or an account of the process of poetic composition, and, second, the completely independent suspicion that the whole activity might conceal a pernicious system for evaluating the merit of a given work of literature. It is important now to sweep these two theoretical minefields with proper care.

In no sense, first of all, have I attempted here to characterize the process by which anyone actually reads a work of literature.¹ My concern has been with justifying an interpretation rather than with reaching one, with intellectual evaluation rather than with heuristic experimentation. Nor is this assertion, which I make with some insistence, in any way compromised by the fact that I have frequently made reference to “the reader” in the course of my analyses.

W. Daniel Wilson recently deplored the “unruly profuseness,” the “tangled mass” of epithets currently applied by critics to the once innocent “reader,” the “jargonistic neologisms” of what has come to be called the “reader-response school” of critical theory.² In a wide-ranging study, he reviewed the content, whether stated or only implied, of such terminological monstrosities as the “narratee” (Gerald Prince), the “optimum reader” (Lowry Nelson, Jr.), and the “addressee” (Manfred Naumann). Fortunately for the uninitiated, Wilson’s final conclusion is that, sensitively interpreted, just three terms adequately characterize the only important and theoretically distinguishable figures postulated by reader-response theorists. “Characterized fictive readers” exist entirely within the text, as for example do the thoroughly exaggerated figures with whom Tristram Shandy carries on a constant dialog in Laurence Sterne’s novel of the same name. “Implied fictive readers,” though no less creatures of the text, cannot be used ironically or satirically in the same way. Intimately linked to “our overall interpretation” of a given work, they embody “the attitudes and judgments demanded of the real reader by the text.”³ Critics generally agree, for example, that one capacity essential to a fully adequate reading of Tristram Shandy is the capacity to laugh with Sterne himself at the more absurd of the “characterized readers” he presents in the course of his rambling narration. Insofar as this ability is actually demanded of the reader of this novel, it will define in one important respect the work’s “implied readership.”

“Both these fictive readers,” fictive in the straightforward sense that they exist solely by virtue of the fictional texts to which they are uniquely relevant, “must be distinguished from real readers,” Wilson then goes on to claim, “. . . even if real readers actualize the implied reader’s role after the work is created.”⁴ Wilson’s theory thus leaves us with just three kinds of “reader”: the characterized, the implied, and the real. Alas, wherever in the preceding pages I have used the term “the reader,” I have been referring to none of these theoretical entities, but to a fourth, a figure that I shall tentatively call the “native reader.”⁵

As the name is supposed to suggest, the native reader bears much the same relation to stylistic theory that the “native speaker/ hearer” has traditionally borne to transformational linguistic theory and rests on a series of abstractions and simplifications of individual events inescapable if one wishes stylistic theory to achieve a high level of what Noam Chomsky would term “descriptive power.” Major advances in linguistics became possible after Chomsky advocated temporarily putting to one side the attested linguistic behavior of individual informants and taking as data instead what all speakers of a given language seemed to agree to be the intuitive core of their linguistic knowledge. My introduction of the native reader depends on a precisely analogous conscious idealization in stylistic studies.⁶ In itself this step has been justified, I would maintain, by the critical and technical insights that this study has been able to achieve. Where such practical advantages do not seem to accrue from positing my particular brand of native reader, however, the assumptions that the term encodes are easily undone or held in abeyance.

As a necessary condition for the stylistic theory that it enables, but also paradoxically as a creature of that theory, the concept of the native reader involves specific idealizations in each of the areas addressed by my analytical model. Since technical stylistics depends heavily on linguistic theory, as has been argued several times, a rather straightforward assumption for us to make is that one component, so to speak, of my native reader will in fact be the linguist’s native speaker/hearer. Idealization in the technical sphere is thus virtually defined a priori. To complete a fully adequate characterization of the native reader, though, would also require parallel abstractions in the perceptual and even the interpretive subfields, where a consensus on what material should be regarded as idiosyncratic and on methods for achieving appropriate generalizations is far less likely to be forthcoming.

Fortunately, if the general logic underlying my development of a native readership is accepted, then it may prove unnecessary ever to formulate such criteria precisely; a “complete characterization” may be exactly what is not called for. For, as I argued in chapter 1, to isolate the native reader from the specific stylistic case that he “feeds” immediately implies for him an extratheoretical reality that endangers his very usefulness. I find myself increasingly convinced, therefore, that the generalizations one makes about “the reader” in the course of a stylistic analysis are ultimately justified en bloc by the critical insights they promote. Certainly, losses may be incurred as a result of this process of idealization, and Stanley Fish and Barbara Herrnstein Smith are right to demand that we be alert for such “slippage.” But as Paul Armstrong notes, this problem is not unique to stylistics:

[Every] hermeneutic standpoint has its own dialectic of blindness and insight—a ratio of disguise and disclosure that stems from its presuppositions. To accept a method of interpretation is to . . . gamble . . . that the insight its assumptions make possible will offset the risk of blindness.⁷

Detailed caviling about excessively (or inadequately) sophisticated claims that some stylist may have made for the reader he assumed when constructing his argument thus misses the mark unless the caviler also addresses what Armstrong calls the criterion of “efficacy,”⁸ the overall gains or losses that have resulted for our understanding of the text under discussion or for our grasp of the role of language in literature as a whole.

It is of course precisely this flexibility, which I view as inherent in this field of scholarship, that leads me to believe that the procedural obstacles that Fish creates for himself and for stylistics as a discipline in his more recent essays are mere mirages. Fish argues himself into a position where he can no longer live with idealized abstractions (although he used them repeatedly in his earlier work with most impressive results). The “interpretive community,” as I noted in chapter 1, then represents Fish’s belated attempt to resurrect on a different plane that idealization on which his best critical work always depended. He tries to remove his idealized reader from the realm of stylistic theory altogether, and to base that construct instead on sociocultural and general linguistic criteria. Inevitably, this attempt fails since all language “communities” fragment finally into individual speakers, our concepts of “dialect” and “language” themselves resting on pre-theoretical abstractions designed for the sole purpose of facilitating linguistic investigation. The interpretive community, analogously, can exist only as part of a stylistic theory, not independently of it. The end justifying the means, to put it bluntly, is what Fish finds unpalatable, and in the final analysis one has simply to decide whether Fish’s scruples in this area are one’s own.

An altogether different consideration separates my native reader from the reader implied by George Dillon in his latest work. In the preceding chapters, I offered an account of what I might call the native reader’s “stylistic competence.” Dillon, by his own admission, details instead a theory of stylistics based on performance factors—heuristics, pragmatics, and strategies. His theory and my own are thus not incompatible; both assume an idealized reader whose stylistic capabilities can be described or modeled along linguistic lines.

Both may also, in passing, improve our understanding of (though neither of them sets out directly to describe) what Wilson would call the “real reader.” To cite just a single example of how, Fish notwithstanding, this may be so, recall that in the closing paragraphs of chapter 2 I promised further discussion of the role played by nonstandard coordinate structure deletions in Pope’s poetry. That promise followed hard on the heels of an extended theoretical argument. I had proposed specifically that an adequate technical account of Augustan poetic syntax would have to include some statement to the effect that all coordinate structure deletion rules apply freely (that is, in either direction) in texts of this period, uncontrolled by the constraints that affect such processes in standard modern English. This theoretical hypothesis, I argued, easily won out over reliance on a series of other (individually no more radical) analyses that would each explain only a subset of the attested cases of syntactic irregularity. For it permitted the stylist to capture a major generalization by uniting several phenomena that must otherwise be assumed to appear together in Augustan poetry by a most extreme coincidence.

My own informal observations, which several audiences at professional meetings have not found cause to dispute, suggest that we may derive some circumstantial evidence in favor of this purely theoretical technical argument from the literature classroom.⁹ When first introduced to Augustan poetry, most students encounter some difficulty with its syntax. They claim “not to understand” passages whose vocabulary and overall interpretation are relatively straightforward; yet when one disentangles for them the poetic inversions and deletions with which the text is riddled, they routinely admit that the whole is now perfectly clear. What is striking, however, is not merely the fact that syntax should provide the major stumbling block to these students, but the manner in which, in due course, they overcome their hesitancy. Rather than comprehending one line, or even one construction at a time, that is, they almost always struggle along for some time, then suddenly find themselves able to interpret all of the text equally well.

If we assume for a moment that this “informal observation” on my part could be adequately corroborated—a feat which, to my mind, would raise the assessment of “reader-response” to new heights—then the following conclusion would flow from it rather naturally. The available pedagogical evidence would be fully compatible with my own stylistic hypothesis about coordinate structure deletions in Pope’s poetry. My technical analysis requires that only one very general statement predict a rather large variety of nonstandard constructions. If this happens also to be the correct way to look at things practically, then we might expect that for each student mastery of one such construction would entail mastery of all, creating a sort of “threshold” effect in terms of reading strategies of the kind discussed by Dillon in his recent book. Conversely, while my observations of classroom behavior do not absolutely conflict with the alternative, multiple-rule theoretical hypothesis, accommodating those two bodies of data to one another would seem to force us to assume that several disparate syntactic processes just happen to be learned simultaneously by inexperienced readers—a coincidence that would compound the coincidence already noted from a purely theoretical standpoint above.¹⁰

Let me note in closing that stylistics would certainly benefit from many more such technical analyses and from correlations of those analyses with teachers’ problems and teachers’ solutions to those problems. I see every reason to hope that, with enough of this detailed information, stylistics as a discipline may help to improve the teaching of literary comprehension. A full discussion of this and of the many other potential applications of my stylistic theory in “the real world” deserves, however, far more extensive treatment than I can afford it here.

The implications of my theory of stylistics for our assessment of the way in which readers read constitute fertile soil for debate. While mine is not a theory of reading, one might summarize, it certainly has some points to make about reading. It requires little ingenuity, then, to estimate from the preceding discussion my opinion of how far stylistic theory should venture in making statements about how poets write. The stylist, I would claim, need make little (or even no) reference to the poet in assembling his critical case. Such references as are made will generally be matters of critical cliche, and will be made with due regard for the danger of attributing too definite a profile to a specific historical figure on the basis of stylistic data alone. A stylistic argument that sets forth truly original and substantive claims about the psyche of the poet whose works it discusses should be viewed with some caution.

Not that interesting observations can never be made in this general area. Richard Ohmann’s Shaw: The Style and the Man certainly indulges in some psychological speculation, but Ohmann avoids major problems by openly coupling a stylistic analysis of Shaw’s prose with an independently validated psychoanalytical study of Shaw the man. My own analysis in an early paper of Wordsworth’s textual revisions in The Ruined Cottage similarly toyed with psychoanalytical reconstruction.¹¹ My comments too, however, involved corroborating stylistically some general trends in Wordsworth’s philosophical thought that are almost universally accepted. What the stylist must avoid is emerging as a closet analyst or closet biographer, proposing his own character sketches or disputing those of others solely on the basis of his linguistically grounded observations. This risk can be avoided, providing that, as always, we delimit clearly the legitimate scope of stylistic commentary and draw a precise line setting stylistic judgments apart from the psychobiographical information with which we may later associate them. Only within such a framework will the poet assume the correct degree of prominence in stylistic theory as an intriguing but by no means integral aspect of the stylist’s work.

The third and last of my caveats against potential abuses of my stylistic theory concerns the question of evaluative assessment. As I bremarked early in chapter 4, E. L. Epstein, in his essay “The Self-Reflexive Artefact,” sets about integrating a single stylistic concept (mimesis) into a comprehensive overview of the most highly rated literary works from different historical periods. He hypothesizes that mimesis represents the interpretive function of poetic syntax most favored by readers and critics alike during the past two centuries, a conclusion with which, in such very general terms, I would probably want to concur. It is vital to bear in mind, though, that Epstein’s is an essentially observational hypothesis, which describes as accurately as possible a historical state of affairs rather than prescribing some absolute, right-minded criterion of excellence for current or future readers. In terms of its impact on the casual reader, his argument still suffers from two serious practical drawbacks. First, Epstein assumes that we have in place a stylistic model sufficiently sensitive to be relied on when gathering the stylistic data so crucial to his correlation of them with evaluative norms. In the event, my own foray into Dryden’s rhetorical style has shown all too clearly the inadequacies of our current understanding of interpretive strategies in general. We cannot, then, be sure that the works Epstein points to as critically preferred do in fact share the common basis in interpretive effect that he claims for them. Second, Epstein also accepts too easily, I think, the possibility of separating readers’ stylistic evaluation of a text from their assessment of its genre, its philosophy, or even its politics. In short, we need a far more sophisticated stylistic theory than is currently available, not to mention a clearer understanding of the notion of the literary “canon,” before the link between style and literary value can be authoritatively established.

Above all, it must be reemphasized that no aspect of the theory presented in these pages translates straightforwardly into an algorithm for plotting the worth of a given work. Mimesis does not outrank iconicity on some cosmic scale of values; a text that exhibits both perceptual and technical features worthy of analysis is not thereby more noteworthy than one that is, say, technically unexceptional. We are simply a very long way from appreciating fully the role that syntactic style plays in the apportionment of general critical acclaim—a process that must be, to say the least, exceedingly complex.

II

The contentious state of modern critical theory may be blamed for the rather negative tone and defensive thinking of this final chapter. These same features, however, cannot be allowed to prevail to the very end of the study. Stylistic analysis is emerging from a period of some disarray; it remains in need of considerable repair. But it also offers great hope for a positive and productive future. Like its fellow-traveler, linguistic theory, stylistics finds itself at a crossroads, exhausted by the rapid progress of the sixties and the seventies yet uncertain of where the eighties will lead. This air of indecision by no means entails that no road forward can be discerned; on the contrary, the discipline simply needs to reorient itself, to reexamine its bases and goals, and to gather its energies before setting forward once again. This study has attempted to begin such a revitalization. By reviewing the arguments of some of stylistics’ major detractors, I have shown the comparative insubstantiality of their reservations; by analyzing many different passages of poetry, I have suggested a series of avenues for further investigation. But the real excitement still lies ahead, in the work yet to be done developing what must surely become a comprehensive and critically invaluable theory of syntactic stylistics.