I

As his father’s ghost haunts the unlucky Hamlet, so the grisly spectre of mechanistic interpretation clearly stalks through the waking nightmares of many contemporary critical theorists, who seek to exorcise the demon by writing lengthy attacks on stylistic criticism. As is often the case with such obsessions, the potential dangers that these scholars point to represent wholly reasonable fears. I would myself heartily agree that no lover of poetry should let pass unchallenged any suggestion that the meaning of a poetic text might be discoverable merely by plugging appropriate values into a differential equation of linguistic variables. Such eminent good sense, however, verges on irrational paranoia whenever these opponents of the stylistic approach aver that the worst has in fact already happened—that stylists are actively claiming for their method that it is some kind of critical Rosetta Stone that can decipher for us, as if by magic, the meanings encoded in any literary text.

Take, for example, Barbara Herrnstein Smith’s long tirade against the papers that appear in the first part of Roger Fowler’s anthology, Style and Structure in Literature. Each of those papers, Smith asserts, develops a slightly different version of the same basic theoretical “formula,” which she summarizes as follows:

[S]omething in a literary work that is . . . manifest or “surface,” S, bears some relation, R, to something else that is . . . obscure . . . or “deep,” X; therefore, by analyzing S, one may discover X.¹

As a basis for the sweeping indictment of stylistics that follows it, this characterization labors under at least one serious misunderstanding; it assumes the primary goal of all stylistic analyses to be the “discovery” of a text’s “deep” meaning. In Smith’s view, the stylist considers both the analysis of S and the determination of R merely intermediate steps on the path toward an altogether more important end, the revelation of the mysterious, till-then-concealed X. This subtly distorted view of stylistic criticism did not originate with Smith, however. As early as 1973, and primarily in reaction to some of Roman Jakobson’s more extreme pronouncements, Stanley Fish had begun accusing stylists of responding primarily if not exclusively to “the promise of an automatic interpretive procedure.”²

In the paragraphs that follow, I offer a radically different, and I think more accurate, perspective on the stylistic enterprise. Stylistic criticism, I suggest, directs the spotlight not at X, the meaning of a text under scrutiny, but at R, the term in Smith’s formula that represents the relationship obtaining between S and X. What principally fascinates stylists, then, is the “how” rather than the “what” of poetic expression, the “maze” that Irene Fairley suggests lies “between the poem, its configurations of words and sentences, and the reader’s interpretation.”³ This focus, which, I assert, characterizes the best recent work in stylistics, has been described by Jonathan Culler as “an important reversal of perspective” in critical method precisely because it “[assigns] a secondary place to the interpretation of individual texts.”⁴ Its importance for stylistic criticism has also been aggressively championed by scholars such as Donald Freeman, Geoffrey Leech and Michael Short, and Muffy Siegel.⁵ The point still bears reiteration here, however, in view of the frequency with which its simple intent has been misunderstood or avoided by unsympathetic commentators.

Ironically, it is Smith herself who notices one rather telling indication of stylists’ preoccupation with R in the very papers she attacks, though she fails to recognize its full significance. In the course of her discussion, she berates stylists for permitting “R, the relational term,” to become “a word or phrase of exceptional vagueness. For example, S ‘might suggest’ X . . . S ‘embraces’ X.”⁶ Her remarks again echo those of Fish, who had complained bitterly about

statements like this; the verbal patterns “reflect” the subject matter, are “congruent” with it, “express” it, “embody” it, “encode” it, and at one point even “enshrine” it.⁷

But neither critic acknowledges the most obvious implication of the plethora of descriptive terms that both so perceptively detect. The proliferation of theoretical terminology within a single component of any theoretical model surely betrays an area of major emphasis and growth for that theory. Stylists’ persistent gropings for words adequate to describe the many relations they uncover between the linguistic forms and the general interpretations of literary texts do not, that is, constitute lapses into “exceptional vagueness”; on the contrary, they signal these authors’ awareness of the wide variety of ways in which verbal form may affect meaning, and, simultaneously, their frustration at the lack of nonmetaphorical language with which to discriminate among them. The “relational term,” seen as an object for study in itself rather than as a trivial preliminary to the act of deducing an interpretation, forms the very core of the stylistic approach to literature.

Fowler himself proposes just such an emphasis on the techniques by which poets capture meanings in words when he insists that

there cannot be a ‘linguistic criticism’ in the naive sense . . . [of a] mechanical discovery procedure for poetic structure. . . . An urgent priority for contemporary stylistics is [instead] to determine . . . how . . . systems of literary knowledge are coded in the structure of language.⁸

And the same priorities inform E. L. Epstein’s admirable attempt earlier in the same volume to develop a somewhat more specific and practical agenda for interpretive stylistics in his essay “The Self-Reflexive Artefact.” In that discussion, Epstein specifies as his goal

to observe the relationship or lack of relationship between the principles of selection and arrangement operating on the elements of form to see whether they are determined or not determined by the particular state of affairs conveyed by the lexical constellation chosen.⁹

One may, with Smith, regret Epstein’s convoluted exposition in this passage, but his message at least is a simple one. He announces that his paper will aim to improve our working knowledge of how widely disparate, vastly complex human thoughts and emotions (“states of affairs”) are actually “realized in linear and segmental form syntactically.”¹⁰ The method Epstein adopts is to refine and extend a relatively simple classification of such methods of “realization” contained in Donald Davie’s brilliant essay, Articulate Energy, and then to relate the resulting taxonomy to what he considers a dominant trend in our methods of evaluating “Renaissance and post-Renaissance poetry from . . . Western Europe and America.”¹¹ The resulting discussion I find stimulating, informative, and classically “stylistic”; yet it has little if anything to do with the provision of explicit readings for particular texts, a point that seems to escape Smith as she launches her attack on the theoretically far less crucial second half of Epstein’s paper.

This chapter, then, will continue the important task, begun by Davie and Epstein, of characterizing and categorizing the wide variety of relations that may from time to time obtain between technical and perceptual aspects of poetic syntax, on the one hand, and the meanings of the poetic texts in which they occur, on the other. That I shall term all such relations “interpretive” should certainly not be taken to imply, however, that a text is to be interpreted by relying upon them. The interpretive subfield of stylistics within my theoretical model constitutes quite simply the domain within which linguistic and aesthetic statements about the text of some poem may be compared, contrasted, or otherwise correlated with critical assessments of its meaning. To put this point another way, the actual process of achieving for oneself a satisfactory interpretation of that text is wholly independent of, and will generally precede (theoretically if not practically), even the earliest stages of an interpretive stylistic discussion of its technique.

I am tempted at this point to move directly to the enumeration of specific interpretive strategies announced in the preceding paragraph. Were I to do so, however, it might wrongly be inferred that I naively believed that stylists never concerned themselves with what I just referred to as the “process of . . . interpretation” or with the particular readings that evolve from that process. Such is clearly not the case, and it is not hard to reconstruct the logic that allows stylists to associate “interpretive stylistics” as I have strictly defined it above with “interpretation” as it is widely understood by pre-Post-Structuralist critical theorists. “As stylistic theory gradually improves our comprehension of how verbal forms habitually correlate with literary meaning,” that argument runs, “critics will be able to rely with increasing confidence on syntactic evidence as they formulate and defend their interpretive hypotheses.”

Inevitably, this reasoning runs afoul of Fish, who positively derides the attempts of stylists such as J. P. Thorne to “match up grammatical structures with the effects they invariably produce.”¹² Fish has, for once, read aright the general purpose of studies such as the one he is criticizing—to “match up” (or correlate) structures with effects rather than to unearth meanings by way of structural analysis. But this useful insight is marred when his old skepticism reasserts itself in a subtly new form. In the sternly inflexible adverb “invariably” that Fish interpolates into his paraphrase of Thorne’s analytical goal, we may detect again the worrisome presence of a familiar gaunt and ghostly figure; Fish’s ever-present fear of mechanistic interpretation, to be precise, has driven him to overreact once more to a menace that has never actually appeared in the flesh. From what he has read, Fish has inferred that stylists regard their own stylistic evidence as wholly irrefutable, as interpretive “money in the bank.”¹³ This inference (which specifically denies, of course, the spirit of the adverb habitually that I chose when formulating my own paraphrase of the stylist’s manifesto above) I believe to be fundamentally erroneous. In the real world, it seems to me, most stylistic critics remain fully sensitive to the enormous complexity of the interpretive process. Many would probably agree with E. D. Hirsch, who argues that in its earliest stages, far from being linguistically or in any other important sense “invariable,” the whole complex procedure cannot be reduced to any “systematic structure” at all, “because there is no way of compelling a right guess by means of rules and principles.” Only much later, in fact, long after an intuitive reading of the text has already been developed by some indefinable heuristic strategy or group of strategies, does “the systematic side of interpretation begin,” as the critic attempts to distinguish more valid individual interpretations from less valid ones.¹⁴

Within such a comparatively orthodox theory of literary criticism, stylists’ attempts to rebut Fish’s charge of mechanism should take the form of responses to questions of the following kind: What constitutes bona fide stylistic evidence in the second, “systematic” stage of the interpretive process (the stage which Hirsch calls “the job of validation”)? And what weight is to be attached to that evidence (in particular, does stylistic evidence enjoy some kind of privileged, “invariant” status)? To what extent, in other words, is stylistic evidence vulnerable to overthrow in light of other, nonstylistic critical testimony?

In his introductory essay to Style and Structure in Literature, Fowler considers precisely this issue when assessing how papers such as Epstein’s that rely on stylistic methods in pursuing critical goals will be read by the general public. He finds it particularly praiseworthy that “it is natural to argue with [that] paper according to the normal canons of critical discourse.”¹⁵ The “significance of the literary phenomena noticed” in that and other such papers, Fowler suggests, will be compared on equal terms with insights gained using more traditional critical methods, a basis for evaluation with which, as a stylist himself, Fowler appears perfectly content. Elsewhere in the same volume, Freeman offers a more detailed scenario, but one wholly consistent with Fowler’s egalitarian philosophy:

[A]nalysis of a poet’s . . . manipulations of. . . syntactic processes can lay bare the deep form of particular poems—the form controlling metaphor, theme, tone, imagery, and diction.¹⁶

Where one “deep form” determines (“controls”) such diverse features of a literary text as its syntax, tone, and imagery, Freeman surely implies, analysis of any one “surface” feature stands as good a chance of revealing that form as analysis of any other.

In ideal instances of course, as Freeman himself points out, the stylist’s conclusions will turn out to “parallel and reinforce” arguments based on the study of “other strategies of design—metaphor, rhetoric, even meter.”¹⁷ To these kinds of internally motivated arguments, furthermore, one might very well add arguments whose bases lie outside the text (narrowly defined) in its biographical, historical, or sociocultural background. But where agreement and harmony are the ideals, occasional disputes and discord are inevitable. In such cases, I would strongly contend, stylistic critics seek no preferential treatment, acknowledging that an “invariant” relation between syntactic process or syntactic structure and literary interpretation is neither probable nor desirable. They fully anticipate, in fact, that their critical hypotheses will sink or swim as each fails or succeeds in the fundamental critical task of correctly characterizing readers’ informed responses to the work it purports to explain.¹⁸

I do not believe, in summation, that most stylists would wish to dispute the claim which Smith reiterates so insistently:

[A]lthough “the verbal structure” of the poem may direct one’s experience and interpretation of it, that structure cannot “control,” in the sense of unequivocally determine, either of them.¹⁹

But “unequivocal control” exaggerates stylists’ views of the role of syntax in poetic expression just as unfairly as Fish’s notion of “invariance.” (Irene Fairley, to cite only one example, chooses instead the phrase “stimulates and guides” when she attempts to describe the subtle effects that a text’s language has on its readers.²⁰) In the end, therefore, it could be said to be almost irrelevant whether or not one agrees with Smith that “a significant aspect of meaning in the poem . . . is necessarily variable, irreducibly indeterminate”²¹ so long as stylists actually lay claim only to explicating other parallel or perhaps intersecting “aspects” of that same overall “meaning.”

Despite considerable provocation (based often on misreadings or careless readings of their work), stylists have generally succeeded quite well, I submit, in moderating their claims, alleging only that we must allow to the language of any work of literature a measure of influence on the way that it is interpreted—a simple creed that may be found expressed in the italicized sentences with which I opened chapter 1 of this study. Perhaps the preceding discussion and the examples that follow will do something to convince all parties to this dispute of the remarkably small extent to which committed stylists and supporters of Fish and Smith need actually disagree on this important issue.

One final remark. Given the delicately balanced set of methodological assumptions that I have now attributed to stylists, it may often be hard to determine by which of two complementary routes a given stylistic assessment of a literary text actually evolved—even for the analyst himself. Was it his initial achievement of a valid critical insight into poem X that provided one anchor, so to speak, for the interpretive stylistic judgment that he went on to make about it? Or did his detection of a familiar stylistic technique affecting the syntactic form of a passage prompt in its turn a particular reading of the lines in question? Such hermeneutic conundrums, though, are characteristic of the critical process in general and by no means unique to the stylistic approach.²² We may never know which den produced the possum to whose capture we suddenly find ourselves committed (extending for a moment Hirsch’s metaphor from the epigraph to this chapter); but the possum itself is no less substantial for that, and no less worthy a quarry for our hunt.

II

Having thus completed our obligatory detour to consider the complexities of the interpretive process, we may return at last to the central topic of this chapter, the cataloguing of interpretive stylistic effects, narrowly construed. As a first step, let us recognize that many of the technical and perceptual phenomena described in earlier chapters of this study can by all means feature in a text without having any clear effect on, or correspondence with, its overall interpretation. The importance of this simple observation to a fair evaluation of stylistic theory generally will justify, I think, the following detailed consideration of one particular case in point.

In chapter 2 of this study, we examined rather thoroughly Shelley’s syntactically anarchic style in “Adonais.” I shall want to return to that material again shortly. But the technical feats that Shelley performed in that work he had first learned several years earlier, for we find him invoking precisely the same strategies freely in Laon and Cythna (the poem later published, only very slightly revised, as The Revolt of Islam). I have developed elsewhere a rather lengthy technical analysis of these syntactic aberrations in Laon and Cythna.²³ At this point, therefore, I shall rely on only two rather brief analyses of particularly intriguing passages as pro forma support for my basic contention: that this work abounds in constraint-violating syntactic brain-teasers.

The first of my two illustrations represents a striking example, indeed a “textbook” case, of syntactic center-embedding, in which both an adverbial subordinate clause and the relative clause that it dominates are bracketed by lexical material that belongs to their respective matrix sentences:

From that lone ruin, when the steed that panted

Paused, might be heard the murmur of the motion

Of waters,

(CWS I, 334: 200–202)

The most deeply embedded clause in this construction, [S3 that panted], is properly contained within the intermediate clause, [S2 when the steed . . . paused]. S2 itself, however, appears between the bulk of the main sentence, [S1 might be heard . . .], and its preposed locative PP (From that lone ruin).

These three lines thus closely resemble in all relevant respects some of the contorted passages from “Adonais” discussed at length in chapter 2. In their own poetic context, however, they hardly qualify as a major stylistic curiosity. The stanza that they introduce incorporates a variety of syntactic complexities convoluted enough to mask almost completely even the rather considerable problems caused by center-embedding:

From that lone ruin, when the steed that panted

Paused, might be heard the murmur of the motion

Of waters, as in spots forever haunted

By the choicest winds of Heaven, which are inchanted

To music, by the wand of Solitude,

That wizard wild, and the far tents implanted

Upon the plain, be seen by those who stood

Thence marking the dark shore of Ocean’s curved flood.

(CWS I, 334: 200-207)

How many readers, I wonder, succeed on the first or even the second reading in parsing this long-winded sentence? Its problems are both formal and pragmatic. When we first encounter the phrase the far tents (line 205), to consider just a single pragmatic stumblingblock, it betrays no feature that might help us to associate it with any construction then undergoing analysis. Any attempt, furthermore, to resolve this anomaly by assigning that phrase its correct role as the subject of a second main-clause conjunct [the far tents . . . (might) be seen by those . . . ] is promptly thrown into question by the freakish appearance of a completely aberrant comma separating that noun-phrase from its (foreshortened) verb-phrase. Such pragmatic problems beset the reader repeatedly as he completes this stanza of the poem. It is important to realize, though, that the stage was set for this whole syntactic disaster by the formally constraint-violating center-embedded construction with which the stanza—and our discussion of its syntactic complexities—originally began.

Those of my friends whom I have in the past challenged to paraphrase my second example of constraint-violating syntax from Laon and Cythna have for the most part had to resort to pencil and paper in order to establish a plausible analysis. The passage in question occurs at the close of the third canto of the poem, when the hero, Laon, is rescued from a nightmarish end by a mysterious hermit who rows him across a lake to safety:

And the swift boat the little waves which bore,

Were cut by its keen keel, tho’ slantingly;

(CWS I, 298: 300–301)

We must assume, I take it, that the underlying syntactic form of this sentence is as follows:

[S1 [NP1 The little waves [S2 which bore [NP2 the swift boat]]] were cut by its keen keel].

Deriving the surface form actually attested in Shelley’s poem from this structure will involve preposing NP2 (the swift boat) not merely to the left of the relative clause within which it originated (S2), but beyond that, to the left of the head of that clause, NP1 (the little waves), as shown in Figure 22.

FIG. 22

I outlined in the course of chapter 2 the syntactic problems that Shelley causes for his reader whenever he extracts syntactic material from within tensed clauses, particularly when, in addition, those tensed clauses contain overt complementizers such as which. Complexity of both these kinds will inevitably be introduced by Shelley’s inversions in the present passage, as a glance at the accompanying diagram clearly demonstrates. In this particular case, however, the awkwardness of the lines Shelley actually wrote is still further aggravated by the fact that the syntactic movement Shelley permits in the course of its derivation violates a third syntactic constraint, J. R. Ross’s Complex Noun-Phrase Constraint.²⁴ This constraint, simply described, prohibits all movement out of clauses whose syntactic “heads” (their antecedents where relative clauses are concerned) are fully specified lexical items (rather than pronouns or syntactic “dummy” items). A glance back at the diagram that summarizes Shelley’s derivation of his lines in Laon and Cythna demonstrates that such conditions are certainly met here, since NP2 is moved completely out of S2, a relative clause whose head is the lexical material, the little waves. Formally, therefore, the technical purist may arraign Shelley on no fewer than three separate counts as a perpetrator of syntactic disorders in this briefest of passages.

Ironically, it is Shelley’s readers who must serve the concurrent sentences that he has earned as they battle to disentangle the syntactic puzzle he has created for them. Initially, after all, it is virtually impossible for readers not to misanalyze the surface structure they confront in these lines. They quite naturally take NP2 (the . . . boat) to be the subject of S1 and assume NP1 (the . . . waves) to have been generated originally within the embedded clause, S2. These perfectly reasonable assumptions yield a reading which we may paraphrase as follows:

The swift boat, which bore the little waves, were cut by its keen keel.

This represents, of course, an unacceptable reading because of the clauses’ syntax (particularly the failure of subject-verb agreement) as well as their basic semantic uninterpretability. Yet correcting the misanalyses that such an approach involves will require of the same readers that they rely on their semantic intuition of what must have been intended to overcome strong prejudices against permitting constraint-violating movements in the course of syntactic derivations—prejudices which, as I have already noted in prior discussion, may be deeply rooted in the form and functioning of the human psyche.

Passages and constructions as complex as the two discussed in the preceding paragraphs abound in Laon and Cythna. One is tempted, indeed, to apply to this poem comments made about Prometheus Unbound by a reviewer for the infamous Quarterly Review who found Shelley’s work “absolutely and intrinsically unintelligible,” and singled out Shelley’s syntax for particularly virulent attack, asserting that “both the ear and the understanding are disgusted, . . . by awkward and intricate construction of sentences.” Nor is it at all difficult to find more recent critics who have in fact been at least this damning in their treatment of the style of Laon and Cythna. A. Clutton-Brock, for example, remarks of it:

The poem is difficult to read as it is; in blank verse it would probably have been impossible.²⁵

But the very feature of this poem’s form that renders it least attractive to the general reader—the randomness and pervasiveness of the syntactic anarchies it contains—makes it a work of very great relevance and interest to our discussion in this chapter.

Laon and Cythna addresses a wide variety of issues—political, philosophical, religious, and social. A work much like Queen Mab in this respect, its attempts at comprehensiveness may again represent its most damaging flaw as a complete poetic achievement. Yet no one of its many themes, nor even a coherent grouping of some few among them, appears to conjure up in any consistent fashion the syntactically deviant language that I have characterized in the preceding pages. Beside Cythna and Laon themselves, this work also develops a large cast of characters—some named, others merely “a child” or “a Youth.” Again, though, syntactic complexity is restricted to no one voice, at least so far as I have been able to discover. I can certainly conceive, then, of circumstances in which some technical or perceptual aspect of syntactic form, even in a poem as long and as involved as this one, might correspond productively with one of its thematic or narrative emphases. In actuality, I have been unable to detect any such correspondence in the case of Laon and Cythna. In discussing this work, I am repeatedly brought up short after completing my technical account of its style, frustrated in every attempt to offer interesting observations on the interpretive front.

(I do not, of course, view such a limitation as an embarrassment for the stylistic method of critical analysis. As I first suggested in chapter 1, the stylist’s work may be carried on at any of the three analytical levels or by combining observations from any two or all of those levels. No analysis is fairly criticized merely for restricting itself to fewer than three levels. Nor, in principle, do I see any justification for valuing less highly works in which certain technical or perceptual stylistic features display no clear interpretive reflexes. I shall have more to say on this head at a later stage in this study; for now I simply observe that, in approaching works such as Laon and Cythna, the stylist may quite easily find himself with a great deal to say at one level of analysis—in this case the technical—yet be unable to take even the most preliminary steps toward an interpretive generalization.)

This situation contrasts vividly with that obtaining in the case of “Adonais.” In that poem, a high concentration of constraint-violating structures in a single section of the text does at least invite the analyst to consider the possibility that syntactic complexity may be playing a determinate role in Shelley’s composition. Selectivity, or what Geoffrey Leech and Michael Short would call “cohesion and consistency in preference,”²⁶ rather than mere virtuosity or carelessness, seems to be controlling Shelley’s pen in the later of these two poems. Such selectivity merits our special attention, of course, because the additional detection of a principle capable of explaining the perceived distribution of syntactic effects would constitute our first serious candidate for consideration as an interpretive stylistic strategy.

The thematic structure of “Adonais” follows the sequence prescribed for elegiac poems by literary convention. In the opening stanzas, Shelley repeatedly stresses the irrevocability of death and in particular his personal sense of loss upon hearing of the passing of Adonais (Keats):

I weep for ADONAIS—he is dead!

(CWS II, 389: 1)

Lost Echo sits amid the voiceless mountains, . . .

And will no more reply to winds or fountains,

Or amorous birds perched on the young green spray,

(CWS II, 393: 127, 129-130)

He will awake no more, oh, never more!

(CWS II, 395: 190)

However, in the thirty-eighth stanza (and, appropriately enough, with the syntactically and semantically startling remark “Not let us weep . . .”) this elegy makes the anticipated turn towards reconciliation. In a sequence of images carefully chosen to match and to supersede those of the opening section of the poem, Shelley insists on the enduring power of Keats’ creative genius:

Mourn not for Adonais.

(CWS II, 401: 362; emphases in all three of these citations are my own)

He is made one with Nature: there is heard

His voice in all her music, from the moan

Of thunder, to the song of night’s sweet bird

(CWS II, 401: 370-372)

He hath awakened . . .

(CWS II, 400: 344)

From this reorientation stems, at last, the hopefulness and the calm that mark the poem’s perhaps rather unorthodox concluding question, “What Adonais is, why fear we to become?” (CWS II, 404: 459) By the time we reach these closing lines, we realize that Shelley’s views have swung through 180 degrees, from an initial sense of cruel deprivation and hopelessness at Keats’ death to an imaginative recreation of a joyful reunion with his fellow artist, a reunion conceivable only beyond the frustrations and restrictions of mortal existence.

Within this thematic framework, the violations of syntactic constraints discussed in chapter 2 of this study occur in only three, and indeed in three consecutive, stanzas of the poem—Stanzas IV, V, and VI. Nowhere else, despite repeated rereadings of the text, have I been able to isolate a single phrase or clause whose syntactic structure even approaches in complexity the contortedness of these three closely adjacent passages. In stylistic analysis, as indeed in any other critical pursuit, a textual feature thrown into such dramatic relief by its localized distribution within the work as a whole surely demands our careful attention.

Stanzas IV-VI of “Adonais” contribute to the pessimistic early section of that poem a brief historical survey of injustices perpetrated on poets by their sadly unappreciative publics. Stanza IV specifically laments the miserable fate of John Milton, who died “Blind, old, and lonely”; Stanza V describes the failure of later poets’ public reputations to reflect their individual merits (“tapers yet burn through that night of time/In which suns perished”—my emphases); and Stanza VI focuses on what Shelley sees as just one more example of such injustice, Keats’ own unfair treatment as “the loveliest and the last” of the lyricists in Milton’s line.

Those familiar with Shelley’s other writings will recognize in the human imperfection that he laments in these stanzas a typical symptom of the postlapsarian world described at greater length in A Defence of Poetry. That world is one in which, Shelley claims, we mortals simply cannot shake off “the curse which binds us to be subjected to the accident of surrounding impressions” (CWS VII, 137). In the early stanzas of “Adonais,” then, we witness at first hand this world untransmuted by poetic vision, where Shelley, as a representative of humankind, flounders about, unable to see recent events in their proper context or from a sufficiently elevated perspective. And it is in his attempt to convey this particular rather obscure state of affairs poetically, in his anxiety to capture the inadequacy of this all too “familiar world” which, as he says in the Defence, “is a chaos,” that Shelley relies on a syntactic form that is also “a chaos,” full of complexities and convolutions. At this point, early in the poem, as Shelley deliberately places himself on a par with those whose outlook he will later criticize, the wrong-headedness of the mundane standards by which he at first assesses and mourns Keats’ death is suggested obliquely for the reader by the contortedness of the grammatical means by which he expresses his sorrow and anger.

But two forces, Shelley believed, were constantly working to ameliorate Man’s sorry plight: history and poetry. The contemporary poet, aided by both his position in historical time and his aesthetic sense, represented all that was most promising for the future:

It is impossible to read the compositions of the most celebrated writers of the present day without being startled with the electric life which burns within their words (A Defence of Poetry, CWS VII, 140).

It is, of course, this enduring “life” that Shelley celebrates later in “Adonais,” poetic inspiration as a power that reveals once and for all the limitedness of a vision not mediated in this way:

’Tis we, who lost in stormy visions, keep

With phantoms an unprofitable strife,

And in mad trance strike with our spirit’s knife

Invulnerable nothings.

(CWS II, 400: 345-348)

Yet how was Shelley to suggest for his readers, mere mortals and thus inevitably unfamiliar with this poetic super-vision, the character of that amazing higher plane of knowledge? For the clearer, purer, truer vision that great poets such as Keats could alone achieve, Shelley offers, if you like, the syntactic metaphor of a style clear of technical challenge and complexity. Toward the end of the poem his voice rings out boldly in simple clauses:

The soft sky smiles,—the low wind whispers near:

’Tis Adonais calls! oh, hasten thither,

No more let life divide what Death can join together.

(CWS II, 404: 475-477)

This, I would claim, is Shelley’s true “voice of poetry,” a syntax as clean and uncluttered but also as “burning” as the “electric life” that he so much admired in the work of his contemporaries.

If we accept it at all, then, we accept Shelley’s tone of calm in the closing lines of this elegy largely because his use of the English language has, throughout the poem, reflected and reinforced the contrast between the worldly and poetic visions on which his whole philosophy is predicated. An understanding of the crucial roles that syntactic complexity and simplicity play at different moments in this poem is thus at least as useful a critical tool as, say, a thorough historical knowledge of the relationship between Shelley’s poetry and the Defence. Neither kind of information will, of course, guarantee us a more rapid or an absolutely “valid” interpretation of “Adonais.” Both, however, do afford us the chance of a valuable insight into the philosophical assumptions that motivate Shelley’s resigned, if not almost suicidal, tone at the end of this extremely challenging poem.

III

This chapter opened, somewhat combatively, with a defense of the stylist’s right to seek out and to examine correlations between syntactic forms on the one hand and interpretive emphases or contrasts on the other. I immediately conceded that the potential for discovering such a correlation necessarily implied also the possibility that in certain cases no correlation would in fact present itself. Shelley’s Laon and Cythna, or more precisely the pervasive and apparently undirected syntactic complexities in that poem, were adduced as an excellent case in point. Yet now we see that “Adonais,” a poem by the very same author exhibiting certain very similar syntactic properties, belongs in an altogether different category so far as interpretive stylistic analysis is concerned. In this second poem, we can assign to each of the poles of one of the poem’s major thematic contrasts an intuitively apt mode of syntactic expression. Passages describing mortal man’s imperfect vision appear in a syntactic form that confuses the reader by flouting linguistic conventions. For sections of the poem intended to delineate the clearer perspectives of poetic vision, however, Shelley adopts a comparatively lucid, less baffling set of syntactic structures.²⁷ This correlation is one of those positive correspondences between syntax and interpretation that I earlier defined as the proper subject matter of interpretive stylistics. It therefore remains to assign that particular kind of correspondence a suitable name, if only for the sake of convenience, and then to establish more precisely the characteristics that qualify a given interpretive strategy for membership in that subclassification.

In selecting the term mimesis for such interpretive effects as Shelley’s use of syntactic complexity in “Adonais,” I am aware that I run a considerable risk of inviting controversy. I cannot help agreeing with Smith when she objects strongly to stylists’ excessive reliance on the words mime, mimetic, and mimesis.²⁸ But of the few other terms that spring to mind from the literature, icon, iconic, and iconicity suffer from almost equal overuse and abuse. In any case, I shall shortly need to invoke that particular trio to name a second category of interpretive effects—effects that I shall want expressly to distinguish from those observed in “Adonais.” By the same token, the notion of a metaphor, while useful metaphorically, remains notoriously imprecise and poorly understood wherever it is applied. I shall continue, therefore, to rely on the familiar mimesis, offsetting its decidedly checkered history by providing as full a characterization as I can of the term within the theoretical framework that I am developing. The same arguments, mutatis mutandis, will also have to justify my decision to employ the term iconicity for other examples that fill the coming pages. Neither this parallelism, though, nor the fact that I have little choice but to define each term by comparing it with the other, seriously undermines, I believe, the value of both as critical concepts.

The crucial distinction between what I intend by syntactic mimesis and what I intend by syntactic iconicity in the interpretive analysis of poetic style is best captured, I think, in the following pair of wholly artificial sample sentences:

(1) Behind the door, ill at ease, I stood, expecting at any moment to betray my presence, and awaited his return. (MIMESIS)

(2) Behind the door stood a tall man, and behind him, a little girl with bright red hair. (ICONICITY)

In the first of these sentences, we focus our attention on the speaker, and specifically on his mood and his predicament. The surface syntactic structure of the sentence helps to identify that predicament as one full of danger, the speaker’s mood as one of nervous apprehension. PP-Fronting and other syntactic movement processes combine to create a general impression of disjointedness, uncertainty, and tentativeness—an impression that becomes immediately evident if sentence (1) is contrasted with its more standard paraphrase, (1’).

(1’) I stood and awaited his return behind the door, ill at ease, expecting to betray my presence at any moment.

It is my contention that stylists should reserve the term mimesis for just such cases as this, cases in which it is an abstract theme of the passage under consideration with which syntactic features are being alleged to correlate. Our recently completed discussion of syntactic form in “Adonais” constitutes an excellent example taken from a major literary work, since there the abstract concept of flawed human perception provided a thematic correlate for syntactic complexity.

Compare with such cases sentence (2) from the preceding paragraph. Fronting the PPs behind the door and behind him in this construction achieves no dramatic mirroring of any thematic concern of its author. Instead, syntactic movement has here resulted in a surface structure in which the noun-phrase the door occurs, quite literally, as the first item that we encounter as we read through the sentence. The phrase a tall man, in deep structure the sentence-initial subject NP, thus appears “behind” the phrase the door, and the words describing the little girl only “after” that. In this sense, the physical (temporal and spatial) order in which the reader naturally processes these crucial referential NPs precisely matches the physical (spatial) relations that are claimed to hold within the scene described. To see how helpful such an interpretive aid from the syntax may be, one need only contrast sentence (2) with its denotatively parallel but untransformed counterpart, (2’).

(2’) A tall man stood behind the door, and a little girl with bright red hair (stood) behind him.

Here the reader must envision the whole scene on the basis of the sentence’s semantic content; no clues from the syntax will help him to “set the stage.”

Where, as in sentence (2), a correspondence between syntactic form and interpretation is based on physical rather than thematic imitation, I shall use the term iconicity in referring to it. One need not look far for examples of such iconic strategies, since cases of this kind have long figured prominently in the stylistic literature; chapter 4 of Richard Cureton’s dissertation, for a start, contains a challengingly full inventory culled from the works of E. E. Cummings (“high priest,” perhaps, of iconic syntax).²⁹ For those who do not wish to move even that far afield in search of appropriate examples from the works of major poets, chapter 3 of this study itself provided several, though at the time we were focusing almost entirely on those examples’ perceptual, rather than interpretive, analysis. Among other passages discussed there, for instance, was one from Coleridge’s “Reflections on Having Left a Place of Retirement.” As a brief review of that discussion will confirm, the striking concentricity that characterizes the syntactic form of the opening lines of that poem was held to represent a perfect spatial icon for the concave valley, the “little landscape round,” that Coleridge sought to describe. In discussing Wordsworth’s The Ruined Cottage later in the same chapter, I pointed out in a similar vein that the “interleaving” of structurally related syntactic elements at surface structure in the opening lines of that work resulted in an iconic recreation of the “dappled” sunlight that the poem’s narrator studies and describes so carefully. And, still in chapter 3 of this study, I pointed to one other, non-Romantic example of syntactic iconicity, a simple line of Dryden’s cited originally for its instantiation of concentric perceptual patterning:

As you meet it, the Land approacheth you.

(“Astraea Redux”: WJD I, 29: 253)

From our current perspective, this line represents a perfect icon in miniature—not, it is true, of the advancing king’s actual progress, but of the reciprocal movement of both monarch and land-mass that Dryden would have us imagine to have occurred under the august and propitious circumstances he describes. Both the stylistic literature in general, then, and earlier sections of this study furnish plentiful examples of iconic syntax in major texts.

If I now include one further, new illustration, I do so at least in part because it represents something of a personal favorite. In writing to Robert Southey in July 1797, Coleridge transcribed an early version of the poem later published as “This Lime-Tree Bower My Prison.”³⁰ As he then wrote it, the poem included the lines:

My friends . . .

Wander delighted, and look down, perchance,

On that same rifted Dell, where many an Ash

Twists it’s [sic] wild limbs beside the ferny rock . . .

In this passage the surface syntactic order of the constituents within each clause corresponds closely to their probable order in deep structure. At the same time we may note in passing Coleridge’s use of the verb twists in the final clause, a verb which must effectively carry all of the descriptive weight as Coleridge tries to capture for us the ashes’ strangely contorted poses.

This original text was not destined to last long. By 1800, when the poem appeared in the Annual Anthology, the passage excerpted above had undergone several changes:

Friends, whom I never more may meet again, . . .

Wander in gladness, and wind down, perchance,

To that still roaring dell . . .

Where its slim trunk the ash from rock to rock

Flings arching like a bridge;

(CPW I, 179: 6, 8-9, 12–13)

In comparing these two texts, one may by all means commend Coleridge for his intuition that the poem would benefit from having the imagined walking-party actually “wind down” to the waterfall instead of remaining mere spectators high above it. Much may also be said for the characteristically Coleridgean oxymoron “still roaring” which replaces the relatively drab “same rifted” of the 1797 text. Most relevant to our present concerns, however, are the newly introduced syntactic inversions of lines 12 and 13. In the 1800 revision, two correlative PPs (from rock and to rock) have been preposed, and the object NP (its slim trunk) has also been fronted within its clause. These emendations represent an excellent example of iconically motivated syntactic maneuvering. The contortions of the later text’s syntax, to be precise, surely help to convey the twisted form of the ash trees that fill this particular Coleridgean “dell” as they recreate for the reader a linguistic “wild”-ness wholly absent from the syntactically tame early version.

This particular case history differs from others that I have presented in this essay for one very significant reason. For once, it seems to me, the interpretive claim that I have made need not fall back upon a humble appeal for intuitive assent.³¹ For, in this instance, the poet himself, sensing however unconsciously the expressive power now concentrated in the syntactic form of his newly revised lines, also felt free to replace the (once pivotal) verb twists with flings. This decision represents in its own right an aesthetically fortunate move in that it adds a new dynamism and brings out a previously unsuspected horizontal axis to the trees’ arc over the waterfall. But by and in itself, this verbal emendation could never have been so successful had not Coleridge’s syntactic rephrasing of the passage already built convolution into his overall description stylistically. The introduction of flings and its success as an ameliorative revision thus offers valuable corroboration for the hypothesis that “twistedness” becomes in the 1800 text of this poem a function of syntactic rather than lexical choice.

In our final example of syntactic iconicity, then, we see an interplay between “referential or propositional meaning” (the strictly semantic content of the word twists, however that is to be determined) and what might be termed “stylistic meaning” (in this passage, the iconic expressive power of Coleridge’s twisted syntax). By reviewing the variant texts of this poem chronologically, we discover Coleridge’s capacity for shifting the weight of his description, so to speak, from one foot to the other. As iconically charged syntactic style assumes a particular semantic function in the transition from the 1797 to the 1800 texts, so Coleridge’s need to express that same semantic notion explicitly and referentially appears to diminish.

I am of course well aware that many traditional critical theorists and hard-line transformational grammarians alike will have problems in accepting the idea that syntactic form can “mean” in this very definite, almost paraphrasable sense. A single example such as that presented here will scarcely persuade such skeptics, furthermore, and if we restrict our discussion to cases of interpretive iconicity alone, progress will be slow indeed, since few such cases are so dramatically corroborable using textual emendations as clinching evidence. In what remains of this chapter, therefore, I want to turn instead to a different subfield of interpretive stylistics, an area too seldom discussed by contemporary stylists—rhetorical applications of poetic syntax. There, if anywhere, in the use of style as an overtly argumentative tool, we should be able to observe and evaluate syntactic forms’ power to “mean.”

IV

In his satiric poem “The Medall,” John Dryden writes:

The common Cry is ev’n Religion’s Test;

The Turk’s is, at Constantinople, best;

Idols in India, Popery at Rome;

And our own Worship onely true at home:

(WJD II, 46: 103-106)

These four lines offer the reader an inductive generalization (line 103) supported by four items of evidence. In presenting that evidence, Dryden relies on two quite separate techniques to convince his reader that it does indeed provide an adequate basis for the inference he has drawn. He depends first and foremost, of course, on having selected appropriately varied yet complementary illustrations of his thesis. Both geographically and theologically, his four exempla suppose an attractively heterogeneous field of discourse. On these grounds alone, Dryden may hope to carry the day argumentatively, relying only on the simplest semantic interpretation of his lines and on our tacit understanding of common-sense criteria for assessing argument-by-exemplification. If the reader is indeed convinced by these means, he may be said to have accepted what I shall call the propositional argument of Dryden’s lines.

The syntactic form of lines 104-106, however, and especially the exaggeratedly parallel, gapped clauses of line 105, fuel a subsidiary, entirely stylistic argument in the very same passage. Inevitably (and despite the syntactic deletions that have occurred in deriving the precise surface form of these clauses) the reader recognizes Dryden’s repeated use of the simple syntactic formula, “A is best/true in/at B.” Further, he intuitively defines paradigmatic classes for each of the “variables” in that formula—names of religions to occupy position A; corresponding geographical locations for position B. But in positing this syntactic template, the reader unwittingly helps Dryden to achieve his persuasive goal, for the abstract claim that it embodies (“religion x is best at geographical location x”) exactly parallels Dryden’s own summary statement in line 103.

Still more interestingly, I think, this crucial if abstract syntactic formula, once established by the reader, actually hints at examples that Dryden has no room to detail (“Shinto is true in Japan” for instance) and even permits modern readers to project corroboratory cases about which Dryden himself could, in principle, have known nothing (“Mormonism is best at Salt Lake City”). Thus the repetitive syntactic form both unifies and universalizes what would otherwise be effective but isolated corroboratory examples.³²

It is just such a usage of poetic syntax that I wish to term rhetorical.³³ In lines like these, syntactic form plays a direct and clearly defined role in developing the poet’s argument. Its contribution to the presentation of Dryden’s case is so distinct that I was able virtually to paraphrase it propositionally in the paragraph before last. Nor can this particular function of poetic syntax possibly be viewed as either mimetic or iconic. Rather, Dryden’s syntactic choices are designed solely to enable him to promote stylistically an argumentative proposition, a thesis consciously conceived and explicitly advanced.

Let me take a moment here to refine the crucial definition that I am proposing. If iconic interpretive correlations connect syntactic choices with a poet’s efforts at concrete physical description, and if mimesis, as a stylistic strategy, introduces poetic syntax into a broadly expository framework, then poets may be said to be using rhetorical syntactic techniques whenever they choose their syntactic form with an eye to furthering essentially persuasive ends. The rhetorical application of some technical or perceptual feature of syntactic structure may thus be said to enhance the reader’s appreciation of a poet’s intended inferences or deductions rather than improving his understanding of the text’s immediate denotative content.

I certainly do not lay claim to any great originality in isolating this category of stylistic effects for attention. Dillon, in a paper that I have already cited extensively in this essay, proposes three “modes” of syntactic manipulation: the “prosodic” (an aspect of poetic syntax that I have chosen to yield to the metrists), the “mimetic” (a category that conflates my own iconic and mimetic functions), and what he calls “the presentative.”³⁴ The last-named category, for Dillon, includes principally cases in which Topicalization or PP-Fronting may be used to emphasize or highlight the referents of the syntactic items that they prepose. Corresponding, then, to the parallel examples of mimesis and iconicity offered as sentences (1) and (2) earlier in this chapter, sentence (3) would represent the paradigmatic example of rhetorical or presentative syntax:

(3) Behind that door, ladies and gentlemen, lies tonight’s Grand Prize, a 1983 Buick Skylark. (RHETORICAL APPLICATION)

Fronting the PP behind that door in this instance does not evolve out of any attempt to mimic physical movement or arrangement, nor does it constitute stylistic simulation of any thematic concern. Indeed, if we avoid all consideration of the specific communicative context, the transformed syntactic structure of the sentence appears motiveless. Only if we deliberately broaden our view to include the speaker/writer’s communicative intent can we make any sense of the structural frames that he has chosen, as we appreciate his attempt to direct his audience’s gaze in a certain way, to delay their discovery of the nature of “tonight’s Grand Prize,” or most probably to do both these things at once.

Simple rhetorical emphasis of this kind, however, Dillon’s “presentative mode,” always seemed to me a limited and unsophisticated way for great poets to have employed the vast resources of syntax argumentatively. The tendency for past studies of poetic syntax to fall back repeatedly on examples of Topicalization or of PP-Fronting (if indeed they discussed rhetorical applications of syntactic devices at all) might be due, I surmised, to most stylists’ preoccupation with the Romantic and post-Romantic periods of English literature, periods in which direct argumentation in poetry was held in relatively low esteem.³⁵ With this in mind, therefore, I deliberately directed my own attention toward a school of poets whose advocacy of poetic argument has never been in doubt—the Augustans—and in particular toward the widely acknowledged master of that particular genre, John Dryden. With a variety of cases such as that cited from “The Medall” above, I began to mine a rich vein of rhetorical stylistic effects.

The passages that I shall discuss in the following pages all exemplify, then, Dryden’s masterly ability to correlate syntactic form with ratiocinative intent. What distinguishes them from one another are the ways in which the stylistic argument is made to amplify, to coincide with, to throw into question, or even to undercut the propositional argument with which it invariably coexists. (My introductory illustration exemplified only the simplest possible case, in which syntactic argument precisely parallels and reinforces propositional content. But the use of these two terms—indeed the very notion of “stylistic argument” as a concept in its own right—opens up the possibility of far more complex relationships.) We are now in a position to consider a few of these important cases.

The following couplet is taken from what is probably Dryden’s best known satire, “Absalom and Achitophel”:

God’s pamper’d people whom, debauch’d with ease,

No King could govern, nor no God could please;

(WJD II, 7: 47-48)

To determine Dryden’s explicit propositional argument in these lines, we need to set them in context. In line 45 the poet first introduced his reader to the Israelites (or, reading the text allegorically, the British public), characterizing them as “a Headstrong, Moody, Murmuring race.” Their headstrong nature, he now claims in line 48, appears most blatantly in their political and theological fickleness. Propositionally, therefore, this couplet offers an opinion about the Israelite (British) mob—a simple fact to be borne in mind during the discussion that follows.

Perceptually, the syntactic technique of this couplet differs only trivially if at all from that of the lines from “The Medall” that we discussed above. Line 48 of this poem, to be specific, achieves the same close parallelism seen in line 105 of the other. In the earlier case, it will be recalled, I argued that the general syntactic congruence of the conjoined clauses, together with the semantic similarity of the syntactic items that occupied structurally equivalent positions in them, led the reader to take two interpretively crucial steps. First, the reader reformulates Dryden’s overarching generalization about geographical places and the beliefs appropriate to them; then he extends it to embrace a variety of unspecified but logically compatible examples. In so doing, he finds his “stylistic logic” to coincide exactly with Dryden’s explicit induction. But in our second case, no such strategy would occur to the reader. Two of the key terms implicitly paired by the syntactic congruence, govern and please, by no means form a natural semantic class (as for example, did Constantinople, India, and Rome above). As a result, extending this class of predicates so as to predict a third or a fourth member would be, in a practical sense, impossible.

Let us turn then to the other twosome created by Dryden’s choice of parallel syntactic form in line 48—the words King and God. Once again, of course, the reader seeking a ready-made semantic class to encompass these two items will find himself frustrated. Yet the perceptual parallelism in the line remains a powerful influence, suggesting that the poet may himself see some semantic connection between these two items not apparently deducible from everyday usage.³⁶ It is this hypothesis, however tentative at first, that ultimately unlocks the full power of Dryden’s couplet. For in positing a semantic relationship between these syntactic terms (and thus also between their referents) as the stylistic argument of this passage, the reader rediscovers a major theme of “Absalom and Achitophel”—a theme articulated in both the opening and the closing lines of the poem:

. . . several Mothers bore

To Godlike David, several Sons before.

(WJD II, 6: 13-14; emphasis my own)

Once more the Godlike David was Restor’d,

And willing Nations knew their Lawful Lord.

(WJD II, 36: 1030-1031; emphasis again mine)

The syntactic form of the couplet we are presently examining reminds the reader, in effect, that this is to be a poem defending the controversial doctrine of the Divine Right of Kings (not just an innocuous paraphrase of a relatively obscure Bible story) and that in daring to attempt yet another change of monarch, the British (Israelites) are defying powers infinitely greater than they at first realize. Of course, the perceptual parallelism in line 48 of the poem advances such a defense of the Stuart cause only indirectly, merely linking the terms King and God without explicitly equating them. Yet in its context, such a stylistic connection surely takes on the force of an argumentative “proxy,” the echo of an argument already expressed more unambiguously elsewhere and all the more powerful here because of its covert stylistic nature.

For we should remember that at the level of propositional content this couplet is concerned with neither monarchs nor deities, let alone constitutional principles. At that level it remains a simple, if biting, indictment of the Israelite (British) character. This absolutely crucial disjunction of propositional and stylistic arguments in the passage stands out more clearly when one observes that Dryden could have leveled his charge of political and religious willfulness in the following only slightly altered words:

God’s pampered people, whom, debauched with ease,

No king could govern, nor no church could please.

What “gets lost in translation” in my reworking of these verses has nothing to do with their propositional content, everything to do with the couplet’s stylistic “hidden agenda,” its subtle propaganda in support of Charles and James Stuart.

We have now seen two rather different examples of rhetorical syntactic style. In the first, perceptual parallelism suggested a simple semantic generalization precisely supportive of the main argumentative thrust of the passage in which it occurred. In the second, a very similar syntactic form functioned instead to introduce a second level of argument, an echo from elsewhere in the poem working apart from, in addition to, even in spite of, the ostensible role of the passage in its persuasive context. This latter possibility, inherently more surprising, certainly calls for further exemplification. Let us turn, therefore, to one of Dryden’s most high-minded poems of intellectual debate, “Religio Laici.” Here, after all, the cause is one that was dear to Dryden’s heart; on the success of his presentation might have depended in large measure the survival of his church and of his political party, not to mention his own career as a professional poet. Here if anywhere we may expect to encounter the full strength of Dryden’s persuasive arsenal.

Early in “Religio Laici,” Dryden confronts with remarkable candor his greatest single problem in defending a religion of revealed truth. Such a faith, he explains, holds the acceptance of certain doctrines essential for salvation. Yet many men, women, and children have been prevented by accidents of birth beyond their control from ever encountering, let alone accepting, those doctrines. Geographical isolation is only one of several factors that appear to condemn countless souls to eternal damnation in this way. In the face of what appears to be arbitrary cruelty on the part of the God that he is called upon to defend, Dryden apparently falters:

Of all Objections this indeed is chief

To startle Reason, stagger frail Belief:

(WJD II, 115: 184-185)

At first glance, Dryden seems to have conceded this all-important point virtually without a fight. Worse yet, he would seem to have done so in a line (line 185) whose syntactic parallelism sinks to the level of the decorative but semantically tautologous.

Such an interpretation of these lines fails, I believe, to satisfy a careful reader, who expects here a far less wholesale retreat from orthodoxy. To see how this might indeed be so, let us reexamine line 185 more closely. Syntactic parallelism and phonological alliteration do indeed prompt the reader of this line to place the terms startle and stagger in a relation of approximate equivalence that is semantically unremarkable. But a similar equation of the nouns Reason and Belief on the grounds of their (entirely analogous) structural parallelism immediately strikes the informed reader as a most unexpected interpretive consequence of the underlying perceptual pattern. No Augustan poet-philosopher would so easily or carelessly imply an identity between these two very distinct mental faculties.

If, then, relating together the lexical items that occur in congruent positions in these two clauses proves interpretively unhelpful, perhaps we should instead pair up the parallel phrasal constituents, Reason and frail Belief. In that case, we might suppose Dryden’s intended stylistic argument to run something as follows: “Reason is equivalent to (only) a frail kind of belief.” Just as with the example from “Absalom and Achitophel” discussed earlier, this hypothesis should immediately strike a familiar chord. The attentive reader picks up the reverberations of a theme that has sounded throughout this long, profoundly antirationalist poem. It is a theme, for example, expressed in the poem’s rightly famous opening simile:

Dim, as the borrow’d beams of Moon and Stars

To lonely, weary, wandring Travellers,

Is Reason to the Soul: . . .

And as those nightly Tapers disappear

When Day’s bright Lord ascends our Hemisphere;

So pale grows Reason at Religions sight.

(WJD II, 109: 1-3, 8-10)

As this simile so beautifully expresses (and as the couplet that we are currently examining then reemphasizes nearly two hundred lines later), Dryden views reason as belief’s paler, frailer reflection, a source of welcome but ultimately inadequate and purely derivative illumination. This secondary thesis concerning the role of reason in theological argument, then, and not a pompous, despondent, and self-indulgent reiteration of personal reservations about the eternal fate of ignorant heathens, is responsible for the syntactic form of line 185 of Dryden’s poem.

Indeed, if read aright, that same line may be seen to point rather dramatically to the fatal flaw in the counsel of doubt and despair that the passage ostensibly (propositionally) expresses. For, Dryden implies in his equation of reason with frail belief, another kind of religious conviction, a faith not founded on reason alone, will not blanch at problems such as those he has presented. The rout of both reason and frail belief, in other words, by no means spells defeat for that more perfect kind of assurance that, as Dryden later goes on to argue, Anglicanism can provide.

In these lines, then, we see another example of stylistic argument working at a tangent to the propositional argument that the text appears to be taking as its primary emphasis. In this case, indeed, that stylistic argument, once deciphered or intuited, influences in important ways the assumptions that we bring to bear on the passage’s propositional content itself.

I propose to investigate one final example of syntactic style functioning as an independent argumentative force before drawing a few general conclusions. What distinguishes this case from the others we have discussed is that in this passage Dryden plays off against one another his stylistic and his propositional arguments with delightfully comic results. At the climax of Dryden’s satiric masterpiece “Mac Flecknoe,” the “aged prince” Flecknoe praises his heir-apparent, Shadwell, in supposedly glowing terms:

Like mine thy gentle numbers feebly creep,

Thy Tragick Muse gives smiles, thy Comick sleep.

(WJD II, 59: 197-198)

We are fully familiar now with syntactic parallelism of the kind exhibited in the second line of this couplet. A simple formula, “Thy ADJECTIVE Muse gives NOUN,” underlies each of its constituent clauses. But let us again consider the interpretive consequences of this particular parallelism.

Our previous analyses of ratiocinative syntax revolved around the interaction of perceptual parallelism with a rough semantic equivalence—an equivalence either assumed as a given of the standard language (as in the lines from “The Medall”) or imposed on the overall interpretation of the passage by the syntactic form itself (as in the citations from “Absalom and Achitophel” and “Religio Laici”). But line 198 of “Mac Flecknoe” does not conform to either pattern, since the adjectives modifying the identical noun subjects of the two parallel conjuncts [Tragick and Comick (Muse)] are semantically antonymous. As a result, some general interpretive principle dictates, I believe, that we shall expect the same relationship of antonymy to hold between those same two clauses’ objects. In effect, for the evident syntactic parallelism to be matched by an overall semantic equivalence, one pair of opposites must be counterbalanced by another. Or, to put this same observation still another way, we sense a stylistically inspired need to resolve a semantic chiasmus for which three of the four crucial terms have been explicitly provided:

TragicksmilesComick‘X’

---A-----B------B----A-

Finally, if semantic criteria dictate in position ‘X’ some term antonymous to “smiles,” the fundamental syntactic parallelism of the clauses equally certainly requires an unmodified plural noun in that position. Under this dual structural pressure, the reader anticipates as the final word in line 198 one of many terms describing emotional reactions appropriate to the genre tragedy such as “tears” or “frowns.”

Such a reconstruction of this line, or more precisely of the interpretive consequences of its rhetorical syntax, implicitly attributes to Dryden the intention of ridiculing Shadwell for routinely inspiring in his audiences emotions diametrically opposed to those conventionally appropriate to the announced genres of his plays. Such an allegation, I would maintain, is indeed conveyed by the interaction of this couplet’s syntax with its semantics; the indictment of Shadwell would be far weaker than in fact it is were we not thus led up Dryden’s stylistic “garden path.” But on the surface, Dryden astutely interrupts the apparently determined flow of his own line. He fills the all-important position ‘X’ in the chiastic sequence diagrammed above with the noun sleep. Even as the implicit stylistic argument presses us to expect tears at the close of line 198, then, explicit propositional semantics insist that we add to the list of Dryden’s charges against poor Shadwell one further, telling count—that of dullness.

In its context at the frenzied climax of “Mac Flecknoe” this heavily ironic strategy creates a rich variety of comic effects. It is, for instance, Shadwell’s ultimate indignity that he is denied even the chance of pleading that his dramatic—or melodramatic—gifts are substantial, his failures simply the result of their misapplication. Far from merely inverting tragedy and comedy in such a straightforward sense, Dryden explains, Shadwell confounds both in a sense-benumbing hodgepodge of absurdities. Correspondingly, it is his mentor Flecknoe’s sad fate, as ever, that the paean of praise that he originally intended should turn out to be both backhanded and inept, displaying in its clumsy two-level insult that special combination of “Impudence” and “Ignorance” that he had earlier “recommended” to his successor (line 146). For all this satiric richness to be fully effective, however, it is essential to Dryden’s stylistic method that the reader of lines 197-198 first appreciate the role of his rhetorical syntax, syntax which both undercuts and is in its turn undercut by the more prosaic terms of the ongoing literary indictment.

V

In the previous section, I attempted to indicate, however summarily, the expressive potential locked inside the syntax of Dryden’s famed heroic couplets. On the one hand, such analyses carry a considerable weight of implication for Dryden studies in general. On the other, they significantly broaden the horizon for students of stylistic criticism.

A very high degree of critical consensus surrounds Dryden’s role in the history of English versification:

It is the story of a poet who inherited a medium, perfected it, . . . and handed it on.³⁷

Dryden figures consistently as the inventor of the heroic couplet, although, as in this quotation, concern generally centers not so much on his own poetic writings as on those of his predecessors and successors. Dryden’s own role, in fact, is generally reduced to that of a talented intermediary who adapted the methods and discoveries of the former to the needs of the latter. Within these rather narrow limits, agreement is the order of the day. George Williamson identifies Dryden’s source, the “informing force of the couplet,” he says, being “ultimately derived from Latin rhetoric.”³⁸ Critics such as Lillian Feder and K. G. Hamilton then accept that attribution, and proceed to catalog in more detail the specific avenues by which the tradition descended from the classical authors to Dryden himself. And if the origins of the heroic couplet that Dryden “inherited,” and of the syntactic norms that accompanied it, are thus supposedly well enough understood to require only the briefest discussion, so too are the features of his technique that particularly attracted the later Neoclassical poets to whom he “handed it on.” Dryden’s couplets, Hamilton points out, required “a fine organization of words, . . . which in the hands of Pope became an almost geometric manipulation.”³⁹ And this time it is Williamson who concurs, alleging that in the later poet’s style syntactic antithesis became “structural rather than significant” and “passed into the very design of the verse.”⁴⁰

It is at this point that we may begin to notice critics’ apparent aversion to discussing one facet of their topic which one would expect prima facie to be of considerable interest and concern. Dryden, they all agree, differs from Pope precisely in his rejection of excessive or pervasive syntactic balance—purely “structural” or “geometric” arrangements. “The extreme compression and strict organization of Pope is not essential to his [Dryden’s] method,” reiterates Hamilton⁴¹; Paul Ramsey agrees: “it is easy to overrate their importance in Dryden’s poetry.”⁴² Yet amid all this testimony to the effect that Dryden’s heroic couplets were designed to perform some substantive poetic function distinct from, and often more highly valued than, the verbal fireworks typical of Pope’s style, no voice is to be heard explaining precisely what that function might be. Certainly, excellent studies of individual passages have been published; both Hamilton and Ramsey analyze the style of “Absalom and Achitophel” with considerable success.⁴³ Such achievements, however, still leave the broader theoretical questions unanswered—and indeed even unasked. What in general terms did Dryden himself expect syntax to contribute toward poetic expression? And how did he bring about such a contribution in his own works in particular?

Thanks to the interpretive stylistic analyses of particular passages undertaken earlier in this chapter, we now find ourselves in a position to begin to formulate answers to these questions. It would, I suspect, be possible to distinguish Dryden’s heroic couplet style from Pope’s perceptually, perhaps as a function of the density or complexity of the syntactic patterns that each employs. No such measure, however, will be very successful, I suspect, in isolating for the experienced reader the touchstone of either poet’s genius. It will be far more intuitively satisfying, in the final analysis, to rely instead on the interpretive purposefulness, the intensely directed suasive force, of Dryden’s syntactic forms as the characteristic most peculiar to his individual style. For it is this that one always remembers most clearly from reading his work. What for Pope was to become too often mere frosting on the poetic cake is for Dryden an essential ingredient which lends spice and texture to his already tangy argumentative batter. In the analysis of their interpretive application, therefore, rather than in the mere description of Dryden’s perceptual patterns themselves, we can come to understand better the character of this major English poet.

Stylistic critics, too, can benefit from discussions such as those in section IV of this chapter, as their theories and methods flex to meet challenges posed by data as fresh and as unfamiliar as the Augustan couplets considered there. At the very least, our discussion of Dryden’s poetic syntax has demonstrated conclusively that those familiar stylist’s standbys, mimesis and iconicity, fall seriously short of accommodating all of the relationships that may connect syntactic form with literary meaning. Nor can the “emphatic” function be any longer regarded as the sole, or even as the major, rhetorical interpretive strategy. We clearly need a new language altogether to help us capture the delicate nuances possible where syntactic style is employed to enhance argument, persuasion, and logic.

Implicitly, I have already hinted at one direction that future discussions in this area might take. Rhetorical strategies might, I suspect, be usefully classified on the basis of the kind of impact “stylistic argument” has on “propositional content.” To call an effect “corroborative,” for example, might entail that one was claiming a parallelism to exist between the thrusts of these two aspects of the argument of the passage under discussion. “Ironic” strategies would, by contrast, involve one facet of the argument undercutting the other. And “displaced” rhetorical strategies would occur wherever a stylistic argument picked up an argumentative thesis propounded propositionally not in the same passage but elsewhere in the text.

Such details will take considerable time to refine; I have done no more than scratch the surface in this chapter. What is important, I believe, is that these questions be raised, for stylists have two major responsibilities in the subfield of interpretive stylistics at the present stage of its evolution. The first is to affirm (and to reaffirm until it is clearly understood) that stylistic criticism pursues correlations between syntactic forms and their specific interpretation within the unique framework of a given text, not some linguistic passkey to the literature of the world. The second, equally vital, is to acknowledge and admire the richness and variety of those correlations. If all that I have achieved in these pages is to contribute to the “hunt” in which we are all engaged a few interpretive “possums” from a family whose presence was heretofore barely even suspected, then I still hold that effort alone to have been well worthwhile.