From Stress to Stress: An Autobiography of English Prosody. - book reviews

George T. Wright

The poets and the audiences of 1590 and 1600 were in love with the unrhymed pentameters. To us familiar; to them a Newfoundland. They ran riot with the discovery, as an age does when a new creative medium falls to its portion. They heard the beat with rapture. Even when the poets and players subtilised the beat, it must have been still a more powerful presence and a greater source of positive, conscious enjoyment than we in our sophistication will usually let it be.

Herbert Howarth (225)

1

Prosody, as Alan Holder suggests in his effort at Rethinking Meter, is a hotbed of confusion, double-talk, inaccurate analysis, and extravagant claims. The nature of rhythm, and the kind of life it assumes in poetry, are explained or perceived differently by different readers, and as everything else has multiplied in our time, so have theories that try to account for the movement of sound in poems. Stressers and timers, as George Saintsbury called them, have perhaps always been with us, and new perspectives on language and literature - Russian Formalism, structural linguistics, Chomskyan generative theory, New Criticism, free verse, and New Historicism, to mention a few - have spawned approaches to prosody so specialized that their adherents find it increasingly difficult to learn from one another, a condition exacerbated by the habit of dogmatic belief in the rightness of their own views and the inadequacy or absurdity of others'. That seems particularly natural to experts in this field. When I first ventured into this kind of study, coming to it as a teacher of literature, I remember being surprised to find that the books on meter were shelved in the library under Language, not Literature, and my surprise was a sign of my readiness to understand meter as an instrument of aesthetic effect rather than as a feature of language (though now I think of it as both). But the question arises constantly as to whether prosody is 'essentially' or primarily a matter of linguistics or aesthetics, of rhythmic sound or of syntax, of dramatic speech in performance, or of the allocation of major and minor words to strong or weak positions in metrical lines. Without agreement on such topics, and starting from different sets of assumptions, the different schools have developed remarkably different methodologies and, for the most part, regard their rivals with varying mixtures of indifference, hostility, and contempt. A critic like Holder, whose book should probably be entitled Rejecting Meter, has little difficulty in making the internecine wars between and within these schools (especially within the 'classical,"foot-bound' approach to meter) an object of ridicule.

Holder's position, nevertheless, is hardly a serious one, because his amusement at the antics of prosodists in distress is apparently based on the view that they really have nothing to study, that the very notion of meter as a recurrent rhythmic pattern is a delusion, and that the hundreds of scholars and critics who have, over several centuries, made it their study have been occupied with nothing more substantial than the crazy scientists of Laputa whom Swift so effectively satirized in Gulliver's Travels. Much nonsense, though, and equally unbecoming disputes have scarred the history of law, religion, politics, philosophy, and literary criticism, to go no further, but they do no! often lead to the conclusion that the objects of these inquiries do not exist. That informed and intelligent debaters may sometimes express foolish views on a subject does not vaporize the subject.

For readers of poetry (students, scholars, critics, actors), but not necessarily for linguists, whose interest is likely to be more theoretical, the main point of metrical study is to provide coherent and consistent guidelines for following, by ear or voice or eye, the rhythmic patterns that poets have composed in their poems. Traditional English prosody has usually described these patterns in terms borrowed from ancient poetry written in quantitative meters, which are very different from the accentual or strong-stress meters common in English verse. This has resulted in the widespread perception that this system of prosodic analysis fails to fit the body of verse it is supposed to describe and analyze, and virtually all the newer linguistic and some of the literary approaches to English prosody attempt to modify, recast, and reframe the questions, especially to recognize that an accentual language in which the prominence of syllables is based on stresses is utterly different from ancient languages like Greek or Latin in which the prominence of syllables was largely based on length or duration or pitch (though sharpness of accent could also play a role). Even if sixteenth-century English "iambic" verse evidently developed as a kind of accentual equivalent of the Latin iambic pattern, the equivalency is not perfect. After all, if, in theory at least, a long syllable in Latin was equal in duration to two short syllables, that arithmetical equivalence had no counterpart in an English syllabic system, unless in triple-rhythm verse. But the English duple verse we find in Chaucer, and again in Tudor poets of the mid-sixteenth century, was different and was based on an alternation of less stressed and more stressed syllables, sometimes perhaps confirmed by the longer duration of the stressed ones but much more often not. The stressed syllable might simply be said more sharply or louder, but it might last no longer than the unstressed one, and, in fact, as iambic verse came more and more to be written in the mid-sixteenth century, the tune it generated was typically a tripping one in which it was easy to hear a succession of lesser and greater points of stress. The key was the alternation - of sharpness, perhaps of loudness, but not of alternately long-held and quick-spoken syllables. This kind of thing:

Thou Cupide God of love, whom Venus thralles do serve, I yield thee thanks upon my knees, as thou dost well deserve. By thee my wished joyes have shaken of[f] despaire, And all my storming dayes be past, and weather waxeth faire.

(Tottel's Miscellany, 231, 1-4)

Verse as regular as this, fixing its stresses firmly in the even-numbered syllables of the line, raised no immediate practical or theoretical problems. Very occasional variations in stress - major monosyllabic words in unstressed positions or reversals of the lesser-greater pattern especially in the first two syllables of the line - did not substantially disrupt a listener's sense of alternation as the basic rhythmic feature of the line. But when, in the next generation, poets began to employ these and other variations more frequently and to make them a familiar part of the metrical design, the theoretical difficulties began. The alternation still persists as an easily heard phenomenon, but it is subject to modification by an array of deviations from the pattern that have led most poets and metrists in the centuries that followed to view the resulting line as most helpfully described as being composed of "feet," i.e., of segments consisting of a normally lightly stressed syllable followed by a more strongly stressed one (??? / in the usual graphic representation). This may seem at first like just another way of defining alternation, but in practice it allows the scanner of a line to treat each pair of syllables independently, so that both syllables of one pair may be lightly stressed and both syllables of the next strongly stressed. If, then, the third of this series of four syllables is more strongly pronounced than the second (say, ?? \ ^/, in the Trager-Smith four-level system), the principle of alternation will seem to be violated, unless we maintain that what matters is the relative strength within each pair. The problem here, however, for many critics of this system is that we normally do not hear lines in two-syllable segments. But, after all, the "foot" theory does not claim that we do, only that it is because of the structural design (too quick in passing to be consciously picked out but graphically analyzable if we wish to write it out and slow it down) that we accept this slight modification of the alternation principle. Some prosodists maintain, quite reasonably, that when the stricter alternating verse changes into a verse that tolerates and even enjoys the variations made possible by the analysis of the line into feet - notably, the trochaic "inversion" (/?? for ??/) and the pyrrhic-spondaic combination (?? ?? \ /) we are really talking about a different meter (see Cable 114-31). But it seems just as reasonable to maintain that when this happens, as it appears to have done quite palpably with the verse of Sidney and Shakespeare, the alternating meter has simply evolved into a looser form, a form about which it may be said that it tends to function somewhat less as a continuing musical mode of presentation that is pleasing in itself and for its own sake and somewhat more as a base from which expressive variations may occasionally mime the action or feeling conveyed by the words (see James I. Wimsatt).

Although the theoretical misgivings took a long time to find their way into writing about meter, some evidence of the period's uncertainty about the iambic decasyllabic line's structure is to be found in the great variety of metrical styles that poets devised. Spenser's usually regular and mellifluous metric (derived from Chaucer) is vastly different from Jonson's more spare but still regular lyric verse, but Donne's apparently deviant verse style is extraordinarily different from either. Shakespeare's dramatic verse undergoes extreme changes from his early plays to the late. Milton's epic blank verse is in a class by itself - radically enjambed, and following some deviational patterns hardly known earlier, unless in late Shakespeare. The more "correct" iambic verse that developed through the seventeenth century and remained dominant through much of the eighteenth was seen by its practitioners and their critics as a needed corrective to the licenses practiced by their older and wilder precursors - in effect, as a means of saving iambic verse and its permissible range of variations in the alternating decasyllabic pattern from tendencies that would undermine its structure and, by implication, the culture's inherent commitment to order.

By the nineteenth century the foot found deeper trouble. Poets began to write more poems in triple and dipodic meters and to sprinkle extra syllables in the duple feet of iambic poems. The effect was to make decasyllabic verse more accentual and to call into question the syllabic ingredient in its structure. Now, instead of the ear's hearing the intervals between the beats of the line as occupied - and indeed measured - usually by a single syllable, the measurement was increasingly achieved by a stricter keeping of time, regardless of whether one or two syllables intervened between the stressed ones. One result was that critics and readers began to hear "iambic" verse more and more as a stricter accentual system than it had probably been from the mid-sixteenth century to the mid-nineteenth century, a tendency that, as we shall see, has reached a significant critical climax in the metrical theory of Derek Attridge. On the other hand, the subordination of all metrical poetry to an ever-more-popular "free verse" has caused many readers to lose their ability to hear metrical verse as such (or at least to pay much attention to its pattern), a tendency that, as we shall see, has attained a kind of critical climax in the anti-metrical polemics of Alan Holder. Both of these critics, in effect, show what happens as centuries pass and a people loses the tune of an older style of verse, as, even in the Renaissance, most readers had already lost their ability to hear the tunes of Latin verse, a phenomenon which, ironically, has been most masterfully chronicled by Derek Attridge in his splendid account, Well-Weighed Syllables (1974).

2

Attridge's current book, Poetic Rhythm: An Introduction (1995; henceforth PR), is derived from a longer and more detailed study published in 1984, The Rhythms of English Poetry (henceforth REP). This earlier work is an elegant effort to put literary prosody on a sound linguistic basis and yet to explain how the management of metrical forms by English poets enabled them to enhance the expressiveness of their poetry, a subject of primary concern to literary critics. PR has grown out of REP, but it is meant for students and it presents essentially the same view of meter, but more genially, less as a rule-driven system and more as a descriptive account of English meters. To appraise this account, however, we need first to review how REP approached this subject.

In many ways REP was an admirable and successful undertaking, a landmark in practical prosodic analysis. Attridge is a first-rate explainer, and he uses this talent to summarize various approaches to prosody that have been used in our time: "classical" (stressers), temporal (timers), and linguistic (structuralists and generativists). He finds all of these wanting, but he also learns and borrows from all of them. After a fine section on "Rhythm" and the way it operates in English speech and hence in English poetry, REP analyzes how metrical rules operate in English verse and how different contexts and "conditions" work to complicate the elementary structures of English versification.

The formulation of rules would appear to have two purposes: (1) to theorize meter for linguists - i.e., to account systematically for the structured ways in which metrical patterns and linguistic forms intersect in verse; and (2) to help expert readers explain to beginners how metrical poems work. Neither of these is easy, partly because few experts have the credentials needed to throw light on both, and a persuasive account of either depends on some accurate understanding of the other. Even linguists who care for poetry are often not notably sensitive to its aural patterns, and literary metrists are usually limited to what O. B. Hardison called "applied prosody" (xiv) - i.e., the detailed study of line-patterns or of particular lines of verse, usually through scansions that run the risk of being impressionistic rather than, in some sense, objective. On the other hand, the formulation of rules that govern the rhythmical structure of metrical lines is likely to become a project without end, forever seeking an elusive precision. From my own point of view - that of an unreconstructed literary metrist and applied prosodist - it is not rules that will help most readers to get the hang of, say, iambic pentameter, not rules that prescribe and forbid, but orderly and accurate description that tells us what happens rhythmically in poems that use this meter.

Happily, that is what Attridge largely gives his readers in PR. Although the rules he has formulated in REP remain implicit throughout PR, in the later book, since it is intended for students, he does not burden the reader with their neo-Euclidean formulation. Instead, he explains - and does this brilliantly - what rhythmic patterns we are likely to encounter consistently in the metrical lines of English verse, how beats can be heard, and offbeats, and how frequently the beats are arranged in sequences of four, and the four-beat lines in groups of four. The basic terms are made available to readers in an unfailingly intelligent and fluent commentary, about which one's only complaint might be that the author asks a student in his opening pages to master too much abstract material too soon, with too few friendly examples. Of course, a wise teacher can help students through this initial difficulty, and for any teacher of the subject it is a pleasure to read a well-informed guide who is so expert and so sensitive to the movements of rhythmic verse.

He is particularly helpful in showing how the natural "peaks" of spoken language can be converted, in verse, to metrical beats. For "We are all experts in rhythm and meter" (43), and respond from our earliest years to the beats we can hear and reproduce in nursery rhymes and other wholly accessible popular forms of verse, all of which we understand to be intimately connected to the rhythms of the language we are learning to speak. Attridge shows charmingly how easy it is to turn almost any stretch of spoken English into verse (49-52), especially into the 4 x 4 pattern that he argues persuasively is almost anyone's standard notion of what verse is. He makes the important point (overlooked, as he justly remarks, in "many scansions and discussions of verse") "that STRESSES are different from BEATS; they often coincide, but they are not the same thing" (64, his emphasis). So, though his scansion system provides a / for stresses, it provides a different mark, _, to signal a beat; a stress that is also a beat receives both marks: ??. To his great credit, he keeps emphasizing the difference between stresses and beats, insisting that we do not have to say a syllable strongly to recognize it as a beat. "All that is necessary is a distinct pronunciation, allowing each syllable its own rhythmic space" (109). Alas, I shall use this sensible formulation against several of his claims.

What is striking about Attridge's approach to English meter is that, in company with the "timers," he regards accentual or "strong-stress" verse as central and develops a method of scansion based on the beats and offbeats that are clearly audible in this kind of verse. He contends, quite plausibly, that virtually all English verse, except iambic pentameter and other kinds of syllable-stress verse, is essentially four-beat accentual. Lines of hymns and ballads that appear to have only three beats make up the deficit with a fourth, silent one, and almost all other kinds of lines - fourteeners, poulter's measure, most hexameters, limericks, and much verse whose lineation is deceptive - turn out to be written not only in four-beat lines but in four-line stanzas. The 4 x 4 pattern is the basic form for English poetry, the kind of rhythm that every child can recognize and respond to, from "Nursery rhymes, street games, popular ballads, advertising jingles, sports chants, rap lyrics, and . . . other kinds of widely familiar verse" (63). For this verse Attridge develops a notation based on symbols for beats (B) and offbeats (o). Both the analysis and the notation are explained clearly and work well. The perception that this kind of poetry often, even usually, and perhaps almost always, implies measured pauses at the end of shorter lines, and that these lines, so augmented, take their place in a 4 x 4 metrical design, ought to command the assent of most readers, critics, and linguists.

But to give an adequate account of poetic rhythm in English requires an equally convincing analysis of the two main systems of verse that are not subsumed in the 4 x 4 pattern but have produced an extraordinary proportion of the most impressive poetry in the language: iambic pentameter and free verse. For free verse Attridge has to find other terms and methods to study its very different way of moving (more on that later); for iambic pentameter (and a few of its metrical cousins) he adapts the system that has worked so well for accentual verse. But when this beat-driven system is extended to the syllable-stress verse which Attridge acknowledges is less strongly determined by beats (PR 97-100) - traditional iambic pentameter, or, more broadly, accentual-syllabic verse - it seems inadequate to measure, explain, and illustrate what is going on in iambic pentameter, and it inevitably implies that iambic pentameter is a much more accentual meter than it is.

Attridge, of course, is quite convinced of the contrary - that "classical" metrics cannot explain traditional poetry as well as his new system can (REP 10-18 and passim). Traditional prosody "presents a dizzying array" of technical terms (PR, xix), and its foot-scansion "works only for those who already know what is going on." He regards its notation and terminology as not only elitist (xix) but as insufficiently flexible and nuanced, resting as it does on intuitions of performance rather than on rules derived from linguistic evidence about English phonology - i.e., about where to place stress when one reads verse lines aloud or scans them in writing. For Attridge, as for the generative metrists, a meter is based on rules. In PR he mentions generative theory only in his Glossary and describes it there as seeking to define what is metrical and what is not: "the limits of metricality" (219). REP, though it treats generative metrics critically, is more respectful and considers it promising (34-55, 14951). It may be that his respect for this kind of theory has declined with the years, but in neither book does he question the key idea of generative theory, that rules can be formulated which generate "all the acceptable lines in this or that variety of English metre," except to object that their rules "do not, however, generate only such lines." He rightly criticizes the generativists for their relatively crude instruments (sensitive, but to the wrong matters) for analyzing iambic pentameter. He even admits: "There comes a point where rhythmic distinctions are too fine to be captured in manageable rules, and one has to say with Saintsbury, 'The ear must decide'" (152). But his method of analysis in REP is to set out carefully formulated "rules of metre." Although most of these seem valid for accentual verse, with its strong beats and its minor, "offbeat" syllables that are unambiguously minor, for a meter as subtle as iambic pentameter the rules may start out simple enough, but the complexity of this meter's variations leads to increasingly complex explanatory procedures. Attridge runs into the same problem that helped to discredit the generativists: once you start formulating rules, marginal or tricky cases lure you into elaborating sub-rules and rule expansions to cover every possible condition, and the supposedly simple and economical language of rules begins to sound like this:

An implied offbeat may occur only (i) when it is immediately preceded or followed by a non-final double offbeat,(1) or (ii) in observance of an initial inversion condition. (194)

The notation, too, that Attridge has developed for his system, though it is meant as an improvement on the unattractive Greek terminology of classical English metrics, has grown astonishingly complicated. Toward the end of PR (213-14), he lists the symbols used to mark lines of verse for what he calls single- and double-line scansion [see page 156]. And this is the simplified version, reduced in PR so as to be printable on computers from the somewhat more cumbersome notation of REP, which includes ??, ??, ??, and ?? (for double offbeats, implied offbeats, and demoted and promoted syllables, respectively).

But even when the rules have been intelligibly formulated (and the new nomenclature and notation learned), it must be shown how they apply to lines that seem to disobey or evade them, and much of Attridge's earlier book, REP, is spent, in effect, justifying the rules - i.e., showing how they generate iambic lines that seem to Attridge on "the borders of metricality" (233). They raise problems for his system largely because, despite his insistence that "We are all experts in rhythm and meter" (PR 43), a truth that I have no quarrel with, it is not easily demonstrable that everyone is an expert in hearing the rhythms of iambic pentameter. To make good his claims that we can all hear this meter easily, too, he subscribes to dogmas (again, as the generative metrists and other prosodists do) about which kinds of words or syllables in metrical lines must receive stress.

Single-line scansion   Double-line scansion

/                      Above the line
\                      +s
??                     -s
??                     s
x                      (s)
??                     ^s
-
[/]                    Below the line
[x]                    B
/                      b
a/                     ~B
|                      o
R                      "
F                      *
M                      [B]
ANT                    [o]
ARR                    ^
STA
EXT
>

This is treacherous ground. Drama coaches usually want to say, "Read the line naturally," with an awareness both of the beat and of the sense, noting in particular, as the English director John Barton urges his actors (55-56 and passim), the antitheses that are often the key to the intelligible and forceful speaking of Shakespeare's lines. Traditional metrists like myself would nowadays usually add that where the line seems to violate the meter it will often be sufficient to "tilt" the pronunciation of the words just slightly (it will need no more than that) in order to keep the line audibly in touch with the metrical pattern. (The word "tilt" is W. K. Wimsatt's, and the idea has made excellent sense to most traditional readers of verse; see, for example, Cable's approving words about it, 123-27.) Such tilting strikes Attridge as a violation of normal stress pattern, which for him derives from the linguistic ability to distinguish two classes of English words, content words and function words. Content words (nouns, verbs, adjectives, and adverbs) convey "a full meaning by themselves"; function words (prepositions, articles, conjunctions, etc.) "depend on other words for their meaning" (PR 27). Attridge is no doubt aware how inadequate these definitions are - as if, for example, the noun "priority" conveyed more meaning than the preposition (or conjunction) "before." He accepts them, presumably, because he wants to train students to see major words as the likeliest candidates for stress, and minor ones as long shots. But this procedure then compels him, as it has other metrists, to take a lot of it back, for it turns out that, more often than anyone might expect, "content words" are not stressed. In certain circumstances their claims to stress are less pressing than those of other syllables, or they come where the metrical beats cannot conveniently fall, where the language does not need them for beatwords - they may get some degree of stress, but their stresses do not count as beats in, say, the five-beat line; whereas function words may get to occupy a beat, even without being strongly stressed. Here are two examples from Browning's The Ring and the Book, as Attridge scans them in PR (252). (Beats are underlined, whether they appear on strongly or weakly stressed syllables.)

(1) Dip a broad melon-leaf that holds the wet (2) Presently, though my path grew perilous

These lines seem to me impeccably scanned and their delivery can be made metrically coherent by only a little tilting, but Attridge, whose theory does not permit him to resort to a notion of tilting, has to explain why, in a normal delivery of these lines, what seem like stronger words (broad, grew) do not attract a beat and what look like weaker syllables (though, -ous) do. To explain this, Attridge uses, as metrists of other schools have done, the terms demotion and promotion, terms which seem to me particularly unfortunate (like substitution and inversion) because they imply that someone (but who? the poet? the reader? Attridge?) has, at some definite moment, downgraded or upgraded a syllable in a line (or substituted or inverted an already written foot). But lines of verse get written as they are written; poets do not write their lines as perfectly regular icons and then promote or demote certain syllables. They may fiddle with the words till they sound right, but they do not demote or promote syllables. What do the terms mean, anyway? Not, apparently, that the words are spoken more softly or loudly than we would normally speak them - Attridge specifically denies that (REP 168; PR 71, 75) - but that we see with our eyes and note with our sense of English grammar that a major (or a minor) syllable is sitting in the line where it may not belong. The sense of grammar is apparently what counts. What has happened here is that Attridge, like many of his peers, has first "promoted" all the stressed syllables in the language's content words (i.e., declared them eligible, by class or nature, to carry a beat) and "demoted" all the syllables in the language's so-called function words (declared them, by the same standard, ineligible to carry a beat); and then he has found it necessary to demote some of the first (promoted) group when they actually appear in verse lines, and to promote some of the second (demoted) group. But the whole idea of promotion and demotion is cumbersome, abstract, and unnecessary. Why not take the language as it comes, recognizing (1) that much of the time major words do get stress but that frequently they do not because other words they keep company with sometimes seem to have a better right to it, and (2) that poets, observing this, have placed the stress-attracting words, whether minor or major ones, in the positions that normally take stress? Especially in the packed lines we meet in much poetry the expectations we have for the language generally may riot hold; the poets themselves may find it interesting and expressive, or interesting and rhythmically attractive, to violate our expectations and to let words we have categorized as major or minor resist or attract stress in response to what the meter and the language, working in concert, ask them to do. "The ear must decide."

For many lines of English accentual-syllabic verse, Attridge's system of scansion works well enough, given his procedures for promoting and demoting, and his pages are rich with perceptive observations on language and poetry. But there are two kinds of line especially on which he differs from traditional metrists in his scansion. One of them is the line that many readers and critics have testified to hearing as a series of ascending stresses, sometimes numbered 4-3-2-1. (See Jespersen 125, Winters 94, W. K. Wimsatt 775, and Cable 123-27.) Since 4 is stronger than 3, and 2 than 1, many metrists have thought the set of four syllables intelligible as, in effect, two iambic feet, though iambic in different ways or on different scales. But for Attridge, though he admits that there may be "a slight blurring of the rhythm between the adjacent nonstress and stress" (182), the four syllables are to be heard as an "introductory double upbeat and concluding pair of beats" (183). T. V. F. Brogan has questioned Attridge on this "metrical figure" (11-12), giving as examples such lines as Pope's:

(3) On her white breast, a sparkling Cross she wore and Sidney's: (4) That in one speech two Negatives affirm (5) With how sad steps, 0 Moon, thou climb'st the skies

As Brogan points out - very sensibly, in my view - these syllables may be read, in an iambic pentameter context, as pyrrhic and spondee, in which the two syllables of both feet, though close to each other in strength, fit the iambic pattern of less stress-more stress. What matters is the relative degree of stress (lesser followed by greater) within each pair. Attridge in reply shows how strongly accentual his reading of this meter is: "I hear a line with a certain number of rhythmic peaks," not one "with five syllable pairs" (12). Whereas Brogan argues for paying attention to meter and meaning in deciding just how stresses should fall, Attridge replies, "My evidence is English phonology" (12) - and, in effect, that settles the matter: major words deserve more stress than minor ones. With his students, he prefers, he says, "to create an awareness of fundamental rhythms based on beats and offbeats, before moving to the syllabic tradition of analysis which divides lines up into pairs" of syllables (18). But Brogan hears beats and offbeats, and so do I; we just do not require the beats always to fall on so-called major or content words. We do not require it because the language does not require it; our evidence, too, is English phonology, but an English phonology unencumbered by (elitist?) dogmas about compulsory stressing of privileged syllables.

What Attridge says here in 1987 is much more absolute than his words in REP about "metrical subordination": "the specific conditions of a particular line . . . can induce any monosyllable or disyllable to lose all or some of its stress" (235). By that standard, if one can show a likelihood that lines of the type Attridge refers to here were not read his way but tended to keep the initial beat on one of the first two syllables of the four, then one may read these lines differently from Attridge, and many metrists and readers have done so through the centuries. Indeed, in the face of strong evidence to the contrary - in the published scansions of metrists and literary critics, in recordings by actors and poets, and in other kinds of evidence that attest to the way earlier poets and critics read this metrical figure, the burden of proof is still on Attridge to show convincingly why the phonological patterns of prose should overweigh the metrical pattern when all the metrical pattern needs in order to be realized is a slight, very slight, hardly detectable augmentation of emphasis on one syllable and diminution of force on another, just enough to make sure that even without strong stress the beat will be felt on the second syllable, not on the third.

In taking this stand, Attridge appears to be legitimizing a reading of lines like these that will skip lightly over the first two syllables and come down sharply on the third: On her white breast; That in one speech; With how sad steps (the audible steps, of course, disappear with this reading). And the same lurch is to be heard in thousands of other lines that English poets have written - I suspect, without ever imagining that readers, much less professional metrists, would countenance such a lurch. He scans, for example, Keats's line from the "Ode on a Grecian Urn" like this (PR 120):

(6) Thou foster-child of silence and slow time

a reading that lurches where it ought to slow down over the last four syllables. Then there is the often debated line from Shakespeare's Sonnet 30 (with Attridge's scansion [REP 261]):

(7) When to the sessions of sweet silent thought

And there are lines like these from Arnold's "The Scholar-Gypsy," which Attridge sets for students in PR and then scans himself in the "Suggested responses to the exercises" (248):

(8) Screen'd is this nook o'er the high half-reap'd field(2) (9) Through the thick corn the scarlet poppies peep (10) And round green roots and yellowing stalks I see (11) Of bloom on the bent grass where I am laid (12) And the eye travels down to Oxford's towers

Each of these lines includes the kind of four-syllable combination that traditionally has been called a pyrrhic-spondee: x x \/. The first two syllables lack one of Attridge's major words, whereas the last two syllables are both major. Not even round is major enough to entitle it to a beat; we may have trouble suppressing it, but it is only a lowly preposition, and that settles it. In Attridge's speech-world, the major syllables must receive stress (and the beat) because they are used to receiving stress in the speech of native speakers. Discussing a similar line from Crabbe's The Village (I, 112) in REP, which he gives a double-line scansion,

(13) And a bold, artful, surly, savage race

Attridge says that "there can be no question about the scansion" because the first two syllables "are clearly to be regarded as -s" i.e., unstressed (216).

But of course there can be a question; the first two syllables may normally go unstressed, but in lines like these a beat (and perhaps a stress) may still be heard on one of them. We can also establish the questionableness of his ruling by finding phrases in poems that do not distribute stresses as Attridge says they must. To begin with, it seems almost self-evident that the exercise lines quoted above (8-12) are susceptible of another reading. Three of the phrases have almost exactly the same syntax: they begin with a preposition followed by the, a monosyllabic adjective, and a noun (or the first syllable of another adjectival phrase). Many readers would therefore have no difficulty placing a little more emphasis on the preposition (especially when the preposition is a location marker: o'er, through, on) and a little less on the adjective, with the result that the first beat of the phrase falls on the first-syllable preposition and the second on the fourth syllable:

Through the thick corn, etc.

The whole series thus constitutes a milder version of the familiar trochee-spondee pattern, though this, as we shall see, is a major reason why Attridge will have nothing to do with it. But Attridge's own Promotion Rule allows it: "an unstressed syllable may realise a beat when it is not adjacent to a stress in the same line" (REP 168). That rule would allow beats on "Through" (in 9) and on "And" in (12) and (13), and would justify those readers and actors who have heard or read a slight stress on the "And" in lines like these from Tennyson's "Morte d'Arthut":

(14) And the long ripple washing in the reeds (117) (15) And the third time may prosper, get thee hence (130) (16) And the long glories of the winter moon (192)

And if on those line-initial syllables, why not when the metrical figure occurs in an echoing phrase later in the line? Otherwise, the echo will not be heard:

(17) On the waste sand by the waste sea they closed. (Tennyson, "The Passing of Arthur")

Attridge, however, scans it like this (PR 120):

On the waste sand by the waste sea they closed.

Did Tennyson's voice really lurch its way through such lines? Though I remain doubtful, it seems to me not entirely impossible, but I do not believe for a moment that Pope made a habit of doing the same lurching when he began lines in his Essay on Man with such phrases as: "From the green myriads," "To the first good," "In the same temple," "In the small circle," "As the small pebble." Pope was a stickler for correct technique, and most poets at least from Dryden to Johnson would have been scandalized by anapests leading to monosyllabic feet. But if we are supposed to hear the lurching pattern in these lines, why not in these: "When his lewd father," "When the loose mountain," "Lo, the poor Indian," and many others? The notion that such traditional poets would tolerate an anapest followed by a monosyllabic foot seems remarkably anachronistic.(3)

Consider another line of Tennyson (from "In Memoriam," VII, 12):

(18) On the bald street breaks the blank day

It looks, on the face of it, as if the second half-line echoes the first, but this can only be heard if one recognizes that for Tennyson's ear (the best in English poetry, according to Auden), "On" was heard to be stressed almost as strongly as "breaks":

On the bald street breaks the blank day

If phrases like these can be scanned as I suggest, at least as a permissible option, why should they not when they occur in midline or at line's end? Here, with some suggestions for scansion, are some early examples, written at a time when, as almost everyone agrees, the sixteenth-century line was still very regular:

(19) That cannot take a mouse as the cat can (Wyatt) (20) With sodein noise thondred on the left hand (Surrey) (21) They his bare neck beheld, and his hoar hairs (Grimald)

Or these from "Dulce Bellum Inexpertis" by Gascoigne, our best authority for the view that he and his contemporary Tudor poets would not suffer gladly the lurching movement that displaces the beat from an even-numbered syllable:

(22) But who they be now herken and take heede (23) Oh Epitaphe of honor and high happe (24) And he that keepes his Majesties great seale

Later, the poets do this kind of thing often. Here are two, among many, from Herbert:

(25) As the day lessens, and his life with it (26) Finds his crackt name at length in the church-glasse

Sidney has plenty of such lines; Jonson uses them to speak of Shakespeare:

(27) And though thou hadst small Latin and less Greek (28) He was not of an age, but for all time

Shakespeare's own work is full of them. These are from the Sonnets:

(29) But as the marigold at the sun's eye (25:6) (30) All days are nights to see till I see thee (43:13) (31) So long as youth and thou are of one date (22:2)

In all of these the first or second syllable of each italicized pair might imaginably be given a little more stress - just a little more! - and the anapest disappears. Many readers of 29-31 would find it easy to give some slight additional stress to at, till, and are; some would find it more satisfying to stress I or of a little more prominently.

Other lines from the Sonnets which can easily be made to avoid the lurch by giving just a shade more stress to the second of the four syllables in question, or, even without more stress, speaking the second syllable without hurrying:

(32) Whence hast thou this becoming of things ill (150:5) (33) Thy bosom is endeared with all hearts (31:1) (34) The world will be thy widow and still weep (9:5)

Contemporary poets, or even some earlier twentieth-century ones, might hear these italicized words as lurches, but to attribute them to much earlier poets is hardly credible, and, in my view, Attridge does no service to students in making them believe that earlier poets really heard the lines that way. Even if he himself does not adopt the lurch in reading all these lines, it is hard to see how he can discourage his students from doing so when the second syllable of such a group is a preposition or a pronoun (or the last syllable of a polysyllabic word) - that is, when even though it belongs to an unfavored grammatical category it can in speech sometimes compete with, say, adjectives for prominence. Or second-syllable copulative or auxiliary verbs may compete with what may seem like more important words. In all such cases Attridge himself may not choose to stress the third syllable, but his position invites his students or others to perform the lurching movement that takes the beat on the third syllable, not the second. Consider how all of these lines from Shakespeare's Sonnets would sound with that lurch to the italicized syllables over the two previous ones:

(35) And beauty making beautiful old rhyme (106:3) (36) The argument all bare is of more worth (103:3) (37) But do not so; I love thee in such sort (36:13, 96:13) (38) Say that thou didst forsake me for some fault (89:1) (39) Was it the proud full sail of his great verse (86:1) (40) Which eyes not yet created shall o'er-read (81:10) (41) In me thou seest the glowing of such fire (73:9) (42) Blessed are you whose worthiness gives scope (52:13) (43) Till I return, of posting is no need (51:4)

I venture to say that this is not Shakespeare's tune. It would be easy to adduce dozens more from Shakespeare's poems, and hundreds from the plays. Notice that in such lines a reader who jumps from the sixth to the ninth syllable over two intervening ones classed as minor, unstressed, and negligible is likely to put more stress on the ninth syllable than on the tenth: old rhyme, great verse, some fault, such fire, and so on. What this kind of reading can do to the subtlety of Shakespeare's argument in these poems does not bear thinking of.

Or consider Milton. In "Lycidas," I make out about seventeen places where the Attridge lurch would be called into play, among them:

(44) Together both, ere the high lawns appeared (25) (45) Batt'ning our flocks with the fresh dews of night(29) (46) Fanning their joyous leaves to thy soft lays (44) (47) But the fair guerdon when we hope to find (73) (48) Set off to th' world, nor in broad rumor lies (80) (49) But swoln with wind, and the rank mist they draw (126) (50) But that two-handed engine at the door(*) (130) (51) Ye valleys low where the mild whispers use (136) (52) On whose fresh lap the swart star sparely looks (138)

* This is one of Attridge's examples.

And so on. It should be pointed out that if we lurch through (52) we lose: entirely any sense we might have of the unusual combination in one line of what linguists call proclitic and enclitic phrases - the adjective and noun that in one case occupy a weak and then a strong position, and later a strong and then a weak position. You can hear that difference if you do not rush the line; if you lurch through the phrase, you may find yourself giving six beats to the line, failing to recognize that "star" occupies what Attridge calls an offbeat. Attridge's principles would probably oblige him to read the line as having six beats, because the adjective-noun phrases are the same grammatical construction and if we stress both members of one such pair - so he tells us (see below) - we must do the same with the other.

Lurching through some of the other lines (44, 45, 47, 49, 51) may also put a reader in jeopardy of speaking a four-beat line, "the unwanted triple effect" (REP 261), something that Attridge worries about with lines of this structure and others (256-64), because his unwillingness to hear a beat on the second syllable of a pyrrhic before a spondee results in many more double offbeats than seem audible to a more traditional listener. In what seems to me a related argument, Attridge notes how rarely poets use what he calls stress-final pairing with linkage (265), the pattern shown in (54) of four successive syllables in which the first two are weakly stressed and the next two more strongly stressed, but in which the middle two form a single word:

(53) And growing still in stature the grim shape (54) Remembering not, retains an obscure sense.

For Attridge, these two lines do not sound very different from each other (269) because in both he hears the beat on the ninth syllable, not the eighth. He notes that the second pattern is much less frequent than the first, according to various statistical studies, that apparently both poets and metrists have more trouble with it. But he fails to draw the likely inference: that poets, as he says, feel more uncomfortable with the second pattern because they find it goes against the grain to elevate the usually unstressed syllable of a disyllabic word to a beat position while the other, usually stressed, syllable does not receive a beat, but that they do not find it so awkward to hear even a word like "the" taking a beat when the next syllable is a fairly important word like an adjective. This suggests that, despite what he says about "lexically-determined"' stress, poets and other critics must hear the eighth-syllable beat in pyrrhicspondee combinations like (53) much more often than he does.

Attridge may respond that to let the beat fall on the minor syllable rather than the major one in such cases violates the phonological rules that govern normal English speech. But even in ordinary speech different structures may surprise our expectations. Natural speech has varying ways of observing the usual demands of English phonology. For one thing, pace Attridge, it may adjust to the tilting requirements of iambic verse; but even Attridge admits that there are many combinations where words of major grammatical categories may be less strongly stressed than so-called minor ones - in contrastive-stress situations, for example ("I was walking from town, not toward town"). (Compare Shakespeare's "On this side my hand, and on that side yours" [Richard II, 4.1.183]; or Auden's "The earth turns over; our side feels the cold" [144] - better than "our side feels the cold.") Phrasal verbs, too ("stand up," to use Attridge's example), may seem to displace the stress patterns that our categorizing habits lead us to expect. When news anchors on our networks speak the phrase, "When we come back . . ., "as they do time after time, night after night, they never put significant stress on "come." They always stress "back," and they variously stress either "When" or (less often) "we." In the few similar phrases Attridge scans in these two books, he usually ignores the voice's option of reducing the expected stress on the verb:

We shall grow old (PR 257), Pleas'd, out of Heaven, shalt look down (REP 246) glide o'er the dusky green (REP 263), bends toward the east (REP, 263)

preferring in the first two examples to perform the standard lurch, in the last two to read a standard but by no means obligatory trochee. Subtle as Attridge's analysis usually is, the language of verse seems to me still more subtle and capable of accepting in its stride even the unusual formations exemplified by (53) and (54).

One reason Attridge is almost obliged to rule out some of the alternative readings of the lines we have been reviewing is that, if he is somewhat suspicious of these, he is even more aggressively suspicious of the kind of line that would result from scanning lines 3, 8-9, 11-20, 25-26, 29-31, 44, 47, 49, and 51 as I suggest, because to do so would legitimize the dreaded trocheespondee combination, and Attridge does everything he can to scotch this metrical figure, focusing at length (230-39) on lines like these by Pope and Shakespeare:

(55) Damn with faint praise, assent with civil leer (56) For precious friends, hid in death's dateless night

It disconcerts him that "faint" in (55), which ought to receive not only a stress but a beat, must be "subordinated" to "praise" in this formation. Later, in an article in Eidos, acknowledging that "Metrical subordination has remained a part of my published account of English verse with which I have never been very happy" (3), he formulates for such combinations, which he finds disturbingly numerous, a rule he calls "double-offbeat demotion":

A stressed syllable, or an unstressed syllable and a stressed syllable (in either order), or two stressed syllables, may realize an offbeat between two stresses, or after a line-boundary and before a stress. (5)

It is a measure of Attridge's willingness to reconsider some weak points in his theory that he can now accept some optional readings that use "metrical subordination" to achieve what traditional metrists would call a trochee and spondee, though he oddly attributes such metrical disturbances not to the pattern of stresses and beats involved but to the double offbeat. Such readings, he warns us in a phrase reminiscent of the worries of Halle and Keyser, "may push the line to the borders of metricality" (233). An interesting line in Hamlet "takes one to the limits of regular metrical form" (234). Earlier, another line, for another reason, "totters on the edge of unmetricality" (185). He admits that "the specific conditions of a particular line . . . can induce any monosyllable or dissyllable to lose some or all of its stress" (235). Yet, rather than tolerate lines that are similar to (55) and (56) but run a "spondee" over punctuation, he resorts to the fiction that expert writers like Browning, Milton, and Spenser preferred to hear lines of six beats where their verse form called for five (237). I give my scansion above the line, Attridge's below:

(57) Wretched man, wretched tree; whose nature weak (58) Weep no more, woeful shepherds, weep no more(4)

For one line, his rigidities lead him into this unseemly passage:

(59) The deep groves and white temples and wet caves (Browning, "Pauline")

One of the three adjectives would have to be metrically subordinated to achieve a five-beat line, but as all are parallel they all demand the same degree of emphasis. To pick one out arbitrarily would be invidious, and a momentary expansion of the metre to accept six beats is the only linguistically acceptable solution. (238; my emphasis)

Such language makes it sound as if we are talking about human rights and not metrical stress. To argue this way ignores the hardly disputable fact that poets like Browning did not momentarily expand a meter because they could not think of a way to say what they wanted to say within it. Attridge's argument postulates some wholly fictional rule about the English language's need to observe an egalitarian balance among adjectives, which ten minutes of listening to any native speaker should show is dubious. If you speak Browning's ten syllables quickly and "distinctly" - Attridge's word about other combinations (109) - but without much variation in stress, you can keep the sense of a more or less alternating rhythm and avoid worrying over whether you have violated any linguistic or metrical rule. In some puzzling lines Attridge allows for "rhythmic blurring" (238 and passim), his own version of what earlier metrists used to call "hovering accent" (182). But too often he seems bound by his frequently expressed insistence that the language has to follow some supposed rule of pronunciation. In discussing metrical subordination in the pattern discussed above (his "double-offbeat demotion"), he notes, to his surprise (232, and again in Eidos [4]) that many poets use it; it "has therefore become one of the available alternatives enshrined by convention at the start of the iambic line, in spite of its potential disruptiveness" (233). The obvious inference is that poets and their readers have not usually found the pattern as disruptive as Attridge does, nor have they needed, in the lines cited above, to invent an extra beat. Clearly, he ought to recognize that his difficulty in hearing the tune is not a sufficient reason to excommunicate such lines or to regard them as marginally metrical. Rather, it is a reason to become suspicious of his own methods.

As for the democracy he posits as existing between parts of speech that look as if they had equal weight, I might note that at one time I collected instances in Tudor poetry, and in Sidney, Marlowe, and early Shakespeare, of lines that began with repeated monosyllables. They were very frequent, and yet this was a time when the poets, as almost everyone agrees, were writing a very regular verse, a verse that could be depended on to alternate the degree of stress from one syllable to the next. Gascoigne, the critic who stated most unambiguously that this alternating practice should guide poets and whose verse is still known for its regular alternating structure, uses "Pray, pray" several times in "The Steele Glas"; Surrey has "Flee, flee," in his translation of the Aeneid; Sackville writes "Cure, cure" in his "Induction" to The Mirror for Magistrates; Marlowe uses several phrases that way, most memorably Faustus's "See, see where Christ's blood streams in the firmament"; and my list of Shakespeare's similar phrases in metrical lines, merely from early plays, includes twenty-two different repeated words: among others, Aye, Nay, Go, No, See, Down, Hark, Mount, Lord, Fool, and even Sans. Attridge's belief that such repeated words must be spoken with exactly the same force (255), though the first might be subordinated metrically to the second, seems less persuasive than that the poet who wrote them probably expected the second to be said differently from the first - more weightily or emphatically or sharply, as we are likely to do in our own speech. Perhaps it should be noted in passing that, while Attridge's sensitive analysis of accentual poetry suggests that he is probably an excellent reader of verse, among his hundreds of cited lines in REP (and his practice is similar in PR) only eleven are of iambic pentameter in Shakespeare's plays, and there are no such citations from any other verse drama. Whatever students may learn from either text, it is evidently not intended as a guide for actors.

In the Eidos article referred to above, Attridge makes a remark that almost gives the whole point away. He notes that "rhythmically, there is not a great deal of difference between 'You that poor Petrarch's long-deceased woes' [Sidney, Astrophel and Stella] and 'You that in Petrarch's. . .'" (4). Accordingly, Attridge now incorporates this variation into his rules (see the rule quoted above, p. 166) But if the rhythm is about the same, what does it matter if the word occupying the third syllable is a preposition or any of a group of adjectives, which Attridge lists in order of increasing importance and hence "of increasing rhythmic irregularity" (4), sweete, great, fine, young, dumb, shrewd? But Attridge's conversion is only partial: when the variation requires the reader to hear it straddling two phrases, as in 57-59 above, he still thinks the correct solution is to hear these as six-beat lines, no matter how strange it seems that Sidney or Spenser would ever toss such long lines into a steady stream of pentameters. Traditional metrics should have no difficulty with these lines, easily understanding that the second "foot" is a spondee,(5) unusual only in that it bridges two phrases, something that Donne's verse does habitually.

The root of the difficulty here is clearly Attridge's unwillingness to agree that in reciting verse the voice may ever tilt the lines a little in order to augment the stress on syllables that in prose or speech would not attract it, or to diminish the stress on syllables that in prose or speech would receive it.

(1) He can accept ?? / (is as a death) because the second syllable, having no competition for stress, is perceived as a beat, though the voice does nothing extra to attract it (REP 168).

(2) He can accept ?? / \ / (The long day wanes) because this pattern often occurs in actual speech (168).

(3) He accepts most forms of / ?? \ / (Damn with faint praise) reluctantly because this variation occurs so often that he has to acknowledge it as a conventional form of metrical subordination.

But where the stress pattern is ?? \ / he cannot accept the idea that it might sound very much like ?? / \ / (On whose fresh lap) or / ?? \ / (Through the thick corn) - i.e., like milder versions of (2) and (3), where in both cases the beat in the first foot falls on an apparently weaker syllable. Yet these differences are often slight, and it would seem that a sure way out of the awkwardness he senses is to acknowledge that combinations like these can often, in effect, masquerade as (2) and (3) in traditional English verse.(6)

For free verse, Attridge, recognizing its fundamental difference from metrical verse, offers in PR what seems to me a useful but incomplete strategy of analysis. He is aware that the term covers many different varieties of verse and he is sensitive to the kinds of rhythmic movement that can be found in poets like Pound, Eliot, Williams, Rich, and Lawrence. But here, too, one may object that his insights - into the way phrases move us through a poem - tend to freeze into categories (movement toward, movement away from, a static moment (not a happy term), and moment of arrival), analytical terms (statement, extension, anticipation, arrival), and a graphic notation of capitalized abbreviations (ANT, ARR, STA, and EXT), to the point where he has students scanning for these phenomena as they scan lines for their metrical form. This procedure is interesting; some kind of phrasal or syntactic analysis helps us see why some phrases have special force. But this does not seem essential or crucially helpful, as metrical scansion is crucially helpful to our hearing of rhythm. Without it, we might go astray; without this sort of phrasal scansion, we still have no problem reading the lines and sentences. Metrical scansion needs to be graphic; this does not.

Analysis along these lines, conducted with Attridge's usual intelligence and lucidity, can illuminate the way different poets tend to construct their sentences, speeches, and verse paragraphs. So, one might add, would traditional ways of diagramming sentences or otherwise inspecting the grammar of poets. When we classify phrases according to this four-type system, we lose the possibly greater clarity and interest we might get from knowing just what sorts of grammatical phrases we are talking about - participial, adjectival, vocative, nounal, predicate, or whatever. It might help more to know whether a writer's sentences are periodic or cumulative, hypotactic or paratactic, whether they break with the line-breaks or against them, whether the poet uses sentence fragments, parallel constructions, rhetorical balances and contrasts, or unbalanced or inverted structures, and so on. Attridge may respond that we are still free to notice these things, too, but it remains unclear why the four elements of phrasing he highlights here deserve the centrality he gives them. In any case, it seems a cumbersome procedure to inflict on students, and it implies that the material being subjected to analysis - the movement toward and away from some kind of central assertion, has something like the same kind of audible patterning that metrical verse lines have. But it does not; its patterning is visible, and it bears the same relation to audible English metrical poetry as Attridge, in Well-Weighed Syllables, once pointed to as existing between that poetry and the visually apprehended quantitative verse of certain sixteenth-century writers, for whom "quantity" was something not perceived in the sound of words but "found in the minds of the Elizabethans, trained in the application of a complex set of rules to the graphic embodiment of lines of verse" (160).

3

Compared to other great poets in English, such as Spenser, Marlowe, Shakespeare, Donne, Milton, Pope, Tennyson, Hopkins, Whitman, Hardy, and Yeats, critics have often thought of Wordsworth's verse as relatively pedestrian, undistinguished, and undistinctive. Saintsbury's judgment, quoted by Brennan O'Donnell as typical, has been shared by many readers: "In no great poet does prosody play so small a part" (6). But O'Donnell argues persuasively, in The Passion of Meter, that Wordsworth, though certainly a metrical "conservative," studied closely the metrical craft of his great predecessors, thought intently and with considerable originality about the part meter plays in significant verse, and used meter in imaginative ways to amplify and reinforce his narrative and lyrical themes. The passion of meter interplays with the passion of syntax to form a strong verse language, which for Wordsworth either moderates the life-passions that form the subject of poetry or, in more contemplative modes, provides that passion. O'Donnell's scrutiny of the poet's many statements about meter makes clear how important Wordsworth thought it in the composition of verse and what varied uses he found for it. Any reader will find much wise and patient discussion of the art of prosody in this subtle and thorough examination of Wordsworth's metrical theory and practice.

In Wordsworth's poetry, O'Donnell argues, meter usually figures as an instrument of balance and counterweight to other elements in the poem, or else as a means of reinforcing or signaling certain tensions that are at work. In the first instance, it may operate to give dignity, through an elaborate verse form, to a narrative told in colloquial diction; "The Sailor's Mother," for example, "confronts the reader from the outset with a markedly incongruous relationship between humble subject and elaborate stanza form" (53). Similarly, in "The Thorn," Wordsworth's "principle of similitude in dissimilitude" develops a contrast "between slow speech and rapid meter" (66). O'Donnell also shows well the gradual process by which Wordsworth sloughed off the antithetical style of Pope's heroic couplets and learned to write couplets in a style of his own, considered "harsh" by early reviewers: lines that "tend to focus attention on the phrase, rather than the line, as the chief structural unit" (108). O'Donnell makes a case for the great metrical variety of Wordsworth's poems in Lyrical Ballads (115-78) and claims that in many poems Wordsworth's choice of meter and particular use of it are purposeful, subtle, and effective: as O'Donnell puts it in his rather scholastic language, "even the most ostensibly simple and predictable of Wordsworth's stanzas will reward careful attention to subtle tensions between the metrical frame and diverse realizations in actual sonic and rhythmic impulses" (154). In another long chapter, O'Donnell analyzes Wordsworth's increasingly skillful use of a Miltonic blank verse, especially in such different modes as "Tintern Abbey" and "Michael," on both of which his commentary is excellent.

This analysis is managed with considerable grace and eloquence, though, given Wordsworth's vast poetic production, O'Donnell does not attempt to examine all of it, or even a large sample. He deals mainly with some early poems, especially "An Evening Walk" and "Descriptive Sketches," with the stanzaic verse of Lyrical Ballads, and with the same volume's blank verse. That means that not much, if anything, is said about Wordsworth's later stanzaic poems, his odes, his hundreds of sonnets, or the vast production of blank verse poems that include The Prelude and The Excursion. Instead, he focuses on the early poems in which Wordsworth was establishing the principles that guided his metrical art for the rest of his career. This is a sensible decision, though one cannot help feeling disappointed at so little reference to most of the major poems on which his reputation rests.

O'Donnell likewise recognizes the importance of the poetry that most influenced Wordsworth: Pope for his early verse, then Cowper and other eighteenth-century poets, and Milton for the blank verse of his middle and later career. He notes, too, Wordsworth's interest in taking his place among the pantheon of English poets, including the Elizabethans and especially Shakespeare, but he says little about any specific debts he owes them. He frequently observes Wordsworth's use, for example, of line-ending polysyllables (like piety or humanity) that give the last beat of a pentameter line a rather weak stress, and he notes the incidence of this feature in Milton, but he never connects it to earlier poets like Marlowe and Shakespeare, in whose work it is habitual. Nor does he give credit to Shakespeare's late dramatic verse for having suggested to Wordsworth the advantages of closing one verse paragraph (or, in Shakespeare, one speech) in midline and beginning the next one there, a detail of verse architecture that Wordsworth was the first nondramatic poet after Shakespeare to use as an occasional resource in his major blank verse poetry.

O'Donnell's sympathy with Wordsworth's creative prosody is no doubt responsible for the clarity and understanding with which he sets forth its principles. But it also constitutes one major difficulty I found in the book - namely, that O'Donnell is insufficiently critical of Wordsworth's practice. He shows quite brilliantly how the poems reflect and embody - but also how they seem to justify - Wordsworth's theory of meter. His precise analysis is often concerned with what he admits are "minute effects" (224, 152, 206), and be occasionally writes a suggestive phrase like "insofar as it is effective" (89), or "this relatively wooden use of metrical contrast" (90), but usually he is content to explain Wordsworth's rationale for choosing a particular meter or for using some "disruptive" metrical variation, and the rationale is often so intelligent, and so intelligently explained, that we are invited to infer that the tactic is successful. But readers of Wordsworth's day and ever since have found that the combination of diction and prosody in "The Idiot Boy," for example, does not work, i.e., does not produce that "pleasurable excitement" that Wordsworth was hoping for, and O'Donnell's expert presentation of Wordsworth's rationale for it (163-69) is no more convincing today than Wordsworth's was in 1800. The best one can say for a lot of O'Donnell's defenses of mediocre poems is that they are very eloquent, but they persuade us that Wordsworth did think this way about the poems, not that the poems have any notable merit. Useful as this book is and helpful as its author's studies are, they might have been even more valuable if O'Donnell had clarified for us where Wordsworth's theory works and where it does not, if with some critical acuteness along with his penetrating theoretical research he had shown us why, where, and to what extent the poems Wordsworth constructed so carefully fell short of his ambitious plans for them, and why others succeeded. O'Donnell does not really account for the great difference in merit between, for example, "The Idiot Boy" or "The Mad Mother" and, say, the Lucy poems.

As the book continues, in fact, the author is drawn more and more into graceful overstatements about the success of Wordsworth's metrical procedures. The last two lines of "A slumber did my spirit seal" can

bring into sympathy the physical impulses of a reader in such a way that he may feel in the approximation and departure from the basic 4 x 4 rhythm both the power of a force extrinsic to individual expression and the human impulse to resist or escape such power. (154)

Ultimately, meter is conceded a power to register awarenesses that are fully cognitive, as when we are told that "Hart-Leap Well"

incorporates the very impulses that it seeks eventually to rectify, forcing a recognition at the level of minute and pleasurable rhythmic and sonic patterns that human pleasure and pride is indeed . . . inextricably intertwined with "sorrow of the meanest thing" (and not only the meanest) "that feels." (163; my emphasis)

By the last page the grandiose claims take the following form, which can hardly be more than a pious hope, if we understand it to include all of Wordsworth's verse, the mediocre along with the truly impressive:

In Wordsworth's metrical art, each sound, syllable, word, verse, stanza, or verse paragraph is represented as simultaneously an expression of spontaneous impulse and a fulfillment of a single controlling system of organization. (248)

And all his poems "may be understood as magnificently and diversely articulated expressions of a single system of 'tones and numbers'" (248). They are all, that is, "Wordsworth," as Allen Ginsberg used to think of all his lines as "Ginsberg." But this is undiscriminating praise, and many readers will long for a manner of criticism that will distinguish the really good poems from the clunkers.

But if O'Donnell's book, fine as it is in its clarification of Wordsworth's metrical principles and aims, suffers from this excessive deference to Wordsworth, it suffers equally from its excessive deference to his principal metrical mentor, Derek Attridge. He adopts Attridge's system absolutely, with all its elaborate terminology, from "initial inversion condition" (16) to "stress-final pairing with linkage" (233), and with those quaint schoolmasterish devices of "promotion" and "demotion" to explain the way Wordsworth places minor or major syllables in major or minor metrical positions. This wholehearted acceptance commits him to Attridge's insistence, modified in the theory of both writers but not in their practice by the flexible notion of "optionality" (33), that two minor syllables followed by two apparently more important ones must be treated as an occasion for lurching through the first two in order to get to the second pair, as in the numerous examples given earlier. Wordsworth's later feeling of unease at his own line

(60) Impressed on the white road; - in the same line

is used (32-35) to substantiate the view that this poet must have spoken such lines as Attridge and O'Donnell evidently wish us to do, jumping over the minor syllables to stamp on the major ones.

Impressed on the white road; - in the same line

But whether Wordsworth actually did so remains in doubt. In a letter to John Thelwall, Wordsworth called this "the most dislocated line I know of in my writing" (qtd. on 32), but we do not really know whether he thought it dislocated because he heard in it the lurching movement that Attridge and O'Donnell hear, or because in reciting it - in "chaunting" it, according to Hazlitt - he adhered to the metrical pattern and gave to each "the" a heavier stress than he would have done in saying such words in conversation, so that the meter "dislocated" the normal conversational rhythm. O'Donnell does not even consider this second possibility; it is evidently not an option. At the same time, following Attridge, he argues for a principle of "optionality" that allows us to be aware of alternative readings of many lines. In "My heart leaps up," for example, he suggests that we may read the verbal phrase iambically, stressing "up," or (in effect) spondaically, giving a stronger emphasis to "leaps" as well as to "up." This is perfectly sensible, but O'Donnell, in this case as in many others, exaggerates the difficulty of the line, claiming that "leaps" "dislocates the verse" because it is "in an offbeat position" (36). But such variations happen all the time in traditional iambic poetry, and their effect is not to dislocate a line but to vary the iambic pattern a little.

Despite the apparent generosity of their principle of optionality, Attridge and O'Donnell never concede that it is acceptable in a reading of iambic verse to give to a very minor word in stressed position enough weight to allow it to retain our sense that it occupies a metrical beat when it is competing with a word that belongs to a more privileged grammatical category. English phonology, they claim, insists that when we come across such lines we perform the metrical lurch. O'Donnell's book is full of such lines, and his procedure is usually to decide on the lurching scansion and then to explain its significance. But if the scansion is wrong or dubious, the explanation of its significance is bound to be, too. If Wordsworth and many of his readers did not hear the lurch that is second nature to Attridge and O'Donell, then the elaborate inferences these critics draw from their analysis of it will not be valid. The "disruption" that O'Donnell hears in such lines will not be as disruptive as he claims, and the options that he offers a reader can be extended much further or in different directions than he appears willing to grant. Here is how he scans three lines of "Yew-Trees" (213-14):

There is a Yew-tree, pride of Lorton Vale, Which to this day stands single, in the midst Of its own darkness, as it stood of yore

O'Donnell does not offer as an optional reading an "inversion" in the stress pattern of "There is," though my own voice finds that sound at least as natural as stressing "is." For the second line, he judges that its opening "stress-final formation" is especially "disruptive because of the relative weakness of 'this'" (214), but the simplest way of solving this problem either does not occur to him or is forbidden by some rule:

Which to this day stands single, in the midst

Instead, he offers this option, then dismisses it for complex reasons:

Which to this day stands single, in the midst

He prefers, then, to believe that Wordsworth started two successive lines with almost identical lurching rhythms and that he did so in the interest of preserving "rhythmic complexity." In this case, as in most others involving the lurch pattern, O'Donnell somehow manages to claim some kind of subtlety for what he argues are Wordsworth's metrical procedures and to ignore the banality or clumsiness of the rhythm he is trying to persuade us is Wordsworth's. All too often he makes large claims for metrically unexceptional verse, and Attridge's terminology can sometimes sound ludicrously clinical, as when "Stoops her sick head, and shuts her weary eyes" is said to illustrate a particular form of the "initial inversion condition" (94).

Occasionally, O'Donnell is willing, like Attridge, to believe that Wordsworth may have preferred to scan some lines as hexameters rather than subordinate one stressed syllable to another (94, 214). He persuades himself, too, that "[f]ive-beat" and "pentameter" "may be used . . . interchangeably" (16); so may "iambic" and "rising rhythm." These are seemingly harmless points, but what they imply is that lines may be classified as iambic by their deep structure, not by their surface rhythm. It means that no matter how many double or implied offbeats, say, they include (e.g., x / x x / x x / x x / x x / or x / / / / x/) they remain perfectly iambic, that Robert Lowell's five-beat son - nets in History and Notebook are metrically of a piece with Sidney's and Shakespeare's.

Elsewhere, alert to the significance of midline punctuation, O'Donnell not so alertly identifies such punctuation as requiring pauses. This is a reasonable procedure when one wants to establish statistical norms for poets over great stretches of their poetry, as critics like Ants Oras and Marina Tarlinskaya do; in such cases readers will understand that "pauses" does not necessarily mean pauses but may mean only syntactical breaks. But when one discusses individual lines one ought to be on one's guard against assuming that every piece of punctuation actually requires a pause in delivery. In "Tintern Abbey," for example, it is hard to imagine a delivery of these lines,

If this Be but a vain belief, yet, oh! how oft,

that would actually pause after "belief," after "yet," and after "oh!" and still keep any sense of pentameter rhythm. But O'Donnell does not shrink from aestheticizing the "triple pause" as helping "to create the impression of a mind engaged in a process of discovering - and making - the meanings it articulates" (192), as if one could not claim something similar not by counting the punctuation marks but by gauging the syntactical self-interruption and the surprising exclamatory way of resolving a sentence begun with an if-clause.(7)

In short, many of O'Donnell's readings of particular lines are questionable, and the claims that he makes for the significance of the minute variations that he discerns are thrown into doubt if he is not describing these minute variations accurately. His reliance on Attridge's system is almost a test-ease for that system, and, in the judgment of this reader at least, the system does not come out very well. Both O'Donnell and Attridge have many slighting things to say about traditional metrics and metrists; Saintsbury especially comes in for O'Donnell's scorn, and he often writes as if all later metrists who still think there is some validity in the traditional metrical analysis had never advanced an insight that might take us beyond Saintsbury's imprecise formulations and sweeping judgments. But it is still possible to read some lines as including pyrrhic-spondee combinations without behaving "as if each two-syllable unit were detachable and modifications within a 'foot' did not affect the structure of other pans of the line" (34). Detachable, indeed! Who ever believed that? Whatever their faults, many traditional metrists have written of meter's subtle but elusive effects with considerable insight, in the hope - it is O'Donnell's, too - of learning more about how poems and poets work.

4

Alan Holder would disagree. His book, Rethinking Meter, brings charge after charge against traditional prosodists who hear in lines of most English verse before 1900 a distinct metrical structure. Making a show of reasonable inquiry, which cannot really survive the wisecracking and sarcasm that distinguish his polemical style, Holder assembles their more extreme statements, reports their disagreements with one another, and notes their inability to settle even elementary issues in the study of poetic meter. That the field is filled with casualties is beyond debate, and Holder has a good time picking at the wounds. He has read a great deal of prosodic criticism, and he writes clearly, though with a poisonous pen. His first three chapters survey the follies of traditional metrists as they go about their absurd and irrelevant enterprise. The fourth exposes their misguided attempts to analyze free verse as if it were still metrical. The last three show us Holder's idea of the kind of "phrasalist" and intonational analysis he thinks might be far more useful for our study of the sound of poems than metrical analysis can be - useful for free verse but also for traditional metrical poetry.

Holder is unsparing in his contempt for traditional metrical study and the academic critics who practice it. He is hard on their inconsistencies, their arrogance, their over-reading, and often with good reason; but a large part of his irritation with them is for hearing more than Holder hears. Where they hear something metrical going on in poems, Holder claims to hear nothing but phrasing: words, words, words. The view he evidently takes is that those older poets who produced hundreds of thousands of ten-syllable lines heard them only as sequences of phrases; they never heard a meter at all. Or else he thinks that even if they did, the meter is not of much sonic importance compared with the rhythm of phrases. All those prosodists who have thought the great strength of much English poetry was its way of letting the phrasing somewhat alter the meter and letting the meter sometimes tilt the phrase are considered by Holder to be quaint enthusiasts or academic bullies, essentially wasting their time and the time of their students, when they might be focusing on phrasal arrangements in order to tell us something really useful about verse.

What is strange about this position is that it apparently never occurs to Holder that he might be mistaken, that all those fairly bright scholars and critics who say they have been hearing something metrical in poetry that he does not hear might really be pointing to something genuine. Almost alone among modern readers of older poetry, he hears none of its meter and he assumes, therefore, that it is not there - rather like a blind man telling us that looking at a sculpture is of no interest; the only way to judge it is by touch.

Part of the problem is that Holder does not want to hear meter. Meter to him is a "tyranny" (134), a way of imposing uniformity on sentences and phrases, of reducing the interesting language of poetry to "a system" (65), to an "a priori model" (233), to an "arithmetic regularity" (116), to a "totalitarian 'scheme'" (65). He jeers at metrists' "hunger for a system" (64), at their "prosodic foot fetishism" (29; also 63), at the "Mystique of Meter" (22), at "that Napoleon of notions, the foot" (30): "the poem is treated, in effect, as a mechanism for grinding out clonelike lines" (48), as if it were the intention of metrists to take the fun and the life out of poetry. Doggedly skeptical about metrical pattern anywhere, he frequently refers to metrists' findings as "the alleged meter" (this or similar phrases on 38, 53, 57, 63, 85, 182, 199, etc.) In Holder's highly colored account, metrists are conducting a kind of war against lines of poetry, to "bring them to heel" (84), "to tamper with the stress values of a line's syllables" (84), or to make them "toe the mark" (85). Willfully dishonest and distortive (82), we persecute the poor line of verse, which "never had a chance of breaking out of the iambic mold" (88). Our aim, he charges, is to make feet all alike, to save the system from irregularities (47). Determined as we are to find feet even in the most unlikely phrases, we ignore performance and find what we are "disposed to find" (83). We bend speech "patterns into a presupposed shape" (266); we allow feet to "dismember words" (269n35). Intent on "making the line safe for iambicity" (90), we "capture" lines "for the cause of iambicity through the intervention of . . . strategems," because "all is fair in love and prosody" (89).

On the other hand, our zeal for discovering patterns of metrical tension leads us to engage in "indiscriminate variation-fondling" (63) with the result that "iambic pentameter is awash in variations" (61). It is part of Holder's auditional disability, after all, that if he cannot hear metrical patterns he cannot hear metrical variations from the pattern. So he interprets our seeking both pattern and variation in verse lines as "yet another instance of a meter-minded prosodist wishing to have things both ways" (63). When we do venture to enjoy the "creative equilibrium" (94, a term of mine which he quotes) that a poet like Shakespeare in his late dramatic poetry sets up between the line and the phrase, Holder somehow takes this to mean that we see phrase and foot as "inimical" (93, his word, his italics), as "a simple, mutual enmity" (256n31). This reading is understandable, because if he cannot hear beats he cannot hear the interplay between a verse line's metrical pattern and its phrasing. He thinks it peculiar that anyone should hear in the language of poetry a meter and a phrasing at once, as if musicians do not do something similar all the time, staying conscious of the beat and yet giving full attention to the melody coursing through the measures. But musical analogies make Holder extremely uncomfortable (53, 54, 96, 100-01), a natural enough response for a theorist who cannot hear a metrical beat.

For most traditional metrists, including Attridge, the disparity between the metrical pattern of a line and the speechlike way of saying it generates some degree of tension, but for Holder this is a fiction. "I would argue," he asserts pugnaciously, "that in performing the poem, what we say is what we get, and all that we get" (52). He hears no tension from beats or meter; verse is like prose in its rhythms. This is a defensible position for a critic of much free verse, and the basis for many of Holder's judgments. His hostility to meter and metrists, though it may spring from his metrical deafness, owes something to his feeling that metrists and the very concept of meter are antagonistic to free verse, and to his wish to defend it from its prosodic enemies. He has a point, though he characteristically overstates it. He is right in noticing that many metrists try, often absurdly, to find the ghost of traditional meters in free-verse poems, as if they could not bear to let free verse be free, and his chapter on "The Haunting of Free Verse" is easily the best in his book. But he warms to the suggestion of one free-verse scholar that traditional verse, too, "ought to be read without the application of traditional scansion" (114), that is, without a sense of its meter.(8) And this overreaction to the follies of metrists in reading metrical pattern into free verse results in the equally flawed position that governs Holder's later chapters. If it is wrong to read free verse by the metrical standards of older verse, it is equally wrong to read older verse by the nonmetrical standards of free verse.(9)

Disbelieving in meter, Holder never uses the word "beat" except when he quotes metrists with whom he disagrees. One measure of his hostility to meter is that in summarizing Attridge's prosodic theory he misses its most striking feature, the analysis of virtually all English poetry except iambic pentameter as accentual, a view that leads Attridge to develop a notation based entirely on beats and offbeats. Holder mentions this notation but regards it as nothing more than "a terminological shift" (65), apparently not realizing the profound difference between a traditional scansion based on stressed and unstressed syllables (accentual-syllabic) and a scansion based on beats (accentual). Holder decides Attridge's system is just "essential traditionalism" (67). This misreporting is typical. Either through misunderstanding or deliberately, he rarely makes a good-faith effort to explain, without prejudice, the aim of a metrist's writing before he begins his work of demolition. Even when he occasionally comes close to acknowledging vaguely that there might be such a thing as meter at work in much poetry - after all, he says he wants to rethink it - it does not emerge as a rhythm that he hears. He wants only to redefine meter as phrasing, which of course is something else. It may be that, not hearing beats in verse of any kind, he cannot grasp what Attridge and other metrists are about and so cannot give a fair or judicious account of accentual or accentual-syllabic verse - indeed, cannot resist ridiculing anyone who professes to hear metrical patterns in poetry.

His only effort to deal with accentual verse comes when he accuses Wimsatt and Beardsley of wrecking the natural phrasing of Blake's line, "Ah, Sunflower weary of time." It is revealing that Holder quotes only the first line of Blake's poem, so he does not give a reader a chance to hear how steadily Blake allows a fairly regular beat to be heard throughout the poem. Here is the first stanza:

Ah Sun-flower! weary of time, Who countest the steps of the Sun, Seeking after that sweet golden clime Where the traveller's journey is done

Wimsatt and Beardsley point out what would seem elementary to most of us, that the beats in this highly accentual stanza can be heard on "Sun-", "wea-", and "time" (and later on "count-", "steps," and "Sun," etc.). Holder accuses them of ignoring the secondary stress on "-flow-" as well as the expressive "Ah," and of breaking into "word boundaries and syntactical groupings" (27) with their villainous foot divisions:

Ah Sun | flower wea | ry of time,

Clearly, he does not grasp the elementary point that in accentual verse some otherwise stressable syllables do not occupy the beat. That does not mean the syllable disappears; it gets said pretty normally (in my saying), but you do not hear a beat on it. But Holder views what he calls, in discussing another poem, this "supression [sic] of normal stressing" (27) as taking us to "the heart of the morass that is meter country" (26-27).

The traditional notation that Wimsatt and Beardsley use is troublesome to many readers, and their tone at times is admittedly magisterial, but their markings constitute a graphic means of pointing to a pattern that is certainly audible to many readers. Even those who find it disturbing to suppress a portion of the possible speech accent in "Ah," "-flow-," "Seek-," and "gold-" are likely to recognize the recurrent beat defined by the stresses on the other major words and their stressed syllables. Unless you do not want to hear it, you cannot miss it. Holder misses it.

His missing it makes his intentions clear enough. It is not that he simply proposes we study additional dimensions of sound (those of phrasing and intonation) that metrists have traditionally ignored or undervalued. He wants us to "disestablish the standard notion of meter" (16) and to recognize what a tyrannical and wrongheaded enterprise metrical analysis has been from the beginning. "The concept of meter stands between us and our poetry" (16), he tells us, and all the evidence of this book, including its stumbling efforts to rethink or redefine meter, makes clear that he wants to scrap it. But his inability to hear any kind of recurring rhythm in poems, and his sense of outrage at the metrists who keep "imposing" it on innocent lines of verse, lead him apparently to take the unusual view that it is the metrists, not the poets, who have put the meter in poems. With a few exceptions here and there, the poets have not scanned their poems or placed foot dividers in their lines. They have simply presented phrases and lines to their readers, who should take them as they come and not invent metrical restraints to hem them in and make them "toe the mark" (85). But of course this is nonsense. Metrical scholars may disagree with one another, may propose unconvincing scansions and theories of scansions, may badly misread what poets have written (see the next section). But the evidence is clear that almost all older poets writing in English typically wrote lines that kept some sort of recurring beat, and a critic like Holder who blames the metrists for the poets' metrical patterns is trafficking in absurdities. (When he comes to offer his own pitch-scansions, however, he usually suggests that he is merely recognizing patterns in the line that the poet has put there.) Holder never quite says he does not think there is such a thing as meter, and he has one sentence that suggests he understands that it must exist ("There is, after all, a centuries-old tradition of distinguished metrical poetry" [127]), but his redefinition of Shakespeare's meter in Sonnet 116 (see below) is pretty conclusive evidence of his inability to hear anything like what most earlier listeners have heard.(10)

Surely a metrical critic so scornful of others will see to it that his own work is free from errors and contradictions? Unfortunately, Holder's book not only frequently misrepresents other critics' views, but it is full of mistakes and riddled with misunderstandings. He thinks that a feminine ending constitutes a foot (27-28); on the other hand, he does not spot the feminine ending in the line, "That looks on tempests and is never shaken," but thinks "shaken" scans as a trochee (180). He does not know that "spirit" was often pronounced as a monosyllable in the Renaissance (27-28). He misdefines Hopkins' term "out-rides" as "extra syllables within a line" (80); he does not seem to know that Auden's poem "Streams" is syllabic (77). He does not understand that Attridge's demotions and promotions often do not, in Attridge's view, involve any change of stress, or that most of the adjustments that metrists make for metrical variations are slight. He cites Josephine Miles's interest in "phrasal poetry" in support of his own position (157-58), but he seems not to know that she is using the term in an entirely different sense.

As his almost obsessive comments on the subject make clear, he does not understand that graphic foot dividers are an analytical convenience and do not cut into the sound of phrases in some artificial way. Inconsistent as ever, in his own late-chapter effort to show how he would scan lines, he gives us this sort of thing (230). The poem is by Sylvia Plath; the levels are levels of pitch:

Then                                   blue
                  sub
          the           stanceless

For King Lear's line, "Why should a dog, a horse, a rat have life," he proposes this arrangement of leaping pitches:

        dog,                 horse                rat
a                     a                      a

One may pity the poor actor who, under Holder's direction, is not permitted to vary the pitch of the three words denoting animals, and is obliged to come back to the low pitch-level of "a" every time. Lear's line of nevers (214) is shown in the same obsessive two-tone:

Ne          ne           ne           ne          ne
     ver          ver          ver          ver         ver!

Apparently, words "dismembered" (269n35) this way please the same sensibility that is offended by foot divisions and frequently pleads for "respecting word integrity" (180). And these two-part inventions begin to look suspiciously like those "clonelike" lines and feet (48), those "countable clones" (116) he hated in other people's scansions.

Holder is not only prescriptive in such pitch-scansions; he is much more prescriptive than most metrical scansions are, which legitimize a range, even a class, of possible readings, whereas Holder's are so precise, so particular, that they seem to legitimize only one. Most scannings of most verse lines permit a reader to vary the pitch, the speed, the phrasing according to the reader's' judgment, but Holder's precise delineation of pitch-levels seems, as he admits, to map a very personal way of speaking or hearing lines.

One cannot help noticing that Holder is guilty of many of the charges for which he indicts traditional metrists. As these pitch-scansions show, he uses a notation for free verse that breaks up the words of a line at least as much as traditional foot dividers do, indeed tosses the words around the page. He is angry that metrists break up the line with foot dividers, but intonational patterning, he claims, gives us "a basis on which to break up a given utterance into parts . . . intonation units" (203). If prosodists are always "playing the old game of looking for foot-defined variations and trying to link them to an expressive effect" (99), Holder is doing the same with his fanciful pitch-scansions. He complains about Jonathan "Holden's animus against free verse" (121), but shows at least an equal animus against traditional metrical analysis. He indicts traditional prosodists again and again for "fuzziness" and sloppy thinking, but his own analysis is equally flawed. Noticing "a pattern of increasing concreteness in the sequence of nouns" in a line, he conveniently omits one of the four nouns in the sequence (89). He misquotes a line he is analyzing (224). He accuses Marjorie Perloff of using weak instances to condemn a whole field, but this is his own strategy throughout the book. He accuses Annie Finch of finding dubious patterns in Whitman's verse (117-19) but is equally fanciful in detecting in a Whitman line an effort "to dramatize the notion of a 'cluster' by clustering stresses" (119). His rhetoric is always loaded: "if ordinary pronunciation has to be deformed, so be it, say the system-mongers" (47); "Epstein and Hawkes have turned out to be metrical good old boys" (33); "her study keeps us stuck in the swamp of traditional scansion" (70). He is willfully distortive of the views of those he condemns, as when he pretends to think that Wimsatt and Beardley's strictures about syllables carrying greater or lesser degrees of stress in duple verse "mean the dissolution of the anapest" (25) in triple verse. Throughout his book he is just as assertive and absolute in his judgments of what is "sane" and sensible as any of the metrists he attacks.

He charges traditional scanners with a "passion for abstract pattern" (44); the work of one represents "the apogee of allegiance to abstraction" (45); but he has to admit that "the designated relative pitch levels" of his own "intonational schematizings" "represent something of an abstraction (but of a lesser degree of abstraction . . . than the schemes produced by foot scansion" (229). But how so, one wonders? Holder does not explain. There is something in him of the child who thinks that his own transgressions are lovable. He despises metrists' tendency to prescribe readings, and then he prescribes his own: words set in lines must be spoken more slowly than "if they were set as prose" (140); and "I have to . . . insist that a pause be registered at the end of [every] line and that it be actual, however slight" (148). He decides that Lear's line of nevers demands "four distinct pauses" (214). In fact, he boasts that his intonational "profiles" give us "a truer representation of Lear's words, 'Thou'lt come no more' than what traditional scansion will give us, a pair of lambs" (216). But his raised profile,

                       come no
Thou'lt                                   more

besides being almost ludicrously unconvincing and unnuanced, with its mere two levels of pitch, deprives any rich-voiced actor of options that a more reticent traditional scansion leaves open. That is, despite all his reading of metrical theory, he does not understand that a scansion is not normally an analytical record of an individual reading but a quite generalized analysis of a class of readings within which a good actor or reader will find one that will be effective. It should enable performances, not script them.

Among the metrical critics whom Holder excoriates or mocks for their analytical absurdities are Saintsbury, Wimsatt and Beardsley, Shapiro and Beum, Turner and Poppel, Halle and Keyser, Epstein and Hawkes, Attridge, Tarlinskaya, John Thompson, Suzanne Woods, Dennis Taylor, Richard Cureton, Edward Weismiller, Paul Ramsey, Paul Kiparsky, Sandra Gilbert, Marjorie Perloff, Thomas Cable, Jonathan Holden, and myself. He has some good words to say about a few of these, but most of the time he does not try to give a fair and accurate account of what such critics are trying to do. Instead, he usually prefers to discuss their most extravagant, wayward, or unwary pronouncements and to omit any explanation of the context in which they were made. So, in scoffing at a little article I once wrote on the meter of the poem, "Shall I die?" which has been attributed to Shakespeare, he fails to cite any of the poem's lines that might have made my view of the poem's meter plausible; instead, he quotes that part of my article in which I deliberately cited some of the most bizarre lines. I did this in order to show that some lines of the poem were very hard to read as I was suggesting, and I was not applauding the poet for putting the reader in the position of having to wrench the lines' natural sound in order to keep them to the metrical pattern. I speculated that whoever wrote the poem may have been experimenting to see if he really could adhere to a cretic pattern for several stanzas, and if the result seemed odd in some places we are free to blame the poet. Holder gives his reader no sense of this context, and when he attacks other metrists, I either know or suspect that he has sometimes taken their statements out of context and attributed to them views that they certainly do not hold in the form in which he presents them.

When Holder speaks of dead metrists, he does not usually moderate his scorn; living ones he tends to approach with more courtesy before he starts cutting them to pieces. He devotes many pages to the refutation of my own metrical studies, and I should acknowledge that he has some good things to say about my Shakespeare's Metrical Art ("a formidable work, a learned labor of love. . . . I do not think traditional prosody has anything better to offer us" [81]). Nevertheless, he deplores my "mind-set" (81) and calls me "a stresser pure and simple" (255n27). His treatment of my work is distortive in almost every detail, full of willful misunderstanding and misreporting, and failing at every point to offer an accurate account of my claims and arguments. He attributes views, positions, and feelings to me that I do not hold (e.g., "The apparent trochee of the third food [sic] . . . appears to disturb Wright"(11) [91]) and terms that I do not use (demote, promote [91], "demonstrated" [89]); and his wisecracking phrases ("meter-maintenance scanning" [91], "The Principle of Disinterested Meddling" [90], "he who has declared the limits of phrasalism" [172]) establish the tone of jeering and sneering that is almost constant in his discussions of traditional metrists.

In the course of disputing my readings of Shakespeare's dramatic verse lines in Shakespeare's Metrical Art, Holder offers this scansion of Sonnet 116, "with only stressed syllables marked (intralinear phrasal divisions have been indicated by the use of extra spaces)" (176):

Let me not to the marriage of true minds 1 Admit impediments. Love is not love 2 Which alters when it alteration finds, 3 Or bends with the remover to remove. 4 Oh no! it is an ever fixed mark 5 That looks on tempests and is never shaken, 6 It is the star to every wand'ring bark 7 Whose worth's unknown, although his height be taken. 8 Love's not Time's fool, though rosy lips and cheeks 9 Within his bending sickle's compass come. 10 Love alters not with his brief hours and weeks, 11 But bears it out even to the edge of doom. 12 If this be error, and upon me proved, 13 I never writ, nor no man ever loved. 14

(176-77; line numbers added)

Holder argues at some length that the "forced pause" (178) he hears after "it" in line 3 results from the

semantic hitch at the point where "it" meets up with "alteration," requiring a pause at an awkward place, after a pronoun functioning as a subject, which normally one would be inclined to join with its adjacent verb as soon as possible. (Here, it seems to me, is a real source of"tension," that is, between the way we would normally handle a pronoun-as-subject and the way we need to handle it here.) (178)

Later, among other bizarre suggestions for the "handling" of pitch in this sonnet (he almost always gives disyllabic words different pitches for each syllable [222-24](12)), he even proposes a "raised pitch" for "it" in line 3 (222). As my students have testified for decades, and as every recording I know of an actor reading the sonnet confirms (the ones I have heard lately are by Simon Callow and John Gielgud), line 3 offers no difficulty to any normal reader. It moves easily and smoothly through the series of somewhat greater and lesser stressed syllables. I have never known anyone to have trouble with the line until Holder appeared on the metrical scene and invented grotesque reasons for not taking the line simply as it comes. Instead, he "handles" the line in a fashion that I think most people will regard as preposterous, involving as it does an extremely unspeechlike stress on "it" and a rationale that is at least as wayward as any metrical misreading his book points to on the part of traditional prosodists. Indeed, Holder's way of scanning this line may serve as a measure of how seriously his rethinking of meter is to be taken.

I could be wrong, but I suspect that hardly anyone who has heard the alternating syllabic pattern familiar in English verse from hundreds of thousands (or millions) of poetic lines is likely to think Holder's an adequate substitute for some traditional reading of such lines. And it should be clear that he is proposing it as a substitute scansion, not as an arrangement of the poem's phrasing that is intended to supplement a metrical scansion. The stresses in his version combine with the unstressed syllables to form what he calls a looser version of a meter he prefers (no wonder!) to call decasyllabic: 9-11 syllables; 4-5 stresses in a line; normally beginning with an unstressed syllable and "using at least one such syllable to separate stressed syllables" (175). But of course this is no meter at all. It might imaginably be a meter if his scansion came close to adhering to his definition, but for Sonnet 116 he finds four lines (4, 9, 11, 14) that do not even qualify as having 4-5 stresses (one has 7!); and even if the notion of using "at least one [unstressed] syllable to separate stressed syllables" could really help the ear to identify a meter, he violates that prescription thirteen times by my count in fourteen lines. Some meter! It is not a meter because it has no audibly recurring pattern of beats, but such a pattern is what Holder's ear is apparently unable or unwilling to hear. His scansion of Sonnet 116 is his only effort in 240 pages to scan any English verse of the kind long recognized as iambic. What is more revealing, his only attempt to deal with accentual verse, that other great mode of English poetry, comes in the course of his criticism of Wimsatt and Beardsley's reading of Blake's "Ah Sunflower! weary of time," and even then, as I have said, he tries to "handle" only the first line of it. To repeat, because it may seem unbelievable: In a book called Rethinking Meter, this author actually scans only one line of accentual English verse and fourteen lines of iambic pentameter!(13)

It is worth noting that, whereas most metrists are not troubled at finding eleven syllables in line 6 of Sonnet 116 because of the feminine ending, Holder rather congratulates himself on being able to reduce that number to ten. He argues that the word never "undergoes a standard contraction" (180) here (but apparently not in line 14). In fact, never appears sixteen times in the Sonnets and does not ever need to be reduced to a single syllable in order to make a line metrical. In Shakespeare's plays the word is much more often given its two-syllable pronunciation than it is contracted, at least if one relies on the usual metrical test. Holder here simply does not know what he is talking about. I often get the feeling that he has read much more traditional metrical criticism than traditional poetry, a charge that our next author will certainly not need to face.

5

If Holder's New Age scansions are embarrassingly bad, so is the sketchy book on meter cast forth by the prolific scholar, translator, and poet Burton Raffel. His subtitle, The Autobiography of English Prosody, implies that the verse lines Raffel quotes, scans, and sometimes annotates will do the job of explaining how poets writing in English have used meter. Unfortunately, they cannot do this on their own, and Raffel's meager commentary (only about eighty pages out of 185) is superficial, cranky, and unreliable. He claims to be merely reporting "the facts" (xii, 111, 119, 123, and elsewhere), but his comments constantly take the form of praise and blame. The book is a slapdash production full of scansion errors, misunderstandings, eccentric judgments, plain blunders, and strangely chosen exhibits. Raffel ignores English dramatic verse altogether, chooses remarkably uncharacteristic poems to represent Donne and Auden (among others), declines to comment on Poe, gives us eight lines on Wordsworth and two on Yeats. He cannot recognize that the Dylan Thomas poem he scans is syllabic, the Masefield one dipodic. When he runs against a line he cannot scan, his typical response is to suggest that the poet has been incompetent, instead of suspecting that his own ear may be faulty. Unable to find likely readings of poems by Nashe and Jonson, he accuses them of "clumsiness" (58, 60). The Jonson line is especially revealing. Raffel scans it (using capitalized syllables to show stress, a method likely to encourage singsong readings and mute the subtle effects of pyrrhics and spondees): "FOR / whose SAKE, / hence-FORTH, / all his VOWES / be SUCH," instead of just letting the meter suggest a quite smooth reading: For WHOSE / sake, HENCE- / forth, ALL / his VOWES / be SUCH (59). "The awkwardly trisyllabic fourth foot" is used to establish "Jonson's frequent prosodic clumsiness" (60), and both can now be taken as "facts."

This kind of foolishness - misperceived lines leading to blanket judgments of poets' or even centuries' whole poetic production - is typical of the book and might lead even a patient reader to adopt the Holderian strategy of condemning metrical analysis wholesale for its "Disinterested Meddling." The heroic protagonist of this autobiography is what Raffel calls the "Chaucerian Compromise," i.e., accentual-syllabic verse as practiced by the best poets from Chaucer through Pope. He praises Dryden and especially Pope for their "deft" (75) and "amazingly varied" (80) placement of caesuras, but he cannot even report accurately where they place them. The passage of Pope he quotes actually has a break in phrasing after the fifth syllable in five successive lines (amazingly varied?), but Raffel's list misreports three of them. What matter? Soon after Pope the prosodic roof falls in: the introduction of trisyllabic feet by nineteenth-century poets is treated by Raffel as a major metrical crime. He does not ask, as O'Donnell does, what the poets might have been trying to accomplish with their innovations; instead, he judges them to be "casual, even indifferent (i.e., uncaring)" (103), or lazy. Wordsworth gets off lightly ("consistently conservative" [96]) because Raffel has nothing to say about him; he judges Coleridge's prosody to be "remarkably like Wordsworth's . . . with an admixture of ballad- and folk-style poems in which bi- and trisyllabic feet are blended pretty much indiscriminately, with the folkish trisyllabic predominating" (97). But he is at his most peevish in writing about Shelley. Irritated at one line, he writes: "Scansion is an abstract, almost a meaningless game, when one employs it as Shelley does" (106). Somehow he seems unaware that it is Raffel who is employing scansion, not Shelley. He is similarly dismissive about the metrical practices of Keats, Tennyson, Browning, Hopkins, and every Edwardian he deigns to quote. Judging them all by his one rigid formula (how well they adhere to the principles of the Chaucerian Compromise), he hasn't a clue to what Browning was up to in "Soliloquy of the Spanish Cloister" he misunderstands Hopkins's aims and procedures, and even when he acknowledges that Whitman's verse "should not be scanned" (132, his emphasis), he goes on scanning it and all the other free verse he quotes. He cannot keep the beat in Blake's "The Shepherd" for two whole stanzas (89), and Holder may be pleased to know that Raffel cannot hear the beat in "Ah! Sun-Flower" and considers the poem "metrically . . . jumbled" (90) - evidence either of metrists' incorrigible disposition to disagree (Holder's point) or of Raffel's tin ear (my point). Unlike Holder, Raffel has apparently not read Wimsatt and Beardsley on this poem; their article does not appear in his "Works Cited and Recommended," which also omits standard works by Thompson, Tarlinskaya, Attridge, Cable, Suhamy, myself, and others, or any of John Hollander's work except his brief and charming Rhyme's Reason. This is an author who thinks that his sensitive ear can make up for his lack of knowledge - a woefully misguided view.(14)

6

One crucial issue in all these books is performance, a term that is understood differently by different schools of readers. Actors, poets, and most readers of verse, including Holder, are principally concerned with how to speak or hear the lines, with their performance on specific occasions. Even if they look for guidance in the form of general principles, they suppose that such principles will help them speak or hear lines with understanding of their rhythmic design.

Linguists, in contrast, have usually been concerned to study, at a higher level of abstraction, how structures of language interact with metrical, intonational, or other formal structures of poetry, to observe, for example, without commenting on a poem's aesthetic success or failure, the extent of the congruence between stress positions and the stressed syllables of major-category words. It is irrelevant to them how emphatically or lightly a line's stressed syllables are stressed, or whether minor syllables in stressed positions are given additional stress or major syllables in weak positions are "demoted" in sound as well as in status. How a line is sounded is often not of interest to them at all. They are particularly wary of having their analyses thrown off by particular performances of lines. In theory, at least, and rightly, their studies maintain a comfortable distance between the abstract structure of a metrical line and actual performances of it.

In practice, however, they have to attend to performance in at least two senses. First, they must take into account the speech habits of native speakers of the language, modified perhaps by information they may have about the actual pronunciation of early performers (poets, actors, or others). They have to notice which syllables of polysyllabic words normally take the main stress, which syllables in short phrases of different sorts are most likely to attract stress, and so on. They are also inclined to generalize about the probability of major-category words being stressed more than minor-category words, and they eventually have to notice exceptions and alterations of such generalized features of the language under special conditions of phrasing. As they refine their account of the way a language is spoken, they have to be guided by speakers' performance in this sense, by the way native speakers normally speak and modify their speech according to the needs of particular situations.

But if they recognize and feel comfortable with this peculiarity of language - that it will change its rules or practices under the pressure of different speaking situations - linguists often feel distinctly uncomfortable with the notion that in verse situations another kind of pressure may be brought to bear on the stress patterns of English, the pressure that the meter may exert - in Wordsworth's phrase, "the passion of metre." But this pressure or passion is felt differently by different performers, poets, and metrists (to tilt or not to tilt?), and it is understandable that linguists are reluctant to make pronouncements about a matter that varies so much from one reader or performer to another. Understandable, too, that they should want to stipulate, therefore, that particular performances, no matter how brilliant (or perhaps especially if they are brilliant) should not be counted as relevant.

On one side, then, we have linguists, skeptical about performance, but having to acknowledge that the phrases and sentences we find in lines of verse usually are derived from a spoken language.

On the other side, at its extreme position, Holder, skeptical about meter, and refusing to grant that lines have beats, concerned with performance, but either a poor transcriber of its details or a poor performer himself.

Or are they both on the same side? Both not knowing how to handle the pressures (the passion?) of meter.

Attridge seems in some way caught in the middle, wanting the system and order that linguistic analysis might provide, but also concerned with effective reading. It is understandable, then, that, wanting to accept the insights of linguists about the relative importance of major words, he should devise a system designed to show the metrical poetry of English as built on the placement of major syllables in major positions; but it also makes sense that, as a presumably effective reader of verse himself (everything in his books suggests that he must be a very good one), he should want some justification for adjusting his system to the little changes that seem called for if one is to read lines convincingly as being at once English and verse. For him, therefore, the crucial problem is "metrical subordination," and it is not surprising that he feels uncomfortable with it. There is the problem in a nutshell: whether and to what extent the meter can exert such pressure on our normal English habits of speech as to tilt a little the prose or speech pattern, just enough to make the verse pattern pleasurably audible. But as Holder refuses to hear beats, Attridge refuses to let the voice tilt, though he grants that the English iambic ear, so to speak, may sometimes perceive a tilt (REP 164-68).

My own view is that much of the difficulty here could be resolved if linguists and actors, scholars and ordinary readers, would understand that the differences that matter for the performance of lines of iambic pentameter verse are usually very slight. One does not go lumbering (or lurching) through lines of verse, making huge variations between what count as stressed and unstressed syllables. Small increases or decreases of emphasis do not much change the basic pattern of alternating degrees of strength in a succession of English syllables. English is an accentual language, and there are some speakers (the late sports announcer Howard Cosell was a notable example) who habitually pounce on the major stressed syllables in word-groups. But as Attridge reluctantly admits, the difference between "You that poor Petrarch's" and "You that in Petrarch's" does not really amount to much. It should not be enough to make one change one's understanding of either phrase's metrical pattern in the line. Since good readers will often differ in the way they distribute stress, it makes sense to allow for such differences and to prescribe as little as possible, or at least not to think that one can resolve all the differences by a more and more precise accumulation of rules and conditions. As Aristotle said wisely a long time ago: "Our discussion will be adequate if it has as much precision as the subject matter admits of, for precision is not to be sought for alike in all discussions. . . . lilt is the mark of an educated man to look for precision in each class of things just so far as the nature of the subject admits" (1.3.1094b).

One of the problems of traditional metrics has been its inability to describe metrical patterns in language (and graphic notation) that would convey lucidly and interestingly the rhythmic patterns of English poetry. Readers irritated with the clumsiness of metrical description have justifiably complained that the formal study of meter often stands, as Holder tells us, between poems and their readers. Despite the continued overreading or misreading that Holder rightly disparages, I like to think that the situation is improving, that, at least in the hands of some sensitive readers, tactful modifications of the older formulas are making the system seem more flexible and are helping readers to hear with greater insight and pleasure our older poets' metrical designs. By the same token, linguists who analyze metrical structures in their own way, or literary scholars who ground their metrical theories in linguistic data and theory, need to be told when those theories seem to conflict with the evidence of ordinary phonological experience, with the testimony of poets, or with the practice of competent actors and professional readers of verse.

In line with this modest perspective, I recommend some words written in a useful and unpretentious book by Delbert Spain, Shakespeare Sounded Soundly (1988). This is essentially a "Handbook for Students, Actors, and Directors," and it contains very good advice, usually presented in plain English and a minimum of jargon. Spain has this to say as a general statement on the relation of dramatic poetry to ordinary speech:

In the writing or the reading of poetry, there is always some imposition of the rhythmic form that is not quite inherent in the words. We make adjustments in the pulses and stresses when we read that may be so inconspicuous that we do not realize we are doing it. There are cases where we may need to notice such adjustments and make them arbitrarily. If we get heavy handed with them, we create an effect of singsong that we do not want. Our goal is a happy compromise, enough delicate touch of stress applied in the line to make it reveal its proper poetic rhythm without impairing the good sense and spirit of the words. (20)

Spain keeps steadily in mind the problems that face actors in the theater, especially their suspicion of declamatory verse and their all-too-frequent habit of trying to speak dramatic poetry as if it were prose. His book is designed to give them the knowledge they need - about Elizabethan pronunciation, about iambic pentameter and Shakespeare's patterned departures from it - to do justice to the poetry he describes as "language on parade" (1). In trying to help actors to solve the practical problems of verse-speaking, he summarizes only as much metrical theory as they will need, for he knows that if the actors do not give life to the verse, it will not be alive for the audience. The actors must sound it "soundly," must send it forth invisibly for everyone in that circumscribed globe to hear. In the theater there are no feet, no lines, no scansion marks or marks of punctuation. But we hear the words arriving in our ears; we hear a succession of syllables, some softer, some louder or sharper; we hear perhaps a pattern as they scurry or tumble by us; and if the language is verse and the actors speak it, we hear the poetry.

Notes

1 If I understand him, I think what Attridge means here is: when the next or the previous offbeat is a non-final double one. Offbeats are normally followed and preceded by beats, not by offbeats. But it seems to me a problem for Attridge's terminology that he often writes as though a verse line's character were defined by its offbeats rather than its beats.

2 Attridge specifically reads a virtual offbeat in this line but not in the others (248-49), apparently because the grammar is different in (8), though to my ear it sounds just like the others if I try to speak the line as I understand Attridge wants it spoken.

3 To be sure, Attridge does declare from the start in REP that "since my interest is primarily in the singleness of [the English accentual-syllabic] metrical tradition - in the capacity, that is, of the modern reader to engage directly with rhythmic forms produced over the past six hundred years - I have deliberately ignored its historical dimension" (vii). But if earlier poets did not hear lines as Attridge thinks contemporary poets and readers do, his account is bound to be flawed.

4 John Crowe Ransom long ago suggested an at least equally plausible scansion for this line: Weep no more, woful shepherds, weep no more (190). This seems an appropriate way of recognizing that older poets enjoyed placing the same phrases in different metrical positions in a line and often did so with great expressive force, as Milton does here.

Attridge constantly faults "classical" metrics for being unable to do more than label a line's subtle effects with "trochee-spondee" or "pyrrhic-spondee" and so on (PR 142; REP 168,232, 234, and elsewhere). But this is a bad rap. His own eloquence about rhythmic effects lies not in his elaborate notation but in his commentary, and that recourse is open to traditional metrists as well. He also makes traditional scansion seem more wooden than it is because when he tries his hand at it he declines to employ the secondary stresses that many traditional metrists now use to show an intermediate range of stress between stressed and unstressed.

5 Note that spondee throughout this review means spondaic lamb. We understand that in English it hardly ever happens that two adjacent stressed syllables will be stressed exactly the same. Similarly with pyrrhic.

6 Listening lately to John Gielgud's recent recording of most of Shakespeare's Sonnets, I noticed that he appears to have four different ways of treating the kind of double offbeat that I think of as a pyrrhic-spondee combination. I heard him definitely lurch twice, on the phrases, "And the firm soil" (64:7), "And for that sorrow" (120:2), possibly also on one other, "that the world's eye" (69:1). His much more usual practice was to give either the first or second syllable of the four a substantial stress, either primary or secondary, enough to assure that it would be heard as a beat. Examples: You should live twice (17:14), Nor shall death brag (18:11), To the wide world (19: 7), That in black ink (65:14), From this vile worm (71:4), as my poor name (71:11), Or as sweet seasoned (75:2), etc. What is perhaps most interesting is that Gielgud sometimes gives the first two syllables a distinct pronunciation that seems to my ear strong enough for the second one to retain the beat, but then raises the stress-level of the next two syllables, an elocutionary resource that I believe accounts for the effective speaking of many lines in the theater, but that metrists have paid little attention to. Is he doing this, as I believe, because the line seems to demand it, or because he is consciously treating the figure as comprising two very different kinds of feet, both iambic? This is how he reads these phrases, among others: But from thine eyes (14:9), It is my love (61:10), to thy fair flower (69:12), On your broad main (80:8), So try great gift (87:11), is of more worth (103:3). Gielgud's performance does not prove that the phrases must be spoken his way, but it ought to be harder to argue, in the face of such readings by an extraordinarily experienced actor, that the classification of minor syllables as belonging to an unstressed category resolves the question of where we can hear the beat in such lines.

7 Two other matters of detail are worth mentioning. (1) Despite O'Donnell's assertion (133), no part of "The Thorn" is dipodic. (2) O'Donnell has some elaborate charts that graphically map instances of assonance and "alliteration"; by the latter term he apparently means what many critics would call consonance, the repetition of consonance sounds anywhere in words, not especially at the beginning of stressed syllables.

8 If by this phrase he means simply that we should read traditional verse without laboriously assigning metrical values to every syllable, who would disagree?

9 Despite the justice of some of Holder's criticism of critics who apply traditional prosody to free verse, every claim that a free verse poem is "haunted" by older poetry should not be dismissed out of hand. It is obviously true of some of Eliot's major poems, and one of Attridge's best passages in PR analyzes the extent to which Adrienne Rich's "Night Watch" "derives its rhythmic quality from its existence on the borders of regular verse" (172).

10 Holder's anger at a metrical "system" that bullies the phrases makes me wonder why he does not feel an equal outrage at grammarians. Is not grammar as much of a system as meter? Does not it similarly break up phrases into fragmentary parts of speech? Shouldn't we regard the objects of verbs as oppressed by their transitive masters? Should we sympathize with adjectives and adverbs, that can only modify something? Or is it the case that poets can use both phrases and grammar at once, can hear the same words operating at once in a phrasal, a grammatical, and a metrical context, and with some kind of pitch-patterns, too?

11 Holder's book has its share of typographical errors, and its index is unreliable.

12 An example (222):

         bends
                                  mov
                                                         move
Or                with the re              er to re

Does that sound natural to anyone?

13 Here we might adapt one of Holder's own sentences about another critic: "The sonnets of Shakespeare, already abused by prosodists, suffer additional scansional indignity at the hands of" Holder (70).

14 In a later article in CEA Critic, Raffel makes extravagant claims to hear "prosodic signals" in short or long lines. As we might expect, he has read no one else who has written on the subject; he mis-scans some lines badly; and he attributes remarkably explicit "messages" to mere metrical variations: "a signal of her unwilling but enforced acceptance of the nurse's inadequate pace of operation"; "It reveals Juliet's trust in the absolute reciprocity of her and Romeo's mutual affection" (56).

Other Works Cited

Aristotle. Nicomachean Ethics. Trans. W. D. Ross. Introduction to Aristotle. Ed. Richard McKeon. New York: Modern Library, 1947. 308-543.

Attridge, Derek. "'Damn with Faint Praise': Double Offbeat Demotion." Eidos 4 (Aug. 1987): 3-6.

-----. The Rhythms of English Poetry. London: Longman, 1982.

-----. Well-Weighed Syllables: Elizabethan Verse in Classical Metres. Cambridge: Cambridge UP, 1974.

Auden, W. H. The English Auden: Poems, Essays, and Dramatic Writings, 1927-1939. Ed. Edward Mendelson. New York: Random, 1977.

Barton, John. Playing Shakespeare. London: Methuen, 1984.

Brogan, T. V. F., and Derek Attridge. "On the 'Crescendo Foot.'" Eidos 4 (Dec. 1987): 11-12, 18.

Cable, Thomas. The English Alliterative Tradition. Philadelphia: U of Pennsylvania P, 1991.

Hardison, O. B., Jr. Prosody and Purpose in the English Renaissance. Baltimore: Johns Hopkins UP, 1989.

Hollander, John. Rhyme's Reason: A Guide to English Verse. New Haven: Yale UP, 1981.

Howarth, Herbert. "Metre and Emphasis: A Conservative Note." Essays on Shakespeare. Ed. Gordon Ross Smith University Park: Pennsylvania State UP, 1965. 211-227.

Jespersen, Otto. "Notes on Metre." The Structure of Verse: Modern Essays on Prosody Ed. Harvey Gross. Greenwich, CT: Fawcett, 1966. 111-30.

Oras, Ants. Pause Patterns in Elizabethan and Jacobean Drama: An Experiment in Prosody. Gainesville: U of Florida P, 1960.

Raffel, Burton. "Metrical Dramaturgy in Shakespeare's Earlier Plays." CEA Critic 57.3 (1995): 51-65.

Ransom, John Crowe. "The Strange Music of English Verse." Perspectives on Poetry. Ed. James L. Calderwood and Harold E. Toliver. New York: Oxford UP, 1968. 184-90.

Shakespeare, William. The Sonnets. Perf. John Gielgud. Caedmon, 1963, 1996.

Spain, Delbert. Shakespeare Sounded Soundly: The Verse Structure and the Language. Santa Barbara: Garland-Clarke, 1988.

Tarlinskaya, Marina. English Verse: Theory and History. The Hague: Mouton, 1976.

-----. Shakespeare's Verse: Iambic Pentameter and the Poet's Idiosyncrasies. New York: Peter Lang, 1987.

Wimsatt, James I. "Rhyme / Reason, Chaucer / Pope, Icon / Symbol." MLQ 55 (1994): 17-46.

Wimsatt, W. K., Jr. "The Rule and the Norm: Halle and Keyser on Chaucer's Meter." College English 31 (1970): 774-88.

----, and Monroe Beardsley. "The Concept of Meter: An Exercise in Abstraction." PMLA 74 (1959): 585-98.

Winters, Yvor. "The Audible Reading of Poetry." The Function of Criticism: Problems and Exercises. Denver: Swallow, 1957.79-100.

Wright, George T. "The Meter of 'Shall I Die?'" Eidos 3 (1986): 6, 11-12.

-----. Shakespeare's Metrical Art. Berkeley: U of California P, 1988.

George T. Wright is professor emeritus of English (University of Minnesota, Twin Cities), and lives in Tucson, Arizona. His publications include Shakespeare k Metrical Art (1988) and two award-winning articles: "Hendiadys and Hamlet" (PMLA, 1981) and "The Lyric Present: Simple Present Verbs in English Poems" (PMLA, 1974).

COPYRIGHT 1997 Northern Illinois University
COPYRIGHT 2004 Gale Group