Versification: an interdisciplinary journal of literary prosody
Filename: MeterInEnglish_597_MahoneyRev.html (corrected V.30.97>
29 May 1997, Vol. 1, No. 1

What Is a Meter?
A Review of Meter in English: A Critical Engagement, edited by David Baker

Anne Mahoney
Boston University

Copyright © Anne Mahoney 1997
Received: 28 May 1997; Published: 29 May 1997

KEYWORDS: poetic meter, English (modern), poetic rhythm, scansion.

REVIEW OF: Meter in English: A Critical Engagement, ed. David Baker. Contributors: Robert Wallace, Eavan Boland, Annie Finch, Dana Gioia, Rachel Hadas, Charles O. Hartman, Robert Hass, Margaret Holley, John Frederick Nims, David J. Rothman, Timothy Steele, Lewis Putnam Turco, Barry Weller, Richard Wilbur, Susanne Woods. 1996. University of Arkansas Press, Fayetteville. pp. 368. ISBN 1-55728-422-9, 1-55728-444-X (paper)

Readers of poetry in English are aware that some of this poetry is metrical, like Shakespeare's sonnets, and some is not, like Whitman's "Song of Myself." Students of meter have also observed that the metrical poems involve several different-sounding metrical patterns, different patterns of stressed and unstressed syllables. These patterns are conventionally called the meters of English poetry: iambic, trochaic, anapestic, dactylic. In Meter in English, Robert Wallace argues that these conventional "meters" are really all the same meter, accentual-syllabic, and lie on a continuum from strict iambics to loose lines with mixed feet. He raises the further question whether two other poetic forms, accentual and syllabic, can really be called meters at all. The 16 essays in this book discuss the nature and conventions of English meter, how to scan it, and how to teach it. Wallace's main intentions are "to clarify some of the methods of scansion, to simplify (or make more appropriate) some of the vocabulary of metrics, and to streamline our ways of thinking about the fundamental structure of meter in English." (p. xi)

The book is in the form of a symposium. It begins with a substantial essay (40 pages) by Wallace in which he makes ten proposals about English meter. Each of the other contributors then takes the floor to respond. The editor's guidelines to the contributors were simply that their essays should address each of the ten proposals as well as respond generally to Wallace's ideas. They do so in a variety of ways, some treating each proposal in turn and some writing a unified essay incorporating discussion of the proposals. Finally, Wallace sums up in a 55-page response to the responses, modifying some of his earlier statements and standing by others. The format does give the flavor of attending a panel discussion. As I read the essays one after another I could feel tension mounting: how will Wallace respond? What will be the consensus on these points? It's not often that a book of metrical theory has a plot.

Although the book is written by and for practicing poets and metrists, it would also be a useful supplement for a class at the stage of discovering that meter is not always as simple as basic textbooks make it out to be. There is a solid bibliography on prosody and metrics, compiled by the editor with the aid of the contributors, and many of the essays also have references for further reading. There are two indexes, one for the discussions of the ten proposals in the various essays and one for the authors cited, but no general index. The only small criticism I would make of the format of the book is that poems quoted in the individual essays are not always identified: some of the contributors do, some do not. While most of the poems used as examples are familiar, it would be a convenience to the reader to note, for example, that

To a green thought in a green shade

is "The Garden" line 48, not just that it is by Marvell.

The ten proposals range from relatively innocuous questions of terminology to the very definition of meter itself. The first seems to be one of the innocuous ones: (1) Instead of the term "feminine ending," we should say simply extra-syllable ending (p. 5). The desire for simpler language is one of the themes of Wallace's essay. As he points out in his second essay (p. 345), some students may be uncomfortable with what they perceive as gendered, hence sexist, language. Of course it's not, but even students who know some French or Spanish may not connect "feminine ending" or "feminine rhyme" with the form of the feminine gender in the Romance languages unless the connection is pointed out to them. Margaret Holley makes this point: "A simple statement of the origin of these two terms [masculine, feminine] can forge the link in a student's or reader's mind between modern English and one of the languages from which it arose" (p. 164). Other contributors point out that whatever term we choose, we must emphasize that this is a normal variation in iambic pentameter lines and can be used for expressive effect. Timothy Steele mentions Shakespeare's Sonnet 20, "A woman's face with Nature's own hand painted," in which the subject is a man gifted with feminine beauty, and every line has a feminine ending. Strikingly, many of the contributors argue for keeping the traditional term. Annie Finch proposes "falling ending" (p. 70), shorter than "extra-syllable ending" and also self-illustrating. She points out that the corresponding term "falling rhyme" can also apply to trochaic lines in which the final unstressed syllable isn't extra. Both Charles O. Hartman and Lewis Putnam Turco propose calling the last foot of an iambic line with feminine ending an amphibrach, which of course it is. Hartman (pp. 113-14) finds that this confuses students, however, who seem to discover amphibrachs in the middle of lines where they are not wanted. He also points out that if we are to add the amphibrach for this purpose, we will also need the second paeon and the palimbacchiac, for anapest and spondee plus extra syllable. I agree that "feminine ending" is not the best term, because it is not descriptive in English. Perhaps English scansion could take over "pendant" and "blunt" from classical scansion, for "feminine" and "masculine" respectively.

The second proposal also deals with terminology. (2) For an omitted first syllable of a line, we should use the term anacrusis (p. 6). Responses here were divided between those who were unwilling to use a perfectly good term from classical metrics in the opposite of its conventional sense and those who were willing to accept this to avoid the gruesome image of a beheaded line. Wallace argues that "anacrusis" should apply to both an added syllable and an omitted one at the beginning of a line, and notes that added syllables are rare; he is "not convinced that anacrusis, in the sense of added initial syllables, exists at all in English verse" (p. 345, his italics), and would simply call the first foot an anapest in the case of a line like Merchant of Venice V.1.169

And so riveted with faith unto your flesh.

I suspect that making one word carry two opposite meanings is unnecessarily confusing. If "acephalic" is too Greek, and "headless" is too gory, why not "omitted-syllable beginning" to match "extra-syllable ending"? In fact, I disagree with Wallace that terms like "acephalous line," "decapitation," and "initial truncation" "both overstate and seem pejorative" (p. 6). The point to impress upon students is that, as in the case of the pendant ending, both acephaly and anacrusis (in the strict sense) are normal variations in English lines. Where is the harm in a slightly colorful word like "headless"?

With proposition three we get into the meat of Wallace's argument. This proposal says (3) Quantities are not a basis for meter in English (p. 12). That is, despite various poets' experiments, meter based on long and short syllables does not work in English. There was near-unanimous agreement on this statement, with only Holley dissenting. She points out "Since quantitative verse has clearly been written in English, I don't see the value of arguing that it does not exist" (p. 166), but accepts it only "to honor the undertakings of others, however few, who have found it interesting" (p. 166). Most of the contributors, like Wallace himself, correctly point out that quantities are not part of the English language as they are in Greek or Latin: you can't look up the quantity of a syllable in the dictionary. We should distinguish quantitative meter from accentual adaptations of classical meters, however, a point that John Frederick Nims makes (p. 176). Longfellow's Evangeline, for example, is written in dactylic hexameter based on stresses, not quantities, and there are many other examples.

Much more controversy attends the next proposal: (4) Syllabics is not a meter in English (pp. 12-14). Wallace does not intend to deny the existence of syllabic verse. The question here is whether counting the number of syllables in a line is sufficient to create a meter. Wallace's arguments are that we do not hear syllable count, because number of syllables has no definite relationship to patterns of stress. Moreover, some of the major examples, notably Marianne Moore's "Poetry" (in the longer version, not the final three-line version), are not strict about syllable count. The variations may be deliberate and expressive, or they may be merely "sloppy" (discussion and examples, pp. 317-22). "Syllabics," concludes Wallace, "is not a meter, but the mere idea of a meter" (p. 321). It is a kind of free verse. For Wallace, English meter must be based on patterns of stressed and unstressed syllables: "Meter is thus a system of measurement, conventional but natural to the language, which makes the rhythmic units of line more or less predictable to a reader or hearer of voice. What meter measures is speech" (p. 7). That is, meter must involve measuring or counting something that is inherent in the language, and must allow the hearer or reader to recognize the pattern. It is the second part of this definition that causes the problem. Certainly syllabic poetry involves counting entities that are "natural to the language." Is it possible to hear the count of syllables? Is this necessary to a definition of meter?

Several contributors agree with Wallace that syllabic verse is a form or technique, but not a meter. Finch, for example, agrees that "meter in English (including accentual meter) counts the most conspicuous feature in English-language poetry and in the English language, which is accent." (p. 71) Hadas concurs that we do not hear syllable count, and that syllabics can be a useful tool for the poet without necessarily being of concern to the reader (pp. 99-101). Woods agrees that "metrical poetry derives from a conventional counting of derivable stress patterns" (p. 284). Hass's discussion (pp. 135-41) contrasts Moore's stanza patterns with stichic syllabic poems by Kenneth Rexroth ("The Signature of All Things") and John Logan ("On Reading Camus in Early Spring"). In both cases there is a pattern, but not as "addictive" as the pattern of verse with strong stresses. The reader or hearer perceives the general shape of the stanza or the line, just as we perceive the complicated rhyme-schemes of seventeenth-century stanza forms. Hass would call syllabic verse a form like blank verse or the Spenserian stanza: a way of organizing a poem, but not a meter.

Others disagree. Rothman puts this side of the argument well: "Syllabic meter in English is a compelling measure because it is clear, simple, consistent, and regulates phonemic flow, albeit minimally" (p. 207). He goes on to point out that it can be difficult to recognize accentual-syllabic meters by ear, especially given unrhymed lines with much enjambment, but this does not mean that, say, Paradise Lost is not written in meter. In other words, counting syllables is as valid as counting stresses, and whether the listener can hear it is not important. Hartman points out that if a poem is written in strict syllabics ("the strictness made necessary by the system's unfamiliarity is its main drawback," p. 117), a reader can verify the meter by counting up syllables. "If a poet subjects syllable counts to a rule, and the reader has some reason to know this (if, for instance, that poet's practice becomes famous), then a syllabic meter is possible." (p. 117) Holley points out that the inaudibility is part of the point, one of the reasons why a poet like Moore chooses to write syllabics. As long as the meter can be counted or measured, it is a real meter. "The relative quietness of the syllabic mode and the predominance of other music does not mean that such a poem [specifically Moore's "A Carriage from Sweden"] is not metered, that it has not been written in the discipline of a pattern of measure." (p. 157) Weller, comparing syllabics to shaped poems and other strictly visual patterns, notes that "it isn't self-evident this measure needs to be aural rather than visual" (p. 268); Gioia similarly notes that syllabic verse may sound unmetered but has a perceptible visual shape (p. 80).

Some of the contributors argue that syllable count is, or can be, audible. Nims says that it is for short lines, like the five-syllable lines of Sylvia Plath's "Mushrooms." While most people cannot hear longer lines, "no matter: the poets themselves clearly had their numerical meters in mind" (p. 177). Turco points out that Moore herself seems to have written syllabic verse by ear (p. 257), and notes that if the French, the Welsh, and other people can hear the syllabic verse of their traditions, surely Anglophones can too (p. 258). Of course, as Gioia notes (p. 79), French verse is helped along by end-rhyme, and Moore uses rhyme to mark line-endings on occasion too.

Finally, Boland argues that Moore's "Poetry" "is an aggressive and disruptive dialogue with historic iambic meter." (p. 50) For her, a definition of meter that focuses on the canonical forms, ignoring such disruptive dialogues, excludes much of what makes our metrical practice grow and evolve. Even Donne's poetry, canonical now, "conducted a dialogue with the meters of the sixteenth century which cannot be excluded from any final sense of his achievement" (p. 50).

The fifth proposal raises a similar question. (5) In modern English, accentual meter does not exist (pp. 14-22). By "accentual meter" Wallace means lines in which the stressed syllables are counted but the unstressed are not. For him, "counting only stresses offers no meaningful predictability and is, ultimately, hopelessly subjective" (p. 14). In fact, if the number of stresses actually is held constant, Wallace will scan as iambic with substitutions, and if it is not, Wallace considers the poem to be in free verse. He distinguishes this sort of verse from the meter of Old English, where the rules of alliteration clearly indicate which syllables count as stressed, and would prefer to call the latter "Old English meter" or "alliterative meter" (p. 21). He points out that it is often possible to disagree about the number of stresses, and uses Eliot's "The Waste Land," line 18, as an example. He reads this line with six stresses and six feet (p. 16):

       /                 /                                 /                         /       /                                 /
I read, | much of | the night, | and go | south in | the winter

Later (p. 322) he quotes a scansion by Harvey Gross, omitting the stress on 'much' and 'go' to get a four-beat line. Meanwhile, Richard Wilbur (p. 280) has scanned it as pentameter:

        /           /                       /                          /                      / 
I read, | much | of the night, | and go south | in the win|ter

If the same line can be scanned with four, five, or six stresses, we do not have a predictable pattern here. Of course this is an extreme example, since "The Waste Land" is not really an accentual poem (the opening is "obviously more a parody than anything else," says Rothman, p. 209). More standard examples, such as Yeats' "Easter 1916," sound iambic to Wallace. He calculates a ratio of 2.5 iambs to each anapest in the poem, with the highest concentration of anapests in the first stanza and the most iambs in the fourth. Moreover, he points out that there are some lines with more or fewer than three stresses:

         /         /         /                     /

A horse-hoof slides on the brim (l. 51)

           /                     /

Enchanted to a stone (l. 43)

This makes sense in an iambic context, where line 51 is x /| / / | x x / and line 43 is x /| x (/) | x /, but does not work in an accentual poem (pp. 316-17; I have supplied the explicit scansions).

Whereas proposal 4 did not argue that syllabics do not exist, the argument here is that there is no such thing as "accentual" meter, but instead everything that might be called "accentual" is really loose iambics. As before, some of the contributors agree and some do not. Says Gioia, "Wallace is hopelessly wrong in asserting that accentual meter does not exist in contemporary English" (p. 80), going on to note that "his analysis of the subject is nonetheless illuminating. By skeptically examining statements and scansions ..., he correctly points out the vagueness and inconsistency of most critical discussions of accentual meter" (p. 80). Gioia believes it is not possible to scan accentual meter by means of ordinary feet, only by lines and by strong stresses. He notes, perceptively, that accentual verse is more at home in popular verse, while literary poetry prefers accentual-syllabics (p. 84); this is also a distinction between forms that are primarily heard, like most popular forms, and those that are primarily read. Gioia scans a stanza of Auden's "September 1, 1939" (pp. 85-86) and observes that while every line has three stresses, only one of the eleven lines is a regular iambic trimeter. This poem is regular as an accentual poem, almost unrecognizably loose as an accentual-syllabic poem. Other contributors (Holley, Nims, Steele, Turco, Weller) cite Hopkins and his "sprung rhythm" as a form of accentual verse, though in this case I am inclined to agree with Wallace: "In a consideration of accentual meter, Hopkins is a special case. 'Sprung rhythm,' whatever it is, was written with all of Hopkins's eccentric machinery of paeons and outriding feet in reference to accentual-syllabic meter" (p. 325). Susanne Woods, however, interprets the proposal more nearly like the fourth one: "I am inclined to believe that much late-twentieth-century poetry is actually more accentual than accentual-syllabic, but whether such poetry should be considered 'metrical' is probably a matter for debate" (p. 291n4). That is, she asserts the existence of accentual verse, which is what Wallace denies, but proposes that it is non-metrical in the same way as syllabic verse.

Does it simplify our metrical theory or our teaching of meter to say that poems like "Easter 1916" and "September 1, 1939" are in loose iambic meter rather than accentual or strong-stress meter? Wallace believes it does. Hadas applauds, saying Wallace is "on the side of clarity, and a foe to obfuscation" (p. 102). Weller agrees that students may find accentual meter difficult to recognize (p. 270, quoting Wallace p. 20), unless there are external clues. Wilbur, who has not contributed a full essay to the book but has allowed some of his correspondence with Wallace to be quoted, tells Wallace "It seems to me that you make a good case against 'accentual meter'" (p. 279), and points out that readers do disagree about where strong stresses fall.

In the next proposal, (6) Anapests and dactyls are legitimate substitutions in the iambic norm of English meter (p. 22), Wallace returns to the nature of accentual-syllabic iambic meter. Speaking strictly of iambs, almost everyone agrees that anapests are legitimate substitutions; Finch cautions that dactyls (and trochees) can only be substituted at particular places in an iambic line (p. 71). She and Holley both object to the phrase "the iambic norm" here, responding also to proposal 8 (see below). Wallace himself revises the present proposal in his final essay, having concluded that dactyls are not permissible in iambics. Taken simply at face value, the idea that an anapestic foot may be substituted for an iambic is not particularly controversial, and this proposal can, I think, be accepted independently of proposal 8.

With proposal seven, we get into the terminology for scanning English poetry. (7) We should drop the pyrrhic foot and accept in its place the double-iamb, as one of the six foot-terms necessary: iamb, trochee, anapest, dactyl, spondee, double-iamb (pp. 22-28). There are several separate points here: do we ever scan a foot of iambic meter as two unstressed syllables (a pyrrhic)? Is the double-iamb (by which Wallace means x x / / a useful foot for scanning iambics? Is the given list of foot-terms the correct one? This proposal looks at first to be as simple as the first two, merely suggesting suitable descriptive language. In fact the double-iamb is a misfit, leading to four-foot pentameters. Moreover, as several contributors point out, we cannot assume that every pyrrhic is followed by a spondee. Finch (p. 72) quotes and scans

x   x     x      /      x      /     x      /    x  x       /

It is | the star | to ev'| ry wan|dering bark (Shakespeare, Sonnet 116, line 7)

"The first foot here is not a trochee; to scan it as such would obscure the earnest, hopeful, thwarted stretch of the second syllable towards a stress, and the consequent increase in energy that finally accompanies the awaited stress on 'star.' Only the pyrrhic-iamb combination could embody this dynamic" (p. 72). Nims similarly (p. 189) quotes additional lines from Shakespeare showing pyrrhic-iamb sequences; Weller (p. 272) has an example from Keats. Hartman, on the other hand, conjectures that "every possible pyrrhic either (1) is heard as an iamb, ... or (2) occurs immediately before a spondee" (p. 114); he would presumably read the first foot of the line above as an iamb or a trochee.

Another objection raised to the term "double-iamb" is that this term has also been used to refer to an iambic dipody, x / x / . The contributors who raise this point prefer "ionic," the Greek term for the foot. There is little discussion about the suggested list of feet. In his final essay, Wallace decides that there are really no dactyls, but that the bacchiac is useful in a triple-rhythm context. In response to Finch's argument, he proposes adding the fourth paeon to the list. Now we have two double feet, and Shakespeare's pentameter degenerates into a tetrameter. I do not see the utility of these feet for scanning iambic verse. Wallace is correct to add the bacchiac, though, which has the same sort of function in triple rhythm as the spondee does in duple.

The terminological and pedagogical discussion around proposal seven was merely the calm before the storm. With the next proposal we are back to fundamental issues as in proposals four and five. (8) Anapestic, trochaic, and dactylic meters do not exist in English (pp. 28-30). Wallace is perfectly definite about this, and means exactly what he is saying. For example, "I count the first stanza of 'How They Brought the Good News from Ghent to Aix' as comprising eighteen anapests, four iambs, two spondees. So, 'heavily anapestic'" (p. 29). That is, Browning's poem is not in anapestic meter, it is in iambic, even though most of the feet are anapests. Even "The Song of Hiawatha" is not trochaic, it is iambic, with many omitted first syllables and consistent pendant endings. In fact, Wallace holds that much of the poem can be scanned as trimeter, "if one forgets the meter and doesn't hit the first beats with a pile-driver" (p. 336). And as for dactyls, as we've already seen, Wallace ultimately comes to deny the existence of dactyls in English at all.

Not surprisingly, most of the contributors disagree strenuously with this proposal, as do I. One of Wallace's key points in his rejection of syllabics as a meter is that they cannot be heard. The differences among iambic, trochaic, and triple meters certainly can be heard; no one listening to "How They Brought the Good News from Ghent to Aix" (Browning) or "The Night Before Christmas" (C. Moore) or "'O where are you going?' said reader to rider" (Auden) would hear any of these as "iambic with substitutions"-they are all in triple meter. Trochaics, too, are different from iambics. Although many trochaic lines are catalectic, and thus end with a stressed syllable as iambic lines do, they are not iambic; they have a very definite falling rhythm. As for the non-catalectic trochaics of "Hiawatha," it seems to me somewhat perverse to treat them as iambics with an unusual feature at each end rather than as perfectly normal, textbook examples of trochaics.

Finch has experimented with writing in non-iambic meters. "At first," she says (pp. 59-60), "I found it extraordinarily difficult to conceive of a poem in indeterminate shape in a non-iambic meter (though I had written some sapphics), much less to sustain the rhythm; the poems would transform themselves into iambic pentameter or die on the page. I spent several years in the process of training my poetic ear (which had originally been trained in free verse and then in iambs) in meters other than iambic." If the non-iambic meters are merely variants of iambic, why would any ear-training be necessary? Finch goes on to say that "each of the non-iambic meters, also, has its own character, music, and history, however subtle or intermittent" (p. 61): irrationality for dactyls, supernatural and exotic subjects for trochees. Gioia demonstrates the difference between iambic and trochaic meter, in particular, by noting that iambic is much more flexible, and trochaic allows very few liberties. "There are only two common variations in English trochaic verse - iambic lines and elided triple feet" (p. 91). That is, a poem in trochaic meter may allow a line in very strict iambic meter, which might also be called a trochaic line with anacrusis, or it may allow a dactylic foot provided the two unstressed syllables are actually pronounced as one: "barberry" in "Where the tangled barberry-bushes" ("Hiawatha," introduction, l. 103) would be pronounced as "barb'ry."

Triple meter is much more complicated. While the two varieties of duple meter are generally clearly separable (not always, though: consider Milton's "L'Allegro"), the three possible forms of triple meter, dactylic, amphibrachic, and anapestic, frequently blur into one another. One of the running examples through Meter in English is Auden's "'O where are you going?' said reader to rider," first quoted and scanned by Wallace (p. 23). Wallace believes the poem is best scanned with anapests and iambs, but shows how it could also be scanned as amphibrachic. The difference between the two scansions would be the foot-divisions, not the marking of stresses. In this poem, as in many in triple meter, the line-breaks run counter to the foot-divisions: between the last stress of the first line and the first stress of the second, there are the expected two unstressed syllables, but the line-break comes between them, indicating that they should belong to different feet. If we allow for dovetailing, we can scan the first and second lines together, and the third and fourth, and there is no need for amphibrachs. If it is necessary to make a choice, I would label this poem 'dactylic,' predominantly a falling rhythm with the same sort of insistent, pounding sound as trochees. Yet Wallace's preferred anapestic scansion also works. Nims quotes Hopkins on this: "it is very hard to tell whether to scan by dactyls, anapests, or amphibrachs." "And probably not worth the trouble, since the meter will sound the same no matter what we call it," he adds (p. 188). "Exactly!" as Wallace says when he quotes this line, p. 298.

The proposal to call all the trochaic or triple-meter poems "loose iambic" or "iambic with substitutions" provokes strong reactions. Nims calls it "a more serious [than the seventh] and more preposterous proposition" (p. 189). Rothman: "Wallace's eighth principle makes little sense to me" (p. 213). Turco: "I say, 'Bah, humbug, Bob!" (p. 261) Hartman points out that although this system has fewer different kinds of meters, it is not a simplification for students, because it tends to "minimize the significant difference that any reader feels and hears, consciously or not" (p. 111) between poems in different meters. To instruct students in an exclusively iambic regime, and then hand them [Blake's] 'Ah Sun-flower' is simply sadistic. He goes on to make the essential observation that

Iambic poems tend to resolve passages of rhythmic complication in strongly iambic lines, anapestic poems resolve on anapestic lines, and trochaic poems on trochaic lines. Whether we can identify these points of resolution or release of metrical tension in a poem partly determines whether we can grasp its movement as a whole - which is surely the ultimate goal of metrical and rhythmical analysis and training. (p. 112)
In other words, no matter what metrical key a poem modulates into, it will return to its own tonic meter in the end. Consider for example the first quatrain of Yeats's "Leda and the Swan":

  x   /      x        /        x      /       /           /     x      /

A sudden blow: the great wings beating still

  x   /         x      /   x   x    /        x      /         x    /

Above the staggering girl, her thighs caressed

   x    x    /          /         x      /            /      x    x     /

By the dark webs, her nape caught in his bill,

  x      /         x     /      x        /      x     /    x      /

He holds her helpless breast upon his breast.

Line three struggles; in line 4, Leda is pinned and helpless, and the meter returns to perfect regularity.

Wallace's point, however, is not entirely unreasonable: there is a unity or a family resemblance in all the English accentual-syllabic meters, and there are places (precisely the poems that might be called accentual, I suspect) where they shade into each other. As he puts it in the final essay, "our different rhythms flow into and out of each other without a seam" (p. 331). He observes, agreeing with Steele, that trochaic and triple meters seem less flexible than iambic (pp. 335-6, 242). Trochaic, in particular, encourages an exaggerated reading with an insistent stress on the stressed syllables, always including the first syllable of the line even if it would not normally be stressed. The running example here is a "formulaic" line from "Hiawatha,"

The blue heron, the Shuh-shuh-gah

Our stereotype of trochaic meter requires heavy stress on the two definite articles here. Wallace would prefer to hear:

  x        /         /    x       x       x         /        x

The blue | heron, | the Shuh-shuh-| gah

His stresses are of course more natural, but his foot-division is unlikely. He has made the line an iambic trimeter with one trochaic substitution and a pendant ending. Since the rest of the poem is in tetrameter, surely this line, with these stresses, would also be trochaic tetrameter, with an iambic substitution in the first foot and a pyrrhic in the third. Anapests, too, can run away with the reader.

All things considered, I find it difficult to hear trochaics or triple meter as merely an extreme variation of iambic meter. It is true that these are all accentual-syllabic meters, and that individual feet may be substituted from time to time, depending on the base meter, the position in the line, and the genre. Nonetheless, they sound different, and they have different traditions in English literature.

The last two propositions are once again less fundamental. Next comes (9) We should never use four degrees of stress for scanning (p. 30-77). Wallace is here specifically objecting to scansion systems like that of Trager and Smith (An Outline of English Structure, G. L. Trager and H. L. Smith, Jr., University of Oklahoma Press, 1951), which mark four levels of stress and insist that no two adjacent syllables can have the same stress level. It is the latter point that is the problem: with such a rule, there can be no pyrrhics, no spondees, no anapests, no dactyls. The objection is not that speech has only two degrees of stress, for in fact it has far more than two or four. The problem is that meter is, for Wallace, intrinsically binary. It can be appropriate to note that a particular syllable in a stressed position is in fact weakly stressed (only worth 2, say, not a full 4), but this should be a supplement to ordinary scansion, not a replacement. As regards practical scansion, Wallace's proposal is sensible. He himself uses three degrees: unstressed, stressed, and weakly-stressed, or promoted unstressed. Most of the contributors, except for Woods, Steele, and Rothman, agree. Woods and Steele use four degrees in their own scansions. Rothman considers Wallace's rule "far too absolute" (p. 215), and points out that whether or not we agree with the work of the structural linguists, we have a responsibility to make it available to our students. "The point," he says (p. 215), "is not to agree with the linguists, who often disagree violently with each other, but to foster exchange with them, to create dialogue." Literary prosodists need to know how language works, and "to teach the subject in a historically and theoretically rigorous way" (p. 216).

Finally, the tenth proposal is a corollary to the ninth. (10) The spondee is a good, and fairly frequent, foot in English (pp. 37-41). If we use a binary system of scansion, we will occasionally have two stressed syllables together. Steele and Woods, preferring the four-stress system, do not permit true spondees; for them, one of the two stresses must be more stressed than the other. Everyone else, using a binary system, allows spondees, although Nims appropriately cautions "A foot, yes. But there is no such thing as spondaic meter" (p. 195).

Wallace sums up his proposals in a single sentence: "There is one meter in English: accentual-syllabic, and its base is always iambic." (p. 22) He clarifies this a bit in the final essay, saying that the tone of "always iambic" is misleading. Nonetheless, "In the continuum of equally valid feet that make our meter, the iamb will continue to be accounted the norm. The most frequent foot, it is also the most rhythmically neutral" (p. 295, his emphasis). One of his main goals is to simplify the way English meter can be presented to students, a desirable and humane idea. This is the reason behind the terminological proposals, and although I would not always choose the same words, I believe he is choosing the right concepts to label.

The other goal is to simplify the way poets and metrists conceive of English meter, and here I believe he is going too far. What is important at first, though, is not so much the answers as the questions. What, exactly, is meter in English? When is a poem metrical? How regular does it have to be? I think that both syllabics, which can be very strict, and accentual meter, which is necessarily quite loose, are meters for English, and I do believe in non-iambic accentual-syllabic meters. In reading Meter in English, however, I have had to test and examine those beliefs. The great merit of the book is not merely that it presents a theory of English meter but that it forces readers to formulate their own theories. The symposium format is particularly useful here. It's all too easy to read a book by a single author and accept arguments unquestioningly, or verify that the argument is internally consistent without looking critically at the underlying assumptions. It is very difficult to do that with a book whose individual chapters contradict and argue with each other. The book, as a whole, does not present the "right answer" to any of the fundamental questions it raises. Instead, it presents fifteen different answers and leaves the reader to judge among them. Wallace's essays on their own would have been valuable; the book is even more thought-provoking in its present form.