On Rhyme

It is sometimes heedlessly supposed that verbal consonance is a matter so elementary and self-evident as not to require discussion, but few “self-evident” issues prove indisputable, and this is not one. In his Essay on French Verse, Jacques Barzun discriminates five kinds of rhyme: feeble, sufficient, rich, Norman, and licentious. Of the rich rhyme he observes, “The consonants preceding the vowels must be alike: sommeil and vermeil, armant and charmant. The rhyme is still richer when the two words sound exactly alike: bois (wood) and (je) bois (drink). In English, this total sameness would make no rhyme at all; in French, it is felt as agreeably surprising in its juxtaposition of remote ideas—which is what Victor Hugo implies in calling it a ‘jumped-up pun.’” Such perfect identity of sound, as distinct from meaning, is precisely what Théodore de Banville recommends in stating: “Vous ferez rimer ensemble, autant qu’il se pourra, des mots très-semblable entre eux comme SON, et très-different entre eux comme SENS.” But in English, too close an identity of sound (weight / wait; some / sum) strikes the ear as banal, obvious, and flat, unless employed in a sestina where the remorseless repetition of terminal words is charitably relieved by such homophones.

In his introduction to The Selected Poems of Marianne Moore, T. S. Eliot has some characteristically shrewd things to say about rhyme. I want to quote a single paragraph of his, interrupting him at certain points to illustrate his comments.

In the conventional forms of rhyme the stress given the rhyme tends to fall in the same place as the stress given by the sense. The extreme case, at its best, is the pentameter couplets of Pope. Poets before and after Pope have given variety, sometimes at the expense of smoothness, by deliberately separating the stresses, from time to time; but this separation—often effected simply by longer periods or more involved syntax—can hardly be considered as more than a deviation from the norm for the purpose of avoiding monotony.

(Here, in a pre-Pope example, are some lines from Christopher Marlowe’s Hero and Leander, with enjambment and subordinate clause to lighten, however slightly, the weight of the rhymed couplets:

A post-Pope example may be supplied by Robert Browning’s “My Last Duchess,” with its own enjambments, to be sure, but also the effects of a brilliantly discontinuous catalogue, the very randomness of which enforces the Duke’s point and mutes, or perhaps voids, the resolutions usually provided by mating rhymes:

It may be noted, by the way, that in this poem of fifty-six lines, none but the fourth and final ones end in a full stop.)

“The tendency of some of the best contemporary poetry,” Eliot continues, “is of course to dispense with rhyme altogether; but some of those who do use it have used it here and there to make a pattern directly in contrast with the sense and the rhythm patterns, to give a greater intricacy. Some of the internal rhyming of Hopkins is to the point.”

(What Eliot is leading toward are those less and less obtrusive rhymes, not only buried “internally” but slighter by their transitory modesty, confined on occasion—in Marianne Moore, for instance—to an article, a mere schwa. They rejoice in lying unnoticed unless by the most patient and discerning reader. An example may be found in Brad Leithauser’s “In Minako Wada’s House,” of which the first four stanzas are quoted here:

A casual reading of the first stanza would recognize the rhyming of the third and sixth lines and would plausibly assume that the “internal” rhyme of away and day in the fourth and fifth lines was simple inadvertence, the sort of thing that happens commonly in everyday speech. Indeed, the phrase “folded away all day” doesn’t call much attention to itself as a musical chiming. But we go on to observe that in each ensuing stanza the terminal word of the fourth line rhymes with some word in the interior of the fifth, some of these fleetingly: for example, “crickets that are fed fish . . . ,” where the auxiliary verb, a slight thing, rhymes with jar. This slightness of rhyme consorts gracefully with the miniature elements of which the poem speaks. Such rhymes are often unnoticed and thus are not felt as “constraints,” which is the way rhyme is sometimes regarded; and this is partly due, though only partly, to the freedom of position allowed the rhyming word in the fifth line. I must add that only when I moved to Washington did I encounter anyone, in this case a building contractor, who pronounced “niche” as though it rhymed with “MacLeish,” a pronunciation that would have thrown a monkey wrench into Leithauser’s poem. Few poets have resorted to such furtive, concealed rhymes as W. H. Auden, whose disconcerting pleasure it was to rhyme an unaccented with an accented syllable, as Ransom did in the quatrain cited above:

Robert Graves claims that the rhyming of accented with unaccented syllables gives “the effect of uncertainty, incompleteness, suspense,” and it seems to bear something like the audible effect of some slant rhyming, a deliberate dissonance; though in the hands of some poets, it confers the effect of a tin ear. Auden for his part was, in the course of time, to become still more cunning. Poems like “Pleasure Island,” “Music Is International,” and “The Duet” rhyme the penultimate syllable of one line with the final syllable of the next, a technique that yields friendship / weekend as rhymes. In “The Managers” he simply reverses the order, rhyming the final syllable of the odd-numbered with the penultimate syllable of the even-numbered ones, which affords, among other curiosities, vows / flowers and sees / policemen. But it is time to return to Eliot.)

“(Genuine or auditory internal rhyme must not be confused with false or visual internal rhyme. If a poem reads just as well when cut up so that all the rhymes fall at the end of lines, then the internal rhyme is false and only a typographical caprice, as in Oscar Wilde’s Sphinx.)”

Oscar Wilde’s poem appears to the eye to be composed in unrhymed octameter distichs:

It takes no great shrewdness (the last line quoted is a dead giveaway, with its caesura falling so neatly in the middle, between two drearily parallel phrases) to determine that each of these octameter distichs can easily be recast as tetrameter quatrains, rhyming abba:

Eliot is right in calling this “typographical caprice,” though many of E. E. Cummings’s effects come to much the same thing. What we dislike about the lines by Wilde derives from our sense that he may believe his music is achieved by mysterious subtlety, when in fact it is too nakedly apparent, unredeemed by any real linguistic interest. But Eliot’s point about typographical setting raises some interesting puzzles. Arthur Golding’s 1567 translation of Ovid’s Metamorphoses is framed in English fourteener rhymed couplets, though every once in a while he extends his rhyme to cover three consecutive lines:

Though laid out differently on the page, the last three lines are identical in meter and rhyme pattern with “The Walrus and the Carpenter.” Continuing his discussion of internal rhyme, Eliot declares, “This rhyme, which forms a pattern against the metric and sense pattern of the poem, may either be heavy or light—that is to say, either heavier or lighter than the other pattern. The two kinds, heavy and light, have doubtless different uses which remain to be explored. Of the light rhyme, Miss Moore is the greatest living master; and indeed she is the first, so far as I know, who has investigated its possibilities.” He goes on to instance rhymes produced by hyphenation, and thus extremely transitory:

I leave off quoting both Miss Moore and Eliot, who has brought me to what, adapting a term from Gerard Manley Hopkins, may be called rove-over rhyme. The thirty-first stanza of Hopkins’s great and deeply serious poem “The Wreck of the Deutschland,” rhymes unconfessed of them with the breast of the and borrows the needed “m” sound that would complete the rhyme from the word at the beginning of the following line, Maiden. More daringly still, and in the same stanza, he rhymes Providence with it, and, borrowing the “S” of Startled from the beginning of the following line. Clearly this liberty, this vocal athleticism and audacity, can be justified by an impetuosity and rapidity of movement, both narrative and meditative, that urges the poem headlong through enjambments, accelerated by heightened devotion and sprung rhythm. The stanza deserves quotation, but it may be useful in advance to gloss its stanzaic form. It occurs in the second part of a two-part poem, the two parts differing ever so slightly in formal detail. Part One concerns the poet’s conversion from the Anglican to the Roman Catholic faith. Part Two is about the shipwreck. The stanza form of both parts is identical, except that the first lines of Part Two are longer by one foot than those of Part One. The rhyme scheme throughout is ababcbca, and the respective line-lengths (always allowing for the discrepancies permitted by sprung rhythm) run 3, 3, 4, 3, 5, 5, 4, 6.

It should be noted that, since lines 1, 3, and 8 rhyme (a pattern maintained throughout this thirty-five-stanza poem), Hopkins has linked pain, for the, vein for the, and grain for thee, thus mating the muted, inconspicuous article with the elevated, hieratical pronoun. This is startling, and resembles, I think, some of the light rhyming of Marianne Moore.

The exoticisms of such rhyming practices could legitimately lead to considerations of light verse, where virtuosity parades itself with ostentations of ingenuity. The field is too vast to compass, reaching as it must from Swinburne’s limerick “There was a young girl from Aberystwyth,” backward to Winthrop Praed and Thomas Hood, onward through Charles Calverley, Lewis Carroll, W. S. Gilbert, James Kenneth Stephen, and Owen Seaman, to American song lyricists, Cole Porter being only one of these, and then spreading in every direction. But, abandoning any such ambitions, we may consider the narrower domain of deliberate and of naive incompetence. Both furnish pleasures, though of different kinds, and although incompetence is common enough, it rarely achieves the distinction of the ludicrous so carefully culled in The Stuffed Owl and its sequel, Pegasus Descending. From the first of these, we may savor some crudités from the poetry of Julia Moore, the Sweet Singer of Michigan.

Ineptitudes here only partially concern rhyme, but in their aggregate they so much amused the public of her day (she died in 1920), including Mark Twain, who commended her work in the press, that her book ran to three editions. That she was an untutored farmer’s wife who was not only deficient of ear and without skill as a poet but who took at face value the critical approval, often conferred by commentators convulsed by efforts to smother their giggles, adds a note of pathos and a flavor of shame to our pleasure in her work. In this she resembles the singer Florence Foster Jenkins, who enjoyed the same sort of barely concealed ridicule in the world of music.

But there is also a poetry of mock-incompetence, whose chief modern exemplar is Ogden Nash. Though a superbly deft artificer of complex verse forms, Nash most commonly adopted the composition of poems in rhymed couplets in which the first line tends to be long and the second incomparably longer, as though the poet, desperate to find some rhyming mate, is obliged to push on, awkwardly tottering toward his distant goal through barricades of necessary exposition and grammatical necessity, like the insomniac in the following lines:

So much has this kind of rhyming (feigned incompetence) become Nash’s trademark that its distinguished genealogy is usually forgotten.

There are still earlier examples, provided by Jonathan Swift. Shakespeare, a supremely accomplished maker of verse, was therefore capable of foisting the authorship of wonderfully incompetent doggerel upon his buffoons, most famously in the “tedious brief scene of young Pyramus / And his love Thisby: very tragical mirth.” Perhaps less well known are the lines spoken by Holofernes about Moth, a boy who, in mute charade, represents one of the Nine Worthies in his infancy, in Love’s Labour’s Lost.

John Dover Wilson called the rhymed couplets spoken by the Player King and Queen in Hamlet “deliberately commonplace . . . so as to provide a rest for the audience after the excitement connected with the dumb-show and the prologue.”

From these varieties of inferior poetry we must turn in another direction and consider the fact that partly, perhaps, because much poetry secures its rich and powerful effects without rhyme—Homer, Virgil, Paradise Lost, William Wordsworth, Walt Whitman, and many others—rhyme is relegated to the province of decorative adornment, an inessential luxury that may be regarded as effete (coinciding with its identification with light verse) or dated, a throwback to medieval preoccupations with arbitrary impediments, self-imposed constraints, spiritual and imaginative submission to rigorous laws and self-mortifying discipline; or else with the tinkling verse of troubadours and minnesingers. Such a view of rhyme as fundamentally ornamental fails to do justice to its versatility, for when properly integrated into a poem of serious character, it becomes an operational part, a functional instrument of the work. A poem may usefully be thought of as a gestalt or congeries of elements assembled in delicate balance; and although any given poem may dispense with some of these elements, the ones employed ought to cooperate smoothly enough to give the effect of a happily conceived ensemble. Of the single element of rhyme, Robert Graves observed, “it must come unexpectedly yet inevitably, like presents at Christmas, and convey the comforting sense of free will within predestination.” However lightheartedly Graves may resort to this theological language (a slightness indicated by “presents at Christmas”), what he is saying is not mistaken or trifling. It is put analogously, with regard to ornamental elements in architecture, by John Ruskin. “Observe,” he writes in The Stones of Venice,

that the value of this type does not consist in the mere shutting of the ornament into a certain space, but in the acknowledgement by the ornament of the fitness of the limitation;—of its own perfect willingness to submit to it; nay, of a predisposition in itself to fall into the ordained form, without any direct expression of the command to do so; an anticipation of the authority, and an instant and willing submission to it, in every fibre and spray; not merely willing but happy submission, as being pleased rather than vexed to have so beautiful a law suggested to it, and one which to follow is so justly in accordance with its own nature. You must not cut out a branch of hawthorn as it grows, and rule a triangle round it, and suppose that it is then submitted to a law. Not a bit of it. It is only put in a cage, and will look as if it must get out, for its life, or wither in the confinement. But the spirit of triangle must be put into the hawthorn. It must suck in isoscelesism with its sap. Thorn and blossom leaf and spray must grow with an awful sense of triangular necessity upon them, for the guidance of which they are to be thankful, and to grow all the stronger and more gloriously. And although there may be a transgression here and there, and an adaptation to some other need, or a reaching forth to some other end, greater even than triangle, yet this liberty is to be always accepted under the solemn sense of special permission, and when the full form is reached and the entire submission expressed and every blossom has a thrilling sense of its responsibility down to its tiniest stamen, you may take your terminal line away if you will. No need of it any more. The commandment is written in the heart of the thing.

Tony Tanner, who cited this remarkable passage in his fine book Venice Desired, remarked that it accommodates architectural elements to “laws at once natural and religious,” both of these domains having been of great concern to Ruskin. One detects somewhere lurking behind his paragraph the medieval dictum that perfect freedom is perfect obedience to perfect order. There are those who will find this attitude, and the way it is stated, mystical, fuzzy-minded, and silly. Among other things, while nominally writing about carved stone, Ruskin makes free with pathetic fallacies, permitting ornamental details to “acknowledge” the “fitness” of limitations, express a “predisposition” to submit to “ordained forms,” a submission not only “willing but happy,” and so on. Ruskin himself, in his earlier Modern Painters, had been highly suspicious and condemnatory of the pathetic fallacy, and it clearly has something to do with capacities for empathy and with the imagination itself. “Art, being a thing of the mind, it follows than any scientific study of art will be psychology. It may be other things as well, but psychology it will always be,” declares Max Friedländer, and the question such an assertion raises concerns whose psychology is involved. That is to say—with specific regard to rhyme—are these elements to be regarded as “constraints,” limiting the freedom of the poet? Or are they useful devices that, astutely linked with metrical patterns, can provide the raising of expectations (as regards an anticipated rhyme), the calculated delay or even disappointment of such expectations (as, for example, George Herbert provides in “Denial”), surprise at the fulfillment of the rhyming sound, either by the use of an exotic, unexpected word (often a feature of light verse) or because the chiming lines are so widely spaced that the suspended sound, awaiting concordance and resolution, has all but been forgotten by the mind’s ear? Further still, are they—continuing to regard rhymes as psychological devices—simply a way of preoccupying some part of the poet’s mind, liberating it to unconscious fluencies or filling it with unforeseen suggestions? Can “Christians” and “resistance” taken together propose some line of thought? Further still, and nearer to Ruskin’s paragraph, does the matching of rhymes satisfy some curious but deeply human craving for a formal order that is meant in homage to some universal order we need to posit, if not out of reverence, then out of dread of its possible absence? Or is there something in the human psyche that delights in order “for its own sake,” wherever it may be found, in symmetries, harmonies, repetitions, and resolutions? And can an answer to these questions be found that would apply equally well to nursery rhymes, Alexander Pope’s couplets, and a Stephen Sondheim song? We should not be surprised if, among the poets themselves, we find a few who make light of the matter:

This is the twenty-first stanza of Byron’s Beppo, a poem of ninety-nine such stanzas that from time to time indulge in acrobatic, virtually contortionist rhyming. Here we meet the heroine, and given the libertine nature of the tale to be unfolded, already hinted at in one of the poem’s epigraphs, we may take it that her pseudonymity is emphasized for several reasons: (1) gentlemanly discretion, linked by literary tradition to the practice of the Latin poets and their Renaissance heirs, who bestowed fictive names on their loved ones to protect them from scandal or from the jealous reprisals of their husbands; (2) a masquerade appropriate to the spirit of Carnival, during which the events of the poem take place; (3) a slightly provocative gesture by the poet (“Her real name I know not”), whose very protestations of ignorance invites us to suspect that he knows more than he will be telling and raises the whole question of how fact and fiction are to be played with; (4) the proposed name of “Laura” may have been chosen to contrast ironically with Petrarch’s Laura, an icon of purity; and (5) the final couplet suggests that the name, from the poet’s point of view, is a mere poetic convenience, chosen possibly because at some future point he will find it expedient to rhyme it with “Flora,” “aura,” “begorra,” or “Gomorrah” (some rhymes, like “fedora” or “Floradora,” were not available in 1818). He would, of course, have been quite willing to use any of these. As it happens, he didn’t use any of them, and from this fact we may surmise that when he was composing his twenty-first stanza he may have been keeping open his rhyming options, or else that when he says he has chosen the name “because it fits into my verse” he means only, and lamely, that it doesn’t violate his meter; though we know perfectly well that if meter is his only consideration, “Polly;” “Gerty,” or “Suky” would have worked just as well. The topic invites our attention because it deliberately raises the issue of how formal considerations may require the poet to reorganize the literal world to make it conform to artistic goals.

Just as lightly, here is a stanza from Part One of Auden’s Letter to Lord Byron, in which the poet pretends to anxiety about whether his publishers will be irked or angered by his frequent Byronic digressions.

When we too strongly feel the coercions of rhyme we know (if all else, meter especially, is flawlessly handled) that we are reading light verse; whereas coercive rhyme can be one of the blatant signs of poetic incompetence, which may be why some not altogether skillful poets avoid it. But we must turn from this levity to more serious matters.

John Dryden, addressing the earl of Orrery, for whom a model of the instrument, a replication of planetary order, invented by George Graham, was made by an instrument maker named Rowley and dedicated to the earl, defends rhyme by approving Sir Philip Sidney’s claim that it aids memory and, on his own authority, adds that it discourages prolixity. At a certain early stage in their careers, Eliot and Pound, at Pound’s suggestion, applied themselves to the study and imitation of Théophile Gautier as a corrective to the indiscipline of fashionable vers libristes. Eliot produced his quatrain poems, while Pound was famously to write:

“Rhyme” here surely means poetry itself, the whole kit and caboodle, its metrics, stanza form, traditional dignity and reserve, classical grace and decorum. This may be inferred from previous lines, as well as from the poem and doctrines of Gautier’s ars poetica that Pound means to honor:

(Johann Wolfgang von Goethe, too, in Lied und Gebilde, was drawn by the idea of the relation between verse and sculpture.) Stone is resistant, demanding; it lacks the fluidity or limpness of the “prose kinema,” the duplicability, hence susceptibility to forgery, of a “mould in plaster,” those factory-cast statuettes or bland cinematic tales that “the age demanded.” And “demand” is obviously for Pound part of the “supply and demand” language of modern economic parlance, in which art has become merely a servile commodity. The carven skill of Gautier’s verses are only emphasized by the brevity of his lines, bringing his rhymes close together, and by the strictness of his form. The likening of verse to sculpture involves not only the challenge, the difficulty, but the creation of something manifestly durable: the final word of his poem is résistant.

The formality here exemplified implies a relation between rhyme and metrical form, a matter about which there is no orthodoxy. Usually, in a stanzaic poem, each line is mated with at least one other line; and if the stanza pattern involves an odd number of lines, the common practice rhymes the odd-numbered final line with one of the previous lines, unless deliberate dissonance is desired, as in George Herbert’s “Denial,” cited earlier. There is a curious exception to this rule that I think can best be explained musically. In his book on the history of music, Donald J. Grout reports that “Luther himself wrote many chorale verses, for example, the well known Ein’ feste Burg (‘A mighty fortress’); it has never been definitely established that Luther wrote the melody of this chorale (first printed in 1529), though the music is generally attributed to him.” The chorale’s text is a nine-line stanza, with its first eight lines matched in this rhyming pattern: ababccdd. The remaining final line is unrhymed with anything, either in the original German or in the familiar English translation, which is fairly faithful to the sense and absolutely trustworthy as regards the rhyming pattern. And if we look merely at the printed words, the final line stands out, in terms of word-music, like a sore thumb.

This curious phonic discrepancy is concealed, or perhaps we may say healed and harmonized, by the musical setting, which recalls the exact notational phrase of music that occurs in the second and fourth lines, thereby producing not only a welcome cadence but an echo, and thus suggesting a rhyme that is, in fact, not there.

Poets have written from time to time about their struggles with rhyme, Ben Jonson (in “A Fit of Rime Against Rime”) being one of the earlier ones, and even Chaucer complained (in “The Complaint of Venus”) of the “skarsete” of rhyme in English, while in Sature, Eugenio Montale, in a poem called “Le Rime,” wrote (in William Arrowsmith’s translation):

In the original, the first and last lines truly rhyme (delle / quelle), and other lines employ slant rhymes, disguised or outwitted by the poet, or cunningly sneaked by: supportano / allontana / ardono. The Italian language simply has more rhyming words than English, allowing it to perform rhyming feats that either would be impossible in English or would sound to us like a tour de force, while quite unstrained and easygoing in Italian. Jacopone da Todi composed a manifestly serious, not to say solemn, poem called “De la incarnazione del verbo divino.” It is written in quatrains with the first three lines rhyming with one another, while the fourth line hangs suspended, apparently rhyming with nothing, like the last line of Luther’s hymn. But it turns out that the stanza’s fourth line rhymes with the fourth line of the following stanza, and furthermore with the terminal line of every ensuing stanza, and there are eighteen of these. The poet admittedly permits himself three repetitions (Signore, amore, and fiore), but the effect is persuasive in the redundant manner of Gregorian chant. This effect is virtually unavailable in English. Multiple rhyme either gives the effect of levity (as in Hardy’s “The Respectable Burgher on the Higher Criticism,” which parades thirty-six consecutive rhymes of the same sound) or else a sense of nervous obsessiveness approaching hysteria, as in Sylvia Plath’s “Daddy.” And in “The Ballad of Reading Gaol,” Oscar Wilde, through an excess of rhyme, contrives to undermine the very sympathy he must be trying to enlist.

The labored irony here is due, at least in part, to excessive rhyme and the repetition of words: lutes / flutes; sweet / feet. Dance is mentioned four times, sweet (a hazardous word) twice. The crude contrast between heedless bliss (of a slightly Bacchanalian kind, with dancing to the accompaniment of flutes; the lutes seem thrown in for rhyme) and hanging by the neck until dead is made grotesquely melodramatic by these chimings, and the ironies are almost adolescent. An 1811 Dictionary of the Vulgar Tongue gives “dance upon nothing” as a standard formula for someone who is hanged.

I have found myself put off by the rhymes of one stanza of Edwin Arlington Robinson’s otherwise deeply moving and powerful poem “Eros Turannos.”

What bothers my ear is the role played in this stanza by rhyme, conjoined to a catalogue of perilous abstractions and more or less mechanically grammatical parallelisms. The four-syllable mating of inaugurates / reverberates alternating with the three-syllable mating of confusion / illusion / seclusion is conspicuous enough. The parallel construction of the first two lines with the two that follow, down to identical articles and prepositions, identical placement of participles, nouns, and verbs, is so annoyingly efficient and neat that we are prevented from investing any sympathy in the plight of the woman whose unhappy fate is being described, and we are diverted by a sense that approval is being invited in behalf of the symmetries of the language. Such laborious rhyming, enforced upon the reader’s notice by other constructive elements, recalls us to Eliot’s careful discrimination between heavier and lighter rhymes, and alerts us to the serious problem of how much attention rhyme ought to call to itself. There are those poems that rhyme only occasionally and irregularly, of which more in a moment. But because we are now discussing rhymes as obtrusive and conspicuous, a word perhaps ought to be said about bouts-rimés. The story goes that a French poet of no great consequence, named Dulot, reported in outrage to his friends that his home had been burglarized, ransacked, and that some of his most valuable papers were missing, including some three hundred sonnets. No one believed that he could have written so many, and under the pressure of skeptical questioning he admitted that he had worked out only the rhyme-schemes of these and planned to fill in the poems at some later time. This was thought to be quite comic, and it became a fashionable parlor game. Someone proposes the terminal rhyming words of a sonnet, with which none of the other participants have any foreknowledge. Each player is invited to compose a sonnet on the spot, using the rhymes in the exact order in which they have been proposed. Judgment is awarded on the basis of speed and merit, and sometimes the competition is made especially challenging by improbable and far-fetched rhymes. Dante Gabriel Rossetti was reputed to be unusually good at these games, able to produce a respectable sonnet in five or six minutes.

Lighthearted and perhaps frivolous as the game may be, it points to some serious issues. Rhymes are not merely limiting constraints but clearly also suggestive elements that can lend direction to a poet’s imagination. If he or she is a bad poet, sooner or later we will feel a sense of bondage by the rhymes. If skillful, our feelings are likely to approve different effects suitable to varied poetic strategies. Sign / dine is not a striking rhyme until adopted for his purposes by Pope: “The hungry Judges soon the sentence sign / And wretches hang that jurymen may dine.” This confrontation of pleasure and execution is not so remote from the lines of Wilde quoted above, but it is far more effective by virtue of its terseness and economy, to say nothing of its smooth incrimination of the justice system, which Wilde’s lines make no mention of, at least explicitly. But the illustration of well-deployed rhymes is a thankless task, their fittingness being so integrally dependent on the poetic framework of which they are a part.

Even the degree of integration of rhymes will vary from poem to poem. And with Matthew Arnold’s “Dover Beach” and Thomas Hardy’s “Last Look Round St. Martin’s Fair” we encounter a liberty of rhyming that could have been found as early as the seventeenth century in France but was something of a novelty in English. These are rhymes that occur according to no fixed pattern and at the ends of lines of varying and unpatterned length. After the regularities of the towering Victorians, this might almost be called “promiscuous rhyming.” It occurs also in the early poems of Eliot: in “Prufrock,” “Portrait of a Lady,” and “Rhapsody on a Windy Night,” and he would continue this practice right through segments of Four Quartets, while varying his style with greater or lesser strictness and attaining a rich variety thereby.

This reference to Eliot’s virtuosity of rhyming recalls me to something quoted earlier in his remarks about Marianne Moore. He speaks of rhyme “which forms a pattern against the metric and sense pattern” of a poem, and this contrariety, the more striking in poems of formal regularity, affords a pleasure akin to syncopation, in which internal movements oppose one another in some artistically satisfying way. And I would like to illustrate this with some stanzas of a poem by James Merrill called “The Blue Grotto.” The poem is composed in five-line trimeter stanzas, rhyming ababa, and I will confine my comment to the first four stanzas, addressing them two at a time.

The rhymes of the first stanza are modest and without ostentation. The first line involves enjambment, running smoothly into the second and lightening the rhyming burden of the transitional preposition “into.” Impasse is a curious word, a French-English half-breed, subject, accordingly, to a number of pronunciations, one of which, with the dictionary’s authority, provides a faultless rhyme with glass in the fourth line. But because it appears at the end of the second, we may approach it with a gingerly uncertainty about how it is to be sounded. The approved French pronunciation would rhyme it with Alsace, making glass as a rhyme sound flat, nasal, and “American.” The poet could, of course, have intended just this effect. Indeed, poets have sometimes surprised us by delayed indications of what pronunciations they intend. Here, for example, is a double dactyl by George Starbuck, titled “Said,” in which the quasi-literacy of the speaker is surprisingly revealed.

Merrill’s rhymes in his first stanza, though not simple, are confined to single syllables. But in his second stanza the second and fourth lines offer disyllabic rhymes, and unexpected ones at that. The rhyming of the noun fact with the transitive verb thwacked and the intransitive verb react provides a subtle and important pleasure, amounting to a rule of thumb about rhyming: those rhymes are best which mate different parts of speech, like into / foreknew / blue in the first stanza. Of these first two stanzas we may go on to remark that as a world-renowned tourist attraction, the Blue Grotto at Capri enjoys a celebrity different from Hollywood stardom, and more akin to the reverence accorded an operatic diva. Its attractions, spread by word of mouth, travel posters, and postcards, are known to multitudes who have never been there. And such grandeurs are always a challenge when subject to someone’s first personal experience. There is always a possibility that what we have allowed ourselves to hope for in imaginative anticipation and what we come up against as “fact” will fail to coincide. And this can mean either that the place or experience is not all it’s cracked up to be or else that we ourselves, our perceptions, sensibilities, are somehow at fault. As Edmund White reports of a conversation with David Kalstone, “If I’d say I’d decided I didn’t really like Goethe, David would say coldly, ‘Goethe is not being judged.’ ” Just who is under scrutiny, subject, object, or both, together with the self-conscious posturings engendered by this “test,” is a probing part of what this poem is about. The first two stanzas balance upon the fulcrum of reputation and actuality, first felt from the point of view of the visitor, but then, as “expectations / Dispersed.” to be replaced by the real thing, we seem to shift to the almost personified, foot-tapping impatience of the place itself, awaiting, like a Metropolitan diva, the tribute of customary ovation; and then, once again refocusing on the visitors’ sense of the pressures imposed upon them by the force and majesty of the occasion. Returned by the final line of the second stanza to the viewpoint of the visitors, we are now to examine their individual reactions.

The dexterity of rhyming is impressive. The third stanza offers disyllabic rhyming throughout, while the fourth employs it in the second and fourth lines. Water / Pater is a witty sight-rhyme. But to avoid laboring the matter, it seems to me that Merrill’s rhyming in this poem operates like those cut-out shapes, fins, or terminal blobs of a Calder mobile, each contributing its own carefully calculated weight that must be reckoned in achieving the balanced yet flexible effect of the whole. That is, the rhyme is experienced as working within a field of forces that include meter and syntax, line-length, word order, diction, grammar, and tone. Diane gets one line; Don gets two; Jon and Jack share three lines, with Jon getting the larger portion, while Janet is allotted three lines to herself. These disparities are displayed against the formal regularity of the tight trimeter five-line stanzas with their demanding rhyme-scheme, each calling for the triple chiming of first, third, and fifth lines, interspersed with the less demanding harmonies of the second and fourth. And this, I think, is what Eliot was admiring in the work of Marianne Moore when he wrote of “rhyme, which forms a pattern against the metric and sense pattern of the poem,” and which I have characterized as syncopation. Merrill’s stanza pattern, made the more demanding by the brevity of his lines that congest the rhyming words into very close quarters, is imposed upon an amusing catalogue of quirky reactions to the testing sight (or test-site) of the famous grotto, and with psychological insight reveals, like an inkblot test, more about the observer(s) than the observed.