Navigation – Plan du site

In Defence of the Dolnik: Twentieth-Century British Verse in Free Four-Beat Metre

Défense du Dolnik: la poésie britannique du vingtième siècle en mètre variable à quatre temps forts
Derek Attridge
p. 5-18


Un certain nombre de poètes britanniques du vingtième siècle ont écrit des poèmes en en mètre variable à quatre temps forts, norme qui s’inscrit dans une longue tradition de poésie anglaise mais qui ne peut pas être analysée correctement en termes de prosodie traditionnelle. Une manière opératoire de désigner ce type de pied et le mot “dolnik, emprunté à la tradition de la prosodie russe qui l’utilise couramment. Cet article se penche sur des poèmes d’Auden, de Hardy, de MacNeice, de Betjeman et de de la Mare composés en dolnik afin de d’établir la gamme d’effets que peut produire ce pied et d’en souligner la permanence dans la poésie de la première partie du vingtième siècle

Haut de page

Texte intégral

1W.H. Auden wrote his contemporary ballad “As I Walked Out Last Evening” in 1937, and published it in 1940. It begins with a conventional ballad opening that quickly establishes the familiar ballad rhythm, though soon the setting is revealed as the modern city:

  • 1 I quote this poem, like all my examples, from the Norton Anthology of Poetry. I do this to highligh (...)

As I walked out one evening,
Walking down Bristol Street,
The crowds upon the pavement
Were fields of harvest wheat.
And down by the brimming river
I heard a lover sing
Under an arch of the railway:
“Love has no ending.”
I’ll love you, dear, I’ll love you
Till China and Africa meet,
And the river jumps over the mountain
And the salmon sing in the street1

2The lover’s protestations are met with a stern reminder from the city’s clocks of the power of Time, ending with two stanzas of direct appeal which manage to fuse raw despair with a remnant of faith in life and humanity, in a rhythmic form whose easy swing is an important part of their intensity and memorability:

O look, look in the mirror,
O look in your distress:
Life remains a blessing
Although you cannot bless.
O stand, stand at the window
As the tears scald and start;
You shall love your crooked neighbour
With your crooked heart.

After this, the poem ends with a calm but bleak coda:

It was late, late in the evening,
The lovers they were gone;
The clocks had ceased their chiming,
And the deep river ran on.

3The question I want to ask in this essay is: what is the best way of describing the metre of this poem? I should say right away that I don’t believe there is only one correct metrical description of any line; different types of description are good for different purposes. For my own part, I have always had pedagogical effectiveness in mind in discussing metre and rhythm, rather than the search for human universals or the annotation of minute phonetic details. In writing about rhythm, my central question has constantly been: how can we talk about these matters in a way that makes sense to those who are not specialists?—which means, for me, how can we best tap into the considerable rhythmic capability of the ordinary man or woman, girl or boy, who heard nursery rhymes as a child, enjoys the occasional dance, and listens from time to time to popular music?

4The most likely result of an attempt to scan these stanzas in a traditional way as accentual-syllabic metre would be something like the following:

5Having divided the lines into feet it would be necessary to explain that lines 1 and 5 contain trochaic substitutions; that line 6 begins with a pyrrhic substitution followed by a spondaic substitution; that line 7 begins with an anapaestic substitution; that lines 1, 3, 5 and 7 end with amphibrachs; that lines 3 and 8 are headless; and that line 6 contains a monosyllabic foot. (An alternative scansion would separate the final syllable of the odd-numbered lines, thus increasing the proportion of iambic feet at the cost of adding an anomalous half-foot.) Alternatively, we could try to scan it as trochaic verse, which would make line 3 entirely regular, and the last line regular but for a missing final unstressed syllable.

6There is not much to choose between these two scansions. The iambic scansion gives us 13 iambs and 11 other types of foot; or 17 iambs and 7 other types of foot plus four extra syllables. The trochaic scansion yields 15 trochees, 6 other disyllabic feet, and 9 monosyllabic feet or extrametrical syllables. Most recent developments in generative metrics also use a concept closely allied to the foot: for example, Nigel Fabb and Morris Halle, in Meter in Poetry: A New Theory, argue that all regular metres, in all languages, are made up of “groups”, each of which has one syllable as its “head”. In their discussion of English metres, they do make a sharp distinction between strict and loose forms: in the latter, the stress maxima determine the division into groups, rather than the other way round. In other words, in loose metres we perceive the beats first, or at least a certain number of them (a “maximum” in loose metres is defined as “the syllable bearing the word stress, . . . except when it is immediately preceded or followed in the same line by a syllable carrying greater stress” [68]), and the organisation of the line falls into place around them. This is a distinct advance on previous generative theories, which imply that all lines are made up, first and foremost, of a series of positions, which are then matched against the syllables—as if readers began with an abstract idea and then looked for sonic material to fill it. (Although it’s pleasing to find a generative theory that pays this degree of attention to the demands of performance, it’s a rather surprising concession, given Fabb and Halle’s resolute insistence that “the metrical grid of a poem is distinct from the grid of its recitation, which. . . consists of the assigning of a melody to a text”, and their assertion that “it is fundamental to our approach that metrical form is not given the same representation as performed (or perceived) rhythm or any other aspect of performance” (39). This refusal to take performance, or even perception, into account renders theories such as Fabb and Halle’s of little use in the literature classroom.)

7But why should we be attempting to discover, or impose, feet (or headed groups) at all? Auden has not written a poem involving a highly complicated set of variations on a standard accentual-syllabic metre, but a very familiar rhythm that the ear immediately latches onto. When we feel the rhythm of the lines in our musculature, as we do all strong rhythms, we don’t feel iambs or trochees, but a sequence of beats and offbeats, falling into the temporal patterns we’ve heard thousands of times since our early childhood. The following example is one way of representing this familiar rhythm:

8This scansion provides a visual representation of the movement of the verse, a movement that encourages the voice to bring out the regularity of the rhythm, including the natural insertion of pauses at the end of each line to allow for the fourth unrealised beat of the elementary four-line, four-beat-per-line structure. The many variations are shown as they are experienced: different ways of beginning the line—on a stressed syllable that takes the beat, or with one or two unstressed syllables leading up to the beat; moments of tension when the absence of a realised offbeat induces a compensatory lengthening of the first of two stressed syllable (“look, look”, “stand, stand”, and especially, since there is no punctuation to aid the reader, “tears scald”); and the patterned alternation of endings on a beat and an offbeat which runs throughout the poem. Because of the strongly rhythmic character of the verse, it is quite natural to give “your” in “your distress” and “With” at the start of the last line a degree more emphasis than they would receive in ordinary speech, so there’s probably no need to introduce the idea of promotion here. Auden has taken the traditional form of the ballad, and although his version is not intended to be sung, has retained the type of freedom between beats that characterised the sung ballad (because, of course, when there is a melodic structure to provide the rhythm, it’s easy to add or subtract unstressed syllables). The result is a poem with a strong narrative drive that can produce moments of powerful poetic resonance. (Auden’s other ballads from this period, “Miss Gee” and “Victor”, also written in 1937—the “black ballads”, I’m tempted to call them—exploit the jauntiness of the same form to create a pair of comic horror stories.)

9There have, of course, been many suggestions about naming this kind of verse; “loose iambics” perhaps being the most popular (though as we’ve seen, it sometimes wouldn’t make a huge difference to call them “loose trochaics”). But a name that distinguishes this verse-form more clearly from accentual-syllabic verse is desirable, in order to signal clearly its identity as a recognisable metrical genre in its own right, and one that doesn’t require any agonising about whether to divide up the line into iambic or trochaic feet. Although I hesitate to endorse the introduction of yet more foreign-sounding terms into prosodic studies, I find the term employed by scholars of Russian prosody, dol’nik, a useful one. Without the apostrophe in English, it’s short and memorable, and signals the translinguistic nature of the form in question, which is a staple in Russian and common also in German. (In Romance languages, where the syllable carries more of the rhythmic weight of the verse, such variations in the numbers of syllables tend to have the opposite effect, producing less, rather than more, regularity.) Marina Tarlinskaja, a pioneer in demonstrating the importance of the dolnik in English verse and its relation to its Russian and German equivalents, uses statistical methods in Strict Stress-Meter in English Poetry to show the distinctiveness, across the long history of English verse, of a variety of metre characterised by its use of both single and double offbeats (to employ my terminology), or both monosyllabic and disyllabic intervals (to employ hers). Tarlinskaja takes iambic metre to be the norm from which dolnik verse deviates, and finds that in a large sample of four-beat English verse there is a frequency gap between, on the one hand, verse with up to 10% of its offbeats double and, on the other, verse that has 20% or more of its offbeats double. In other words, there is a large amount of four-beat verse with a very limited amount of variation between single and double offbeats, verse which Tarlinskaja is willing to call “loose iambic”; and a large amount of four-beat verse with extensive variation in the offbeats, which is the form she calls “dolnik” or, less happily, “strict stress-metre”. She does not find a great deal of verse in between these two categories, suggesting that poets gravitate towards one or other type of verse, instinctively regarding them as different, and as having different poetic capabilities.

10Others have taken up the campaign to recognise the dolnik as a distinctive form. Thomas Cable, for instance, arguing against the notion that “strong-stress metre” is foreign while “accentual-syllabic” metre is native, has shown that the bulk of the medieval Harley lyrics are in a metre it would be appropriate to call dolnik. Martin J. Duffel, in A New History of English Metre, makes much use of the term—though to my mind he illustrates the danger that if it is employed too widely it loses its capacity to name a highly distinct verse form. That distinct form, I would argue, has the following characteristics:

  1. It derives its characteristic rhythmic swing from the four-line, four-beat structure;

  2. It increases that rhythmic swing by employing a variety of offbeat types, including single unstressed offbeats, double unstressed offbeats, single stressed (that is, demoted) offbeats, and implied or virtual offbeats (where the absence of a syllable demands compensation from the reader). Occasionally, double offbeats may include a stressed syllable, but such is the insistence of the metre that in performance these will usually lose most of their stress.

  3. It may avail itself of the particular freedom of the opening of the line, beginning without an offbeat, or with any of the types of offbeat just mentioned. Depending on the choices made at this point, it can feel like a version of iambic or of trochaic verse, or neither.

11To say that dolnik verse derives its rhythmic swing from the four-line, four-beat structure is not, of course, to say that all examples have groups of 16 realised beats; Auden’s poem, for example, has three realised beats per line. But they all capitalise on the rhythmicality and familiarity of the four-by-four form, even when they use lines of two, three, six, seven or eight beats. (Five-beat verse with varying offbeats seems to me to be a different form—a loose pentameter, that doesn’t engage with the strong four-beat rhythm—though Tarlinskaja includes it in the category of dolnik verse, while noting that it is very rare.) And the reason for emphasising, in my second point, that the variation in offbeats must be such that it increases the rhythmic insistence of the strict form is that not every variation will do this; some variations will reduce that rhythmicality or have little effect on it. Imagine if Auden had written:

O look, look in the mirror,
In your gloomy distress;
Life clutches at straws
For those who bother to bless.

Technically, I suppose one would have to call this a dolnik if someone perpetrated it; but it’s not in the spirit of the dolnik, since it is actually less rhythmical than the equivalent lines in strict iambic or trochaic verse would be, as this version indicates:

O look inside the mirror,
O look in your distress;
Your life remains a blessing
Although you cannot bless.

12Note that the strictness of the metre produces a more rigid rhythm, one that is less likely to induce a tapping of the hand or a nodding of the head. Tarlinskaja has provided some statistical evidence for the preferences shown by the dolnik, which no doubt relate to the differences in rhythmic insistence produced by variations in different positions. For instance, in English four-beat verse the incidence of double offbeats between the third and fourth beat is much lower than is the case between the first and second or between the second and third. Further study is needed to explain how these choices increase or decrease the strength of the rhythm.

13Of course, poets are never bound by formal categories, and there exist many poems that are hard to classify as either dolnik or accentual-syllabic (or, in the other direction, strong-stress). But there is a large body of English verse that conforms to the principles I’ve outlined. The nineteenth century was the great period of the dolnik—Coleridge could be said to have ushered it in with “Christabel”—but there are many fine examples in twentieth-century British poetry, Auden’s contemporary ballad being just one example. I’d like to look at a few instances, to demonstrate the continuing vitality of the dolnik and the range of uses to which it has been put in this body of verse.

14Like all verse-forms, the dolnik is capable of a variety of different realisations. Most commonly, its lines begin on an offbeat and end on a beat, giving an iambic or anapaestic feel to the rhythm; and usually there are not enough double offbeats to tilt the predominant movement from duple to triple. A master of the form—as he was a master of many metrical forms—was Thomas Hardy, and his poem “Neutral Tones” will provide an outstanding example of this type of dolnik. (I’m cheating a little in including it, since the poem was published two years before the twentieth century, and in fact written a good many years before that.)

We stood by a pond that winter day,
And the sun was white, as though chidden of God,
And a few leaves lay on the starving sod,
—They had fallen from an ash, and were gray.

Your eyes on me were as eyes that rove
Over tedious riddles solved years ago;
And some words played between us to and fro—
On which lost the more by our love.

The smile on your mouth was the deadest thing
Alive enough to have strength to die;
And a grin of bitterness swept thereby
Like an ominous bird a-wing. . . 

Since then, keen lessons that love deceives,
And wrings with wrong, have shaped to me
Your face, and the God-curst sun, and a tree,
And a pond edged with grayish leaves.

15The poem has 60 offbeats. 33 of them are single and 27 are double—not enough to promote a triple rhythm, though enough to prevent the poem from falling into a comfortable iambic rhythm. (The fourth line may be read with a rare triple offbeat, if we don’t reduce “fallen” to a monosyllable.) Three of the single offbeats are stressed syllables, accommodating themselves to the rhythm through demotion, something that is always possible when a stress falls between two other stresses: “few leaves lay” (line 3); “then, keen lessons” (line 13); and, most powerfully, “God-curst sun” (line 15). Where God had merely chided the sun in line 2, his rage at his creation has now so increased in ferocity that he curses it, and the stressed offbeat contributes to that strength. (I should emphasise that “demotion” does not imply a performance with reduced stress: it is a feature of English verbal rhythm that under the right conditions a syllable may be given its normal stress while functioning in the metre as an offbeat.) This use of single, stressed offbeats is a characteristic of many dolnik poems, causing the rhythm to modulate momentarily into a weightier but still very insistent sequence; it was a favourite device of Swinburne’s, and an important ingredient in the infectious swing of Elizabeth Barrett Browning’s “A Musical Instrument” (“What was he doing, the great god Pan. . . ”). Four of the double offbeats in Hardy’s poem include a demoted syllable—“solved”, “played”, and “lost” in stanza 2, none of which troubles the rhythm greatly, and in the last line, “edged”, which does, I think, contribute to the uneasiness of the inconclusive conclusion. By following three lines of four beats with one of three, rather than the more usual 4-3-4-3 pattern, Hardy ends each stanza with a dying fall. His use of the dolnik here pulls it away from the musicality and liveliness it often exhibits towards a speaking tone of peculiar hollowness—as if the cheerful rhythms had become merely dead echoes.

16I want to return to Auden for an example of a poem with a predominantly falling rhythm which makes use of the dolnik freedom at the start of the line, though it remains very regular thereafter: the well-known “Lullaby”. Here is the second stanza (the lower case b indicates a promoted syllable—in other words, an unstressed syllable which, because of its position between two other unstressed syllables, or at the beginning or end of a line, can function metrically as a beat while remaining without stress):

17Approached in terms of foot-prosody, this stanza is a puzzle: half the lines begin with an iamb and half with a trochee. In the rest of the poem, trochees dominate, so that may settle the question for our foot-prosodist—though we still have the problem that occurs with most so-called “trochaic” tetrameter, that the “trochaic” lines, as well as the “iambic” lines, end on a beat, and therefore their last foot is defective. (Unless, of course, they are actually headless iambic lines.) If we forget about iambs and trochees, we can say that this is duple verse with four realised beats per line which displays the dolnik’s freedom at line-beginnings but eschews it thereafter. The effect is of great simplicity of rhythm, tending towards a chant but kept from it by the varying line-openings, the extensive use of promotion, the ten-line stanza with its intricate rhyme-scheme, and the frequent enjambments.

18A different rhythmic quality is produced by Louis MacNeice’s use of the dolnik in “Bagpipe Music”. This is the last stanza of the poem, and I’ve indicated the rocking dipodic rhythm by using the lower case b for the less emphatic beats that alternate with the stronger beats:

19Here we have the typical dolnik variations between beats—no syllables, one syllable, and two syllables—and at the opening of the line—15 of the poem’s 34 lines begin on an offbeat, the remainder on a beat. Unlike Hardy, MacNeice avoids stressed syllables in the offbeats, keeping the rhythm light and ensuring that the infectious dipodic rhythm emerges clearly: another case in which the jauntiness of the rhythm contributes to the dark heart of the poem, in which a series of Scottish characters exhibit their cheerful improvidence while the social and political barometer keeps falling. (The poem, written, like Auden’s ballads, in 1937, was published in 1938.) In a poem published three years later, “London Rain”, MacNeice uses the three-beat dolnik with great success to formulate a philosophy of acceptance.

20It’s not possible to discuss the twentieth-century British dolnik without referring to John Betjeman, the best of whose dolnik poems is “Death in Leamington”. Yet again, the lilt of the dolnik is counterpoised against the grimness of the narrative. Nurse tries in vain to wake her patient for five-o’clock tea, and the poem ends as follows:

Do you know that the stucco is peeling?
Do you know that the heart will stop?
From those yellow Italianate arches
Do you hear the plaster drop?

Nurse looked at the silent bedstead,
At the gray, decaying face,
As the calm of a Leamington evening
Drifted into the place.

She moved the table of bottles
Away from the bed to the wall;
And tiptoeing gently over the stairs
Turned down the gas in the hall.

21Again, there is the typical dolnik variation between single and double offbeats both within the line and at the start (though most lines begin with a single offbeat). Line-endings are more regular: feminine and masculine alternate (as in “As I Walked Out”), except in the three cases where the fourth beat of the third line is realised, in which case the ending is masculine—“soul”, “cheeriness”, and “stairs” (though one might prefer to read “cheeriness” as an unusual triple ending). The sense of conclusion in these stanzas is enhanced by some departures from the most common patterns of the poem up to this point: the first is the only one in which every line begins with a double offbeat, leading into the three heartbreaking questions; the next stanza ends with the only beat-initial line in the poem, encouraging us to linger on the first word—“Drifted into the place”—and the penultimate line is one of the rare full four-beat lines, taking its time over the Nurse’s actions—“And tiptoeing gently over the stairs”—and leading into the plangent simplicity of the final line.

22My last example is a poem that has featured in many popular anthologies but has not fared terribly well in high literary circles: Walter de la Mare’s 1912 poem “The Listeners”. I’m sure I’m not the only one who encountered it at a young age, and found it instantly—and persistently—haunting. The brief story it told was part of its allure: the unnamed Traveller who knocks, for unknown reasons, at the door in the forest and receives no answer, unaware that he is being heard by a “host of phantom listeners”. But de la Mare’s brilliant exploitation of the dolnik rhythm surely played its part: catchy, musical, and with a rhythmic character not quite like that of any other poem. This is the memorable opening:

“Is there anybody there?” said the Traveller,
Knocking on the moonlit door;
And his horse in the silence champed the grasses
Of the forest’s ferny floor;
And a bird flew up out of the turret,
Above the Traveller’s head:
And he smote upon the door again a second time;
“Is there anybody there?” he said.

23De la Mare uses the full range of variants available to dolnik verse: not only the usual single and double offbeats, but several triple offbeats (as in “anybody there”); line-openings on the beat or with a single or double offbeat; masculine and feminine line-endings; and lines of 3 or 4 beats, arranged in two “short measure” 3-3-4-3 quatrains. This is not ballad-metre—we don’t find ourselves caught up in a swinging song-like movement, and in fact as the poem proceeds the patterns of beats become less easy to perceive—but is a fine example of a poet drawing on the age-old resources of dolnik verse to create something splendidly new.

Haut de page


Cable, Thomas, “Foreign Influence, Native Continuation, and Metrical Typology”. Approaches to the Metres of Alliterative Verse, ed. Judith Jefferson and Ad Putter, Leeds: Leeds Texts and Monographs.

Duffell, Martin J., A New History of English Metre, London: Legenda, 2008.

Fabb, Nigel & Morris Halle, Meter in Poetry: A New Theory, Cambridge: Cambridge UP, 2009.

Ferguson, Margaret, Mary Jo Salter & Jon Stallworthy, The Norton
Anthology of Poetry, Fifth Edition, New York: Norton, 2005.

Tarlinskaja, Marina, Strict Stress-Meter in English Poetry, Compared with German and Russian, Calgary: U of Calgary P, 1993.

Haut de page


1 I quote this poem, like all my examples, from the Norton Anthology of Poetry. I do this to highlight the inadequacy of the anthology’s essay on “Versification”, which provides no way of understanding the metre of the poems I discuss.

Haut de page

Pour citer cet article

Référence papier

Derek Attridge, « In Defence of the Dolnik: Twentieth-Century British Verse in Free Four-Beat Metre », Études britanniques contemporaines, 39 | 2010, 5-18.

Référence électronique

Derek Attridge, « In Defence of the Dolnik: Twentieth-Century British Verse in Free Four-Beat Metre », Études britanniques contemporaines [En ligne], 39 | 2010, mis en ligne le 11 février 2016, consulté le 27 juin 2017. URL : ; DOI : 10.4000/ebc.2797

Haut de page


Derek Attridge

University of York.
Derek Attridge is Professor of English at the University of York, with research interests in several different fields, including poetic rhythm, Joyce, South African literature, and literary theory. In the area of poetic rhythm, he is the author of Well-Weighed Syllables: Elizabethan Verse in Classical Metres (Cambridge, 1974), The Rhythms of English Poetry (Longman, 1982), which introduced the highly influential beat-offbeat approach to metrics, Poetic Rhythm: An Introduction (Cambridge, 1995), and, with Thomas Carper, Meter and Meaning: An Introduction to Rhythm in Poetry (Routledge, 2003). Other major publications include Peculiar Language: Literature as Difference from the Renaissance to James Joyce (Cornell and Methuen, 1988), Joyce Effects: On Language, Theory, and History (Cambridge, 2000), The Singularity of Literature (Routledge, 2004), J.M. Coetzee and the Ethics of Reading: Literature in the Event (Chicago and Natal, 2004), and How to Read Joyce (Granta, 2007). He is also the editor or co-editor of Post-Structuralist Joyce (Cambridge, 1984), Post-Structuralism and the Question of History (Cambridge, 1987), The Linguistics of Writing: Arguments between Language and Literature (Manchester and Routledge, 1987), The Cambridge Companion to James Joyce (Cambridge, 1990; second edition, 2004), Acts of Literature by Jacques Derrida (Routledge, 1992), Writing South Africa: Literature, Apartheid, and Democracy 1970-1995 (Cambridge, 1998), Semicolonial Joyce (Cambridge, 2000), and James Joyce’s “Ulysses”: A Casebook (Oxford, 2004). His most recent book is Reading and Responsibility: Deconstruction’s Traces (Edinburgh, 2010); forthcoming work includes Theory after “Theory”, co-edited with Jane Elliott (Routledge, 2011).

Articles du même auteur

Haut de page

Droits d’auteur

Licence Creative Commons
Études britanniques contemporaines est mise à disposition selon les termes de la Licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Haut de page
  • Logo Presses universitaires de la Méditerranée
  • Logo ERIH +