By ‘caesura’ I refer to the intralineal syntactic break. In some ways it would be better to call it that, or at least some less unwieldy alternative – but the caesura’s currency is, alas, far too wide. Like most other prosodic terms it belongs to the language of classical quantitative prosody, which is where we might have left it.

General remarks on the caesura are perhaps of limited value, as its effectiveness will very much depend on the extent to which its deployment enhances the meaning and performance of the individual line. Each case will be different. While the caesura has metrical consequences, in English it is not (unlike its French counterpart) a metrical effect per se; the caesura only becomes one when its occurrence is regular (as that which divides the Anglo-Saxon hemistich). The caesura is the product and evidence of an expressive tension between the poet’s speech and their metrical template; together with enjambment, it is often a sign of the kind of formal-syntactical disagreement that we read, quite correctly, as evidence of strength of feeling. The symmetry of the template provides a contrastive means to make salient the naturalism and authenticity of our speech; it does this by foregrounding the asymmetric nature of the often complex, hypotactic syntax (the language of argument, consideration, qualification, conditionality, parenthetical aside) that we need as proof of both the vulnerable humanity and intellectual sophistication of poetic thought. Not only will speech-clauses run expressively counter to the lineal template; the template may also insist on pauses and hesitations that the speech, if presented in an unlineated form, would simply not imply. Thus poetry’s expressivity is increased via the friction, grinding, tension and fracture introduced through the disagreement between lineal and stanzaic form and speech. It may therefore be regarded as a higher-level metrical disagreement of the kind we see in lower-level metrical stress position/lexical stress tensions. (Indeed there’s some mileage in considering enjambment and caesura forms of higher-level metrical tension.)

It’s probably not worth giving a systematic account of the caesura, since it’s an anti-systematic feature of the line whose purpose is to generally to help it sound more organic and improvisatory. Though it need not interrupt the metre the way prosodic variation does, it shapes metre to its will, inserting pause, silence, breath, cadence and inflection; it naturalises and softens metre’s more mechanical effects to give the line a plausible, spoken-word verisimilitude. I feel it should emerge just as unselfconsciously in poetic composition; the caesura should be the mere product of the agitation between the isometric, recursive fixity of the template and the fluid rhythms of real speech, not a matter of conscious deliberation. (Unless it is a feature of the template.) Its successful effects are achieved by virtuosity alone; no one I know thinks of ‘where to put in a caesura’, the way they might sometimes do the line-break.

The caesura appears to ‘insert’ a pause or a silence within the line, but in reality merely is a pause that arises through a syntactic not a metrical exigency. It generally falls between clauses, and is almost always marked by punctuation or by some other typographical convention (such as the space employed by certain Beat and Black Mountain poets). My certainty here is controversial. Standard definitions also allow for caesurae to be merely inferred by the reader, and ‘placed’ at unmarked pauses on phrase-boundaries. However, such latitudes lead to immense confusion, as well as the subjective diagnosis of pauses the poet frequently did not intend. When poets want a pause, silence or a breath, they will generally indicate it. The caesura should only rarely be inferred or divined from the syntax where no typographical prompt is given – and even then, only as the stylistic elision of a punctuation mark. (Unless, of course, it forms a diagnosis of deliberate omission or incompetence.) But if a poet appears have run two clauses together, they probably didn’t want you to pause for breath.

Like the line-break, the caesura can be classified as masculine or feminine if the last syllable before the indicated pause is strong or weak.1 (Examples are from Richard Wilbur’s ‘The Mind Reader’.2)

       Lobs up a blink of light. The sun-hat falls (masculine)

                                 x  /      |    x   /

[…]

There would be obfuscations, paths which turned (feminine)

                                 x   /  x   |   /        x          /

Much as we saw with the line ending and line-beginning, the caesura accommodates metrical variation more smoothly than would be possible without it:

The mind is not a landscape, but if it were,

                              /       x   |   x   /   x   /

Yet seem, too, to be listening, lying in wait

                                /   x   x   |   /   x   x   /

The notion that a pause may fill up the missing part of a metrical line belongs to discredited temporal theories of metre, but it is a mistake to dismiss this out of hand. It holds perfectly well that within the metrical line-unit a pause can occupy the place of a weak syllable with little disruption. It’s natural to have the caesura occasionally stand in for a w; a comma, say, is a genuine temporal gap, and can be easily read as a syllable-substitute, as it makes the reader physically perform the weak placeholder. This coincidence of omitted syllable and punctuation also lengthens and emphasises the caesura, and to a degree ‘times’ it:

[…] And realisation of it rages out

In furnace-fear when we are caught without

People or drink. Courage is no good:

            x   /   [x]   /   x   /   x   /

It means not scaring others. Being brave

Lets no one off the grave.

(PHILIP LARKIN, ‘Aubade’)4

However, classifying these variations as hypercatalectic, acephalic and so on strikes me mere trainspotting, and gives the false impression that the caesura is as metrical as the hiatus and the ghost metron. It can, however, be exploited for metrical effect.

The variation of caesural placement is the sign of a supple and adaptive syntax, evidence that the template has not dictated the kind of sentences the poet writes, but instead – like enjambment – encouraged some productive resistance between syntax and lineal frame. As I’ve mentioned, the pattern of concurrence and disagreement between phrase length and line length is closely analogous to the lower-level relationship between lexical stress and metrical position; caesura is a kind of free variable operating between the two. It mitigates against any metrical extremes: too little syntactic resistance to the template can produce an uncomfortable isometrism that results in something we can experience as robotic speech.

In i.p. and longer lines, the caesura affords us a means of quietly entertaining the reader through variation. (In shorter lines – because the neat coincidence of line-break and phrase-length is far easier to casually engineer, something readers instinctively understand – its overuse is more conspicuous.) As usual, our default is variation within repetition. Lines which do nothing but vary and disagree – i.e. which constantly enjamb, constantly vary the placement of the caesura – become highly predictable. The virtuosic and sparing use of enjambment and caesura is a sign that the line has a robust purpose: the poet feels strongly enough about something to go against the grain of the metre. It is perceived by the reader unconsciously as another level of interest in the language, providing them with unexpected places to breathe and pause, lengthening and slowing the line as it reaches the phrase boundary; it is one of the quietest ways of ‘ringing the changes’. The analogy doesn’t survive close inspection, but the poet might think of it as strengthening the poem in the way that irregular verticals strengthen brickwork: a poem whose vertical mortar-joints are aligned is less strong that one which varies them.

We might make a rough classification with the terms masculine, feminine, medial, terminal and initial (though the last three are inevitably vague). If we also take account of line endings, we can quickly account for the poem’s lineal variation. For now I’ve just marked the lines as end-stopped, half-stopped (i.e. enjambing at a clause boundary) or enjambed, though using my ‘cline of disruption’ will give more granularity to the description.

(This is E.B.B. giving another masterclass in how one manages caesural and line-end variation within an unforgiving formal frame; poets will learn much from a careful study of her technique.) We might then abstract the following common-sense rules regarding caesurae in medium-length lines (i.e. of 8 to 12 syllables):

a) medial caesurae are generally the least conspicuous and are felt as least disruptive, since they occur at the furthest distance from both the line beginning and line ending.

b) caesurae nearer the end of the line are generally more disruptive, since the line-break will then occur soon after the start of the new phrase, whether it is enjambed or not.

c) caesurae at the start of the line are generally the most disruptive, since we don’t expect the line to be interrupted so early after its inauguration, especially if the previous line was end-stopped. (This is borne out by the classic miscounting error, where a stress is dropped in the line immediately following an enjambment, if that line contains an initial caesura.)

The relative frequency of the initial, medial and terminal caesurae roughly reflect these rules.

As discussed elsewhere, the use of metre and rhyme can mean fixing some elements of the line earlier in the compositional process than others, and the syntax then has to prepare for their effortless fall. Syntax is the capacious variable that restores the combinatorial possibilities that formally fixed elements can drastically reduce, especially in our post-poetic diction era, when hyperbaton is regarded as an embarrassing anachronism. (Rhymed poems often produce more varied caesurae than unrhymed poems – and far more than in free and irregular-lined verse, which often finds itself shaped to its clause-lengths, since the form offers no particular resistance to its doing so.) In formal verse, a fluid and adaptable syntax will often make use of caesurae to naturalise speech which otherwise might draw attention to itself.

1 An old convention for describing ‘strong’ and ‘weak’ endings that I observe here with reluctance. I will, however, attempt to reclaim the words for neutral use, and remind readers that they originate not in a sexual power-metaphor but in a quirk of French, where words of feminine grammatical gender have the tendency to end in a more lightly stressed syllable, and words of a masculine in a more heavily stressed. Passing the buck, but take it up with them.

2 Richard Wilbur, The Mind Reader: New Poems (San Diego, CA: Harcourt Brace Jovanovich, 1976).

3 Michael Donaghy, Collected Poems (London: Pan Macmillan, 2014).

4 Philip Larkin, Collected Poems, ed. Anthony Thwaite (London: Faber & Faber, 2003).

5 Elizabeth Barrett Browning, Selected Poems, ed. Marjorie Stone and Beverly Taylor (Peterborough, ON: Broadview Press, 2009).