Setting about a guide to poetics throws up some simple, central and yet terrifyingly big questions.  When, for example, does enjambment (re-)start in English verse?  Enjambed lines of verse are frequent in Old English, but much less common in pre-1350 poetry.  Donka Minkova’s excellent introduction to ‘The Forms of Verse’ in Peter Brown’s A Companion to Medieval English Literature and Culture c. 1350-c.1500 (Wiley-Blackwell, 2007) asserts that before Chaucer ‘Middle English verse was end-stopped, meaning that each line-ending coincided with a major syntactic break – the end of a clause or phrase’ (p. 187).  Chaucer is claimed as the re-originator of a ‘new and unexpected’  (p. 188) verse innovation, namely enjambed or run-on lines ‘in which a syntactic […] unit straddles two lines’.  She gives as examples House of Fame 349–50 and 582–83.

Like many generalisations, this is broadly true but immediately set me off looking for exceptions.  Minkova is right that enjambment is rare in pre-Chaucerian verse.  Often Early Middle English poets fit one unit of sense into one line-unit.  Below is an extract from The Owl and the Nightingale (lines 13ff) in which the sentence runs across six lines.  Each line is a self-contained unit of sense, and one could remove any single line after the first line containing the subject and main verb without rendering any of the rest of the remaining lines incomplete.  At the end of each line, we feel that each clause or grammatical unit is complete and that the sentence could stop there.

If you compare the next example from the late-thirteenth-century romance Havelok (see lines 2736ff below), you can see a longer sentence running on over nine lines.  And what is different here, partly because this is action rather than description, is that there are several lines which, as we read them, are grammatically incomplete.  We can feel that the third line, ‘he smote him on the shoulder so’ is not complete, we expect a ‘that’, a conclusion or result of this blow.  We know that ‘He broke open more of his chain-mail rings’ requires a ‘than’, a comparison to fulfil our expectations of this ‘more than something’ structure.

Here you can see that the Havelok author does not think in units of an individual line or pair of lines in a couplet.  He is thinking beyond the couplet, stretching his ideas across longer enjambed units.  This is a more sophisticated, aurally and cognitively interesting verse, where sense units are in tension with units of form, rather than always being in neat parallel.  As you can see from the last line, the poet does not end this sentence on the second line of the couplet.  This sentence ends on the first of the pair of rhyming lines.  So this sentence, and this unit of the narrative ends, but the open-ended couplet tells us that something else will come, that Havelok’s retaliation is imminent, that the fight is not over.  The couplet ends with a beginning: ‘Tho was Havelok swithe wo […]’, and thus we have already moved into the next moment in the narrative.

The author of Havelok, of all the pre-Chaucer poets I have read so far, makes the most interesting use of enjambment.  Below is another example.  This is the moment at which Havelok brings his wife Goldeboru before sixty English earls to see if they will recognise her as queen of English.  Here is enjambment in its strictest sense, units of sense which carry on beyond the end of the couplet.

In the second sentence, the poet is happy to begin the main clause on the last word of the line, splitting the noun phrase (alle the English men) across the line break, and in the next line splitting the prepositional phrase from the verb across the line break.  These splittings create a tension between the completion of the couplet and the completion of the syntactical unit, a kind of enjoyable lopsidedness.  It does not seem to have caught on widely and become the imitable feature that Chaucer’s enjambment was.  But it is there, a hundred years before the House of Fame.