The alliterative line

So last time I did this I talked a bit about the five-beat line (5BL), the workhorse for Kin-Bright. The poem has other metres, some of them very different.

Today, I'll talk about the alliterative verse. Readers might be less familiar with this, so I'll try to keep things clear! You don't need to know any of this, at all, to enjoy the poem, I should say. I'm just slamming it out here in case people're interested. Unless / until someone yells at me, I'll continue filing these as game design, on the basis that these are the technical systems within which I'm working.

1. History

I model the alliterative verse of Taru high speech on the later Middle English (ME) alliterative verse surviving from the second half of the fourteenth century, and from the start of the fifteenth.

Later ME alliterative verse comes from Old English alliterative verse, the stuff of the poem we now anachronistically call Beowulf. We can see this descent in, for instance, some of the specialised poetic vocabulary that appears in both, but not elsewhere in ME.

However, the later alliterative verse doesn't work in quite the same way. And although Old English alliterative verse has more modern fame, later ME alliterative verse survives in significantly greater amounts: the longest example, John Clerk's Destruction of Troy, runs to 14,000 lines, or four-and-a-bit Beowulfs.

I don't follow all of the tenets of later ME alliterative poetry—in part because we're still slowly reconstructing them!—but the above might offer useful context for what follows.

2. Metre

A basic and crude model works as follows. The line falls into two halves, separated by a regular metrical caesura, that is, a rule-bound, required pause:

for the sack of our city | sealed us fearless

We can call these two halves the a-verse and b-verse, if we want (and I do want).

You can divide vocabulary into open-class words and closed-class words. Closed classes of words accept new members only slowly, often over centuries; open classes of words accept new members quickly. Closed classes are groups like prepositions and conjunctions. Open classes are groups like nouns and verbs. This distinction is quite like 'content words' versus 'function words', and you can think of it like that if that helps!

Now for some components. It is a crude but workable starting model to say that:

• The most naturally-stressed syllables of open-class words form lifts (/); these are the sites of metrical prominence.
• A single naturally-unstressed syllable is a short dip (x).
• Two or more naturally-unstressed syllables—the number doesn’t matter, the plurality does, the key contrast here is one against many—form a long dip (x…x).

Remember that lifts are a matter for syllables, not whole words: the ter- of terror forms a lift, but the -ror forms either a short dip or, when followed by other naturally-unstressed syllables (such as the), a long dip.

A-verses

The first half of the line, the a-verse, typically has—whatever else it also has!—two lifts and two long dips:

O protector of Tar

(x…x/x…x/)

turning to terrify

(/x…x/x…x)

Some a-verses have three lifts:

or taste now with terror the tip

(x/x…x/x…x/)

A-verses typically also avoid b-verse patterns. Which brings us to the b-verse. The b-verse has tighter rules, in keeping with the general tendency of historical forms across Indo-European languages to tighten metrically towards the line's ending, the principle of closure.

B-verses

The second half of the line, the b-verse, has exactly two lifts, and exactly one long dip. The one long dip must come before the first lift or between the lifts—never after the second lift. The b-verse always ends with a lift and a short dip (/x).

There are therefore exactly four legitimate b-verse patterns:

1. x…x/x/x
2. x…x//x
3. x/x…x/x
4. /x…x/x

of a Taru weapon!

(x…x/x/x)

so to gaud weddings

(x…x//x)

have failed us and broken

(x/x…x/x)

shaking our foundations

(/x…x/x)

This is the basic model. Some open-class words, most commonly verbs, can be demoted to form dips, or parts of dips.

Syntactically, later ME alliterative verse tends towards complete lines, unlike Old English alliterative verse, which included many more free-standing half-lines and more happily placed major syntactic breaks at the metrical caesura.

As you can hear, alliterative metre differs from the 5BL not only in fairly superficial matters such as the beat count—four against five—but also in its fundamental workings: we don't expect alternation, sites of prominence can abut as they typically don't in verse from the 5BL tradition, and at base alliterative verse makes patterns out of natural lexical stresses rather than establishing an on–off template that not infrequently promotes naturally-unstressed words to beat status. Alliterative verse differs from the 5BL but also from the whole family of alternating metres within which the 5BL sits, the family onto which the idiotic, catastrophic, very-ungood sixteenth century eventually imposed a bunch of inappropriate Grecian and Latinate vocabulary to annoy and confuse everyone.

But what about the actual alliteration?

3. Alliteration

Later ME alliterative verse demanded more rigorous alliteration than Old English alliterative verse.

The first three lifts in the line alliterate with each other, and the fourth lift doesn’t alliterate with the others (AA|AX); occasional variations (e.g. AB|AB) are possible but must stay occasional:

now you must shepherd them forth, | and shield them for the future

(A line in which I deliberately stretch the long dips, because it's the end of a speech.)

Lift syllables alliterate, and not all lift syllables are word-initial, so decisive, for instance, alliterates on its naturally-stressed middle syllable, and would alliterate with sickle. All lift vowels with no preceding consonant onset in their syllable alliterate with each other, regardless of value: oaf alliterates with equal.

For listeners, the arrival of the non-alliterating fourth lift, and its fully-predictable lift-and-short-dip (/x) pattern, signals the line's end.

4. Modernity

Plenty of other modern poets have revived alliterative verse, in various ways. Not so very many have had a crack at later ME alliterative verse, in part because most examples have fewer readers than the big Old English hits, and in part because we only really started to nail down what's even going on in later ME alliterative metre in the last four decades (and the hunt continues: scholars disagree a bunch over the a-verse).

Someone could, in theory, write thousands of lines of modern English poetry in alliterative verse; probably, out there somewhere someone has. But it's a real challenge: I mentioned specialised poetic vocabulary earlier, and since we now lack that breadth of synonyms, the alliteration presents a harder task—especially if you're doing full three-lift alliteration rather than the less demanding Old English model.

I'm lazy, so I'm not attempting that. But I find alliterative verse useful to indicate the elaborated formal speech used by the Taru brass—so up themselves, the poor godawful things—and I like writing it, and there's precedent for shuttling between alternating and alliterative metres in the poem we now call Sir Gawain and the Green Knight.