Forgotten Ground Regained
Varieties of Alliterative Meter
I recently read Geoffrey Russom’s book, The Evolution of Verse Structure in Old and Middle English Poetry: From the Earliest Alliterative Poems to Iambic Pentameter. It is a technical work, and you need a decent background in linguistics and medieval literature to follow its arguments easily. But it is an important work and has significant implications for poets who want to write alliterative verse. Russom has also written an article, “Poetic Form”, which presents the outlines of his theory in a more accessible form. While his isn’t the only theory of how poetic meter works (that is, after all, a whole field of study in its own right), Russom’s theory offers the most convincing explanation I have ever read about why alliterative verse died out in the fifteenth century. And that, in turn, provides insights into what is going on in the work of modern poets who have chosen to write in alliterative forms. So I am going to take the time to sketch how Russom’s theory makes sense of the history of alliterative meter and use it to explain some of the variations in alliterative structure that we see among modern-day poets.
Classical Head-Stave Meter
Before I dive into the details it will be useful to review the distinctive characteristics of Germanic alliterative verse, the kind that shows up in Old English poems like Beowulf or the Old Norse Poetic Edda. The only native accounts of alliterative meter we have — accounts written by practicing alliterative poets — are provided in the first instance by Snorri Sturluson’s Hattatal (Gade, 2017), and developed further by poets and scholars in Iceland, the only country where Germanic alliterative verse has survived as a continuous literary tradition. (Adalsteinsson, 2014, provides a fairly exhaustive list of references.)
The native Norse/Icelandic tradition describes alliterative verse as being organized around ‘staves’ (stressed, alliterating syllables) that create what modern scholars call the alliterative ‘long line’. Each long line consists of two short lines, or ‘half-lines’. Each half-line contains two (occasionally, three) strongly stressed syllables, or lifts, and sequences of unstressed syllables, or dips. The first half-line in each pair (termed the a-verse by modern scholars) can have either one or two alliterating stresses. The first stress in the second half-line (termed the b-verse by modern scholars) must alliterate. The final stress in the b-verse must never alliterate. Thaliarchus’ poem ‘Farewell’, printed earlier in this issue, provides a straightforward illustration. Here are three lines from this poem:
I have underlined the strongly stressed syllables and bolded the alliterations required by the meter. In the native alliterative tradition, the alliterating stave in the b-verse is called the head-stave because it is the key to the alliterative structure of the line. Everything leads up to the head stave. The alliterating stress(es) in the a-verse are considered ‘props’ that function to support the head-stave.
There is a lot more to be said about traditional alliterative meter, but for present purposes, I will highlight two: (i) In the older Germanic languages, primary stress almost always falls on the first syllable of a word. That is, of course, where the alliteration goes. (ii) While it has plenty of rhythmic flexibility, older Germanic poetry tends toward a falling rhythm (strong-weak) rather than a rising rhythm (weak-strong). For example, in Old English alliterative verse, the most common half-line rhythm is what German scholar Eduard Sievers (Sievers, 1885) termed ‘Type A’ (strong-weak-strong-weak). The Norse/Icelandic alliterative tradition records a similar intuition. In traditional Icelandic metrical analysis, each line is divided into feet. Odd-numbered feet are considered ‘heavy’ and even-numbered feet are considered ‘light’, and there is a strong preference to anchor alliteration on heavy rather than light feet (Jónsson, 1892, as explained by Ringler, 1996). Russom argues that the linguistic characteristics of old Germanic languages make word-initial alliteration prominent and favor a falling rhythm. Which means that all an alliterative poet really needed to know was to use alliteration to join the most prominent words in the two halves of the line. The metrical patterns that resulted fell naturally out of the normal rhythms of the language.
Russom's Theory of Alliterative Meter
Russom starts with the assumption that poetic patterns are based on linguistic patterns. Certain things are naturally prominent – sentence, phrase, word, and syllable edges, stressed syllables, and repeated sounds. Russom argues that poetic units correspond naturally to linguistic units that organize these patterns of prominence. Poetic beats prototypically correspond to syllables, poetic feet, to words, half-lines, to phrases, and lines, to clauses or sentences. The idea is that the most natural poetic patterns align with the most common, natural patterns in the language.
In Russom’s account, the prototypical Old English half-line was a phrase prototypically composed of two (inflected) words. Because English words were almost always stressed on the first syllable, and the end of the word was an unstressed inflectional ending, the default rhythm of the half line was therefore a falling rhythm (Type A). This falling pattern was reinforced by the fact that Old English sentences typically ended with a verb, which received weaker emphasis than subject or object nouns. So the end of the Old English or Old Norse line was the place that naturally got the weakest stress. And alliteration only works rhythmically if you alliterate on the most strongly stressed syllables.
In short, Russom’s theory claims that the typical Old English alliteration pattern (aa / ax) followed from typical Old English rhythms. We can describe the default poetic rhythm in the following diagram:
- Poetic Line
- / \
- / \
- A-verse B-verse
- (strong) (weak)
- / \ / \
- / \ / \
- Foot Foot Foot Foot
- (strong) (weak) (strong) (weak)
- / \ / \ / \ / \
- Lift Dip Lift Dip Lift Dip Lift Dip
- (s) (w) (s) (w) (s) (w) (s) (w)
In this kind of theory, deviations from the basic pattern are possible so long as they are not too great. Poets found it easy enough to use words with different rhythms in many of the feet. But alliterating on the final stress would suggest a rising rhythm for the entire line.
As I understand Russom’s theory, that is why the head-stave was so important. Alliterating on the head-stave, and not on the final stress, maintained a sense of falling rhythm from one line to the next.
Of course, Modern English is very different from Old English. A rather large chunk of the Modern English vocabulary consists of words borrowed from French, Latin, and Greek, where the stress falls toward the end of the word. English has lost most of its inflectional endings (which came at the end of the word) in favor of using grammatical function words – mostly articles and prepositions, which are usually placed before the content word they modify. And English clauses have shifted almost entirely to Subject-Verb-Object word order, which places the natural emphasis at sentence end. All of these factors militate toward rising rather falling rhythm (think: iambs and anapests instead of dactyls and trochees.)
This is why Russom thinks alliterative verse died during the fifteenth century. Head-stave meter only makes sense if the underlying rhythm is a falling one. But the modern English poet has to work hard to keep a falling rhythm going. Thaliarchus’ poem, earlier in this volume, illustrates some of the techniques that poets can use to do so.
To begin with, Thaliarchus’ poem contains a lot of compound nouns, many of them original — catch pair, bus stop, bold-clutch, clock-hand, love-pair, etc. Compound nouns have a falling rhythm, with the added bonus that Old English poets used them to create kennings, giving the poem a distinctively Old English feel.
Thaliarchus’ poem also takes advantage of (often archaic) patterns of grammatical inversion, which again create falling stress patterns:
- "in bold clutch to abide", "now bites clock-hand", "now short grows queue", "woeful the love pair"
- "winters not many", "together can muster", "for them gapes week-span", "break-ups stormy"
- "sorrow not little", "did we bliss grapple"
Not coincidentally, many of these inversions are necessary to get an aa/ax alliteration pattern. The poem also suppresses function words (especially articles) where it would be more natural to include them (e.g., one would normally say “a” or “the” clock-hand), also making it easier to keep a falling rhythm.
Notice also that in this poem, no sentence ever ends at the end of the line. Ending sentences between half-lines is normal enough in Old English, but in modern English it has the added benefit of putting the naturally strongest sentence-final stress in the a-verse. The overall effect of these stylistic choices is purposely archaic. It produces the rhythm and feel of Old English poetry in a modern English setting. But as such, it is not likely to be adopted by poets who favor a more contemporary style.
In his book, Russom argues that in Middle English (which had transitioned partially to modern English patterns), these kinds of archaisms played an important role in maintaining the traditional head-stave alliteration pattern. But the literature on Middle English alliterative verse also suggests that there were rather strict constraints on the b-verse that (I believe) had the effect of keeping stronger emphasis on the head-stave and keeping a sense of falling rhythm at the end of the line. There could be at most one “long” (multiple-syllable) dip in the Middle English b-verse, and it had to appear either before or after the head-stave(Duggan, 1986; Inoui & Stokes, 2012). And the final foot nearly always had a falling rhythm — typically, a final -e that was no longer pronounced by the end of the fifteenth century (Duggan, 1988).
I have found in my own writing that it is much easier to make an a(a)/ax alliteration pattern sound natural if an iambic pattern is avoided and the head stave is reinforced by a long (or secondarily stressed) dip. One can also maintain the rhythm that head-stave meter requires by carefully controlling sentence length, structure, and punctuation. The excerpt from Rahul Gupta’s Arthuriad which comes at the end of this issue is a case in point – the way Gupta organizes his sentences to create an incantatory effect also makes the endings of lines much less prominent to the ear. However, Russom’s theory has further implications for modern English alliterative verse.
If Russom’s theory is correct, modern English has a natural predilection toward rising rhythms. Since alliterative verse places alliteration on the strongest stresses, and the final stress of the line is, by default, the naturally strongest stress in modern English, modern English rhythms should therefore predispose poets to adopt an alliterative meter in which alliteration falls obligatorily on the final stress.
Let us take the diagram I provided earlier, but invert the rhythm, so that weak precedes strong at every level of analysis. If we do that, the last stress in the b-verse – let us call it the tail-stave – should be the most likely to alliterate. I contend that we see exactly this tendency in alliterative verse written by modern poets.
Poetic Line / \ / \ A-verse B-verse (weak) (strong) / \ / \ / \ / \ Foot Foot Foot Foot (weak) (strong) (weak) (strong) / \ / \ / \ / \ Dip Lift Dip Lift Dip Lift Dip Lift (w) (s) (w) (s) (w) (s) (w) (s)
Alex Rettie’s poem “The Future” (published in this issue) is a clear example of pure tail-stave meter. Here is how I analyze it:
In this poem, the tail stave always alliterates, and there is always at least one matching prop in the a verse. Perhaps in recognition of the tail-staves’ importance, Rettie has placed them in short lines of their own. The rhythm and syntax that result are entirely consistent with a colloquial, modern style.
My poem "Housebreaker" was my first experiment in writing in tail-stave meter:
- Housebreaker
- I woke without light – I sensed, not alone;
- half-rose, reaching out, pulse rushing
- Through arteries and veins, but the room was empty.
- At the base of the stairs a creaking board
- halted me, but I heard nothing. Trust a housebreaker
- to tread softly, stand quiet when the household stirs!
- In the kitchen, a clatter: In flashlight-beam, a kitten,
- Reminding me that some murderers focus on mice.
I am less sure about the lines where I alliterated the 1st stress with the 4th. Allowing two lifts between alliterating staves seems like a stretch, though Maryann Corbett does it, too, in lines like “memory of a love ‧ that crumpled to malice.” It might be better to require no more than one non-alliterating lift between alliterating staves, much as in traditional Icelandic metrics.
In any case, it is also noteworthy how often modern English poets alliterate on the final stress even when they are working to create an Old English mood or responding to Old English material. Maryann Corbett’s poem in this issue is a case in point. By my count, 24 of 44 lines alliterate on the tail-stave. And there are many, many such examples among the modern English alliterative poems I have collected on alliteration.net.
We can sum up what I have found in this study as follows: While it is possible to write beautiful, effective alliterative verse in head-stave meter, the rhythms of the language may encourage modern English poets to adopt a different alliterative pattern, a tail-stave meter predicated on rising rather than falling rhythms.
References
- Adalsteinsson, Ragnar Ingi. Traditions and Continuities: Alliteration in old and modern Icelandic verse. Rejkjavik: University of Iceland Press, 2014.
- Duggan, H. N. The shape of the b-verse in Middle English alliterative poetry. Speculum, 61(3), 1986, 564-592.
- Duggan, H. N. Final"-e" and the Rhythmic Structure of the B-Verse in Middle English Alliterative Poetry. Modern Philology, 86(2), 1988, 119-145.
- Gade, Kari Ellen. ‘ Snorri Sturluson, Háttatal’ in Kari Ellen Gade and Edith Marold (eds), Poetry from Treatises on Poetics. Skaldic Poetry of the Scandinavian Middle Ages 3. Turnhout: Brepols, 2017, p. 1094. <https://skaldic.org/m.php?p=text&i=1376> (accessed 31 July 2024)
- Inoue, Noriko, and Myra Stokes. "Restrictions on dip length in the alliterative line: The a-verse and the b-verse." The Yearbook of Langland Studies 26, 2012, pp 231-260.
- Jónsson, Finnur. Stutt íslenzk bragfræði. Kaupmannahöfn: Í prentsmiðju S. L. Möllers [Möller & Thomsen], 1892.
- Ringler, Dick. Jónas Hallgrímsson: Selected Poetry and Prose. Retrieved from https://digicoll.library.wisc.edu/Jonas/Jonas.html
- Russom, Geoffrey. The Evolution of Verse Structure in Old and Middle English Poetry: From the Earliest Alliterative Poems to Iambic Pentameter. Cambridge: Cambridge University Press, 2017. https://www.cambridge.org/core/books/evolution-of-verse-structure-in-old-and-middle-english-poetry/D7AD46EBF974D3B02DB02BA8F415CB8C
- Russom, Geoffrey. "Poetic Form", 2017. Downloaded from https://literaryuniversals.uconn.edu/2017/10/07/poetic-form-2/
- Sievers, E. Zur rhythmik des germanischen alliterationverses. I. Beiträge zur Geschichte der deutschen Sprache und Literatur, 10, 1885, pp. 209–314.
First Published in Forgotten Ground Regained, New Series, Issue 4, Fall, 2024
Copyright © Paul D. Deane, 2024 No part of this site may be used or reproduced in any manner for the purpose of training artificial intelligence technologies or systems