- The Destruction of Troy's Different RulesThe Alliterative Revival and the Alliterative Tradition1
Almost simultaneously, Nikolay Yakovlev, Thomas Cable, and the University of Bristol's Ad Putter, Judith Jefferson, and Myra Stokes have come to similar conclusions about what kinds of syllables are and are not allowable in the final weak positions, or dips, in a-verses and b-verses in Middle English alliterative poetry.2 The Bristolians in particular have convincingly pursued the status of historical final -e and the requirement of a final, unstressed dip at the end of the b-verse.3 It is Yakovlev's "non-schwa principle," however, that captures the underlying phonology best and, more importantly, demonstrates how this principle applies to other weak positions in both half-lines.4 Through it we see why the long line must end on a single syllable with schwa, and the emerging principle, then, that nonschwa vowels cannot occur in certain weak dips or the final dip of the b-verse. The case is closed, so to speak, on an argument that began more than twenty years ago, when Cable and Hoyt Duggan independently came to similar conclusions about the rules of the b-verse. Cable argued that the b-verse must end in a single, unaccented syllable, whereas Duggan argued that the b-verse may end in such a syllable. Cable's theory that the poets conserved historical final -e has been confirmed, and through Yakovlev's work we now understand the phonological principle underlying not only this final dip, but also other dips in the Middle English alliterative long line. Thus is Wimsatt [End Page 162] and Beardsley's famous description of Middle English alliterative meter finally and firmly countered.5
But we have not yet managed to describe what constitutes a metrical line in Middle English alliterative poetry: the a-verse's meter remains elusive, though the discoveries made by each of these metrists offer important inroads into discerning its rhythm, and we can see that the poets of the West Midlands shared the London poets' phonology. If Yakovlev's theory of the vowel as a diagnostic test for the kinds of syllables allowed in the dips of the long line holds up on further investigation, it will simplify and clarify our understanding of the meter. The nonschwa principle is probably a more workable feature than Putter, Jefferson, and Stokes's "secondary stress," a term that remains vexed despite their best efforts to resolve it. Yakovlev does not explicitly consider secondary stress in his rules, though he recognizes this issue as both a linguistic and metrical phenomenon requiring more work.
Some basic terminological distinctions are germane in any discussion of meter, and the following definitions are foundational to the argument presented in this article. Foot meters such as the octosyllable, the decasyllable, and iambic pentameter create regular rhythms followed in every line, in that the naturally varying linguistic stress levels of English are reduced with few exceptions to a binary of alternating ictus and nonictus. The exceptions, such as the inverted first foot in a line of iambic pentameter, never disrupt the basic melody of the rhythm. These regular, unvarying meters are deductive, in that the audience quickly deduces the rhythm and then imposes it on each of the lines so that prepositions and auxiliary words—so-called functional items—may be promoted, and nouns and adjectives—so-called lexical items—may be demoted. The most deductive meter in English poetry is the nursery rhyme, where the singsong meter absolutely trumps word meaning, and any child can learn it. In sum, the language conforms to a metrical template in deductive verse.
The meters of Old and Middle English alliterative verse, on the other hand, are inductive; the metrical rhythm depends upon linguistic stress and emerges from it. Each line conforms to the same rules, but there is no single pattern generated.6 Thus, each line is a rhythmical surprise in that [End Page 163] the audience never knows what the rhythm will be. In other words, the meter conforms to linguistic stress; more importantly, there is not a fixed number of metrical stresses per line as there necessarily is...