Old English Feet

In this chapter I show that the Beowulf poet carefully avoids lines whose halves are identical in terms of stressed (x) and stressless (.) syllables. The half-line (x.x.), for instance, is happily paired with anything but another (x.x.), despite the fact that (x.x.) is by far the commonest half-line type in the poem. Statistical analysis shows that this avoidance is not by chance and suggests that having identical adjacent half-lines is unmetrical in the poem. The same appears to be the case for half-lines identical in weight: a half-line with heavy and light syllables that runs (HLHL) does not pair with one that runs (HLHL). The results are obtained by looking at each and every syllable in the poem, regardless of stress, morphological status, or position within the line. This shows that Beowulf meter strictly regulates every syllable in the poem. Theories that ignore pre-tonic syllables (Bliss 1958), syllables that occur in prefixes (Russom 1987), stressless syllables in general (Keyser 1969, Fabb & Halle 2008), or stress in general (Golston & Riad 2001) cannot account for this fact. An adequate theory of OE meter must include stress, quantity, and the avoidance of identical half-lines. 2. Variation We may begin with a metrical scansion of the first part of the poem, where the first syllable of every lexical root (noun, verb, adjective) is stressed (x) and all others are stressless (.): (1) Beowulf lines 1-11 Hwat, wē gār-dena in geār-dagum, . . x x . . x x . 1 þēod-cyninga þrym gefrūnon, x x . . x . x . 2 hū ðā æþelingas ellen fremedon! . . x . . . x . x . . 3 oft Scyld Scēfing sceaþena þrēatum, . x x . x . . x . 4 monegum mǣgþum meodo-setla oftēah, x . . x . x . x . . x 5 egsode eorlas, syððan ǣrest wearð x . . x . . . x . x 6 fēasceaft funden; hē þæs frōfre gebād, x . x . . . x . . x 7 wēox under wolcnum weorð-myndum þāh, x . . x . x x . x 8 oð þæt him ǣghwylc ymb-sittendra . . . x . x x . . 9 ofer hron-rāde hȳran scolde, . . x x . x . x . 10 gomban gyldan; þæt wæs gōd cyning! x . x . . . x x . 11 Listen up! We have heard of the power of the kings of the spear-Danes, how the nobles won their honors! Oft Scyld Scefing took the mead-benches from many enemy nations’ troops, terrified warriors. After he was first found destitute, he found this consolation, he grew up under clouds, he prospered in honor, until each of the surrounding nations on the sea-lanes had to obey him paying tribute; that was a good king! Two things should be mentioned before we go on. First, I’ll have nothing to say about the important issue of alliteration, only because it falls outside the scope of this article. Second, only primary stress is recorded in the scansions—secondary is sometimes taken to be metrically significant in Old English meter and I do not mean to deny that it could be; but to test scansion empirically one needs to have a simple and replicable method of scanning and I have chosen to scan only primary stresses as ekses. Looking at the patterns of stressed and stressless syllables, it should be clear that there is no meter in the usual sense of a simple rhythmic pattern repeated a number of times. We do not find four trochees here or five iambs or six dactyls or the like. Rather, there seems to be a fairly haphazard distribution of stressed and stressless syllables. But a closer look at the data reveals an anti-pattern: the Beowulf poet strenuously avoids lines whose first and second halves are identical in terms of stress. The point of the meter is to have rhythmic variety between the two halves in every line, more Coltrane than Chopin. Note that in the eleven lines above, none has a line