Fulk (2007) emphasizes the importance of probabilistic reasoning for analysis of Old English meter. Accepting all manuscript verses as authentic would burden a metrical theory with false evidence, blurring important distinctions and obstructing important discoveries. Doubtful verse patterns with low frequency should be excluded from consideration during initial attempts to formulate a theory. As an inherently plausible theory is refined, it may accept some anomalies as rare but acceptable departures from metrical norms. It is unimaginable, however, that a valid theory would accept all manuscript verses.

Here I would like to recommend a kind of probabilistic reasoning that goes beyond questions of what does or does not occur. As Halle and Keyser have observed, a poet's audience is “capable of distinguishing not only metrical from unmetrical lines but also more complex metrical lines from less complex lines” (1971: 142). Ideally, a theory of meter will posit gradations of complexity among acceptable lines and the most complex lines should have the most restricted frequencies (Hayes, Wilson, and Shisko 2012). In Old English meter, which defines acceptable stress patterns at the level of the verse (or half-line), theorists will also want to consider how the metrical complexity of a verse affects its placement within the line. Besides providing a more comprehensive account of the meter, a theory sufficiently robust to predict the distribution of verse types and their various linguistic realizations will depend less crucially on the precision of scribes.

One aid to assessment of metrical complexity is the universal principle of closure: adherence to metrical norms becomes stricter toward the end of a metrical unit (Hayes 1983: 373). Applied to Old English poetry, this principle predicts that the closing half of the alliterative line (the b-verse) should be less complex than the opening half (the a-verse). Complex verses should be identifiable not only because they have restricted frequency but also because they tend to occur in the opening half of the line. Even with this valuable aid, it is no simple task to isolate the influence of a metrical norm. There are several important norms and they apply within a wide variety of verse types.