Many instances of plural number marking in Old English disyllabic neuter a-stem nouns appear uncertain. This is due partly to a want of additional empirical evidence regarding what appears to have been a tension between a high vowel deletion process, by which some disyllabic neuters containing a long root vowel failed to attach the nominative/accusative plural number marker -u, and several analogical extension processes which resulted in irregular attachments of the plural markers -u, -ø, and others. This apparent unpredictability, however, is also due to a lack of agreement about how best to subclassify many disyllabic a-neuters. Various scholars have addressed the problem of the allomorphy at issue here, but their grouping criteria have differed and no one scheme has proven truly satisfactory (cf. Brunner - Sievers 1965; Dahl 1938; Campbell 1959; Wełna 1996). Consequently, determining which disyllabic a-neuters attached the u-plural allomorph regularly and which attached it by analogy as well as which of these neuters suffixed the ø-allomorph regularly and which did so analogically is trouble-some. In an attempt to augment our understanding of allomorphy in the plural morpheme of the Old English disyllabic neuter a-stems, this paper analyzes more than 300 plurals culled from both Early and Late Old English texts, and it proposes, unlike previous treatments, that token frequency was crucial to the analogical processes which so often determined plural marker selection in these nouns.
