Advances in the Automatic Lemmatization of Old English: Class IV Strong Verbs (L-Y)

The morphological features of an inflectional language like Old English (OE), which also presents generalized spelling inconsistencies, limit the use of lemmatizing and tagging tools that can be applied to natural languages. Consequently, the development of Natural Language Processing (NLP) models, which crucially depend on lemmatized corpora, is slowed down. Against this background, this article develops a lemmatizer within the framework of Morphological Generation that allows for the type-based automatic lemmatization of OE class IV strong verbs (L–Y). The lemmatizer incorporates a set of algorithms to account for features of inflectional, derivational, morphophonological and diatopic variation. The generated forms are automatically compared with Taylor et al. (2003) and Healey et al. (2004) to confirm their attestation and are assigned a lemma. Overall, the research proves successful in setting up form-lemma associations, while highlighting areas of ambiguity and mismatches. The main conclusion of the article is that taking the route of automatic lemmatization with this methodological framework will contribute to the field of OE lexicography by both lemmatizing attested inflectional forms and by identifying areas for manual revision.