Adverbs Revisited: Enhancing WordNet Coverage of Adverbs with a Supersense Taxonomy

Abstract--WordNet offers rich supersense hierarchies for nouns and verbs, yet adverbs remain underdeveloped, lacking a systematic semantic classification. We introduce a linguistically grounded supersense typology for adverbs, empirically validated through annotation, that captures major semantic domains including manner, temporal, frequency, degree, domain, speaker-oriented, and subject-oriented functions. Results from a pilot annotation study demonstrate that these categories provide broad coverage of adverbs in natural text and can be reliably assigned by human annotators. Incorporating this typology extends WordNet's coverage, aligns it more closely with linguistic theory, and facilitates downstream NLP applications such as word sense disambiguation, event extraction, sentiment analysis, and discourse modeling. We present the proposed supersense categories, annotation outcomes, and directions for future work. As a primary lexical class, adverbs perform a range of semantic functions, from answering fundamental questions about an event, such as how it was performed (manner), when it occurred (temporal), or to what extent a property holds (degree), to expressing speaker attitude, discourse stance, and logical relations between propositions. Despite this semantic richness, adverbs have long occupied an ambiguous and often marginalized position in linguistic classification, frequently described as a "residual" or "wastebasket" category [9, 20]. Words are often assigned to this category not because they share definable grammatical properties, but because they fail to conform to the morphological and syntactic criteria of nouns, verbs, adjectives, prepositions, or conjunctions.