Tiedekunta/Osasto - Fakultet/Sektion
Laitos - Institution
Yleisen kielitieteen laitos/Englannin kielen laitos
Tekijä - Författare
Anttila, Arto
Työn nimi - Arbetets titel
Conversion in English: How to treat categorical homonymy in the lexicon
Oppiaine - Läroämne
Yleinen kielitiede/englantilainen filologia
Työn laji - Arbetets art
Aika - Datum
huhtikuu 1990
Sivumäärä - Sidoantal
88 + 14
Tiivistelmä - Referat

Systematic part-of-speech ambiguity and morphological syncretism are phenomena characteristic of English. A word like "love" is categorically ambiguous in several ways: N, V; V_present, V_imperative, V_infinitive, V_subjunctive and possibly more. Besides causing practical disambiguation problems in automatic language processing (DeRose 1988, Karlsson 1990), such ambiguity often seems theoretically undesirable and an artefact of description rather than a real phenomenon of English.

This work explores the possibility of eliminating categorical ambiguity from descriptions. It is observed that a systematic application of classical structuralist discovery procedures (Harris 1946, 1951) to English would suggest a representation where the lexicon contains a great number of partially specified "content words" instead of fully specified nouns and verbs, and a few heavily specified function words which fill in the missing syntactic information in syntactic constructions. The proposal is embedded in a feature-based framework that borrows techniques from GPSG and Categorical Grammar.

Avainsanat - Nyckelord
Säilytyspaikka - Förvaringställe
Muita tietoja