1997 IDontBelieveInWordSenses
Jump to navigation
Jump to search
- (Kilgarriff, 1997) ⇒ Adam Kilgarriff. (1997). “I Don't Believe in Word Senses.” In: Computers and the Humanities, 31(2).
Subject Headings: Word Sense, Dictionary, Word Sense Disambiguation Task.
Notes
Quotes
- Key words: ambiguity, disambiguation, lexicography, polysemy, word sense
Abstract
- Word sense disambiguation assumes word senses. Within the lexicography and linguistics literature, they are known to be very slippery entities. The paper looks at problems with existing accounts of ‘word sense’ and describes the various kinds of ways in which a word’s meaning can deviate from its core meaning. An analysis is presented in which word senses are abstractions from clusters of corpus citations, in accordance with current lexicographic practice. The corpus citations, not the word senses, are the basic objects in the ontology. The corpus citations will be clustered into senses according to the purposes of whoever or whatever does the clustering. In the absence of such purposes, word senses do not exist.
- Word sense disambiguation also needs a set of word senses to disambiguate between. In most recent work, the set has been taken from a general-purpose lexical resource, with the assumption that the lexical resource describes the word senses of English/French/: : :, between which NLP applications will need to disambiguate. The implication of the paper is, by contrast, that word senses exist only relative to a task.
1. Introduction
- Word meaning is of course a venerable philosophical topic, and questions of the relation between the signifier and the signified will never be far from the theme of the paper. However, philosophical discussions have not addressed the fact of lexicography and the theoretical issues raised by sense distinctions as marked in dictionaries. We often have strong intuitions about words having multiple meanings, and lexicography aims to capture them, systematically and consistently. The philosophy literature does not provide a taxonomy of the processes underpinning the intuition, nor does it analyse the relations between the word sense distinctions a dictionary makes and the primary data of naturally-occurring language. This is a gap that this paper aims to fill.
- I show, first, that Cottrell’s construal of word senses is at odds with theoretical work on the lexicon (section 2); then, that the various attempts to provide the concept ‘word sense’ with secure foundations over the last thirty years have all been unsuccessful (section 3). I then consider the lexicographers’ understanding of what they are doing when they make decisions about a word’s senses, and develop an alternative conception of the word sense, in which it corresponds to a cluster of citations for a word (section 4). Citations are clustered together where they exhibit similar patterning and meaning. The various possible relations between a word’s meaning potential and its dictionary senses are catalogued and illustrated with corpus evidence.
2. Thesis and Antithesis: PracticalWSD and Theoretical Lexicology
2.1 THESIS
- NLP has stumbled into word sense ambiguity.
- Within the overall shape of a natural language understanding system – morphological analysis, parsing, semantic and pragmatic interpretation – word sense ambiguity first features as an irritation. It does not appear as a matter of particular linguistic interest, and can be avoided altogether simply by treating all words as having just one meaning.
- NLP has not found it easy to give a very principled answer to the question, “what goes in the lexicon”.
- As machine-readable versions of dictionaries started to become available, so it became possible to write experimental WSD programs on the basis of the dictionary’s verdict as to what a word’s senses were (Lesk, 1986; Jensen andBinot, 1987; Slator, 1988;Veronis and Ide, 1990; Guthrie et al., 1990; Guthrie et al., 1991; Dolan, 1994). Looked at the other way round,WSD was one of the interesting things you might be able to do with these exciting new resources.
2.2 ANTITHESIS
- Since the publication of Metaphors We Live By (Lakoff and Johnson, 1980) and Women, Fire and Dangerous Things (Lakoff, 1987), there has been one approach to linguistics – cognitive linguistics – for which metaphor has been a central phenomenon. Metaphor is, amongst other things, a process whereby words spawn additionalmeanings, and cognitive linguists are correspondingly interested in polysemy. Lakoff’s analysis of the polysemy of mother is hugely cited.
7. Conclusion
- Following a description of the conflict between WSD and lexicological research, I examined the concept, ‘word sense’. It was not found to be sufficiently well-defined to be a workable basic unit of meaning.
- I then presented an account of word meaning in which ‘word sense’ or ‘lexical unit’ is not a basic unit. Rather, the basic units are occurrences of the word in context (operationalised as corpus citations). In the simplest case, corpus citations fall into one or more distinct clusters and each of these clusters, if large enough and distinct enough from other clusters, forms a distinct word sense. But many or most cases are not simple, and even for an apparently straightforward common noun with physical objects as denotation, handbag, there are a significant number of aberrant citations. The interactions between a word’s uses and its senses were explored in some detail. The analysis also charted the potential for lexical creativity. The implication for WSD is that word senses are only ever defined relative to a set of interests. The set of senses defined by a dictionary may or may not match the set that is relevant for an NLP application.
- The scientific study of language should not include word senses as objects in its ontology. Where ‘word senses’ have a role to play in a scientific vocabulary, they are to be construed as abstractions over clusters of word usages. The non-technical term for ontological commitment is ‘belief in’, as in “I (don’t) believe in ghosts/God/antimatter”.One leading lexicographer doesn’t believe in word senses. I don’t believe in word senses, either.
,