[Corpora-List] WordNet vs Ontology

John F Sowa sowa at bestweb.net
Wed Oct 1 18:32:42 CEST 2014

On 9/29/2014 8:51 PM, Jim Fidelholtz wrote:
> The gist is that, in my probably undereducated view, nearly all polysemy
> (in the usual sense of 'all the definitions under the same headword')
> originates in metaphor.

If you said 'metaphor and metonymy', that would cover the majority of cases. There are other terms for related phenomena. But the fact that those words come from ancient Greek is a reminder that many very smart people have been analyzing the phenomena for a very long time.

> I believe once we have some sort of handle on metaphor and its role in
> the semantics/pragmatics of language use, many of the Kilgarriff et al.
> complaints about 'word senses' will be found to be at least partially
> resolvable. Then again, I'm an optimist.

I partly agree. But I'd like to add two newer terms that may help focus the analysis.

The first is 'microsense' by Alan Cruse: It's compatible with the Atkins-Kilgarriff complaint, and it's related to Wittgenstein's point that the meaning of a word shifts with the language game. Instead of saying that word senses don't exist, Cruse would say the number of microsenses of a word type is open-ended and they may shift in subtle ways from one use (token) to another.

The second is 'meaning potential' by Patrick Hanks: He cites Cruse, Kilgarriff, Atkins, Wittgenstein, and many others. He doesn't use the term 'microsense', but taken together, the two terms provide a theory (Hanks calls it TNE -- Theory of Norms and Exploitations) where the meaning potential for each word type, guided by its norms, determines a cloud of microsenses for the possible exploitations.

But clouds are rarely static. Potentials, norms, and exploitations evolve with changes in culture, knowledge, and fads. As Peirce said "symbols grow": for any word token, the listener's microsense may be more developed, less developed, or just different from the speaker's.

For many documents, the author's intended microsense might not be the ideal "gold standard". For example, consider a newspaper report about some new invention, scientific discovery, or political event.

References below.

John ______________________________________________________________________

Cruse, D. Alan (2000) Aspects of the micro-structure of word meanings, in Ravin & Leacock (2000) pp. 30-51.

Cruse, D. Alan (2002) Microsenses, default specificity and the semantics-pragmatics boundary, Axiomathes 1, 1-20.

Hanks, Patrick (2013) Lexical Analysis: Norms and Exploitations, Cambridge, MA: MIT Press.

Ravin, Yael, & Claudia Leacock, eds. (2000) Polysemy: Theoretical and Computational Approaches, Oxford University Press, Oxford.

For CSP's observation, google Peirce and "symbols grow".

More information about the Corpora mailing list