Page MenuHomePhabricator

Why are Senses not first-class entities?
Closed, InvalidPublic

Description

Since a Sense can be translated then the sense effectively applies to more than one Lexeme.

Likewise, a Sense can be tied to the same Sense on another Lexeme.

A Sense can (and ought?) represent a concept, however, some Senses cannot directly represent a concept (i.e. on verbs), because the sense is the concept.

Given these to things, why are Senses tied to Lexemes at all? Why not make the Sense a first-class entity that is referenced from a Lexeme? I imagine the referenced/created Sense would still be "embedded" in a Lexeme, but I don't see why it couldn't be reused on other Lexemes (especially when the Gloss can be translated as well)? or is there no circumstance (or very rare) where a Sense would be the exact same for multiple Lexemes?

Event Timeline

I think the design of Wikidata lexeme was inspired by the work around wordnets and Ontolex. In the wordnet world, a sense is the intermediate step between a lexeme and a concept/synset, see Ontolex specification where the sense is called LexicalSense https://www.w3.org/2016/05/ontolex/

In DanNet, "bil" is word-11004644 which has a sense wordsense-21007098 that is then tied to synset-1507 which is the concept of car. https://tools.wmflabs.org/ordia/L36385-S1 shows the synonyms for the sense of "bil", e.g., "øse" and "dytbil", each which can be tied to language style.

In the world of wordnet and Ontolex, the gloss, however, is on the synset/concept level rather than the sense level.

Thanks for your reply! I don't even understand my own question anymore, but I appreciate your response. :)