RE: ontolex:usage

Dear Christian,
Thanks for bringing this up.
I' m also in favour of option 1. To reinforce the argument, there are also cases of syntactic behaviour of words used only in a dialect, and would benefit from being described with some lexinfo:usage subproperty.
I have also noticed that there 's an ontolex:usage property (https://www.w3.org/2016/05/ontolex/#usage) with the same definition as lexinfo:usage. What's the relation between the two?
Thanks in advance,
Penny

From: Christian Chiarcos <christian.chiarcos@gmail.com>
Sent: Wednesday, May 4, 2022 4:27 PM
To: public-ontolex <public-ontolex@w3.org>
Subject: ontolex:usage

Dear list members,

I would like to elicit some feedback on the suggestion  to use ontolex:usage for elements other than ontolex:LexicalSense. This would be a backward-compatible revision of the core model (dropping the domain restriction of ontolex:usage, no rewording or diagram revision needed).

In recent OntoLex-Morph discussions (https://docs.google.com/document/d/1qdYg1jvf-YR8bmfKHE3kyRXAJvacTu1zE2P1PM4AZgM/edit?usp=sharing), we discussed possible uses of ontolex:usage, resp., its subproperties defined in LexInfo. The context of this discussion is that we characterize inflected forms (!), morphemes (LexicalEntries!) and (optionally) morphological rules (no OntoLex core counterpart) in terms of their usage features (archaic, dialectal, formal register, etc.).

In principle, this is possible with ontolex:usage. However, ontolex:usage is defined as a property of ontolex:LexicalSense (thanks to Penny Labropoulou for pointing that out), but some of the subproperties (defined in LexInfo) are either not restricted to senses (but can be equally applied to, say, forms) or are even explicitly addressing lexical units other than senses.

In terms of the information ontolex:usage already provides, there is very little in these properties that wouldn't be useful to have for lexical forms or lexical entries, as well. And even looking on some definitions for these sub-properties (e.g., lexinfo:domain: "usage marker which identifies the specialized field of knowledge in which a lexical unit is mainly used"), this seems to have been intended (otherwise, why "lexical unit" and not "lexical sense"?).

lexinfo:socioCultural is one point in case: "usage marker which identifies the use of a given lexical unit by particular social groups and/or in certain types of communicative situations depending on their level of formality". This does not involve semantics, but pragmatics, so the difference between formal and informal speech shouldn't even exist for senses (~ context-independent meaning). For some instances, even the name indicates that it's not for a lexical sense, e.g. "archaicForm".

Basically, there are three options:
1. dropping the domain restriction of ontolex:usage,
2. deprecating ontolex:usage in favor of a novel (and more flexibly defined) lexinfo:usage, or
3. duplicating ontolex:usage and its subproperties under a novel property with the same function (but different domain).

My preference would be option 1, I could live with 2 (although I'm not sure it has any benefit over 1), but I have strong concerns against option 3. In the end, this would unnecessarily inflate the vocabulary and create inconsistencies between otherwise equivalent information in different datasets.

Ideas?

Best,
Christian