tekinged / missing

The repository where the tekinged.com committee tracks and defines missing words. Anyone can join!
https://tekinged.com
Apache License 2.0
5 stars 0 forks source link

Listing all verb forms #182

Closed johnbent closed 1 year ago

johnbent commented 1 year ago

Listing all verb forms created by johnbent@gmail.com on 2017-01-04 10:20:29

johnbent commented 1 year ago

johnbent@gmail.com replied,

In another thread we discussed how we include the word okedii but not the word okedau. Here's a bit of discussion about why this is true. Please chime in with suggestions and comments about this if you have any.

In terms of which verb forms are added, here's an email I sent recently to Jim about how difficult this would be in addition to how it might make the website harder to use. The problem is that there are way too many forms of each Palauan verb!

We have thought about adding all forms beyond just the 3rd person forms. Currently we only add when they're irregular. I wanted originally to add all of them but Lewis Josephs talked me out of it. Correctly I think. Each verb has the base verb and then 5 hypothetical prefix forms (k-, l-, m-, d-, ki-) and 10 suffix (none for imperfect, then nine perfective: -k -au -ii -d -mam -miu -ir and one for plural inanimate objects that sort of doesn't have a suffix) so that's a total of 60 forms for each verb. Then double that to 120 forms for each verb if you want to add plural forms. Adding them to the displayed list would hurt the usability of the site and how do we find them all anyway? Because there are also variant forms of each of the above; for example, sometimes it's -mam and sometimes it's -emam.

If we could somehow gather them, it would be useful to add them to the database as hidden words that aren't displayed but can be searched to find the root word. Although, if they're regular, it'd be better maybe to teach the rules to the website and have it say, 'that looks like the hypo 3rd person singular of mengang' or something.

johnbent commented 1 year ago

palau371@gmail.com replied,

what do you think about extracting of basic forms which have v.inch. and v.pred?

johnbent commented 1 year ago

johnbent@gmail.com replied,

Can you give two examples please? I think most of us don't know the Josephs terminology.

johnbent commented 1 year ago

johnbent@gmail.com replied,

Can you give two examples please? I think most of us don't know the Josephs terminology.

johnbent commented 1 year ago

johnbent@gmail.com replied,

Can you give two examples please? I think most of us don't know the Josephs terminology.

johnbent commented 1 year ago

palau371@gmail.com replied,

omkobk: 

basic form: obkobk

inchoative: obkobkang

predictive: obkobkung

i think about the basic forms too much, i guess the are derived, but not inflected, maybe we should separate them. but i'm not sure about it yet

johnbent commented 1 year ago

palau371@gmail.com replied,

omriid:

basic form: obriid

inchoative: obridang

predictive: obridung


i think, basic forms having inchoative, predictive and reduplicated forms can be extracted. 

johnbent commented 1 year ago

palau371@gmail.com replied,

i think we can choose the most unpredictable verbs and figure out how to deal with them? OMES group looks monstrous now. 

johnbent commented 1 year ago

johnbent@gmail.com replied,

Maybe we could make '-UES' be a root word and call it v.hyp.suffix or something. Link it to OMES, and then list the kues,mues,lues,etc. under it?

johnbent commented 1 year ago

misc@leitmotiv.org replied,

I like the idea of "roots," but I'm afraid that the cost/benefit tradeoff is the hard part...

Finding the root is often an exercise in morpho-phonological analysis.  You have to look at multiple forms (not just verb forms but also derived nouns, etc.) to find a common root that can be used as a basis for all of those surface forms -- each one is like a puzzle.  So despite the usefulness of roots, my opinion (though maybe I'm missing something) is that it would be too time-consuming/error-prone to do this on a large scale.

That being said, I've been giving some thought to trying to come up with an automatic morphological analyzer (using some tools from Xerox/PARC research center).  This could aid in automatic identification of roots.  But this would be quite a large project (I'd love some help), and success is not guaranteed.  Anyone (who knows somebody) with both Palauan language and computer programming expertise and wants to discuss further, feel free to open another thread on Debugle and I'm happy to provide some pointers to the documentation/tools.  :)