modeling verb classes

[Brendon]

I'm running into trouble trying to model inflection for verb classes in Abawiri (Indonesia: Lakes Plain). Abawiri has (at least) 4 verb classes that take somewhat different TAM suffixes. My problem mainly has to do with the optionality of the perfective/imperfective suffixes.

• Class 1 takes an obligatory perfective or imperfective suffix, followed by other optional TAM suffixes.
• ker-u 'break-PFV
• ker-i 'break-IPFV'
• Class 2 takes an allomorph of the perfective suffix if another suffix follows, otherwise no perfective. No imperfective suffix.
• toi-j-e 'cut-PFV-DECL'
• toi 'cut.PFV/IPFV'
• Class 3 takes no perfective or imperfective.
• fwrì 'blow.PFV/IPFV'
• Class 4 is the same as Class 1 except that with one of the second-slot TAM suffixes, the imperfective (obligatory in Class 1) cannot occur. 
• kàr-u 'carry-PFV'
• kàr-i 'carry-IPFV'
• kàr-ro 'carry-CONT' (compare Class 1 ker-i-ro 'beak-IPFV-CONT')

In Flex, I initially assigned verb lexical entries to 4 inflection classes but found that this didn't work since I'm not dealing with lexically conditioned allomorphy. My second attempt was to create 5 subcategories of the Verb category, and for each build a separate template. A problem is that most of the TAM suffixes are used in multiple verb classes. When I add a suffix to more than one template, Flex automatically creates an additional sense in the lexical entry of that suffix for each position it holds in a template. The result is a messy set of multiple senses in affix lexical entries.

A related complication with this second approach is the directional prefixes, which apply to all verbs regardless of verb class. I created a template with these prefixes at the main Verb category level, but the parser doesn't seem to allow this template to apply, followed by the nested template in the appropriate subcategory. Alternatively, if I add the prefix slot to every subcategory template, I will get (at least) 4 senses for each prefix lexical entry.

I haven't tried inflection features yet. I think this method could solve some of the problems (e.g. constrain continuous, irrealis, habitual, etc to only occur after imperfective and not perfective). But I don't think it could handle the perfective/imperfective suffixes that are obligatory in some classes and ungrammatical in others. I could tag verb stems and suffixes with the appropriate inflection features. If the perfective/imperfective slot is obligatory, but all the suffixes in it can only go with inflection feature Class 1, what will happen to a Class 2 verb?

Any advice would be most appreciated.

Brendon

----

Brendon Yoder, CPhil

UC Santa Barbara and SIL International

[Kevin Warfel]

Hi Brendon,

 

If I’m understanding your explanation correctly, my approach would be what you described as your second attempt, with subcategories of Verb and a separate affix template for each subcategory, since the differences between the classes are structural. However, I don’t have the aversion that you do to having multiple senses in the affix entries, and that is a direct consequence of this approach.

 

I wonder if a satisfactory solution for you lies in adopting the subcategory approach but then finding a way to address what you dislike about the fact that the affix entries have multiple senses. What is it about the multiple senses in these entries that you don’t like? Maybe I or someone else could suggest a way to work around that.

 

-Kevin

 

Kevin Warfel

Associate Dictionary and Lexicography Services Coordinator

a.k.a. Dictionary Development Coordinator

SIL International

 

Current technology makes it possible to provide those translating into just about any language with both a dictionary and a thesaurus in the target language, the standard tools of the trade for professional translators, so why are mother-tongue translators in minority languages still expected to do their work without these tools?  Ask me about Rapid Word Collection after reading about it at rapidwords.net.

--
You are subscribed to the publicly accessible group "FLEx list".
Only members can post but anyone can view messages on the website.
To change your status, please write to flex_d...@sil.org.
You can join this group by going to http://groups.google.com/group/flex-list.
---
You received this message because you are subscribed to the Google Groups "FLEx list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to flex-list+...@googlegroups.com.
To post to this group, send email to flex...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/8c9def49-00a6-421d-b7d7-c8e57529333d%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Brendon]

Hi Kevin

Thanks for your reply. 

There are a couple of reasons why I don't like the idea of having multiple senses in entries for affixes that occur in multiple templates. The main reason is that I think this would make working with verb inflections in Flex cumbersome. In Abawiri, after the perfective/imperfective inflection there are 2 additional slots with maybe 7-10 TAM suffixes. The form and function of these suffixes remains pretty much consistent regardless of verb class. An approach with multiple templates would mean up to five senses in each TAM suffix entry. But it would be more accurate to have a single sense for each since they (largely) have a single grammatical function. Doing manual interlinear parsing would get difficult because I'd have to make sure I'm choosing the right sense in each case. And dictionary outputs would get cumbersome. 

Andy Black suggested offline that I could use a combination of inflection classes and inflection features. Here's a summary of this approach as I understand it:

• Verbs are tagged with inflection classes
• The PFV and IPFV suffixes have allomorphs tagged with inflection classes
• There is a single verb template (or possibly separate perfective and imperfective templates, but both in the Verb category)
• There is a custom inflection feature TAM that is used to constrain co-occurrence of IPFV and the other TAM suffixes.
  

Thanks to both of you for your suggestions.

Brendon

----

Brendon Yoder, CPhil

UC Santa Barbara and SIL International