Print

Print


On Sun, 21 Apr 2013 18:22:33 -0400, Anthony Miles <[log in to unmask]> wrote:
>
>>Eh, one has flexibility.  In vocabulary listings in particular I've taken a variety of different approaches.  For instance,
>> in  http://akana.conlang.org/wiki/Kib%C3%BCl%CA%8Ci%E1%B9%85/Lexicon ,
>>I list verbs as a pair of bases (nonpast and past), but I list nouns as their unsuffixed form and indicate (here by
>> superscripts) consonants of the base whose identity is unclear in the citation form.  
>
>>Alex
>
>R: Due to the high incidence of aspectually-determined suppletion in Siye, I've taken to using the citation form si/ye, which indicates that /siye/ is the base noun, /si/ is the imperfective root and /ye/ the perfective root. Unfortunately, the participles for the most part use one or the other root, so the citation form  "a/tom" "to engage in commerce" indicates an imperfective/perfective split, but the participle /amakim/ "market, place where selling occurs" uses the imperfective root, while /tomkasunaki/ "goods" uses the perfective root. /tomkasunaki/ must use the perfective root tom- because the "resolved to" suffix -ka requires it. So the verb ends up under A, the first noun ends up under A before the verb, and the second noun under T, far away from the related words. It's not perfect, but otherwise I have to enter many verb roots twice. If anyone has a better system, I'm all ears either here or off-list.

A similar problem exists for nouns derived from verbs by prefixing or with noun-verb compounds.

I'm currently trying to decide on an approach for Apr21. Verbs can have up to 3 stem phases, e.g. qot, qta, qoda, depending on what affixes appear (as in qot "he asked', oqtam "you asked me", qodam "he asked me") and nouns may also vary, with some changing class prefix for the plural and some may add a possessor suffix. The variants actually follow a set of rules, so they can be derived from the root (*CVCV or *qota in the example).

--
Jeff