Re: Classes of cmavo

[And Rosta <a.rosta@UCLAN.AC.UK>]

Colin:
> What is interesting about this classification, is that I have found it
> quite hard to find a grammatical characterisation of selsmuma'o that
> excludes brivla! It seems to me that, while cuvma'o are clearly a
> different kind of animal from selsmuma'o and brivla, on both structural
> and semantic grounds it is difficult to distinguish the class of brivla
> from any other selma'o - except that it happens to be much larger.

Aren't {du}, {co`e}, {go`i} both cmavo *and* brivla? Or do brivla
exclude selbri valsi that are cmavo?

Either way, I think the essence is that cmavo form a closed class,
while cmevla, lujvo and fu`ivla are open classes. The cmavo/nae`e
cmavo distinction is more a lexical one than a syntactic one.

Interestingly, gismu and rafsi also form a closed class, though, so
the difference between gismu and cmavo must be purely or primarily a syntactic
one.

--And