Re: Classes of cmavo

[John Cowan <cowan@locke.ccil.org>]

la .and. cusku di'e

> > No, it's a morphological one.  Cmavo are of CV['[V]] form, or
> > certain expanded forms, but in any case with only one C.
> > Brivla invariably have more than one C.
> 
> That seems to be more a phonological condition.

Well, no.  Phonologically, "'" is a consonant, but it is not a 
C (a morphological term).

> I am sure it is the
> most robust definition, but I doubt that it echoes most people's
> intuitions about the essence of cmavohood.

Doubtless.

> BTW, can cmavo be used as cmene? I know I could licitly be {la rocta}
> (la rosta sounds like I am an opera diva),

Not so clear.  In my view, one may not introduce new gismu into the
standard language, even when they are used for names: a complete
Lojban parser would be entitled to report "rosta" and "rocta" as
errors.  AFAIK this is not definitively settled, though.

> but could I be, say, la fo`o`o,

Yes, because the minute you introduce an experimental cmavo, you
are outside the standard language, and extended syntax may be in effect.

> or la dai?

Definitely not.  This string is ungrammatical, because it contains
only standard cmavo and can't be parsed by the standard syntax.
(To be precise, "dai" is attached as an indicator to "la", and then
there's no name to parse at all.)
 
> And what is the status of hybrids like "smi`i" (Smithy), "ma`iu"
> (Matthew), "Aga`a" (Agatha)? Are they valid lojban words?

"ma'iu" is an experimental cmavo: see above.

".aga'a" is the compound cmavo ".a ga'a", and so ungrammatical as a name.

"smi'i" would be valid iff it were a valid fu'ivla, but it trivially
fails the "slinku'i test": "pu smi'i" becomes the lujvo "pusmi'i", and
lujvo have precedence over fu'ivla, so "smi'i" is not a Lojban
word of any kind.

-- 
John Cowan	http://www.ccil.org/~cowan		cowan@ccil.org
	You tollerday donsk?  N.  You tolkatiff scowegian?  Nn.
	You spigotty anglease?  Nnn.  You phonio saxo?  Nnnn.
		Clear all so!  'Tis a Jute.... (FW 16.5)