From: "Karapcik, Mike" <KarapcM@...>
> | How plausible is it for tones to have a grammatical
> | rather than lexical meaning? For instance, where many
> | languages have articles, this one would have different
> | tones instead. Or perhaps verb number or tense would be
> | indicated thus.
>
> It occurs in natlangs.
[snip]
> I believe it's the book "Describing Morphosyntax" that gives an
> example from an African language where tone pattern is the only thing to
> distinguish between two different tense/aspect states of the same clause
> (person and number doesn't change). I don't have the book with me right now,
> unfortunately....
One of the first examples in the book, in a discussion of how morphemes are not
necessarily linked to separable "pieces of form":
éyétá emutí
remove.meat pot.ACC
"She removed meat from the pot."
eyéta emúti
remove.meat.MID pot.NOM
"The pot is 'de-meated'."
The language is Maasai. ('e' and 'u' in the example use the [E] and [U] IPA
symbols).
*Muke!
--
http://www.frath.net/