Re: Word Construction for a New Conlang

On Thu, 8 Jul 1999, Kristian Jensen wrote:

> Boudewijn Rempt wrote:
> >A lot of modern, descriptive grammars have a good and full chapter on
> >the phonotactics of the language they describe - you could do worse
> >than trying to find Solnit's _Eastern Kaya Li_, Rutgers' _Yamphu_ or
> >Ikroro's _The Kana Language_. These really differ from the old-style
> >grammars for classical languages.
>
> I might try and get these grammars for myself so that I sorta have a
> guide of how to describe the phonology of Boreanesian. Question: What
> language families do these languages belong to? Since its Boudewijn
> suggesting this, I'm guessing they are Tibeto-Burman or something?
>
Kana is a Kegboid language (Greenberg puts it in the Cross-River group,
but, well, I'd distrust any grouping put forward by a megalocomparationis
like hime). It's spoken in Nigeria. The others are Tibeto-Burman:  Yamphu
is one of the complex-pronominalizing Kiranti languages (by far the best
grammar I've read in the past year: 632 pages of pure delight), and Kayah
Li is a Karen language from Burma/Thailand. The grammar of Kayah Li is
rather interesting, but also quite expensive, especially since the 380
pages are set in a rather large font.

Dirk, has your Shoshone grammar already been published? That should be
interesting too, especially qua phonology.

One grammar to avoid like the plague is Holmer's _A parametric grammar
of Seediq_. This 'linguist' has amassed an enormous amount of material
about one of the dying aboriginal languages of Taiwan, and what does he
do? He uses it as a mere example for his interpretation of GB Grammar,
and doesn't so much as prove even one little text. It's completely useless
as a description of a language.

Boudewijn Rempt  | http://www.xs4all.nl/~bsarempt