Talk:Oligosynthesis Project Morphology

• • • Sandhi: a question***

It seems to me that with a restricted lexical root repertory like in this project, compactness of complex derivates may become critically important.

I thought of one thing that could help with that: allowing CCV- and -VCC transformations of CVC-roots in certain types of derivates. (For me, that would make the morphology look less monotonous, too.)

However, to make full use of such transformations, a set of sandhi rules seems to be necessary that would produce consonant clusters pronounceable word-initially and word-finally. (E. g. initial clusters like td- or gk- don't look quite OK, with me.)

It apears that such sandhi (at least if they output only sequences of already established phonemes) would form a complex and idiosyncratic system which, as it seems, must be designed by one participant.

That's why I'm asking. Too big a portion, perhaps critical for shaping the language, from just one person. Is that OK? Or might spoil the game for others?

--Basilius 07:14, 19 December 2006 (PST)

Firstly, I think that a sandhi system could be created piecemeal by a number of collaborators. If you create a wordform that looks awkward, you could then devise a phonological rule that simplifies it. Doing it this way might in fact produce a more complex and idiosyncratic system than one devised by a single person.

Secondly, remember that I've only specified phonemes so far. A given phoneme could have different phonetic realisations in different environments, which gives you more freedom when creating sandhi.

--PeteBleackley 03:36, 8 January 2007 (PST)