Locatives and wa- problems.

[Rankin, Robert L.]

I think you'd be wrong.  By accent placement rules and by morphological analysis the GL and BL clusters count as two syllables.  The little phonetic tics are immaterial.  Fortunately or unfortunately the Gs all go back to full syllables, mostly KI while the Bs of the BL clusters all go back to WA or WI.  All were morphemes also.  Ordinarily the prehistory of these things might not matter, but the accent rules still seem to be able to treat the Gs and Bs as morae for purposes of assigning stress synchronically.  This is especially true of Hochunk which, assuming Ken Miner was right, is a mora counting language.  I'm guessing that Dakotan is too.
________________________________

> Phonetically, there is a schwa in there.  But phonologically, I would count glV- as one syllable.

Actually, that's backwards.  Phonetically BLV and GLV may form single syllables but phonologically they count as two for the reasons cited above.  \

It gets worse, of course.  If the structure is CVglV the syllabification rules for Siouan languages assign the /g/ phonetically to the second syllable along with the initial member of all other CC clusters.  I remember telling an Australian linguist that and being laughed at because he believed that syllable boundaries could be derived from a "universal."  It's all very messy, but it's a fact that CL clusters can behave as two syllables for various phonological purposes and perhaps as single syllables for yet other purposes.

Bob
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/siouan/attachments/20130909/e698f915/attachment.htm>