Locatives and wa- problems.

[Rankin, Robert L.]

> Ok, but in a purely synchronic phonological analysis of Lakota (disregarding history and morphology), you would have to say that these CCV syllables are one syllable, starting with a cluster.  Right?  That is the analysis of Boas and Deloria, Carter, Shaw.

If they actually make such assertions I think it's from a Eurocentric view of the syllable.  It's what we were taught in grade school.  I used to do an experiment in phonology class asking how many syllables were in the word TRAIN (as in choo choo).  All Americans said "one".  Spanish and Portuguese speakers said "two" and Korean and Japanese speakers all said "four".  Everyone relied on training and on their native languages, etc.  The question to ask is how those linguists handled accent on words beginning with GL or BL (incl. MN) clusters.

Here's an interesting question though:  What percentage of words beginning with BL, MN or GL carry accent on the second vowel to the right counting from the beginning of the word?  In Kaw the answer shows pretty clearly that those CC clusters are counted as having an extra syllable.  These are nearly all initial syllable accented words.  How would Pat, Dick, Trudy or others handle that "linguistically significant generalization"?  Should they just ignore it?

> I know I am a little bit out of the loop in Siouan studies, but is there a more modern study of Lakota phonology showing that these are synchronically 2 syllables?

I understand what you're saying, but there's no such thing in modern phonology as an analysis that completely disregards rules of accent placement/shift, morpheme/word  boundaries, or even the actual identities of morphemes.  In addition, although we can say that we are disregarding "history", various rules/constraints are always affected materially by the relative chronology of changes in the phonological system that took place centuries ago.  We just invent purely synchronic ways of talking about such information.  In standard generative phonology it was extrinsic rule ordering along with tags within rules like "minus rule 159" or "plus French loan", etc., later on it was lexical compartmentalization, rules were replaced with constraints and so forth.  I lost track after the beginning of "optimology", but no matter what the model, the data will require consideration of all these characteristics of language.

As linguists became disillusioned with endless phonological theorizing, phoneticians reasserted themselves and we were treated to the "IPA fetish" and 20 minute Power Point spectrogram presentations desperately condensed into 3 minutes because of laptop glitches at LSA/SSILA.  They liked to ignore the complexities of phonology to concentrate on machines and minutiae.

> I don't mean to go into a big argument regarding this, but I am just curious.

Oh, don't worry, I don't take these discussions personally.  I just think it's important to air all the issues for the sake of all the eager young students who read our stuff.  :-).

Best,

Bob


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/siouan/attachments/20130909/17bc30b8/attachment.htm>