Funny W

[Rankin, Robert L]

> I've been assuming that they were parallel too, but I'm a little mystified on how *W => w across Dakotan, while *R => Da. d, Na. n, and La. l.  By that division, I would have expected *W => Da. b, Na. m, and perhaps La. w.  And I think Winnebago also has *W => w, while *R => d, doesn't it?

It would be nice if parallels were perfect, but the original r/w were typically in different phonological contexts.  I suspect that's a good part of it.  Doesn't *W turn out as [b] in some Dakotan dialects?  (And in LA it's [b] before /u/.)  As I recall, you have doublet instrumentals wa/ba and maybe wo/bo??  Check both Buechel and Riggs.

> Why laryngeals?

Actually more than one kind of consonant can be involved, but h and ? are the ones that pull disappearing acts and remain the most likely candidates in those cases where no other conditioning factor can easily be identified.  As I say n the handbook article, one has to be very careful not to use such things as "finagle factors".

> 'Snow' would be an example ('Spring' too,I think) with *wa-wa > *w-wa > *Wa.

For Dakotan it's wa, for OP it's ma, and for Osage/Kaw I believe it's pa/ba.  Where are we finding the *wa-wa combination?  Southeastern?  And how do we know it reflects the primitive state, rather than just being a reduplication or something?

I think the assumption here was that #wa- was the absolutive and formed the nominal as opposed to a verb.  It is the absolutive that most strongly tends to undergo syncope (along with the 1st sg. wa-), leaving a [b].

> *w+glottal would be the sort of thing we've discussed before on the list with regard to the verb ?oo 'to wound, shoot at and hit'.  In the 1st person you would have *w(a)-?oo. Unfortunately I've never found all the conjugated forms of this verb in most of the languages.  In Dakotan, analogy as reintroduced the full wa-prefix.  Hi?u is another case that David pointed out, with hibu in the 1st person sg. (b is the allophone of /w/ that occurs preceding /u/ in Dakotan.)

> I'm not following.  Does this have to do with *W, or are we talking about a separate *w+glottal development here?

Hard to say, there are so few cases of W.  It may result from Cw or wC, where C includes certain consonants and laryngeals.  Someday maybe we'll collect enough examples to be sure.  Or maybe Blair's Catawba will elucidate more of the puzzle.

> *w+r gives all those bl- stems (Omaha bdh-).  What is happening is that ordinary [w] and [r] are assimilating a feature from an adjacent consonant or sonorant that is causing them to obstruentize one degree.

> So *w+r => MVS bl/bdh, and *p+r => Dakotan bl, other MVS *R ?

My own analysis is that there aren't any *pr sequences.  I think there may have been one or two of the 'flat' terms that looked as though they MAY have had a sequence /para-/ or something similar that had collapsed, but I think all the bl clusters go back to *wVr with syncope of the V.  In most cases the identity of the wV- is fairly clear with the inanimate *wa- or animate *wi- absolutives being the primary culprits.  The other instances are 1st sg. *wa-.  The main problem with the putative pr- is that I can't identify the /p/ part as a morpheme.  The problem is messy, so there is room for more than one hypothesis, certainly.

I'll be in Oklahoma the next coupla days.

Cheers,

Bob