Funny W. More cold water.

[Rankin, Robert L]

> wa- 'absolutive'
> wi- 'absolutive' for many animates
> wa- '1st sg. actor'
> As I've said before, those would be my guesses.
>  All three have tended to undergo syncope.

> Yielding a stop, such as p ?  Or do other values occur after the syncope?

[m] before /?/ or maybe nasal vowels
[b] before /r/ (or l, d, dh, n, etc.)
[p] before /h/, and voiceless obstruents such as /p, t, k, s, s^/

>  *R is the second member of a *w-r cluster in many cases.

Is this *w-r cluster leading to *wR actually attested as such, or do we only have cases where the *w has developed full oral closure, as in Dakotan blo/bdo/mdo situations?

[b] and [m] are allophones or, in some languages, former allophones, of /w/.

>  Other instances aren't so clear, e.g., the doublet demonstratives, *re: and *Re:.  I suppose some sort of reduplication might have been involved in Re:, but that's speculation.  

I like the idea of reduplication.  Given the rather parallel nature of *W and *R, could both of them have arisen from syncope of *w- and *r- prefixes against an initial of the same type?  I.e., *w-w > *W, *r-r > *R ?
 
That's my thought, but unfortunately I don't see a lot of evidence for reduplicated deictic particles.  Then again, I haven't looked for them.
 
Bob