How widespread is Sound symbolism

[Robert L. Rankin]

> would be interesting to look at the data from Boas' Dakota again to
> see whether /s/ and /sh/ were more common than corresponding /x/ in
> shifted forms statistically.

I think you'll find that if only two of the three places of articulation
participate in symbolic sets (NOT aug./dim. symbolism, but the "real"
pan-Siouan phenomenon), that it will the the dental and palatal.  There
may be a few exceptions, but that's how it seems to me.

Bob