Re: Question about Central Mountain Languages

> I don't see the mechanism itself as artificial, as long as
> it arose from some diachronic process at some point. Also,
> does the same derivational mechanism hold through all 12,000
> years of history, and across different languages in the
> family? That would be very surprising in natural languages
> (although certain trends can last a long time).
The processes of IN formation do change over time.

> Hmm. Why would they be deliberate changes, rather than
> (unconscious) analogy? I understand that they don't work as
> regular sound changes, but analogy could do that.
For analogy you need forms to imitate; Wiyot sound history offers none, and indeed
the sound symbolic changes are at odds sometimes with historical changes among the
same group of sounds.  It's not analogy.  (Something similar occurred later in
Cheyenne; most Proto-Algonquian *p and *k are lost in Cheyenne, but one of the few
places they're retained is in diminutive versions of words, while the augmentive
forms loose *p and *k.  It's semantically-conditioned sound change that functions
as sound symbolism; the Algic-Wiyot phenomenon seems to be the same thing.)