Re: Basque onddo

From: stlatos
Message: 70453
Date: 2012-11-14

--- In cybalist@yahoogroups.com, "Brian M. Scott" <bm.brian@...> wrote:
>
> At 2:11:56 PM on Wednesday, November 7, 2012, stlatos wrote:
>
> > --- In cybalist@yahoogroups.com, "Brian M. Scott"
> > <bm.brian@> wrote:
>
> >> At 6:26:41 PM on Sunday, November 4, 2012, stlatos wrote:
>
> >>> The above opt. needn't look strange, since a similar one
> >>> is needed for mucho / muy no matter what the middle stages
> >>> were.
>
> >> Nothing optional is needed here.
>
> > Wrong. There's Por. muito, abutre, cutelo vs (O)Sp muyt,
> > buitre, cuchiello, with no reg. apparent in either (or
> > between).
>
> >> One is the regular outcome of (U)LT- when the T remains
> >> syllable-initial; the other is the regular outcome when
> >> the T becomes final, as in MULT(U) > muyt, later muy.
>
> And yet your (O)Sp. examples are consistent with this.
>
> Never mind; I was mildly curious about what you'd say, but I
> don't consider you much more capable of serious discussion
> than I do Tavi. A large fund of lexical data is no
> substitute for understanding of what to do with those data.
>


Since you said you got this from Penny, I looked in his book and saw that he didn't claim regularity at all. He doesn't even give an order for all the changes, so discussion of what occurred is made difficult, but if I must assume changes like (using the same type of occasional *intermediates he gave (moyto, etc.)):

*
cuneum
conyo
conYyo
cunYyo
cunYo

cuño


but

*
cico:nia
ciconya
ciconYya
cicoynYa

cigüeña


Older stressed oy/oi > ue in modern Sp. His opt. (non-)metathesis seems to correspond exactly to muyt vs mucho. That is, if cuneum didn't change ny>yn but cico:nia did, there's no reason to assume (non-)metathesis is based on syllabification (specifically, yCV or CyV have no restrictions; only yCC > CyC is excluded).


However, if my opt. ui vs u in Por. instead corresponds to oi vs ui/uy in (O)Sp., then the pos. exists that all ny > yn first, and oy > uy was the opt. change, or some sim. intermediate, then uy > u in Por. but uy stayed ui/uy in Sp. (with ytY > ycY opt. (or in standard: met. then pal.) in Sp, too), vs oi > ui in Por. but oi > uE > (approx. pronunc.) in Sp. This chain would include opt. met. or opt. yñ > ñ (corr. to ñy > ñ / ny), among others.


This corresponds to opt. such as *nux > *noks > *noys > nuez (z analogical < *noce), indicating ks and kt patterned w Ls and Lt. However, if the changes are as in standard (all C>y in these clusters first (before pal.)), this still creates the need for opt. in C as well as the C or? V changes mentioned above (I'll discuss this later).


Moreover, if standard rec. allows that only Ty > pal. occurred, it creates many problems in ordering, including how the seemingly close Sp and Por correspond with vastly dif. changes and ordering:


*
pisce:s
peskes
pekses
peyses
pesyes
pesYyes
pes^yes
peys^es

peixes


would need no less than 3 instances of met., while muito had none. I'll give more ex. and discuss this later.