Re: Origin of Spanich /ch/ and /j/

Isaac Penzev wrote:
>
>Can anybody help me to find the origins of Spanish phonemes /ch/ and /j/
for
>my Arabo-Romance project?
>
>I know that /ch/ in many positions originates from consonant clusters *-ct-
>like in noche < *nocte and *-lt- like in escuchar < *a(u)scultare. But how
>do we get all those _chiquitas_ and _muchachos_?
>
>As for /j/. I know it hides several Old Spanish phonemes: /S/, /Z/, /dZ/. I
>can fugure somehow that in e.g. _dije_ it was /S/ comparing with
Portuguese:
>dije < *dixe ["diSe] < *dissi < *dixit. But what on earth made /L/ turn
into
>jota? What stages had the process? I'm comparing Sp. _ojo_ to Po. _olho_
and
>VL _oclu_ and get lost in doubts...
>
W. D. Elcock, _The Romance Languages_, Faber 1960, or any more recent
summary of Vulg.Latin > Romance should go into the subject.

Basically, it seems *l palatalized > L (Elcock uses IPA lambda) in the case
of *-liV- (e.g. filius > *fiLu) which then reduced > *j, then ("by the 12th
C") >affricate **dZ [or perhaps just *Z, as in modern dialects], then
devoiced **S, then to velar /x/-- this last change apprently circa XVI C.
After initial *k, *p and somtimes *f (clamare > *cLamar, plenum > *pLenu).
Then the old initial Cons. was lost, and the /L/ remained (llamar, lleno).
According to Elcock, Aragonese still preserves the Cl- pronunciation, and a
sub-dial. even has Clj-.  Cf. the similar development in Italian, *kl- >
kj-, *pl- > pj etc.(chiamare, pieno etc.)

Sometime after these developments, old *-ll- was palatalized to yield mod.
Cast. /L/ "thereby filling the place vacated in the internal consonant
system; it underwent no further change until recent times......it has tended
to be reduced to [j]" (so calle [kaje], and note dial. [kaZe]-- while Pablo
Flores (Argentine) usually transcribes the sound with [S]).  IMO this would
appear to be what's called a "persistent rule" in phonology-- i.e. the same
things keep happening over and over, but at different times.

Contrast Portuguese, where *-liV- > **ljV, and stopped-- so olho, Span. ojo.
Whereas in the initial clusters, e.g. *kl- > **klj- > **0lj-, but then
followed part-way the path that word-medial **lj took in Spanish:  *lj > Z
or dZ > S, so chamar (llamar), chuva (lluvia) etc.  (I don't recall offhand
the fate of *-ll- in Port., though probably it remains /L/ |lh| )

An interesting set of almost parallel developments.  It suggests that the
change spread from a single center, but the rules were adopted in different
order, and with slightly different environments, in Span. vs. Port. (as
often happens in cases of rule-borrowing, or so we're told)---

Common to both:  Palatalization of *-li- and certain cases of *Cl- (with
loss of the initial C).  Then Spanish
1. affrication/devoicing of medial **lj (with later changes > velar fric.)
2. retention of initial *lj-
(and note that dialects are continuing to apply Rule 1 to instances of /L/)

versus Portuguese
1. retention of medial *lj
2. affrication/devoicing of initial *lj
(and unless I'm mistaken, Port. lacks further development of these sounds)

As to _why_ VL/Proto Romance *l palatalized...alas, why does any sound
change take place?  Who knows?  But it must have been a common feature at an
early period (in Western Romance, I don't know about Romanian), in order for
it to be found to greater or lesser extent, in Span/Port, Italian and
French.

 Hope this helps-- but you should probably check with more up-to-date
sources than Elcock (or me!).......