Yivrian /l/ (was Re: syllables)

Just so everyone knows, there's at least 3 levels of historical derivation
in Yivrian, with 4 living languages that I know something about to
compare:

                      Proto-Yivrian
                      /          \
                Old Yivrian    Proto-Hiksilipsi
                /    |   \                |
           Praci Yivrian Tzingrizil     Hiksilipsi (various dialects?)

The relationship of Old Yivrian vis-a-vis its descendants shown here is
much like the relationship of Old Church Slavic to the modern slavic
languages. OCS/OY are not truly the ancestors, but they're very close and
provide the best evidence for what the proto-language was like.

First Nik said:

> I'd analyze that grouping of [l] as proof that [l] is descended from
> something else.  It's in the same group as nasals, so one possibility is
> that [l] is derived from [n].  Yoruba, for example, uses [n] and [l] as
> allophones.
A good hypothesis. But Old Yivrian (OY) already has an /n/, and
Proto-Yivrian already has two n-like sounds, symbolized *n and *N, along
with *l. It'd be hard to get another coronal nasal in there.

(For the curious, PY *n > OY r, and thus into the descendants, while PY *N
> OY n. In H., PY *n, *N > n, nd. Thus, PY *N most likely is /nd/ or
something similar.)

Then John suggested:

> Retroflex [d] would be a logical candidate: it is a stop, it would be
> isolated in OY, and the reverse transformation is known to have happened
> in natlangs, e.g. Latin ll > Sardinian d.d.
This is a good suggestion. I can't think of any argument against it,
except for the objections I'll raise in the next section.

Steg said:

> Could [l] in Old Yivrian have come from some less sonorous sound, for
> instance, a lateralized [d], in Really Old Yivrian?
What do you mean by a lateralized [d]? A [d] with a lateral release? That
might be the best suggestion yet.

> What else is weird about the Old Yivrian /l/?
Basically just its sonority. The evidence for the Obstruent L theory runs
like this:

In OY, we have consonant clusters of obstruent + glide: /tj/, /tw/, /tr/,
etc. But we have no /tl/, /pl/, etc, and we *do* have /lj/ and /lw/. This
all suggests that in OY /l/ at least behaves as a stop, if it wasn't
actually articulated as one. Deciding the actual articulation is
tough--all of the actual children have [l]'s for OY /l/, but they all do
great violence to the /lj/ and /lw/ clusters. Interestingly, /lj/ has
become a *fricative* in each of the languages: Praci [Z], Yivrian [j\]
(voiced palatal fricative), Tzingrizil [K\] (voiced lateral fricative).

Going back to PY, we find that almost all roots are CVC, except for a
couple that end in NC. Here the hard nasals *M and *N pattern as stops,
but so does *l, e.g. *gaml "foot". This also tends to support the view
that *l is a stop in PY. BUT PY also has syllabic consonants, which
include the usual suspects *m, *n, *ñ, but also includes *l, e.g. *k.l
"all, every" and *t.l "smooth". None of the other stops can be syllabic.

Synthesizing all of that, perhaps in PY the *l has two allophones: one, a
lateralized /d/, as Steg suggested, but with a purely lateral allophone
when it was syllabic? Hmmm.

Jesse S. Bangs jaspax@u.washington.edu
http://students.washington.edu/jaspax/
http://students.washington.edu/jaspax/blog

Jesus asked them, "Who do you say that I am?"

And they answered, "You are the eschatological manifestation of the ground
of our being, the kerygma in which we find the ultimate meaning of our
interpersonal relationship."

And Jesus said, "What?"