Phoenician Indonesia (Was: etymology of ketchup/catsup?

Barry Garcia wrote:
>Oh that page is funny, especially the vocabulary bits. I just love the
>connection of "asam" with "asem". Not surprisingly it looks alot like the
>Tagalog root for sourness, "asim" (maasim - sour) .
    As well it should, since Ml. asam, Tag. asim, Jav. as@m (and many
others, tho Dempwolff cites only 1 (irreg.) Oceanic cognate) regularly
reflect AN *qasem (_e_ represents schwa) or in Zorc's more recent version
*qalsem-- the *-l- or ?*-al- may have been an optional infix; other PI
langs. retain it e.g. Iloc. alsem, Bicol alsum, Ibanag attam.  The *-l- may
be a feature of Proto Philippine (assuming ALL (or most) PI languages have a
common source, still a controversial view).  Zorc develops Tag. as follows:
*qalsem > **qa( )sem > á:sim-- though, ignoring other PI forms, it could
equally well reflect *qasem.  The difficulty with the Egyptian connection
("alloyed metal"), needless to say, is that it requires a certain leap of
semantic faith, as the AN word seems to have meant only 'sour'.

> Tagalog in fact will
>often change /e/ to /i/: babae - babae + ng - babaing. >
    My impression is the /e/ is rather rare in native Tag. vocabulary;
historically this word was probably **babai, with sporadic lowering of the
unstressed /-i/ in a word of some frequency ('woman').  Lots of Spanish
words with /e/ were borrowed in Tag. and elsewhere with /i/-- likely the
earliest borrowings, before massive borrowing from Span.(and later Engl.)
introduced phonemic /e/ and non-final /o/.

>Anyway, question for romilly: does "araw" (day, sun in Tagalog)  descend
>from the AN "waRi", if so, i'm wondering how it got to "araw" (any ideas?,
>though i know you focus in Indonesian).>
    Nope.  There were at least 2 AN words for 'day/sun'; the other was
*qaljaw--
that *-(a)l- "infix" again, mainly attested in the PI and Formosa, lots of
IN langs suggest an optional nasal, ?*qa(n)jaw; along with difficult *j.
Again, Zorc proposes PPh *aljaw > Pre-Tag **aldaw > **a( )daw > á:raw.  The
normal Tag. reflex of *j is (intervocalic) -l-, final -d, apparently blocked
in this case by the early cluster.
Note Bisayan adlaw, with regular metathesis of **-ld-.
It's possible *qaljaw referred to daylight, daytime, while *waRi was more
specifically sun/sunlight; its modern reflexes frequently mean 'to dry in
the sun' or some such, but that's speculation.  Both words are widely
attested in all the major areas (Formosa, PI, IN, Oceanic).