My query:
>>Another curiosity:  Arabic borrowings with original /g/  occasionally turn
>>up with Ml/In /dZ/, written "j". No examples come to mind at the moment,
>>sorry.  Some peculiarity in the pronunciation of Ar. /g/?   Where would
>Arab
>>traders in the period ca. 900-1300 C.E. have come from?  IIRC mainly from
>>the Hadramaut, so perhaps a dialectal pronunciation?


Basilius wrote:
>Can you give any examples? What do you transcribe as /g/? 'Ghain'?


Not _ghain_ I'm sure; more likely this-- i.e. the 3d letter, Hebrew gimel---
>Early arabic /g/ ('jin') has became /d_Z/ in the modern educated
>pronunciation and in most dialects (sometimes with further change
>to /Z/). The old sound value of this letter is attested by many
>borrowings, both *from* and *to* Arabic. So there may be a confusion
>here.

Examples are hard to present, since generally I've noted the correspondence
in miscellaneous articles; my dictionary doesn't cite Arab. originals.  But
here is one probable ex.:  _gaib_ mysterious vs. _ajaib_ miraculous,
amazing.  Perhaps also: _jahanam_ Hell (if equiv. to Biblical Gehenna?).
There are rather few visibly Arabic words in the _g_ listings; many under
_j_, but that can equally well reflect borrowing of /z/.

An additional problem is that some regional langs. borrow Ml. /j/ as /g/.

I'm almost ashamed to say, most univerities in this country don't require or
even urge a course in Arabic for those concentrating in Indonesian
historical linguitics.
So occasionally, one finds (especially) lexicostatic lists of regional
languages treating reflexes of Ml. _pikir_ to think (Arab. fikr?) as if it
were a native word.