Sanskrit R in Tagalog

Waruno Mahdi mahdi at FHI-Berlin.MPG.DE
Mon Nov 22 19:41:27 UTC 1999

> [stress]; q = final (phonemic) glottal stop
> Waruno Mahdi says: "_l_ is the regular Tagalog and Cebuano reflex of _r_ in
> loans from Malay (also Sanskrit loans which were mostly probably acquired
> via Malay)."
> I beg to disagree.

Jean-Paul, that actually wasn't "my" grand proposition, but something
which has been more or less established for quite a while. There's
an old but still highly useful publication of John Wolff on this
subject (1976, Malay Borrowings in Tagalog, pp. 345-367 in C.D. Cowan,
         O.W. Wolters [eds], Southeast Asian History and Historiography.
         Ithaca-London: Corenll Univ. Press)
not to mention other publications of various authors on Philippine
historical linguistics, in which the proposition might have been
implied or mentioned en passant.

>    First, my impression is that the great majority of Sanskrit loanwords in
> Tagalog were borrowed directly from Sanskrit, not via Malay, otherwise such
> terms would have a Malay form. For instance, Sans. _buddhi_ > Tag.
> _bud[hiq]_, but Malay _budi_ "thought, conscience etc.". It would have been
> impossible for Tagalog speakers to reconstitute the /h/ from the Malay form.

We should be careful not to jump to conclusions too quickly in such
instances. Reflection of Sanskrit lexical items in Malay dates back from
some two thousand years ago, and involved a host of Malay dialects and
styles. The failure of modern Malay to reflect the _h_ does not mean that
there were no earlier forms in which it was reflected. I don't have my
Old Malay sources at hand at the moment, but I remember for instance
that modern _tetapi_ "but" is reflected something like _tathApi_ (_A_ is
long _a_) in Old Malay epigraphy of the Sri Vijaya period.

Furthermore, the retention of such modern doublets like Malay
   _bagi_ / _bahagi_ "divide" (Sanskrit _bhAgi_ "divide"),
   _basa_ / _bahasa_ "language" (Sanskrit _bhAs.A_ "speech, language
                                  _s._ is _s_ with dot underneath)
suggests existence at some time in the past of intermediate forms
with _bh_.

One should also bear in mind, that _dh_ in Sanskrit and in Tagalog
is not the same. In Sanskrit the cluster forms the initial of a
syllable, i.e. it is an aspirated _d_. In Tagalog on the other hand,
if I understand corredtly, the _d_ and the _h_ belong to two different
syllables, i.e. the _h_ is not an aspiration of the _d_, but is the
entire initial of a syllable. Historically, in authentic (i.e. not
borrowed) vocabulary, such internal clusters in Tagalog developed
as a result of deletion of an intervening schwa (i.e., if this had
been not a borrowing but an authentic Tagalog word, the "protoform"
would have been **budehiq). This is however ALMOST the typical
way, in which Malay words with internal consonant clusters were
transmitted throughout the archipelago. Thus Malay _kerbau_ "water
buffalo" is reflected in many languages as if there had been a
"protoform" **kerebau with epenthetic schwa (see Tagalog _kalabao_).
The reason I say "ALMOST typical" is that schwa in loans from Malay
is typically reflected in Tagalog as _a_ (which would not have deleted,
but would have led to *budahiq in Tagalog). However, we have
no instances of this internal schwa before _h_, so that it is
theoretically possible, that this internal schwa got deleted before
_h_ in Tagalog before shifting to _a_ when in other positions.

>    Second, from my corpus, mainly based now on Noceda y Sanlucar (1754,
> 1860), _Vocabulario de la lengua tagala_, Sans. /r/ is reflexed as /l/ in
> Tagalog, not as /r/.

If you read what I wrote once more, you'll see that that's exactly what
I said, that Sanskrit and Malay _r_ is reflected in Tagalog (and Cebuano)
as _l_. Your many examples indeed confirm this quite impressively.

The role of Malay in the furtherance of lexical items of Indic origin
throughout insular Southeast Asia has been underestimated, I think, and
there is a lot of interesting evidence to substantiate it.
One is the retention of the same particularities in the Malay reflection
of Sanskrit in other languages of the Archipelago.

e.g. Sanskrit _upavAsa_ "fast (not to eat)" > Malay puasa "id."
The reflection is regular for Malay (tendency to drop initial vowel
in words with more than 2 syllables, neutralization of antepenultimate
vowel to schwa with subsequent fusion with a following glide to a
high vowel, in this case pav_._ > paw_._ > pew_._ > pu_._).
Cebuano has _puqasa_ with the same phonological particularity
(I don't have a Tagalog reflex in what I have at hand at the moment).

Also: Sanskrit _cukra_ "vinegar" > Malay _cuka_ "id."
Loss of _r_ as a result of simplification of the consonant cluster
_kr_ (there's an interesting article by Bob Blust on the general
phonological procedure, I don't have it at hand, but I think it's
1982, An Overlooked Feature of Malay Historical Phonology, Bulletin
of the School of Oriental and African Studies 45:284-299; I'll
look it up at home and correct this if I'm mistaken).
Tagalog and Cebuano both have _sukaq_ "vinegar", with the same
phonological particularity.

One other interesting feature in the Tagalog and Cebuano reflexes of
Sanskrit items is the very frequent postglottalization of originally
final vowels. This is difficult to explain if the borrowing had been
directly from Sanskrit, but would be understandable if the Malay
dialect spoken by sailors sailing the route through the Java Sea and
Strait of Makassar to the Philippines had automatic postglottalization
of final vowels. Indeed, this happens to be the case in Brunei Malay
and Banjarese (the Malay dialect of East and Southeast Kalimantan),
and Sander Adelaar reported numerous instances of unexplained final
glottal in Betawi / Jakarta Malay -- all of which lie on that sailing
route (I made this suggestion in endnote #83 in W. Mahdi, 1994, Some
Austronesian Maverick Protoforms with Culture-Historical Implications I,
Oceanic Lingusitics 33:167-229; see there p. 214).

There are still further points one could bring forward, but this is
already much too long and I have a bad conscience vis-a-vis some
list subscribers with slow internet connections, so I'll stop here
with apologies to those subscribers.

I see you've amassed an impressive amount of material and have started
on some very profound processing of the data. I'm looking forward very
much to see the publication I hope you'll make on all that.

Salut,   Waruno

Waruno Mahdi                  tel:   +49 30 8413-5411
Faradayweg 4-6                fax:   +49 30 8413-3155
14195 Berlin                  email: mahdi at
Germany                       WWW:

More information about the An-lang mailing list