Marathi repha

Madhav Deshpande mmdesh at UMICH.EDU
Sat Nov 1 13:02:57 UTC 1997

It is not certain that the two Rephas of Marathi have anything to do with
Dravidian.  At least I have not heard or read that argument.  But one can
probably make a generalization that the words directly borrowed from
Sanskrit show the Sanskrit Repha.  However, the reverse is not true.
There are non-Sanskritic words which can have the Sanskritic Repha as
well.  For example daRyaa "ocean", but daR'yaa "valleys".  Here the first
word has the Sanskritic Repha, (though it is a Perso-Arabic loan?).
Generally the palatalized and aspirated consonants in Marathi are confined
to contexts derived from purely Marathi morphophonemics, such as
plural-formation or formation of the oblique forms etc.  The original
words themselves could indeed be Sanskritic.  For example, consider darii
"valley", which is Sanskritic, but has its plural daR'yaa with the
non-Sanskritic Repha.  This leaves the first short vowel Laghu "light".
But this is a common pattern not limited to Repha.  For instance, nadii
"river", plural nadyaa "rivers".  Though this word in Marathi is written
the same way as the Sanskrit instrumental singular of nadii, the Marathi
word is pronounced differently from the Sanskrit word.  The Marathi word
has its first syllable laghu, while the Sanskrit word has its first
syllable guru.  Thus, the standard Marathi orthography hides a lot of
phonological distinctions.
        All the best,
                                Madhav Deshpande


> At 07:51 AM 10/31/97 -0500, you wrote:
> >Ashok's observations are correct in that a cluster beginning with R' does
> >not make the preceding vowel heavy.  Phonologically a combination like R'y
> >can indeed be treated as a palatalized R.  In Marathi, this is not the
> >only palatalized consonant.  Many other consonants can be palatalized, but
> >are written as clusters without making the previous vowel heavy, e.g.
> >vahyaa "notebooks".  There are similarly nasal consonants which get
> >aspirated and are written as clusters, without making previous vowel
> >heavy, e.g. panha "drink made from green mango".  There is a discussion of
> >this phenomenon in Aparna Jha's booklet on Marathi phonology published by
> >Deccan College many years ago.
> >        Madhav Deshpande
> >
> Maharastrians are counted among the pancha drAviDAs ( five Dravidian groups).
> Is the presence of two `rEfAs' in their language something to do with their
> draviDian connection?
> sarma.

More information about the INDOLOGY mailing list