Dravidian substratum in India

From: naga_ganesan@...
Message: 9744
Date: 2001-09-24

An aspect of Dravidian substratum in India. From
Indology at Liverpool arhives.- N. Ganesan


The Prakritic consonant assimilation may be due to a substratum
influence of Dravidian originally spoken in North India: when
people speaking a language with few or very simple consonant
clusters speak another language with complicated consonant
clusters such as Sanskrit, this is what tends to happen. I
think the Tamil assimilation provides a parallel showing
that this is indeed a likely hypothesis.

With best regards,
Asko Parpola

Original Posting is archived at Indology website:

Tamil and Prakrit Consonant Assimilation

Sanskrit consonant clusters, 'rm' and 'rN' change to
Prakrit 'mm' and 'NN' respectively. In Vedic, there are
only very few words with consonant groups 'mm'. Later in
time, most Sanskrit words with 'rm' transform into 'mm'.
This simplification process through assimilation can be
observed in Old Tamil also.


Sanskrit | Prakrit
karNa (ear) | kaNNa
varNa (color) | vaNNa
kIrNa (scattered) | kiNNa
pUrNa (full) | puNNa
parNa (leaf) | paNNa
nirNaya (decision) | niNNaya

Sanskrit | Prakrit
karman (action) | kamma
dharma (law) | dhamma
varman (armour) | vamma
gharma (heat) | ghamma
carman (skin) | camma
kUrma (tortoise) | kumma

Compunds with nir-(without)
nirmathana(churning)| nimmathana (Pali)
nirmala (clean) | nimmala (Pali)
nirmUla (rootless) | nimmUla (Pali)


In old Tamil, parallel processes of assimilation can be observed
from "rum" to "mm" and from "run" to "n2n2' (taking "n"
as ta. dental n and "n2" as ta. alveolar n).

"rum" -> "mm"
1) porumal (plumpness/abundance) -> pommal
2) cerumAn2 (leather worker) -> cemmAn2 (cf. ceruppu = sandals)
3) perumAn2 (great man) -> pemmAn2

"run" -> "n2n2"
1) nerunal (the day before) -> nen2n2al

This assimilation is present partially in certain word pairs
1) parutti (cotton), pan2n2al 2) karumpu (sugarcane), kan2n2al.
Campantar (7th century CE) sings "karunal paravai kamaz kAzi"
Can this *karunal taken to mean sugarcane??
karnATu appears to be a back formaton from kan2n2ATu (boulder/hill

Is this "rm to mm" and "rN/rn to NN/nn" assimilation observed in
both Aryan an Dravidian languages, an area defining feature
like echo words, reduplication, onomotopoeic words, etc.,?
If not, is this observed in other language families?

Comments are welcome,
N. Ganesan