1.7 Dr
avidian and Indo-Aryan
39
verbal participles as heads of subordinate clauses, changing finite verbs into verbal
adjectives (relative participles) before noun heads, extensive use of echo words, and the
use of classifiers and quantifiers (the last one Emeneau’s exclusive discovery), a feature
of Tibeto-Burman languages shared by Northeastern Indo-Aryan and transmitted to
Dravidian. Emeneau (1965) examines the occurrence of retroflexes
.
t,
.
d,
.
r,
.
n,
.
l in several
languages of the Iranian group – Balochi, Pashto, Ormuri, Parachi, Yidhga, Sanglechi–
Ishkashmi, Wakhi – and the unique Burushaski. He says ‘the isogloss then runs roughly
north to south through Afghanistan and Baluchistan ...Bilingualism, involving Indo-
Aryan languages must be the answer’ (Emeneau 1965; repr.1980b: 128–30). In further
studies, Emeneau (1969a, 1971a, 1974b) discusses syntactic parallels between PD
∗
-um
‘also, and, ev
en, (with numerals) inde
finiteness’ and Skt. api with the same meaning
range (see section 8.4.1 below), the Dravidian use of the past participle derived from the
verb ‘to say’
∗
an-/
∗
en-/
∗
in- (now in my reconstruction
∗
aHn-) and the use of Skt. iti as a
quotative particle,
and using phonological strategies in distinguishing male and female
members of various castes and subcastes etc. (Sjoberg 1992: 510).
Andronov (1964b) thinks that the replacement of negative verbs by ‘special negative
words’ is a feature of Indo-Aryan found in Dravidian. William Bright (1966) notices a
phenomenon, similar to the Proto-South Dravidian rule iu> eoin the environment [a
(umlaut, vowel harmony), which he calls ‘Dravidian metaphony’, also in some Modern
Indo-Aryan and Munda languages extending from Assam to Ceylon which he calls a
‘linguistic area’ (see Krishnamurti 1969b: 324). Kuiper (1967) particularly discusses the
use of iti ‘thus’ as a complementizer of onomatopoetic expressions from the Vedic times
as strikingly Dravidian in origin.
Most writers (Bloch, Emeneau,
Kuiper) are agreed on
the use of the absolutive form of the verb, the gerund, as head of non-finite clauses as
a typical Dravidian feature of Indo-Aryan syntax. Masica’s typological study of shared
features (1976) has led him to isolate at least four, typically marking the South Asian
area from the rest, namely retroflex consonants, Skt. api/Dravidian -um meaning ‘even,
also, and, indefinite’, dative-subject constructions,
and
finally echo-words. He has a
qualified ‘yes’ for nine features, because typologically they extend to much larger areas
outside South Asia. These include the conjunctive particle, morphological causatives
etc. (187–90) (see Krishnamurti 1985: 224).
Emeneau has continued to add to the ‘areal features’ in his papers on onomatopoetics
(1969a; this is not specifically South Asian), onomastics (1978), and intensives (which
he prefers
to call
‘expressi
ves
’now;
1987a). He has further looked at different kinds
of phenomena in ethno-semantics, which have an areal bias, in his papers ‘ “Arm” and
“leg” in the Indian linguistic area’ (1980b: 294–314), and ‘The right hand is the “eating
hand”: an Indian areal linguistic inquiry’ (1987b).
In a comprehensive and well-documented paper, Andr´ee Sjoberg (1992) discusses the
impact of Dravidian on Indo-Aryan. She has added to the observations of Emeneau and