Let me add a few more cents’ worth.

The idea that Indo-Aryan, including Sanskrit, fundamentally differs from Dravidian in its syntactic typology, though sanctioned by a certain “tradition” in South Asian linguistics, is problematic on several counts.

First, the only area in which (most of) Modern Indo-Aryan can be said to be robustly right-branching is that of Complement structures (marked by ki/ke/jo/je) — a relatively recent phenomenon, reflecting Persian influence. (Some of the languages also are beginning to adopt postnominal, center-embedded relative clauses, most likely based on the English model.) 

Earlier Indo-Aryan is essentially left-branching, but with a fair amount of word-order (and not just phrase-order) freedom, depending on genre. This freedom is often contrasted with the supposedly very rigid structure of Modern Dravidian. However, Susan Herring has furnished excellent evidence from Old Tamil attesting to a fair amount of phrase-order freedom, includiing structures that are verb-initial, with all other elements extraposed to the right. I have observed similar freedom (through interviews) in a colloquial variety of Modern Kannada, and I suspect that the rigidity attributed to the modern (literary) languages is an effect of diglossia, similar to the relatively more rigid sentence structure of modern literary German, as contrasted with colloquial or dialectal varieties. 

Another feature shared by traditional Indo-Aryan, including Sanskrit, and Dravidian is the fact that relativization can be encoded both in nonfinite form, through (relative) participles, and in finite form, through relative-correlative structures. The latter had for a long time been considered to be Indo-Aryan borrowings in Dravidian (and some linguists considered the Indo-Aryan counterparts to somehow reflect an innovation, triggered by contact with Dravidian); but starting with research by Ramasamy, Lakshmi Bai, and Steever, it has become clear that the structures are indigenous to Dravidian, and I have presented arguments and evidence that the Sanskrit/Indo-Aryan structures are inherited from Proto-Indo-European. 

There is, however, one important behavioral difference: Dravidian relative clauses can only precede their correlative counterparts; in Sanskrit/Indo-Aryan (and PIE) they can either precede or follow (and in some cases both precede and follow, one on each side). Whether this should be considered evidence for “right branching” or “mixed right- and left-branching” might be subject to debate (fueled, in many cases, by theory-internal or even ideological considerations). What is important to know, however, is that relative-correlatives are a common phenomenon in languages that would otherwise be considered “rigid” SOV/verb-final or left-branching.

One more thing that is worth bearing in mind. Recent publications by Pilot-Raichoor (with Murugaiyan) suggest that prehistoric (i.e. pre-Old Tamil) Dravidian may have been very different in its morphology and, by implication, in its morphosyntax, with traces still observable in Old Tamil and elsewhere. At this point, these publications do not seem to have been subjected to the (obligatory) vetting process; from personal experience I know that many Dravidologists are skeptical. Nevertheless, this work suggests that we need to be cautious concerning claims about the syntactic similarities and differences between Indo-Aryan and Dravidian at, say, 1500 BC.

All the best,

Hans Henrich Hock

On 18 Aug 2015, at 07:58, Robert Zydenbos <zydenbos@uni-muenchen.de> wrote:

George Hart wrote:

Of course, Sanskrit compounds can seem
difficult if one’s native language does not mimic their syntax. Both
Hindi and Sanskrit are right-branching, whereas Dravidian is
left-branching.

Please allow me a bit of nit-picking. If by 'left-branching' we mean
that, e.g., attributes precede the substantives to which they refer,
then both Sanskrit and Hindi (and all the rest of the so-called
Indogermanic / Indo-European languages of India, i.e., 'Indo-Aryan') are
quite left-branching indeed. The so-called 'genitive' in Hindi (which is
actually a kind of adjective, inflected kā-ke-kī according to the gender
and case of the following substantive) already illustrates this.

Of course it is possible for genitives in Sanskrit or (very rarely)
Hindi to follow the substantives to which they refer, esp. for metrical
reasons in verse. But it seems that also in Sanskrit prose, genitives,
as attributive words, as a rule precede that to which they refer – which
is precisely not the tendency in a language such as Latin, which has a
more clearly right-branching tendency. I think that this syntactic
feature is one more bit of evidence that the Indo-Aryan languages were
heavily Dravidianized already from their earliest historical beginnings,
as F.B.J. Kuijper and others have pointed out.

Naturally, people who speak those languages find
Sanskrit compounds, which are left-branching like Dravidian languages,
somewhat difficult.

Indeed the internal structure of samāsas is left-branching, which may
explain why an author like Rāmānujācārya in Tamiḻnāḍu sometimes uses
very long compounds such as are uncommon among philosophical authors.

Stella Sandahl wrote:

The Sanskrit compound is not at all as complicated as students (and even teachers) like to think.

Indeed. Though there may be statistically determinable average limits of
quick comprehension among ordinary readers, I am sure that it is largely
a matter of what one is accustomed to.

RZ

_______________________________________________
INDOLOGY mailing list
INDOLOGY@list.indology.info
indology-owner@list.indology.info (messages to the list's managing committee)
http://listinfo.indology.info (where you can change your list options or unsubscribe)

_______________________________________________
INDOLOGY mailing list
INDOLOGY@list.indology.info
indology-owner@list.indology.info (messages to the list's managing committee)
http://listinfo.indology.info (where you can change your list options or unsubscribe)