LINGUIST List 2.786

Tue 12 Nov 1991

Qs: Inflection/Derivation, Sinhalese

Editor for this issue: <>

Directory

1. David Stampe, 2.770 Query on order of inflection vs. derivation
2. BROADWELL GEORGE AARON, sinhalese

John Nerbonne <nerbonnedfki.uni-sb.de> asks whether the Russian
reflexive suffix s'/s'a, which follows agreement suffixes, isn't an
exception to Greenberg's 28th universal, viz that derivation comes
between root and inflection. If it's a true reflexive (I'm not a
Slavist), I'd question whether it is derivational. My favorite rule
of thumb for whether a morphological process is derivational is
whether it can be avoided by paraphrase. Reflexives proper usually
aren't avoidable when Subject = Object.
(Here's a query of my own. In languages with S and O marking on the
verb, there is one paradigm for cases where S is not equal to O, and
another for cases where S is equal to O ("reflexive"). But I've never
found such a language with ANY way of expressing cases where S properly
includes O, or vice versa, e.g.
	I saw us in the mirror.			*ourselves
	We (including you) discussed you.	*yourself
Why is reflexive defined as complete rather than partial equality?)
For another possible exception to Greenberg's universal, antecedent
clauses in conditional sentences in many head-last languages are
formed by nominalizing a fully inflected sentence, and then postposing
a postposition meaning "if". E.g. in Sora, a Munda language (India),
 anin-ji dOng-nAm kong -l -e-n -ji-dEn, gd-t -Am
 he -PL ACC -you shave-PA-3-NOM-PL-if, cut-NONPA-you
 = If they shave you, you'll get cut.
The if-form, -dEn, is postposed to a plural, -ji, which is suffixed to
a nominalizer, -n, which is applied to the (almost) fully inflected
clause anin-ji dOng-nAm kong-l-E-n. (Normally the 3 PL suffix -ji,
agreeing with the subject, would come next. Instead, it is added to
the nominalized clause, which is at least less surprising since it is
also the 3 PL suffix.) In any event, the nominalizer is certainly NOT
between the verb root and all its inflection.
However, since it is not the root that is being nominalized, but the
verb and its clause, it seems to me that the nominalizer is (almost)
exactly where it OUGHT to be. How else could you say this in Sora? I
have to confess that I can't think of any understandable paraphrase.
So is this all inflection? Or are the characteristics we try to line
up under the headings of DERIVATION vs INFLECTION simply out of line
here? Maybe we should erase the headings and start over.
As I recall, Uma Subramaniam of OSU gave a paper at CLS about 1985 on
related difficulties that Tamil presents for this "universal".

In a recent piece on Sri Lanka, the author claims 'scholarly opinion is
divided on whether Sinhalese is an Indo-European or a Dravidian language.'
This sounds like outright nonsense to me -- I have never heard anyone
suggest that Sinhalese is Dravidian. But I don't work in that part of
the world. Could someone more familiar with these languages tell
me whether there is really any difference of opinion on the Indo-
European status of Sinhalese?