|Title:||Clitics and agreement|
|Description:||Re: LINGUIST 16.1000 http://linguistlist.org/issues/16/16-1000.html
Further to Grev Corbett's summary posting on clitics and agreement affixes,
I pointed out in Everett (1996) that clitics and agreement affixes are in
complementary distribution in most cases. In fact, there is a universal
(depending on your analysis of ergativity) which I suggested in that 1996
1. Clitic-Agreement Implication:
a. Subject Clitics -- Object Clitics
b. Object Agreement -- Subject Agreement
If the agreement apparatus is primarily to track the Privileged Syntactic
Argument (e.g. subject, actor, topic - see Van Valin and LaPolla 1997), then
subjects will get agreement first. NonPSA NPs can also be tracked, but
secondarily. Everett (1996) proposed that both agreement and clitics
correspond to a syntactic node along the lines of GB's AGR node. When it is
a complement to the stem it is an affix. When it is an adjunct it is a clitic.
This not only predicts their complementary distribution and the universal in
(1), but it predicts all the different diagnostics for distinguishing clitics
(morphological adjuncts) vs. affixes (morphological complements)
summarized by Zwicky and Pullum years ago.
Everett (1996) also proposes that pronouns, affixes, and clitics are all
simply realizations of the same types of grammatical features in different
syntactic positions. The historical progression outlined there for this
a. [X...] [pronominal features] = pronoun
b. [X [X...][pronominal features] = clitic
c. [X ... [pronominal features] = affix
There are of course many other ways that have been explored in the
literature for explaining the pronoun-clitic-affix interaction. But I don't
know of any other account (or even notice) of (1) elsewhere.
Everett, Daniel L. 1996. Why there are no clitics. SIL-UTA series in
linguistics, Dallas, Tx.
Van Valin, Robert and Randy LaPolla. 1997. Syntax. OUP.