Antw: Re: O-only agreement - Sign languages

[jess tauber]

In many spoken languages one finds iconic components in the deictic-demonstrative system, and these are then available for further grammaticalization, and as is well known two common sources of case marks are serial verbs and relational nouns. Now, in Kalam-Kobon as the serial verb system has shrunk it has become more iconically organized, and I would guess that this sort of thing happens more than is generally recognized. I have no information as to whether relational noun systems also undergo such evolution, but it would be interesting if it turned out to be so.

It seems to me, after looking at hundreds of languages, that such secondary iconization tends to happen as ideophone-type iconicity dies (as fusion and synthesis progress), applying perhaps first to lexical items, then grammatical ones. What happens next? If I'm right in hypothesizing that ideophonic diagrammatical iconicity itself starts out as a post-indexical effect, and there is a cyclicity to developments of the Peircean triad from iconic to symbolic to indexical and then back to iconic, then use of iconic demonstration in signed languages (both dynamic and static) is easier to understand. 

In my model I have opposed grammaticalization and antigrammaticalization (the latter term now unfortunately claimed also for two completely different theories), where the former is fed by lexicon, and the latter produces ideophones (which eventually feed the lexicon). Cyclicity demands that there must be another link connecting grammaticalization products (which swing towards larger pragmatic use) to materials that eventually coalesce into ideophones. It is clear also that phonological-segmental realization gets minimized at this second juncture (while prosody is maximal).

The use of modeled referential space in signed languages may show that these languages are specialized to make use of this part of the cycle- extensive iconic 'classifier' systems, which may be very similar to the large ideophone systems found in many languages of low fusion and synthesis, would be party to the next natural step in the cycle, utilizing both space and motion in diagrammatic-iconic ways. These legs of the cycle would be opposite, on the cycle, to those linking lexicon to grammaticalized units, and be more akin to the 'antigrammar' part of my model. Apples and oranges.

Jess Tauber
phonosemantics at earthlink.net