<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=windows-1251">
<META content="MSHTML 6.00.2600.0" name=GENERATOR></HEAD>
<BODY>
<P>
<DIV></DIV>
<DIV>I guess what I'm after it this- languages can vary in the proportion of the grammar instantiated by syntagmatic (ordering) and paradigmatic (closed class) elements. Obviously the more grammaticalized the paradigmatic side of things becomes the less syntagmatic the grammar has to be, and thus ordering (or reordering) becomes more open for purely pragmatic purposes, as is often the case for heavily head or dependent marked languages.</DIV>
<DIV> </DIV>
<DIV>On the other hand where well-grammaticalized paradigmatic elements are relatively lacking, syntax has much more of grammatical edge to it. When this is the case, is there more of pragmatic interpretation of the lexical items which might serve grammatical functions? For instance, word class specification seems in some isolating languages to be less predefined in the underlying root lexicon, and more dependent on where in the sentence structure the item shows up. Might the degree of lexical versus grammatical force be more dictated by discourse factors, etc.?</DIV>
<DIV> </DIV>
<DIV>This is what I mean by complementary distribution.</DIV>
<DIV> </DIV>
<DIV>But it goes further than this. In my crosslinguistic work on ideophone numbers and distributions it is clear that as synthetic tendencies in languages grow, any separate class of ideophones shrinks and eventually de-coheres, and many of the surviving roots show up in other word classes. It has recently been demonstrated that at least for one language (work done on Pastaza Quechua by Janis Nuckolls), ideophones can have grammar functions, even if highly circumscribed (in this case marking aspect). But rather than being grammaticalized, I propose that they are instead antigrammaticalized, both processually in their creation and in their functions, since they normally distract listeners from the "just the facts, ma'am" streamlining of normal grammaticalized morphology, and otherwise exhibit properties more suitable for the pragmatic side of the coin.</DIV>
<DIV> </DIV>
<DIV>Ideophones are thus both a source of new lexical roots (with typological shift and reanalysis) and have their greatest use in languages with the least morphology (though the particulars depend on the degree of verbiness versus nouniness of the language, and there are both adverb like ideophones as well as adjective like ones (as in languages such as Ulwa)). The facts further suggest that the pragmatic side things has a more lexical (or in this case even prelexical) focus in languages with weak grammaticalization. </DIV>
<DIV> </DIV>
<DIV>To me the most fascinating aspect of all this is that ideophones, in languages with very large numbers of them, are themselves paradigmatically organized, in that slight internal changes yield definable slight shifts of meaning. This form-internal paradigmaticity (part of their phonosemantic transparency to analysis) degrades as the class shrinks, and form-external grammatical paradigms grow (complementarity again?).</DIV>
<DIV> </DIV>
<DIV>Are their basic word order correlations as well? I've noticed that many left-headed/right-branching languages (I'm specifically referring here to verb-left languages) have relatively small ideophone classes- instead there is often "internal syntax"- complex patterns of reduplication, infixation, and augmentative/diminutive shifting of vowels, consonants, or both. The functions of these seem mostly related to aspect, either spatial or temporal, and not to any higher level grammaticalization. Yet in many cases normal lexical roots (often underspecified as to word class) are themselves to some extent phonosemantically transparent.</DIV>
<DIV> </DIV>
<DIV>Right-headed/left-branching languages, on the other hand, tend in my 300-language sample to be quite different. Internal syntax is relatively rare and/or uncomplicated, but on the other hand there are often many ideophones. Normal lexical roots tend to be on the opaque side (except for those shifting in from the ideophone class).</DIV>
<DIV> </DIV>
<DIV>Ideophones and internal syntax normally don't seem to mix, despite the fact that both are iconic. More complementary distribution? I'd be interested in hearing opinions about how "grammatical" internal syntax is when highly developed in languages (for instance, in some of the Salishan languages it seems choice of reduplicational patterns is either more lexicalized, or more situationally (thus pragmatically) grounded).</DIV>
<DIV> </DIV>
<DIV>I suspect much more is going on. But nobody is going to look unless they suspect something is there to be found.</DIV>
<DIV> </DIV>
<DIV>Best,</DIV>
<DIV>Jess Tauber</DIV>
<DIV> </DIV>
<DIV> </DIV>
<DIV>...Jess E. Tauber</DIV>
<DIV>--- <A href="mailto:phonosemantics@earthlink.net">phonosemantics@earthlink.net</A></DIV>
<DIV>--- EarthLink: The #1 provider of the Real Internet.</DIV>
<DIV> </DIV>
<P></P></BODY></HTML>