Aymara's time line

[jess tauber]

Hi- I just spent a large part of the weekend looking up morpheme correspondences in languages along the west coast of the Americas that may have utilized  the same or a similar grammaticalization chain, though of course this will not answer whether these ended up with any sort of reverse oriented time axis given cultural assimilations in the meantime.

It would be interesting to see whether any other properties of these languages associate with similar sourcing of grams- for instance the many American west coast languages which had well-developed augmentative/diminutive shifting for existing lexemes, yet curiously very few 'classical' ideophones. Aymara and Yahgan are both like this. There may be some motivation to look for a 'back door' route to explain some of this- normally ideophone roots infuse the lexicon with fresh sound symbolic resources, but shifting does not seem to be part of the deal- rather such word-class change brings with it change FROM modification (as in ideophones) to relatively direct reference (as in lexical roots).

It may be that augmentative/diminutive shift affects derivational morphology FIRST- there are for instance many languages with both diminutive shift AND a diminutive affix cooccurring, but fewer with just the shift. It has always bothered me- why develop an affix if shift is unambiguous? But of course if the development started with the affix and then worked its way into the modified root, the affix becomes redundant, and can be dropped. Interestingly, this would jibe with what is believed by historical linguists to be a possible track of development of glottalization on obstruents (when not borrowed as for instance in Quechua, from Aymara and other Jaqi languages). Phonological systems with glottalized obstruents are extremely more common westward than eastward in the Americas, as are those with more than minimal aug/dim shifting. There seems also to be, in these languages, a tendency to allow consonantal identity to historically vary much more than the vocalic melody- perhaps a consequence of aug/dim shifting and selection of different alternants in different daughter languages. And these vocalic melodies in turn have strong sound symbolic associations- almost as if they carried the major semantic 'theme' of the lexeme. Opposite of the usual situation as in Eurasian languages?

On the face of it these and other language properties and processes might appear rather far removed from relevance. But then again stranger things have been discovered. If pragmatic information (such as affective/attitude notions in aug/dim shift) start to infect the lexicon through grams (rather than the reverse), this would imply a development in the direction opposite that expected. The actual morphemes involved in such shifts are already quite 'worn away' - down to individual free features in many cases, glottalic and other. They help to pump up a numerically impoverished lexical rootstock, one perhaps with eroded phonology (a less capable source of new grams?)- pointing to things that have been said about polysynthesis by various scholars. One may see similar things going on in certain analytical languages (such as Matisoff's Lahu). Pragmatics ends up informing the lexicon by phonological distinctive feature 'micromanagement' rather than the creation of entirely new lexical roots, as in the case of ideophones.

Jess Tauber