Aymara's time metaphor reversed? Yahgan says....

[jess tauber]

Source: http://www.sciencedaily.com/releases/2006/06/060613185239.htm

Apparently Aymara is the only (so far) documented language whose speakers view the past as being ahead and the future behind. The terms NAYRA 'eye, front, sight' is grammaticalized to mean 'past', while QHIPA 'back, behind' is used for 'future'. 

The piece notes ambiguities in such metaphors in languages such as English, and the possibility that the evidential system may be a cause of the reversed mapping in Aymara.

Comments:

Yahgan, a couple of languages 'down' on the 'left' in Chile, may have cognate terms. For instance TELLA 'eye, face' may match Aymara NAYRA. USHPA 'back, behind' may match QHIPA. Yahgan evidentiality, from the manuscript sources I have, does not appear to be as obligate as in Aymara, but there are a good number of forms. Yahgan simple past morpheme is -de: (e: tense, relatable to TELLA?), simple future -u:a (u: tense, relatable to USHPA?).

USHPA in Yahgan is also grammaticalized in the language to mean 'after X'. In the now extinct 19th century Yahgashaga dialect (the basis for the vast majority of writings on the language) it was a free or encliticized morpheme following the full or pronominal subject NP (SOV, SVO orders normal). Tense morphemes, on the other hand, are suffixed to the verb stem, after aspect, but before mood affixes. Bound evidentials are variably placed within TAM.

While there is no OBVIOUS grammaticalized form of TELLA in this position, it is interesting that there is the form TU:LA, meaning 'if X'. I don't know enough about the grammaticalization chains in this area to be able to say with any authority that TU:LA is 'from' TELLA, but perhaps one of you can tell me? Yahgan has many lexical doublets where there is opposition of meaning carried by vowel alternants. Or TU:LA may relate to U:LA meaning 'don't' 

The suffixed evidentials include (but are not limited to) -MUSH 'hearsay' (from MVRA 'to hear, listen', V is schwa), -MIN 'visual evidence' (no simplex but is AMIN 'look, see!' with prefixed mild imperative A-), -TIKALVRA 'distant past visual evidential' where -VRA may be a form referring to 'contrariness to expectation'.

The -TIKAL- component may have relatability to TELLA, as also may be the form -DVGA-, having a 'past' meaning. From the historical viewpoint, I have elsewhere (and here?) claimed possible relation to Salishan languages, also on the Pacific coast (but in northern North America) as well as to Chemakuan, and further south Chumashan families.

The Chemakuan 'evidence' implies that the doubled -LLA in TELLA originated in -L- followed by a velar or uvular fricative (other forms support this). One sees similar things in the other families I include in the comparison. Thus, IF true, then perhaps the -K/G- elements seen in the above Yahgan forms are explained (with perhaps reordering in -TIKAL-).

I do not know whether -MUSH (and source MVRA) is relatable to USHPA, either etymologically or by convergence. I had previously speculated that -VRA (also -ARA) 'contrary to expectation' or 'new information' WAS relatable to -MUSH/MVRA. And in terms of etymologies within Yahgan, there are MANY different lexical forms beginning with USH- referring to the lower back or rear, its functions on the body, or products of same. In Salishan the lexical suffix -APS/UPS has the same range of reference.

The suffixal (extended) TAM system in Yahgan appears to be relatively young Most of the relevant morphemes still retain corresponding lexical entries Positionally the clitic-second (really a chain of slots) group also often has matching terms in the suffixal system. For instance -MUSH 'hearsay' as suffix is identical in form to (')MUSH in the clitic chain set. Further grammaticalization has pushed the latter MUSH towards simple modality, seen as well in the surviving dialect 130 years after the grammar was laid out for the extinct one by Thomas Bridges.

The first time I'd ever heard of the reversed time metaphor in Aymara was during a talk given by Eve Sweetser (who not surprisingly is the coauthor of the study cited above) during the Santa Barbara LSA Linguistic Institute a few years 'back'. As I haven't read the source article yet I don't know whether other local South American languages were included in the study. Joseph Greenberg of course was famous (or infamous depending on one's vantage) for attempting to lump, er.. group many of the Pacific coast languages under the 'Andean' rubric. It would be interesting to see whether they (and perhaps other languages/families further afield) share this way of organizing their grammaticalization chains. Since the report on the article says that younger bilingual generations appear to be switching over to the dominant European time metaphor, it is possible that in the past the region was far more uniformly 'reversed', with viewpoints doing 'the 180' one by one after the Conquest. Such changes are probably completely lost now in the mists of history.

Of more theoretical import, perhaps, would be the finding that directionality of time metaphor might depend in part on 'directionality' of grammaticalization? Where tense (if they actually ARE tense) morphemes come from may be determined in part by such a directionality, if there is one, implicit (perhaps hidden?) within the overall scheme of possible canalizations.

So if any of you aren't having too much fun vacationing comments would be of use. Thanks.

Jess Tauber
phonosemantics at earthlink.net