Study of obscure Amazon tribe sheds new light on how language affects perception (fwd)

[jess tauber]

Dan Everett, whose work was ultimately the basis for this report, posted this (see below my contribution) a few days ago on LINGUIST- I suggest that readers take a look particularly at the paper link on his website:   Cultural constraints on grammar(http://ling.man.ac.uk/info/staff/DE/cultgram.pdf)

My own take is that the Piraha~ situation is symptomatic only of its morphosyntactic/phonological status, which prevents deep hierarchicalization- in its syntax, in its morphology, in its phonology, and in parallel with cultural knowledge organization (complexity of tools, depth of generational knowledge of kin, myth, etc.). This status may in fact be a normal fact of life in the morphosyntactic cycle when heavily polysythetic languages with a lot of morphological fusion break down towards the isolating/analytical type again. One sees in various parts of this phase of the cycle languages with low numerical specification, tendency towards borrowing pronouns (either from external sources or from internal ones such as distance demonstratives), lack of syntactic embedding, perhaps a trend towards serialization, and ultimately monosyllabification, which helps to re-complexify the phonology and also may give rise to tonal and/or vocalic register, and in extreme cases, clicks.

Grammatical complexity in such languages is completely overshadowed by complexity in the pragmatic side of the coin, and possibly also in dominance of prototypically right-brain processes over left-, prosody over segmentality phonologically, and so on.

Piraha~ is in the early-mid stages of such breakdown, if my hypothesis is correct. Information in such a language is radically distributed over the entire discourse, and also not limited mostly to the speech channel, and thus one would expect a great deal of paralinguistic effect to accompany it. Nothing in Everett's paper (and see also the story text:   http://ling.man.ac.uk/info/staff/DE/panther.pdf) really seems to contradict such an analysis.

One symptom to look for in languages in this part of the cycle (grossly) is a strong tendency towards phonological monosegmentality of morphology, as well as incorporation of adverbial and higher level materials into it (the residue of former polysynthesis).

Jess Tauber


Message 1: Piraha numbers, counting, and culture
Date: Fri, 20 Aug 2004 10:09:10 -0400 (EDT)
From: Daniel Everett <dan.everett at man.ac.uk>
Subject: Piraha numbers, counting, and culture

Many readers of this list may have seen today's CNN report on the
Science article which was published yesterday on Piraha
(http://www.cnn.com/2004/TECH/science/08/19/science.counting.reut/index.html).
The Science article reports on research carried out by Peter Gordon,
along with me and my wife, Keren, several years ago to test my view
that Piraha had no counting. Gordon confirmed that indeed Pirahas have
no concept of counting and further refined my original ideas on the
matter.

Gordon's conclusion in Science is that Piraha offers support for the
Whorf hypothesis. While I believe that this is plausible, my own view
is that the lack of counting must be seen in the larger cultural
context and that when thus viewed in conjunction with the lack of
color words, the lack of embedding, the simplest kinship system ever
documented, and various other characteristics, a different,
non-Whorfian picture emerges. The basic conclusion I reach is that
culture constrains grammar in ways many of us have not previously
imagined. I take this to be an argument against, for example,
Universal Grammar, at least the more widely-accepted versions of it.

Anyone interested in reading on this further (and the Piraha case does
seem to be getting a lot of attention from various countries, based on
my email folder this morning), my paper "Cultural Constraints on
Grammar in Piraha", currently under review, is available from my
University of Manchester website: at
http://ling.man.ac.uk/info/staff/DE/DEHome.html