Prosody-pragmatics rich language- the case of Piraha~?

Prof. Dr. Wolfgang Schulze W.Schulze at
Thu May 25 16:19:06 UTC 2006

Dear Aya, dear Funknetters

>The linguistic cycle is well described by Dixon in THE RISE AND FALL OF
Let me first add that the Linguistic Cycle (or: Grand Cycle, as we 
usually term it) is well described already in the tradition of 19th 
centruy German grammarians. Here, I do not want to turn to the question 
of which position Proto-Indo-European did take in this cycle (I think 
that Dixon referred to that layer of Proto-IE that is reconstructed by 
external evidence. The internal reconstruction of IE clearly opens the 
option for a position more close to four o'clock (isolating via 
agglutination)). Still, one should note that Dixon's clock can go into 
both directions: For instance, Ossetian and Tokharian both show clear 
evidence for the re-agglutination of a former fusional paradigm 
*without* pasing through the isolating stage. In fact, the Grand Cycle 
is just a tendency (with massive recursions), not a *must* for the 
dynamics of language paradigms. 

>The question seldom addressed by functional linguists is which typology is
>likely to have come first. It's an endless cycle, so finding a language at
>any particular clock position proves nothing about how many cycles it has
>undergone previously.
I admit that I have diffculties to understand what Aya means by 
'language' in this context. Perhaps, it is better to term it a 
'linguistic-communcative tradition', because this would help us to avoid 
terming e.g. French and Indo-European 'one [and the same] language'. In 
this sense, the seemingly endless Grand Cycle goes together with the 
fact that language always is always is the reflex of a foregoing stage: 
A 'linguistic-communicative tradition' hence has only *one* starting 
point, that is 'language origin'. Note that here, I do not mean that all 
languages are genetically related - this seems not to be provable (out 
of methodological and theoretical considerations). But what we can say 
is that all language-based communicative systems are originally (!) 
related. The same logically holds for the Grand Cycle: It's starting 
point is immediately related to language origin. However, just as we can 
hardly say anything safe about the early seconds of language origin, we 
cannot describe the position, which language took on Dixon's clock at 
the period of language origin. In other words: The Grand Cycle is 
nothing but a heuristic approach to language change in the period of 
observable and reconstructable historical stages.

I think that it will be difficult ever to describe how language looked 
like in its first stages (in the first seconds of language origin). From 
a modern point of view, it sounds logical to assume that in these 
seconds, language had been 'word-based' (or: isolating), but this is 
nothing but a projection of current linguistic thinking and 
categorization. Maybe that language was (by that time (!)) totally 
different from what we today know about linguistic 'types'. For 
instance, if we start from a phrase-based model, linguistic categories 
as we know them today would have emerged from massive processes of 
re-analysis and paradigmatization of more 'iconic' (or: idiosyncratic) 
utterances. If this is true, the Grand Cycle would have started to 
become a 'tendency' in language change at a later stage. Maybe that the 
Cycle was - in its first run - fed by a number of factors different from 
those we use to describe for the Cycle itself.

Turning to the Piraha~ case, I do not see the necessity to set up a new 
type of the (fortunately dismissed) Stadial Theory (N. Marr) as 
suggested by Aya's wordings. There is no proof that a sociological 
habitus can be immediately related to a linguistic type. There is no 
convincing evidence that growing interactional complexity favors 
linguistic complexity. Nor does exist evidence for the contrary.  
Linguistic complexity is primarily a question of phenomonology. We 
linguists are used to describe something as complex, because we are 
trained in thinking in building blocks and in modeling the world 
thereafter. But we cannot say that a language is 'by itself' (or: 
ontologically) more complex than another (the assumption of complexity 
is immediately coupled with (more or less scientific) perception). In 
other words: Complexity bascially is a meta-linguistic criterion, which 
should be very carefully referred to in terms of a tertium comparationis 
when relating world 'objects' (such as a sociological habitus and a 
language type).

Finally, I think that it is rather dangerous to start from the 
assumption that

>these people were not previously members of highly complex trading
>societies who later returned to a more primitive lifestyle, 
and to conclude that

>complex morphosyntax is the more conservative typology. This
>indicates that our original starting point was far from isolating.
Maybe that say 15.000 years ago, the complexity rate of languages was 
much higher than thousands of years later, but this does not necessarily 
imply that massive complexity was given for the many earlier years, too. 
15.000 years is a little bit of nothing in the history of mankind and 
(perhaps) in the history of language change.

Prof. Dr. Wolfgang Schulze
Institut für Allgemeine und Typologische Sprachwissenschaft  (IATS)
[General Linguistics and Language Typology]
Department für Kommunikation und Sprachen / F 13.14
Ludwig-Maximilians-Universität München
Geschwister-Scholl-Platz 1
D-80539 München
Tel.:     ++49-(0)89-2180 2486 (secretary)
             ++49-(0)89-2180 5343 (office)
Fax:     ++49-(0)89-2180 5345
E-mail: W.Schulze at

More information about the Funknet mailing list