LINGUIST List 15.2237

Thu Aug 5 2004

Disc: Re: Comments on things no language does

Editor for this issue: Naomi Fox <>


  1. Harald Hammarstr�m, Re: 15.1782, Disc: New: Comments on things no language does

Message 1: Re: 15.1782, Disc: New: Comments on things no language does

Date: Thu, 5 Aug 2004 07:06:29 +0200 (MEST)
From: Harald Hammarstr�m <>
Subject: Re: 15.1782, Disc: New: Comments on things no language does

Re: Linguist 15.1782, Linguist 15.1816, Linguist 15.1941 

> Date: Wed, 2 Jun 2004 23:32:54 -0400 (EDT)
> From: Everett Daniel 
> Subject: Comments on things no language does
> I have recently finished an article, on my website, about some
> properties of Piraha culture and grammar that seem unique: no
> grammatical number, no numerals, no counting, no morphologically...

I'd like to point out that it's quite unlikely that the absense of
numerals is unique to Pirah�, given that there are many other more or
less credible references in the literature (I'll list some of the more
interesting cases below). The difference, thanks to Everett(s)[1] and P.
Gordon[2], is that Pirah� is well-documented.

In an otherwise excellent article Geoffrey P. Smith states that "The Macro-Ge
language (sic!) of Brazil is the single case of a language reported to
have number words only for 'one' and 'many'" (p. 6 of [3]) but gives no
source. He could be referring to e.g Kren�k or Canela-Krah� as most of the
other J� languages seem to have attested numerals.

For Canela-Krah�, Diana Green writes "A l�ngua Canela (da fam�lia
ling��stica j�), por exemplo, n�o tem termos num�ricos espec�ficos;
limita-se a termos gerais tais como: 's�', 'um par', 'alguns' e 'muitos'
(Jack Popjes, comunicac,�o pessoal 1983; Popjes & Popjes 1986)." [4] p.
181. The source must be personal communication because Popjes & Popjes
1986 (in _Handbook of Amazonian Languages_, ed. Derbyshire & Pullum), as
far as I can see, does not feature such a claim.

For Kren�k, an even stronger case for the absense of numerals is made
by Loutkotka. he compares independent attestations of vocabulary from
different tribes and concludes that "Les mots d�signant les nombres sont
excessivement primitifs dans la langue des Botocudos. En g�n�ral on ne
compte pas dans cette langue, car il n'existe qu'un seul nombre, c.-�-d.
l'unit�" (p. 125 [5]). However, from the discussion it is also clear that
they, as opposed to the Pirah�, were familiar with the concept of
precise integral counting for small values (with aid from fingers).

Likewise, an early grammar of Chiquito reports ".. etama, uno; ominama,
algunos, pocos; auc�r�, muchos; ana��a, todos. ... Para decir dos, tres,
etc. muestran dos � tres dedos, y dicen: omina hane, son algunos de esta
suerta como estos dedos. ... No se puede en chiquito, ni contar dos, tres,
quatro, etc., ni decir segundo, tercero, etc., sino es acaso por mil
rodeos y circunloquios, de los cuales ya no se usa porque han aprendido
los Indios � contar en castellano, que es mas f�cil y mas claro .." (p. 19

Another genetically unrelated Amazonian language is conjectured by
Aikhenvald & Dixon as "It is likely that Jabut� originally had no numbers. In
the contact situation 'one' is given as uici (we have no information about its
original meaning). For 'two', je-bo is used, involving the verb root -bo
'be equal' and the reflexive-type prefix je-..." (p. 358 [7]). Of course, I
would not disqualify its numeral status just because it has a transparent
etymology, or assert that there was no other word for 'two' before the
latter expression became current. But the original source, an MA Thesis
from Uni. Campinas [8], is in accessible to me so I cannot try to check if its
truth-value semantics or frequency is different from e.g English 'two'.
Also, it leaves the attested Jabut� numerals (1, 2 and 4 (= 2 + 2)) in [9]
(p. 50) unexplained!

Outside Amazonia, there are some scattered vague assertions that the Veddas
(aboriginals of Ceylon) had no numerals, such as Wijesekera "They do not
seem to have words to express numbers or period of time" (p. 104 [10]).
But a check in the original source [11] (pp. 84-91 and p. 107) reveals
that this is just an absense-of-evidence is evidence-for-absense-fallacy
topped with a bit of hearsay.

A very tricky case is Bernatzik's account of the Yumbri of [12]. He
explicitly declares that they have no other distinction than 'one'
vs. 'many' in counting e.g items, days or years. On the one hand he is fairly
clear on the epistemological basis of his gatherings throughout the book. On
the other, as Rischel puts it "He was clearly very much inclined to conclude
that phenomena about which he had failed to retrieve information did not
exist in the culture, and that the Mlabri were not capable of reasoning
about matters unless he could persuade them to communicate such thoughts
to him" (p. 40 [13]). Rischel's research (published 1995) on what he calls
Minor Mlabri, reveals the numerals 1, 2 and 3 (expressed by 2 and an
adverb 'plus one' at the end of the NP!) in that language (p. 146 [13]).
However, Rischel's Minor Mlabri (as well as Major Mlabri) are shown to be
closely related to "Yumbri", but _not_ identical. Neither could he find
evidence of an ethnographic or genealogical identification between Minor
Mlabri and Yumbri.



[1] Daniel L. Everett 2004, _Cultural Constraints on Grammar and Cognition
 in Pirah�: Another Look at the Design Features of Human Language_,

[2] Peter Gordon 2003, _Numeral Cognition Without Words: Evidence from

[3] Geoffrey P. Smith 1988, _Morobe Counting Systems_ in Papers in New Guinea
 Linguistics 26:1-132.
[4] Diana Green 1997, _Diferen\cas entre termos num�ricos em algumas l�nguas
 ind�genas do Brasil_ in Boletim do Museu Paraense Em�lio Goeldi, S�rie
 Antropologia 12(2):179-207.
[5] C^estm�r Loutkotka 1955, _Les Indiens Botocudo et leur Langue_ in
 Lingua Posnaniensis V:112-135.
[6] L. Adam and V. Henry 1880, _Arte y Vocabulario de la Lengua Chiquita con
 algunos textos traducidos y explicados compuestos sobre manuscritos
 in�ditos del XVIII^o siglo_, Librairie-�diteur J. Maisonneuve, Paris
 (Biblioth�que Linguistique Am�ricaine VI).
[7] Alexandra Aikhenvald and R. M. W. Dixon 1999, _Other Small Families
 and Isolates_ in The Amazonian Languages (eds.) R. M. W. Dixon and A.
 Aikhenvald, pp. 341-383, CUP. (Cambridge Language Surveys)
[8] N. N. Pires 1992, _Estudo da Gram�tica da l�ngua Jeoromitxi (Jabuti):
 aspectos sint�ticos das cl�usulas matrizes_, MA Thesis, Universidade
 Estadual de Campinas.
[9] C^estm�r Loutkotka 1963, Documents et vocabulaires in�dits de langues
 et de dialectes sud-am�ricains_ in Journal de la Soci�t� des
 Am�ricanistes LII:7-60.
[10] Nandadeva Wijesekera 1964, _Veddas in Transition_, M. D. Gunasena & Co,
 Ltd., Colombo.
[11] H. Parker 1909, _Ancient Ceylon: An account of the aborigines and of
 part of the early civilisation_, Marwah Publications, New Delhi.
[12] H. A. Bernatzik 1938, _Die Geister der gelben Bl�tter: Forschungsreisen
 in Hinterinden_, Bruckmann, M�nchen.
[13] J. Rischel 1995, _Minor Mlabri: A Hunter-Gatherer Language of Northern
 Indochina_, Museum Tusculanum Press, University of Copenhagen.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue