The recent issue of The New Yorker has a detailed and thought-provoking 
article on Dan Everett's controversial thesis that the language of the 
Piraha, a remote Amazonian tribe, does not use recursive embedding. [*] 
Perhaps even more remarkable, Everett proposes that the Piraha do not 
generalize and thus do not utilize abstract concepts. Instead they live in 
the moment and thus speak only about immediate or specific observations. 
However, could the following interpretation of a Pirahan statement reported 
in the same article be a counterexample to such an immediacy thesis? 
 
As an example of the immediacy thesis, when a group of Piraha were shown a 
King Kong movie, Everett made the observation about their comments: 
''They're not generalizing about the character of giant apes. [...] They're 
reacting only to the immediate action on the screen with direct assertions 
about what they see.'' [page 137] However, any suggestion that such is a 
universal trait of Pirahan speech seems contradicted by a proposition made 
by a Pirahan at another time, which Everett translated as: 
 
 ''Monkeys go to the jungle.'' [page 133] 
 
The context of the utterance of that proposition was during a formal test 
where the subject, a Pirahan, was to indicate in which direction an 
animated monkey head floats on a computer screen, to the upper-left or 
upper-right corner. Those two were the only options. But when asked in 
which direction he anticipates the monkey head will float, the Pirahan did 
not understand the question and answered instead: ''Monkeys go to the 
jungle.'' [page 133] That case illustrated the difficulty of communicating 
with the Piraha. 
 
But it seems clear from the described context that the Pirahan pointed to 
referents beyond the given monkey immediately before him on the screen, 
referring instead to the abstract set of monkeys and to a universal 
behavior of set members. An important aspect of the context was that the 
Pirahan's answer, 'Monkeys go to the jungle', lacked the ability to allow 
the desired polar distinction to match a left or right distinction. Why? 
Precisely because the statement references *all* monkeys by its implication 
that no monkey should be found that does not go to jungle. As such, the 
Pirahan's proposition could be further translated into first-order 
predicate logic as: 
 
 Monkeys go to the jungle. 
 
 For all x, if x is a Monkey, then x goes to the Jungle. 
 
 Ax( Mx -> Jx ) 
 
So it seems that the proposition, 'Monkeys go to the jungle', describes a 
characteristic trait of members of the set called 'monkeys' and as such is 
a clear example of generalizing (universal generalization) and thus of 
conceptual abstraction. Note too that the abstraction defines a 
second-order predicate because the set 'monkeys', being a first-order 
predicate, is a subset of the set of entities that 'go to the jungle'. My 
sideline hunch would be to suspect that even if the Piraha lack discrete 
words for 'all' and 'some', they still manage to communicate such 
distinctions. For example, by the given proposition: ''Monkeys go to the 
jungle.'' 
 
Now of course its likely that something has been lost in the cited 
translation. For example, perhaps the author of the article did not 
accurately quote Everett's interpretation, or perhaps Everett's 
interpretation was hasty. However, it just seems clear to me in the 
described context of its utterance the proposition ignores the given monkey 
and points instead to (1) the set of which the given animated monkey was a 
member and (2) a generalized trait of all set members. 
 
Also, in arguing that the Piraha are intelligent Everett notes that they 
''understand the behavior of local animals and how to catch and avoid 
them.'' (page 131) But does not such understanding necessarily entail 
generalizing types of behaviors to abstracted sets of specific types of 
animals? If so, then if a Pirahan sees an animal of type A, she attributes 
to it characteristic behaviors of class-A animals -- if x is an animal and 
is of type A, then x is ... (list class traits). Indeed, that seems to be 
exactly what we observe in the proffered example above. If the Pirahan can 
conceive of characteristic traits of specific sets of animals, surely they 
are necessarily generalizing and abstracting. And if so, then it seems 
improbable or even unimaginable that their language should fail to 
communicate such abstracted conceptions among community members. 
 
That proposed counterexample to the immediacy thesis aside, the New Yorker 
article is a great read and Everett raises many good observations with 
exciting alternative views, and in so doing does us all a favor by 
subjecting the popular view to skeptical inquiry. The scientific process! 
And even if the Piraha do generalize and abstract, as I suggest above, the 
overall picture sounds like abstraction is sufficiently rare in both 
Pirahan words and existence that this surely points to important 
anthropolinguistic facts to be uncovered, especially related to the 
intimate connection between language and perception. 
 
~Ian Goddard 
 
''The limits of my language mean the limits of my world.'' Wittgenstein 
 
[*] Colapinto, John. ''The Interpreter.'' The New Yorker (April 16, 2007) 
118-37. 
http://www.newyorker.com/reporting/2007/04/16/070416fa_fact_colapinto