Has anyone read the actual article referenced in this one:  http://www.newscientist.com/news/news.jsp?id=ns99996303 ?  It describes research done with members of a tribe in Brazil having no words for numbers greater than two.  At least in this summary of the research article, the point seems to be that the study, which showed that members of this tribe had difficulty with the concept of numbers much greater than two—as demonstrated by an ability (or lack thereof) to replicate numbers of objects, numbers of sounds, etc., indicates that because they have no word for the concept of, say "four," they THEREFORE have no understanding of it.  Is this really the researchers’ premise?  And how does this study prove it?  Why is it a lack of vocabulary that they are sure limits the thought?  Surely it is more likely to be the other way around…

That was an interesting article, overdone.

I’ve also heard that "One, two, and many" quote applied to unspecified Native American tribes.

According to another article, on Excite.com’s Odd News feed (Great Excuse for Poor Math Skills), there are only about 200 Piraha and they live in tribes of 10 to 20, so they really don’t need large numbers.  They do seem to have a hard time in a matching game using more than three objects. It appears that everything happens in threes for them, not in fours as for the Navajo.    

Curiously, while the Piraha adults had trouble learning (matching) large numbers, their children did not!  

Finally, the article states:

"Not only do the Piraha not count, but they also do not draw," Gordon wrote. "Producing simple straight lines was accomplished only with great effort and concentration, accompanied by heavy sighs and groans."

Actually, drawing a straight line is probably a fine-motor skill that other cultures practice from early childhood on.  I’m sure the Piraha are better at hunting and tool-making than the the average Westerner.

Another article I found following some related links, Ancient shell jewellery hints at language, makes what I see as a related point:

"They [the shell bead makers in the article] must have had a way of describing the symbolic message that the beads represent," he says. "If I were Catholic and you were not, and I wanted to explain to you how rosary beads are used to count prayers, it would be very hard to do that with sign language, or a grunt and a groan - you wouldn’t get the message at all."

Grunts aside, it would also be hard to do even with speech without the words for "count" and "prayer," for instance, just as it seems to be hard for the Piraha to comprehend the concept of "four" and higher numbers.

After reading the article it sounds to me like a simple matter of learning certain concepts in childhood that we take for granted.  I would also say that, to a point, language does influence thought, but it does not necessarily limit it.  

To understand the problems the adults had with concepts of large numbers culture, education, lifestyle, and a myriad of other concepts must be factored in.  They are hunter gatherers which means they spend little time on what we might deem "standard" education.  Also, they probably tend to deal with concepts of "enough" and "too much" instead of worrying that they only have 512 yams this harvest as opposed to 525 last time.  When your life rests upon physical labor output not winning a debate you tend not to sweat small things like counting large numbers.

Personally, I think there are some holes in the logic of the experiment and some important variables being omitted.

If you go to http://www.languagelog.org/ and search them for piraha, you will find quite a lot of information from qualified linguists.

We’ve looked at this topic before.  Please refer to the following thread

Cheers, BNJTOKYO

The lack of numbers greater than «2» among the Pirahã didn’t come as a surprise to me, but the lack of matching ability (among adults) did. I’ve always just assumed (on the basis of no evidence whatsoever) that peoples whose languages lacked a developed numerical concept would compare amounts the way mathematicians compare infinities—by matching them. Peter Gordon’s work would indicate that this is not necessarily the case, and that the lack of a number concept in the language does lead to difficulty in dealing with quantity, at least among (a majority of) people of a certain age. It would be fascinating to know at what age the «window» for learning concepts for larger numbers and matching processes closes ! Perhaps it’s not such a bad idea—or perhaps it is ?—that arithmentic is taught in primary school….

Henri