Piraha, a lanugage without recursion?

[EdEarl]

I just saw a documentary about the Piraha, an Amazonian culture, whose language apparently has no recursion. For example it has no conjunctions, so you cannot say Mary and Mike drank water. It has started a war among linguists, with Chomskyians and others getting pretty hostile, kind of like the war between theists and atheists...not very scientific in some cases. There is no linguistics forum, so I didn't know where to post this.

 

There are a couple of references to the Piraha in Religion; they do not have a belief in god. The linguist who reported this language went there as a missionary whom the Piraha converted him to atheism.

 

It was also about the Piraha culture, so I guess this could be posted in sociology, The only social controversy is whether the Piraha should be forced to modernize or left untainted by modernization.

 

The Piraha also had no concept of numbers.

 

I thought it was a very interesting documentary.

 

http://www.youtube.com/watch?v=HqkQJiDXmbA

 

 

[Stanstead]

The documentary is startlingly slanted, and almost every fact in it has been disputed (with evidence) somewhere or other. For example, on the topic of Piraha religion, and other parts of the story told in this documentary (like why the researcher Everett has been barred entry to the Piraha villages), you might view the interview here: http://www.youtube.com/watch?v=J3jWI4cPRMg. On whether Piraha has recursion or not, and whether the Chomskyans have actually been "hostile", or simply brought forward counterevidence, you can read the exchange that starts here: http://ling.auf.net/lingbuzz/000411 (with links to reply and counterreply). More recently, independent researchers have done work on the language and cast the "no recursion" claim into doubt:

 

http://www.nature.com/news/war-of-words-over-tribal-tongue-1.10595

http://ling.auf.net/lingbuzz/001095

http://www.letras.puc-rio.br/pag_profs/documentos/cilene/artigos/AbstractPirahaRodrigues&Sandalo.pdf

 

Hope this is helpful to you.

Edited May 18, 2013 by Stanstead

[EdEarl]

The video titled "A Conversation with Augusto and Yapohen Pirahã" is very interesting. It shows the difficulty of first contact with another society, and how different people involved in these situations have different impressions of the interactions. I cannot discern how to what degree the Pirahã answers to questions were politically motivated rather than factual.

 

I have no personal knowledge about the controversy between Chomsky and Everett. I am not an expert on the Chomsky Hierarchy, but I have worked on computer languages and believe it to be pragmatic. Whether people process language in this manner or not, I do not know.

 

I don't believe analysis of all the worlds languages will settle the issue, and I know of no proof that people do process language using recursion. That's my opinion, which is worth nothing.

Edited May 18, 2013 by EdEarl

[Popcorn Sutton]

I was under the impression that recursion is simply an endless "while" loop such as "while 0 == 0: listen and generate output".

 

Btw, there is a linguistics forum, which can be found at www.lingforum.com.

I've probably done well over 10,000 hours of research in linguistics and that is my impression. Basically, it may have at once been a category error, which was challenged, and is now being reduced to, if you're alive, then your running a recursive procedure.

[EdEarl]

Recursive and iterative loops are a bit different. While, do, for, perform, etc. are iterative. An example recursive loop follows:

 

factorial(n) = if (n==0) return(1) else return( factorial(n-1) * n);

 

I think that is correct, I hate to release code without testing it.

 

Recursion works because the function call (e.g., factorial) pushes its argument onto a stack, and return pops the stack. Thus, a recursive loop uses additional memory during each function call, and it is possible to exhaust available memory causing premature termination (an error).

 

Iterative loops require a counter, which has a limited capability to store the iteration value; a condition that also causes an error when the iteration value exceeds limits.

 

I am no expert, but AFAIK the brain does not use either iteration or recursion, except when doing an algorithm mentally. When following an iterative loop, most people are forced to keep track of an iterator and other data on paper or otherwise external to their brain. Similarly, mentally calculating a recursive algorithm generally requires careful book keeping.

Edited May 30, 2013 by EdEarl

[Popcorn Sutton]

I am not going to put your code to the test because in this context, I do not believe it has relevance (although in some sense I do, but I will leave that aside for now). I am familiar with the term "recursion" as it is used specifically within a linguistic context. Chomsky wants a recursive generative procedure. I translate this to the desire for a program to continually think and make sense while doing so. If this is achieved, as he states in his "60+ years of generative grammar" on YouTube, then the only problem is to add more memory. If you add more memory, the program will keep running. He explicitly states that "if you take a bunch of data and you throw a big computer at it and do a bunch of boring and totally irrelevant statistical analysis, something interesting will pop out" although he follows up by saying that the chances of that are slim to none.

 

I, personally, have been working on this problem computationally for over 5 years although I didn't find the motivation to start coding until approximately 2 years ago (all using Python 2.7). The reason was because I thought I had the answer (and still do), AFAIK it is all about focal points and pattern recognition. From what I currently have, I see that I am on the right track and I can do the calculations by hand to provide an output that the program would produce given a set of data.

 

A recursive procedure would look something like this (with about 100 or more lines of code between the input and the output)

 

While 0 == 0:

n = 0 + 1

print n

 

1

2

3

4

5

6

7

 

etc.

 

But as I have stressed on these forums and on the linguistics forums, to say that something is truly recursive is contingent. The reason is that it implies infinity, and, on the basis of observation, infinity is logically invalid.

Edited May 30, 2013 by Popcorn Sutton