Apr 17 2007
Language, Structure, Information, Knowledge
All of sudden, I have heard so much about an ongoing academic debate: a linguistic professor is challenging whether Chomsky’s tree is really that universal in all languages.
I first heard of Dan Everett on a Sunday morning when I forgot to turn off the alarm clock set for the day before. NPR’s weekend edition has a nice piece of how an Amazonian tribe “converted” a missionary into an atheist professor and inspired him to challenge Chomsky.
I was a dirt poor immigrant student struggling with libido, money, grades and finding a job when I first learned the Chomsky’s theory in my Compiler class. I almost failed the class. It was not until all those temporal issues were settled did I realize what I had missed: one day I hit an epiphany and became a convert. I even bought his political arguments (the return on that was much lower though).
Hearing someone challenging Chomsky was enough to keep me awake on a gray Seattle weekend morning. I did some research on the topic but couldn’t understand all the original documents flying back and forth among Everett and his challengers.
A day later came the Apr 17th issue of the New Yorker that has a featured article (c) about this very topic. It made things much easier for me to understand but I guess I am still missing the part I don’t understand.
As described by the NYKR article, the central question is whether recursion is part of the “universal grammer” that is innate to human beings. Chomsky thinks that vocal communication is not unique to human, but recursion is. It seems so far, all the languages in history has bore him out until Dan Everett met the Piraha tribe in the Amazon jungle.
What Everett, a gifted linguistic expert and a former Evangelical missionary, found is that:
1. The Piraha tribe is a capable people: they have a history, a body of knowledge that enable them to survive in nature and, most importantly, they seem to have fostered a tradition or a culture–not in the forms of dances or rituals but in how they think. For example, they don’t trust outsiders and are resistant to outside influences.
2. Yet their language seems to be extremely primitive: it does not have words for directions or numbers.
The whole affair is enormously fascinating. I have so many questions that I don’t even know where to start. For example, Everett seems to suggest that the Pirahas do not use, nor do they understand, abstraction: they have to deal with something concrete and immediate. For example, a boy can build a model airplane but only when there is a real one in front of him. Instead of using directional words, they use physical landmarks (e.g. turn to the river). What I don’t understand is, if this is true, how can they pass knowledge around? How does a father tell his son how to evade a jaguar? Does he have to do so in front of a live one?
Other questions include: is Everett’s discovery really enough to unset Chomsky’s universe? Or maybe the Piraha case is an exception of the close coupling of culture and language?
More importantly, is there a parallel between language structure and knowledge structure? Theoretically speaking, if a body of knowledge can be described by a langugae, the knowledge should conform to the Chomsky normal form as well. For example, let’s say how to work with a spreadsheet is a body of knowledge, then there a tree-like form that encompasses all the necessary information (e.g. feature list) and sub-knowledge (e.g. what is a spreadsheet). Interestingly enough, a business guru, Barbara Minto, advices businessmen to use a “pyramid structure” to carry out effective communications (i.e. to convey/exchange a body of knowledge).
This kind of link back to what I have always thought as the “meta-concept” in a language. To put it simply, a meta-concept is an abstraction of a collection of lower level concepts until one hits physical senses. For example, the colors red, green, blue are concepts derived from physical senses and the concept/word “color” is an abstraction of all visible colors. Color, texture and rigidity may be further abstracted as the concept of “attribute” (of an object)… and so on.
My argument is, the more high level meta-concepts a language espouses, the more powerful the language is (e.g. more capable of delivering complex concepts). This is perhaps too self-explanatory to make a difference. However, that is exactly why the Piraha case is so strange. Something is not quite right here: either humans can survive with limited language ability or the whole language-knowledge relationship is wrong: e.g. the Pirahas can use limited vocabulary and recursion-less structure to express complex meanings. Either way, it doesn’t make a whole lot sense.
Read a pretty interesting article:
Or competing chimps would solicit support from others to upstage a rival.
, is politicking still governed by ethics or morality? If someone (I mean humans here) still insists he lives by his principle, where does this principle come from? In other words, what is the justification of this principle? Of course, this discussion only makes sense in a rational and secular setting. One can always invoke religion and the divine, but he probably doesn’t even consider himself an animal anyway.
Isn’t this stuff fun??
I am implicitly agreeing with Karl Marx and his materialism: politics is built on a material infrastructure. Except that he deemed this infrastructure is only the modus of production at a given historical moment. What he missed, I posit, is Freud’s discovery of physiology of human consciousness. In other words, Marx, without the benefit of knowledge of psychology, didn’t realize that the so-called subjective matters–history, society, culture, tradition, morality–are just as material as machines and capital.