Tuesday, November 6, 2007

Learning ASL: when you are neither American nor... human

An African chimpanzee, the first nonhuman ever believed to have acquire human language, has died recently at a Washington research institute. Her name was Washoe, and according to the article "An African chimpanzee and its language of signs," she knew and could reliably and dynamically used a vocabulary of about 250 American-Sign-Language words by the time she died.

Reading this article, I immediately wondered if Washoe truly understood and used the signs taught to her, or was simply taught through rewards and punishments to associate a sign and the behavior of signing that sign with specific contexts. In other words, did she truly learn how to express herself with language, or did she simply learn signs the way dogs learn a new trick? The Gardners--the family that first taught Washoe the use of sign language, definitely taught her the sign for "more" as a rewarded behavior. Noting that when she's tickled, Washoe brings her hands together in a similar way a human would when signing "more," the Gardners shaped this behavior into the correct sign for "more," rewarding Washoe each time she performs it correctly by playing with and tickling her. According to the Gardners, however, Washoe seems to learn not only the behavior for, but also the concept of "more." They claim she understood she could sign "more" "to get more of anything, including food, games, and books." Also, she seemed to have learned the sign for "toothbrush" not through any operant teachings, but only by observing others around her signing the sign for "toothbrush" whenever they hold one.

Another key word in the claim is "human language;" Washoe was the first nonhuman to learn "human language." What is the integral difference between human language and animal language? Here is what I found on the general characteristics of human language:
  • Arbitrariness: no relationship between a sound or sign and its meaning
  • Cultural transmission: language passes from one user to the next
  • Discreteness: discreet units (words, for example) are used in different combinations to create meaning.
  • Displacement: can be used to communicate ideas that are not in the "immediate vicinity"
  • Duality: language has and works well on both the "surface level" and the "semantic (meaningful) level"
  • Metalinguistics: language can be used to discuss language itself
  • Productivity: "a finite number of units can be used to create a very large - though often mistakenly described as 'infinite' - number of utterances"
It is obvious no animal language possesses all these complex qualities of human language. How are we to decide, however, that the portion of "human language" that Washoe learned is not a simple enough portion that it is only as complex as the advanced animal language that a primate like Washoe is already capable of. Washoe most definitely did not learn to use words as part of sentences, or to understand words on both the surface and the meaningful level. According to the Gardners, however, Washoe had "the ability to combine signs in novel and meaningful ways." For example, she"referred to her toilet as DIRTY GOOD and the refrigerator as OPEN FOOD DRINK, even though the scientists around her always called them POTTY CHAIR and COLD BOX." But even then, how are we to know for sure that chimpanzees like Washoe do not already possess these cognitive linguistic capabilities in their own communication? Instead of comparing Washoe's cognitive abilities with those of humans, then, I think we should only observe her only to understand the possible extents of animal cognition in itself.


An interesting fact: A researcher in the 1970s studied animal language with a chimpanzee named Nim Chimsky, and failed to replicate in a more classical experimental setting the results the Gardner achieved with Washoe. Linguist Noam Chomsky is also very skeptical of the Washoe's language skills, believing that there are no scientific support to the claim that she truly "spoke" human language.



Sources I researched for this article:
  1. Washoe - Wikipedia
  2. Animal Language - Wikipedia

1 comment:

Steve said...

nice post! i am curious about that set of human language features you listed. where does that list come from? who wrote it? do you think it might be based around a specific theory of language, or did the authors really investigate EVERY human language? can you think of a way in which a human language might fail to possess one of those attributes?