Unformatted text preview: Study of Language
Study of Language
COMD 2050 Chapter 4: Animals and Human Chapter 4: Animals and Human language
Do animals really “understand” words like humans? Or do animals produce a particular behavior in response to a stimulus? Chimpanzees and language
Chimpanzees and language The chimpanzee has 99% of its basic genetics in common with humans.
Nonhuman primates do not have a physically structured vocal tract suitable for producing human sounds.
In early attempts to teach chimps human language, people raised chimps in their home. Washoe
Washoe Beatrix and Allen Gardner taught ASL to Washoe (raised like a human).
After 3.5 years, she used signs for 100+ words – airplane, baby, banana, window, woman, you.
Produced 23 word phrases – gimme tickle, more fruit, open food drink. All of which were requests.
Created her own signs (bib, water bird for swan).
Seemed capable of having basic conversations (questionanswer).
Understood many more signs than she could produce. Sarah
Sarah Ann and David Premack used a set of plastic shapes representing words to communicate.
Did not learn ASL, lived in a cage, trained with food rewards.
The shapes/words could be arranged in order to form sentences (vertical).
Symbols were arbitrary.
Could produce “Mary give chocolate Sarah.”
Could understand complex structures – if/then. Lana
Lana Duane Rumbaugh and Sue Savage at Yerkes Labs in Atlanta
Used an artificial language “Yerkish” a set of symbols on a large keyboard and computer. Lana would press the symbols on the keyboard to produce language (similar to logographic codes: wordwriting.
“please machine give water” – typical request form.
Did Lana really understand “please” as a human would v. working a machine? Nim Chimpsky
Nim Chimpsky Herb Terrace taught him ASL under controlled conditions, records, and video taped all sessions.
After 2 yrs, he knew many signs (words) and produced many 2 word combinations – more drink, give banana (requests).
Like Washoe, Nim seemed to be developing language similar to the way a young child would. Nim and Washoe: another look
Nim and Washoe: another look Longer utterances were repeats of simpler utterances, not expansions.
Rarely used ASL to initiate interactions, only responded to teachers signing.
Repeating signs used by teachers. Chimps produced prompted repetitions.
Seemed to be producing behavior (signing, arranging chips, pressing symbols) only to get rewards. Signing was not linguistic behavior. Hans, Buzz and Doris
Hans, Buzz and Doris Clever Hans used hoofbeats to answer math questions and to tap out the alphabet. People that were asking him questions were giving him visual cues that he was responding to.
When Doris was signaled by a light, she would tell Buzz across a wall how they could get a snack. However, Doris signaled Buzz even when he was not in the tank.
Behaviors were conditioned responses. The controversy
The controversy Explanation of language behavior in animals: cues provided by trainers and the conditioned response behavior of animals.
Gardners showed that Washoe could produce correct signs to identify objects in pictures without humans present. Nim was treated as a research animal and his researchers were not fluent in ASL. The controversy
The controversy Gardners stressed the fact that their chimps were raised inan environment that promoted play and interaction with fluent sign users.
Moja, Pili, Tatu, and Dar learned sign mainly from hearing people born to deaf parents.
They used sign language with each other and with Washoe, even with no humans present. Learned earlier than Washoe and faster.
Loulis (adopted by Washoe) developed 50+ signs without human training. Sherman, Austin and Kanzi
Sherman, Austin and Kanzi Sherman and Austin became the first two chimps to communicate with each other using Yerkish. (Yerkes Lab – Sue SavageRumbaugh).
While Matata was being trained to use Yerkish, she always had her nursing baby Kanzi with her. Matata was not successful, but Kanzi started spontaneously using the symbol system. He had only been exposed to the language, never directly taught, but he saw the language in active use. Kanzi
Kanzi Developed a large vocabulary – 250+ forms.
By 8 years of age, he could understand spoken English associated with Yerkish symbols at the level of a 2.5 year old human child.
He could ask for his favorite movies “Quest for Fire” about primitive humans and “Greystoke” about the legend of Tarzan. The barest rudiments
The barest rudiments Chimps could interact with other chimps and humans with a symbol system.
They could not perform on a level comparable to a human child at the same age – the chimps were always much older.
We don’t always have a crystal clear, completely objective and noncontroversial definition of what counts as using language.
Yule feels that chimps do have the barest rudiments of language. Questions
Questions Can chimps be taught to produce human speech sounds? What would be the problem?
In Sarah’s vocabulary, the color red was represented by a grey plastic shape. If Sarah could use this plastic shape to convey the meaning ‘red’, which property did her language have?
What did Terrace conclude was the reason that chimps use of sign language was not real language?
What was the key element in Kanzi’s language learning? ...
View Full Document
- Fall '08
- chimps, Great ape language, Washoe