Transmission II: How much of language is cultural, how much biological? In particular: Can non-human primates learn language from humans? Where are we in the course? We have just looked at how culture is linked to biology in the area of growth and body size? What is the role of cuisine and diet and (we might add) life style in relationship to phenotypic unfolding of the genotype? These processes of transmission involve interaction between biology and culture. Now we want to look at studies of how much of language (and culture) can be transmitted across species. How much of the exaggerated prominence of social learning in the human species is a result of biological adaptation for such learning, how much is the accumulation of culture? What is at stake? If chimps or bonobos or gorillas can be taught human language, then language is cultural, i.e., humans are NOT uniquely biologically programmed for language. If non-human primates cannot be taught language, then some part of language must be dependent on the specific biological endowment of human beings, an ability to learn language (and other aspects of culture). Outline 1. Can chimps be taught to speak English? 2. What are the principles of inference through which communication takes place in the animal world? 3. What are the distinctive properties of human language as a means of communication? 1. Can chimps (or other non-humans) be taught to speak English? Recall that chimps do appear to have rudimentary culture. And to review: • the human-type family with its role of husband/father is not essential to culture. • some transmission can be achieved through the mother-child bond without the husband-father role. However, it is one thing to ask whether rudimentary culture could be achieved without human-type families. Another to ask whether complex culture (including language?) would be possible without human-type families. We’ll see that the more successful attempts to teach language to non-primates typically involved placing the animals into human family situations. Preliminary observation: Some aspects of language are obviously cultural, that is, they are socially learned. 6-7,000 languages on the planet (6,059-7,300) A child brought up among speakers of any one of those languages will learn that language and no other; The child will speak that language without an accent, and speak it fluently, barring physical impairments; Languages differ considerably in their lexicons, sound patterns, word morphology, and syntax, as well as in patterns of discourse. Historical overview of attempts to teach non-human primates to speak English: 1909 Clinical evaluation of a trained chimp ("Peter") by L. Witmer; claim that the chimp had learned to say "mama." 1916 W.H. Furness claims to have taught an Orangutan to say the words "papa" and "cup." 1930s Research by Louise and Winthrop Kellog raised a baby orangutan (named Gua) together with their son Donald (then 9 1/2 months old). When Donald was 12 1/2 months old, he could comprehend 20 verbal requests; Gua evinced comprehension of 21. Thereafter, Donald began to outstrip Gua. 1940s Keith and Cathy Hayes attempted to teach English to a chimp named Vicki. [Popular movies based on these experiments, e.g., "Bedtime for Bonzo" (starring Ronald Reagan)] Bedtime for Bonzo (1951) Reagan plays a psychology professor who treats a chimp, Bonzo, as his child for a nature vs. nurture experiment. Bedtime for Bonzo is a silly, cute film. The movie has been cited as Reagan's most ridiculous career move, but the film displays his comedic ability. http://www.usatoday.com/life/columns/video/vid10.htm QuickTime™ and a Cinepak decompressor are needed to see this picture. Observations Vicki’s speech: 1. Chimp is able to learn some fragment of human speech; evidence here for transmission of culture from humans to chimps, however fragmentary. 2. Chimp produces recognizable words "papa," "mama," and "cup," although the sounds she produces are distinct from their English counterparts: English [p] [m] [k] [a] Vicki bilabial click voiceless [m] [x], i.e., velar fricative voiceless [a] Observations Vicki’s speech: 3. Vicki produces the words in response to verbal cues. 4. However, she never learns to speak like a human. Question: Why can't chimps learn to speak English? One popular answer: their vocal tract anatomy is different from that of humans (Philip Lieberman, Jeffrey Laitman). [Note: I don't think that is probably the principal reason]. QuickTime™ and a decompressor are needed to see this picture. QuickTime™ and a H.263 decompressor are needed to see this picture. Alex the Parrot (research of Irene Pepperberg, U of Arizona): • Capable of distinguishing fine grained-sound differences, for example, "key" versus "pea" = [ki] versus [pi] • Capable of extremely clear pronunciation of words and phrases, including a clear differentiation of [ki] from [pi]. • These distinctions are like those that form the basis of what linguists call "phonemic" (as opposed to "phonetic") contrasts. • Alex is learning all of this from his human teachers; hence, this is further evidence for the transmission of a porition of language (words and phrases) to a nonhuman species. 2. What are the principles of inference through which communication takes place in the animal world? Associationalist psychology of the 19th century proposed two: 1. Contiguity: you can reason from something (a sign) to something else (the meaning) by asking with what else that something is or tends to connected physically (that is, through space-time co-presence). For example, a smell might be associated with a particular animal; a vocal call might be associated with the kind of animal that produces it). The other is: 2. Similarity: you can reason from something (a sign) to something else (the meaning) by asking what else that something resembles in its physical appearance. (photograph might resemble the thing photographed; one situation might remind you of another; one individual (for example, of a species) might resemble another). Semiotics = the study of signs in relationship to their use. index = a sign whose meaning is inferred through contiguity icon = a sign whose meaning is inferred through similarity Much of human linguistic usage is indexical or iconic "how are you" ------------------> "Fine, and you" "what's that?" + -------------------> ”a cup" cup visible Much of human communication with animals is through indexes (and icons). "how many?" + ---------------------> "two" two objects visible "what color bigger?" + -----------------------> "yellow" larger of two objects is yellow Conlusion: This is not to belittle the remarkable findings of Irene Pepperberg in her research with the African Gray Parrot, Alex, or that of other researchers with non-human primates. One anthropologist who studies language as it is actually used (Michael Silverstein - U Chicago) has suggested that the vast majority of human communication is through indexical and iconic signaling. Research has demonstrated that a considerable amount of human culture can be passed to non-human species. But the question remains of whether they are actually acquiring the kind of language that all human cultures have. 3. What are the distinctive properties of human language as a means of communication? The ability to generate new signs that another individual has not heard (or seen) and to have that individual be able to correctly interpret those signs. The ability to correctly interpret a sign that you have not before heard (or seen). What makes this possible? Answer = the combinatorial character of human languages morphology = the combination (in accord with rules) of elementary meaningful units (called morphemes) into words. re + work + er "again" "work" + "one who" "one works something over again." Some languages (Eskimo languages, Chinook, Kuna, etc.) rely heavily on morphology; others makes little use of it (Chinese); English is not terribly big on morphology. What makes this possible? Answer = the combinatorial character of human languages syntax = the combination (in accord with rules) ofwords into sentences. the + cat + sat + on + the + mat "The cat sat on the mat." Beyond teaching spoken English to non-human primates - I 1960s Beatrice and Allen Gardner attempted to teach sign language (AMSLAN = American Sign Language) instead of spoken English to the chimp Washoe; Washoe learned more than 150 signs; claim that she could use some novel sign combinations. Her adopted son Loulis is said to have acquired 50 signs from her, without human intervention. Beyond teaching spoken English to non-human primates - II 1960s David Premack with 6 year old chimp Sarah -demonstrated that she had ability (with colored plastic tokens instead of speech) to understand and use some grammar: pluralization, "and", "if...then" constructions. 1970s+ Francine Penney Patterson with the gorilla Koko (and later Michael); claim that Koko uses 1,000 plus gestural signs. An Internet Chat with Koko the Gorilla This is the transcript of the AOL Live Internet Chat with Koko the gorilla. Which took place on April 27th, 1998, at 7pm EDT. In honor of the month long celebration of Earth Day. This event was the first ever live inter-species internet chat with Koko the gorilla. Brought to you courtesy of H.E.A.V.E.N. (Helping Educate, Activate, Volunteer and Empower via the Net), The Envirolink Network, The Gorilla Foundation, and AOL Live. These pictures are of Koko's life, from the time she was baby being raised by Dr. Penny Patterson, to the years growing up at the Gorilla Foundation. In the past 26 years, she has raised kittens on her own, she has learned American Sign Language...with a working vocabulary of 500 words. Today, she has her own computer. The Gorilla Foundation was founded in 1976, to promote the protection and preservation of gorillas. Project Koko, which began in 1972, is the primary focus of the foundation. HaloMyBaby is the moderator of the chat on AOL, DrPPatrsn is Koko's friend and trainer, and LiveKOKO is Koko the gorilla. Scroll down to read the full transcript... ***************************************************** http://www.geocities.com/RainForest/Vines/4451/KokoLiveChat.html Welcome, Dr. Patterson and Koko, we're so happy you're here! DrPPatrsn: HaloMyBaby: LiveKOKO: DrPPatrsn: HaloMyBaby: LiveKOKO: DrPPatrsn: LiveKOKO: HaloMyBaby: DrPPatrsn: LiveKOKO: DrPPatrsn: LiveKOKO: HaloMyBaby: DrPPatrsn: http://www.geocities.com/RainForest/Vines/4451/KokoLiveChat.html You're welcome! Is Koko aware that she's chatting with thousands of people now? Good here. Koko is aware. I’ll start taking questions from the audience now, our first question is: MInyKitty asks, Koko are you going to have a baby in the future? Pink We've had earlier discussion about colors today Listen, Koko loves eat Me too! What about a baby? She's thinking... Unattention She covered her face with her hands....which means it's not happening, basically, or it hasn't happened yet. I don't see it. That's sad! In other words, she hasn't had one yet, and she doesn't see it happening. She needs several females and one male to have a family. In our setting it really isn't possible for her to have a baby. Beyond teaching spoken English to non-human primates - III 1970s Herbert Terrace, work with chimp named Nim Chimsky; became a critic of ape language research; claimed that Nim could not use syntax. Responded only to indexical cues. 1980s+ Sue Savage-Rumbaugh with the bonobo Kanzi; used computer keyboard. Claimed that, while Kanzi first used indexical cues, later developed true language. Could respond accurately 70% of the time to a novel sentence communicated to him. Conclusion I: Ideological positions of researchers and scholars Can non-human primates learn language from humans? YES, we have demonstrated that bonobos, chimps, and gorillas are capable of learning human language; they have, in the process, passed the threshold into humanness. NO, the research to date does not show the ability of chimps to acquire morphology or syntax in any consistent way. Much of the research can be explained through indexicality and iconicity, or through the "clever Hans" phenomenon (Thomas Sebeok). Conclusion II: Realities of research to date Can non-human primates learn language from humans? SORT OF, bonobos, chimps, and gorillas show evidence of being able to acquire through social learning many (500 +) individual signs and to pass some of those signs onto their offspring (hence, they have some culture). They also show some ability to produce and interpret novel sign combinations. Hence, they have something like an incipient or protolanguage ability. NOT REALLY, the research to date on non-human primates is highly suggestive. Apes clearly have the ability to acquire culturally constituted signs in some quantity, but their ability to produce and interpret novel sign combinations has not been fully establish. The apes make too many errors in syntax tests to have confidence that they have made the transition to language. QuickTime™ and a H.263 decompressor are needed to see this picture.