sign in sign up
<<
Home , Ai (chimpanzee), Chantek, Chimpanzee, Kanzi, Koko (gorilla), Lana (chimpanzee)

Current Comments” EUGENE GARFIELD INSTITUTE FOR SCIENTIFIC INFORMATION* 3501 MARKET ST PHILADELPHIA, PA 19104

When the Speak, Linguists Lfsten. Part 1. The Studies

Number 31 August 5, 1985 “Come open, “ “Hurry drink milk,” Tokyo, noted that apes do not lack the “You chase me,” and “Please machine needed for language. Rath- give milk” are all parts of conversations er, the most significant difference is be- between and apes. Since the late tween the and vocal 1960s psychologists and anthropologists tracts.z Philip Lieberman, Department have launched intensive projects to of , Brown University, Provi- teach great apes—such as the chimpan- dence, Rhode Island, reported differ- zee, , and —a form of ences in the structure of the oral cavity that is comparable to and less tongue mobility in human language. The apes are large tail- compared to humans. s In the chimpan- less that are classified just be- zee, the tongue is longer and narrower low humans in the evolutionary tree. Re- than the human tongue. The placement search on “ linguistics” and the of the chimpanzee’s tongue and its ma- different approaches to it have resulted neuverability within the oral cavity pre- in debate among the ape language re- vent the chimpanzee from producing the searchers and among linguists as to the full range of sounds necessary for definition and uniqueness of human lan- human , guage. This first part of a twopart essay Of the early efforts to teach spoken focuses on the ape language projects. language to a chimpanzee, the project Part 2 will discuss the controversies sur- carried out in the 1950s by Keith J. rounding them. Hayes and Catherine Hayes, then of the The earliest language experiments Yerkes Laboratory of Biology, were attempts to teach Orange Park, , was the only one to young home-raised apes. These at- to achieve even an approximation of oral tempts failed for many reasons. Accord- language. The Hayeses raised their ing to Winthrop N. Keflogg, professor chimpanzee, Vicki, in a home environ- emeritus of experimental psychology, ment similar to the setting in which hu- Florida State University, Tallahassee, man parents raise their chfldren. How- when the home-raised chimpanzee is ex- ever, Vicki never articulated more than posed to the same linguistic and social three or four words—’’Mama,” “Papa,” environment as a human child, the ape “Cup,” and something that sounded like displays little evidence of the vocal imi- “Up.” And these were learned with great tation common in children. 1 The long difficulty, over a six- period, with period of and prattling that hu- the Hayeses shaping Vicki’s mouth in man babies undergo seems to be a pre- some instances so she could articulate requisite for subsequent articulation of the sounds.4 words. The ape language studies and the en- Fred C.C. Peng, professor of linguis- suing controversies really began in 1969 tics, International Christian University, when R. Alen Gardner and Beatrix T.

293 Gardner, Department of Psychology, fore a box that the trainer opened to University of Nevada, Reno, introduced reveal the picture of an object projected —the first language-trained on a screen. The trainer asked Washoe chimpanzee-to the worlds In 1966 the the name of the object, and two observ- Gardners acquired Washoe, a 10 ers who did not know which object was month-old female chimpanzee born in shown on the screen recorded her re- the wild. Based on their observations of sponse, The Gardners reported about 90 spontaneous used by chimpan- percent between the observ- zees and of the proclivity chimpanzees ers,9 and that Washoe identified 70 to 80 show for , the Gardners decid- percent of the items correctly. 10Howev- ed to teach Washoe a gestural form of er, even when Washoe made an error, language, American Language she generally identified the item as (ASL), rather than spoken language. something in the same category as the ASL is a that is used by pictorial representation. For example, deaf people in . It con- an animal picture was often identified as sists of a set of manual configurations another animal. and gestures that correspond to particu- These vocabulary tests served several lar words or concepts and has its own purposes. One was to show that the grammatical structure.b chimpanzee can use sign language to Although Washoe lived in a trailer in communicate information. Another ob- the Gardners’ backyard, she was raised jective was to show that the chimpanzee in a social environment “similar” to that could use to refer to natural-lan- of a human child except that ASL was guage categories. For example, they the only means of communication. No wanted to show that the chimpanzee spoken language was allowed in Wa- could use the ASL sign for “” to iden- shoe’s presence. ASL training took place tify pictures of . According to the in the context of Washoe’s daily activi- Gardners both objectives were accom- ties. Assisted by Roger S. Fouts, then a plished with Washoe and with other graduate student in their laboratory, and chimpanzees in their later studies. 10 other assistants, the Gardners taught Ten months after the project began, Washoe sign language by molding her when Washoe knew approximately eight into the proper ASL shape, a signs, she began to combine two or more method they found more practical than signs. Washoe was then between 18 and waiting for Washoe to spontaneously 24 months old, which is close to the age produce the sign made by the trainer.T when children begin to use two-word At the same time, Washoe was shown combinations. These combinations al- the actual item represented by the ASL lowed her to ampliiy the meaning of the word. They repeated this procedure un- single-sign utterances. As an example, til Washoe made the sign spontaneously. the Gardners noted that Washoe signed Later, she learned signs by observing “Listen dog” when she heard an unseen and imitating. As a criterion for deter- dog barkings mining that a sign was indeed a part of The Gardners maintained diaries of Washoe’s vocabuhq, the Gardners re- Washoe’s use of language to estabfish a quired that she use it appropriately and record of the signs, the context of their spontaneously each day for 15 consecu- use, and interchanges between the train- tive days. After 14 months of training, ers and the chimpanzee. These record- Washoe knew and used 13 signs. By the ings, made by a skilled observer tran- end of the third year, Washoe’s vocabu- scribing Washoe’s signs into English lary had increased to 85 signs.g words, also provided a diary of her use of To test Washoe’s vocabulary, the signs in combinations so they could be Gardners devised a double-blind exper- compared with records of human chil- iment in wh]ch Washoe was seated be- dren.

294 The Gardners reported that the the Gardners reported that at this young phrases Washoe used were appropriate age, the chimpanzees used signs with to her referent. For example, Washoe variation in form and in appropriate referred to soft drink as “Sweet drink. ” variations in a basic context. 10.14 An The Gardners noted that she used sign example is Pili’s use of the sign “More” language for requests, and she answered when his water bottle was taken away or questions about objects and events. And after the trainer stopped playing with Washoe initiated many of these conver- him. sations with questions and statements of These chimpanzees, like Washoe, her own. At the end of the S1-month pe- were able to extend the use of signs they riod during which the Gardners worked learned for a few items, such as flower or with Washoe, the chimpanzee correctly cat, to include all flowers and cats. The used a total of 132 signs. 11 chimpanzees seemed to divide the world The Gardners’ work with Washoe into conceptual categories just as hu- ended in 1970. However, thk was not the mans do. For example, the Gardners re- end of Washoe’s language training or of ported that , on seeing a large cut the Gardners’ involvement with signing of fresh beef on the kitchen table, signed apes. Washoe was moved to the primate “Meat.” Moja had not been taught to in- center at the University of Oklahoma in clude both fresh and cooked meat in the Norman, where Fouts continued her meat category. The transfer from language training. The Gardners’ work cooked to raw was spontaneous. 10 with Washoe demonstrated that ASL The Gardners were interested in the can be a suitable means of commu- ability of the chimpanzee to respond to nication for the great apes and that questions, because there is extensive chimpanzees are capable of performing data on this developmental aspect of and comprehending certain language child language. Children generally are operations. able to respond to “What,” “Where,” In the next several , the Gard- and “Who” questions by the age of 21 ners extended their work to include four months. With the new group of young other chimpanzees-Moja, Pili, Tatu, chimpanzees, this aspect of language de- and Dar. To make a more valid compari- velopment could be studied. The Gard- son between in ners found that Pili and Tatu responded children and chimpanzees, language with the appropriate sign to questions training for these began in early containing “What” and “Where” at 18 infancy. One significant difference be- months of age. Between ages two and tween this second project and Project three years, Moja and Tatu responded Washoe is that the ASI., teachers for this with proper names and pronouns to group were all fluent ASL signers. Both “Who” questions. 12 Moja and Pili began to make recogniz- Ape language researchers wondered able ASL signs when they were about whether a young chimpanzee could three months old. 12 Although thk may learn ASL from its mother, as human seem early for language development, children learn language from their moth- parental reports in~lcate that it is not ers. At Central Washington University, dramatically ddferent from the age at Ellensburg, Fouts and colleagues stud- which deaf children produce their first ied the transfer of ASL skills between signs. Hilde S. Schlesinger, Center on the Gardners’ chimpanzee Washoe and Deafness, University of California, San Washoe’s adopted child, . When Francisco, and Kathryn P. Meadow, this project began, Washoe was 15 years Gallaudet Research Institute, Gallaudet old and her vocabulary consisted of 180 College, Washington, DC, cited a paren- signs. After eight days with Washoe, tal report of a deaf infant signing for milk Loulis was observed to produce his first at around five months of age. 13 In fact, ASL sign. The sign was the name for one

295 of hk human caretakers. Fouts and col- chimpanzees are better able to solve cer- leagues expected Loulis to acquire ASL tain types of problems than his chimpan- from Washoe by imitation. However, zees who have not received language they report evidence that she actively training. In a 1983 article in l?eha r+ taught Loulis the sign for “Eat/f ood.” iom[ and Bmin Sciences, a journal “Washoe stopped signing and took that also publishes peer review com- Loulis’s in hers and molded it into mentary, Premack noted that language the food sign configuration and touched training enhances the chimpanzee’s abil- it to his mouth several times.” The fact ity to solve problems involving abstract that none of the humans had used ASL judgment, such as analogical reasoning. in Loulis’s presence supports the argu- Spatial abilities, such as the of ment that Washoe was his only teacher. items in space, are not improved by lan- At a little over two years of age, Loulis guage training. 17 had a vocabulary of 17 signs. 15 In 1972, a chimpanzee named , at Around the same time the Gardners the Yerkes Regional Primate Research began working with Washoe, another Center, , Georgia, learned to ape language project was under way at communicate by using a computer the Psychology Department, University and a “language” her trainers called of California, Santa Barbara. David Pre- “.” The original Yerkes Primate mack, now at the University of Pennsyl- Center, located in Orange Park, Florida, vania, Philadelphia, taught his chimpan- was founded by comparative psycholo- zee, , a language based on plastic gist in 1929; in 1961, it shapes symbolic of words. lb Each became part of the National Institutes of “word” had a unique color and shape. Health Primate Research Centers Pro- For example, an apple was represented gram and moved to in by a small blue triangle. These plastic Atlanta. At Yerkes, Lana’s specially de- shapes were backed with metal and used signed computer keyboard, mounted on on a magnetic slate. Premack equated the wall of her enclosure, initiaUy con- the act of placing the plastic word on the tained 25 word keys. Each key was board with . marked with a unique symbol to distin- Premack began working with Sarah guish it from the others and when Lana when she was over 5 years old. Her train- pressed the key it lit up to indicate that it ing required that she learn to associate had been pressed. All of Lana’s language the plastic symbol with the object it rep- interactions were recorded by the com- resented. Although Premack reports puter. The trainers’ responses to Lana that Sarah was able to do this with an ac- appeared on a separate “receive” display curacy of over 80 percent, he does not situated above her keyboard. indicate the length of training required Psychologists Duane M. Rumbaugh to achieve this goal. The initial vocabu- and Timothy V. Gill, Georgia State Uni- lary of object names was extended to in- versity, Atlanta, along with my dear clude words for action. Sarah ultimately friend Ernst C. von Glasersfeld, Depart- learned to arrange the shapes on the ment of Psychology, University of board to create sentences. She also Georgia, Athens, found that after six learned to follow rules of word order in months of training, Lana was able to her sentence construction that suggest- read the beginnings of familiar sentences ed a mastery of elementary .lb and appropriately complete them, 18 Today, Premack’s work with chimp- Like Sarah, Lana was also required to anzees no longer focuses on language, follow basic rules of Yerkish . 19 but rather on general cognition. Since She communicated by selecting the ap- language is not only a for communi- propriate words and pressing the keys cation, but also a tool for thinking, Pre- serially as in a sentence, in appropriate mack believes his language-trained word order. Lana used her computer to

296 request food and answer simple ques- panzee. Unliie common chimpanzees tions. 18 A little over two and one-half that require extensive training to acquire years after the computer language proj- language skills, their pygmy chmpan- ect began, Lana had a vocabulary of 80 zee, , learned to use Yerkish with words.m no diiect instruction.zs As an infant, Rumbaugh and his graduate student, Kanzi played in the laboratory whiIe the Gill, observed that, once Lana under- researchers taught his mother the stood the basic concept of naming, she Yerkish system of symbolic communica- began to spontaneously request the tion. At two and one-half years of age names of objects for which she had no Kanzi spontaneously began using several name. Or, like Washoe with ASL, she in- symbols correctly. Six months later he vented her own names by combining the was using nearly 30 symbols appropriate- symbols on her keyboard .21 ly and demonstrated an abdity to under- The LANA Project was extended to stand their symbolic meaning. However, study language-like communication be- the most profound difference the re- tween two young chimpanzees using searchers found between the two chim- Yerkish. E. Sue Savage-Rumbaugh and panzee species was Kanzi’s ability to un- colleagues, also of the Yerkes Center in derstand spoken English, a skill that Atlanta, taught two young chimpanzees, seems to elude Sherman and Austin.zs Sherman and Austin, the meaning of Another study, using a computerized computer-key symbofs for individual method similar to that used in the food items and to request the food by Yerkes projects, examined mathemati- pressing the appropriate keys.zz The two cal abilities in the chimpanzee. Tetsuro chimpanzees spontaneously requested Matsuzawa, Department of Psychology, and received specific food items from Primate Research Institute, Kyoto Uni- one another, indicating that they were versity, Inuyama, Japan, found that the able to exchange information by using chimpanzee was able to identify arabic the computer symbols. numbers by selecting the on the An important question in ape lan- keyboard corresponding to the number guage studies is whether the animals un- of objects displayed. The chimpanzee derstand the meaning of their symbolic was afso able to correctly identify the ob- communication or whether they learn ject’s color and category label, such as contextually appropriate usage. Savage- “Dog” for alf dogs.zb Rumbaugh tested this by putting f~d in The Gardners’ ASL studies with containers that required to be Washoe have been extended to include opened. One animal had the tool kit and other great apes. For her doctoral dis- the other the food container.~.zd In sertation, developmental psychologist order to gain access to the food, the Francine G. Patterson, then of Stanford chimpanzee had to correctly identify the University, began working with the young necessary tool and request it from the female gorilla . Patterson showed other animal. Sherman and Austin were that chimpanzees are not unique in their able to do this with an accuracy of about ability to use a human form of communi- 90 percent.zd According to Savage- cation. Koko was one year old when Pat- Rumbaugh and colleagues, these experi- terson began ASL trainhg, using tech- ments demonstrate that the apes are niques similar to those described by the learning the basics of language and not Gardners.2y,28 Patterson established two just responding to stimuli; that is, the criteria for a sign to be considered part chnpanzees have a mental image of the of Koko’s vocabulary: a) it must be re- object represented by the symbol.~ corded by two independent observers, Savage-Rumbaugh and colleagues and b) it must be used spontaneously found large dtiferences in the communi- and appropriately on at least half the cative abilities of two species of chim- days of a given month. z? After three

297 years of training, Koko’s vocabulary orangutan is an interesting subject for consisted of 127 signs.m (p. 87) When language training since anatomic evi- she was between five and six, Koko’s dence ind~cates that the orangutan vocabulary consisted of 246 words. 29 is more asymmetric, and therefore more Like the ASL-trained chimpanzees, like the , than the of Koko also used signs in combinations. the other great apes. Marjorie LeMay, For example, Patterson reported that at Department of Radiology, Massachu- about 14 months of age, the gorilla setts General Hospital, Boston, and the signed “Gimme food’ in response to a late Norman Geschwind, Harvard Neu- drink held out of her reach.zT In a Na- rological Unit, Boston City Hospital, tiona[ Geographic article, Patterson studied this problem. They found that noted that, compared to the ASL- indirect measures of the planum tempo- trained chimpanzees, the gorilla is rale, a left-hemisphere language area in cahner and more deliberate. “Koko the human brain, indicate that the left seems to resort to the sign language side is significantly larger than the corre- more often to express herself, and she sponding right-hemisphere area in the discusses a wider range of activities.”~ orangutan. Although there is a degree of One of the hallmarks of human lan- asymmetry in the chimpanzee and goril- guage is displacement, or the ability to la, it is most pronounced in the orang- refer to past emotional states or utan.ss A larger left hemisphere in the events.~ Patterson claimed that Koko is human brain is associated with left- able to refer to past experiences without hemisphere specialization for language. actually experiencing them at the mo- Miles reported that first ment. As an example, Patterson queried signed after one month of training, and the gorilla about a biting incident that that his sign-language acquisition was had occurred the day before. When comparable with that of Koko and asked, “What did you do to Penny?” Washoe. Chantek’s first sign was fol- Koko responded, “Bite.”m This ability lowed by a new, additional sign about has also been found in the chlmpan- every 25 days. He spontaneously pro- zee.lb duced two-sign combinations in the sec- Patterson also has trained a male go- ond month of training, and three-sign rilla named in ASL. Koko and combinations in the tenth month. Miles Michael are able to communicate with suggested that Chantek initiates commu- each other in sign language. However, nication with hk trainers at a higher rate Patterson noted that Koko signs to than that reported for chimpanzees.sl In Michael more frequently than Michael 1983 after three and one-half years of signs to her. In fact, there is evidence training with criteria for sign acquisition that Michael learned some of hk sign similar to those of Patterson, Miles re- language from Koko.zs (p. 175) Re- ported that Chantek acquired a vocabu- searchers at the Gorifla Foundation, lary of 56 signs. Today Chantek has a V* W ootilde, California, founded by Pat- cabulary of 125 signs.3z Miles noted that terson in 1977 to support her work, hope production of ASL signs by the orang- that eventually Koko and Michael will utan is slower and more deliberate than produce a young gorilla and teach it ASL that of the chimpanzee. sl Another inter- as Washoe did with Loulis. esting aspect of thk research is the ob- Another ape language project is being served propensity of the orangutan to carried out by anthropologist H. Lyn produce ASL signs with h;s feet. In hk Miles, University of Tennessee, Chatta- book on orangutan behavior, Terry L. nooga, with a seven and one-half year- Maple, School of Psychology, Georgia old orangutan named Chantek.sl ,32The Institute of , Atlanta, sug-

298 gested that thk may be a reflection of the species.” It did not carry through to any arboreal, or tree-dwelling, adaptations 1984 research fronts. The four core pa- of the orangutan.~ pers that identified thk research front These ape language studies provide were not heavily co-cited in 1984. These information on linguistic behavior as a include the original ape language paper measure of intelligence and provide a by the Gardners,s Rumbaugh’s bookzl greater understanding of the continuity on the LANA Project, the book by Grif- between human and nonhuman intelli- fin,ss and a paper by psychologists Herb gence. However, ape language research S. Terrace, , New raises more questions than it answers. York, L.A. Petitto, Harvard University, The primary question concerning lan- Cambridge, and R.J. Saunders and T.G. guage-trained apes is whether they use Bever, also at Columbia. language in the human sense. Another The paper by Terrace and colleagues question focuses on the ape’s awareness describes Project Nim, an ape language of the meanings of the words it uses. In study. From thk study, the authors con- 1976 Donald R. Griffin, Rockefeller cluded that, while apes learn words, they University, New York, addressed the is- do not produce new and original sen- sue of animal awareness in hk book The tences.~ This research convinced the Question of A nima[Awareness.35 In ad- Columbia group that the great apes do dition, ape language studies have stimu- not use language in the human sense. lated linguistic research into the essence Others also question the claims made by of human language and the differences ape language researchers regarding the between it and the natural communica- linguistic abilities of their research sub- tion system of the great apes. jects. These criticisms, which raise ques- Original research on ape language tions about the methodology and inter- studies is limited by the availability of re- pretation of the ape language studies, search subjects and the special facilities will be discussed in the next part of this required to house them. Consequently, essay. the number of original research papers I would like to make a special point in published on thm subject is limited. thanking the numerous scientists who Twenty-one articles, published between took the time to referee this essay. The 1976 and 1984, were retrieved from the of their criticisms, however, is Arts & Humanities Citation IndexTM perhaps the best indicator that animal (A& HCITM). SciSearch@ identified 43 linguistics is as yet a soft science. Nor- articles published on ape or primate lan- man W. Storer, then of the Social Sci- guage studies from 1974 to 1984. Social ence Research Council, New York, used Sci.Search@ identified another 122 arti- the terms soft and hard science to differ- cles on ape language studies published entiate the “social” from the natural from 1972 to 1984. While these two data- sciences.37 Perhaps the level of contro- bases reflect the somewhat arbitrary sep- versy in a field is one of the best indica- aration of the Science Citation Index” tors of softness. (SCP ) and the Social Sciences Citation Index@ (SSCF ), thk is not the case in our research fronts. They are derived ● ☛☛☛☛☛ from a combined SCZ/SSCI file. Based on citation analysis, only one distinct research front (#83-6410) deal- ing with ape language experiments was My thanks to Linda LaRue for her found—’’Animal awareness, cognition help in the preparation of this essay. and language in chimpanzees and other 019851s1

299 “Now that we’ve got this wrapped up, I’d like to get into math.”

REFERENCES

Kefbgg W N. Communication and language in the home-raised chimpanzee. Science 162:4237, 1968. 2. Peng F C C. Lmguiatic potentials of nonhuman primates. (Peng F C C, cd.) Sign fanguage and larwua~e acauisifion in man and ape. Boulder, CO: Westview Press, 1978. u. 20324, 3. Lkhe=ari P. Pfiate vocalizations and human linguistic abtity. J. Acoust. .%.. Amer. 44: 1574-S4, 196S. 4. Hayea K J & Hayes C. The intellectual development of a home-raised chimpanzee. Prwc. Amer. Phil. Sot. 95:105-9, 1951, 5. Gardner R A & Gardner B T. Teaching sign farrguage to a chiniparrzee, Science 165:664-72, 1%9, 6. Kffnm E S & Befhrgl U. The $igm of language. Cambridge, MA: Harvard University Press, 1979. p. 193. 7. Fouts R S. Use of guidance in teaching sign language to a chmpanzee ( troglodytes), J. Comp. Physiol. Psychol, 8(I5 15-22, 1972. 8. Gardner B T & Gardner R A. Twwway communication with an infant chimpanzee, (Schrier A M & StoUnitz F, eds.) ffekrvior of nonhuman pn”rnates. New York: Academic Press, 1971, Vol. 4, p. 117-S4. 9 ------Comparing the early utterances of chifd and chimpanzee. (Pick A D, cd.) Minnesota Sympo$ia on C/fi/d Psychology. Mimeapolis, MN: University of Minnesota Press, 1974. Vol. 8. p. >23. 10. Gardner R A & Gardner B T. A wxabulary test for chnpanzees (Pan troglodytes). J. Comp. Psychol, 98:3S1-404, 1984. 11. ------ and language acquisition. Ann. IVY A cad. Sci. 309:37-76, 1978. 12, Gardner B T & Gardner R A. Two comparative psychologists look at language acqutiltion. (Nefson K E, cd.) Children k language, New York: Gardner Press, 19S0, Vol. 2. p, 331-69, 13. Schlm.ingcr H S & Meadow K P. Sound and sign. Berkeley, CA: University of California Press, 1972. p. 69. 14, Gardner R A & Gardner B T. Early signs of language in child and chimpanzee. Science 187:752-3, 1975.

300 15. Fonts R S, Hfrsch A D & Fonts D H. Cultural transmission of a human language in a chmpanzee mother-infant relationship, (Fitzgerald H E, Mrdhs J A & Gage P, eds. ) Child rrurturance, New York: Plemrm Press, 1982. Vol. 3. p. 159-93. 16. Premack D. Langrrage in chimpanzee? .%ierrce 172:808-22, 1971. 17. ------The of man and beasts. Behav. Bruin Sci. 6:125-67, 1983. 18. Rumbrmgh D M, Gfff T V & von Gfasersfeld E C. Reading and sentence completion by a chmpanzee (Parr). Science 182:731-3, 1973, 19. voir Glaaerafeld E. The Yerkish language and its automatic parser. (Rumbaugh D M, ed. ) Language teaming by a chimpanzee. New York: Academic Press, 1977. p. 91-120. 20. Rumbangh D M & Gfff T V. Language and tbe acqrri.dtion of language-type skills by a chimpanzee (Pan). Ann. NY A cad. Sci, 270:9123, 1976. 21------Lana’s acquiabion of language skifk. (Rumbaugh D M, cd. ) Language learning by u chimpanzee. New York: Academic Press, 1977. p. 165-92. 22. Savage-Rumbaugh E S, RrrmbarrgfrD M & Boysen S. between two chimpanzees (Pan troglodytes), Science 201:641-4, 1978. 23. Savage-Rumbaugh E S, Rumbmrgb D M, Smith S T & Lawson 1. Reference: tbr. linguistic essential. Science 210922-5, 1980. 24. Savage-Rrrmbaugb E S & Rrrmbaugb D M. Language analogue project, phase II: theory and tactics. (Nelson K E, cd.) Children> language, New York: Gardner Press, 1980. Vol. 2, p. 267-307. 25. Savage-Rrrmbaugh E S, Sevcfk R A, Rumbaugh D M & Rubert E. The capacity of animals to acquire languagtx do species dtiferences have anything to say to us? Phil. Tmns. Roy. Sot. London B 308( 1315): 177-85, 1985. 26. Matauzawa T. Use of numbers by a cbinrpanzee. iVafure 31S(6014):57-9, 1985. 27. Patterson F G. The gestures of a gorilla: language acqrridtion in another pongid. Bmin Lang. 5:72-97, 1978. 28. Pattemon F& Lfmfen E. Theeducation of Koko. New York: Holt, Rinehart and Winston, 1981.224P. 29. Patiemon FG. Innovative uwsofhnguage byagoflla: acasestudy. (Nekon KE, ed.) Childrvnk farrguage. New York: Gardner Press, 1980. V01.2. P. 497-561. 30. ------. Conversationa with agorilla. Nat. Geogr. 154:438-65, 1978. 31. MEa HL. Ap=andlanguage: tieseamb forcommunicative competence, (de Luce J& Wifder HT, eds.)Lurrguage irrprimote.r. New York: Sprirrger-Verlag, 1983. p. 43-61. 32------The development of signs of reference in an orangutan. Unpublished paper presented at themeeting oftieherican Awmiation forthe Advancement of Science. 26-3l May 1985, Los Angeles, CA. 33. Le~y M& Geschwfnd N. Hemispheric differences inthebmins ofgreataF. Bmin Behav. Evo[. 11:48-52, 1975. 34. Maple TL. Omrrg-utun behavior. New York: Van Nostmnd Reinhold, 1980. P. 211. 35. Grffffn DR. Theque@iorr ofarrima/awarenem. New York: Rockefeffer University Press, 1976.135 p. 36. Terrace H S, Petftto L A, Sanders R J & Bever T G. Can an ape create a sentence? Science 206:891-902, 1979. 37. Storer N W. The hard sciences and the soft. IJuf/. Med. Libr. Awn. 55:75-84, 1967.

301