Report on SAGA2/COE Symposium



Probing the Realities of Ape Language

Sue Savage-Rumbaugh and William M. Fields


     Ape language work at GSU's Language Research Center (LRC) has produced important advances in the understanding of apes and their potential for linguistic processes since the first keyboard was presented to the chimpanzee (Pan troglodytes) Lana in 1972. At that time, little was known about the perceptual and cognitive capacities of great apes and it was considered unlikely that they would be capable of discriminating the small 2 dimensional printed patterns ("lexigrams") intended to serve as words. The first studies left no doubt that Lana could discriminate lexigrams visually, and that she could learn the simple ordering rules sufficiently well to apply them to novel sequences. Lana could also associate different symbols with various real world people, places, and things. The computer-collected data demonstrated that imitation was not the basis of her performance.
     The second generation of language studies with the lexical-keyboard system attempted to compensate for some of the perceived inadequacies in Lana's semantic performance. Her errors had revealed that while she grasped the combinatorial rules of her syntax, she often did not consistently apply semantic content. Consequently, the "meaning" of words came under intense focus, and receptive understanding, along with object labeling, became an important component of the linguistic instruction. The social aspect of language was also enriched far beyond what had been the case for Lana and lastly, in place of working with a single subject, efforts were concentrated upon communications between two co-reared male chimpanzees, Austin and Sherman (2 and 3 years of age).
     The work revealed that symbolic communication of a high level, with the use of an abstract code and with mutual understanding and cooperation, was possible between non-human creatures. It also revealed that the semantic processing of the symbolic components of the communicative system was not just lexically based and dependent upon stimulus-response associative phenomena. It was instead, semantically grounded and functionally abstract. Finally, it illustrated, for the first time in the field of animal language, the critical components of listener comprehension and listener cooperation.
     The third generation of language studies at LRC began with bonobos (Pan paniscus). Bonobos were selected because their naturally occurring gestural and vocal capacities appeared to be more complex than those of chimpanzees. At first this research effort focused upon replicating the chimpanzee studies with a mother-infant pair, Matata and her 6-month-old son, Kanzi. Matata's progress was slow in that she acquired only 8 symbols across a 3-year training period. Her son Kanzi however, learned the symbols even though he was not being trained. His success with observational learning led to the abandonment of training orientedapproach. All further effort focused upon the natural, rather than the acquired, acquisition of language. Lexigrams were combined with speech in semi-naturalistic forest setting which fostered communication and eschewed intentional training.
     Symbol acquisition and symbol comprehension proceeded much more readily under these conditions, and was made possible by the spontaneous appearance of a high level of spoken language comprehension. For the first time, contingencies associated with symbol use and acquistion were noncritical and understanding at a grammatical level emerged. These findings were replicated with an additional bonobo, Mulika.
     The fourth generation of language studies addressed the question of whether or not the capacity to acquire language skills without the assist of intentional training was unique to the bonobo or was instead a function of the new methodological approach that had been developed with Kanzi and Mulika. Two apes were again co-reared, but this time they were different species and the co-rearing began at birth. Neither Panpanzee (the chimpanzee) or Panbanisha (the bonobo) were exposed to training of any kind, nor did they observe training being given to Matata. They were immersed with a naturalistic environment as newborn infants, which included daily trips to the forest and round the clock exposure to spoken English accompanied by the presentation of lexical symbols. This work revealed that the species variable did not account for the differences in linguistic competency displayed by Kanzi and Mulika. Panzee was able to acquire lexical symbols and speech comprehension without exposure to any systematic training. She was however, delayed in her symbol acquisition relative to Panbanisha and Mulika, and her responses and symbol use were more concrete and less varied. In addition, her understanding of novel sentences was very limited, however the co-rearing only lasted until age4 and had it continued longer she might have demonstrated facility equal to that of Panbanisha in this regard.
     The current series of studies is centered around Panbanisha and her son Nyota and addresses cultural issues of bi-species rearing rather than biological platforms or methodological variables. The previous 27 years of language rearing efforts has led to the inescapable realization that no logical understanding of what has come to be called ape language can be obtained without fully acknowledging the cultural domain in which the exchanges take place. More importantly, no understanding of ape behavior can be achieved without recognizing the role of individuals within a socio-dynamic culture. Finally, a thorough grounding of knowledge requires a grasp of the cultural substrate which serves as the hidden platform for all communicative action. Achievement of such an understanding will take us far beyond what has traditionally been conceived of as the scientific study of other species; for speaking and sharing a culture with another species moves beyond investigating them---it moves directly into the realm of sharing the responsibility for life with them.
     The emergence of this new theoretical perspective can come about only through a bi-species co-construction of reality. That is, it will require the joint mental construction of what are seen --by members of both species -- as the social flow of connected events. This is because language becomes manifest only through jointly understood "events." The perception of action as "events" occurs only within an inter-individual subjective framework of connectedness which operates to structure space and time. It is through such "connectedness" that we are able to perceive human noise-based utterances as "language". That is to say, the noises which we make could not be perceived as carrying representational value if we did not subjectively construct these abstractions regarding the actions of others and the links between different actions, in a common manner. It is in understanding how we come to do this that we will finally grasp how it is that culture is fashioned through language, and language, in turn is fashioned through culture.



Back



Copyright (C) 1999- COE International Symposium
saga@pri.kyoto-u.ac.jp