Development of Language Skills in Bonobo and Chimpanzee Article Essay Can you help me writing a 3 pages assignment?You can send it back to me by Thursday m

Development of Language Skills in Bonobo and Chimpanzee Article Essay Can you help me writing a 3 pages assignment?You can send it back to me by Thursday midnight.The instructions are in the pdf file.Thank you. See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/232443743
The development of language skills in bonobo and
chimpanzee-I. Comprehension
Article in Language & Communication · April 1995
DOI: 10.1016/0271-5309(95)00001-7
CITATIONS
READS
58
303
2 authors:
Karen Brakke
Emily Rumbaugh
Spelman College
Bonobo Hope
28 PUBLICATIONS 657 CITATIONS
95 PUBLICATIONS 3,321 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Language, Tools and Culture in Apes View project
Assessment of Student Learning Outcomes in Psychology View project
All content following this page was uploaded by Karen Brakke on 11 October 2017.
The user has requested enhancement of the downloaded file.
SEE PROFILE
Pergamon
Language & Communication, Vol. 15, No. 2, pp. 121-148, 1995
Copyright © 1995 Elsevier Science Ltd
Printed in Great Britain. All rights reserved
0271-5309/95 $9.50 + 0.00
0271-5309(95)00001-1
THE D E V E L O P M E N T OF L A N G U A G E SKILLS IN B O N O B O AND
CHIMPANZEE–I. COMPREHENSION
KAREN E. BRAKKE and E. SUE SAVAGE-RUMBAUGH
The species most closely related to our own are the bonobo (Pan paniscus) and the chimpanzee
(Pan troglodytes), who diverged from the hominid evolutionary path as little as five to eight
million years ago (Diamond, 1988; Miyamoto et aL, 1988). The two species of Pan, in turn,
shared a common ancestor two to three million years ago (Diamond, 1988).
Although chimpanzees and bonobos have demonstrated many complex cognitive skills (e.g.
Boysen and Berntson, 1990; Premack, 1976; Rumbaugh et al., 1993; Toth et al., 1993), their
capacity to acquire and use language has been a topic of controversy (Terrace et al., 1979;
Vauclair, 1990). Part of the difficulty in assessing language skills in apes arises from the fact
that they do not speak. Even if they possessed all of the cognitive requisites for language, the
structure of their vocal apparati would not permit articulate speech (Duchin, 1990; Lieberman,
1991). Thus, in order to examine language production in the great apes, investigators have
turned to sign (Fouts, 1973; Gardner and Gardner, 1969; Miles, 1983; Patterson, 1978) or to
artificial symbol systems (Premack, 1971; Rumbaugh, 1977; Savage-Rumbaugh, 1986;
Savage-Rumbaugh et al., 1986), each of which entail motor and/or vocabulary limitations for
the non-human subjects. Manual signs such as those used in American Sign Language (ASL),
for example, often require fine motor coordination that proves awkward for less dextrous ape
hands. Also, ape subjects using non-sign symbols are exposed to at most several hundred
symbols that they can produce (Premack, 1976; Savage-Rumbaugh et al., 1986) rather than the
thousands of words heard by most children as they mature. Evaluation of the apes’ productive
skills is in large part, then, determined by modality limitations.
Comprehension may provide a better index of a species’ competence for language than does
production. As far as we know, no peripheral anatomical limitations apply to apes’ ability to
comprehend human speech. The ear structure and the sensory-perceptual auditory pathways of
the two genera appear to be very similar (Ankel-Simons, 1983); speech comprehension
differences between species, therefore, are likely to reflect CNS functioning. By evaluating
which of the receptive capacities underlying linguistic competence are well developed in
species closely related to our own, and which are not, we may be able to better understand how
humans came to be the adept linguistic creatures that we are.
Language comprehension in humans
Developmental psychologists are familiar with the premise that comprehension ‘leads’
production (Goldin-Meadow et al., 1976; Huttenlocher, 1974). Comprehension of language
Correspondence relating to this paper should be addressed to Karen E. Brakke, Language Research Center, Georgia
State University, University Plaza, Atlanta, GA 30303-3083, U.S.A.
121
122
KAREN E. BRAKKE and E. SUE SAVAGE-RUMBAUGH
begins to be evident around 8-10 months of age (Bates, 1993), and a child’s receptive
vocabulary early in the language-learning process is likely to be larger than its productive
vocabulary (Bates, 1993; Goldin-Meadow et al., 1976; Snyder et al., 1981).
Although language comprehension begins to appear early in life and does not inherently
require the fine motor abilities that speech and sign productive skills do, it is by no means a
simple or automatic phenomenon. The development of receptive language skills in our species
requires that several minimal conditions be met. Using spoken language as an example, the
individual must be able to distinguish different phonemes, then relate them to one another as
words. Meaning must be attached to the words, and finally the words must be systematically
related in order for the individual to derive the aggregate meaning of the utterance. Underlying
these linguistic skills are cognitive capacities such as the ability to integrate temporally or
spatially distinct pieces of information (Bates, 1993; Diamond, 1991), to mentally represent
absent entities (Piaget and Inhelder, 1969), and to establish part-whole relations (Langer,
1986). The context of the utterance and the communicative intent of the speaker are also
essential components of successful understanding of others’ messages (Bruner, 1983; Lock,
1980).
In addition to cognitive requisites, appropriate environmental conditions must also be extant
for language-learning to occur. The infant cannot learn language unless it is exposed to it. It is
becoming increasingly evident that the development of language competence is heavily
supported by the people and culture surrounding the infant (Bruner, 1983; Lock, 1980). Use
of child-directed speech (Berko-Gleason, 1989; Ferguson, 1964; Snow, 1977) and joint
attention routines (Bakeman and Adamson, 1984; Bruner, 1975; Trevarthen and Hubley,
1978), for example, highlight the communicative process and engage the infant’s attention.
Parents encourage children to participate in and respond appropriately to utterances during
daily activities long before they expect to hear similar utterances from the infants (Bruner,
1983; Goldin-Meadow et al., 1976).
Comprehension, then, arises in humans during normal day-to-day interactions between
infants and others. Children are not taught language by drilling repeated trials of words in
order to get a food reward. This must be considered when studying language competency in
other primates. In order to evaluate skills of related species at a level that permits comparison
to our own, we must provide rearing environments as similar as possible to those found in
human cultures. This includes providing linguistic and social input that is modeled after the
way human adults interact with infants around the world.
Language comprehension in bonobos
At the Language Research Center (LRC), two bonobos (Kanzi and Mulika), learned to use
and understand symbols in many different communicative capacities without explicit training
(Savage-Rumbaugh et al., 1990; Sevcik and Savage-Rumbaugh, 1994). Their instruction was
much less structured than that of other apes who were taught to use symbols in repeated-trial
or reward-based paradigms (Premack, 1971; Rumbaugh, 1977; Savage-Rumbaugh, 1986).
Language-learning by these bonobos involved observation of human models who used the
symbols as they interacted with the apes throughout the day. The young bonobos were asked
to attend to and participate in the daily activities of the laboratory (including play and
structured tasks), but were not required to produce symbols in order to receive food or other
rewards.
COMPREHENSION IN PAN
123
The two bonobos spontaneously developed receptive language skills and readily
comprehended several English words under blind test conditions (Savage-Rumbaugh et al.,
1993; Sevcik and Savage-Rumbaugh, 1994). The older of these bonobos has, in fact,
demonstrated the ability to understand several hundred novel complex spoken sentences
(Savage-Rumbaugh et al., 1993). Furthermore, the comprehension of symbols generalized to
symbol production in these subjects (Savage-Rumbaugh et al., 1990). In short, the two
bonobos were reared much like children and their symbol acquisition evidenced many of the
same characteristics as that of young humans.
Language-learning and routines
Central to the bonobos’ learning of natural language was participation in interindividual
routines accompanied by adult language use. As outlined in Savage-Rumbaugh et al. (1993),
these routines are ‘more or less regularly sequenced sets of interindividual interactions that
occur in a relatively similar manner on different occasions’ (p. 25). These interactions are
embedded within a larger context. For example, within an ‘evening’ context when the lab is
quiet, there will occur routines involving milk preparation, quiet play, diaper changing, and
grooming until the infant falls asleep. Each of these routines, in turn, is made up of a sequence
of acts of components such as–for milk preparation–putting the lining in the bottle, opening
the milk formula, pouring the formula into the bottle, adding water, putting the nipple on, and
finally drinking the milk.
Within these routines, each participant has a role that from day to day stays relatively
consistent but over the course of time may shift, with the infant becoming more and more
active in initiating the components of the routine. Initially, however, the adult participant
directs the flow of the interaction and assigns a role to the infant that consists of paying
attention and indicating readiness to continue with the next component of the routine. Salient
components or transitions of the routine are behaviorally and linguistically (via speech and/or
lexigram) encoded by the adult. Such ‘marking’ of the components appears to be most
effective when it occurs prior to the actual event (Savage-Rumbaugh, 1991). Talking about
what we are going to do, before we do it, allows all participants to predict what will happen
next. Once able to predict upcoming events, the infant can begin to exert some control over
how those events unfold.
Quite often, components of a ‘standard’ routine will be added or modified. If, while
preparing milk, for example, it is discovered that the box of bottle liners is empty, a new box
will have to be located and opened. Linguistic marking becomes especially important at these
times when the routine is altered. By providing information that might otherwise be lacking,
the verbal information allows all of the individuals involved to coordinate their behavior so
that continuation of the routine may proceed smoothly (see Savage-Rumbaugh, 1991, for a full
discussion of the contributions of routines to language-learning in apes and children).
Species or rearing differences?
Prior to the study with bonobos at the LRC, two male chimpanzees, Sherman and Austin,
were taught referential use and comprehension of visuographic symbols (‘lexigrams’) via
explicit trial-based paradigms. Each communicative function–requesting, labeling, and
comprehension of others’ utterances–was taught separately and required many repeated trials
for the chimpanzees to master. Productive skills did not generalize to comprehension without
explicit training (Savage-Rumbaugh, 1986), and comprehension of symbols, once learned, did
124
KAREN E. BRAKKE and E. SUE S A V A G E – R U M B A U G H
not extend to speech (Savage-Rumbaugh et al., 1985b). Although Sherman and Austin
appeared to understand some verbal requests presented to them on a regular basis, when
compared on a formal single-word receptive language task (‘Give me the x’), the two
chimpanzees demonstrated no comprehension of spoken English divorced of contextual
cues.
The data from Sherman and Austin on one hand and Kanzi and Mulika on the other
suggested a possible species difference in the ability to understand complex speech (SavageRumbaugh et al., 1985a). If confirmed, this difference would be a point of distinction between
the two species of Pan and might provide important clues to the evolution of human
language.
The comparison between the bonobo (Kanzi and Mulika) and chimpanzee (Sherman and
Austin) subjects was compromised by two differences in their rearing histories, however. The
chimpanzees were 1% to 2’~ years old when they started symbol training, while the bonobos
were 4-6 months old when they were first exposed to symbol use and English-speaking
caregivers. It may be that early exposure and attachment to humans is a critical component for
developing speech comprehension in chimpanzees.
There is some evidence that this may be the case. Fouts et al. (1976) reported appropriate
responses to at least 10 English words in a chimpanzee trained in ASL. Earlier studies with
chimpanzees Viki (Hayes, 1951) and Gua (Kellogg and Kellogg, 1933) suggested that some
speech comprehension may have developed in ‘home-reared’ subjects, but these reports were
based on data collected without contextual controls and so are inconclusive.
The second difference between the LRC studies involving the chimpanzees on the one hand
and the bonobos on the other was the language-learning environment experienced by the apes
after their inclusion in the project. The bonobos learned symbol use observationally, with an
emphasis on symbol comprehension. The chimpanzees, on the other hand, were first taught
symbolic request skills in a structured, trial-based paradigm. As Savage-Rumbaugh et al.
(1990) noted, it ‘may be that an early emphasis on productive symbol t r a i n i n g . . , may, in
itself, interfere with acquisition processes that rely on observation’ (p. 250).
In order to evaluate both productive and receptive language differences in the two species
of Pan, then, the next step was to raise a member of each species together from an early age,
providing them with the opportunity to learn language skills observationally and emphasizing
the comprehension of verbal and visual symbol use instead of requiring symbol production
from the outset.’ This is the same paradigm used with the older bonobos, Kanzi and Mulika.
If under such conditions the chimpanzee fails to develop speech comprehension skills similar
to those that bonobos have exhibited, then this species difference may represent a key, abrupt
development in the phylogeny of linguistic competence. If at the other extreme, however, the
chimpanzee and bonobo profit equally from observational learning, this would testify to the
importance of early exposure to a linguistic culture in fostering comprehension skills even in
species other than our own.
This comparison has now been effected with a single bonobo and chimpanzee pair and the
results of the study relevant to comprehension are reported herein. Two general aspects of the
subjects’ receptive language development provided the foci of the present study: their initial
entry into the system (i.e. how they first came to respond appropriately to utterances) and the
emergence of competence (i.e. the growth of their skills responding to single words and
complex utterances).
COMPREHENSION IN P A N
125
Method
Subjects
Panbanisha, the bonobo (Pan paniscus), was born at the Language Research Center (LRC)
on 17 November 1985. She was reared by her mother, Matata, until she was 7 weeks of age
when her care was transferred to human caregivers. Panpanzee, the chimpanzee (Pan
troglodytes), was born at the main station of the Yerkes Regional Primate Research Center on
31 December 1985 (the first chimpanzee born there after Panbanisha’s birth) and transferred
to the LRC on 8 January 1986. From that day through the end of the study period (October
1989), Panbanisha and Panpanzee were nearly always together and in the company of at least
one human caregiver, except for the last months of the study when the subjects slept with older
bonobos at night. The subjects formed close attachment bonds with the caregivers who worked
with them consistently.
Rearing and communicative environment
Panbanisha and Panpanzee were reared together in an environment similar to that in which
many human children are raised, but one that took into account the activities and interests
to which young apes are prone. Eating, sleeping, bathing, changing diapers, traveling in
the woods and climbing trees, social interaction with other apes and humans, and quiet
‘work sessions’ (during which various assessments might be administered) filled the young
apes’ day.
The different routines of the day, interspersed with occasional twists or surprises, lent
structure to the young apes’ lives and also created an optimal environment for communication
about onging and upcoming events. Such an environment was considered crucial to the
emerging theory of non-human language acquisition that had been established a few years
earlier with the bonobos Kanzi and Mulika (Savage-Rumbaugh, 1991; Savage-Rumbaugh et
al., 1986). Panpanzee and Panbanisha were included in all aspects of laboratory life, and every
effort was made to convey information to them with lexigrams as well as with spoken English.
It was in this way, through observation of adult models and subsequent comprehension of
utterances, that the apes developed language skills.
Throughout the study, caregivers spoke to the subjects in English, much as they would
speak to human infants. No specific procedures were followed in speaking with the infants, but
speech could often be characterized as ‘child-directed’ (also known as ‘motherese’) (BerkoGleason, 1989), especially when the subjects were very small. Inflection, tone of voice, eye
contact, and gestures were all used as necessary to communicate the desired message during
daily interaction.
The principal difference between communication with these apes and communication with
human infants was that, whenever possible, visuographic symbols (called ‘lexigrams’) were
used to supplement or complement the English utterances. These symbols had been created
and assigned meanings during earlier study with other apes (Rumbaugh, 1977; SavageRumbaugh et aL, 1986). Each lexigram was glossed as the equivalent of an English word, such
as ‘milk’, ‘dog’ or ‘good’. Approximately 256 lexigrams were available for use throughout the
study period. The symbols were arranged on faceplates throughout the lab but they
sometimes also appeared on individual photographs of the referents or were mounted
individually in appropriate places (for example, the ‘bedroom’ lexigram was taped to the
bedroom door).
126
KAREN E. B R A K K E and E. SUE S A V A G E – R U M B A U G H
Data collection procedures
Daily data. Throughout the 4 years of the co-rearing project, the caregivers who
accompanied Panbanisha and Panpanzee throughout the day recorded virtually all instances of
symbol production and symbol and speech comprehension by each subject.: This was done in
order to provide an exhaustive record of all language-related behavior produced by the apes
that would be amenable to several types of analyses relevant to different aspects of language
acquisition. The Appendix provides examples of notes collected with the subjects during the
study period.
The daily data provided by the apes are similar to those of diary studies with human
children. Each entry was handwritten in a notebook as soon as possible after its occurrence
along with codes representing communicative function and behavioral concordance or
response, as well as contextual notes describing the context of the utterance or receptive
behavior. The entries were later entered onto a computer database, from which the receptive
data reported in this manuscript were extracted.
The receptive entries recorded in this way represent instances during daily interaction in
which a subject’s comprehension of caregivers’ utterances was evident. The exchanges
occurred withi…
Purchase answer to see full
attachment