SCIENTIFIC

mER-- OFFPRINTS

Teaching Language to an Ape

by Ann James Premack and David Premack

SCIENTIFIC
s/ilYIERI~
OCTOBER 1972
VOL. 227. NO.4 PP.92-99

’
_’
II
- PUBLISHED BY W. H. FREEIVMN AND COMPANY &jO MARKET STREET, SAN FRANCISCO, CALIFORNIA 94104

Copyright 0 1972 by Scientific American, Inc. All rights reserved. Printed in the U.S.A. No part of this offprint may be reproduced by any mechanical, photographic or electronic Process, or
in the form of a phonographic recording, nor may it be stored in a retrieval system, transmitted or otherwise copied for public or private use without written permission of the publisher.
 -
c y
at

Teaching Language to anApe l ‘1

Sarah, a young chimpanzee, has a reading and writing vocabulary
of about 130 Qords.” Her understanding-goes beyond the meaning
of words and includes the concepts of class and sentence structure

by Ann James Premack ind David Premack

0
ver the past 40 years several ef- derstand about 100 words, but she never recently Allen and Beatrice Gardner
forts have been made to teach a did try to speak them. In the 1940’s have taught their chimpanzee Washoe
chimpanzee human language. In Keith and Cathy Hayes raised a chim- to communicate in the American Sign
the early 1930’s Winthrop and Luella panzee named Vicki in their home; she Language with her fingers and hands.
Kellogg raised a female chimpanzee learned a large number of words and Since 1966 in our laboratory at the Uni-
named Gua along with their infant son; with some difficulty could mouth the versity of California at Santa Barbara
at the age of 16 months Gua could un- words “mama,” “papa” and “cup.” More we have been teaching Sarah to read
and write with variously shaped and
colored pieces of plastic, each repre-
senting a word; Sarah has a vocabulary
of about 130 terms that she uses with a
reliability of between 75 and 80 percent.
Why try to teach human language to
an ape? In our own case the motive was
to better define the fundamental nature
of language. It is often said that lan-
guage is unique to the human species.
Yet it is now well known that many oth-
er animals have elaborate communica-
tion systems of their own. It seems clear
that language is a general system of
which human language is a particular,
albeit remarkably refined, form. Indeed,
it is possible that certain features of hu-
man language that are considered to be
uniquely human belong to the more gen-
eral system, and that these features can
be distinguished from those that are -
unique to the human information-proc- -
essing regime. If, for example, an ape
cau be taught the rudiments of human
language, it should clarify the dividing
line between the general system and the
human one.

T chimpanzee
here was much evidence that the
was a good candidate for
the acquisition of language before we
began our project. In their natural en- (
vironment chimpanzees have an exten-
sive vocal “call system.” In captivity the
SARAH, after reading the message “Sarah insert apple pail banana dish” on the magnetic chimpanzee has been taught to sort pie-
board, performed the appropriate actions. To be able to make the correct interpretation tures into classes: animate and inani-
that she should put the apple in the pail and the banana in the dish (not the apple, pail and mate, old and young, male and female.
banana in the dish) the chimpanzee had to understand sentence structure rather than just Moreover, the animal can classify the
word order. In actual tests most symbols were colored (see iEZustration on opposite page). same item in different ways depending
2
 NOUNS

SARAH MARY PAIL . DISH

CHOCOLATE APPLE BANANA APRICOT RAISIN

VERBS

GIVE TAKE INSERT WASH

:ONCEPTS/CONDITIONALS

SAME DIFFERENT NO-NOT

NAME OF COLOR OF
A ? IF-T&N

ADJECTIVES (COLORS)

RED GREEN

PLASTIC .SYMBOLS that varied in color, shape and size were cept. A “Chinese” convention of writing sentences vertically from
chosen as the language units to be taught to Sarah. The plastic top to bottom was adopted because at the beginning of her training
pieces were backed with metal so that they would adhere to a mag- Sarah seemed to prefer it. Sarah had to put the words in proper se-
netic board. Each plastic symbol stood for a specific word or con- quence but the orientation of the word symbols was not important.
3
 SAME DIFFERENT

CONCEPTS “SAME” AND “DIFFERENT” were introduced into were placed before her and she was given plastic word for “same”
Sarah’s vocabulary by teaching her to pair objects that were alike and induced to place word between the two objects. She was also
(top illustration). Then two identical objects, for example apples, taught to place the word for “different” between unlike objects.

DiFFERENT SAME SAME DIFFERENT

THE INTERROGATIVE was introduced with the help of the con- tion mark” was placed between two objects and Sarah had to re-
cepts “same” and “different.” A plastic piece that meant “ques- place it with either the word for“same” or the word for “different.”

SAME DIFFERENT - - ?-

NEW VERSION OF THE INTERROGATIVE was taught by ar- [Object Al the same as?” or “What is [Object A] different from?”
ranging an object and plastic symbols to form questions: “What is Sarah had to replace question marker with the appropriate object.

4
 on the alternatives offered. Watermelon her reach. To obtain the fruit Sarah now between pairs of objects. Thjs ~1~~~~~
is class&xl as fruit in one Set of alterna- had to put the plastic piece on a “lan- shape (which bore no resemblance to the
tives, as food in another set and as big guage board” on the side of her cage. usual kincl of question mark) made the
in a third set. On the basis of these dem- (The board was mngnetic and the plastic; question es$icit rather than illlplicit, as
onstrated conceptual abilities we made square was backecl witli a thin piece of it had been in the simple matelling tests.
the assumption that the chimpanzee steel so that it would stick.) After Sarah When tJ)e interrogative element \vas
could be tnuSht not only the names of had learned this routine the fruit was placed L)etweeu a pair of cups, jt lne;lllt;
specific n~emlxxs of a class but also the changed to an apple and she had to “What is tlze relation between cup A and
names for the classes themselves. place a blue plastic word for apple on cup B?” The choices provided Sar:lJl
It is not necessary for the names to be the board. Later several other fruits, the were the plastic words “same” and “dif-
vocal. They can just as well be based verb “give” and the plastic words that ferent.” She learned to remove the in-
on gestures, written letters or colored named each of them were introduced. terrogative particle and substitute the
stones. The important thing is to shape To be certain that Sarah knew the correct word [see to)) Illnstrntiotz OIZ op..
the language to fit the itlfolmation-proc- meaning of “give” it was necessary to p05itc fMgc]. Sarah was able to transfer
essing capacities of the chimpanzee. To contrast “give” with other verbs, such what she had learned and apply the
a large extent teaching language to an as “wash,” “cut” and “insert.” When word ‘same” or “differeilt” to numerous
animal is simply mapping out the con- Sarah indicated “Give apple,” she was pairs of objects that had not been used in
ceptual structures the animal already given a piece of apple. When she put her training.
possesses. By using a system of naming “Wash apple” on the board, the apple Any construction is potentially a ques-
that suits the chimpanzee we hope to was placed in a bowl of water and tion. From the viewpoint of structural
find out more about its conceptual world. washed. In that way Sarah learned what linguistics any construction where one
Ultimately the benefit of language es- action went with what verb. or more elements are deleted becomes a r
periments with animals will be realized In the first stage Sarah was required question. The constructions we used
in an understanding of intelligence in to put only one word on the board; the with Sarah were “A same A” and “‘A dif-
terms not of scores on tests but of the name of the fruit was a sufficient indi- ferent B.” Elements in these construc-
underlying brain mechanisms. Only then cator of the social transaction. When tions were removed and the deletion was
can cognitive mechanisms for classify- names for different actions-verbs-were marked with the interrogative symbol;
ing stimuli, for storing and retrieving in- introduced, Sarah had to place two Sarah was then supplied with a choice
formation and for problem-solving be words on the board in vertical sequence. of missing elements with which she
studied in a comparative way. In order to be given an apple she had to could restore the construction to its fa-
The first step in teaching language is write “Give apple.” When recipients miliar form. In principle interrogation
to exploit knowledge that is already were named, two-word sentences were can be taught either by removing an
present. In teaching Sarah we first not accepted by the trainer; Sarah had element from a familiar situation in the
mapped tile simple social transaction of to use three words. There were several animal’s world or by removing the ele-
giving, which is something the chimpan- trainers, and Sarah had to learn the ment from a language that maps the an-
zee does both in nature and in the labo- name of each one. To facilitate the imal’s world. It is probable that one can
ratory. Considered in terms of cpgnitive teaching of personal names, both the induce questions by purposively remov-
and perceptual elements, the verb “give” chimpanzees and the trainers wore their ing key elements from a familiar situa-
involves a relation between two individ- plastic-word names on a string necklace. tion. Suppose a chimpanzee received its
uals and one object, that is, between the Sarah learned the names of some of the daily ration of food at a specific time
donor, the recipient and the object being recipients the hard way. Once she wrote and place, and then one day the food
transferred. In order to carry out the act “Give apple Gussie,” and the trainer was not there, A chimpanzee trained in
of giving an animal must recognize the promptly gave the apple to another the interrogative might inquire “Where
difference between individuals (between chimpanzee named Gussie. Sarah never is my food?” or, in Sarah’s case, “My
“&fary” and “Randy”) and must perceive repeated the sentence. At every stage food is?” Sarah was never put in a situa-
the difference between donors and re- she was required to observe the prop- tion that might induce such interro-
cipien ts (between “Mary gives Randy” er word sequence. “Give apple” was ac- gation because for our purposes it was
and “Randy gives Mary”). In order to cepted but “Apple give” was not. When easier to teach Sarah to answer ques-
be able to map out the entire transac- donors were to be named, Sarah had to tions.
tion of giving the animal has to distin-- identify all the members of the social
guish agents from objects, agents from transaction: “Mary give apple Sarah.” t first Sarah learned all her words in
one another, objects from one another A the context of social exchange. La-
and itself from others. r he interrogative was introduced with ter, when she had learned the concepts
Is‘ the help of the concepts “same” and “name of” and “not name of,” it was
r ‘he trainer began the process of map- “different.” Sarah was given a cup and possible to introduce new words in a
1 ping the social transaction by plac- a spoon. When another cup was added, more direct way. To teach her that ob-
ing a slice of banana between himself she was taught to put the two cups to- jects had names, the plastic word fol
and Sarah. The chimpanzee, which \vas gether. Other sets of three objects were “apple” ancl a real apple were placed on
then about five years old, was allowed given to her, and she had to pair the the table and Sarah was required to put
to eat the tasty morsel while the tr;iiner &o objects that were alike. Then she the plastic word for “name of” between
looked on affectionately. After the trnlls- was taught to place the plastic word them, The same procedure was repeated
action had become routine, a lnnguagc for “same” between any two similar ob- for banana. After she had responded car-
element consisting of a pinlc plastic jects and the plastic word for “different” rectly several times, the symbol for “ap-
SC]UWe WklS placed Close to Sarull wlijle unlike objects. Next what plc” and a real banana were placed on
the slice of banana was moved l~eyolld to a question mark was placed ;hc table and Sarah had to put “not
5
 written construction: “Sarah t&e apple’?
Mary give chocolate Sarah.” Sarah was
provided with only one plastic word: the t
conditional particle. She had to remove
the question mark and substitute the
conditional in its place to earn the apple
and the chocolate. Now she was pre-
sented with: “Sarah take banana ? Mary
no give chocolate Sarah.” Again only the
conditional symbol was provided. When
Sarah replaced the question mark with
the conditional symbol, she received a
banana but no chocolate. After several
such tests she was given a series of trials
on each of the following pairs of sen-
tences: “Sarah take apple if-then Mary
give chocolate Sarah” coupled with
“Sarah take banana if-then Mary no give
chocolate Sarah,” or “Sarah take apple
if-then Mary no give chocolate Sarah”
coupled with ‘Sarah take banana if-then
Maly give chocolate Sarah.” r
At first Sarah made many errors, tak-
ing the wrong fruit and failing to get her
beloved chocolate. After several of her
strate,& had failed she paid closer
attention to the sentences and began
choosing the fruit that gave her the
chocolate. Once the conditional relation
had been learned she was able to apply
it to other types of sentence, for euam-
pie “Mary take red if-then Sarah take
apple” and “Mary take green if-then
Sarah take banana.” Here Sarah had to
watch Mary’s choice closely in order to
take the correct action. With the paired
TEACHING LANGUAGE WITH LANGUAGE was the next step. Sarah was taught to put sentences *‘Red is on green if-then Sarah
the symbol for “name of” between the word for “apple” and an apple and aIso between
take apple” and “Green is on red if-then
the word for “banana” and a banana. She learned the concept “not name of” in the same
way. Thereafter Sarah could be taught new nouns by introducing them with “name of.” Sarah take banana,” which involved a
change in the position of two colored
cards, Sarah was not confused and per-
name of” between them, After she was applied by Sarah in linguistic construc- formed well.
able to perform both operations correct- tions.
ly new nouns could be taught quickly
and explicitly, The plastic words for
In English the conditional consists of
the discontinuous elemknts “if-then,” As aconcepts
preliminary to learning the class
of color, shape and size
“raisin” and “name of” could be placed which are inconvenient and conceptual- Sarah was taught to identify members of
next to a real raisin and Sarah would ly unnecessary. In symbolic logic the the classes red and yellow, round and
learn the noun. Evidence of such learn- conditional consists of the single sign square and large and small. Objects that
ing came when Sarah subsequently re- 3, and we taught Sarah the donditionnl varied in most dimensions but had a par-
quested “Mary give raisin Sarah” or set relatidn with the use of a single plastic ticular property in common were used.
down “Raisin different apple.” word. Before being given language train- Thus for teaching the word “red” a set of
An equally interesting linguistic leap ing in the conditional, she was given dissimilar, unnamed objects (a baI1, a toy
occurred when Sarah Iearned the predi- contingency training in which she was car, a Life Saver and so on) that had no
cate adjective and could write such sen- rewarded for doing one thing but not property in common except redness were
tences as “Red color of apple,” “Round another. For example, she was given a put before the chimpanzee, The only
shape of apple” and “Large size of ap- choice between an apple and a banana, plastic word available to her was “red.”
ple.” When asked for the relation be- and only when she chose the apple was After several trials on identifying red
tween “Apple is red ? Red color of ap- she given chocolate (which she dearly with a set of red objects and yellow with
Pl err and given “same” and “different” as loved). “If apple, then chocolate, if ba- a set of yellow objects, Sarah was shifted
choices, she judged the sentences to be nana, then no chocolate” were the rela- to trials where she had to choose be-
the same. When given “Apple is red ? tions she learned; the same relations :ween “red” and “yellow” when she was
Apple is round,” she judged the sen- were subsequently used in sentences to shown a colored object. Finally com-
!.znces to be different. The distinctions teach her the name for t/he conditional 9etely new red and yellow objects were
between similar and different, first relation. iresented to her, including small cards
learned with actual objects, was later The subject was introduced with the hat were identical except for their color.
_-- - -
&
 Again she performed at her usual level of a piece of fruit and two plastic words, using actual fruits in one test and only
accuracy. she was required to put the correct word fruit names in the other. Sarah’s choices
Sarah was subsequently taught the for the fruit on the board before she was between the words were much the same
names of shapes, “round” and “square,” allowed to eat it. Surprisingly often, as her choices between the actual fruits.
as well as the size names “large” and however, she chose the wrong word. It This result strongly suggests that she
“small.” These words formed the basis then dawned on us that her poor per- could generate the meaning of the fruit
for teaching her the names of the class formance might be due not to errors names from the plastic symbols alone.
concepts “color of,” “shape of” and ‘size but to her trying to express her prefer- We obtained clearer evidence at a
of.” Given the interrogative “Red ? ap- ences in fruit. We conducted a series of later stage of Sarah’s language training.
ple” or “Yellow ? banana,” Sarah was re- tests to determine her fruit preferences, In the same way that she could use
quired to substitute the plastic word for
“color of” for the question mark. In
teatihing class names a good many sen-
tences were not written on the board but
were presented as hybrids, The hybrid
sentences consisted of a combination of
plastic words and real objects arranged
in the proper sentence sequence on SARAH MARY
Sarah’s worktable. Typical sentences
were “Yellow ?” beside a reaI yellow
balloon or “Red ?” beside a red wood
block. TAKE GIVE
The hybrid sentences did not deter
Sarah in the least. Her good perform-
ance showed that she was able to move
APPLE CHOCOLATE
with facility from symbols for objects to
actual objects. Her behavior with hy-
brid constructions recalls the activity of
young children, who sometimes combine IF SARAH
spoken words with real objects they THEN
are unable to name by pointing at the
objects.
Was Sarah able to think in the plastic-
word language? Could she store infor-
mation using the plastic words or use
them to solve certain kinds of problem
that she could not solve otherwise? Ad-
ditional research is needed before we SARAH MARY
shall have definitive answers, but Sarah’s
performance suggests that the answers
to both questions may be a qualified yes.

0
To think with language requires being TAKE NO
able to generate the meaning of words
in the absence of their external repre-
sentation. For Sarah to be able to match
BANANA GIVE
“apple” to an actual apple or “Mary”
to a picture of Mary indicates that she
knows the meaning of these words. -It
does not prove, however, that when she IF CHOCOLATE
is given the word “apple” and no apple
THEN
is present, she can think “apple,” that is,
mentally represent the meaning of the
word to herself. The ability to achieve SARAH
SuCh mental representation is of major
importance because it frees language
from simple dependence on the outside
world. It involves displacement: the
ability to talk about things that are not
actually there. That is a critical feature
of language.

T- hederstand
CONDITIONAL RELATION, which in English is expressed “if.. . then,” was taught to
hint that Sarah was able to un- Sarah as a single word. The plastic symbol for the conditional relation was placed between
words in the absence of two sentences. Sarah had to pay attention to the meaning of both sentences very closely in
their external referents came ear-y in her order to make the choice that would give her a reward. Once the conditional relation was
language training. When she was given learned by means of this procedure, the chimpanzee was able to apply it to other situations.

7
 1
l \a

“name of to learn new nouns, she was l

able t0 use “Color Of” to leaIn the nameS

of new COlOrS. For instance, the names L.

“brown” and “green” were introduced
in the sentences “Brown color of choco-
late” and “Green color of grape.” The
RED ? APPLE only new words at this point were
“brown” and “green.” Later Sarah was
confronted with four disks, only one of
which was brown, and when she was
instructed with the plastic symbols
“Take brown,” she took the brown disk.
Since chocolate was not present at any
YELLOW BANANA
time during the introduction of the color
name “brown,” the word “chocolate” in
CLASS CONCEPT OF COLOR was taught with the aid of sentences such as “Red ? apple”
the definition must have been sufficient
and “Yellow ? banana.” Sarah had to replace the interrogative symbol with “color of.”
to have Sarah generate or picture the
property brown.
What form does Sarah’s supposed in-
ternal representation take? Some indica-
tion is provided by the results of a test
of ability to analyze the features of an
object, First Sarah was shown an actual
apple and was given a series of paired
YELLOW COLOR OF comparisons that described the features
of the apple, such as red v. green, round
V. square and so on. She had to pick the
descriptive feature that belonged to the
apple. Her feature analysis of a reaI ap-
ple agreed nicely with our own, which is
evidence of the interesting fact that a
RED NO - COLOR OF BANANA -
chimpanzee is capable of decomposing a
complex object into features. Next the
NEGATIVE CONCEPT was introduced with “no-not.” When asked “Yellow 2’ apple” or
apple was removed and the blue plastic
“Red ? banana,” Sarah had to replace interrogative symbol with “color of” or “not color of.”
triangle that was the word for “apple”
was placed before her and again she was
given a paired-comparison test. She as-
signed the same features to the word
ALTERNATIVE FEATURES that she had earlier assigned to the ob-
ject. Her feature analysis revealed that
it was not the physical properties of the
word (blue and triangle) that she was de-
scribing but rather the object that was
represented by the word [see bottom il-
lustration at left],

T sion
o test Sarah’s sentence comprehen-
she was taught to correctly fol-
low these written instructions: “Sarah
insert apple pail,” “Sarah insert banana
pail,” “Sarah insert appIe dish” and
“Sarah insert banana dish.” Next instruc-
tions were combined in a one-line ver-
tical sequence (“Sarah insert apple pail
Sarah insert banana dish”). The chim-
panzee responded appropriately. Then
the second “Sarah” and the second verb
“insert” were deleted to yield the com-
pound sentence: ‘Sarah insert apple pail
banana dish.” Sarah followed the corn--
plicated instructions at her usual level of
FEATURE ANALYSIS o f an actual apple and the plastic word for “apple” was conducted. accuracy.
Sarah was shown an apple or the word and made to choose from alternative features: red The test with the compound sentence
or green, round or square, square with stem or plain square and square with stem or round. is of considerable importance, because it
Sarah gave plastic word for “apple” same attributes she had earlier assigned to apple. provides the answer to whether or not

8
 Sarah could understand the notion of to say that he recognizes the various lev- to require of Sarah what one would re-
constituent structure: the hierarchical els of sentence organization: that the quire of a human adult. Compared with
organization of a sentence. The correct subject dominates the predicate and the a two-year-old child, however, Sarah
interpretation of the compound sentence verb in the predicate dominates the ob- holds her own in language ability. 11~
was “Sarah put the apple in the pail and jects. fact, language demands were made of
the banana in the dish.” To take the cor- Sarah that would never be made of a
rect actions Sarah must understand that
“apple” and “pail” go together but not
S arah had managed to learn a code, a
simple language that nevertheless in-
child. Man is understandably prejudiced
in favor of his own species, and members
“pail” and “banana,” even though the cluded some of the characteristic fea- of other spedies must perform Herculean
terms appear side by side. Moreover, she tures of natural language. Each step of feats before they are recognized as hav-
must understand that the verb ‘iosert” the training program was made as sim- ing similar abilities, particularly lan-
is at a higher level of organization and ple as possible. The objective was to re- guage abilities. Linguists and others who
refers to both “apple” and “banana.” Fi- duce cornpIes notions to a series of sim- study the development of language tend
nally, Sarah must understand that she, as ple and highly learnable steps. The same to esaggerate the child’s understanding
the head noun, must carry out all the ac- program that was used to teach Sarah to of language and to be extremely skepti-
tions. Xf Sarah were capable only of link- communicate has been successfully ap- cal of the experimentally demonstrated
ing words in a simple chain, she would plied with people who have language language abilities of the chimpanzee. It
never be able to interpret the compound difficulties caused by brain damage. It is our hope that our findings will dispel
sentence with its deletions. The fact is may also be of benefit to the autistic such prejudices and lead to , new at-
that she interprets them correctly. If a child. tempts to teach suitable languages to
child were to carry out the instructions In assessing the results of the esperi- animals other than man.
in the same way, we would not hesitate ment with Sarah one must be careful not
The Authors theory. More exotic birthplace than
Shanghai, namely Aberdeen, S.D.”
- ANN JAMES PREMACK and DAVID
PRElMACK are respectively a free-lance
writer and professor of psychology at
Bibliography
the University of California at Santa SYNTACTIC STRUCTURES. Noam Chom-
Barbara. Mrs. Premack received her sky. Mouton & Co., 1957.
Sachelor’s degree at the University of THE GENESIS OF LANGUAGE. Edited by
\/linnesota and was an early member of F. Smith and G. A. Miller. The
,he ape-teaching group. “Born and M.I.T. Press, 1966.
*aised in Shanghai,” she writes. “Pas: BEHAVIOR OF NONWAN PRI~WTES:
donately devoted to ballet. Enjoy the VOLS. III-IV. Edited by Fred Stoll-
:ompany of three teen-age children, one nitz and AIlan M. Schrier. Academic
:errnan-shepherd puppy and a myna Press, 1971.
rird.” David Premack received his LANGUAGE IN CHIMP~~VZEET David Pre-
tachelor’s, master’s and doctor’s degrees mack in Science, Vol. 172, No, 3985,
rom the University of Minnesota. “In- pages 808-822; May 21,1971.
reduced to chimpanzee on first job at A FIRST LANGUAGE: THE EARLY STAGES.
‘erkes Laboratories of Primate BioIogy Roger Brown. Harvard University
1 Florida in 1955,” he writes. ‘Also Press, in press.
rork on reinforcement and learning
 Study Guide PreparedbyJOHNP.J.PINEL, UNIVERSITYOFBRITISH COLUMBIA

TEACHING LANGUAGE TO AN APE SQENTIFIC OFFPRINT 549

AMERXAN
Ann James Premack and David Premack OCTOBER 1972

I. SUMMARY

The authors have taught a chimpanzee named Sarah to read words in a more direct way. Subsequently, Sarah was trained
and write with variously shaped and colored pieces of plastic, in the uses of adjectives and conditional and interrogative
each representing a word. Why try to teach human language statements. In order to teach her the concepts of color, shape,
to an ape? The motive in this case was to better define the and size she was taught to identify members of the classes red
fundamental nature of language. Language is a general sys- and yellow, round and square, and large and small. Objects
tem of which human language is a particular, albeit remark- which varied in most dimensions but had one of the above
ably refined, form. By teaching language to a chimpanzee it properties in common were used for this purpose. In teaching
may be possible to identify those aspects of language that are class names many of the sentences were not “written” on the
uniquely human. Moreover, such experiments may reveal the board but were presented as hybrids consisting of a combin-
nature of the chimpanzee’s conceptual world and facilitate ation of plastic words and rea1 objects. For example, a typical
the comparative study of cognitive processes. hybrid sentence was “Yellow?” beside a banana. Her perfor-
The first step was to exploit knowledge that Sarah already mance showed that she was able to move with facility from
had; they mapped out the social transaction of giving, which symbols for objects to the real objects.
is something that the chimpanzee does both in nature and in Was Sarah able to think in her new language? Could she
the laboratory. In order to map out the entire transaction of store information using the plastic words and use it to solve
giving, the animal had to distinguish agents from objects, problems that she could not have otherwise solved? Addi-
agents from one another, and objects from one another. Sarah tional research is required for unequivocal answers to these
initially had to put a pink plastic square on a “language questions, but several aspects of Sarah’s performance indi-
board” mounted on the side of her cage in order to receive a cate that the answers may be a qualified yes. For Sarah to
slice of banana. Later several other fruits, the verb “give,” and match the word “apple” with an actual apple indicates that
the plastic words that named each of them were introduced. she knows the meaning of the word, but it does not mean that
To be certain that Sarah knew the meaning of “give” it was she can think apple when she is presented with the word
necessary to contrast “give” with other verbs. When she put alone. The ability to achieve such mental representation is
“wash apple” on the board, she did not receive the apple; it important because it frees language from direct dependence
was placed in a bowl and washed. At this stage Sarah had to on the external world. The hint that Sarah could use words in
place two words, “give apple,” on the board to receive the the absence of their external referents came in a test where
apple. When recipients were named, three words were re- she was given a piece of fruit and two plastic words. While the
quired; identification of the donor required yet another word. task was to put the correct word for the fruit on the board, she
At every stage she had to observe the proper word sequence. frequently put up the wrong word. Subsequent tests indi-
At first Sarah learned all her words in the context of social cated that she was trying to communicate her preference in
exchange, but later, when she had learned the concepts of fruit. This strongly suggests that Sarah could generate mean-
“name of” and “not name of,” it was possible to introduce new ings of fruit from the symbols alone. .

II. GLOSSARY

autistic - individuals who indulge in wishful thinking or phantasy to a patho- constituent structure - the hierarchical organization of a sentence.
logical degree. dispkzcement - the ability to talk about things that are not currently present.
call system - a series of vocalizations used by some animals for intraspecific language board - a sheet of steel on which Sarah placed her magnetized
communication. words.

III. ESSAY STUDY QUESTIONS

1. What was the objective of the authors’ work and to what extent did they 6. A chimpanzee is capable of decomposing a complex object into features.
accomplish it? Explain.
2. Why did the authors teach Sarah to communicate with plastic chips rather 7. What evidence is there that Sarah could think and solve problems in plastic
than vocal patterns? Is this really language? word language? - -
3. Until animals are taught language it will not be possible to study their 8. How was Sarah trained to understand conditional statements?
cognitive processes for classifying stimuli, for storing and retrieving infor- 9. What procedures were used to train Sarah to form concepts of color, shape,
mation, and for problem solving. Discuss. and size?
4. In most psychological experiments many subjects are tested but the 10. How was Sarah trained to understand questions?
Premacks’ conclusions are based on data from only a single organism. Dis- 11. Sarah could have learned more words if the symbols were similar to the
CUSS. objects they represented. Why did the authors not use this strategy?
5. The test with the compound sentence was of considerable importance.
Explain.