Max-Planck-Institutfiir Psychiatrie, Miinchen

A NewTclestimulationTechniquefor the Study

ofSocialBehavior ofthe Squirrel Monkey
MANFRED MAURUS
Speciflc behavior elements and reactions such as vocalization, orientation, affective
reactions,genitalresponses, urination,salivation,changesinEKG,breathingfrequency,
and hippocampal activity and motoric responses could be elicited invarious com-
)I
binations by electrical brain stimulation. Some ofthese reactions are important ele-
~
ments ofsociallv signilicant behavior.
Inordcrtoinvestigatesuchquestionsas cansocialbehaviorbcelicitedbystimulation
l! of specific brain structuresandif so, howstimulatedandnon-stimulatedanimalsreact
I
within their social group in various situations, all animals must be free moving. In
addition, itis important that a receiving device does not produce major changes in
i. Ii
appearance ofthe animals.
q
A remote control receiver including batteries that can be worn on the head ofa
500 g to 900g animalwitha brain weightof26 g must be small andlightweight. In
order to reduce electrolytic brainlesions from repeated stimulation it is necessary to
use a biphasicpulse.
Anapparatus whichfulfills these basic requirements willbe described.
Twoindependenttransmitters controlthenegativeandpositivepulse, respectively,
ofa biphasic stimulus ofthe calibrated receiver. The transmitters are controlled by
square wave generators. Each pulse lasts as long as the appropriate transmitter runs,
thus frequency, pulse duration, train durationand pulse delay can be entirely con-
trolledat the transmitter. Theamplitudemust, for thetime being, be manuallyseton
the receiver butis independentofthefield strength. Full currentflows over the elec-
trodes only when the fleld is above the minimum operating level ofthe receiver.
otherwise there is nomeasurable current flow. Leakagecurrentofthe entire receiver
is ca.4mA.Thelifeforasetofbatteriesis aboutoneyear.Thebatterieslastca. 100hrs.
withcontinuousbiphasicstimulation(30cps, I mspulsedurationandmaximumampli-
tude). The entire receiver including batteries, battery case and twenty-pole plug for
electrodeconnections weighs ca.II g.
:

'I
DepartmmtofZoology, UniversityofWisconsin
Madison,Wisconsin53706
CulturalAcquisition ofa SpecifIc Learned Response
Among Rhesus Monkeys
GORDON R.STEPHENSON"
A. Introduction
KAWAMURA \1959) andlaterKAWAI (1963),intheirdiscussion of field studies ofJapa-
nese macaques, described behaviors which were peculiar to certain troops ofthese
monkeys. Foodhabits, usually learned byinfants from their mothers, were found to
bespecificforeachtroop.Within atroop, theremaybesub-populationswhichexhibit
particular habits. For example. in 1956, females and juveniles of the provisioned
KoshimatroopexhibitedSweetPotatoWashingBehavior,whileperipheraladolescent
males did not. KAWAI (1965) suggests that reduced social interaction between these
parts ofthe troop makes it unlikely these peripheral males will acquire Washing
Behavior. IrANI (1958) has referred to the above behaviors as culture.
WASHBURN and HAMBURG (1965), from observations by DEVORE, describe an in-
cident which had profound and long term effects upon the subsequent behavior of
membersof ababoongroup. Priortotheshooting oftwomembers, the group could
easily be approached in a car. The authors comment, .. . it is very unlikely all the
animals in the group saw the shooting. .. buteight months later this group ...still
couldnot be approached, even though the animals must have seen cars almost daily
in theintervalii . Theexperienceof someoftheanimalshadapparentlybecomepartof
thewholegroup'sbehavior,aneventsuggestiveofculturaltransmission.
JOSLIN, FLETCHER and EMLEN (1965) examined the reactions ofrhesus monkeys to
snakes. Theyfound thatwild-reared animals wouldnot reach towardalivesnake or
snake-like model to retrieve a food reward, while lab-reared animals, in the same
situation, showed little hesitancy in reaching for the reward. In explanation, they
proposed that,in thewild,young rhesus monkeys learn tofear snakes from the reac-
tionsof oldermembersofthetrooptowhichtheybelong;lab-rearedmonkeys,never
exposedtosnakes orto adultreactions towardsnakes, donotacquire theresponse. In
thisview,thetransferofinformationunderlyingresponsesof thewild-rearedmonkeys
couldbe regardedas culturalinthesense ofItani.
I regard culture as the constellation ofbehaviors characteristic ofa single social
group, behaviors which are transgenerational and socially learned by individuals as
members ofthe group. The present report describes an attempt to apply controlled
laboratory methods to a social learning situation suggested by the above field and
GORDON R . STEPHENSON
laboratory observations. My particular interest was whether the learned avoidance
behavior of a conditioned monkey toward a conditioning object could induce a
lasting effect on the behavior of a second monkey toward that same object.
My warmest thanks go to Professor John T. EMLEN for discussion and guidance
which led to posing the question, to Professor Harry F. HARLOW, Dr. Bruce K. ALEXAN-
DER and personnel of the Primate Laboratory, Department of Psychology, University
ofWisconsin, for guidance in expediting the study, and to my wife, Gayle, for assist-
ance in gathering the data and patience during their analysis.
B. Method
l. Subjects
Eight lab-reared rhesus monkeys (4 males and 4 females, Macaca lI1ulatta) about
Cluee years of age served as subjects. They had been raised with their mothers and had
had much play experience with each other during their first 2% years oflife. Primate
>:
Laboratory personnel regarded them as socially normal. No conditioning experiments
tI,
had been performed on these subjects prior to the study reported here.
-.1,
l. I
1j!
I
2. Apparatus
I
Subjects were taken by transport cage from their separate home cages in colony
rooms to the apparatus (Fig. r) in the testing room. White noise masked most of the
I extraneous noises in the test environment. Observations of each subject's behavior in
h
:
the presence of an object were made through a one-way window and recorded con-
tinuously with a keyboard activating nine channels of an Esterline-Angus recorder.
Data recorded were location of the subject and the object in the apparatus, incidence
of vocalization, and incidence and duration of visual exploration, ambulation, stereo-
typic movements, and manipulation of the object. When two subjects were observed
at once, a trained assistant simultaneously recorded the second subject's behavior with
similar keyboard. General comments and impressions were noted at the end of each
seSSIOn.
Only the data for manipulation of objects are considered in this report. Manipula-
tion ranged from apparently passive contact of the subject's hand with the object, to
active investigation, including mouthing and chewing. Objects were plastic kitchen
utensils (seive, cup, mustard jar, etc.) of different sizes and colors; all elicit similar
responses from the subjects under similar circumstances.
3. Procedure
Subiects were assigned to unisexual pairs. Each subject, either alone with an object,
or together with its partner and an object, was observed 111 sessions of [5 minutes
duration. A total of 45 test and control sessions was conducted with each subject. The
critical test sequences were towards the end of thIS series and contained four sessions as
follows:
Cultural AcqUIsition of a Specific Learned Response Among Rhesus Monkeys
28 1
OBJECr
AIR JET
Fig. r: Steel cage is which all sessions were conducted. The 7 x 2 X 2 ft. angle-iron frame was
covered with heavy duty hardware cloth except for one side of steel slats. Prior to a
entry into the test room and apparatus, one object was placed on the' board at the masonite end
I. One subject of a pair, while alone in the apparatus, was punished with an air
blast each time it started to manipulate an object. Two or three air blasts generally
sufficed to terminate all manipulation tendencies toward this object.
2. This subject was again observed alone with the object to test for retention of the
learned response.
3 The subject, now a demonstrator, was observed with its naive partner in the
presence of the conditioning object. No air blasting was done.
4 The naive subject was observed while alone with its partner's conditioning object.
The remaining sessions were conducted to gather control data.
An initial block of r6 sessions served to habituate subjects t() the test apparatus and
procedure , to determine the manipulation tendency of each individual, and to deter-
mine the effect of social interactIOn on an individual's manipulation tendency. Sub-
jects were exposed to either Object I or to Object 2 in these sessions. A subject was
exposed to Object I only when alone in the apparatus and was exposed to Object 2
both alone and when together with its partner. Each subject had four seSSIOns with
Object L six sessions alone with Object 2, and six sessions together with its partner and
Object 2.
GORDON R.STEPHENSON 282
Tocontrolfor thepossibilityof dailyvariationin asubject'stendencytomanipulate
anobject, andforanychangesinthistendencyduringthethreemonthsof study, each
subjectwasobserved whilealonewithObjectI in13 sessions systematicallyscheduled
throughtheentireperiodofstudy, bothbeforeandafterthecriticaltestsessions.Eight
of thesesessionswerescheduledso thateachsubjectwasinasinglesessionandallof the
subjects hadtheir session onthe same day. Tocontrolfor changes inthe effect ofthe
presence ofasecond subject on asubject's manipulationtendency, additionalcontrol
data about each subject were gathered in two sessions alone with Object2 and four
sessions together with its partner and Object 2 scheduled over the remainder ofthe
,tudy.
Todetermine the effect ofnoveltyper se ona subject's manipulationtendency and
thepossibleeffects of its reactionstonovelty on asecondsubj ect' smanipulationtend-
ency,thesequenceof foursessionslistedaboveas thecriticaltestsequencrwasfIrSt run
withoutreinforcementas acontrolseries.Thiswas followedwithintwoweeks bythe
testseriesinwhichthenovelobj ectwasreinforcedforthedemonstratorsubjectwitha
blastofair as described above.
To increase the number oftests oftransfer ofinformation, the respective roles of
>: subject A andB ofapairwere assigned and conducted inonewayandthenreversed
1 (SetsIandIIinFig. 2a). Thisprocedurewasfollowed for boththenovelty(non-rein-
forced controlseries) and thereinforced obj ect tests senes (Fig.2b) .
Fig.2 : Sequence ofSessions withNovelObjects
a) Control: Response to a Novel Object
Set I Set II
.._ ----
A A A + B B B B B + A A
alone alone together alone alone alone together alone
- -_.__.__..
Object 3 Object 4
b) Test: Transmission ofa Learned Response Towarda Novel Obj ect
Set I Set II
A A A + B B B B B + A A
alone* alone together alone alone* alone together alone
Object 5 Obj ect 6
* air blast whenthesubject attemptsto manipulate the object.
Fig. 2: Format ofsequences ofsessions with novel objects for all pairs ofsubjects. (a) In the
controlsequence, subjectswereputwithnon-reinforcednovelobjects.(b) In thetestsequence,
thedemonstratOrsubjectwas punishedwithairblasts at themomentitwasabouttotouchthe
obj ect. Sincethedemonstratordidnotattempttomanipulatetheobjectduringitssecond ses-
sionwithit,nofurther air blastswereadministered,andthe demonstratorwasconsideredsuc-
cessfully conditioned to avoid this specific object
CulturalAcquisitionofa SpecificLearnedResponseAmongRhesusMonkeys 283
C. Results
1. Controldata
In theinItialblockofr6sessions, malesmampulatect 8.5%andfemales mal11pulated
35.4%of thetotaltimeavailable.Averagesoverthis blockforindividualmalesranged
from 2.5% to 12.9% and for individual females from 22.4% to 58.1%. This sex
difference inmanipulationtendencies was maintained throughout thestudy
A subject' smanipulationtendency, relative to the tendencies ofthe othersubject'>.
remained similar. session to session, throughout the study. Subjects were ranked tor
theiramountof manipulationineachof theten sessionsintheinitialblockwhilealone
with either Obj ect I orObj ect ::.. Concordance- ofthese rankings was tested by the
Kendall method (Siegel, 1956) and was found Significant (p < 0.01). Rankings of
subjectsineachof eightoftheremaining 13 sessions alonewithObject r also showed
concordance(p < o.or) .
Whilethere was daily variationH1 a subjectsmanipulatIOn tendency, each subject
tended to haveits high and lowamounts ofmanipulation on the same daysthat the
othersubj ects hadtheirrespectivehighs andlows Tensessions ofeach subiect when
alonewithObject I wererankedfor amountofmal11pulatlon. Concordance ofthese
rankingsforallof thesubjectswas foundslgmflcant(p < o.or).
The manipulation tendency ofa subject was afTected by the presence of second
subject.FemalesmanipulatedObject2abouttwiceas muchwhentogether(62.7%)as
when alone (30.1%) . Individually, every female manipulated Object 2 more when
with its partner than when alone. Males manipulated Object 2 more when together
with another subject(11.4'0) than when alone (7.6%). As individuals, twomales in-
creased and twomales decreased theiramountofmanipulationwhenwiththeir part-
ners fromtheamountwhentheywerealone.
Therelativemanipulationtendencyof thesubjectswhenwiththeirpartnersdidnot
differ from that when alone. The rankings ofsubjects for total manipulation of
Object2whilealone andwhiletogetherwereverysimilar (SpearmanRankCorrela-
tionCoefficient rs = 0.95, P < o.or).
Subjectsmanipulatednovelobj ects morethanfamiliar objects.Whenallsessionsfor
each subjectwhilealonewith an objectwererankedfor amountofmanipulation, the
foursessionswhenasubject was alonewithan objectfor thefirst timewerenotuni-
formlydestributedovertheranking,buttendedtobeinthe highpartofit(X2 = 10.75,
P< 0.02). When together-sessions were ranked for amount of manipulation, the
distributionof sessions withnovelobjectswasmoreuniformovertheranking.
Subjectsmanipulatednovelobjects more ofthe available timewhen togetherwith
their partners than when alone. Females manipulated such objects 65.7% when to-
gether and 38.8% while alone.Males manipulated suchobiects38.7%when together
and 32.9% whenalone.
Noveltyhad more effectonmales thanon females inthe together-sessions.Among
femal es,thepercentof availabletimespentmanipulatingwasaboutthesamefornovel
objects (65.7%) as for familiar objects (67.4%) . Among males, novel objects were
manipulated more 3 9 . 2 ~ ~ ) thanfamiliar obj ects (13.r%).
GORDON R.STEPHENSON 284
2. Testsfor transfer of information
Comparison of thetimecourseof manipulationof anobj ectduringthefoutsessions
of thecrIticalte5tsequencewithareinforced novelobiectandthefoursessions ofthe
control "equence wIth a non-remforced novel object is presented for the males in
figures 3 and 4 and for the females in figures 5 and 6. In the critical test sequence,
designed to test for transfer ofa learned response, the learned response toward the
reinforced object was transferred with lasting consequences from the demonstrator
subjecttothenaivepartnerinthreecases;allof theminvolvedmales(Fig. 3, SetsIand
II; Fig.4, SetIJ. Inthreeothercases, allfemal es, theinformationwas nottransferred,
but instead. the avoidance response ofthe demonstrator was gradually extinguished
as thenon-conditi.onednaivepartnerproceededtomanipulatetheobject(Fig. ), SetI,
FIg.6. Sets1andm.Datafortheremainingtwosequenceswereinconclusive.
Allsubjects appearedtohavebeensuccessfully conditionedto avoidthereinforced
object during their first session as demonstrator in the critical test sequences; no air
blastswerereqUIredtopreventmanipulationbythedemonstratorinthesecondsession
whileitwas alone withtheobject.Behaviorinthethirdsession, whiletheconditioned
df'monstrators were pairedwith their n;1ive partners, varied according to the sex of
thesubjectsas notedabove. Intwoof themalesequences, thedemonstratorexhibited
what wasmterpretedas athreatfacial expressionwhileinafear posture. As thenaive
parmer (Subject AI) approached the conditioning object in the third sequence with
males(Fig. 3, SetII),thebehavoirof thedemonstrator(SubjectBl ) wasmoredefmite
\1 as described below:
The demonstrator subj ect, upon being placed in the apparatus, retired to the end
"j
furthestfromhis conditioningobject. Thenaivesubject,whenplacedintotheappara-
J tus tenseconds later, flISt went to and mounted the demonstrator subject, thenpro-
,,',
ceededtotheobjectwithouthesitation,lookedatitandslowlyreachedouttowardsit.
At this moment, the demonstrator subject came from the far end ofthe apparatus,
grabbedthenaivesubjectbytheBanks andpulledhimaway. Theobjectremainedin
placeontheboard. Thedemonstratorthenturnedandsexuallypresentedtothenaive
partner. The naive partner, although it subsequently approached the object several
times, didnottouchtheobjectin theremaining 14 minutesofthesession.
In thethreecases involving females, the demonstratorsubjectremainedawayfrom
the obj ect, butmadenoovertmoves tointerfere.
D. Discussion
Datafromsixof eightcriticaltestsequencesdesignedtotestfortransferof intorma-
between socially normallab-reared monkeys demonstrate that behavior ofone
memberofa pair toward anovelobject can alter the behavior ofits partner toward
that objectwithlasting consequences.
In thethreefemalesequences,theinBuenceof thenaivesubjectonitsdemonstrator
partner appeared to be through simple observational learning (c.f. HALL, 1963) in
which the demonstrator, known to have been successfully conditioned to avoid a
specific obj ect, observed its partner manipulating that object, and subsequently, also
manipulated the obj ect. hANI'S (1958, p. 86) originalreferentfor cultureinmonkeys
CulturalAcquisitionofa Specific LearnedR esponseAmong R hesus Monkeys 285
r OGE TH ER NAJV(
_. ]1 Ifii:"''"i.1.IIII ,>e"'",";'":II.
L
,::C I
A, A,
::lIiliL.L1Ju=
;::C"=JI .. I
-...J
o 81 81
! [ ! I ! [ [ ! ! I !!I ! 1 t ! III I j ! I ! 1 ( ! I !! ...1....lLLJ I I I I I I I I \ I I !) I "I ! I I \ \ I 1 ! ! I I I
4 8 12 16 j 4 8 12 16 I 4 f:I 12 [6 I 4 8 12 16
SUCCESSIV E MIN UTE I NTER VA LS
Fig. 3:ManipulationofnovelobjectsbymalesubjectsA, andBt . Histogramsrepresentamount
ofmanipulationin each one-minute interval ofthe session. Time available formanipulation
in the first andsixteenthintervals ofthesession totals one minute. Stars indicate blasts ofair
to punishattempts to manipulatea specific object
r--' I
IJI .. 11i[IIIW t,J] I,
oL11 I I
,,01--
A,
I
I
. I n'
o <
i


- A,
_ ___, J A2+ Bz
l' *
:r: i::i a Az
rL
"::WIII. II
-l
.LU.... '-----.,-:.,,------J
':l... " I[ "' -J
l
o B2 B2
...l_.L.:_--L.LL i .-LLLL_:_"':' I II I
I 4 S 12 16 12 16 12 16
SUCCESSI VE MINUT E INTERVALS
Fig. 4: Manipulation ofnovel objects by male subjects A2 and B2 While subjectB did not
manipulatethenon-reinforcednovelobjectmuchas naivesubjectinSetI, hewouldnottouch
his demonstrator partner's conditioning object
NAIVE '-
r - ---
J
- l
L____I
-8,- .

II

A,
. 1 1 II1 I 1 I I I
12 16
GORDON R.STEPHENSON 286
TOGETHER
I
It I
- 1m
I I
'::J . ' Ie
<
Ill l.
j
i7+- A3 A3
,::11111.1" II II 1_
,::1 .. 11 "- I Hji
IIII1
a5 a, 9 ,+A3
! 11I 1 11 1 11!!!!1 1111 1 1111 I 1 I 1I II II II I I 1 1 1!! [;[ '111 1111111 1
12 16 S 12 16 I 12 16 4 S 12 J6
SUCCESSIVE MINUTE INTERVALS
Fig.5: Manipulation ofnovel objects by female subjects .1,andB,
NA I VE
OEMONSTRATOR I I DEMONSTRATOR I I' 1 I I . i
'''T ,
11
'H"
1
' , , I ,li il, , . ,
"
of
L II , .
I !
UJ, ,.
-'
1
;'. lm -"'.Lru
,.g I,u'.. 'I I J[ JI.
::. :::0 5 *
, * I A. A. t
0 1* A.
u

,::c l.
' :: [ " I
o -
a. a.
! ! J I I ' I I [ I I II ! ! [ I I t 1 J [ I I I 1 II I I I I I ! I I ' [
12 )6 12 16 I 4 8 12 16 16
SUCCESSIVE M[ NUT E INT ERVALS
Fig. 6: Manipulation ofnovel objects by female subjectsA, and B4
was the constellation offood habits acquired by Japanese macaques through a social
learningprocesscloselysimilartothis.Newfoods(FRISCH, 1959) as wellas newobjects
(MENZEL, I966) appeartohavesomenegativevalencefortheseferalmonkeys.Through
observationallearning, some negativevalences associatedwithspecific objects(edible
as wellas others) can be overcome.
CulturalAcquisitionofaSpecific Learned ResponseAmong RhesusMonkeys 287
In the three male sequences, the demonstrator subject altered thenaive subject's
behaviortowarda novelobjectwhichhad previously beennegativelyreinforcedfor
thedemonstrator. Controldataindicated thateachsubjecthadacharacteristiclevelof
manipulationwhichpersistedwithlittlevariationthroughoutthestudy,thatthetend-
encytomanipulatewas enhancedby noveltyoftheobject, andthat, for twoofthese
subjects, presence ofa second subject further enhanced the tendencyofeach subject
to manipulate. In view ofthis control data, the observed alteration ofthe naive' s
behaviorwasunexpected. Iconsidertransferofinformationintheseinstancesdifferent
from the observationallearning described above for females, and homologous with
admonitionintheacquisitionofculturebyhumansas discussed by E.T.HALL (1959).
A demonstratorsubjectpullinghis naive partnerawayfrom the demonstrator' s con-
ditioning object is admonition. Similarly, I regard the interactions between feral
Japanese macaques in response to novel objects described by MENZEL (I966, p. I34)
asadmonition: Onmorethan6 occasions, females withinfantsshowed clearly that
somecautionunderlies theirindifference [toobjects] whentheypulled theiroffspring
awayfromanapproach toanobject.
The case is more difftcult to interpret when no physical contact is observed. No
vocalizationwas heardinthetogether-sessionswithnovelobjects.Experimentalstud-
ies(MILLER etal; 1959, I966) suggestthatcommunicationof affectcan bemediatedin
rhesus monkeysthroughfacialexpressionandposturealone.Withrespecttothenovel
object in my studies, however. the referent ofcommunication was nonsocial and
specific, andremainedspecificinsubsequenttestsessions whenthesubjectwas alone.
Less overtbehaviorsappeartohavebeenoperatingandresultedinamoresubtleform
ofadmonition.
Recognition ofadmonition as a form ofcommunication extends the modes of
acquisition ofculture which heretofore have been considered available to species of
M acaca andFapia. Admonitioncanoperatebetweenindividuals,andcouldconceivably
operate between a single individual and the rest ofthe group, particularly from a
dominantindividual to subordinates.
Admonition could be operating complementarily with observational learning in
Imanishi's (1957, p. 2) notion ofenculturation. For example, IMINISH! (1960) and
MIYADI (1964) concludethatthestatus ofa givenindividualinthesocial hierarchyof
aJapanesemonkeytroopis atleastpartiallydependentonandlearnedbyothermon-
keys inthetroopinrelationto thesocialstatus ofthatindividual'smother. HALL and
GOSWELL (I964) discuss suchacquisitionintermsof sociallearning. Theacquisition of
fear responseswithinferal troops, as wellas tostimuliexternal to thetroops, couldbe
mediated throughobservationallearningincombinationwithadmonition.
Admonitionin thetests describedinthis reportwas observedonlyinmales; obser-
vationallearning, onlyinfemal es. Thesamplesizeis smallandnosharp sexdifference
shouldbeinferred. Itmaybepertinenttonotethat,inthesample,femaleshadagreater
tendencytomanipulatethan didmales. Fromrankcorrelation, thetendencies of indi-
viduals to manipulateobjectsappearedto berelativelyconstantthroughoutthestudy.
Manipulation ofthe conditioning object by the naive partner in the critical test se-
quence while alone subsequent to the together-session was highly correlated withits
tendency to manipulate objects in the control sessions. Perhaps punishment during
conditioningwasnotsevere enoughto preventobservationallearning byorto educe
admonitionfrom females inthis study.
288 GORDON R. STEPHENSON
E. Summary
The- social learning situation suggested by field and laboratory observatlons to be
involved in the acquisition of culture bv individuals in groups of macaque monkeys
studied by contr()lled laboratory method,. In six of eight critical tests sequences
with pam of sociall)' normal lab-reared rhesus monkeys. behavior of one subiect
toward an object altered. with lasting conse-quences. the independent behavior of a
second subject toward that same object. Observational learning and admonition are
as two types of information transfer between subjects which mediate
:he- acquisition of culture in monkeys.
Bibliography
fRISCH.J.E.. I959: Research on Primate Behavior in Japan. American Anthropologist 61 (4).
584-596.
HALL. E. T .. I959: The Silent Language. Double Day and Company, Inc., New York.
HALL. K. R . L.. 1963' Observational Learning on Monkeys and Apes. British]. of Psychology
54 (3). 201 - 226.
HALL. K. R. L. and M.J. GOSWELL, 1964: Aspects of Social Learning in Captive Patas Mon-
I keys. Primates 5 (3-4). 59-70 .
IMANISHl. K., 1957: Identification: a Process of Enculturation in the Subhuman Society of
Macaca fuscata. Primates r (1), 1-20.
IMANISHl, K .. 1960: Social Organization of Subhuman Primates in Their Natural Habitat.
:
q
J
'I Current Anthropology I (5-6). 393-407.
, Ii I
.
ITANI, J.. 1958: On the Acquisition and Propagation of a New Food Habit in the Natural
Group of the Japanese Monkey at Takasaki-Yama. Primates 1 (2), 84-98.
JOSLIN, J.. H. FLETCHER and J. T. EMLEN, 1965: A Comparison of the Response to Snakes of
Lab- and Wild-reared Rhesus Monkeys. Animal Behaviour 12 (2-3),348-352.
KAWAI. M., 1963: On the Newly-acquired Behaviors of the Natural Troop ofJapanese Mon_
keys on Koshima Island. Primates 4 (I). II3-II5.
KAWAI, M ., 1965: Newly-acquired Pre-cultural Behavior of the Natural Troop of Japanese
Monkeys on Koshima Island. Primates 6 (1), 1-30.
KAWAMVRA, S., 1959: The Process of Sub-cultural Propagation among Japanese Monkeys.
Primates 2 (I), 43-54.
MENZEL, E.W., 1966: Responsiveness to Objects in Free-ranging Japanese Monkeys. Beha-
viour 26 (1- 2), 130-150.
MILLER, R.E.,J. H. BANKS and H. KUWAHARA, 1966: The Communication ofAffects in Mon-
keys : Cooperative Reward Conditioning. J. Genetic Psychology 108, 121-134.
MILLER, R . E., J. V. MURPHY and J. A. MIRSKY 1959: Relevance of Facial Expression and
Posture as Cues in Communication of Affect Between Monkeys. AMA Arch. of General
Psychiatry I. 480-488.
MIYADI, D ., 1964: Social Life ofJapanese Monkeys. Science J 43 (3608), 783- 786.
SIEGEL. S., 1956: Nonparametric Statistics for the Behavioral Sciences. McGraw-Hill Book
Company, Inc.. New York.
WASHBURN, S.L. and D.A. HAMBURG, 1965: The Stndy of Primate Behavior . In Primate
Behavior. Field Studies of Monkeys and Apes. (I. Devore, cd.) Holt, Rinehart and Winston.
New York.
Institute for Comparative Biology and San Diego State College
The Learning Skills of Primates: a Comparative Study
of Apes and Monkeys
DUANE M. RUMBAUGH1 and CAROL MCCORMACK
The purpose of this paper is to assess the learning abilities of groups of great apes, gib-
bons, macaques. and squirrel monkeys on a variety of visual discrimination tasks which
range in difficulty from simple discrimination learning through object-quality learning
set (LS ; HARLOW, I949) , discrimination reversal (DR) , and oddity concept (OC) .
Though there exists a vast literature on the subject ofprimate phylogeny as it relates to
capacity for learnmg a wide vanety ot complex tasks, there has been no study in which
the genera represented by the above listed groups have been directly compared within
a single experimental program. The major barriers to such an experiment have been
access to the animals in numbers and access to appropriate test facilities. Through the
cooperation of the San Diego Zoological Society, it has been possible for the authors
to overcome these barriers and to collect an extensive body of comparative primate
data, major portions of which constitute material for the present paper.
Method
Subjects (Ss)
All ape and macaque Ss are identified and coded in Table T, which also includes
summary data of adaptation and simple discrimination learning data collected in Ex-
periment I described below. All apes, except two chimpanzees (C-6 and C-rr) and
one orangutan (O-II), either had been on exhibit at the Children's Zoo, for various
periods of time prior to serving in the experiment, or had been raised as pets. the latter
being true ofall gibbons. None ofthe macaques had this kind ofhistory, which provid-
ed for close and sustained contact with people at an early time in life, but without ex-
ception they were well-adapted to being maintained at the Zoo before serving in the
program. Squirrel monkeys Ss (Saimiri sciureus) , data from which are reported in
Phase 3 of Experiment II described below, were the only Ss not tested at the Zoo. They
were tested in the laboratory at San Diego State College where they had been main-
tained several weeks prior to experimentation. All Saimiri Ss were adult males (captured
near Letecia, Colombia) estimated to be about three years old.
1 This study program supported by grants from the National Science Foundation to the
first author. Without the cooperation and help ofthe San Diego Zoo's staffand keepers, it would
have been impossible to conduct the studies herein reported. Professor AUSTIN RIESEN'S critical
comments of an earlier version of this paper are gratefully acknowledged.
'9 Primatologen r966