ORIGINAL PAPER

Perception of biological motion in common marmosets(Callithrix jacchus): by females only

J. Brown • G. Kaplan • L. J. Rogers •

G. Vallortigara

Received: 16 March 2009 / Revised: 6 December 2009 / Accepted: 6 December 2009 / Published online: 6 January 2010

� Springer-Verlag 2010

Abstract The ability to perceive biological motion (BM)

has been demonstrated in a number of species including

humans but the few studies of non-human primates have

been relatively inconclusive. We investigated whether

common marmosets (Callithrix jacchus) are able to per-

ceive biological motion, using a novel method to test non-

human primates. Marmosets (7 male and 7 female) were

trained to remove a cover from a container and look inside

it, revealing a computer screen. Then they were presented

with images on this computer screen consisting of a novel

BM pattern (a walking hen) and 4 manipulations of that

pattern (a static frame of this pattern and inverted, scram-

bled, and rotating versions of the pattern). The behavioural

responses of the marmosets were recorded and used to

assess discrimination between stimuli. BM was attended to

by females but not males, as shown by active inspection

behaviour, mainly movement of the head towards the

stimulus. Females paid significantly less attention to all of

the other stimuli. This indicates the females’ ability to

attend to biological motion. Females showed slightly more

attention to the inverted BM than to the static, scrambled,

and rotating patterns. The males were less attentive to all of

the stimuli than were the females and, unlike the females,

responded to all stimuli in a similar manner. This sex

difference could be due to an inability of males to recog-

nise BM altogether or to a lesser amount of curiosity.

Considered together with the findings of previous studies

on chicks and humans, the results of the present study

support the notion of a common mechanism across species

for the detection of BM.

Keywords Common marmoset � Biological motion �Sex difference � Moving dot patterns

Introduction

The ability to distinguish living organisms from other

objects in the environment is crucial for all animals and it is

an aspect of perception that must be performed accurately

and rapidly. We know relatively little of how different

species make these discriminations but distinguishing

specific types of movement seems to be essential, as first

shown in humans. Johansson (1973) discovered that human

actions presented as moving dots representing only the

motion of the major joints of the human body could be

identified veridically by human observers, and this per-

ceptual phenomenon has come to be known as biological

motion (BM) perception.

BM perception by humans is not restricted to the

detection of the motion of humans. Pavlova et al. (2001)

showed that humans can recognise dogs and birds pre-

sented as point-light displays and Mather and West (1993)

showed that BM patterns are recognised as depicting

moving animals even though the species cannot be iden-

tified. Added to this, the direction of locomotion can be

Electronic supplementary material The online version of thisarticle (doi:10.1007/s10071-009-0306-0) contains supplementarymaterial, which is available to authorized users.

J. Brown � G. Kaplan (&) � L. J. Rogers � G. Vallortigara

Centre for Neuroscience and Animal Behaviour,

School of Science and Technology, University of New England,

Armidale, NSW 2351, Australia

e-mail: gkaplan@une.edu.au

J. Brown

e-mail: julian.brown20@yahoo.com

G. Vallortigara

Centre for Mind/Brain Sciences, University of Trento,

Corso Bettini 31, 38068 Rovereto, Italy

123

Anim Cogn (2010) 13:555–564

DOI 10.1007/s10071-009-0306-0

extracted from spatially scrambled BM patterns repre-

senting no known species or action (Troje and Westhoff

2006). These findings suggest that BM perception in

humans involves a mechanism that detects BM in general

(i.e. is not species-specific). Moreover, this perceptual

ability appears to be functional in early life (Fox and

McDaniel 1982). Interestingly, Simion et al. (2008) found

that human infants as young as 2 days old showed a

spontaneous preference for BM over non-biological

motion. Since the BM pattern that the authors presented

was unfamiliar to the infants, these findings suggest the

existence of an inbuilt mechanism for the detection of BM

in general, which may be an evolutionary adaptation

serving to orientate the infant towards a potential caregiver.

Hence, it is possible that this perceptual mechanism might

be present in other species.

Recent findings have shown that BM perception is not

unique to humans. Blake (1993) successfully trained cats to

discriminate a BM pattern of a walking cat from a number

of patterns without biological motion (random motion,

Brownian motion, positionally scrambled motion, and

phase-scrambled motion). A number of studies (Dittrich

et al. 1998; Omori 1997; Omori and Watanabe 1996) have

found evidence for the existence of BM perception in

pigeons. Using the tendency of newly hatched precocial

birds to approach conspicuous stimuli, Regolin et al.

(2000) and Vallortigara and Regolin (2006) found evidence

of BM perception in chicks and Yamaguchi and Fujita

(1999) found evidence of it in quails. The most convincing

evidence for non-human animals having a BM perceptual

mechanism similar to that found in humans has come from

a study showing that newly hatched chicks lacking any

previous visual experience show a spontaneous preference

to approach BM patterns over patterns without biological

motion (Vallortigara et al. 2005). This preference was

expressed regardless of the species depicted (both a

walking hen stimulus and a walking cat stimulus were

used), thus indicating the presence of a general mechanism

for BM perception.

Considering the evolutionary distances between cats,

birds, and humans, these findings suggest that BM per-

ception may be an evolutionarily ancient mechanism.

Surprisingly though, the few results so far obtained

would suggest that non-human primates may not possess

the ability to distinguish BM. Tomonaga (2001) and

Parron et al. (2007), testing chimpanzees and baboons,

respectively, found that, although subjects could dis-

criminate BM from non-biological motion, they could not

necessarily recognise the BM stimuli veridically. How-

ever, it is possible that the methodological procedures

used in these studies on primates did not allow accurate

assessment of the subjects’ sensitivity to the biological

content of the stimuli. Both studies used conditioning

procedures and this may have influenced the stimulus

features on which the subjects based discrimination. In

the case of the Tomonaga (2001) study, the relative

familiarity of stimuli appeared to act as the basis for

discrimination, whereas in the Parron et al. (2007) study

different subconfigurations of the same BM pattern were

used for discrimination by different subjects. Considering

that Oram and Perrett (1994) have found single cells in

the superior temporal polysensory area of the macaque

temporal cortex that respond specifically to BM, the

ability of non-human primates to respond behaviourally

to BM deserves further research using a different

experimental approach.

The aim of the present study was, therefore, to investi-

gate the ability of common marmosets to perceive BM

using a new method. We tested marmosets using non-

rewarded trials in an apparatus modified to suit the

behavioural characteristics of the species.

Methodology

Subjects and housing

Seven male and seven female common marmosets (Calli-

thrix jacchus) were tested. The mean (±SD) age of females

was 6 ± 1.5 years (range from 2 to 14 years). The mean

age of males was 5.7 ± 1.4 years (range from 2 to

13 years). The marmosets were housed in same-sex family

groups of 2–4, apart from 3 that, due to social aggression,

were housed individually but in close proximity to others.

Each cage (1 9 2 9 2 m) contained nest boxes and several

structures for climbing and playing, such as tunnels,

branches, and perches. A series of runways linked the

home cages to indoor rooms (4 9 4 9 3 m) that contained

many structures for climbing and to outdoor cages

(1 9 1 9 2 m). The marmosets received natural light via

skylights, direct exposure to the sun in the outdoor cages

and they were exposed to ultraviolet light (350–390 nm) in

their home rooms for 30 min per day. Feeding took place

once daily between 12:00 and 14:00 h and food provided

was sufficient to last for 24 h. Lights in the rooms con-

taining the home cages were turned on at 07:00 h and off at

19:00 h and temperature was controlled between 18� and

30�C. For further details of care of the colony see Kaplan

and Rogers (1999).

The marmosets had continuous visual contact with

conspecifics and daily contact with human carers. They

also had access to outdoor cages that were in close prox-

imity to a small, open area where they could occasionally

see birds and other animals passing by. Most of the mar-

mosets had also had considerable previous experience with

visual stimuli in similar experimental settings, but the

556 Anim Cogn (2010) 13:555–564

123

stimuli were entirely different from the ones used in the

current paradigm (e.g. Clara et al. 2007).

All tests were conducted in the indoor rooms. The

marmoset being tested in the room could be observed

through a one-way mirror from a darkened ante-room, in

which the experimenter was located. The presentation of

stimuli was controlled from this ante-room.

Stimuli

All stimuli were unfamiliar to the marmosets and this

was important because it meant that the marmosets could

not rely on familiarity to distinguish between them but,

instead, had to rely on informational features specific to

each stimulus. Five stimuli were used and they included;

a walking hen pattern, a static frame of this pattern, and

inverted, scrambled, and rotating versions of the walking

hen. The walking hen pattern was chosen because none

of the marmosets had any previous experience with this

particular species, let alone its BM pattern, and it was

the purpose of this study to search for evidence sug-

gesting the presence of a perceptual mechanism for the

detection of BM in general. Furthermore, common

marmosets may be predisposed to respond to particular

BM patterns of animal species with which they have

interacted in the wild over evolutionary time. Therefore,

the use of a motion pattern of a natural predator or prey

item (the hen being neither to a marmoset) may reveal

only a response to a particular species rather than to BM

in general.

The animation sequences were obtained using the soft-

ware program Macromedia Director (Version 6.0) and they

consisted of sets of 13 bright yellow dots (95.71 candelas

[cd]/m2) seen against a black background (0.03 cd/m2).

The visual angle of each dot measured 1� 440 3800 at a

viewing distance of 16 cm. Animation sequences were

matched for average velocity (54 pixels/s) of each of the 13

dots. Each set of points of light occupied a window of

119 9 108 pixels on the bottom of the computer screen;

the actual visual angle of the window measured 35� 480 4500

(height) and 46� 330 2300 (width) at a viewing distance of

16 cm.

The walking hen animation was obtained by carefully

locating, frame by frame, each of the 13 points of light on

the main joints of the digitalized image of the video

recording of a real animal (Fig. 1; see also Video S1).

Twenty-three frames were required to cover an animal’s

entire step sequence, and then the digitalized sequence was

looped and projected onto a computer screen after sub-

traction of translation components. As a result, the display

was stationary in the central window of the screen descri-

bed previously, but moved as if the hen was walking on a

treadmill. All the other foil sequences were also produced

by looping a 23-frame animation.

The static frame was randomly selected from the

walking hen animation. Since it contained no motion

information, it was used to test the sensitivity of the

marmosets to the motion information contained in the

stimuli.

The inverted version of the walking hen (see Video S2)

presented the pattern in an unnatural orientation and, thus,

the dynamics of this pattern were unnatural (i.e. it did not

conform to locomotion of an animal under the force of

gravity). This stimulus was included to test whether the

dynamic features of a BM pattern that are imposed by

gravity are a necessary perceptual constraint in BM per-

ception by marmosets.

The scrambled hen display (see Video S4) was obtained

by consistently displacing each point of light in each frame

of the walking hen sequence by 1 cm (i.e., by a visual

angle of 3� 340 3400 at a viewing distance of 16 cm). Each

point could be displaced either up, down, right, or left, at

random. Although displaced compared to its position in the

walking hen display, each single point of light in the

scrambled hen animation retained the same motion char-

acteristics (i.e., the same trajectory and velocity) exhibited

by that point in the walking hen. As a result only the

reciprocal positions of the 13 points of light differed

between the walking and the scrambled hen animations.

The scrambled hen display even occupied the same win-

dow on the screen as the walking hen. Spatially rearranging

the starting position of each dot while maintaining indi-

vidual motion characteristics meant that global motion was

completely disrupted but local motion was maintained.

This resulted in the stimulus being biologically incoherent

while still containing some biologically relevant informa-

tion. Thus, this stimulus was used to test sensitivity of the

marmosets to the biological coherence of the motion

pattern.

Finally, there was a solid, rotating hen stimulus (see

Video S3) that was produced by randomly selecting a

single frame of the walking hen stimulus and moving it

rigidly about its vertical axis, thus producing a solid,

rotating, hen-like object that contained no biologically

relevant motion information. It was, therefore, used to test

the marmosets’ sensitivity to the motion of a biological

organism.

These stimuli were presented on a vertically placed

400 mm Diamondtron NF Diamond View monitor with a

refresh rate of 75 Hz.

During presentation of the stimuli, each dot returned to

its starting position at the end of its sequence of movement.

This allowed individual stimuli to be run in a continuous

loop for an entire session.

Anim Cogn (2010) 13:555–564 557

123

Apparatus

The apparatus (Fig. 2) consisted of the computer monitor

(400-mm screen). Its screen was covered completely by a

sheet of black cardboard, except for a circular area 80 mm

in diameter located in the centre of the lower half of the

screen. The stimuli (controlled from outside the testing

room by a laptop computer) were displayed on the monitor

within the circular area. A hollow viewing container

without a base or top (150-mm-long cardboard cylinder,

100 mm in diameter) was fitted over the visible area of the

screen on one end and, on the other end, we made a

removable lid to cover the cylinder completely. Removal

of the cover was thus necessary to view the screen. Prior to

testing, the marmosets were trained to remove this cover in

search of food.

The apparatus was placed on a small table (75 9 42 cm)

70 cm from the ground. In front of the apparatus, on the

table, a small wooden platform (30 9 32 cm) covered in

hessian sacking was placed. The apparatus was located in

the indoor room and the marmoset being tested could be

viewed through a one-way mirror. The apparatus was also

covered in hessian to reduce interest in its form.

Experimental sessions were recorded by three video

cameras. There were two Panasonic SDR-H40 SD/HDD

video cameras, one placed above the apparatus and the

other placed to one side of the apparatus. The third camera

was a miniature camera (USB PC Camera MPC-003) and

this was placed inside the viewing container, directly below

and in front of the screen so that it did not obscure the

subject’s view of the screen. It recorded the face of the

marmoset as it looked into the container.

Testing procedure

Prior to testing, the marmosets were trained in isolation to

remove the lid from the viewing container in search of

mealworms. Training involved a stepwise process in which

the criterion to be met before advancement to the next

stage was successful retrieval of a mealworm. Stage 1, the

marmoset observed as a mealworm was placed inside the

viewing container, which was left uncovered. Stage 2 was

the same as stage 1 except that the cardboard cover was

placed over the container as the marmoset approached it

but before it had removed the worm (i.e. so that the mar-

moset had seen the worm being placed in the container

before it had to remove the cover to retrieve it). Stage 3, the

worm was placed in the container and covered prior to the

marmoset being allowed into the testing room. Once a

marmoset had successfully completed stage 3, it was

considered to be trained and ready for participation in the

experiment.

During testing each marmoset was tested alone in seven

sessions (one session for each stimulus and two training

sessions). Individual marmosets experienced a period of

24–50 h between each session to avoid loss of motivation

Fig. 1 The figure shows a

sequence of frames taken from

the walking hen biological

motion stimulus depicting a

single step of the hen

558 Anim Cogn (2010) 13:555–564

123

to look inside the container. Testing occurred between

09:00 and 12:00 h each day. Each stimulus consisted of

moving dots, each dot returning to its starting position at

the end of its sequence of movement. This allowed indi-

vidual stimuli to be run in a continuous loop for an entire

session. The order of presentation was randomised. Two

training sessions (the computer monitor was turned off and

a food reward was placed inside the container) were ran-

domly inserted between the testing sessions to maintain the

motivation of the marmosets to remove the lid from the

cylinder and look inside it.

The testing procedure followed the following steps: (1)

Depending on the type of session the computer monitor

was either switched on to display one of the stimuli

(testing session) or switched off and food was placed

inside the viewing container (training session). (2) A

cover was placed on the viewing container and all three

cameras were switched on. (3) Then, the marmoset was

allowed to enter the testing room and it was free to

inspect the room and the testing apparatus. (4) If a mar-

moset did not approach the apparatus and had not

removed the cover after 30 min in the room, the session

ended and was not recorded. A session was recorded

when the marmoset voluntarily removed the cover and

looked into the container. The recorded session ended

when the marmoset left the platform.

Scoring

All sessions were video recorded, and the behaviour was

then scored from the digitally recorded footage using

frame-by-frame analysis from the overhead and side cam-

eras (the miniature camera was used to increase the accu-

racy of the judgment of gaze direction).

Viewing (looking at the screen) was scored when the

gaze was (1) directed towards the visible portion of the

screen inside the viewing container and (2) not obscured by

the viewing container.

Fig. 2 The figure (above)

shows the apparatus from the

side. The sequence of images

(below) shows a marmoset (1)

approaching the apparatus (2)

removing the cover from the (3)

viewing container, and (4)

peering inside

Anim Cogn (2010) 13:555–564 559

123

Active screen inspection, as opposed to simply looking

at the screen, was said to occur when the marmoset repo-

sitioned its head relative to the screen while maintaining

visual contact with the screen. Active investigation was a

combination of three types of behaviour; concentrated

inspection (the head moved perpendicular to the screen in a

forward or backward direction relative to the screen, with

the face sometimes being brought to, and remaining, within

1–2 cm of the screen), parallax (the head moved parallel to

the screen in a left or right direction relative to the screen),

and head-cocking (the head was rotated to various degrees

about the rostro-caudal axis of the head; Kaplan and

Rogers 2006; Rogers et al. 1993). Scores were recorded in

two ways: (1) the number of events of active screen

inspection behaviour (i.e. a score for each instance of each

type of component behaviour) and (2) the total time of

active screen inspection.

Searching around the outside of the container was

recorded when a marmoset searched the area in close

proximity to, but outside, the viewing container as judged

by gaze direction (i.e. occurring when the marmoset’s gaze

was directed towards the black cardboard that covered the

area of computer monitor outside of the viewing container

or to the external surface of the viewing container).

Direction of gaze was judged by determining the hori-

zontal and vertical orientation of the marmoset’s head. The

separate views of the three different cameras were com-

pared to determine an accurate score.

Inter-rater reliability was assessed from the scores of

three raters, one of whom was condition blind. The intra-

class correlations for the duration and number of events of

viewing behaviour were 0.96 and 0.87, respectively, and

the intraclass correlations for the duration and number of

events of active inspection were 0.86 and 0.88,

respectively.

Statistical procedure

The data were first tested for normality. Transformations

were used in an attempt to normalise data that were

skewed. First, the scores for each type of behaviour were

analysed to see whether there was an effect of order of

presentation of the stimuli, using a repeated measures

ANOVA for normally distributed data, and Friedman tests

for data that were skewed. Only if no significant effect of

order of presentation was found were the scores analysed

further for an effect of stimulus, using a repeated measures

ANOVA for normally distributed data (post hoc ANOVAs

and t-tests were also applied), and Friedman tests for

skewed data (post hoc Wilcoxon signed ranks tests were

also applied). The alpha value was 0.05. The statistical

program used was SPSS 17.0.

Results

Training

The mean number of trials for all marmosets to complete

training in the cover removal task was 7.8 ± 0.9 and the

mean number of trials for males and females separately

was 9.0 ± 1.6 and 6.7 ± 0.9, respectively. There was no

significant difference in the amount of training required

by males and females to learn this task (t9 = 1.25;

P = 0.24).

Time spent viewing the stimuli

The effect of order of stimulus presentation on viewing

time was found to be significant (x2 = 12.647; P = 0.01).

This effect reached statistical significance between the first

and second presentation (Z = -2.355a; P = 0.02) but not

between the second and third (Z = -0.565a; P = 0.57),

third and fourth (Z = -0.471b; P = 0.64), or fourth and

fifth (Z = -1.445b; P 0.15). Viewing was not analysed

further due to this effect of order.

Duration of active inspection of the stimuli

There was no significant main effect of order (F(4,48) =

0.802; P = 0.53) on the duration of active screen

inspection. Hence, this behaviour was further analysed

to test for an effect of stimulus (repeated measure) and

sex. No significant main effect of sex was found

(F(1,12) = 1.650; P = 0.22). However, a significant main

effect of stimulus was found (F(4,48) = 4.533; P = 0.003)

and a significant interaction between stimulus and sex

(F(4,48) = 3.530; P = 0.01). Separate ANOVAs were

then performed for the males and females. These ANO-

VAs revealed a significant main effect of stimulus for

females (F(4,24) = 7.052; P = 0.001) but not for males

(F(4,24) = 0.557; P = 0.696) (see Fig. 3). Based on pre-

vious studies with chicks (e.g. Vallortigara et al. 2005)

and humans (Simion et al. 2008), it was expected that the

BM stimulus would elicit a different response to the

other four stimuli. An ANOVA limited to the static,

inverted, scrambled, and rotating stimuli revealed no

significant effect of stimulus for the females

(F(3,18) = 1.907; P = 0.165). Therefore, the heterogeneity

observed in the duration of active screen inspection

preformed by the females was due to the BM stimulus

(see Fig. 3a). It was also found that males spent signifi-

cantly less time engaged in active inspection of the BM

stimulus than did females (two-tailed t-test, t12 =

-2.172; P = 0.05).

560 Anim Cogn (2010) 13:555–564

123

Number of events of active inspection of the stimuli

As there was no significant effect of order on the number of

events of inspection behaviour (F(4,48) = 1.005; P = 0.41),

the effect of stimulus was tested. No significant main effect

of sex was found (F(1,12) = 2.292; P = 0.16). However, a

significant main effect of stimulus (F(4,48) = 4.163;

P = 0.01) and a significant interaction between stimulus

and sex (F(4,48) = 2.856; P = 0.03) was found. The pattern

of results was the same as for duration of inspection

behaviour. ANOVAs revealed that there was a significant

main effect of stimulus for females (F(4,24) = 6.420;

P = 0.001) but not for males (F(4,24) = 0.802; P = 0.536).

Further analysis using an ANOVA limited to the static,

inverted, scrambled, and rotating stimuli revealed no sig-

nificant main effect of stimulus for the females

(F(3,18) = 2.055; P = 0.142), showing again that the BM

stimulus elicited a different response than the other four

stimuli.

Components of the inspection behaviour

The three components of active screen inspection by the

females were analysed separately (Fig. 4). Head-cocking

contributed relatively little to this composite score.

Approximately the same number of events of concentrated

inspection and parallax head movement were performed on

viewing (determined using two-tailed paired-samples

t-tests) the static (t6 = 1.00; P = 0.36), inverted (t6 =

0.00; P = 1.00), scrambled (t6 = -1.00; P = 0.36), and

rotating stimuli (t6 = -1.00; P = 0.36). However, the BM

stimulus elicited significantly more concentrated inspection

than parallax head movement (t6 = 3.33; P = 0.02).

ANOVAs revealed that there was a significant main effect

of stimulus on the number of events of concentrated

inspection when all stimuli where included in the analysis

(F(4,24) = 7.331; P = 0.001), but not when BM was

excluded (F(3,18) = 1.432; P = 0.266). It should be noted

that 90% of concentrated inspection consisted of move-

ment towards the screen or the face being brought to within

1–2 cms of the screen. It is also interesting to note that the

females performed significantly more parallax head

movements during presentation of the inverted stimulus

than during presentation of the static stimulus (t6 =

-2.657; P = 0.04).

Searching outside the container

There was no significant main effect of order (F(4,48) =

1.033; P = 0.40) on the amount of time that marmosets

spent searching around the outside of the container. As a

result, the effect of stimulus was tested but there was no

significant main effect of stimulus (F(4,48) = 0.792;

P = 0.54).

Discussion

The scores of time spent viewing the stimulus were con-

founded by the order of presentation of the stimuli: more

Fig. 3 The means and standard errors of the scores of the duration of

active screen inspection behaviour during presentation of the

biological motion (BM), static (SBM), inverted (INV), scrambled

(SCR), and rotating (ROT) stimuli are presented for females (a) and

males (b)

Fig. 4 The means and standard errors for the number of events of

each of the components of active screen inspection are presented;

concentrated inspection, parallax, and head-cocking movements of

the head (see text). The scores are for females only during

presentation of the biological motion (BM), static (SBM), inverted

(INV), scrambled (SCR), and rotating (ROT) stimuli. Data points

marked a are significantly different from those marked b, but only for

concentrated inspection

Anim Cogn (2010) 13:555–564 561

123

time was spent viewing the first stimulus presented than

any of the stimuli presented subsequently. This indicates a

response to novelty and subsequent habituation. Active

screen inspection and time spent looking outside the con-

tainer was not affected by order of presentation. Time spent

viewing may, therefore, reflect assessment of an unfamiliar

situation, whereas active inspection and searching may be

used to gain information about a stimulus regardless of the

familiarity of the situation in which it is encountered.

The scores of active screen inspection behaviour

revealed that the female marmosets were sensitive to (1)

the motion information contained in the stimuli, shown by

the fact that they could distinguish the BM pattern from a

static frame of this image (2) the natural dynamics of

animal locomotion, since they could distinguish the BM

pattern from its inverted version (3) the biological rele-

vance of a motion pattern, since they could distinguish the

BM pattern from a solid rotating hen, and (4) the biological

coherence of a motion pattern, since they could distinguish

the BM pattern from its spatially scrambled version. Dis-

crimination of the latter stimulus was particularly impor-

tant because it required the marmosets to be sensitive to

both the local kinematics of the stimulus dots as well as the

relative spatial locations of the dots. Thus, the marmosets

responded to the global configuration (or at least some sub

configuration) of the walking hen display.

The most commonly performed component of inspec-

tion behaviour was concentrated inspection of the screen.

The higher percentage of ‘perpendicular’ movements in the

direction towards, rather than away, from the screen per-

formed by females during presentation of the BM pattern

suggests that these movements may have corresponded to

viewing the stimuli in detail. Since these movements

towards the screen often resulted in the marmoset’s face

coming to, and remaining within 1–2 cm of the screen and

marmosets are known to rely on olfaction (Lazaro-Perea

2001), it is possible that they might have been trying to

gather olfactory information from the walking hen stimu-

lus. No other stimulus received a similar response.

After the BM pattern, the inverted BM stimulus received

most inspection by the females (see Fig. 3a), and it

received significantly more parallax head movements than

the static frame. As parallax head movements have been

described as a form of visual exploration used to gather

information about depth (Rogers et al. 1993), these findings

suggest that the marmosets may have been trying to

interpret the inverted stimulus. Considering that marmosets

are an arboreal species, and the marmosets in the colony

have often been observed to hang upside down, it could be

that they have seen BM patterns (i.e. human carers and

other marmosets) in this orientation but only when ves-

tibular signals indicate that the observing individual is

upside down. Thus, the incompatibility between vestibular

signal and visual input may have prompted inspection. This

is supported by the finding of Meary et al. (2007) sug-

gesting that when a motion violates the expectations of a

human infant, it is viewed for longer than a natural

movement. However, our female marmosets did not view

the inverted BM stimulus as much or more than the BM

stimulus.

The role of vestibular information in BM perception is

questionable since Troje (2003) found that BM perception

in humans relies on a retinal, rather than environmental,

frame of reference. Troje (2003) pointed out that vestibular

information is not necessary in humans since we most often

experience BM patterns in the upright orientation. Never-

theless, a higher level of sophistication might be advanta-

geous in an arboreal species, such as the common

marmoset, enabling it to recognise a BM pattern in any

orientation and then perform an appropriate response (e.g.

flight or fight in the case of a predator or an intruding

conspecific).

It was found that all marmosets were motivated to

search outside the container (presumably for food) to the

same extent regardless of the stimulus being presented or

the novelty of the situation. The searching effort outside

the container remained relatively constant throughout

testing and contrasted with the active inspection of the

stimulus inside the container, which varied significantly

depending on the stimulus being displayed. These different

patterns of searching and inspecting behaviour indicate that

the marmosets were responsive specifically to the change

in conditions inside the container with the presentation of

each new stimulus. Moreover, as there was no sex differ-

ence in searching around the container, the observed sex

difference in active screen inspection was not related to the

level of motivation to search the apparatus in general.

The sex difference observed in active inspection of the

BM stimulus was probably not related to the known sex

differences of colour vision in marmosets: all males and

approximately one-third of females have dichromatic col-

our-vision, whereas the remaining females are trichromatic

(Jacobs 1998). Given that Blake (1993) found evidence of

BM perception in cats, a species with dichromatic colour-

vision, the sex difference in colour vision is unlikely to

explain our results. An alternative explanation involves

known sex differences in primates in response to novelty.

Female primates are known to acquire novel behaviour

patterns more readily (Itani 1958; Kawai 1963, 1965;

Kappeler 1987; Bachevalier et al. 1989), solve novel food

tasks more rapidly (Yamamoto et al. 2004), and to be more

responsive to novel objects (Rouff et al. 2005; Visalberghi

et al. 2003) than males. In the present study, it took the

marmosets only a small number of trials to learn the cover

removal task, and males and females did not differ sig-

nificantly in their ability to learn this task. This shows that

562 Anim Cogn (2010) 13:555–564

123

the nature of the task was such that sex differences in the

abilities of the marmosets to solve a novel food task or

acquire a novel behaviour pattern did not influence per-

formance. It is possible that the sex difference found in our

study might be due to a sex difference in response to

unfamiliar objects (Rogers et al. 1993). Considering that

the sex difference was significant only for the BM stimulus,

it would seem that higher levels of curiosity in females than

in males are specific to unfamiliar, moving organisms.

Rouff et al. (2005) found that male macaques are more

interested in threatening objects and less interested in non-

threatening objects than females and suggested that the

sexual dimorphism in some primates in response to novelty

and to threatening objects may be an adaptive division of

social role; males remain vigilant of known predators and

are not distracted by non-threatening stimuli, and females

investigate novel objects as potential new food resources.

To add to this, the colour vision polymorphism that mar-

mosets possess may mean that males are better equipped

than most females to detect camouflaged predators (Saito

et al. 2005), whereas most females are more successful at

detecting coloured fruits and leaves (e.g. Caine and Mundy

2000), which might correlate with social role. However,

these explanations do not fit easily with our results since

they would predict that males should be more attentive to

biological motion that might represent a predator than

should females.

It is also possible that the males were attending only to

individual dots of the stimuli. The global configurations of

the stimuli were most important for discrimination between

the patterns presented since some patterns shared individ-

ual dot motions. If this is the explanation why males do not

attend to the BM stimulus, it is worth noting that autistic

humans cannot process global properties of stimuli before

local properties and have difficulties in perceiving inter-

element relationships as part of a global structure (Brosnan

et al. 2004; and see also Vallortigara et al. 2008) and,

moreover, that tamarins process visual stimuli in this way

(Neiworth et al. 2006). Interestingly, infants with autism,

typically males, fail to recognise point-light displays of

biological motion but are instead highly sensitive to the

presence of non-social, physical contingencies that occur

within the stimuli (Klin et al. 2009).

Based on their study with humans, Troje and Westhoff

(2006) suggested the notion of an evolutionarily ancient

perceptual ‘‘life detector’’. A general perceptual mecha-

nism for the detection of articulated terrestrial vertebrates

that relies on the characteristic movement patterns of such

organisms. A fully developed capacity for BM perception

does not appear in humans until age 5 (Pavlova et al. 2001),

with certain aspects appearing at different stages of

development (Bertenthal et al. 1985, 1987; Booth et al.

2002; Meary et al. 2007). In infants close to birth, a novel

BM pattern in its correct orientation is sufficient for pref-

erential attendance (Simion et al. 2008), possibly providing

the infant with a basic means of detecting a potential

caregiver. Vallortigara et al. (2005) found evidence of a

similar general detection mechanism in chicks, which

supports the idea that this mechanism is ancient. Since the

BM pattern used in the present study was unfamiliar to the

marmosets our results provide further support for this

concept.

Although previous non-human primate studies are not in

agreement with our findings, it seems as though method-

ological procedure, rather than perceptual ability, may be

the reason for this discrepancy. Using our method, it is

clear that female common marmosets, if not males, can

distinguish BM from a static image and a number of other

forms of motion.

Acknowledgments We are grateful to the Australian Research

Council for funding to L.J.R. in support of the marmoset colony at

UNE. This project was part of the requirements of J.B.’s Honours

degree at the University of New England. The housing and testing

conditions of the marmosets were in accordance with the principles

and regulations of the Australian Code of Practice for the Care andUse of Animals for Scientific Purposes (1997) and approved by the

Animal Ethics committee at the University of New England (AEC08/

037).

References

Bachevalier J, Hagger C, Bercu BB (1989) Gender differences in

visual habit formation in 3-month-old rhesus monkeys. Dev

Psychobio 22:585–599

Bertenthal BI, Proffitt DR, Spetner NB, Thomas MA (1985) The

development of infant sensitivity to biomechanical motions.

Child Dev 56:531–543

Bertenthal BI, Proffitt DR, Kramer SJ (1987) Perception of biome-

chanical motions by infants: implementation of various process-

ing constraints. J Exp Psychol Hum Percept Perform 13(4):577–

585

Blake R (1993) Cats perceive biological motion. Psychol Sci 4(1):54–

57

Booth AE, Bertenthal BI, Pinto J (2002) Perception of the symmet-

rical patterning of human gait by infants. Dev Psychol

38(4):554–563

Brosnan MJ, Scott FJ, Fox S, Pye J (2004) Gestalt processing in

autism: failure to process perceptual relationships and the

implications for contextual understanding. J Child Psychol

Psychiatry 45:459–469

Caine NG, Mundy NI (2000) Demonstration of a foraging advantage

for trichromatic marmosets (Callithrix geoffroyi) dependent on

food colour. Proc R Soc Lond B 267:439–444

Clara E, Regolin L, Vallortigara G, Rogers L (2007) Perception of the

stereokinetic illusion by the common marmoset (Callithrixjacchus). Anim Cogn 10:135–140

Dittrich WH, Lea SEG, Barrett J, Gurr PR (1998) Categorisation of

natural movements by pigeons: visual concept discrimination

and biological motion. J Exp Anal Behav 70(3):281–299

Fox R, McDaniel C (1982) The perception of biological motion by

human infants. Science 218(29):486–487

Anim Cogn (2010) 13:555–564 563

123

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 1:84–98

Jacobs GH (1998) A perspective on color vision in platyrrhine

monkeys. Vis Res 38:3307–3313

Johansson G (1973) Visual perception of biological motion and a

model for its analysis. Percept Psychophys 14(2):201–211

Kaplan G, Rogers LJ (1999) Parental care in marmosets (Callithrixjacchus jacchus): development and effect of anogenital licking

on exploration. J Comp Psychol 113:269–276

Kaplan G, Rogers LJ (2006) Head-cocking as a form of exploration in

the common marmoset and its development. Dev Psychobio. doi:

10.1002/dev.20155:551-560

Kappeler PM (1987) The acquisition process of a novel behaviour

pattern in a group of ring-tailed lemurs (Lemur catta). Primates

28:225–228

Kawai M (1963) On the newly acquired behaviours of the natural

troop of Japanese monkeys on Koshima Island. Primates 4:113–

115

Kawai M (1965) Newly acquired pre-cultural behaviour of the natural

troop of Japanese monkeys on Koshima Islet. Primates 6:1–30

Klin A, Lin DJ, Gorrindo P, Ramsay G, Jones W (2009) Two-year-

olds with autism orient to non-social contingencies rather than

biological motion. Nature 459:257–261

Lazaro-Perea C (2001) Intergroup interactions in wild common

marmosets, Callithrix jacchus: territorial defence and assessment

of neighbours. Anim Behav 62:11–21

Mather G, West S (1993) Recognition of animal locomotion from

dynamic point-light displays. Percept 22:759–766

Meary D, Kitromilides E, Mazens K, Graff C, Gentaz E (2007) Four-

day-old human neonates look longer at non-biological motion of

a single point-of-light. PLoS ONE. doi:10.1371/journal.pone.

0000186

Neiworth JJ, Gleichman AJ, Olinick AS, Lamp KE (2006) Global and

local processing in adult humans (Homo sapiens), 5-year old

children (Homo sapiens), and adult cotton top tamarins (Sagui-nus oedipus). J Comp Psychol 120(4):323–330

Omori E (1997) Comparative study of visual perception using

Johansson’s stimuli. In: Chase S, Watanabe S (eds) Pattern

recognition in humans and animals. Keio University, Tokyo, pp

27–30

Omori E, Watanabe S (1996) Discrimination of Johansson’s stimuli in

pigeons. Int J Comp Psychol 9:92

Oram MW, Perrett PI (1994) Responses of anterior superior temporal

polysensory (STPa) neurons to ‘‘biological motion’’ stimuli. J

Cogn Neurosci 6(2):99–116

Parron C, Deruelle C, Fagot J (2007) Processing of biological motion

point-light displays by baboons (Papio papio). J Exp Psychol

Anim Behav Process 33(4):381–391

Pavlova M, Krageloh-Mann I, Sokolov A, Birbaumer N (2001)

Recognition of point-light biological motion displays by young

children. Percept 30(8):925–933

Regolin L, Tommasi L, Vallortigara G (2000) Visual perception of

biological motion in newly hatched chicks as revealed by an

imprinting procedure. Anim Cogn 3:53–60

Rogers LJ, Stafford D, Ward JP (1993) Head cocking in galagos.

Anim Behav 45:943–952

Rouff JH, Sussman RW, Strube MJ (2005) Personality traits in

captive lion-tailed macaques (Macaca silenus). Am J Primatol

67:177–198

Saito A, Mikamki A, Kawamura S, Ueno Y, Hiramatsu C, Widayati

KA, Suryobroto B, Teramoto M, Mori Y, Nagano K, Fujita K,

Kuroshima H, Hesagawa T (2005) Advantages of Dichromats

over Trichromats in discrimination of colour-camouflaged

stimuli in non-human primates. Am J Primatol 67:425–436

Simion F, Regolin L, Bulf H (2008) A predisposition for biological

motion in the newborn baby. Proc National Acad Sci

105(2):809–813

Tomonaga M (2001) Visual search for biological motion patterns in

chimpanzees (Pan troglodytes). Psychologia 44:46–59

Troje NF (2003) Reference frames for orientation anisotropies in face

recognition and biological-motion perception. Percept 32:201–210

Troje NF, Westhoff C (2006) The inversion effect in biological

motion perception: evidence for a ‘‘life detector’’. Curr Biol

16:821–824

Vallortigara G, Regolin L (2006) Gravity bias in the interpretation of

biological motion by inexperienced chicks. Curr Biol

16(8):R279–R280

Vallortigara G, Regolin L, Marconato F (2005) Visually inexperi-

enced chicks exhibit spontaneous preference for biological

motion. PLoS Biology. doi:10.1317/journal.pbio.0030208

Vallortigara G, Snyder A, Kaplan G, Bateson PPG, Clayton NS,

Rogers LJ (2008) Are animals autistic savants? PLoS Biology

6:208–214

Visalberghi E, Janson CH, Agostini I (2003) Response toward novel

foods and novel objects in wild Cebus paella. Int J Primatol

24(3):653–675

Yamaguchi MK, Fujita K (1999) Perception of biological motion by

newly hatched chicks and quail. Percept 28:23–24

Yamamoto ME, Domeniconi C, Box H (2004) Sex differences in

common marmosets (Callithrix jacchus) in response to an

unfamiliar food task. Folia Primatol 45:249–254

564 Anim Cogn (2010) 13:555–564

123

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.