Newborns’ recognition of changing and unchanging aspects of schematic faces

Newborns� recognition of changingand unchanging aspects of schematic faces

Chiara Turati and Francesca Simion *

Dipartimento di Psicologia dello Sviluppo e della Socializzazione, Universit�aa degli Studi diPadova, Padova, Italy

Received 28 January 2002; revised 13 September 2002

Abstract

The present study investigated newborns� ability to discriminate, recognize, and

learn visual information embedded in the schematic face-like patterns preferred at

birth. Four experiments were carried out using the visual-paired comparison para-

digm. Results indicated that newborns discriminated face-like stimuli relying on their

internal features (Experiments 1 and 4) and recognized a perceptual invariance be-

tween face-like configurations in conditions of low (Experiment 2) and high-percep-

tual variability (Experiment 3) of their inner elements. Altogether, data show that the

presence of the preferred structure that schematically defines a face, displaying a trip-

let of elements in the correct locations for eyes and mouth, does not constitute a limit

that constrains newborns� face learning processes.

� 2002 Elsevier Science (USA). All rights reserved.

Keywords: Schematic faces; Recognition; Newborns; Perceptual invariance

A number of studies have shown that newborns orient their gaze more

frequently and look longer toward well-contrasted schematic visual patterns

representing face-like rather than non-face-like configurations (Goren,

Sarty, & Wu, 1975; Johnson, Dziurawiec, Ellis, & Morton, 1991; Macchi

Cassia, Simion, & Umilt�aa, 2001; Maurer & Young, 1983; Mondloch et al.,

1999; Valenza, Simion, Macchi Cassia, & Umilt�aa, 1996). Some authors

J. Experimental Child Psychology 83 (2002) 239–261

www.academicpress.com

* Corresponding author. Fax: +39-49-8276511.

E-mail address: [email protected] (F. Simion).

0022-0965/02/$ - see front matter � 2002 Elsevier Science (USA). All rights reserved.

PII: S0022 -0965 (02 )00148-0

mail to: [email protected]

maintain that newborns� preference for face-like patterns is elicited by the

specific natural spatial arrangement of the features within a face and is med-

iated by a subcortical mechanism (i.e., Conspec) solely responsible of direct-

ing newborns� gaze toward any visual pattern characterized by the presence

of three high-contrast blobs in a triangular formation corresponding to eyesand mouth (Johnson & Morton, 1991). More recently, re-evaluating and

re-casting the proposals of the so-called sensory hypothesis 1 (Banks &

Ginsburg, 1985; Banks & Salapatek, 1981; Gayl, Roberts, & Werner,

1983; Kleiner, 1987, 1990, 1993; Kleiner & Banks, 1987), other authors sug-

gest that newborns� face preference might be determined by more general

visual structural properties that face-like configurations share with other

non-face-like visual patterns: the cumulative and combined effects of these

properties would assure a gain value to face-like as compared to non-face-like stimuli, inducing newborns� visual preference (Acerra, Burnod, & de

Schonen, 2002; Simion, Macchi Cassia, Turati, & Valenza, 2001; Turati,

Simion, Milani, & Umilt�aa, in press).

However, irrespective of whether the visual preference for face-like pat-

terns is determined by the specific structure of the face or by non-specific

structural properties, a still unsolved question concerns whether newborns�preference for the overall arrangement of the features allows or precludes pro-

cessing of the components embedded within face-like configurations. Morespecifically, it is still not known how far the presence of the preferred structure

that schematically defines a face, displaying a triplet of well-contrasted iden-

tical elements in the correct locations for eyes and mouth, constrain newborns

to process the overall face-like pattern, without attending to the perceptual

differences and similarities of the inner elements within the pattern.

According to the well-known model of the development of face process-

ing proposed by Johnson and Morton (1991) (Morton & Johnson, 1991; but

see also de Schonen & Mathivet, 1989), the Conspec mechanism becomes in-hibited, at about 6 weeks of age, by cortical mechanisms able to encode and

acquire information about individual faces and to discriminate between dif-

ferent exemplars of faces. Before this age, because Conspec triggers atten-

tion onto the overall schematic structural organization of the face,

learning processes of the inner features might be limited by the detection

of the whole geometry of the face. This possibility is tenable also if new-

borns� face preference is induced by non-specific structural properties (Ac-

erra et al., 2002; Simion et al., 2001; Turati et al., in press). It is possiblethat the preferred structural information that biases newborns� visual atten-

tion toward face-like patterns inhibits discrimination and recognition of the

inner elements comprised within the overall configuration.

1 The sensory hypothesis maintains that faces do not represent a special class of stimuli for

newborns and that they are preferred simply because their psychophysical properties match

those of the infants� sensory channels.

240 C. Turati, F. Simion / Journal of Experimental Child Psychology 83 (2002) 239–261

The issue concerning newborns� ability to process the facial inner ele-

ments becomes relevant in light of a crucial study with real faces that re-

vealed that 3-day-old infants become unable to discriminate and recognize

their mother�s face from a stranger�s face when headscarves mask the hair

and hairlines—that is, when only visual information concerning the innerpart of the face was provided (Pascalis, de Schonen, Morton, Deruelle, &

Fabre-Grenet, 1995; see also Bartrip, Morton, & de Schonen, 2001). Based

on this evidence, newborns� real face recognition abilities seem to rely upon

visual information concerning external face features (i.e., hair, hairline, and/

or outer contour) rather than inner features or configuration. Therefore,

both the newborns� preference for their mother�s face (Bushnell, 2001; Bush-

nell, Sai, & Mullin, 1989; Field, Cohen, Garcia, & Greenberg, 1984; Pascalis

et al., 1995) and newborns� recognition of a stranger�s face (Pascalis & deSchonen, 1994) might be explained as indexes of the capacity to recognize

salient perceptual differences in the external face characteristics, and not

as signs of recognition of inner facial features.

This possibility appears plausible because visual acuity and sensitivity to

contrast at birth are very poor (Atkinson, Braddick, & Moar, 1977; Norcia,

Tyler, & Hamer, 1990). So, newborns might not be able to detect the inner

features of a real face because of their low contrast and salience. However,

recently, it has been shown that newborns prefer photographs of real faceswith eyes opened rather than closed (Batki, Baron-Cohen, Wheelwright,

Connellan, & Ahluwalia, 2000), with direct rather than averted gaze (Far-

roni, Csibra, Simion, & Johnson, 2002) and with attractive rather than un-

attractive inner facial features (Slater et al., 1998, 2000). These results

suggest that some perceptual information embedded within a face is already

detected and processed very few days from birth.

Nevertheless, newborns do not process sufficient information about the

internal configuration or the inner features in order to discriminate and rec-ognize a well-experienced face, such as the mother�s face, when presented

without the outer contour. 2 As a consequence, the question of what infor-

mation newborns actually process and encode when they discriminate and

recognize a face remains still open. In front of this messy landscape, a first

step to address this issue is to test with the well-contrasted schematic face-

like pattern preferred at birth how far the inner elements within the pattern

can be attended to, learned, and discriminated.

2 Also, it is interesting noting that in the study by Pascalis et al. (1995) the outer contour was

completely removed. In contrast, in studies observing a visual spontaneous preference for the

inner part of a face, the outer contour was present but paired. Thus, as an anonymous reviewer

suggested, it is possible that face outer contour plays a superordinate role that lacks when a

scarf masks the hairline.

C. Turati, F. Simion / Journal of Experimental Child Psychology 83 (2002) 239–261 241

This topic appears even more intriguing when evidence on newborns�learning abilities with stimuli other than faces is taken into account. The ca-

pacity to detect, discriminate, and learn elements embedded in well-con-

trasted complex visual non-face stimuli has been well documented at

birth. Newborns are able to discriminate the shape of both global and localforms comprised in geometric hierarchical arrays (Macchi Cassia, Simion,

Milani, & Umilt�aa, 2002), and to differentiate between inner elements sur-

rounded by identical salient frames (Farroni, Valenza, Simion, & Umilt�aa,

2000). Moreover, newborns perceive a single visual stimulus as being the

same throughout variations in orientation, size, and slant, yet discriminating

the different appearances presented (Granrud, 1987; Slater, Mattock, &

Brown, 1990; Slater, Mattock, Brown, & Bremner, 1991; Slater & Morison,

1985). Finally, recent evidence suggests that, a few days from birth, infantsare able to detect and recognize perceptual similarities among different sim-

ple visual forms, provided that they belong to broad or global level percep-

tual categories and that the perceptual properties that define categories are

sufficiently salient to be easily detected. In particular, when closed versus

open geometric shapes are presented, newborns appear able to form broad

categories of discriminable stimuli, showing a preference for the novel-cate-

gory exemplar rather than for the familiar-category exemplar (Quinn, Sla-

ter, Brown, & Hayes, 2001; Turati, Simion, & Zanon, 2002). Overall,evidence is clear in showing that, when well-contrasted non-face-like visual

patterns are used, well-grounded learning abilities come into sight very early

in life.

Based on the reported evidence, some important questions appear still

unsolved: Are the learning processes already documented at birth with

non-face-like patterns also present with schematic face-like configura-

tions? When presented with well-contrasted face-like patterns with a

triplet of identical shapes in place of eyes and mouth, are newborns ableto discriminate and perceive perceptual similarities between the shapes of

the triplet elements? Or does attention to the triplet that schematically

defines a face preclude discrimination and recognition of the inner ele-

ments?

The goal of the present study was to investigate whether neonates are ca-

pable of discriminating salient changes and extracting perceptual invari-

ances concerning the inner elements of face-like configurations. Four

different experiments were carried out using the visual habituation or famil-iarization technique and manipulating the shape of three identical inner

features located in the appropriate locations for eyes and mouth within a

head-shaped, two-dimensional white form. Newborns� capacity to discrimi-

nate face-like configurations relying upon their inner components (Experi-

ments 1 and 4) and to recognize a perceptual invariance between face-like

configurations in conditions of low (Experiment 2) and high-perceptual var-

iability (Experiment 3) of their inner elements was tested.


Experiment 1

The goal of Experiment 1 was to determine whether infants are able to

discriminate two schematic face-like configurations on the basis of their in-

ternal features. More specifically, newborns� ability to discriminate closed-shaped inner features (squares vs diamonds) and open-shaped inner features

(crosses vs Xs) was tested. It was predicted that, after being familiarized with

a face-like pattern, newborns would show a novelty preference for the face-

like configuration with novel rather than familiar inner features.

Method

Participants

Eighty-eight healthy, full-term newborn infants were selected from the

maternity ward of the Pediatric Clinic of the University of Padova. Twelve

participants did not complete testing because of fussiness or drowsiness, 14

infants were excluded from the final sample because they showed a strong

position bias during the preference test phase, and 4 participants were re-

moved because of a technical error (a push button did not work). So, the

final sample consisted of 58 infants (30 males, 28 females), randomly as-

signed to one of two groups (31 to Group 1, 27 to Group 2). All infantsmet the screening criteria of normal delivery, a birth weight between 2550

and 4000 g, and a 5-min Apgar score above 7. They were tested the hour be-

fore the scheduled feeding time. Their ages at the time of testing ranged ap-

proximately from 24 to 80 h (mean¼ 52 h). Informed consent was obtained

from their parents.

Stimuli

The stimuli were head-shaped, head-sized, two-dimensional white forms,about 15 � 25 cm, with three identical black elements in the appropriate lo-

cations for the eyes and mouth regions (see Fig. 1; also see Morton & John-

son, 1991; Valenza et al., 1996). At a viewing distance of about 25 cm, each

stimulus subtended about 34� of visual angle horizontally and about 57� ver-

tically. Four different types of inner elements were used. Two of them can be

described as closed forms (i.e., squares and diamonds), whereas the other

two can be described as open forms (i.e., crosses and Xs). Each inner feature

was planned to cover an overall surface area of approximately 7 cm2. Eachsquare element was 2.8 cm on each side and subtended about 6�. To enhance

the difference between the stimuli, the diamond elements were not simply 45�rotations of the squares and the X elements were not simply 45� rotations of

the crosses. The horizontal diagonal of each diamond was 3.5 cm (about 8�)and the vertical diagonal was 4.2 cm (about 10�). Each cross was composed

of two identical perpendicular rectangles that were 3.5 cm (about 8�) on one

side and were 1.2 cm (about 3�) on the other. Each X-shaped element was


composed of two parallelograms, 3 cm (about 7�) in height and 1.2 cm

(about 3�) in width, placed as to form an X. The stimuli were projected

on a screen at a distance of approximately 7.5 cm (about 17�) from thecentral fixation point.

Apparatus

Each infant was placed on the experimenter�s lap, in front of a tracing-pa-

per screen, at a distance of about 25 cm. The experimenter holding the infant

was blind with respect to the hypotheses under test. The infant�s eyes were

aligned with a red flickering LED located in the center of the screen. The

LED was used to attract the infant�s gaze at the start of both the habituationand preference test phases. The LED subtended about 2� of a visual angle

and, when turned on, blinked at a rate of 300 ms on and 300 ms off. The

stimuli were provided by two carousel slide projectors controlled by a Mac-

intosh Quadra 900 computer and located behind the tracing-paper screen.

To prevent interference from irrelevant stimuli, plain white curtains were

drawn on both sides of the infant.

Procedure

As soon as the infant was apparently at ease and his/her gaze was prop-

erly aligned with the central flickering LED, a second experimenter started

the habituation phase by pressing a key on the computer keyboard. This au-

tomatically turned off the central LED and activated the slide projectors. An

infant control habituation procedure was used (Horowitz, Paden, Bhana, &

Self, 1972; Slater, Morison, & Rose, 1985).

During the habituation phase, two identical face-like configurations were

projected bilaterally, one on each side (i.e., left and right) of the centralLED. Bilateral rather than central presentation was chosen for two different

Fig. 1. The stimuli used in Experiment 1.


reasons. First, at birth photoreceptors in the central fovea are very imma-

ture, this resulting in a poor vision in the central area of visual field (Abra-

mov et al., 1982; Atkinson & Braddick, 1989). Second, when newborns look

at a centrally presented stimulus, it is difficult for an observer to decide if

they are actually looking at the stimulus or if they simply do not move theireyes from the central position.

Infants assigned to Group 1 were presented with face-like configurations

with closed inner features (i.e., squares or diamonds), whereas infants as-

signed to Group 2 were shown face-like configurations with open inner fea-

tures (i.e., crosses or Xs). The familiarization stimulus was counterbalanced

across participants. Half of the infants assigned to Group 1 were familiar-

ized with a face-like configuration with square-shaped elements and the

other half were familiarized with a face-like configuration with diamond-shaped elements. For half of the infants in Group 2, the inner elements dur-

ing the habituation phase were cross-shaped, for the other half the elements

were X-shaped.

An observer, na€ııve to the hypotheses being tested and to the stimuli pre-

sented, recorded the duration of each fixation on the stimulus by pressing a

push button that was connected to the computer. Because during the habit-

uation phase the same stimulus was presented on the left and on the right,

the amount of looking was recorded irrespective of the side. A look-awaycriterion of 2 s was used to determine the end of each fixation. In order to

be sure that this criterion was strictly respected, the software was planned

so that it automatically compacted two consecutive fixations that were not

separated by a time interval of at least 2 s.

The stimuli remained on the screen until the habituation criterion was

reached. The infant was judged to have been habituated when, from the

fourth fixation on, the sum of any three consecutive fixations was 50 per-

cent or less than the total of the first three (Horowitz et al., 1972; Slateret al., 1985). When the habituation criterion was reached, the stimuli

were automatically turned off and the central flickering LED was turned

on.

As soon as the infant�s gaze was realigned to the central LED, a prefer-

ence test phase started. Each infant was given two paired presentations of

the test stimuli. During each presentation, infants assigned to Group 1 were

simultaneously presented with two face-like configurations, one with square-

shaped features and one with diamond-shaped features. A configurationwith cross-shaped inner features together with a configuration with X-

shaped inner features was presented to infants assigned to Group 2. The

two-paired stimuli were always shown in both left and right positions, the

position being reversed from presentation 1 to presentation 2. The initial

left–right order of presentation was counterbalanced across participants.

The central LED flickered between the first and the second presentation

but did not flicker when the test stimuli were shown.


During the preference test phase, the observer, blind to the hypotheses

being tested and to the stimuli presented, recorded the duration of infant�sfixations on each stimulus by pressing two different push buttons depending

on whether the infant looked at the right or the left position. Each presen-

tation lasted when a total of 20 s of looking to the novel and/or familiarstimuli had been accumulated.

All testing sessions were video-recorded. Videotapes of eye-movements

were subsequently codified by a second observer unaware of the stimuli pre-

sented. The mean estimate of reliability between observers was 0.92 (Pear-

son correlation), so the recording procedure was considered reliable.

Results

Preliminary statistical analyses showed no significant effect or interactions

involving the distinct shape of closed features (squared- or diamond-shaped)

or open-features (cross- or X-shaped) presented during the habituation

phase. As a consequence, data were collapsed across this factor.

Two separate t tests for independent samples were applied to compare to-

tal fixation times and number of fixations to reach the habituation criterion

for the two groups (i.e., Group 1, closed features and Group 2, open fea-

tures). In the comparison, neither variable attained statistical signifi-cance.The average total fixation time was 104.06 s (SD ¼ 59:11) for Group

1 and 87.33 s (SD ¼ 51:50; tð56Þ ¼ 1:14; p > :20, two-tailed) for Group 2.

The average number of fixations was 9.39 for Group 1 and 8.48 for Group

2, tð56Þ ¼ 1:2; p > :20, two-tailed.

To determine whether newborns discriminate the inner features of the

face-like configurations, a novelty-preference score (percentage) was com-

puted for each infant by dividing the time the novel stimulus was observed

during the two test presentations by the total time over the two presenta-tions that the novel and the familiar stimuli were fixated. The obtained score

was multiplied by 100. Therefore, only scores significantly above 50% indi-

cated a preference for the novel stimulus. t-tests were performed comparing

the preference scores to chance (50%). The mean scores for the novel stim-

ulus were significantly greater than the chance level for both Group 1

ðM ¼ 64:26Þ, tð30Þ ¼ 3:03; p < :006 (two-tailed) and Group 2 ðM ¼63:84Þ, tð26Þ ¼ 3:37; p < :003 (two-tailed). Additionally, the mean novel-

ty-preference scores for the two Groups did not differ significantly,tð56Þ ¼ :07; p > :90 (two-tailed).

Discussion

The results of Experiment 1 demonstrate that newborn infants discrimi-

nate highly similar face-like configurations on the basis of the shape of their

inner features, independently of whether the features to be discriminated


were closed- or open-forms. This finding indicates that newborns are able to

attend to and to differentiate perceptual changes related to the individual

components of schematic face-like stimuli and that they are not compelled

to process face-like patterns relying exclusively on global information. Ex-

periment 2 will address the question of whether the face-like configurationsthat were differentiated in Experiment 1 could be treated as similar when

compared with other face-like configurations with highly different inner

features.

Experiment 2

The purpose of Experiment 2 was to determine whether newborn infantsare able to extract and recognize an invariant property related to the inner

features of a face-like stimulus. More specifically, the question was whether

newborns are able to treat as similar the inner features that they discrim-

inated in Experiment 1. A modified version of the recognition-memory

procedure was used (Quinn & Eimas, 1996), in which both face-like stimuli

presented in the preference test phase were novel. For one of them, the in-

ner components were highly different from those of the familiarized config-

uration, whereas for the other they were highly similar. It was predictedthat if newborns were able to extract and recognize a basic perceptual in-

variance related to the inner features of a face-like configuration, during

the test phase, they would prefer a novel face-like configuration with

highly different inner features over a novel face-like configuration with in-

ner features highly similar to those of the habituated stimulus configura-

tion. Alternatively, if newborns were not able to extract and recognize a

basic perceptual invariance, they would perceive the two test stimuli as

equally novel. Therefore, they would not show any novelty preference inthe test phase.

Method

Participants

Ninety-two healthy, full-term newborns were tested. Seventeen partici-

pants were discarded from the sample because of fussiness or drowsiness,

15 because of position bias during the preference test phase, and 1 infantwas removed because of a technical error (a push button did not work).

The final sample consisted of 59 newborns (28 males and 31 females), ran-

domly assigned to one of two groups (30 to Group 1 and 29 to Group 2).

The screening criteria were the same as those used in Experiment 1. At

the time of testing, infants had an average age of 60 h (range 24–82 h). They

were tested during the hour preceding the scheduled feeding time. The

parents were informed and gave their consent.


Stimuli, apparatus, and procedure

Stimuli, apparatus, and procedure were the same as in Experiment 1.

During the habituation phase, infants were familiarized with two identical,

bilaterally presented face-like configurations with either square- or dia-

mond-shaped elements as inner features (Group 1) and with either cross-or X-shaped elements as inner features (Group 2).

During the preference test phase, each infant was given two-paired pre-

sentations of novel test stimuli, one being highly different from the familiar

one, the other being highly similar. More specifically, infants familiarized

with square- or diamond-shaped features (Group 1) were presented with

two novel face-like configurations, one with highly similar inner features

(i.e., diamonds or squares, respectively), the other with highly different inner

features (i.e., crosses). In the same way, when the stimulus in the habituationphase was a configuration with cross- or X-shaped features (Group 2), two

novel configurations were presented during the preference test phase, one

with highly similar features (i.e., Xs or crosses, respectively) and the other

with highly different features (i.e., squares) (see Fig. 2).

Results

To determine whether total fixation time and number of fixations toreach criterion in the habituation phase differed as a function of type of in-

ner components of the stimulus (i.e., closed- or open-shaped), two separate t

tests for independent samples (Group 1 and Group 2) were applied, one for

each dependent variable. Mean total fixation times to reach the criterion did

not significantly differ for the two groups (Group 1, 76.91 s, SD ¼ 40:69;

Group 2, 77.55 s, SD ¼ 49:08; tð57Þ ¼ :05; p > :50, two-tailed). Mean number



of fixations was 8.3 for Group 1 (SD ¼ 2:78) and 7.96 for Group 2 (SD ¼1:72; tð57Þ ¼ :55; p > :50, two-tailed).

To determine whether newborns were able to extract and recognize an in-

variant property related to the inner features of face-like configurations, a

preference score for the highly different configuration (percentage) was com-puted as in Experiment 1. Two one-sample t tests were applied to determine

whether infants� visual preferences during the test phase differed significantly

from a chance level of 50%. Both infants assigned to Group 1 (M ¼ 66:37,

SD ¼ 27:64; tð29Þ ¼ 3:24; p < :003, two-tailed) and to Group 2 (M ¼63:55, SD ¼ 29:20; tð28Þ ¼ 2:50; p < :02, two-tailed) had preference scores

for the highly different configuration significantly greater than chance. A t

test for independent samples indicated that the mean novelty-preference

scores for the two groups were not significantly different (tð57Þ ¼:38; p > :70, two-tailed).

Discussion

The results are clear in showing that newborns looked longer at the face-

like stimulus with internal features that were more different from those of

the familiar face-like configuration. This means that the novel-schematic

face with similar-shaped inner features was treated as more alike the familiarschematic face than a novel-schematic face with dissimilar internal features.

That happened even though, as demonstrated in Experiment 1, there were

discriminable changes between the two face-like patterns. Newborns ignored

some of the perceived differences and responded to schematic faces in terms

of similarity. As a consequence, evidence has been provided that newborns

detect and recognize an invariant perceptual property related to the internal

features of well-contrasted face-like configurations.

Nevertheless, in the present experiment, newborns� ability to detect a per-ceptual invariance was investigated by habituating newborns to one single

exemplar of a face-like configuration. That is to say, newborns had to detect

a perceptual similarity between two stimuli: The habituated face-like config-

uration and the face-like configuration with highly similar inner features

shown during the test phase. Therefore, the within-category variability

was extremely low because of the small number of exemplars that repre-

sented the perceptual category (i.e., 2). Moreover, the within-category

face-like configurations displayed a strong resemblance, because newbornsmight have perceived diamonds and squares, as well as crosses and Xs, as

identical features with different orientations. Both these factors (i.e., the

small number of within-category exemplars and their strong resemblance)

might have represented a facilitating condition that might have enhanced

newborns� performance. Experiment 3 was designed to test newborns� ability

to perceive perceptual commonalties among face-like configurations in a

condition of higher perceptual variability of their inner elements.


Experiment 3

The goal of Experiment 3 was to investigate whether newborns are able

to manifest the capacity to recognize a perceptual similarity when a greater

number of face-like stimuli with different components were displayed in thefamiliarization phase.

To address this question, infants were familiarized with three different ex-

emplars of face-like configurations belonging to the same perceptual cate-

gory and then tested with a new exemplar from the familiar-category

paired with a novel-category exemplar. It was predicted that, as a result

of familiarization, newborns would recognize a basic perceptual similarity

among the inner components of the familiar category and prefer the nov-

el-category exemplar rather than the familiar-category exemplar.

Method

Participants

The participants were healthy, full-term newborn infants selected at the

maternity ward of the Pediatric Clinic of the University of Padova. Thir-

ty-nine infants were tested but 14 of them were excluded from further anal-

ysis. Nine participants did not complete testing because of fussiness ordrowsiness and 5 infants showed a position bias during the preference test

phase (they spent more than 80% of the time looking in one direction).

So, the final sample consisted of 25 infants (11 males and 14 females), ran-

domly assigned to one of two groups (13 to Group 1, 12 to Group 2). All of

them met the screening criteria of normal delivery already described in Ex-

periment 1. Infants were tested during the hour before the scheduled feeding

time. Their ages at the time of testing ranged approximately from 24 to 90 h

(mean¼ 57 h). Informed consent was obtained from their parents.

Stimuli

The head-shaped, head-sized, two-dimensional white forms with three

identical black elements in the appropriate locations for the eyes and mouth

regions used in the two previous experiments were used as stimuli (Morton

& Johnson, 1991; Valenza et al., 1996). Eight different types of inner ele-

ments were displayed (see Fig. 3). Four of them were closed forms (i.e.,

square, circle, triangle, and diamond) and four were open forms (i.e., cross,X, an open geometric figure that looked like a toy windmill, and an open

geometric form that looked like an irregular X-shaped figure). Each inner

element was planned to cover an overall surface area of approximately

7 cm2. Squares, diamonds, crosses, and Xs were identical to those already

described in the Stimuli section of Experiment 1. As for the circle, its radius

was 3.2 cm (about 8�). The triangle was 3.4 cm (about 8�) in height and its

base was 3.9 cm (about 9�). The two open features not used in previous


experiments had a maximum height and maximum width of 3.3 cm (about

8�). The extremities of the arms of the element that looked like a toy wind-

mill were 0.8 cm (about 2�). Those of the irregular X-shape form measured

0.6 cm (about 1�).

Apparatus

The apparatus was identical to that used in Experiments 1 and 2.

Procedure

During the familiarization phase, three 20-s trials were administered in

which three different face-like configurations with closed- or open-shaped el-

ements as inner features were displayed. More specifically, infants assigned

to Group 1 were familiarized with different face-like configurations with



closed inner components and infants assigned to Group 2 were familiarized

with different face-like patterns with open inner components, so that three

face-like configurations, one for each trial, were presented. Each configura-

tion was shown bilaterally, on both sides of the central LED.

An experimenter recorded the duration of each fixation on the stimulusby pressing a push button that was connected to the Macintosh computer.

A look-away criterion of 2 s was used to determine the end of each fixation.

To be sure that this criterion was strictly respected, the software was

planned so that it automatically compacted two consecutive fixations that

were not separated by a time interval of at least 2 s. When the 20-s familiar-

ization criterion was reached, the stimuli were automatically turned off and

the central flickering LED was turned on. As soon as the infant was re-

aligned to the central LED, the subsequent trial began.At the end of the three familiarization trials, a preference test phase

started. The procedure of the preference test phase was identical to the

one used in Experiments 1 and 2. Each infant was given two-paired presen-

tations of the test stimuli. During each presentation, infants were presented

with two face-like configurations, one with new closed components and one

with open components as inner features. The order and sequence of presen-

tation of the 4 closed components patterns (3 for familiarization, 1 for pref-

erence testing) were controlled across participants employing a balancedLatin square design. So, the familiar-category exemplar presented in the test

phase was not always the same for all participants. On the contrary, the nov-

el-category exemplar was maintained constant: A face-like configuration

with cross-shaped inner features was presented to newborns belonging to

Group 1, while a face-like configuration with square-shaped inner features

was shown to Group 2.

Results

To determine whether newborn infants perceived a perceptual similarity

among the face-like configurations, a novelty-preference score (percentage)

was computed for each infant as in Experiments 1 and 2. That is, the time

the novel-category face-like configuration was fixated during the two presen-

tations of the preference test phase was divided by total time that the novel-

and familiar-category stimuli were fixated, and the score was multiplied by

100.A one sample t test was applied to determine whether novelty-preference

scores differed significantly from the chance level of 50%. The comparison

was significant, showing that preference scores for the novel-category

face-like configuration were above chance (M ¼ 68:64, SD ¼ 24:33;tð24Þ ¼ 3:83; p < :001, two-tailed) and indicating that newborns preferred

the novel-category exemplar rather than the familiar-category exemplar

never seen before.


To determine whether both infants familiarized with face-like configura-

tions with closed-shaped inner features (Group 1) and infants familiarized

with open-shaped inner features (Group 2) showed a novelty preference

for the novel-category exemplar significantly different from the chance level

of 50%, two separate one-sample t tests were applied, one for each Group.The mean scores for the novel stimulus were significantly greater than the

chance level for both Group 1 (M ¼ 65:77; SD ¼ 20:78; tð12Þ ¼ 2:74;p < :02, two-tailed) and Group 2 (M ¼ 71:75; SD ¼ 28:29; tð11Þ ¼ 2:66;p < :03, two-tailed). Additionally, a t test for independent samples revealed

that the mean novelty-preference scores for the two Groups did not differ

significantly (tð23Þ ¼ �:61; p > :50, two-tailed).

Discussion

Evidence obtained seems to indicate newborns� ability to recognize com-

mon perceptual characteristics shared by face-like stimuli that differ in the

shape of their inner components. When familiarized with face-like configu-

rations with either closed or open forms, in the test phase newborns looked

longer at the configuration markedly different from those they were famil-

iarized to, that is, to the configuration with novel-category inner forms.

Thus, the present findings suggest that, at birth, infants are able to forma perceptually driven category that includes face-like configurations with

closed components as different as squares, triangles, circles, and diamonds

but excludes face-like configurations with open components such as

crosses. Similarly, newborns are also able to represent a category of face-

like configurations with open-shaped inner features (e.g., crosses) that

excludes face-like configurations comprised of closed-shaped inner features

(e.g., squares).

However, a possible alternative interpretation might consider the resultsobtained in the present experiment as a consequence of newborns� failure

to discriminate among within-category members, rather than as an index

of an underlying capacity to detect perceptual similarities among exemp-

lars of the same category. In fact, as an anonymous reviewer pointed

out, evidence from Experiment 1 showed that newborns were capable of

discriminating schematic face-like stimuli containing very similar closed

inner components, such as a diamond and a square, but it did not show

that they could discriminate between circles and triangles, and betweenthese forms and squares and diamonds. An identical criticism might be

applied to newborns� ability to differentiate among open-shaped inner fea-

tures. As a consequence, Experiment 1 does not serve as a proper control

for Experiment 3, because it does not provide sufficient evidence that in-

fants can discriminate between the internal features presented during the

habituation phase. Experiment 4 was designed to test this possible alterna-

tive explanation.


Experiment 4

The goal of Experiment 4 was to test whether newborn infants could dis-

criminate between exemplars selected within each of the two perceptual cat-

egories taken into account in Experiment 3 (i.e., closed- and open-shapedinner features of face-like configurations). Each infant was familiarized with

an instance from one perceptual category and then tested with the familiar

instance paired with a novel instance from the same familiar category.

If within-category inner features are discriminated, then it is possible to

conclude that the results obtained in Experiment 3 are indicative of new-

borns� ability to group different face-like configurations relying on the per-

ceptual similarities of their inner elements. Conversely, if within-category

perceptual discrimination is not evident, then Experiment 3 provides onlyevidence of newborns� distinction between a single undifferentiated face-like

configuration with closed-shaped inner features and a second configuration

with open-shaped inner features.

Method

Participants

Thirty-five healthy, full-term infants were recruited in the nursery of thePediatric Clinic of the University of Padova. Three were removed from the

study because they became too fussy or cried, 2 were discarded because of a

technical error (a push button did not work), and 4 were removed because

they showed a strong position bias, looking more than 80% of the time to

one direction. So, the final sample consisted of 26 newborns (12 females

and 14 males) who met the screening criteria of normal delivery described

in Experiment 1. Infants randomly assigned to Group 1 (n ¼ 13) were tested

with face-like configurations with closed-shaped inner features, while thoseassigned to Group 2 (n ¼ 13) were examined with open-shaped features (see

Procedure below). Babies were tested during the hour preceding the sched-

uled feeding time only if they were awake and in an alert state. Their age at

the time of testing ranged from 24 to 72 h (mean¼ 68 h). Informed consent

was obtained from their parents.

Stimuli and apparatus

The same stimuli used in Experiment 3 were presented (see Fig. 3).The apparatus was identical to that employed in all the previous experi-

ments.

Procedure

The procedure was similar to that used in Experiment 1. During the

habituation phase, one single face-like configuration with closed- (Group

1) or open-shaped inner components (Group 2) was displayed bilaterally.


Subsequently, during the preference test phase, infants were presented with

the familiar face-like configuration paired with a novel face-like exemplar

from the same category.

The stimuli presented during the habituation phase were counterbalanced

between infants. All the possible familiar versus novel instance pairings weretested; as a consequence the pairings were different for each infant. For in-

stance, three participants assigned to Group 1 were habituated with a face-like

configuration with circled-shaped inner features. Of them, one was tested with

circled features paired with squared features, one with circled features paired

with diamond features and the third with circled features and triangular fea-

tures. An identical procedure was applied for all the face-like configurations

displayed, with either closed- or open-shaped features (see Quinn et al.,

2001, for the same procedure).

Results

Mean total fixation time to reach the criterion in the habituation phase

was 104.6 s for infants assigned to Group 1 (i.e., habituated to closed inner

features) and 123.6 s for those assigned to Group 2 (i.e., habituated to open

inner features). A t test for independent samples revealed that mean total fix-

ation times did not differ significantly (tð24Þ ¼ �:67; p > :50, two-tailed).Also the number of fixations in the habituation phase was not significantly

different for the two Groups (Group 1, 8.31, SD ¼ 2:21; Group 2, 9.69,

SD ¼ 3:77; tð24Þ ¼ �1:14; p > :20, two-tailed).

To determine whether newborns were able to discriminate among face-like

configurations with closed- or open-shaped inner components, fixation times

during the test phase were transformed into percentages as in previous exper-

iments. A novelty-preference score was computed for each infant, so that

scores significantly above 50% indicated a preference for the novel stimulus.t-tests were performed comparing the preference scores to chance (50%).

A mean novelty-preference score significantly greater than the chance level

was obtained when newborns� performance was collapsed across groups

(M ¼ 62:97; SD ¼ 29:91; tð25Þ ¼ 2:21; p < :05, two-tailed). This finding

indicates that, overall, newborns were able to discriminate between face-like

configurations with different inner features belonging to the same category.

To establish whether face-like with both closed- and open-shaped inner

features were discriminated, two one-sample t tests were computed consid-ering separately Groups 1 and 2. The mean scores for the novel stimulus

were significantly greater than the chance level for Group 1

(M ¼ 73:35; SD ¼ 21:85; tð12Þ ¼ 3:85; p < :003, two tailed) but not for

Group 2 (M ¼ 52:58; SD ¼ 33:94; tð12Þ ¼ :27; p > :50, two-tailed). Addi-

tionally, a t test for independent samples revealed that the mean novelty-

preference scores for the two Groups tend to differ significantly

(tð24Þ ¼ 1:85; p ¼ :07, two-tailed).


Discussion

Overall, evidence from Experiment 4 shows that newborns are able to dis-

criminate the face-like within-category exemplars presented in Experiment

3. This capacity appears well established in the case of closed-shaped com-ponents, while it vanishes when face-like configurations with open-shaped

inner components are presented. Two different interpretations might be pro-

posed to explain this pattern of results. A first possibility might take into ac-

count the limits of newborns� visual system that might succeed in

differentiating X-shaped from cross-shaped features (Experiment 1), but

might fail to discriminate the details of more complex open inner features.

A different possibility might refer to the intrinsic greater difficulty in process-

ing open than closed features, already observed with adults (Elder & Zucker,1993; Handel & Christ, 1969; Saarinen & Levi, 1999; Treisman & Gormican,

1988) and 4–8-year-old children (Cohen & Haith, 1977; Tada & Stiles-Da-

vis, 1989), despite equivalent amounts of contour, comparable familiarity

and visibility.

In any case, Experiment 4 revealed newborns� ability to differentiate be-

tween exemplars selected within the category of face-like patterns with

closed-shaped features. Based on this result, it is possible to conclude that

newborns� preference for the novel-category exemplar observed in Experi-ment 3 demonstrates infants� capacity to extract and recognize perceptual

similarities among discriminable exemplars of face-like configurations be-

longing to the closed category. Conversely, in the case of face-like configu-

rations with open-shaped components, the novelty preference for the

stimulus belonging to the novel category found in Experiment 3 should be

ascribed to newborns� failure to discriminate among the different within-

category members presented. In fact, the results obtained in Experiment 4

did not confirm the presence of this discrimination ability in newborns.

General discussion and conclusion

Evidence gathered in the present study demonstrates that newborns are

able both to discriminate face-like configurations and to recognize their per-

ceptual similarities relying on the shape of their inner features. Results of Ex-

periment 1 showed that newborns discriminate between face-like stimuli onthe basis of their inner components, indicating that they are not constrained

to process the overall face configuration without attending to the shapes of

its inner elements. Experiment 2 demonstrated that newborns treated two

face-like configurations with similar components as more alike than two

face-like configurations with non-similar components, showing that they are

able to extract, process, and recognize a perceptually invariant property

shared by the stimuli.


Experiment 3 showed that, in conditions of higher within-category vari-

ability, newborns are able to form a perceptual category of face-like config-

urations with closed-shaped inner features that excludes face-like

configurations with open features. As shown in Experiment 4, the results

of Experiment 3 cannot be ascribed to newborns� failure to discriminate be-tween exemplars with different closed-shaped inner components. So, new-

borns are able to recognize the presence of common perceptual properties

among face-like configurations despite discriminable changes in their closed

inner features. This result confirms and extends previous findings obtained

with geometric stimuli (Quinn et al., 2001; Turati et al., 2002).

A different conclusion emerges in the case of face-like configuration with

open-shaped inner features, because newborns were unable to reliably dis-

criminate the subtle perceptual differences among the four exemplars be-longing to this perceptual category (Experiment 4). As a consequence, in

this case it is not possible to conclude that the capacity to identify perceptual

characteristics shared by several different face-like configurations with open-

shaped components is present a few days from birth. Rather, it is plausible

that, in Experiment 3, newborns perceived the different configurations with

open-shaped features as a single indistinct stimulus.

Overall, our findings demonstrate that the presence of the preferred struc-

ture that schematically defines a face, displaying a triplet of elements in thecorrect locations for eyes and mouth, does not represent a limit that con-

strains newborns to process exclusively the geometry of a face, inhibiting

learning processing of its inner elements. When well-contrasted schematic

faces are presented, infants are able to detect perceptual differences and sim-

ilarities. Thus, attention to the triplet that represents a face does not pre-

clude learning processes concerning the detection of the differences that

distinguish a particular face-like stimulus from a different one, nor learning

processes related to the detection of the common perceptual properties thatbind together discriminable face-like configurations.

Hence, learning processes with well-contrasted face-like patterns do not

appear different from those already well documented in literature in the case

of non-face-like patterns. Many lines of evidence converge to demonstrate

that, in experimental conditions in which newborns� visual abilities are not

impaired by low-level variables, such as contrast or luminance of the stimuli,

newborns manifest the capacity to perceive visual stimulations in a coherent

and organized fashion (Slater, 1998). They discriminate geometric stimuli onthe basis of their form (Farroni et al., 2000; Macchi Cassia et al., 2002;

Slater, 1995), they extract and recognize perceptual invariances (Slater

et al., 1990, 1991; Slater & Morison, 1985) and they form broad perceptual

categories for closed versus open geometric forms (Quinn et al., 2001; Turati

et al., 2002). These perceptual abilities have been shown in the present study

with the face-like stimuli usually employed when face preference at birth is

tested.


Evidence about learning capacity with schematic face-like configurations

might shed some light on newborns� real face recognition abilities. In partic-

ular, the results of the present study exclude the possibility that the reported

limits in newborns� capacity to recognize a real face relying on its inner part

(Pascalis et al., 1995) should be related to newborns� visual preference forthe inner geometry of the face. On the contrary, it is likely that such limits

should be referred to general factors that affects learning of any visual stim-

ulus at birth. For instance, the visibility of the features within a real face is

lower than the visibility of the elements that characterizes schematic black

and white face-like patterns. Therefore, it is possible that, a few days from

birth, infants are constrained by their limited visual capacities to discrimi-

nate and recognize faces relying on their most salient characteristics, that

is the shape of the hair and the outer contour (Pascalis et al., 1995). Basedon this hypothesis, the results observed by Pascalis et al. (1995) might be ex-

plained not by virtue of the presence of the preferred face schema but in

light of general constraints of newborns� visual system.

However, this hypothesis appears unlikely because some studies indicated

that visual abilities at birth are sufficiently proficient to detect the inner fea-

tures of a real face (Batki et al., 2000; Farroni et al., 2002; Slater et al., 1998,

2000). A more convincing proposal might suggest that limitations in real

face recognition abilities at birth are due to the engagement of more com-plex attentional or learning processes. In fact, in the case of real faces, the

inner features (i.e., eyes, nose, mouth, . . .) have different shapes that depend

highly upon their location within the face. In contrast, the three inner ele-

ments within the face-like patterns used in the present study display the

same shape, irrespective from their location. This simplifies discrimination

and recognition because newborns do not need to take into account the lo-

cation of the features in order to compare two different face-like patterns.

Conversely, the differentiation between two real faces appears more difficultbecause it requires comparing shapes of inner elements with a reference to

their location. Further research might challenge this point by testing

infants� learning abilities with well-contrasted face-like patterns displaying

inner elements with different shapes.

Acknowledgments

The research was supported by an EC Biomed Grant (No. BMH4-97-

2032) and by a grant from the Ministero dell�Universit�aa e della Ricerca Sci-

entifica e Tecnologica (No. 9911C68488-003). We are deeply indebted to

Prof. F. Zacchello, Dr. B. Dalla Barba, and the nursing staff at the Pediatric

Clinic of the University of Padua for their collaboration. We are grateful to

the editor and the anonymous reviewers for their suggestions on earlier ver-

sions of this paper. We also thank S. Bettella and S. Massacesi for writing


the software, C. Umilt�aa, E. Valenza, and V. Macchi Cassia for their helpful

comments, M. Bonati, C. Menazza, L. Zanon, G. Tomasoni, and L. Zulian

for assistance with infant testing. Special thanks are due to the children who

took part in the study and to their parents.

References

Abramov, I., Gordon, J., Hendrickson, A., Hainline, L., Dobson, V., & LaBossiere, E. (1982).

The retina of the newborn human infant. Science, 217, 265–267.

Acerra, F., Burnod, Y., & de Schonen, S. (2002). Modelling aspects of face processing in early

infancy. Developmental Science, 5, 98–117.

Atkinson, J., & Braddick, O. (1989). Development of basic visual functions. In A. Slater &

G. Bremner (Eds.), Infant development (pp. 7–41). Hove, UK: Erlbaum.

Atkinson, J., Braddick, O., & Moar, K. (1977). Development of contrast sensitivity over the

first 3 months of life in the human infant. Vision Research, 17, 1037–1044.

Banks, M., & Ginsburg, A. P. (1985). Infant visual preferences: A review and new theoretical

treatment. In H. W. Reese (Ed.), Advances in child development and behavior (Vol. 19, pp.

207–246). New York: Academic Press.

Banks, M., & Salapatek, P. (1981). Infant pattern vision: A new approach based on the contrast

sensitivity function. Journal of Experimental Child Psychology, 31, 1–45.

Bartrip, J., Morton, J., & de Schonen, S. (2001). Responses to mother faces in 3-week to 5-

month-old infants. British Journal of Developmental Psychology, 19, 219–232.

Batki, A., Baron-Cohen, S., Wheelwright, S., Connellan, J., & Ahluwalia, J. (2000). Is there an

innate gaze module? Evidence from human deonates. Infant Behavior and Development, 23,

223–229.

Bushnell, I. W. R. (2001). Mother�s face recognition in newborn infants: Learning and memory.

Infant and Child Development, 10, 67–74.

Bushnell, I. W. R., Sai, F., & Mullin, J. T. (1989). Neonatal recognition of the mother�s face.

British Journal of Developmental Psychology, 7, 3–15.

Cohen, K. M., & Haith, M. M. (1977). Peripheral vision: The effects of developmental,

perceptual, and cognitive factors. Journal of Experimental Child Psychology, 24, 373–394.

de Schonen, S., & Mathivet, E. (1989). First come, first served: A scenario about the

development of hemispheric specialization in face recognition during infancy. European

Bulletin of Cognitive Psychology, 1, 3–44.

Elder, J., & Zucker, S. (1993). The effect of contour closure on the rapid discrimination of two-

dimensional shapes. Vision Research, 7, 981–991.

Farroni, T., Csibra, G., Simion, F., & Johnson, M. H. (2002). Eye contact detection in humans

from birth. Procedings of the National Academy of Science, 99, 9602–9605.

Farroni, T., Valenza, E., Simion, F., & Umilt�aa, C. (2000). Configural processing at birth:

Evidence for perceptual organisation. Perception, 29, 355–372.

Field, T. M., Cohen, D., Garcia, R., & Greenberg, R. (1984). Mother-stranger face

discrimination by the newborn. Infant Behavior and Development, 7, 19–25.

Gayl, I., Roberts, J., & Werner, J. (1983). Linear system analysis of infant visual pattern

preferences. Journal of Experimental Child Psychology, 35, 30–45.

Goren, C., Sarty, M., & Wu, P. (1975). Visual following and pattern discrimination of face-like

stimuli by newborn infants. Pediatrics, 56, 54–549.

Granrud, C. E. (1987). Size constancy in newborn human infants. Investigative Ophthalmology

and Visual Science, 28(Suppl.), 5.

Handel, S., & Christ, R. E. (1969). Detection and identification of geometric forms using

peripheral and central viewing. Perception and Psychophysics, 6, 47–49.


Horowitz, F. D., Paden, L., Bhana, K., & Self, P. (1972). An infant control method for studying

infant visual fixations. Developmental Psychology, 7, 90.

Johnson, M. H., & Morton, J. (1991). Biology and cognitive development. The case of face

recognition. Oxford, England: Basil Blackwell.

Johnson, M. H., Dziurawiec, S., Ellis, H., & Morton, J. (1991). Newborns� preferential tracking

of face-like stimuli and its subsequent decline. Cognition, 40, 1–19.

Kleiner, K. A. (1987). Amplitude and phase spectra as indices of infants� pattern preferences.

Infant Behavior and Development, 10, 49–59.

Kleiner, K. A. (1990). Models of neonates� preferences for facelike patterns: A response to

Morton, Johnson, and Maurer. Infant Behavior and Development, 13, 105–108.

Kleiner, K. A. (1993). Specific versus non-specific face recognition device. In B. de Boysson-

Bardies, S. de Schonen, P. Jusczyk, P. McNeilage, & J. Morton (Eds.), Developmental

neurocognition: Speech and face processing in the first year of life (pp. 103–108). New York:

Academic Press.

Kleiner, K. A., & Banks, M. S. (1987). Stimulus energy does not account for 2-month-old

infants� face preference. Journal of Experimental Psychology: Human Perception and

Performance, 13, 594–600.

Macchi Cassia, V., Simion, F., & Umilt�aa, C. (2001). Face preference at birth: The role of an

orienting mechanism. Developmental Science, 22, 892–903.

Macchi Cassia, V., Simion, F., Milani, I., & Umilt�aa, C. (2002). Dominance of global visual

properties at birth. Journal of Experiment Psychology, 131, 398–411.

Maurer, D., & Young, R. (1983). Newborns� following of natural and distorted arrangements of

facial features. Infant Behavior and Development, 6, 127–131.

Mondloch, C. J., Lewis, T. L., Budreau, D. R., Maurer, D., Dannemiller, J. L., Stephens, B. R.,

& Kleiner-Gathercoal, K. A. (1999). Face perception during early infancy. Psychological

Science, 5, 419–422.

Morton, J., & Johnson, M. H. (1991). Conspec and Conlern: A two-process theory of infant

face recognition. Psychological Review, 98, 164–181.

Norcia, A. M., Tyler, C. W., & Hamer, R. D. (1990). Development of contrast sensitivity in the

human infant. Vision Research, 30, 1475–1486.

Pascalis, O., & de Schonen, S. (1994). Recognition memory in 3- to 4-day-old human neonates.

Neuroreport, 5, 1721–1724.

Pascalis, O., de Schonen, S., Morton, J., Deruelle, C., & Fabre-Grenet, M. (1995). Mother�s face

recognition by neonates: A replication and an extension. Infant Behavior and Development,

18, 79–85.

Quinn, P. C., & Eimas, P. D. (1996). Perceptual organization and categorization in young

infants. In C. Rovee-Collier & L. P. Lipsitt (Eds.), Advances in infancy research (vol. 10,

pp. 1–36). Norwood, NJ: Ablex Publishing Corporation.

Quinn, P. C., Slater, A. M., Brown, E., & Hayes, R. A. (2001). Developmental change in form

categorization in early infancy. British Journal of Developmental Psychology, 19, 207–218.

Saarinen, J., & Levi, D. M. (1999). The effect of contour closure on shape perception. Spatial

Vision, 2, 227–238.

Simion, F., Macchi Cassia, V., Turati, C., & Valenza, E. (2001). The origins of face perception:

Specific vs non-specific mechanisms. Infant and Child Development, 10, 59–65.

Slater, A. (1995). Visual perception and memory at birth. In C. Rovee-Collier & L. P. Lipsitt

(Eds.), Advances in infancy research (vol. 9, pp. 107–162). Norwood, NJ: Ablex.

Slater, A. (1998). The competent infant: Innate organisation and early learning in infant visual

perception. Visual, auditory, and speech perception in infancy. In A. Slater (Ed.), Perceptual

development (pp. 105–130). Hove, UK: Psychology Press Ltd.

Slater, A., Bremner, G., Johnson, S. P., Sherwood, P., Hayes, R., & Brown, E. (2000). Newborn

infants� preference for attractive faces: The role of internal and external facial features.

Infancy, 1, 265–274.


Slater, A., Mattock, A., & Brown, E. (1990). Size constancy at birth: Newborn infants�responses to retinal and real size. Journal of Experimental Child Psychology, 49, 314–322.

Slater, A., Mattock, A., Brown, E., & Bremner, J. G. (1991). Form perception at birth: Cohen

and Younger (1984) revisited. Journal of Experimental Child Psychology, 51, 395–406.

Slater, A., & Morison, V. (1985). Shape constancy and slant perception at birth. Perception, 14,

337–344.

Slater, A., Morison, V., & Rose, D. (1985). Habituation in the newborns. Infant Behavior and

Development, 13, 183–200.

Slater, A., Von der Schulenburg, C., Brown, E., Badenoch, M., Butterworth, G., Parsons, S., &

Samuels, C. (1998). Newborn infants prefer attractive faces. Infant Behavior and Develop-

ment, 21, 345–354.

Tada, W. L., & Stiles-Davis, J. (1989). Children�s analysis of spatial patterns: an assessment of

their ‘‘errors’’ in copying geometric forms. Cognitive Development, 4, 177–195.

Treisman, A., & Gormican, S. (1988). Feature analysis in early vision: Evidence from search

asymmetries. Psychological Review, 95, 15–48.

Turati, C., Simion, F., Zanon, L. (2002). Newborns� perceptual categorization for open and

closed forms. Presented at the 13th Biennial International Conference on Infant Studies.

Toronto, Canada.

Turati, C., Simion, F., Milani, I., Umilt�aa, C. (in press). Newborns� preference for faces: What is

crucial? Developmental Psychology.

Valenza, E., Simion, F., Macchi Cassia, V., & Umilt�aa, C. (1996). Face preference at birth.

Journal of Experimental Psychology: Human Perception and Performance, 22, 892–903.


Newborns’ recognition of changing and unchanging aspects of schematic faces

Documents