Neuropsychologia 83 (2016) 6375

Contents lists available at ScienceDirect

Neuropsychologia
journal homepage: www.elsevier.com/locate/neuropsychologia

Visual expertise for horses in a case of congenital prosopagnosia

Nilly Weiss, Elite Mardo, Galia Avidan n
Department of Psychology, Ben-Gurion University of the Negev, POB 653, Beer-Sheva 84105, Israel

art ic l e i nf o

a b s t r a c t

Article history:
Received 21 December 2014
Received in revised form
27 July 2015
Accepted 28 July 2015
Available online 29 July 2015

A major question in the domain of face perception is whether faces comprise a distinct visual category
that is processed by specialized mechanisms, or whether face processing merely represents an extreme
case of visual expertise. Here, we examined O.H, a 22 years old woman with congenital prosopagnosia
(CP), who despite her severe decits in face processing, acquired superior recognition skills for horses. To
compare the nature of face and horse processing, we utilised the inversion manipulation, known to
disproportionally affect faces compared to other objects, with both faces and horses. O.H's performance
was compared to data obtained from two control groups that were either horse experts, or non-experts.
As expected, both control groups exhibited the face inversion effect, while O.H did not show the effect,
but importantly, none of the participants showed an inversion effect for horses. Finally, gaze behaviour
toward upright and inverted faces and horses was indicative of visual skill but in a distinct fashion for
each category. Particularly, both control groups showed different gaze patterns for upright compared to
inverted faces, while O.H presented a similar gaze pattern for the two orientations that differed from that
of the two control groups. In contrast, O.H and the horse experts exhibited a similar gaze pattern for
upright and inverted horses, while non-experts showed different gaze patterns for different orientations.
Taken together, these results suggest that visual expertise can be acquired independently from the
mechanisms mediating face recognition.
& 2015 Elsevier Ltd. All rights reserved.

Keywords:
Expertise hypothesis
Domain specic
Face recognition
Gaze behaviour
Visual perception

1. Introduction
1.1. Characteristics of face perception
Faces have distinctive evolutionary and social signicance and,
therefore, they have long been considered a unique object category. The unique properties of face perception have often been
attributed to holistic processing. Namely, perception which relies
not only on processing of the features comprising the face, but also
on the gestalt of these features (Maurer et al., 2002). Holistic
processing have been investigated using a number of well-known
experimental manipulations including the inversion effect (Farah
et al., 1995; Freire et al., 2000; Yin, 1969), part-whole effect
(Gauthier and Tarr, 2002; Tanaka and Farah, 1993) and the composite effect (Boutet et al., 2002; Farah et al., 1998; Young et al.,
1987; Gauthier et al., 2003). Of most relevance for the present
study is the face inversion effect, indicating the disproportional
decrement in perception (Haxby et al., 1999), recognition (Brooks
and Goldstein, 1963) and memory (Goldstein, 1965; Hochberg and
Galper, 1967; Yin, 1969) for inverted, compared to upright faces
relative to the effect of inversion on processing other object
n

Corresponding author. Fax: 972 8 6428348.

E-mail address: [email protected] (G. Avidan).

http://dx.doi.org/10.1016/j.neuropsychologia.2015.07.028
0028-3932/& 2015 Elsevier Ltd. All rights reserved.

categories. The common interpretation of this nding is that while

upright faces are processed in a holistic, efcient manner which
emphasises the invariant structure of the identity of the face, inverted faces are processed in a featural, piecemeal manner leading
to the typical reduced performance observed in this condition
(Barton et al., 2001; Farah et al., 1995; Rhodes et al., 1993; Rossion,
2009, 2008, but see Susilo et al. (2013)). It has been often argued
that other objects are less prone to the inversion manipulation
because their processing relies less heavily on holistic perception,
compared to faces. Note however that some researchers argue that
the face inversion effect does not necessarily index holistic processing (Murray, 2004; Richler et al., 2011; Richler and Gauthier,
2014; Sekuler et al., 2004). Rather, they suggest that while inverted
faces are processed less efciently than upright faces, this cost in
performance is quantitative rather than qualitative in nature. We
will return to this claim in Section 4 in relation to our results.
An additional and important measure for assessing the nature
of face processing is gaze behaviour and eye movement patterns.
Many studies revealed that when looking at human faces, normal
individuals tend to xate on the eyes and the nose more than on
other features comprising the face (e.g., mouth, cheeks, and chin),
a pattern that was attributed to a holistic processing strategy
(Belle et al., 2010; Schwarzer et al., 2005, 2007). In contrast, an
analytical strategy yields more xations towards the specic informative regions of the face, which are dependent on the task's

64

N. Weiss et al. / Neuropsychologia 83 (2016) 6375

requirements (Schwarzer et al., 2005). This result is consistent

with ndings showing atypical gaze behaviour in individuals with
disrupted holistic face perception due to Autism (Dalton et al.,
2005), acquired prosopagnosia (Bukach et al., 2006a), or congenital prosopagnosia (CP). Pertinent to the present study are
ndings showing that individuals with CP exhibit reduced gaze
preference towards the eyes (Schmalzl et al., 2008), and tend to
rely more heavily on external and dispersed regions of the face
(Schwarzer et al., 2007).
Based on the wealth of behavioural and neural ndings, it is
often assumed that faces comprise a unique object category,
however despite much research, this issue is still contentious.
Specically, some researchers claim that faces are processed by
domain specic distinct cognitive and neural systems which are
different from those devoted to other object categories (Kanwisher, 2000; McKone and Robbins, 2007; Weiner and Grill-Spector,
2013; Yovel and Kanwisher, 2004) and may even be innate (Goren
et al., 1975; McKone et al., 2007). In contrast, others have proposed
the opponent, expertise hypothesis according to which face
perception and representation is not unique, but rather, represents
an extreme case of visual expertise (Bukach et al., 2006a; Gauthier
and Bukach, 2007; McGugin et al., 2012), hence implicating that
processing faces and objects of expertise depend on the same
cognitive and neural mechanisms.
In the current study, we present a case study of an individual, O.
H, who has congenital prosopagnosia (CP), a life-long impairment in
face processing despite normal sensory vision and normal intelligence and in the absence of neurological history, who nevertheless reported having an expertise for horses. The goal of this
study is to test whether this individual indeed presents visual expertise in a visual category not related to faces, and if so, whether
such expertise would be associated with holistic processing, as has
been suggested in other studies (see below). These unique circumstances, in which face perception is impaired while visual expertise naturally emerged in a different visual domain, allow us to
disentangle the cognitive mechanisms mediating visual expertise.
1.2. Processing objects of expertise
As outlined above, some researchers have suggested that faces
represent a special case of acquisition of visual expertise (Bukach
et al., 2006a; Gauthier and Bukach, 2007). According to this view,
one would expect that other objects of expertise would share
some common characteristics with faces related to the underlying
mechanisms mediating expertise. Below, characteristics of face
processing will be compared to those of objects of expertise as
they appear in several studies, and subsequently, some evidence
for a possible dissociation between face processing and processing
objects of expertise will be presented.
1.2.1. Holistic processing for objects of expertise
Given the centrality of holistic processing in face perception, a
major question is whether objects of expertise are also processed
holistically. This question has been addressed in a number of studies
and the results thus far are conicting. For example, in a study
conducted by (Diamond and Carey, 1986), subjects who were dog
experts (i.e., breeders and judges in dogs' competitions) exhibited
an inversion effect when presented with pictures of dogs, while in
contrast, non-experts exhibited no difference in performance between upright and inverted photos of dogs. Another study that
examined holistic processing for natural objects of expertise,
documented the composite effect for cars, in car experts, and the
effect was correlated with the extent of expertise (Gauthier et al.,
2003). Along similar lines, Wong et al. (2009) showed that even in
expertise that was acquired in a short period of laboratory training
(in contrast to real-life expertise), participants acquired holistic

processing as evident by a composite effect obtained for the articial, trained objects (Ziggerins). Another study by Gauthier and
Tarr (1997) attempted to show congural processing for another
category of articial objects (greebles) following training, however, this study only revealed congural sensitivity to the greebles,
but did not nd other signature holistic effects such as the partwhole or inversion effects in greeble experts compared to novices. A different set of studies used a basic-level detection task to
examine possible commonalities between faces and objects of expertise. These studies showed interference for face detection in the
presence of objects of expertise, regardless of their task-relevance
(Hershler and Hochstein, 2009; McGugin et al., 2010), suggesting a
common, holistic search mechanism for both stimuli.
In contrast to the ndings described above, Robbins and
McKone (2007) also investigated face-like processing in dog experts and novices but did not nd any evidence for holistic processing in neither experts nor novices (no inversion or composite
effect). Furthermore, a study by Harel and Bentin (2013) undermined the necessity of holistic processing, typically mediated by
low spatial frequencies for objects of expertise, by showing an
advantage for processing images of cars containing only high
spatial frequencies in cars experts. Another study by the same
group (Golan et al., 2013) have demonstrated better visual detection of cars by cars experts compared to non-experts, as opposed
to similar performance for faces by CP individuals (who may be
considered face novices) and a control group, implying different
mechanisms for detecting faces and objects of expertise.
Finally, several studies used eye tracking to examine the underlying mechanisms of expertise (Manning et al., 2006; Donovan
and Litcheld, 2013). In these studies radiologists were required to
search for a pathological nding in a chest x-ray. Expert radiologists
used less dispersed xations, focusing on the regions with higher
probability for locating abnormality, compared to novices. This may
imply a local, but yet, more efcient search strategy for experts.
1.2.2. Dissociation between the perception of faces and objects of
expertise
In addition to studies which examined the extent of similarity
between the mechanisms underlying processing of faces and objects of expertise, some studies reported dissociations between
face perception and objects of expertise in cases of visual agnosia
or acquired prosopagnosia (AP). For example, an AP patient who
showed preserved recognition ability for cars despite his severe
decit in face recognition (Sergent and Signoret, 1992), a farmer
who lost his ability to recognise his cows, but retained his face
recognition ability, following a lesion in occipitaltemporal regions
(Assal et al., 1984), and an AP patient who became a farmer following his brain injury, and developed expertise for sheep, while
having impairments in face perception (McNeil and Warrington,
1993). Other studies described experimental training manipulations that attempted to train AP (Behrmann et al., 2005a, 2005b;
Bukach et al., 2012; Rezlescu et al., 2014) and CP (Duchaine et al.,
2004) individuals to acquire expertise in a new category in spite of
their impaired face processing abilities, thus indicating a distinction between face recognition and the ability to acquire expertise
in a different visual category. Note though, that Bukach et al.
(2012) suggested that the expertise acquired by the AP patient
they trained, was accomplished through more analytic mechanisms and not the holistic mechanisms normally used for face
perception, which are impaired in this patient. Taken together,
these case studies illustrate a possible dissociation between performance associated with face perception, and acquired or maintained abilities in other domains of visual expertise, and therefore
suggest that different behavioural mechanisms may be involved in
face perception compared to the processing of other objects of
expertise.

N. Weiss et al. / Neuropsychologia 83 (2016) 6375

1.3. The present study

To benchmark face perception abilities against perception of
horses and non-facial objects in experts and non-experts, O.H's
performance was compared to a control group of horse experts,
and a control group of non-experts, all showing normal face perception on two standard diagnostic tests. Various tasks were administered to each participant to assess long-term memory for
familiar faces and horses, and short-term memory for faces and
non-facial objects (cars). Critically, we also used an upright-inverted task for faces, and created a corresponding upright-inverted
task for horses to assess differences in performance for faces
compared to horses and to examine the extent of holistic processing for the two categories as would be indexed by the inversion effect. In addition to response time and accuracy measures, we employed eye tracking, which may provide a more
sensitive measure of expertise-related visual processing.
We hypothesise that O.H and the experts control group will
show better performance than the non-expert group in the horse
task. In addition, based on previous studies in CP, we expect to nd
an inversion effect for faces in both control groups, but not in O.H.
Consistently with this account, we expect that both control groups
would exhibit more xation time towards the centre of the face
thus exhibiting signatures for holistic processing. In contrast, O.H
is expected to show a more local scrutiny of the face. As for horses,
in light of the results described above regarding the possible distinction between face processing and processing objects of expertise, we suggest that holistic processing does not necessarily
underlie visual expertise and hence we do not expect to nd an
inversion effect for horses regardless of expertise. Accordingly, we
expect that experts will gaze to more informative regions of the
horse, regardless of orientation, while non-experts will exhibit a
more free-viewing scan pattern towards horse stimuli. In that
context we note that O.H, like many other CPs, showed a general
impairment in holistic perception that is not restricted to faces
(Tanzer et al., 2013), thus further suggesting that holistic processing would not necessarily underlie her expertise for horses.

2. Methods
2.1. Participants
2.1.1. O.H, a case study of CP
O.H is a 22 years old female, a Psychology undergraduate student
at the time of testing, who reported having life-long difculties with
face perception and memory. She reported numerous embarrassing
situations in which she failed to recognise people whom she should
have known. Aside for the face perception difculty, she reported
having a distinctive ability to identify horses which she acquired
through her intensive experience with horse riding, treatment, and
training. She has been working with horses since she was 7 years old
and she still continues to do so, thus having 15 years of experience.
A representative incidence she had described to us occurred in a
horse riding competition, in which she was able to recognise a horse
she remembered, but not his owner, who she was supposed to know.
O.H is right handed according to her self-report, she has normal vision, and does not have any known neuropsychological or neurological decits. Upon her interactions with the research team (two
students in this team are trained clinical psychologists), she did not
show any evidence of any social decits. Her difculties with faces
were formally assessed with standard experiments (see below).
2.1.2. Horse experts
Fifteen women ranging in age from 20 to 29 (M 25, SD2.5) were
recruited through social media, and personal acquaintance to serve as a

65

matched control group for O.H. The inclusion criteria, in addition to age
and sex, was based on the extent of their experience with horses which
included years of horse riding, treatment, and training, and if applicable
also participation in competitions. The number of years of experience
reported by this group ranged from 7 to 23 years (M12.9, SD 4.99).
All of the horse experts reported right-hand dominance except for one
which was left-handed, and all had normal or corrected to normal vision, and reported no neuropsychological or neurological decits.
2.1.3. Non-expert controls
Fifteen female students from Ben-Gurion University of the
Negev, age 2128 (M23, SD 2) participated in the experiment
in return for course credit. There was no age difference between
the two control groups. All reported right hand dominance and
normal vision and were not experienced with horses.
2.2. Diagnostic tasks
2.2.1. Face perception tasks
To make sure that all participants had normal face perception,
as opposed to O.H, all participants completed two diagnostic tests
commonly used in the literature.
2.2.1.1. Famous face questionnaire. This test examines long-term
memory for known faces (described in detail in Avidan and
Behrmann (2008)). The questionnaire contains 56 photos of celebrities, intermixed with 56 photos of unknown people (celebrities from other countries), arranged in a table. Participants are
required to answer the questionnaire, by providing the name or
detailed context of the famous person, or indicate if the face is
unfamiliar. Response time is unlimited.
2.2.1.2. Cambridge Face Memory Test (CFMT) (Duchaine and Nakayama, 2006). The test examines short term memory for unfamiliar faces and is widely used for diagnosis of congenital/developmental prosopagnosia. The test contains faces shown across
varying views and lightings and has three levels of difcultly.
2.2.2. Non-face tasks
2.2.2.1. Horse trivia questionnaire. The horse experts and the CP
subject reported having years of experience with horses. In order
to quantitatively asses their expertise and make sure that the
participants in the non-expert group were indeed novices in this
domain, we devised a horse trivia questionnaire, which included
15 questions about horses and the realm of horse riding. The
questionnaire was composed with the aid of a horse expert who
did not participate in the study (see Supplementary data).
2.2.2.2. Horse perceptual expertise test. In order to substantiate
horse expertise in the visual domain as well, we have created an
independent horse breed experiment. The experiment consisted of
60 pairs of horses, each horse was presented for 1000 ms, and
following presentation of the two horses, the participants were
required to respond, using the mouse, whether or not the two
horses within a pair were from the same breed. The horses we
used were coloured pictures of 120 different horses in prole view,
in various breeds that were downloaded from the internet. Different images were used for the perceptual expertise test, and for
the experimental task outlined below. The horses were cropped
from their natural scenery, and displayed on a white background,
using GIMP software. The experiment was programmed in Java
and was sent to the participants by e-mail.
2.2.2.3. Horse recognition test. To further verify O.H's superior visual recognition skills for horses; we composed a familiar horse
questionnaire, corresponding to the famous face questionnaire.

66

N. Weiss et al. / Neuropsychologia 83 (2016) 6375

Fig. 1. An example of the stimuli presented in a single trial in the upright/inverted face (1.A, 1.B) and horse (1.C, 1.D) experiments. Three stimuli are presented in a pyramid
conguration: a target stimulus is presented at the top and two stimuli, an identical one, and a distractor are presented at the bottom. Participants are asked to choose, by a
button press, whether the matched stimulus is on the left or right side of the target.

The questionnaire contained 26 photos of horses presented in a

PowerPoint presentation. Thirteen of the pictures were of horses
from the farm where O.H works, and which, according to her selftestimony she knows well, and another 13 photos were of horses
that are similar in appearance to the known horses but are from
other farms, unknown to O.H. All pictures were taken by the research team and were edited and converted to grayscale in Adobe
Photoshop. For each picture, O.H was required to write the name
of the horse, if she recognised it, and/or its breed, or alternatively,
to indicate that she did not know the horse.
2.2.2.4. Object perception task. The Cambridge Car Memory Test
(CCMT; Dennett et al., 2012), which is a car recognition test matched to the CFMT in format was used to assess general object recognition skills in all participants. Note that during the experimental sessions, the participants were tested on additional tasks,
unrelated to the present study.
2.3. Experimental tasks
2.3.1. Stimuli and apparatus
During the experimental tasks, eye movements were recorded
using a video-based desktop mounted eye tracker (Eye Link1000, SR
Research, Ontario, Canada). Stimuli were presented on a 19-in. LCD
monitor at a resolution of 1024  768 pixels with a refresh rate of
60 Hz. The horse stimuli were 60 grayscale prole view photos of
various breeds of horses that were photographed in 3 different horse
farms in Israel by the research team. Using Adobe Photoshop software, the horses were cropped from their natural scenery, and displayed on a grey background, and were all facing the left side of the
screen. Note that we used different horse pictures than the horses
used for the horse perceptual expertise test. In all horse tasks, we
used images of whole horses, rather than images of horse faces based
on input we obtained from several horse experts (who did not

participate in the study). These experts claimed that horse recognition involves information embedded in the body and not just the
head and hence seeing a photograph of a horse without its body
would be less authentic, and even peculiar. Moreover, presentation of
the whole object of expertise (even in the case of animals such as
dogs and birds) is consistent with other studies which investigated
holistic processing of objects of expertise (Diamond and Carey, 1986;
Gauthier and Tarr, 1997; Robbins and McKone, 2007; Wong et al.,
2009). The face task was adapted from (Behrmann et al., 2005a,
2005b), stimuli were 60 grayscale images of faces in frontal view,
with neutral expression (30 males, 30 females), containing no diagnostic or salient cues, scanned from a 3D laser and obtained from
Bulthoff and Troje (Max Planck Institute, Tubingen, Germany (Troje
and Blthoff, 1996)).
2.3.2. Procedure
Participants performed two experiments with an identical design, rst an upright-inverted face experiment, and then an upright-inverted horse experiment. Eye movements were recorded
during both experiments while participants' head was positioned
in a chinrest to minimise head movements.
Participants sat approximately 70 cm from the screen. At the
beginning of each experiment, a 9-point calibration cycle and drift
correction were conducted to ensure recording accuracy with spatial resolution of 0.01 of visual angle. Before each trial, the subjects
were required to press the space bar while looking at a central
xation point to collect data about slow drifts of the eye tracker. If
drift error was more than 2, a new calibration protocol was initiated. The xation point disappeared with the onset of each trial.
Each experiment contained 30 trials, in each trial three stimuli,
arranged in a pyramid-like conguration were shown, such that a
target stimulus was presented at the top of the screen, and two stimuli were presented below, one which was the same as the target,
and the other which served as a distractor (see Fig. 1). The

N. Weiss et al. / Neuropsychologia 83 (2016) 6375

Table 1
Z scores of the CFMT as calculated based on age-matched control subjects reported
in Bowles et al. (2009), Z scores on CCMT as calculated based on data reported in
Dennett et al. (2012), and Z scores of famous faces questionnaire as calculated
based on control subjects included in Avidan and Behrmann (2008). Signicant
differences are highlighted in bold. Note that O.H was more than 2 standard
deviations of the normal range on the two face diagnostic tests, but normal on the
CCMT.
Participant

CFMT (Z
score)

Famous faces questionnaire (Z score)

CCMT (Z
score)

O.H
Horse expert group
(mean 7 std)
Non-expert group
(mean 7 std)

 2.16
 0.007
(1.15)
0.02 (1.3)

 2.45
0.71 (0.42)

 0.48
0.53 (1.09)

0.63 (0.46)

 0.44
(0.76)

participants were instructed to choose which of the two stimuli

below the target is identical to the target. The location of the target
was randomly chosen, as trials were presented in a randomized order. Subjects were given unlimited time to respond to each trial, but
were instructed to respond as quickly and as accurately as they can.

3. Results
3.1. Part 1: Diagnostic face tests and general visual perception
abilities
To allow comparison of performance with faces to other objects,
as well as comparison of O.H to the control groups, all raw data
were transformed to Z scores. These values for the CFMT, famous
face questionnaire, and CCMT experiments are presented for O.H,
expert group, and non-expert group in Table 1. As can be seen, for
the experts and non-experts, average scores for these tests did not
signicantly differ from the typical values previously reported in the
general population (Avidan and Behrmann, 2008; Bowles et al.,
2009; Dennett et al., 2012). In contrast, O.H had abnormal Z scores
in the two face tests, as expected in CP (the threshold for abnormal
face perception in the specied tests was dened as 2 SDs below
average similarly to other papers (Avidan et al., 2011; Bowles et al.,
2009; Duchaine and Nakayama, 2006; Rivolta et al., 2012; Russell
et al., 2009)). Additionally, she had a normal Z score in the CCMT,
indicating the specicity of her impairment to face processing.
3.2. Part 2: Establishing horse expertise
3.2.1. Benchmarking general horse expertise: the horse trivia
questionnaire
To examine the level of general horse expertise, an independent
one tailed t-test was conducted on the score of the horse trivia
questionnaire as a function of group. Data on the questionnaire were
available for 13 experts and 15 non-experts (see Table 2 top row for
average scores and signicance levels in the questionnaire). This
analysis revealed a signicant main effect of group, conrming that
the score of the expert group was higher than that of the non-experts. To estimate O.H's expertise in relation to the performance of
the experts and non-experts, we used the modied t-test for single
case study developed by Crawford and Garthwaite (2002) and implemented in SINGLIMS.exe. This statistical approach is often used in
the neuropsychological literature and will also be used below to
statistically compare the performance of O.H and the controls in the
experimental tasks. This analysis revealed no signicant difference
between O.H and experts on the horse trivia questionnaire, but in
contrast, a signicant difference between O.H and the non-experts,
indicating better performance of O.H on this test. To summarise, the
experts group and O.H did not signicantly differ in their knowledge

67

of the horse realm, which was signicantly better than the knowledge of this topic exhibited by the non-experts' group.
3.2.2. Benchmarking visual expertise for horses
3.2.2.1. Independent visual tests: horse recognition and breed distinction. O.H exhibited perfect recognition of the horses residing
in the farm where she works: in the familiar horse questionnaire,
she correctly identied by name, all the horses from her farm, with
no false alarms for unknown horses (13/13 embedded among 13
horses that were unknown to O.H).
The score of the horse perceptual expertise test was calculated
as the percentage of answers in which participants correctly
decided whether the two horses presented in a trial were of the
same breed or not. To examine the difference between groups in
terms of visual expertise, a one tailed t-test was conducted on this
score as a function of group. Results from 10 experts and 13 nonexperts were available (see Table 2 bottom row for average scores
and signicance levels in the test). Despite the difculty level of
this test, the expert group performed signicantly better than the
non-experts. To compare the performance of O.H to the experts
and non-experts, we used SINGLIMS.exe. No signicant difference
was found between O.H and experts, but in contrast, a signicant
difference emerged between O.H and the non-experts. Hence, in
the perceptual test, as well as in the trivia test, O.H and the experts
group did not signicantly differ in their performance, while both
showed better performance than that of the non-experts group.
3.2.2.2. Performance in the upright/inverted horse and face tasks. To
further estimate the level of visual expertise with horses, we
compared the performance for faces and horses within and between groups in accuracy (ACC), and response time (RT), regardless of orientation.
Accuracy: To directly compare the performance of the two control groups, an ANOVA with group (horse experts, non-experts) as
between-participant factor, and stimulus category (face, horse regardless of orientation) as within-participants variables was conducted for ACC. No interaction was found between the two factors
(F(1,28) 0.48, MSE0.0006, p0.49, p2 0.02), and no effect was
found for stimulus category (F(1,28) 0.12, MSE0.006, p0.73,
p2 0.0004), indicating no difference between the face and horse
tasks between groups. Note that since time to respond was unlimited, accuracy was generally very high in the face task M96.9%
(3%) for non-experts, M97.7% (2%) for experts, and for O.H,
M90% (3.8%, SD was calculated across trials). Hence response time
was used as the main dependent variable throughout the analyses
of both upright/inverted tasks. Nevertheless, despite the overall
high accuracy in these tasks, the analysis of the face task revealed
that O.H still exhibited poorer accuracy compared to the horse-experts (po0.01) and the non-experts (po0.05), thus further attesting to her difculty with faces. As for the horse task, the accuracy was again generally high (M96.1% (3%) for non-experts,
M97.8% (2.1%) for experts, and M 93.3% (4%) for O.H). No differences were found between O.H and the non-experts (p 0.19),
however, experts were better than O.H (po0.05) in accuracy.
RT: In order to examine performance by means of response
time, an ANOVA with group (horse experts, non-experts) as between-participant factor, and stimulus (face, horse regardless of
orientation) as within-participants variables was conducted for RT.
SINGLIMS was used to compare the performance of O.H to the
experts and non-experts. One-tailed paired t-test was conducted
on O.H's scores per trial, in order to compare her performance in
the face and horse tasks.
As is evident in Fig. 2, a signicant main effect of stimulus type
was obtained, (F(1,28) 10.8, MSE 74,307, po0.01, p2 0.27), such
that the two control groups were faster for faces compared to
horses. A signicant interaction between stimulus type and group

68

N. Weiss et al. / Neuropsychologia 83 (2016) 6375

Table 2
Scores and signicance levels in the trivia and perceptual expertise tests. Left panel: Percentage of correct answers and signicance of difference between O.H and each of the
control groups as obtained by SINGLIMS.exe. In the right panel, average percent scores are presented for each control group with 7std in parentheses. Signicance of
difference between groups was calculated for independent one tailed t-test. Signicant differences (i.e. p o0.05) are highlighted in bold.
O.H

Controls

%Correct Signicance of difference:

horse experts

Signicance of difference:
non-experts

%Correct: horse
experts

%Correct: nonexperts

Signicance of difference
between groups

Horse trivia test

80

p 0.41

p 0.0008

78 (8)

44 (9)

t(26) 9.46, po 0.01

Horse perceptual expertise test

73

p 0.22

p 0.007

67 (7)

55 (6)

t(21) 4.08, po 0.01

although O.H showed generally fast performance for both faces and
horses, with no RT differences compared to controls for faces (but
note her somewhat reduced accuracy in the face task), she still
exhibited better performance in the horse compared to the face
task, suggesting that her overall fast RTs may be a result of her
generally high capabilities, and stressing the specicity of her impairment to faces. These results may also imply that the experimental face task tailored to assess the inversion effect, was not
sensitive enough to reveal her basic difculties in face processing
that were evident in the diagnostic face tests (and see Duchaine and
Nakayama (2004) for a discussion regarding sensitivity of face tasks
to detect CP). As will be evident below, her results on faces are
markedly different from both control groups when stimulus orientation and eye gaze measures are taken into account, thus conrming a qualitative difference in her face perception.

Fig. 2. Response time for faces and horses (regardless of orientation), in non-experts, horse experts, and O.H. Error bars for O.H were calculated as standard error
across trials, while error bars for controls were calculated as standard error across
subjects in each group. This calculation for error bars is used for all other gures.

emerged (F(1,28) 5.38, MSE 74,307, po0.05, p2 0.28), indicating

that the non-experts were faster for faces than for horses
(po0.001), while for the experts there was no signicant difference
between horses and faces (p0.49). This interaction further conrms the difference between the two control groups and reveals
that the expertise in horses is evident in better visual discrimination for this category while not affecting performance for faces.
When examining the performance of O.H for faces and horses
using t-tests, a signicant effect was found (t(29) 2.2, po 0.05),
indicating that O.H was faster in discriminating horses than in
discriminating faces.
To examine the difference between the performance of O.H's
and the two control groups, we used RT for the face and horse
tasks and compared O.H to the experts and the non-experts
groups separately using SINGLIMS analysis. In the face task, experts and non-experts were not signicantly different from O.H
(p 0.09; p 0.1, respectively). In the horse task, experts were not
signicantly different from O.H (p 0.09), while non-experts were
slower than O.H (p 0.05).
In conclusion, non-experts were faster in faces compared to
horses, but performance of the experts did not signicantly differ
across the two tasks, indicating that expertise is evident in improvement in the category of expertise but does not affect performance in faces. These ndings, which demonstrate the visual expertise of the experts with horses are in accordance with their score
in the horse trivia questionnaire that was presented in part 2 of the
results. As for O.H, she was faster in horses compared to faces.
When comparing her performance to that of controls, we showed
that O.H was faster than the non-experts in the horse task, but did
not differ signicantly from the experts group. Importantly,

3.3. Part 3: Examining the visual processes underlying horse

expertise
In order to examine the possibility that horse expertise in O.H
and in the experts is indeed associated with holistic processing,
we compared the performance for upright and inverted stimuli
(see Section 4 for elaboration on the inversion effect and its relation to holistic processing).
An ANOVA with group (horse experts, non-expert) as betweenparticipant factor, and orientation (upright, inverted) and stimuli
(faces, horses) as within-participants variables was conducted on
RT for faces and horses. In addition, one-tailed paired t-tests, were
used on O.H's scores per trial, to examine whether she exhibit a
difference between upright and inverted stimuli.
The results of the ANOVA analysis revealed a signicant interaction
between stimulus and orientation (F(1,28) 27.2, MSE 94,777,
po0.001, p2 0.49) indicating that both control groups showed better
performance for upright compared to inverted faces (po0.001), thus
exhibiting the expected face inversion effect. However, in horses, both
groups did not show signicant effects for orientation (p0.54). Interaction with group, orientation and stimulus was not signicant
(F(1,28) 0.22, MSE94,777, p0.64, p2 0.007). Using t-tests to examine O.H's data revealed no difference in performance for upright
compared to inverted faces (t(29)  0.96, p0.35), and no difference
between upright and inverted horses (t(29)  1.44, p0.15) (see
Fig. 3.A for raw data of upright and inverted faces, and Fig. 3.B for raw
data of upright and inverted horses).
To further examine the extent of the inversion effect and to
allow its quantication and direct comparison across O.H and
controls, an index measuring the difference between upright and
inverted stimuli was calculated separately for faces and horses as
follows: Inversion index (RT inverted stimulus-RT upright stimulus)/(RT inverted stimulus RT upright stimulus)
Positive index indicates an inversion effect (better performance
for upright compared to inverted stimuli); the bigger the index, the
greater the inversion effect. Negative index indicates better

N. Weiss et al. / Neuropsychologia 83 (2016) 6375

69

performance for inverted compared to upright stimuli and zero

indicates similar performance for both upright and inverted stimuli.
To compare the performance across the two control groups, an
ANOVA with group (horse experts, non-experts) as between-participant factor, and stimuli (faces, horses) as within-participants
variables was conducted on the inversion effect as calculated
above for RT. SINGLIMS analysis was used to compare O.H's inversion effect to the effect of the two control groups.
The ANOVA revealed signicant effects for stimulus
(F(1,28) 42.66, MSE0.008, po0.001, p2 0.6), indicating a larger
inversion effect for faces than horses in both control groups, no
interactions were found between stimulus type and group
(F(1,28) 0.03, MSE0.008, p 0.86, p2 0.0008). SINGLIMS analysis,
comparing O.H to each control group revealed signicant differences between O.H and the experts and non-experts in the inversion effect for faces (po0.05) (see Fig. 3.B). These results indicate
that in contrast to the two control groups, O.H, who has profound
congenital prosopagnosia, does not show a typical inversion effect
for faces. When comparing the inversion effect for horses, no differences were found for O.H and the experts (p 0.39), as well as
the non-experts, (p0.42) (see Fig. 3.C for average performance,
and Fig. 3.D for individual data of each participant).
In summary, for both control groups, better performance was
shown for upright over inverted faces, and no differences emerged
in performance between upright and inverted horses. As for O.H, as
expected based on previous studies with CP, she showed no difference in performance between upright and inverted faces thus
showing the absence of an inversion effect (e.g. Avidan et al., 2011;
Behrmann et al., 2005a, 2005b), and more critically, she showed no
signicant difference between upright and inverted horses. As indicated by the inversion index, an inversion effect was larger for
faces than horses in both groups, larger for both groups over O.H in
faces, and did not signicantly differ between both groups and O.H
in horses, where she did not show an inversion effect.
3.4. Part 4: Eye movements

Fig. 3. Average response time for upright and inverted faces (3.A) and horses (3.B)
in non-experts, horse experts, and O.H. Panel 3.C displays the average inversion
effect, calculated for upright and inverted faces and horses in each group and in O.
H. Panel 3.D presents the individual scores of the inversion effect for each control
participant and for O.H.

In addition to the standard measures of RT and ACC, during the

upright/inverted face and horse experiments we also measured eye
movements. Since most of the gaze time was spent on the target
stimulus (among the target and two distractors, see Fig. 1), we focus
our analyses of these data on the target stimuli presented in each
trial. For each face, four interest areas (IA) were dened (left eye-the
eye located in the left visual eld of the viewer, right eye, nose,
mouth), while horses were divided to six IAs (head, neck, body,
front legs, back legs, tail) (see Fig. 4 for an example of IA's division)
and analyses were conducted on eye movements within these IAs.
Since we used a comparison task and 3 faces were presented on the
screen, the total number of xations might be affected by the back
and forth xations towards each of the faces on the screen. Hence,
the analysis was conducted on two other dependent measures,
which reect the total amount of gaze toward each of the interest
areas (IAs). In order to compare participants total gaze time, we
used normalised measures: percentage of dwell time (the total
duration across all xations in each interest area (IA), divided by the
total duration across all xations), and percentage of the number of
xations (the total number of xations falling within each IA divided by the total number of xations). Since the results obtained
with these two measures were very similar, for the sake of brevity,
we only report the results of the percentage of dwell time.
ANOVA with group (horse experts, non-experts) as betweenparticipant factor, orientation (upright, inverted), and IA as withinparticipants variables was conducted on %dwell time, for the face
task and the horse task separately (since statistical analyses for
face %dwell time was conducted for the average of both control

70

N. Weiss et al. / Neuropsychologia 83 (2016) 6375

Fig. 4. Interest areas delineated for one example of a face (4.A) and a horse (4.B)
stimulus.

Table 3
Means and p values of comparisons between %dwell time of O.H to the averaged %
dwell time of both control groups in the upright and inverted face tasks. Signicant
differences (i.e. p o0.05) are highlighted in bold.
Upright face
O.H
Control
mean
groups
mean
Left eye
Right eye
Nose
Mouth

0.29
0.14
0.16
0.01

0.04
0.02
0.12
0.06

p Value

0.00002
0.0002
0.35
0.25

Inverted face
O.H
Control
mean
groups
mean
0.22
0.12
0.19
0.02

0.08
0.05
0.06
0.02

p Value

0.02
0.02
0.02
0.5

(F(5,140) 2.5, MSE 0.0001, po0.05, p2 0.08), such that non-experts

exhibited greater %dwell time for the head IA in upright compared to
inverted horses (po0.01) and greater %dwell time for back legs in
inverted over upright horses (po0.001), while experts showed no
difference between upright and inverted orientations in these IA
(head: p0.67; back legs: p 0.32). Following this, in order to compare the %dwell time of O.H to that of the non-experts and experts
separately, SINGLIMS analysis was conducted on each interest area. For
upright horses, O.H showed no difference from both experts, and nonexperts in %dwell time towards any of the IAs. Whereas for inverted
horses, O.H showed no difference from the experts in any of the IAs,
however, she showed signicantly greater %dwell time towards the
head and the neck region compared to the non-experts (see Fig. 6 and
Table 4 for the p-values of these comparisons).
Thus, a difference between experts and non-experts was shown
in eye movements towards horses, again further corroborating the
distinction between these two groups. For an upright horse, no
difference was found between the participants in gaze pattern, as
experts, non-experts and O.H showed more %dwell time towards
the head region. However, for an inverted horse, non-experts
tended to focus more towards the back legs region than the experts, and the experts tended to focus more towards the head of
the horse than the non-experts. As for O.H, she showed no difference in gaze pattern from the experts, but showed more gaze
time than the non-experts towards the head and neck region,
when the horse was inverted.

4. Discussion
groups, averaged %dwell time and p values of both control groups
and O.H are presented in Table 3 for upright and inverted face
condition, while in Table 4 the values for upright and inverted
horse task are presented for each control group separately).
In faces, neither signicant effect, nor interactions were found
with group. A signicant interaction was found for orientation and
IA, (F(3,84) 9.558, MSE 0.004, p o0.001, p2 0.25), such that
when the face was presented upright, the %dwell time was bigger
for the IA of the nose compared to other IAs (p o0.001), with no
signicant difference between the two eyes and the mouth
(p 0.1). In contrast, when the face was inverted, %dwell time for
the nose did not differ from other IAs (p 0.37), however, it was
larger for both eyes than for the mouth (p o0.001).
Following these results, SINGLIMS analysis was conducted on
each interest area, to compare the %dwell time of O.H to that of the
controls that were pooled together (average and standard deviation
were calculated for both control groups together, since no effect or
interaction were found with group). This analysis revealed that when
the faces were upright, O.H exhibited greater %dwell time for the left
and right eyes than the controls whereas she had no signicant
difference in %dwell time towards the nose and the mouth from the
controls. When the faces were inverted, O.H exhibited signicantly
greater %dwell time for the left and right eyes and to the nose, with
no difference in %dwell time towards the mouth from the controls
(see Fig. 5 and Table 3 for the p-values of these comparisons).
To summarize, the pattern of eye movements towards faces did
not signicantly differ between the two control groups, but was
different between both groups and O.H. Both control groups
showed a larger %dwell time for the nose region when the face
was upright and for the eyes region when the face was inverted. In
contrast, O.H showed the same gaze pattern for upright and for
inverted faces, as she had larger %dwell time towards the eyes and
the nose regions whether faces were upright or inverted.
As for horses, in contrast to the results obtained for the faces, here a
signicant interaction between group, orientation, and IA emerged,

The current study was aimed to provide a new perspective on the

long-standing debate regarding the expertise hypothesis (Bukach et al.,
2006b; Gauthier and Bukach, 2007) vs. the domain specic hypothesis
(Kanwisher, 2000; McKone et al., 2007; McKone and Robbins, 2007;
Yovel and Kanwisher, 2004) of face perception. The primary subject of
our study was a young woman (O.H), who naturally developed expertise for horse perception owing to her occupation and interest,
despite her severe congenital prosopagnosia. Due to these unique circumstances, we were able to explore whether expertise in another
visual domain can evolve when face processing is drastically impaired.
To allow making direct comparisons to O.H, we tested a group of
participants who had years of experience with horses (horse experts),
and a control group of women who were novices in this domain (nonexperts). In order to match the characteristics of the control groups to
that of O.H, all of the control participants (horse experts and non-experts) were females and aged between 20 and 30 years old.
4.1. Characterizing O.H's general visual abilities
We rst set out to characterize O.H general visual abilities. This
was done by conrming the face decit in O.H by formal testing
and by examining the extent of the specicity of her decit to
faces. As expected, and in accordance with her self-report, O.H
exhibited poorer performance in all the face processing tasks
compared to the two control groups and also compared to data
sets obtained from other control groups used in our previous
studies. Nevertheless, while showing severe decits for faces, she
exhibited normal performance in a standard object processing task
(using cars). Note that based on other studies, similarly to other
CPs, O.H exhibited a reduced global bias compared to controls in
two other tasks examining holistic processing of non-facial stimuli
(Tanzer et al., 2013). These aforementioned tasks indicate the existence of a more general impairment in holistic processing in O.H,
that extends beyond faces. Thus, O.H presents a clean case of
congenital prosopagnosia, with severe face decits spanning visual

N. Weiss et al. / Neuropsychologia 83 (2016) 6375

71

Table 4
Means and p values of comparisons between %dwell time of O.H to the %dwell time of horse experts and non-experts in the upright and inverted horse tasks. Signicant
differences (i.e. p o0.05) are highlighted in bold.

Head
Neck
Body
Front legs
Back legs
Tail

Upright horse
O.H mean Horse experts mean

p Value

Non- experts mean

p Value

Inverted horse
O.H mean Horse experts mean

p Value

Non- experts mean

p Value

0.08
0.05
0.07
0.03
0
0

0.13
0.25
0.12
0.45
0.19
0.36

0.14
0.04
0.04
0.03
0.005
0.007

0.19
0.16
0.24
0.49
0.29
0.34

0.19
0.06
0.04
0
0
0

0.24
0.12
0.4
0.32
0.26
0.29

0.08
0.03
0.08
0.006
0.05
0.01

0.04
0.05
0.24
0.29
0.22
0.27

0.14
0.03
0.03
0.03
0.009
0.003

0.13
0.03
0.06
0.01
0.02
0.02

Fig. 6. Percentage of dwell time for upright horses (6.A) and inverted horses (6.B),
for the six interest areas of the target horse in horse experts, non-experts, and O.H.

Fig. 5. Percentage of dwell time for upright faces (5.A) and inverted faces (5.B), for
the four interest areas of the target face in horse experts, non-experts and O.H.

perception, short- and long-term memory of faces and extending

to holistic processing of faces and non-face stimuli, but without
evidence for a decit in object perception. This prole makes this
case an ideal candidate for examining whether visual expertise,
often attributed to holistic processing, can be developed under
such circumstances (and see Duchaine et al. (2007), Le Grand et al.
(2006) and Susilo and Duchaine (2013) for more discussion on the
specicity of CP to the domain of faces).
Horse expertise in O.H and controls was rst established using a
trivia questionnaire quantifying general, non-visual knowledge related
to horses that was developed in our laboratory. O.H and the expert
control group performed similarly on this questionnaire and their
performance was signicantly better than that of the non-experts.
Moreover, in another visual horse recognition questionnaire, in which
pictures of horses from the farm she works in were mixed with pictures of similar horses from other farms, O.H demonstrated perfect

ability to recognise by name, all the horses from her farm. Finally, in
the experimental task used in the present study, which included discrimination of upright/inverted faces and horses, aside for the generally atypical rapid performance of O.H., she exhibited better performance for horses compared to faces. Non-experts showed better
performance for the face task compared to the horse task, while experts had no difference between face and horse tasks in terms of response time. Thus, the prociency with the two object categories is
expressed in the differential performance for face and horses rather
than in the absolute performance for each of these stimuli.
To conclude this section, we established the face decit in O.H
that was specic to faces and not evident for other objects but was
yet accompanied by a more general decit in holistic processing. In
addition, we formally documented her unusual expertise in horses
which was evident in superior general knowledge regarding
horses, in a perfect ability to recognise horses she knew and in a
superior visual ability to discriminate between unfamiliar horses.
Moreover, we contrasted O.H's abilities with that of two age and
gender matched control groups that differ in their level of experience with horses. Critically, the expert group did not

72

N. Weiss et al. / Neuropsychologia 83 (2016) 6375

signicantly differ from O.H in terms of their experience with

horses.
4.2. Inversion effect for faces and horses
Several studies claimed that similarly to faces, objects of expertise are processed in a more holistic manner than other visual
objects (Diamond and Carey, 1986; Gauthier and Tarr, 1997; Wong
et al., 2009). However, there are studies showing dissociations
between visual expertise and holistic processing (e.g. Robbins and
McKone, 2007). Moreover, there are studies with visual agnosia or
prosopagnosia patients which exhibit dissociation between face
perception abilities and the ability to acquire or maintain expertise
with other objects (Sergent and Signoret, 1992; McNeil and Warrington, 1993; Duchaine et al., 2004; Assal et al., 1984). Given these
contradictory ndings, and the characteristics of O.H's visual
abilities (expertise for horses along with impaired general holistic
processing), our conservative hypothesis was that her visual expertise with horses would not necessarily entail holistic processing
as indexed by the inversion effect and by eye movement patterns.
To examine the visual processes which enable O.H to achieve
her high prociency with horses and to directly compare it with
her performance with faces, we chose the inversion effect. Specically, it is well known that face inversion drastically impairs face
perception (and compared to the effect on other objects) and this
effect is often taken as a marker for holistic processing (Brooks and
Goldstein, 1963; Goldstein, 1965; Hochberg and Galper, 1967;
Rossion, 2008; Yin, 1969 but see Richler and Gauthier (2014). Note
that we have chosen the inversion paradigm over other possible
paradigms investigating holistic processing (e.g., part-whole effect,
composite effect), since this paradigm allowed us to maintain
comparable structures of both stimuli (i.e. faces and horses), taking
into account that the main features of the face are located vertically (eyes, nose, mouth), while features of the horse are mainly
located horizontally (head, body, tail).
As predicted from previous studies, both control groups showed
similar inversion effect for faces that was not affected by expertise
in another visual domain. In contrast to controls, and consistently
with other studies with CP (Avidan et al., 2011; Behrmann et al.,
2005a, 2005b), O.H did not exhibit an inversion effect for faces, a
nding that could be attributed to impaired holistic processing for
faces (recall also her general decit in holistic processing as evident
in other non-face tasks, see top of Section 4).
The critical question then is whether horse expertise in O.H, or in
the expert controls, that was evident in their general performance in
tasks involving horses compared to faces, would be evident in an
inversion effect for horses. In accordance with our prediction, neither
control groups, nor O.H showed an inversion effect for horses. Thus,
expertise can evolve naturally despite a congenital decit in face
processing. To the best of our knowledge, this is the rst case study
showing the emergence of natural visual expertise in CP. The fact that
this ability is not accompanied by an inversion effect is consistent
with her decit with faces and with her general impairment in
holistic processing of non-face stimuli (Tanzer et al., 2013). Furthermore, and contrary to some previous studies that showed holistic
processing for objects of naturally emerging expertise (Diamond and
Carey, 1986), and for expertise acquired articially in a laboratory
(Gauthier and Tarr, 1997; Wong et al., 2009), we show that visual
expertise may also develop in healthy individuals without necessarily
relying on holistic processing mechanism, at least as indexed by the
inversion effect. This latter nding is also consistent with other behavioural studies with healthy individuals in which no evidence for
an inversion effect for objects of expertise was documented (e.g.,
Robbins and McKone, 2007). Notably, using a basic-level categorisation task, Rossion and Curran (2010) showed that car inversion did
affect performance in car experts, and moreover, inversion effect was

correlated with the level of expertise. The authors inferred that experts learn to take advantage of more holistic properties of the object
as a function of expertise. Importantly, the present study demonstrates a case of a participant who, has impaired holistic processing,
and nevertheless exhibits visual expertise for a natural object category, that was acquired since childhood.
Our results imply that there may be another mechanism underlying processing of objects of expertise that differs from the
mechanism underlying face perception and may be more analytical in nature. Some information regarding the nature of such a
mechanism is exemplied by the eye movement data provided in
the present study (see below) but this issue should also be investigated in further research.
4.3. Eye movement patterns as a function of holistic/analytic
perception
4.3.1. Faces
Eye movements were measured during the upright/inverted face
and horse experiments since they may reveal important information
which might have been missed in accuracy or RT measures. Previous
studies that measured eye movements data for upright vs. inverted
faces showed that when looking at upright faces, normal individuals
tend to xate on the eyes and the nose more than on other features
of the face (e.g. mouth, cheeks, and chin), and such a gaze pattern
was attributed to a holistic strategy (Belle et al., 2010; Schwarzer
et al., 2005, 2007). In contrast, when a more analytical strategy was
used, more gaze time was directed towards specic informative regions of the face, as a function of the task (Schwarzer et al., 2005).
In the data of the control groups, we were able to replicate
these ndings and showed that when the faces were upright, both
control groups had longer dwell time towards the nose (i.e., centre
of the face), whereas when the faces were inverted, they exhibited
longer dwell time towards the eyes region over other interest
areas. Note that stimuli were presented for an unlimited exposure
thus allowing natural viewing of the stimuli and nevertheless the
patterns for inverted and upright were different. These results are
consistent with the centre of mass of the face theory, implicating
that xating towards slightly below and between the eyes is optimal for face recognition (Belle et al., 2010; Bombari et al., 2009;
De Xivry et al., 2008; Hsiao and Cottrell, 2008). This theory is
supported by ndings showing that when a face is upright and
participants xate towards this region, they gather information
from multiple face features simultaneously (Belle et al., 2010). We
are aware of the view stating that the processing of inverted
compared to upright faces is due to quantitative rather than
qualitative cognitive differences employed for the two orientations
(Murray, 2004; Richler et al., 2011; Richler and Gauthier, 2014).
However, our converging results from both RT and eye movements
of faces indicate a different pattern of performance in controls for
the upright and inverted orientations and a similar pattern in O.H
for both stimuli. These results may support the existence of qualitative differences in performance for these two stimuli, although
adjudicating on these views was not part of the study (Barton
et al., 2001; Farah et al., 1995; Rhodes et al., 1993).
As for the results obtained for inverted faces, we note that in
contrast to our ndings and those described by Schwarzer et al.
(2005) other studies revealed somewhat different results. For example, Barton et al. (2007) and Rezlescu et al. (2014) showed that
face inversion yielded slow and systematic xations focused on
each region of the face, starting at the top of the stimuli. This
pattern was suggested to result from a bias toward looking at the
top region of a face (eyes in upright faces), which in an inverted
face would correspond to the bottom of the face (mouth and nose).
The discrepancy between these ndings and our results might be
related to the different tasks employed in the two studies. While

N. Weiss et al. / Neuropsychologia 83 (2016) 6375

Barton et al. (2006) used a memory task, we used a direct comparison task that would naturally involve back and forth gaze
shifts between the target face and distractors, instead of a continuous gaze towards a target face. Consequently, this may result
in the more localised gaze pattern obtained in our study also for
the inverted faces. Another fact that might have contributed to this
discrepancy is the difference in the stimuli used; while Barton
et al. (2006) used famous faces that were presented among unfamiliar faces, we used only unfamiliar faces. Relatedly, a different
study has shown that for upright faces, there are more xations
towards the eyes and less to the nose and mouth as a function of
familiarity (Heisz and Shore, 2008). Contrary to our ndings, another study reported no difference in the gaze pattern for upright
as compared to inverted faces (Williams and Henderson, 2007).
In this study the stimuli were also unfamiliar faces, but the task
was again, a memory task, and not a direct comparison task. It has
been suggested that even when the gaze pattern is similar towards
upright and inverted faces, participants may sample different information about the face as a function of its orientation (Belle
et al., 2010). That is, when the face is upright they look towards the
eyes (or the centre of the face) but collect information about the
whole face simultaneously, whereas when inverted, they still look
towards the eyes, but they only collect information around this
narrow band of xation area.
Most critically however, O.H showed a different pattern of results compared to both control groups attesting to her abnormal
face processing. Specically, she had longer dwell time for the eyes
in the upright as well as in the inverted faces. It is reasonable to
assume that normal viewers as well as O.H, have learned that the
most efcient strategy is to look towards the eye region, which is
most informative. As is evident, O.H's gaze patterns are not consistent with the ndings described by Schwarzer et al. (2005), in
which prosopagnosics have a more dispersed gaze pattern towards
a face. Thus, we assume that controls perceive the inverted faces in
an analytic, local manner in contrast to the holistic perception of the
upright face, while O.H perceives both upright and inverted faces
analytically. We also note that when dening our interest areas, we
did not separate the eyes from eyebrows; hence, it could be that
some xations towards the eye region were actually derived from
xations towards the eyebrows, which as mentioned in Barton et al.
(2007), comprise a region that was typically scanned by an AP individual. Moreover, external interest areas (such as hair and ears)
were concealed by a black oval frame, and hence we could not
measure the extent of reliance on these regions, as was described in
Schwarzer et al. (2007). Interestingly, unlike O.H, a number of studies that used eye tracking and other behavioural methods to infer
gaze patterns in AP patients, showed a larger tendency to extract
information from the lower part of the face (e.g. the mouth), rather
than the upper part of the face (e.g., eyes, nose) (Bukach et al., 2008;
Caldara et al., 2005; De Xivry et al., 2008; Ramon and Rossion, 2010;
Rossion et al., 2009). This difference in gaze pattern between the
acquired and congenital forms of the deciency may be an indication of the differences in the underlying mechanisms of CP and AP.
4.3.2. Horses
In contrast to the results obtained for faces, where O.H performance was dissociated from both control groups, here O.H and the
experts showed a similar gaze pattern towards horses that differ
from those of the non-experts, thus providing an additional evidence for their expertise. Specically, O.H. and the experts exhibited
a similar gaze pattern towards upright and inverted horses and
were mostly focused on the head region regardless of orientation.
Similarly, non-experts showed longer dwell time towards the head
for the upright horses, however, they presented a different gaze
pattern towards inverted horses, showing reduced dwell time towards the head, and longer dwell time towards the back legs.

73

Note that although the horse stimuli were facing to the right in
the inverted orientation, experts preferred the head region, while
non-experts gazed more at the back legs, suggesting that non-experts were perhaps affected by horse face direction, preferring the
upper-left side of the stimuli, regardless of orientation, while experts manifested a more structured strategy, preferring the head
which is a more informative region. The opposite gaze patterns are
observed for faces, as O.H (a non-expert for faces) showed a
preference towards an informative region of the face regardless of
orientation, similarly to experts in the horse task, while both control groups (who are experts for faces) show a change in strategy
for upright compared to inverted faces, similarly to the non-experts
in the horse task. These expertise gaze patterns are reminiscent of
those shown by Donovan and Litcheld (2013), and may suggest
that expertise rely on accumulated knowledge associated with object properties, that can assist in constructing a more efcient and
economical object processing strategy. In the case of horses,
knowledge about more informative local regions can improve local
processing, while in the case of faces (and perhaps other types of
objects of expertise), a more holistic pattern is adopted.
As mentioned, studies investigating behavioural expertise are
not unanimous as to whether expertise is manifested by holistic or
featural mechanisms. Importantly, the present study is the rst to
examine eye tracking measures on objects of expertise in comparison to faces and to compare the eye movement patterns in an
expert who also has CP vs. experts with normal face recognition
skills. The dissociation between face perception and objects of
expertise allowed us to show that naturally occurring visual expertise can be acquired independently from the mechanisms
mediating face recognition, and alternatively, can be acquired
through optimisation of local processing techniques. Indeed, some
studies succeeded in showing the inversion effect, that is considered a hallmark of face perception, in natural (Diamond and
Carey, 1986) and articial (Rossion et al., 2002) objects of expertise. However, our results demonstrate that different stimulus
categories may also recruit and optimise different processing
strategies, possibly according to visual and conceptual characteristics of the tested objects and task demands (Harel et al., 2013).
To conclude, it appears that visual expertise may evolve regardless of face perception abilities at least as indexed by the inversion effect. Moreover, visual skill differentially affects eye
movements towards faces and objects of expertise. These ndings
emphasise the dissociation between perception processes of faces
and objects of expertise.

Acknowledgements
The study was supported by an ISF Grant 384/10 to GA. We
would like to thank Mariah Kasyanenko and Michal Tanzer for the
upright-inverted face experiment and to Michal Tanzer for the
many fruitful discussions and innovative ideas. We thank Tzvi
Ganel for comments on earlier versions of this paper. We also
thank Ravit Yahav for her help with photographing horse images.
We sincerely thank Liat Ephraim for her cooperation and extremely valuable help with putting us in contact with the different
horse farms where we photographed the horses and for allowing
us to photograph the horses in Noam farm. We thank all the nice,
patient and cooperative horse ranchers who helped us with the
photo-shootings. We also thank Prof. Alon Friedman for helping us
with creating the horse trivia questionnaire. Last but not least, we
would like to thank all the participants, but mostly O.H, for her
great patience and willingness to take part in this study and for
providing us with her insights on face and horse recognition.

74

N. Weiss et al. / Neuropsychologia 83 (2016) 6375

Appendix A. Supplementary material

Supplementary data associated with this article can be found in
the online version at http://dx.doi.org/10.1016/j.neuropsychologia.
2015.07.028.

References
Assal, G., Favre, C., Anderes, J.P., 1984. Nonrecognition of familiar animals by a
farmer. Zooagnosia or prosopagnosia for animals. Rev. Neurol. 140, 580584.
Avidan, G., Behrmann, M., 2008. Implicit familiarity processing in congenital prosopagnosia. J. Neuropsychol. 2, 141164. http://dx.doi.org/10.1348/
174866407  260180.
Avidan, G., Tanzer, M., Behrmann, M., 2011. Impaired holistic processing in congenital prosopagnosia. Neuropsychologia 49, 25412552. http://dx.doi.org/
10.1016/j.neuropsychologia.2011.05.002.
Barton, J.J.S., Keenan, J.P., Bass, T., 2001. Discrimination of spatial relations and
features in faces: effects of inversion and viewing duration. Br. J. Psychol. 92,
527549. http://dx.doi.org/10.1348/000712601162329.
Barton, J.J.S., Radcliffe, N., Cherkasova, M.V., Edelman, J.A., 2007. Scan patterns
during the processing of facial identity in prosopagnosia. Exp. Brain Res. 181,
199211. http://dx.doi.org/10.1007/s00221-007-0923-2.
Barton, J.J.S., Radcliffe, N., Cherkasova, M.V., Edelman, J., Intriligator, J.M., 2006.
Information processing during face recognition: The effects of familiarity, inversion, and morphing on scanning xations. Perception 35, 10891105. http:
//dx.doi.org/10.1068/p5547.
Behrmann, M., Avidan, G., Marotta, J.J., Kimchi, R., 2005a. Detailed exploration of
face-related processing in congenital prosopagnosia: 1. Behavioral ndings. J.
Cogn. Neurosci. 17, 11301149. http://dx.doi.org/10.1162/0898929054475154.
Behrmann, M., Marotta, J., Gauthier, I., Tarr, M., McKeeff, T., 2005b. Behavioral
change and its neural correlates in visual agnosia after expertise training. J.
Cogn. Neurosci. 17, 554568. http://dx.doi.org/10.1162/0898929053467613.
Belle, G.V., Graef, P.D., Verfaillie, K., Rossion, B., Lefvre, P., 2010. Face inversion
impairs holistic perception: evidence from gaze-contingent stimulation. J. Vis.
10, 10. http://dx.doi.org/10.1167/10.5.10.
Bombari, D., Mast, F.W., Lobmaier, J.S., 2009. Featural, congural, and holistic faceprocessing strategies evoke different scan patterns. Perception 38, 15081521.
http://dx.doi.org/10.1068/p6117.
Boutet, I., Gentes-Hawn, A., Chaudhuri, A., 2002. The inuence of attention on
holistic face encoding. Cognition 84, 321341.
Bowles, D.C., McKone, E., Dawel, A., Duchaine, B., Palermo, R., Schmalzl, L., Rivolta,
D., Wilson, C.E., Yovel, G., 2009. Diagnosing prosopagnosia: effects of ageing,
sex, and participantstimulus ethnic match on the Cambridge face memory test
and Cambridge face perception test. Cogn. Neuropsychol. 26, 423455. http:
//dx.doi.org/10.1080/02643290903343149.
Brooks, R.M., Goldstein, A.G., 1963. Recognition by children of inverted photographs
of faces. Child Dev. 34, 1033. http://dx.doi.org/10.2307/1126545.
Bukach, C.M., Gauthier, I., Tarr, M.J., 2006b. Beyond faces and modularity: the
power of an expertise framework. Trends Cogn. Sci. 10, 159166. http://dx.doi.
org/10.1016/j.tics.2006.02.004.
Bukach, C.M., Bub, D.N., Gauthier, I., Tarr, M.J., 2006a. Perceptual expertise effects
are not all or none: spatially limited perceptual expertise for faces in a case of
prosopagnosia. J. Cogn. Neurosci. 18, 4863. http://dx.doi.org/10.1162/
089892906775250094.
Bukach, C.M., Le Grand, R., Kaiser, M.D., Bub, D.N., Tanaka, J.W., 2008. Preservation
of mouth region processing in two cases of prosopagnosia. J. Neuropsychol. 2,
227244. http://dx.doi.org/10.1348/174866407  231010.
Bukach, C.M., Gauthier, I., Tarr, M.J., Kadlec, H., Barth, S., Ryan, E., Turpin, J., Bub, D.
N., 2012. Does acquisition of Greeble expertise in prosopagnosia rule out a
domain-general decit? Neuropsychologia 50, 289304. http://dx.doi.org/
10.1016/j.neuropsychologia.2011.11.023.
Caldara, R., Schyns, P., Mayer, E., Smith, M.L., Gosselin, F., Rossion, B., 2005. Does
prosopagnosia take the eyes out of face representations? Evidence for a defect
in representing diagnostic facial information following brain damage. J. Cogn.
Neurosci. 17, 16521666. http://dx.doi.org/10.1162/089892905774597254.
Crawford, J.R., Garthwaite, P.H., 2002. Investigation of the single case in neuropsychology: condence limits on the abnormality of test scores and test score
differences. Neuropsychologia 40, 11961208. http://dx.doi.org/10.1016/
S0028-3932(01)00224-X.
Dalton, K.M., Nacewicz, B.M., Johnstone, T., Schaefer, H.S., Gernsbacher, M.A.,
Goldsmith, H.H., Alexander, A.L., Davidson, R.J., 2005. Gaze xation and the
neural circuitry of face processing in autism. Nat. Neurosci. 8, 519526. http:
//dx.doi.org/10.1038/nn1421.
De Xivry, J.-J.O., Ramon, M., Lefvre, P., Rossion, B., 2008. Reduced xation on the
upper area of personally familiar faces following acquired prosopagnosia. J.
Neuropsychol. 2, 245268. http://dx.doi.org/10.1348/174866407  260199.
Dennett, H.W., McKone, E., Tavashmi, R., Hall, A., Pidcock, M., Edwards, M., Duchaine, B., 2012. The Cambridge car memory test: a task matched in format to the
Cambridge face memory test, with norms, reliability, sex differences, dissociations from face memory, and expertise effects. Behav. Res. Methods 44,
587605. http://dx.doi.org/10.3758/s13428-011-0160-2.
Diamond, R., Carey, S., 1986. Why faces are and are not special: an effect of

expertise. J. Exp. Psychol. Gen. 115, 107117. http://dx.doi.org/10.1037/

0096-3445.115.2.107.
Donovan, T., Litcheld, D., 2013. Looking for cancer: expertise related differences in
searching and decision making. Appl. Cogn. Psychol. 27, 4349. http://dx.doi.
org/10.1002/acp.2869.
Duchaine, B., Nakayama, K., 2006. The Cambridge face memory test: results for
neurologically intact individuals and an investigation of its validity using inverted face stimuli and prosopagnosic participants. Neuropsychologia 44,
576585. http://dx.doi.org/10.1016/j.neuropsychologia.2005.07.001.
Duchaine, B., Germine, L., Nakayama, K., 2007. Family resemblance: ten family
members with prosopagnosia and within-class object agnosia. Cogn. Neuropsychol. 24, 419430. http://dx.doi.org/10.1080/02643290701380491.
Duchaine, B.C., Nakayama, K., 2004. Developmental prosopagnosia and the Benton
facial recognition test. Neurology 62, 12191220. http://dx.doi.org/10.1212/01.
WNL.0000118297.03161.B3.
Duchaine, B.C., Dingle, K., Butterworth, E., Nakayama, K., 2004. Normal greeble
learning in a severe case of developmental prosopagnosia. Neuron 43, 469473.
http://dx.doi.org/10.1016/j.neuron.2004.08.006.
Farah, M.J., Tanaka, J.W., Drain, H.M., 1995. What causes the face inversion effect? J.
Exp. Psychol. Hum. Percept. Perform. 21, 628.
Farah, M.J., Wilson, K.D., Drain, M., Tanaka, J.N., 1998. What is special about face
perception? Psychol. Rev. 105, 482.
Freire, A., Lee, K., Symons, L.A., 2000. The face-inversion effect as a decit in the
encoding of congural information: rirect evidence. Perception 29, 159170.
Gauthier, I., Tarr, M.J., 1997. Becoming a greeble expert: exploring mechanisms
for face recognition. Vis. Res. 37, 16731682. http://dx.doi.org/10.1016/
S0042-6989(96)00286-6.
Gauthier, I., Tarr, M.J., 2002. Unraveling mechanisms for expert object recognition:
bridging brain activity and behavior. J. Exp. Psychol. Hum. Percept. Perform. 28,
431446. http://dx.doi.org/10.1037/0096-1523.28.2.431.
Gauthier, I., Bukach, C., 2007. Should we reject the expertise hypothesis? Cognition
103, 322330. http://dx.doi.org/10.1016/j.cognition.2006.05.003.
Gauthier, I., Curran, T., Curby, K.M., Collins, D., 2003. Perceptual interference supports a non-modular account of face processing. Nat. Neurosci. 6, 428432.
http://dx.doi.org/10.1038/nn1029.
Golan, T., Bentin, S., DeGutis, J.M., Robertson, L.C., Harel, A., 2013. Association and
dissociation between detection and discrimination of objects of expertise:
evidence from visual search. Atten. Percept. Psychophys. 76, 391406. http://dx.
doi.org/10.3758/s13414-013-0562-6.
Goldstein, A.G., 1965. Learning of inverted and normally oriented faces in children
and adults. Psychon. Sci. 3, 447448. http://dx.doi.org/10.3758/BF03343225.
Goren, C.C., Sarty, M., Wu, P.Y.K., 1975. Visual following and pattern discrimination
of face-like stimuli by newborn infants. Pediatrics 56, 544549.
Harel, A., Bentin, S., 2013. Are all types of expertise created equal? Car experts use
different spatial frequency scales for subordinate categorization of cars and
faces. PLoS One 8, e67024. http://dx.doi.org/10.1371/journal.pone.0067024.
Harel, A., Kravitz, D., Baker, C.I., 2013. Beyond perceptual expertise: revisiting the
neural substrates of expert object recognition. Front. Hum. Neurosci. 7, 885.
http://dx.doi.org/10.3389/fnhum.2013.00885.
Haxby, J.V., Ungerleider, L.G., Clark, V.P., Schouten, J.L., Hoffman, E.A., Martin, A.,
1999. The effect of face inversion on activity in human neural systems for face
and object perception. Neuron 22, 189199. http://dx.doi.org/10.1016/
S0896-6273(00)80690-X.
Heisz, J.J., Shore, D.I., 2008. More efcient scanning for familiar faces. J. Vis. 8, 9.
http://dx.doi.org/10.1167/8.1.9.
Hershler, O., Hochstein, S., 2009. The importance of being expert: Top-down attentional control in visual search with photographs. Attention, Perception, &
Psychophysics 71, 14781486. http://dx.doi.org/10.3758/APP.71.7.1478.
Hochberg.
Hochberg, J., Galper, R.E., 1967. Recognition of faces: I. An exploratory study. Psychon. Sci. 9, 619620. http://dx.doi.org/10.3758/BF03327918.
Hsiao, J.H., Cottrell, G., 2008. Two xations sufce in face recognition. Psychol. Sci.
19, 9981006. http://dx.doi.org/10.1111/j.1467-9280.2008.02191.x.
Kanwisher, N., 2000. Domain specicity in face perception. Nat. Neurosci. 3,
759763. http://dx.doi.org/10.1038/77664.
Le Grand, R., Cooper, P.A., Mondloch, C.J., Lewis, T.L., Sagiv, N., de Gelder, B., Maurer,
D., 2006. What aspects of face processing are impaired in developmental
prosopagnosia? Brain Cogn. 61, 139158. http://dx.doi.org/10.1016/j.
bandc.2005.11.005.
Manning, D., Ethell, S., Donovan, T., Crawford, T., 2006. How do radiologists do it?
The inuence of experience and training on searching for chest nodules.
Radiography 12, 134142. http://dx.doi.org/10.1016/j.radi.2005.02.003.
Maurer, D., Grand, R.L., Mondloch, C.J., 2002. The many faces of congural processing. Trends Cogn. Sci. 6, 255260. http://dx.doi.org/10.1016/S1364-6613(02)
01903-4.
McGugin, R.W., McKeeff, T.J., Tong, F., Gauthier, I., 2010. Irrelevant objects of expertise compete with faces during visual search. Atten. Percept. Psychophys. 73,
309317. http://dx.doi.org/10.3758/s13414-010-0006-5.
McGugin, R.W., Gatenby, J.C., Gore, J.C., Gauthier, I., 2012. High-resolution imaging
of expertise reveals reliable object selectivity in the fusiform face area related
to perceptual performance. Proc. Natl. Acad. Sci. USA 109, 1706317068. http:
//dx.doi.org/10.1073/pnas.1116333109.
McKone, E., Robbins, R., 2007. The evidence rejects the expertise hypothesis: reply
to Gauthier & Bukach. Cognition 103, 331336. http://dx.doi.org/10.1016/j.
cognition.2006.05.014.
McKone, E., Kanwisher, N., Duchaine, B.C., 2007. Can generic expertise explain

N. Weiss et al. / Neuropsychologia 83 (2016) 6375

special processing for faces? Trends Cogn. Sci. 11, 815. http://dx.doi.org/
10.1016/j.tics.2006.11.002.
McNeil, J.E., Warrington, E.K., 1993. Prosopagnosia: a face-specic disorder. Q. J.
Exp. Psychol. A 46, 110. http://dx.doi.org/10.1080/14640749308401064.
Murray, J.E., 2004. The ups and downs of face perception: evidence for holistic
encoding of upright and inverted faces. Perception 33, 387398. http://dx.doi.
org/10.1068/p5188.
Ramon, M., Rossion, B., 2010. Impaired processing of relative distances between
features and of the eye region in acquired prosopagnosiatwo sides of the
same holistic coin? Cortex 46, 374389. http://dx.doi.org/10.1016/j.
cortex.2009.06.001.
Rezlescu, C., Barton, J.J.S., Pitcher, D., Duchaine, B., 2014. Normal acquisition of
expertise with greebles in two cases of acquired prosopagnosia. Proc. Natl.
Acad. Sci. USA 111, 51235128. http://dx.doi.org/10.1073/pnas.1317125111.
Rhodes, G., Brake, S., Atkinson, A.P., 1993. What's lost in inverted faces? Cognition
47, 2557. http://dx.doi.org/10.1016/0010-0277(93)90061-Y.
Richler, J.J., Gauthier, I., 2014. A meta-analysis and review of holistic face processing. Psychol. Bull. 140, 12811302. http://dx.doi.org/10.1037/a0037004.
Richler, J.J., Mack, M.L., Palmeri, T.J., Gauthier, I., 2011. Inverted faces are (eventually) processed holistically. Vis. Res. 51, 333342. http://dx.doi.org/10.1016/j.
visres.2010.11.014.
Rivolta, D., Palermo, R., Schmalzl, L., Coltheart, M., 2012. Covert face recognition in
congenital prosopagnosia: a group study. Cortex 48, 344352. http://dx.doi.
org/10.1016/j.cortex.2011.01.005.
Robbins, R., McKone, E., 2007. No face-like processing for objects-of-expertise in
three behavioural tasks. Cognition 103, 3479. http://dx.doi.org/10.1016/j.
cognition.2006.02.008.
Rossion, B., 2008. Picture-plane inversion leads to qualitative changes of face perception. Acta Psychol. 128, 274289. http://dx.doi.org/10.1016/j.
actpsy.2008.02.003.
Rossion, B., 2009. Distinguishing the cause and consequence of face inversion: the
perceptual eld hypothesis. Acta Psychol. 132, 300312. http://dx.doi.org/
10.1016/j.actpsy.2009.08.002.
Rossion, B., Curran, T., 2010. Visual expertise with pictures of cars correlates with RT
magnitude of the car inversion effect. Perception 39, 173183. http://dx.doi.org/
10.1068/p6270.
Rossion, B., Kaiser, M.D., Bub, D., Tanaka, J.W., 2009. Is the loss of diagnosticity of
the eye region of the face a common aspect of acquired prosopagnosia? J.
Neuropsychol. 3, 6978. http://dx.doi.org/10.1348/174866408  289944.
Rossion, B., Gauthier, I., Goffaux, V., Tarr, M.J., Crommelinck, M., 2002. Expertise
training with novel objects leads to left-lateralized facelike electrophysiological
responses. Psychol. Sci. 13, 250257. http://dx.doi.org/10.1111/
1467-9280.00446.
Russell, R., Duchaine, B., Nakayama, K., 2009. Super-recognizers: people with extraordinary face recognition ability. Psychon. Bull. Rev. 16, 252257. http://dx.
doi.org/10.3758/PBR.16.2.252.

75

Schmalzl, L., Palermo, R., Green, M., Brunsdon, R., Coltheart, M., 2008. Training of
familiar face recognition and visual scan paths for faces in a child with congenital prosopagnosia. Cogn. Neuropsychol. 25, 704729. http://dx.doi.org/
10.1080/02643290802299350.
Schwarzer, G., Huber, S., Dmmler, T., 2005. Gaze behavior in analytical and holistic
face processing. Mem. Cogn. 33, 344354. http://dx.doi.org/10.3758/
BF03195322.
Schwarzer, G., Huber, S., Grter, M., Grter, T., Gro, C., Hipfel, M., Kennerknecht, I.,
2007. Gaze behaviour in hereditary prosopagnosia. Psychol. Res. 71, 583590.
http://dx.doi.org/10.1007/s00426-006-0068-0.
Sekuler, A.B., Gaspar, C.M., Gold, J.M., Bennett, P.J., 2004. Inversion leads to quantitative not qualitative, changes in face processing. Curr. Biol. 14, 391396. http:
//dx.doi.org/10.1016/j.cub.2004.02.028.
Sergent, J., Signoret, J.-L., 1992. Varieties of functional decits in prosopagnosia.
Cereb. Cortex 2, 375388. http://dx.doi.org/10.1093/cercor/2.5.375.
Susilo, T., Duchaine, B., 2013. Advances in developmental prosopagnosia research.
Curr. Opin. Neurobiol. Soc. Emot. Neurosci. 23, 423429. http://dx.doi.org/
10.1016/j.conb.2012.12.011.
Susilo, T., Rezlescu, C., Duchaine, B., 2013. The composite effect for inverted faces is
reliable at large sample sizes and requires the basic face conguration. J. Vis. 13,
14. http://dx.doi.org/10.1167/13.13.14.
Tanaka, J.W., Farah, M.J., 1993. Parts and wholes in face recognition. Q. J. Exp.
Psychol. 46, 225245.
Tanzer, M., Freud, E., Ganel, T., Avidan, G., 2013. General holistic impairment in
congenital prosopagnosia: Evidence from Garner's speeded-classication task.
Cogn. Neuropsychol. 30, 429445. http://dx.doi.org/10.1080/
02643294.2013.873715.
Troje, N., Blthoff, H.H., 1996. Face recognition under varying poses: The role of
texture and shape. Vision Research 36, 17611771.
Weiner, K.S., Grill-Spector, K., 2013. Neural representations of faces and limbs
neighbor in human high-level visual cortex: evidence for a new organization
principle. Psychol. Res. 77, 7497. http://dx.doi.org/10.1007/
s00426-011-0392-x.
Williams, C.C., Henderson, J.M., 2007. The face inversion effect is not a consequence
of aberrant eye movements. Mem. Cogn. 35, 19771985. http://dx.doi.org/
10.3758/BF03192930.
Wong, A.C.-N., Palmeri, T.J., Gauthier, I., 2009. Conditions for facelike expertise with
objects becoming a ziggerin expertbut which type? Psychol. Sci. 20,
11081117. http://dx.doi.org/10.1111/j.1467-9280.2009.02430.x.
Yin, R.K., 1969. Looking at upside-down faces. J. Exp. Psychol. 81, 141145. http:
//dx.doi.org/10.1037/h0027474.
Young, A.W., Hellawell, D., Hay, D.C., 1987. Congurational information in face
perception. Perception 16, 747759.
Yovel, G., Kanwisher, N., 2004. Face perception: domain specic, not process specic. Neuron 44, 889898. http://dx.doi.org/10.1016/j.neuron.2004.11.018.