Vous êtes sur la page 1sur 7

Facial Perception

Oliver Wilhelm, University Ulm, Ulm, Germany


Andrea Hildebrandt, Ernst-Moritz-Arndt University Greifswald, Greifswald, Germany
Werner Sommer, Humboldt-Universität zu Berlin, Berlin, Germany
Ó 2015 Elsevier Ltd. All rights reserved.

Abstract

In this article, we summarize popular theories of face identity and facial expression processing. Next, we present some general
findings on individual differences in this field including considerations regarding more common and specific face identity
and facial expression processing assessment principles. We then discuss available measures and present a task-taxonomy,
followed by a short summary of key findings in the field. We then present a summary of research results concerning
neural underpinnings in identity and emotion processing. Finally, we conclude with some remarks about the emerging field
of identity and expression investigations and some research heuristics.

Faces provide indispensable information for human communi- gyri provide input to the lateral fusiform gyrus, where invariant
cation. They inform the perceiver about the familiarity and aspects of faces – arguably cues of unique identities – are
identity of others and about those persons’ emotional states processed. This region has been found to respond more strongly
and intentions. The mechanisms of facial information process- to the perception of faces than to the perception of nonface
ing and the respective neural correlates received wide attention objects (Kanwisher, 2000). Based on this and similar findings,
in the last years. Most of this research is based on experimental the so-called fusiform face area (FFA), situated in the fusiform
findings in healthy participants and evidence from persons gyrus, has been postulated to be a specialized module for face
with innate or acquired disorders of face processing. This article perception. If attention is directed to changeable aspects of
will focus on variability in these abilities across persons and the faces, like emotion expression or lip movement, the superior
structure of individual differences in facial information pro- temporal sulci are activated. Visually derived semantic codes,
cessing, after providing a summary of theories built upon biographical, and name information are processed in the
experimental and clinical research. extended system (Haxby et al., 2000). Beyond the core system,
an extended neural system – specialized on different cognitive
functions – is considered to interact with face-responsive
Theories of Face Identity and Facial regions (Haxby et al., 2000). We highlight four such systems.
Expression Processing First, the intraparietal sulcus is considered to be in charge of
processing spatially directed attention and thought to aid the
Functional models of facial information processing (e.g., perception of eye gaze direction. Second, the audiovisual inte-
Haxby et al., 2000; Young and Bruce, 2011) postulate a hierar- gration in speech perception (see the McGurk effect), which is
chical structure, including view-centered pictorial and view- supported by the interplay of the auditory and superior
independent structural encoding of the face, access to stored temporal cortex. Third, during emotion processing the superior
representations of known faces in memory, and to associated temporal sulcus maintains connections with the amygdala,
(biographical) knowledge about the person and his or her insula, and limbic system, which are the neural systems
name. In addition to identity processing, these models allow involved in the processing of emotion. Fourth, access to
predictions about how the face processing system deals with semantic and biographical information during face processing
variable information provided by faces – in particular, appears to be mediated by anterior temporal regions. Taken
emotional expressions, facial movements accompanying together, face processing is supported by multiple brain regions
speech, and gaze direction. Further, socially relevant informa- that are spatially disparate, yet act together in a coordinated
tion that can be derived from the invariant face structure inde- way. The mechanisms of facial information processing are
pendent of the person’s identity – i.e., age, gender, and supposed to be hierarchically organized and actual models
judgments about attractiveness – is mapped in these models assume an interplay between the identity and expression pro-
as visually derived semantic codes. Early theorizing assumed cessing routes. Furthermore, Haxby et al. (2000) emphasize
independent streams of identity versus expression-related that brain systems for facial information processing do not
information. This initial idea was recently modified in favor work in isolation but contribute to other cognitive functions
of partial dependence views (e.g., Calder, 2011; Young and such as object recognition.
Bruce, 2011). However, it is not yet fully clarified how these
systems interact.
Research on the general physiological implementation of Individual Differences in Face Identity
face-based processes distinguishes between a core system and and Expression Processing
an extended system (Haxby et al., 2000). The core system
consists of three bilateral, hierarchically organized brain Neurocognitive theories of face identity and facial expression
regions. During initial stages of face perception, facial features processing describe general mechanisms and the functional
are processed in the inferior occipital gyri. The inferior occipital organization of the facial information processing system. The

676 International Encyclopedia of the Social & Behavioral Sciences, 2nd edition, Volume 8 http://dx.doi.org/10.1016/B978-0-08-097086-8.25060-5
Facial Perception 677

taxonomy, stability, and predictive validity of individual ought to be considered lower order factors in a higher order
differences in these cognitive mechanisms have been rather model of socioemotional abilities. Based upon the process
recently – at least partly – addressed as a set of independent theories delineated in the introduction facial information-
research questions. related decoding abilities may be further subdivided into
Given the neuroanatomic evidence on at least partly distinct perceptual and mnestic aspects of performance. Additionally,
brain localizations of core aspects of face identity and facial it is not yet settled whether or not the perception of
expression processing one might be tempted to infer that the neutral and emotional faces needs to be distinguished on an
abilities to process neutral and emotional faces are highly individual differences level. Importantly, broader theories of
distinct and independent. However, all of the proposed tasks socioemotional abilities that postulate a series of emotion
for measuring facial information processing have attributes of receptive performance components ought to include more or
prototypical ability tests: Participants are willing to do their less specific abilities of processing identity and emotion in
best, there is an objective correct response, and performance the voice, gestures, and postures. Obviously, systematically
is assessed as degree to which responses are correct. It is well deriving ability measures of such components that are
known that any two such measures will covary positively in grounded in basic research and incorporating them into factor
unselected samples – a phenomenon labeled positive manifold models of socioemotional abilities is a major challenge for
in ability research. Therefore, it is unrealistic to expect that (1) future research.
measures of face perception that capture individual differences Apart from distinguishing perceptual and mnestic demands
will be completely independent of general cognitive abilities of tasks challenging either identity or expression processing of
measures and (2) measures of facial emotion processing will faces, there is one more crucial task attribute requiring consid-
be orthogonal to measures of face identity processing. Addi- eration: whether performance is expressed as speed or accu-
tionally, when taking a closer look at emotion recognition tasks racy. In experimental research many of the tasks used to
Adolphs (2002) concluded that emotion identification may be provoke face identity and facial expression processing are
carried out as a combination of processes reflecting perception, easy to solve. Comparisons between experimental conditions
semantic knowledge, and reasoning. The independence and are then usually based on the latency of responses. Within
redundance of performance in a variety of measures is best ability research such measures would be considered processing
investigated by studying their factorial structure. speed tasks because participants from the application popula-
How do we understand the emotional states of others? tion solve a very high proportion of trials correctly and indi-
Lipps (1907) suggested that persons imitate (mimic) observed vidual differences are primarily manifest in the response
affective behavior and that this mimicry elicits a corresponding latency. It is important to note that easy tasks and hard tasks
state in the observer, facilitating or enabling affect under- usually load on different ability factors. Latency-related ability
standing. This idea is echoed in modern embodied simulation factors may or may not arise as a function of the information
theories. Indeed, facial mimicry of emotional expressions is content. Indeed, measures challenging the speed of object
frequently observed, although in some cases only with the processing are essentially perfectly collinear with measures
aid of electromyographic recordings. Mimicry reactions are challenging the speed of face identity and facial expression
often involuntary, automatic, and occur within a fraction of processing (Hildebrandt et al., 2012, 2013). This is not true
a second (Dimberg et al., 2000). The idea that simulation is for measures of face identity and facial expression processing
fundamental for the perception of emotions in others is in scored by the accuracy of responses (Wilhelm et al., 2010).
line with findings that hindering incidental mimicry leads to Given the distinction between speed and accuracy tasks that
performance impairment in emotion perception (e.g., Neal is well established in ability research and the high collinearity
and Chartrand, 2011). However, whether persons who show of speed-related abilities that lack facets of identity or emotion
more incidental mimicry are also better emotion perceivers processing, we recommend avoiding measures of face identity
and whether they do so in all situations is less clear. Hess and facial expression processing exclusively challenging the
and colleagues did not observe such a relationship between speed of performance.
mimicry and emotion perception in young adults (e.g., Hess After narrowing down the face identity and expression
and Blairy, 2001). Therefore, facial mimicry may not be neces- processing ability domains based on well-funded experimental
sary for the classification of prototypical emotional displays research, it is instrumental to consider prior work on socioe-
but may be helpful for social binding or when emotion recog- motional abilities from other fields of individual differences
nition is difficult. In line with the latter view, Künecke et al. research. The so-called jingle fallacy refers to instances in
(2014), using a multivariate approach, showed a positive rela- which constructs are deemed the same, although they only
tionship between the degree of facial corrugator response and share a similar verbal label. Conversely, the jangle fallacy
the ability of emotion perception. refers to instances in which constructs are deemed different
In the popular four-branch theory of emotional abilities, although the only obvious distinction is the verbal label
face-based emotion processing is included as one of the key established within a scientific community. The recently
components of emotional intelligence (Mayer et al., 2008). popularized theory of mind and measures arguably capturing
In the component process models (Scherer, 2007) face-based individual differences in this ability are probably an instance
emotion processing is included as receptive decoding. Wilhelm of a jangle fallacy. The Reading the Mind in the Eyes Test
and colleagues (2010) claimed that face identity perception displays photographs of the eye region of 36 actors and
and recognition are core aspects of such emotional abilities. actresses. Participants are to select which of several response
If emotional intelligence exists (Borsboom et al., 2004) indi- options best matches what those persons think or feel.
vidual differences in perceiving neutral and emotional faces Obviously, the stimuli and the task presented here are akin
678 Facial Perception

to provoking facial expression processing. It is therefore participants are asked to order the morphs according to their
important to solidly clarify the status of individual similarity with the target face. In the CFMT, participants are
differences in such theory of mind tasks. Presupposing asked to learn and – after a very brief retention interval – recog-
adequate psychometric characteristics of the measures, we nize pictures of six faces in different conditions (identical,
would predict very high correlations between theory of mind different angles, different lighting, and degraded through
and emotion identification (i.e., a jangle fallacy). In that visual noise). Both of the Cambridge tests are certainly supe-
case, the theory of mind ability may not need distinct rior to many earlier measures.
theorizing. Despite this somewhat mixed assessment concerning
ability tasks relying on neutral faces as stimuli, recent develop-
ments in the field are encouraging. On the one side, we recently
presented a task battery that allows measuring individual
Measures of Face Identity and Facial differences in perceiving and recognizing neutral faces and
Expression Processing were able to establish and replicate the distinction of these
abilities from object recognition and established ability factors
One way to investigate the status quo of individual differences (Wilhelm et al., 2010). Within the task battery, an additional
research in this area is to go through available ability measures aim was to distinguish between tasks challenging either speed
and search for tasks that use faces as stimuli. We will describe or accuracy of processing – a prominent distinction in intelli-
a few such measures – beginning with tasks challenging iden- gence research outlined above (Carroll, 1993). The multitask
tity processing and proceeding to measures of facial expression approach for measuring face identity processing-related abili-
processing – and highlight their positive and negative attri- ties presented by Wilhelm et al. (2010) includes task para-
butes. It is important to note that we restrict the presentation digms widely used in experimental research on face
here to protypical ability measures; meaning the behavior of processing (Part–Whole, Composite Faces, and the Inversion
participants is assessed by comparing responses to an objective paradigms), different memory paradigms (delayed non-
standard of performance. matching, immediate recall, delayed recall, implicit
In the famous faces test, participants are asked to pick the memory), and over 500 face identities across all tasks. The
correct name out of a few alternatives for each famous face factorial structure of the tests has been established across
shown. As a measure of identity recognition, this type of test a series of independent samples.
is conceptually problematic because it neglects differences With respect to facial expression processing, many
between persons in prior exposure to the face stimuli and measures have an intrinsic component of perceptual speed
knowledge about the displayed famous persons. In the War- because stimulus exposition is limited in time (i.e., Japanese
rington Recognition Memory for Faces Test no control con- and Caucasian Brief Affective Recognition Test (Matsumoto
cerning stimulus attributes irrelevant for face cognition, like et al., 2000)). The popular Diagnostic Analysis of Nonverbal
clothing, hair, or paraphernalia such as glasses, is applied. Accuracy (DANVA; Nowicki and Carton, 1993), which chal-
Therefore, these irrelevant stimulus attributes might be used lenges recognition of happiness, sadness, anger, and fear in
to solve the task. Presumably, this is specifically true for partic- emotional faces, also has a strictly time-limited stimulus
ipants with severe problems in processing faces, such as proso- exposition. The manipulations of expression intensity in the
pagnosic patients. In the Benton Facial Recognition Test, DANVA would probably allow the derivation of accuracy
participants are presented a target photo and are asked to select scores to express performance. However, the tasks have not
the target identity from six simultaneously presented response yet been probed as measures of accuracy.
alternatives. In some items participants need to identify an The Profile of Nonverbal Sensitivity (Rosenthal et al., 1979) is
identical picture, in others they have to pick the correct identity another popular measure, which uses recordings of faces,
when pictures are taken from different angles or in different voices, and bodies as stimuli to assess emotion recognition
lighting conditions. Although stimulus control is much better from multimodal and dynamic stimuli. The use of dynamic
here, participants can (and do) still rely on eyebrows and hair facial stimuli is an interesting aspect of the measure but an
lines in processing faces. The objections raised above also important drawback here is the limitation to only two emotion
apply to most of the available measures of socioemotional categories (positive versus negative).
abilities and learning and recognition of face-tasks included One out of several measures designed for the assessment of
in many intelligence and memory test batteries. Additionally, potential disorders of affect processing is the Florida Affect
in many of these tasks components unrelated to identity and Battery (Bowers et al., 1999), which uses static facial, vocal,
expression processing contribute to performance. and cross-modal stimuli and multiple methods
In order to solve issues with face external stimulus compo- (discrimination, naming, selection, and matching). This test
nents, Duchaine and Nakayama (2006) developed the is arguably too easy for unimpaired subjects.
Cambridge Face Perception Task (CFPT) and the Cambridge The Multimodal Emotion Recognition Test relies on 10 cate-
Face Memory Test (CFMT). These tests include only face gories of emotion expressions but the measure is limited to
internal features and the perception task successfully mini- affect naming as a single observational method. The Geneva
mizes mnestic demands. The CFPT applies morphing: the crea- Emotion Recognition Test (Schlegel et al., 2014) features
tion of new faces through blending two different faces into dynamic and multimodal actor portrayals of 14 emotions –
each other. Target faces are morphed with several other faces amusement, anger, disgust, despair, pride, anxiety, interest,
varying the similarity of the morphed faces with the target irritation, joy, fear, pleasure, relief, surprise, and sadness –
face. The target face is presented along with the morphs and but exclusively relies on affect naming.
Facial Perception 679

Two tasks challenging perceptual matching and explicit A second critical distinction is between speed performance
labeling of expression with satisfying psychometric back- (i.e., latencies of correct responses) and the accuracy of
ground were recently presented (Palermo et al., 2013). Both performance (i.e., proportion of correct responses in tasks
tasks build upon the six basic emotions (anger, disgust, fear, that are so hard that some participants cannot solve them
happiness, sadness, and surprise). The first task applies an regardless of how much time is available). A third important
odd-man-out paradigm (i.e., participants have to indicate categorization refers to predominant perceptual or mnestic
which out of three simultaneously presented pictures demands of tasks.
displays a different emotion). The second task requires Several larger studies critically test these distinctions. For
participants to explicitly identify the emotion in a target face example, based on two larger studies Wilhelm et al. (2010)
by selecting one of six written labels. provided evidence for three factors of face cognition: (1) the
The validity and utility of highly specific information pro- accuracy of face perception, (2) the accuracy of face memory,
cessing paradigms like the Emotional Stroop Test, for example, and (3) the speed of perceiving and recognizing unfamiliar
(Matthews et al., 2003) is still unclear. faces. These factors were replicated in an independent sample
Most of the above approaches do not allow abstracting of younger adults with new identities as face stimuli
from task-specific variance – a crucial step toward studying (Hildebrandt et al., 2013). These studies show a clear distinc-
latent variables (ability estimates) rather than manifest raw tion between speed and accuracy factors, which were nearly
scores. Unfortunately, multivariate approaches are more often completely independent of each other. The shared variance
an exception than the rule in the field. Additionally, stimulus between the performance accuracy and speed of face cognition
sampling seems to be particularly relevant when creating items was less than 5%. This finding has a strong implication for
and compiling tests (Judd et al., 2012). Very recently Wilhelm theories and research on face processing, because it shows
et al. (2014) proposed a task battery for measuring facial that research outcomes can tell completely different stories,
expression perception and recognition. The same distinction depending on whether speed or accuracy is analyzed. Addi-
of speed versus accuracy-based measures for face identity pro- tionally, these studies show unequivocally that perception of
cessing has also been implemented for the expression process- and memory for faces are distinguishable abilities on the basis
ing tasks. The battery includes five tasks for measuring facial of accuracy indicators, but not on the basis of speed indicators.
expression perception and identification that are based on The expectation of a distinction between perception and
the composite paradigm, the face inversion paradigm, memory was based on the assumption stressed in face process-
a measure of emotion discrimination in morphed continua ing models, that at a perceptual stage structural encoding of
between facial expressions that share action units, and a visual faces is central, whereas at a memory stage, the access to face
search paradigm with emotional expressions of different inten- representations stored in memory is critical. Obviously,
sity. The battery also includes recognition measures that memory performance will partly depend on successful percep-
successfully avoid confusing identity and emotion processing tion. Interestingly, such a perception versus memory differen-
and applies different pictures in the learning versus recognition tiation is not feasible on the basis of speed data. Individual
stages. In order to avoid biases arising through applying the differences in the swiftness of recognizing well-learned faces
same methods for performance assessment repeatedly, multi- are perfectly collinear with face perception speed. In order to
variate achievement tests should vary memory cues, viewing further clarify the status of the speed factor of face cognition,
angles, expression intensity, and the like. the convergent validity of this factor was studied. On the level
of latent variables the speed of face cognition was not distinct
from the speed of processing facial emotional expressions and
Summary of Key Findings both facial speed factors could not be distinguished from the
speed of object cognition. Therefore, the swiftness of
Theories of face-based receptive communicative abilities – as performance in these three domains has zero specificity
reviewed above – are mainly originating from the cognitive- (Hildebrandt et al., 2013).
experimental literature and were investigated primarily with With respect to the two factors that describe and explain
experimental, neurophysiological, and neurobiological data. performance in the perception and memory accuracy tasks
However, the structure and predictive validity of mental there is strong and replicated evidence that they need to be
functions such as face perception, face memory, emotion dissociated from general cognition, object cognition, imme-
perception, and emotion memory can only be determined by diate and delayed memory for words, numbers and figures,
studying individual differences in a variety of adequate tasks. and mental speed. In several samples of young adults, a large
Overarching theories based on such data – for example, the amount of variance that is specific for the face cognition factors
four-branch theory of emotional intelligence (e.g., Mayer was replicated. In the first of these studies there was 46% face
et al., 2008) – are conceptually and theoretically perception specific variability and 52% face memory specific
controversial, mainly because their grounding in more basic variability. In an additional study (Hildebrandt et al., 2011),
science is insufficient. Based on the theoretical distinctions covering an age range from 18 to 88 years, there was no moder-
and models reviewed above, we postulate a taxonomic ator effect of age on the relationship between face perception
model of individual differences in facial information and face memory versus general cognition, memory, and
processing that is – arguably – supported by available data. mental speed. Therefore, this study shows the absence of
Following the well-established distinction in the field, a aging-related dedifferentiation between face cognition on
first differentiation of emotional abilities is between recep- one side and several factors of general cognitive functioning
tive (decoding) and productive (encoding) performance. on the other side. Thus, the specificity of face cognition did
680 Facial Perception

not diminish across adult age, despite the expected substantial Once a nontrivial degree of specificity of emotion process-
age-related performance declines. Taken together these results ing-related factors is established, the emotion specificity is an
support the assumption that individual differences in face important research question. It will be interesting to see how
cognition constitute abilities with substantial uniqueness strongly indicators of performance in different emotion cate-
across the adult range. gories (happiness, surprise, fear, sadness, disgust, and anger)
Conceptually, face perception and face memory can be will be related with each other. Although it has been argued
considered necessary but not sufficient for emotion perception that persons may differ in their accuracy to identify specific
(identification) and memory for emotional expressions. Their emotional expressions, because persons may be predomi-
empirical distinctness is an essential question to be addressed. nantly faced with different affects during early socialization
Contemporary neurocognitive models of face processing (e.g., Tomkins and McCarter, 1964), first evidence suggests
describe partial overlap between the identity and expression a lack of emotion category specificity of individual differ-
processing routes. Recently, Palermo et al. (2013) established ences (Schlegel et al., 2014). Multivariate research designs
correlational evidence in favor of a common first step of iden- will be able to provide more conclusive evidence concerning
tity and expression processing. A factorial model that approx- this question.
imates models of facial information processing may have
a nested structure. A theoretical model is depicted in Figure 1.
To simulate the hierarchical structure of processing face- Neural Underpinnings of Individual
related information – differentiated through successive pro- Difference in Face Cognition
cessing stages in neurocognitive models – the model first
postulates a general factor of face perception. This general The neural mechanisms underlying individual differences in
factor captures interindividual variability of face perception face cognition may provide a deeper understanding of why
tasks and accounts for face perception-related variability in people differ in these abilities. Performance measures used to
all other indicator tasks measuring face memory and facial derive ability factors are end products of the information pro-
emotion processing. Second, to represent a partial overlap of cessing occurring within the neurocognitive system. In
the routes processing face identity versus expression that also contrast, neuroscientific methods can provide information
holds at later stages, the factorial model includes a broad about the success, organization, and timing of intermediate
face memory factor that accounts for variability in all memory processing steps that lead to the manifest, end products of
tasks. Importantly, this factor is purified from variance due to performance.
perceptual processing. Two further more specific factors Neuroimaging studies on individual differences in face pro-
account for variability due to processing emotion-related cessing-related abilities are surprisingly rare and confined to
information only. small samples of usually not more than 30 participants. The
The model (1) implies a partial dependency of emotion responsivity in bilateral anterior temporal poles and right ante-
processing from identity processing and (2) allows for speci- rior hippocampus correlated positively with self-rated
ficity of emotion processing over and above identity cognition. preexperimental familiarity of face stimuli. Rotshtein et al.
However, the degree of specificity of emotion processing- (2007) reported that the behavioral impact of second-order
related factors partialled for face identity processing-related (configural) information of faces correlated with the
variance is still an unanswered research question. The postu- activation of bilateral inferior occipital gyrus and right
lated model needs to be tested on the basis of accuracy data fusiform gyrus. Both studies rely only on small samples. For
because speed measures of facial emotion processing and the a mixed group of normal and prosopagnosic participants Furl
speed of face cognition are perfectly collinear with speed et al. (2011) reported that better performance in facial
measures lacking interpersonally relevant content. identity processing was positively associated (1) with face

Figure 1 Schematic theoretical model of facial identity and emotion processing; ovals represent latent variables and squares depict indicators used
to measure the latent variables.
Facial Perception 681

selectivity in fusiform gyri bilaterally and left anterior temporal faces have been associated with better face cognition perfor-
lobe, (2) with peak face selectivity in the right and left mance. The amplitude of the ERE for learned faces was the
individually defined FFA, and (3) with the size of the right FFA. strongest predictor of face cognition. The behavior versus
Efforts to tap the neurocognitive mechanisms underlying ERP latency associations were not face specific, because they
face cognition in large samples using multivariate modeling diminished after partialing general cognitive functioning.
are currently restricted to the analysis of event-related brain A few studies investigated the neural underpinnings of indi-
potentials (ERPs). ERPs are complex responses of the brain vidual differences in emotional facial expressions. Kim et al.
to discrete events, such as the presentation of a stimulus, and (2003) reported that more intense subjective experience of
consist of distinct components that can be related to specific surprise is correlated with stronger amygdala activation as
cognitive subprocesses. Amplitudes and latencies of ERPs measured in fMRI. Using an extreme group design Corden
reflect the intensity and the timing of the processes reflected et al. (2006) found less activation of amygdala, fusiform,
in these components. Therefore, it is of interest whether and anterior temporal cortex in response to neutral faces for
individual differences in specific ability factors are related to participants scoring low on a fear recognition task as compared
amplitudes and latencies of specific performance-relevant to a group with normal scores in this task. An ERP study by
ERPs. Of particular interest here are ERPs that have been Tamamiya and Hiraki (2013) found that N170 amplitude pre-
related to the putative stages of face processes as postulated dicted the recognition accuracy of happy and angry faces.
in the functional models outlined above.
The occipital P100 is related to the processing of domain-
general, low-level stimulus features and is characterized by Conclusions
a positive peak at occipital scalp sites around 75–135 ms; it
is thought to be generated in extrastriate visual brain areas. Research on face identity and facial expression processing is
The N170 is characterized by a negative peak at a rapidly moving field with exciting new developments in
occipitotemporal sites around 150–190 ms, typically larger many areas. These developments include elaborations on
for faces than for other objects, but still unaffected by the neurocognitive models into the more applied field of indi-
familiarity of faces. Therefore, it is considered to reflect the vidual differences. One consequence of this elaboration is
configural encoding of facial features and their integration that more and better measurement instruments will be avail-
into a holistic percept. Various brain areas have been able for research purposes and applied assessment. Research
proposed as sources of the electrical N170, including results emerging from this field also have the potential to
fusiform, lingual, or occipitotemporal gyri. The Early dramatically improve our current concepts of socioemotional
Repetition Effect (ERE) or N250r is usually operationalized abilities. In opposition to ex-cathedra ideas about the status
as the difference wave between ERPs to primed and and organization of socioemotional abilities, an approach
unprimed faces in priming tasks; it is most pronounced at grounded in neurocognitive work on identity and emotion
temporal-parietal sites around 260–330 m and presumably processing is obviously a more solid foundation of
originates in the fusiform gyrus. The ERE and Late Repetition psychological research. Based upon such a foundation,
Effect (LRE) (see below) are more pronounced for familiar as a variety of intriguing research questions can be proposed.
compared to unfamiliar faces and for personally familiar Some of these questions concerning face identity and facial
than for famous faces. Hence, the ERE is associated with the expression processing not addressed in the present
activation of structural representations of faces in long-term contribution include behavior genetics, molecular genetics,
memory and with the identification of familiar faces (e.g., sex differences, relevance for a variety of mental disorders,
Schweinberger and Burton, 2011). Finally, the LRE or N400 developmental trajectories in childhood, youth, and aging,
– also typically observed in priming experiments – has been training and interventions based upon identities and
related to the activation of person-related knowledge in long- expressions, and many more. Research in all of these areas
term memory (e.g., Schweinberger and Burton, 2011). It is hinges upon adequate measurement instruments. Due to the
characterized as the difference wave between ERPs to primed relevance of these instruments, we would like to conclude
and unprimed faces, usually consisting of a parietal positivity with statements that might seem evident, obvious, and
between 330 and 500 m. According to the model for familiar somewhat boring but nevertheless have not been adhered to
face recognition (see above) the neural generators of the LRE sufficiently in the past.
are considered to be part of the extended neural system The unfolding individual differences research agenda in the
involved in the retrieving of person-related knowledge (i.e., field of face identity and facial expression processing is best
anterior temporal cortex, precuneus, posterior cingulate advised to adhere to the following simple heuristics: Human
cortex). abilities should be measured with ability indicators. Self-
Kaltwasser et al. (2014) investigated how these ERP compo- report measures are not ability indicators. Therefore, self-
nents are related with face cognition abilities and with general report measures should not be used to study socioemotional
cognitive functioning. All ERP components could be modeled abilities. Theoretical claims about abilities should be based
as unidimensional reliable latent variables. Several such exper- on evidence that shows adequate coherence and
imentally independent ERPs were only moderately correlated, dependability within measures. Several such measures
suggesting that they ought to be considered indicators of designed to capture the same ability should show coherence
specific, essentially independent neurocognitive processes. and dependability across measures. Once combinations of
Shorter latencies of the N170 component along with shorter established abilities are partialed from such a set of measures
latencies and larger amplitudes of the ERE and LRE for learned a nontrivial amount of residual variance should be present.
682 Facial Perception

This residual variance should predict criteria or outcomes of Kim, H., Somerville, L.H., Johnstone, T., Alexander, A.L., Whalen, P.J., 2003. Inverse
relevance. The presently available evidence summarized in amygdala and medial prefrontal cortex responses to surprised faces. Neuroreport
14, 2317–2322.
this contribution endorses the idea that some of the
Künecke, J., Hildebrandt, A., Recio, G., Sommer, W., Wilhelm, O., 2014. Facial EMG
proposed socioemotional ability factors are distinct from responses to emotional expressions are related to emotion perception ability. PLOS
established ability factors. This is no small achievement ONE 9 (1): e84053.
given the fate of many initially intriguing but soon Lipps, T., 1907. Das Wissen von fremden Ichen. In: Lipps, T. (Ed.), Psychologische
abandoned ideas brought forward in ability research. Untersuchungen. Engelmann, Leipzig, pp. 694–722.
Matthews, G., Roberts, R.D., Zeidner, M., 2003. Development of emotional
intelligence: a skeptical- but not dismissive perspective. Human Development 46,
See also: Emotion and Expression; Emotion, Perception and 109–114.
Expression of; Emotional Intelligence and Competencies; Matsumoto, D., LeRoux, J., Wilson-Cohn, C., Raroque, J., Kooken, K., Ekman, P.,
Emotional Regulation; Emotions, Sociology of; Emotions: Yrizarry, N., Loewinger, S., Uchida, H., Yee, A., Amo, L., Goh, A., 2000. A new
test to measure emotion recognition ability: Matsumoto and Ekman’s Japanese
Methods of Assessment; Facial Emotion Expression, Individual and Caucasian Brief Affect Recognition Test (JACBART). Journal of Nonverbal
Differences in; Factor Analysis and Latent Structure Analysis: Behavior 24, 179–209.
Confirmatory Factor Analysis; Factor Analysis and Latent Mayer, J.D., Roberts, R.D., Barsade, S.G., 2008. Human abilities: emotional
Variable Models in Personality Psychology; Personality intelligence. Annual Review of Psychology 59, 507–536.
Neal, D.T., Chartrand, T.L., 2011. Embodied emotion perception: amplifying and
Assessment; Personality, Trait Models of; Social Intelligence
dampening facial feedback modulates emotion perception accuracy. Social
and Competencies; Social and Emotional Development in the Psychological and Personality Science 2, 673–678.
Context of the Family. Nowicki, S., Carton, J., 1993. The measurement of emotional intensity from facial
expressions; the DANVA FACES 2. Journal of Social Psychology 133,
749–751.
Palermo, R., O’Connor, K.B., Davis, J.M., Irons, J., McKone, E., 2013. New tests to
Bibliography measure individual differences in matching and labelling facial expressions of
emotion, and their association with ability to recognise vocal emotions and facial
Adolphs, R., 2002. Neural systems for recognizing emotion. Current Opinion in identity. PLoS ONE 8, e68126.
Neurobiology 12, 169–177. Rotshtein, P., Geng, J.J., Driver, J., Dolan, R.J., 2007. Role of features and
Borsboom, D., Mellenbergh, G.J., van Heerden, J., 2004. The concept of validity. second-order spatial relations in face discrimination, face recognition, and
Psychological Review 111, 1061–1071. individual face skills: behavioral and functional magnetic resonance imaging
Bowers, D., Blonder, L.X., Heilman, K.M., 1999. Florida Affect Battery, A Manual. data. Journal of Cognitive Neuroscience 19, 1435–1452.
Centre for Neuropsychological Studies, Cognitive Science Laboratory, University of Rosenthal, R., Hall, J.A., DiMatteo, M.R., Rogers, P.L., Archer, D., 1979. Sensitivity to
Florida. Nonverbal Communication: The PONS Test. John Hopkins University Press,
Calder, A.J., 2011. Does facial identity and facial expression recognition involve Baltimore.
separate visual routes? In: Calder, A.J., Rhodes, G., Johnson, M.H., Haxby, J.V. Schlegel, K., Grandjean, D., Scherer, K.R., 2014. Introducing the Geneva emotion
(Eds.), The Oxford Handbook of Face Perception. Oxford University Press, Oxford, recognition test: an example of Rasch-based test development. Psychological
pp. 427–448. Assessment 26, 666–672.
Carroll, J.B., 1993. Human Cognitive Abilities: A Survey of Factor-Analytic Studies. Scherer, K., 2007. Componential emotion theory can inform models of emotional
Cambridge University Press, Cambridge, United Kingdom. competence. In: Matthews, G., Zeidner, M., Roberts, R.D. (Eds.), Science of
Corden, B., Critchley, H.D., Skuse, D., Dolan, R.J., 2006. Fear recognition ability Emotional Intelligence: Knowns and Unknowns. Oxford University Press, Oxford,
predicts differences in social cognitive and neural functioning in men. Journal of pp. 101–126.
Cognitive Neuroscience 18, 889–897. Schweinberger, S.R., Burton, A.M., 2011. Person perception 25 years after Bruce
Dimberg, U., Thunberg, M., Elmehed, K., 2000. Unconscious facial reactions to and Young (1986): an introduction. British Journal of Psychology 102,
emotional facial expressions of emotions. Psychological Science 11, 86–89. 695–703.
Duchaine, B., Nakayama, K., 2006. The Cambridge Face Memory Test: results for Tamamiya, Y., Hiraki, K., 2013. Individual differences in the recognition of facial
neurologically intact individuals and an investigation of its validity using inverted expressions: an event-related potentials study. PLoS ONE 8.
face stimuli and prosopagnosic participants. Neuropsychologia 44, 576–585. Tomkins, S., McCarter, R., 1964. What and where are the primary affects? Some
Furl, N., Garrido, L., Dolan, R.J., Driver, J., Duchaine, B., 2011. Fusiform gyrus face evidence for a theory. Perceptual and Motor Skills 18, 119–158.
selectivity relates to individual differences in facial recognition ability. Journal of Wilhelm, O., Hildebrandt, A., Manske, K., Schacht, A., Sommer, W., 2014. Test
Cognitive Neuroscience 23, 1723–1740. battery for measuring the perception and recognition of facial expressions of
Haxby, J.V., Hoffman, E.A., Gobbini, M.I., 2000. The distributed human neural system emotion. Frontiers in Psychology, Emotion Science 5, 404.
for face perception. Trends in Cognitive Sciences 4, 223–233. Wilhelm, O., Herzmann, G., Kunina, O., Danthiir, V., Schacht, A., Sommer, W., 2010.
Hess, U., Blairy, S., 2001. Facial mimicry and emotional contagion to dynamic Individual differences in face cognition. Journal of Personality and Social
emotional facial expressions and their influence on decoding accuracy. Psychology 99, 530–548.
International Journal of Psychophysiology 40, 129–141. Young, A.W., Bruce, V., 2011. Understanding person perception. British Journal of
Hildebrandt, A., Wilhelm, O., Schmiedek, F., Herzmann, G., Sommer, W., 2011. On Psychology 102, 959–974.
the specificity of face cognition compared to general cognitive functioning across
adult age. Psychology and Aging 26, 701–715.
Hildebrandt, A., Schacht, A., Sommer, W., Wilhelm, O., 2012. Measuring the speed Relevant Websites
of recognising facially expressed emotions. Cognition and Emotion 26, 650–666.
Hildebrandt, A., Wilhelm, O., Herzmann, G., Sommer, W., 2013. Face and object
cognition across adult age. Psychology & Aging 28, 243–248. http://www.faceblind.org/ – Prosopagnosia Research Centers at Harvard University
Judd, C.M., Westfall, J., Kenny, D.A., 2012. Treating stimuli as a random factor in and University College London.
social psychology: a new and comprehensive solution to a pervasive but largely http://www.face-rec.org/ – Face Recognition Homepage.
ignored problem. Journal of Personality and Social Psychology 103, 54–69. http://scienceaid.co.uk/psychology/cognition/face.html – Science aid: face recognition.
Kaltwasser, L., Hildebrandt, A., Recio, G., Wilhelm, O., Sommer, W., 2014. Neuro- http://ukcatalogue.oup.com/product/9780199559053.do – Oxford Handbook of Face
cognitive mechanisms of individual differences in face cognition: a replication and Perception.
extension. Cognitive, Affective, and Behavioral Neuroscience 14, 861–878.
Kanwisher, N., 2000. Domain specificity in face perception. Nature Neuroscience 3,
759–763.

Vous aimerez peut-être aussi