People automatically evaluate faces on multiple trait dimensions, and these evaluations predict important social outcomes, ranging from electoral success to sentencing decisions. Based on behavioral studies and computer modeling, we develop a 2D model of face evaluation. First, using a principal components analysis of trait judgments of emotionally neutral faces, we identify two orthogonal dimensions, valence and dominance, that are sufficient to describe face evaluation and show that these dimensions can be approximated by judgments of trustworthiness and dominance. Second, using a data-driven statistical model for face representation, we build and validate models for representing face trustworthiness and face dominance. Third, using these models, we show that, whereas valence evaluation is more sensitive to features resembling expressions signaling whether the person should be avoided or approached, dominance evaluation is more sensitive to features signaling physical strength/weakness. Fourth, we show that important social judgments, such as threat, can be reproduced as a function of the two orthogonal dimensions of valence and dominance. The findings suggest that face evaluation involves an overgeneralization of adaptive mechanisms for inferring harmful intentions and the ability to cause harm and can account for rapid, yet not necessarily accurate, judgments from faces.emotions ͉ face perception ͉ social cognition T he belief that the nature of the mind and personality could be inferred from facial appearance has persisted over the centuries. References to this belief can be dated back to ancient Greece, Rome, and China (1). In the 19th century, the pseudoscience of physiognomy reached its apogee. Cesare Lombroso, the founder of criminal anthropology, argued that ''each type of crime is committed by men with particular physiognomic characteristics''. For example, ''thieves are notable for their expressive faces and manual dexterity, small wandering eyes that are often oblique in form, thick and close eyebrows, distorted or squashed noses, thin beards and hair, and sloping foreheads'' (2). Lombroso provided his ''scientific'' testimony at several criminal trials.Although modern science, if not folk psychology (3), has largely discarded such notions, trait evaluations from faces predict important social outcomes ranging from electoral success (4-6) to sentencing decisions (7,8). Studies show that people rapidly evaluate faces on multiple trait dimensions such as trustworthiness and aggressiveness (9, 10). For example, trait judgments can be formed after as little as 38-ms exposure to an emotionally neutral face (10). Studies also show that the amygdala, a subcortical brain region critical for fear conditioning and consolidation of emotional memories (11), plays a key role in the assessment of face trustworthiness (12-15).Why do mechanisms for rapid spontaneous face evaluation exist if they do not necessarily deliver accurate inferences? This apparent puzzle from an evolutionary point of view can be resolved b...

No abstract

We thank Amir Goren, Chris Olivola, Chris Said, Crystal Hall, Sara Verosky, and Sean Baron for their comments on previous drafts of this manuscript.

Recent years have seen an increase in the popularity of multivariate pattern (MVP) analysis of functional magnetic resonance (fMRI) data, and, to a much lesser extent, magneto- and electro-encephalography (M/EEG) data. We present CoSMoMVPA, a lightweight MVPA (MVP analysis) toolbox implemented in the intersection of the Matlab and GNU Octave languages, that treats both fMRI and M/EEG data as first-class citizens. CoSMoMVPA supports all state-of-the-art MVP analysis techniques, including searchlight analyses, classification, correlations, representational similarity analysis, and the time generalization method. These can be used to address both data-driven and hypothesis-driven questions about neural organization and representations, both within and across: space, time, frequency bands, neuroimaging modalities, individuals, and species. It uses a uniform data representation of fMRI data in the volume or on the surface, and of M/EEG data at the sensor and source level. Through various external toolboxes, it directly supports reading and writing a variety of fMRI and M/EEG neuroimaging formats, and, where applicable, can convert between them. As a result, it can be integrated readily in existing pipelines and used with existing preprocessed datasets. CoSMoMVPA overloads the traditional volumetric searchlight concept to support neighborhoods for M/EEG and surface-based fMRI data, which supports localization of multivariate effects of interest across space, time, and frequency dimensions. CoSMoMVPA also provides a generalized approach to multiple comparison correction across these dimensions using Threshold-Free Cluster Enhancement with state-of-the-art clustering and permutation techniques. CoSMoMVPA is highly modular and uses abstractions to provide a uniform interface for a variety of MVP measures. Typical analyses require a few lines of code, making it accessible to beginner users. At the same time, expert programmers can easily extend its functionality. CoSMoMVPA comes with extensive documentation, including a variety of runnable demonstration scripts and analysis exercises (with example data and solutions). It uses best software engineering practices including version control, distributed development, an automated test suite, and continuous integration testing. It can be used with the proprietary Matlab and the free GNU Octave software, and it complies with open source distribution platforms such as NeuroDebian. CoSMoMVPA is Free/Open Source Software under the permissive MIT license.Website: http://cosmomvpa.orgSource code: https://github.com/CoSMoMVPA/CoSMoMVPA

Judgments of trustworthiness from faces determine basic approach/avoidance responses and approximate the valence evaluation of faces that runs across multiple person judgments. Here, based on trustworthiness judgments and using a computer model for face representation, we built a model for representing face trustworthiness (study 1). Using this model, we generated novel faces with an increased range of trustworthiness and used these faces as stimuli in a functional Magnetic Resonance Imaging study (study 2). Although participants did not engage in explicit evaluation of the faces, the amygdala response changed as a function of face trustworthiness. An area in the right amygdala showed a negative linear response-as the untrustworthiness of faces increased so did the amygdala response. Areas in the left and right putamen, the latter area extended into the anterior insula, showed a similar negative linear response. The response in the left amygdala was quadratic--strongest for faces on both extremes of the trustworthiness dimension. The medial prefrontal cortex and precuneus also showed a quadratic response, but their response was strongest to faces in the middle range of the trustworthiness dimension.

People rapidly form impressions from facial appearance, and these impressions affect social decisions. We argue that data-driven, computational models are the best available tools for identifying the source of such impressions. Here we validate seven computational models of social judgments of faces: attractiveness, competence, dominance, extroversion, likability, threat, and trustworthiness. The models manipulate both face shape and reflectance (i.e., cues such as pigmentation and skin smoothness). We show that human judgments track the models' predictions (Experiment 1) and that the models differentiate between different judgments, though this differentiation is constrained by the similarity of the models (Experiment 2). We also make the validated stimuli available for academic research: seven databases containing 25 identities manipulated in the respective model to take on seven different dimension values, ranging from -3 SD to +3 SD (175 stimuli in each database). Finally, we show how the computational models can be used to control for shared variance of the models. For example, even for highly correlated dimensions (e.g., dominance and threat), we can identify cues specific to each dimension and, consequently, generate faces that vary only on these cues.

Using a dynamic stimuli paradigm, in which faces expressed either happiness or anger, the authors tested the hypothesis that perceptions of trustworthiness are related to these expressions. Although the same emotional intensity was added to both trustworthy and untrustworthy faces, trustworthy faces who expressed happiness were perceived as happier than untrustworthy faces, and untrustworthy faces who expressed anger were perceived as angrier than trustworthy faces. The authors also manipulated changes in face trustworthiness simultaneously with the change in expression. Whereas transitions in face trustworthiness in the direction of the expressed emotion (e.g., high-to-low trustworthiness and anger) increased the perceived intensity of the emotion, transitions in the opposite direction decreased this intensity. For example, changes from high to low trustworthiness increased the intensity of perceived anger but decreased the intensity of perceived happiness. These findings support the hypothesis that changes along the trustworthiness dimension correspond to subtle changes resembling expressions signaling whether the person displaying the emotion should be avoided or approached.

Many lines of evidence point to a tight linkage between the perceptual and motoric representations of actions. Numerous demonstrations show how the visual perception of an action engages compatible activity in the observer's motor system. This is seen for both intransitive actions (e.g., in the case of unconscious postural imitation) and transitive actions (e.g., grasping an object). Although the discovery of “mirror neurons” in macaques has inspired explanations of these processes in human action behaviors, the evidence for areas in the human brain that similarly form a crossmodal visual/motor representation of actions remains incomplete. To address this, in the present study, participants performed and observed hand actions while being scanned with functional MRI. We took a data-driven approach by applying whole-brain information mapping using a multivoxel pattern analysis (MVPA) classifier, performed on reconstructed representations of the cortical surface. The aim was to identify regions in which local voxelwise patterns of activity can distinguish among different actions, across the visual and motor domains. Experiment 1 tested intransitive, meaningless hand movements, whereas experiment 2 tested object-directed actions (all right-handed). Our analyses of both experiments revealed crossmodal action regions in the lateral occipitotemporal cortex (bilaterally) and in the left postcentral gyrus/anterior parietal cortex. Furthermore, in experiment 2 we identified a gradient of bias in the patterns of information in the left hemisphere postcentral/parietal region. The postcentral gyrus carried more information about the effectors used to carry out the action (fingers vs. whole hand), whereas anterior parietal regions carried more information about the goal of the action (lift vs. punch). Taken together, these results provide evidence for common neural coding in these areas of the visual and motor aspects of actions, and demonstrate further how MVPA can contribute to our understanding of the nature of distributed neural representations.