People are highly skilled at recognizing familiar faces due to the development of identity representations that are invariant to viewpoint and low-level features. However, how such view-invariant representations emerge for unfamiliar faces remains debated. This EEG study examined the neurophysiological basis of face learning, and how viewpoint generalization is shaped by experience (i.e., increasing familiarity) and stimulus realism. Adult participants performed an identity-matching task with novel faces varying along a realism continuum, from photographs to less and more abstract artistic styles (Renaissance vs. Cubism). Identity (same vs. different) and viewpoint (repeated vs. changed) were orthogonally manipulated within face pairs. Early face structural encoding (N170; 150-190 ms) was modulated by face realism and by viewpoint-related perceptual demands. The subsequent N250r (220-270 ms) was sensitive to the formation of viewpoint-invariant identity representations in memory, with different temporal dynamics across face types: viewpoint tolerance increased gradually for photographs with repeated exposure, whereas it emerged earlier for stylized faces. After repeated exposure, a reliable N250r was also observed for unseen, novel views, suggesting that learning supports progressively more stable and potentially more abstract identity representations. These findings are discussed in light of stimulus structure and the role of top-down processes in the formation of view-invariant representations.

This study examined the category specificity of a previously reported reciprocal interference between holistic face and word processing by incorporating body stimuli as a control. Prior work has demonstrated that superimposed faces and words disrupt each other’s holistic processing: faces are processed less holistically when paired with aligned words, and vice versa. Here, we asked whether a similar pattern of interference extends to body stimuli. In Experiment 1, we assessed whether face alignment would influence holistic body processing. We found no difference in holistic body processing between face-aligned and face-misaligned conditions, suggesting that faces do not interfere with holistic body processing. In Experiment 2, we examined whether word alignment affected body processing and again found no evidence of interference. Together, these findings indicate that body processing is unaffected by the alignment of overlaid words or faces, highlighting both the specificity and the limits of shared holistic processing mechanisms across high-level visual categories.

Face typicality and distinctiveness are key facial attributes that influence face recognition performance and the formation of social impressions. The present study aimed to provide normative data for these dimensions, offering a useful resource for face recognition research. Using a 7-point Likert scale, adult participants rated 304 front-facing faces from the Glasgow Unfamiliar Face Database (GUFD) for typicality–distinctiveness. Results indicated that the subjective rating method produced reliable estimates, with meaningful variability observed along the typicality–distinctiveness continuum. Highly distinctive faces were more sparsely represented in the database. These norms can support principled stimulus selection and improved methodological control in empirical research with faces.

What is the impact of dynamic changes in facial visibility on identifying the eye and forehead region? This study examines how wearing or removing a mask affects the ability to visually identify the eyes. We investigate whether these changes impact the recognition of upper facial features and alter sensitivity to the misalignment of the face's upper and lower halves, which disrupts holistic face processing. Results show that removing a mask generally impairs visual identification, suggesting that the perception of the whole face hinders recognition of the upper half. This hindering is evident from the fact that the interference decreases when the face is misaligned. In contrast, the impairment in identification caused by adding a mask to a target face rises from losing original support of the holistic processing, given that it was not diminished when the upper and bottom halves were misaligned. Additional findings show that misalignment negatively affects the identification of faces where a mask was either maintained or added, suggesting that masks may actually help direct attention to relevant facial features, rather than being integrated into a holistic representation. We discuss these results in light of their theoretical and practical implications for visual identification, particularly in the context of dynamic changes to facial appearance.

Faces have been shown to be primarily processed by the right hemisphere, such that there is an advantage in their processing when they are presented in the left portion of the visual field. In the present research, we explore face lateralization in the context of faces in paintings. While prior research in art perception has explored hemispheric asymmetries, the role of a right-hemispheric advantage is still debated. Using a hemifield paradigm, we explored the accuracy of face identification in photographs of real faces, as well as in paintings with different art styles, differing in the extent to which they are realistic (renaissance) or distorted (cubism). Performance for photographs was higher when presented in the left half of the visual field, displaying a right hemisphere lateralization; similarly, a right hemisphere advantage in face processing was found for faces in paintings, regardless of how realistic the art style was. All in all, we found evidence of a right hemisphere advantage across stimuli categories. We consider these in relation to other phenomena in face processing literature (e.g., pareidolia), and discuss implications based on these parallels, namely regarding the time course of face perception.

Faces are the most important social signal in our society. Nevertheless, there is a problem with faces: they are all made up of the same features in the same general order (the eyes are above the nose, which is above the mouth). To process faces one uses a special kind of processing, which is holistic, considering the integration of the face’s features and their relative distances. One may distinguish the recognition of known faces and the processing of unfamiliar faces. Face processing abilities may be lost due to either a lesion or developmental reasons, i.e., prosopagnosia. To further explore these reasons, one could consider pictorial representations of faces—such as faces in paintings. These are particularly interesting because different art styles differ in how realistic/distorted they are relative to real faces, which allows for exploring people’s sensitivity to face-likeness. In a way, individuals are not sensitive to face-likeness. In face matching part–whole tasks, performance does not differ across art styles. Still, individuals are not fully impervious to distortion: early markers of face processing (N170 component) are sensitive to face-likeness, with more realistic (vs. distorted) art styles eliciting responses more in line with those of real faces.

Human face processing relies heavily on configural processing-the ability to perceive the spatial relationships between facial features-rather than on processing individual features in isolation. In this study, we explored configural processing of faces in paintings, comparing artworks along a realism/distortion continuum, including Renaissance, Postimpressionism, Expressionism, and Cubism. To test competing hypotheses regarding a possible modulation by art style in relational processing, we asked how inversion hinders configural processing of faces in paintings. We found gradation in relational processing, with an inversion effect for Renaissance and Postimpressionism face paintings but no inversion effect for Expressionism and Cubism face paintings. Distortion in the positioning of facial features introduced in Expressionism and Cubism is most probably related to the lesser importance of relational processing in these art styles, whereas the higher realism of faces in Renaissance and Postimpressionist paintings fosters the recruitment of mechanisms involved in the processing of real faces.

Previous research has shown that, like faces, words are processed either holistically or through the automatic representation of their parts combined. The automaticity assumed to underlie the holistic processing of words presupposes that individuals have a relatively low level of control over these processes. However, they may also be capable of learning from their environments whether processing words as a whole is the most efficient processing strategy—which would require at least some control over the corresponding processes. In fact, previous research supports this latter account in the context of the holistic processing of faces: when provided a task in which participants should ignore half of a stimuli (the irrelevant part) and pay selective attention to the other half (the target part), the participants become better at ignoring the irrelevant part when it is commonly misleading (i.e., this suggests a response that is different from that of the relevant part in the context of the task). In the present work, we extend these considerations to holistic word processing. Our results support a learned attentional account in the context of holistic word processing. When an irrelevant word part is systematically helpful for the judgment of a target word half, participants engage more in holistic processing (vs. when the irrelevant word half is misleading). This reflects an incidental statistical learning process in which individuals identify the irrelevant word half as either providing helpful or misleading information about the target half.

In recent years, increased attention has been devoted to visual word recognition under a perceptual expertise framework. Because the information required to identify words is distributed across the word, a holistic attentional strategy is optimal and develops with experience. It is, however, an open question the extent to which other information embedded in a word may contribute to word holistic processing, namely sublexical word properties. In the present research, we therefore explore the role of sublexical properties-specifically bigram transition probabilities-in this processing strategy. We used a common task in the holistic processing literature (i.e., composite task) and four-letter disyllabic words, where two of the bigrams reinforce the cohesiveness of each syllable and one of the bigrams reinforces the cohesiveness between the syllables. We found preliminary evidence of a role of these sublexical properties in word holistic processing.

Holistic processing aids in the discrimination of visually similar objects, but it may also come with a cost. Indeed holistic processing may improve the ability to detect changes to a face while impairing the ability to locate where the changes occur. We investigated the capacity to detect the occurrence of a change versus the capacity to detect the localization of a change for faces, houses, and words. Change detection was better than change localization for faces. Change localization outperformed change detection for houses. For words, there was no difference between detection and localization. We know from previous studies that words are processed holistically. However, being an object of visual expertise processed holistically, visual words are also a linguistic entity. Previously, the word composite effect was found for phonologically consistent words but not for phonologically inconsistent words. Being an object of visual expertise for which linguistic information is important, letter position information, is also crucial. Thus, the importance of localization of letters and features may augment the capacity to localize a change in words making the detection of a change and the detection of localization of a change equivalent.

Faces and words are ever-present stimuli in social environments that require fine-grained, efficient discrimination of their constituents in order to acquire meaning. Provided that these stimuli share multiple characteristics, while simultaneously being different visual object categories in important ways, a debate has ensued pertaining to whether their processing can be reduced to a common mechanism or whether each category mobilizes exclusive resources. We thus first present briefly domain-specific and domain-general accounts, as opposing perspectives that highlight the absence and presence of commonalities in face and word processing, respectively. We then focus on how faces and words are processed. While faces are usually associated with holistic processing of facial features, to create a perceptual whole, there is no such consensus pertaining to word processing. Words have been argued to rely on either letter-by-letter processing or in a way closer to that of faces, since they are also objects of expertise. Lastly, we advance the debate by providing an overview of our latest research findings. These findings provide a more direct comparison of face and word processing, by incorporating both stimuli in one task concurrently.

Faces are generally assumed to be processed holistically, that is, features are represented in an integrated fashion. Similarly, pictorial representations of faces (e.g., drawings) have been shown to elicit holistic processing. Some researchers, however, have contested the concept of holistic face processing, suggesting that the perception of a face is no more than the sum of individual face parts. In the present study, we ask whether faces in paintings are processed holistically and, if so, whether this holistic processing is consistent across art styles along the realism-distortion dimension. Additionally, we seek to understand whether other factors, such as interest in art and exposure to art (e.g., visiting museums), as well as general visual recognition abilities, contribute to the potential holistic processing of faces in paintings. We found holistic face processing across stimulus sets, suggesting that holistic processing of faces in art occurs regardless of the characteristics of the art style (i.e., realism/distortion).Moreover, general interest in art showed a marginally negative correlation with holistic face processing. In contrast, general visual recognition abilities correlated positively with holistic processing, suggesting that increased capacity to process purely visual information benefits perceptual integration and grouping.

Behavioral research (Ventura, et al., 2023) suggested that pictorial representations of faces varying along a realism-distortion spectrum elicit holistic processing as natural faces. Whether holistic face neural responses are engaged similarly remains, however, underexplored. In the present study, we evaluated the neural correlates of naturalist and artistic face processing, by exploring electrophysiological responses to faces in photographs versus in four major painting styles. The N170 response to faces in photographs was indistinguishable from that elicited by faces in the renaissance art style (depicting the most realistic faces), whilst both categories elicited larger N170 than faces in other art styles (post-impressionism, expressionism, and cubism), with a gradation in brain activity. The present evidence suggest that visual processing may become finer grained the more the realistic nature of the face. Despite behavioral equivalence, the neural mechanisms for holistic processing of natural faces and faces in diverse art styles are not equivalent.

The other-race effect (ORE) is a well-known phenomenon in which people discriminate and recognize faces from their ethnic group more accurately than faces from other ethnic groups. Holistic processing, or the mandatory tendency to process all parts of an object together, has been proposed as an explanation for the ORE. According to the holistic perspective of the ORE, other-race faces might be subject to weaker holistic processing than own-race faces. However, evidence for this hypothesis is inconsistent. Although it is generally assumed that holistic processing helps the individuation of objects, holistic processing may also come at a cost. Specifically, holistic processing may reduce the capacity to localize changes in the constituent parts of an object, but not in detecting changes to an object as a whole. In the present study, we examined change detection and change localization accuracy for Caucasian and African faces, and houses. Performance was better for change detection than change localization for Caucasian faces. While clear costs of holistic processing for Caucasian faces were thus found, the difference between change localization and change detection was not obvious for African faces. However, childhood exposure to other-race people correlated with change detection for African faces, but not with change localization for African faces. Our results thus show that holistic processing of other-race faces may depend on early contact with other-race people.

The question of whether word and face recognition rely on overlapping or dissociable neural and cognitive mechanisms received considerable attention in the literature. In the present work, we presented words (aligned or misaligned) superimposed on faces (aligned or misaligned) and tested the interference from the unattended stimulus category on holistic processing of the attended category. In Experiment 1, we found that holistic face processing is reduced when a face was overlaid with an unattended, aligned word (processed holistically). In Experiment 2, we found a similar reduction of holistic processing for words when a word was superimposed on an unattended, aligned face (processed holistically). This reciprocal interference effect indicates a trade-off in holistic processing of the two stimuli, consistent with the idea that word and face recognition may rely on non-independent, overlapping mechanisms.

Recently, paradigms in the face recognition literature have been adopted to reveal holistic processing in word recognition. It is unknown, however, whether different measures of holistic word processing share similar underlying mechanisms, and whether fluent word reading relies on holistic word processing. We measured holistic processing effects in three paradigms (composite, configural sensitivity, part-whole) as well as in reading fluency (3DM task: reading aloud high- and low-frequency words and pseudowords). Bin scores were used to combine accuracy and response time variables in the quest for a more comprehensive, reliable, and valid measure of holistic processing. Weak correlations were found between the different holistic processing measures, with only a significant correlation between the configural sensitivity effect and part-whole effect (r = .32) and a trend of a positive correlation between the word composite effect and configural sensitivity effect (r = .21). Of the three holistic processing measures, only one (part-whole effect) correlated with a lexical access measure of 3DM (r = .23). We also performed a principal component analysis (PCA) of performance in the three lists of 3DM, with the second most probably reflecting lexical access processes. There was a tendency for a positive correlation between part-whole bin measure and Component 2 of PCA. We also found a positive correlation between composite aligned in accuracy and Component 2 of PCA.Our results show that different measures of holistic word processing reflect predominantly different mechanisms, and that differences among normal readers in word reading do not seem to depend highly on holistic processing.

We investigated the relationship between holistic processing and face processing using a latent variables approach. Three versions of the composite paradigm were used to measure holistic processing: Vanderbilt Holistic Face Processing Test, a sequential composite matching task, and a simultaneous composite matching task. Three tasks were used to measure face perception and face memory abilities respectively. We had three pairs of tasks such that within each pair (of memory and perception task), the stimuli involved, the requirement for matching across viewpoints, etc., are the same, such that the only difference is whether perception or memory is taxed. There were no significant correlations between the different versions of the composite task. We discovered no evidence to support a distinction between face perception and face memory, suggesting the existence of a general face processing factor. Finally, there was no evidence that holistic processing (as captured by either of the three composite tasks) is predictive of better face processing per se, casting doubts on the role of holistic processing in differentiating different levels of efficiency in face processing.

A dual-route account of holistic processing has been proposed, which includes a stimulus-based and experience-based approach to holistic processing. The bottom-up route was suggested by the observation of holistic processing for novel Gestalt line patterns in the absence of expertise. For words, there is mainly evidence for a late, lexical, experience-based locus of holistic processing with scarce evidence for an early, stimulus-based locus. However, salient early Gestalt information (i.e., connectedness, closure, and continuity between parts) are important for letter and word identification. Thus, there might be an overlap at an early, perceptual processing stage, between Gestalt stimulus-based holistic processing and word holistic processing. In the task we used, words and Gestalt line patterns were superimposed, and we evaluated whether one class of stimuli was processed less holistically when an aligned other class pattern (processed holistically) was superimposed. There was some evidence supporting an early locus for the influence of word processing on Gestalt line patterns, but the interaction between the two stimuli was not reciprocal, which needs further clarification. When an aligned word (processed holistically) was overlaid on a line pattern, the line pattern was processed less holistically. However, when an aligned line pattern (processed holistically) was overlaid on a word, the word was not processed less holistically. This pattern might result from the higher cohesiveness of words and their automaticity and feedback from the lexicon.

data are for Dprime and RT for two experiments with stimul consisting of words and line patterns superimposed. In Exp. 1, participants were instructed to pay attention to words in a word composite task.In Exp. 2, participants were instructed to pay attention to line patterns in a line pattern composite task.

Previous studies had reported qualitatively identical holistic face processing (using the composite task) already in 6-year-olds. In the present study, we evaluated these processes quantitatively by evaluating whether children are less efficient in dealing with and encoding faces. Thus, in the present study we explored the time course of holistic processing in children and adults by manipulating stimulus presentation time. The study composite task was presented for 800 ms. The test composite face was presented either for 183 ms or 800 ms. Our youngest participants, 6-year-olds, process faces holistically with the same efficiency as typical adults.