Holistic processing, as measured in the composite-face paradigm, was disrupted by contextual cues that discouraged the perceptual grouping of the face parts (Exp. 1). Of note is the ability of such cues to impact holistic face perception even when the face parts themselves are aligned, and thus the configuration of the facial features is intact. In addition, the face-specific nature of this effect provided further support for the impact of perceptual grouping cues on specialized face-processing mechanisms.
Furthermore, the orientation-sensitive nature of this effect at short (200 ms), but not at long (2,500 ms), encoding durations is also consistent with an impact of perceptual grouping cues on holistic face perception (Exps. 2 and 3). Specifically, when the encoding duration was sufficient to allow for upright and inverted faces to be processed holistically (Richler, Mack, et al., 2011), we found an effect of the perceptual grouping manipulation on performance for both inverted and upright faces (Exp. 2). However, when the encoding time was reduced so as to no longer allow inverted faces to be processed holistically, contextual grouping cues only impacted upright face performance (Exp. 3). Notably, the impact of perceptual grouping cues on holistic processing of upright faces for both long (2,500 ms) and short (200 ms) encoding durations is consistent with previous reports that holistic processing can occur with as little as 183 ms presentations of upright faces (Richler, Mack, et al., 2011). This ability to predict and manipulate whether inverted and/or upright face perception would be impacted by our perceptual grouping manipulation, using knowledge of the distinct time courses of holistic processing for these stimuli (as previously documented by Richler, Mack, et al., 2011), provides strong evidence against a role of more general stimulus-based differences, unrelated to holistic processing, in driving these effects. Thus, the orientation sensitivity and temporal properties of the effect of perceptual grouping cues on face perception is consistent with the ability of such cues to impact specialized holistic-processing mechanisms.
The effect of perceptual grouping cues on holistic face processing, as indexed via the composite-face effect, offers insight into our understanding of the effect of misalignment on holistic face perception. A number of influential theories of face perception have proposed the existence of a face “norm” or “template” and that stimuli matching this general template trigger specialized face-processing mechanisms (e.g., Leopold, O’Toole, Vetter, & Blanz, 2001; Rhodes, Brennan, & Carey, 1987; Valentine, 1991). In the context of such theories, changing the configuration of the features within a face by misaligning the top and bottom halves presumably impacts holistic processing because the resulting stimulus no longer matches an internal face template. Thus, such stimuli would no longer trigger specialized (holistic) face perception mechanisms. However, the fact that the prototypical face configuration remained intact in our perceptual grouping manipulation suggests that misalignment may disrupt holistic processing not only by impacting the physical match of the face to an internal template, but also by breaking the perceptual grouping of the parts. This finding is not necessarily inconsistent with the existence of an internal face template (but see Richler, Mack, et al., 2011; Richler et al., 2008), but it does suggest that the perceived grouping of the facial features, not just their physical configuration, can interfere with the matching of a stimulus to an internal face template.
Recent work by Taubert and colleagues has suggested that both a template-matching process and a holistic-grouping process contribute to face processing, but on different time scales (Taubert & Alais, 2011; Taubert, Apthorp, Aagten-Murphy, & Alais, 2011). Specifically, the holistic process was suggested to occur before the template-matching process. Notably, abundant evidence has suggested that perceptual grouping mechanisms, such as those impacted by the contextual manipulation used in the experiments reported here, operate very early on in processing (Kimchi & Hadad, 2002). Thus, the potential of multiple distinct mechanisms to contribute to the unique hallmarks of face processing—that is, an earlier holistic-grouping mechanism and a later template-matching process—allows for the integration of the findings reported here with template-matching models of face perception.
Cues discouraging the grouping of face parts might disrupt holistic face perception via their impact on object-based attention. Object-based attention refers to when attentional deployment is guided by object structure rather than only by spatial location, as in the case of spatial attention (e.g., Egly, Driver, & Rafal, 1994). Perceptual grouping cues, via their contribution to the perception of objecthood, play an important role in defining the entities available for (object-based) attention selection. Specifically, indices of object-based attention are influenced by gestalt grouping principles like those manipulated in the present study (Kramer & Jacobson, 1991; Marino & Scholl, 2005; Matsukura & Vecera, 2006). Object-based attention is frequently indexed as facilitated allocation of attention to regions perceived as being part of the same object as a target (Kramer & Jacobson, 1991) or as a pretarget cue (Egly et al., 1994). This has been interpreted as reflecting an automatic spreading (Richard, Lee, & Vecera, 2008; Vecera & Farah, 1994) or prioritization (Shomstein & Yantis, 2002) of attention to regions within an attended object. Furthermore, depending on the task, this spread of attention through an object can facilitate or impair performance, depending on whether the information within regions of the same object is congruent or incongruent with the correct response to task-relevant information (e.g., Kramer & Jacobson, 1991).
One possibility is that holistic face perception, as indexed via the composite-face effect, stems in part from an automatic prioritization of both the task-relevant and -irrelevant face parts, due to the allocation of attention to the face as a unit—that is, object-based attention. Physically misaligning the face parts, thus disrupting the cohesiveness of the face as a unit of selection, might allow attention to more effectively target the task-relevant part. Consistent with this possibility, other manipulations known to disrupt objecthood and thus object-based attention, such as spatially separating different parts of an object or presenting them on different depth planes, has also been shown to disrupt holistic face perception (Taubert & Alais, 2009).
Furthermore, the tasks used to demonstrate object-based attention and those, such as the composite-face effect, used to index holistic face perception bear some striking similarities (Kramer & Jacobson, 1991; Richard et al., 2008). For example, Kramer and Jacobson had participants make a judgment about a target while ignoring adjacent distractors. The distractors, which served as targets on other trials, could be compatible or incompatible with respect to the correct response of the target. The key manipulation was that the distractors and the target could be embedded in the same object or in different objects. Kramer and Jacobson found a robust decrease in the compatibility effect when the target and distractors were embedded in different objects. In addition, this effect was modulated by perceptual cues encouraging (common color) or discouraging (different colors) the grouping of the target and flanking stimuli. In a similar vein, in the composite-face task used to index holistic processing, participants are asked to make a judgment about a target (top or bottom face part), while ignoring an adjacent distractor (the other, task-irrelevant part of the face). The task-irrelevant distractor parts also serve as targets on other trials and could be compatible or incompatible with respect to the correct response for the task-relevant target part. Furthermore, in a key manipulation in the composite task, the distractor is either aligned, thus binding with the target to make a coherent object, or misaligned, thus disrupting object-based attention. In this way, perceptual cues discouraging the grouping of face parts may serve to similarly disrupt the employment of object-based attention, facilitating participants’ ability to selectively attend to the task-relevant face part, resulting in a decrease in indices of holistic processing.
Given the orientation-specific nature of the impact of perceptual grouping cues on holistic face processing, it is notable that experience with a particular stimulus orientation can strengthen the perception of objecthood (Vecera & Farah, 1997). Experience has been identified as a key factor in image segmentation—that is, in determining whether and how perceptual features will be grouped. For example, fragments from upright letters are more strongly grouped than those from inverted letters or nonfamiliar shapes (Vecera & Farah, 1997). Specifically, participants are faster at determining whether two probed locations are on the same or different overlapping shapes when the shapes are upright letters, as compared to inverted letters or nonletter shapes. In addition, participants are faster to make a judgment about two features on separate fragments behind an occluding shape when their experience is consistent with these two features belonging to the same object versus different objects (Zemel, Mozer, Behrmann, & Bavelier, 2002). Evidence that experience- and stimulus-based influences on perceptual grouping occur in approximately the same time frame further supports the potential of experience to have a pervasive influence on the perception of objecthood, and thus on object-based attention (Kimchi & Hadad, 2002). Future studies should further probe the potential role of object-based attention in supporting holistic processing of faces, especially given other recent work also questioning the assumptions underlying traditional accounts of holistic face perception (Gold, Mundy, & Tjan, 2012).
In conclusion, holistic face perception, as indexed via the composite-face effect, is disrupted by cues discouraging the perceptual grouping of face parts. Given the key role that perceptual grouping cues play in determining the units of object-based attention, these findings suggest that object-based attention may contribute to the difficulty that observers experience when trying to selectively attend to parts within faces, which is characteristic of holistic face perception in the composite-face task. Future studies will be needed to determine the nature of the contribution of perceptual grouping to markers of holistic processing more generally.