Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2023 Mar 22.
Published in final edited form as: J Neurophysiol. 2021 Feb 17;125(4):1058–1067. doi: 10.1152/jn.00481.2020

fMRI response to automatic and purposeful familiar-face processing in perceptual and non-perceptual cortical regions

Silvia Ubaldi 1, Scott L Fairhall 1,✉
PMCID: PMC7611704  EMSID: EMS135241  PMID: 33596739

Abstract

Viewing the faces of familiar people selectively activates a distributed network of brain regions implicated in both the perceptual and non-perceptual processing of conspecifics. In this fMRI study, we investigate the influence of depth of famous-face processing on this network, comparing a passive incidental face-processing to a task that required the extraction of identity and biographic information. We observed that the precuneus, ventromedial prefrontal cortex (vmPFC), anterior temporal face patch (ATFP) and the amygdala exhibit a selective response even during incidental face processing. At the same time, face selectivity was enhanced in the lateral anterior temporal lobe (ATL) and the posterior superior temporal sulcus (pSTS) when identity and information extraction was required. Additionally, goal-directed identity and information extraction was associated with a recruitment of inferior frontal gyrus (IFG), while this region was deactivated during passive viewing. Collectively, these results show that components of the extended system reflect both the passive and active retrieval of person-related knowledge, and that the active access to such knowledge may potentially be mediate control circuits in the IFG.

Introduction

Perceiving and knowing about others is key for human interaction and is fundamental to our daily lives. Effective interpersonal interaction involves not only the perceptual processing of faces, but also non-perceptual cognitive processes related to emotional states, social processing, identity and semantic information.

Person-related perceptual and non-perceptual processes are served by a well-characterised brain network divided into core and extended systems (Fairhall and Ishai 2007; Gobbini and Haxby 2007; Haxby et al. 2000). The core system is closely linked to the visual perception of the invariant facial features that allow the perceptual processes that lead to the identification of individuals, as well as the perception of the changeable aspects of faces that allow recognition of facial expression. This core system includes, ventrally, the occipital face area (OFA) and fusiform face area (FFA) and, dorsally, the posterior superior temporal sulcus (pSTS) (Duchaine and Yovel 2015; Haxby et al. 2000). The non-perceptual extended system includes the anterior temporal cortex (ATL), the precuneus, ventromedial prefrontal cortex (vmPFC), the inferior frontal gyrus (IFG) and the amygdala.

Components of the extended system are believed to be responsible for a broad range of cognitive processes related to other people: identity representation (ATL, ATFP), semantic knowledge (ATL, IFG, precuneus), social knowledge and personality traits (precuneus, vmPFC, ATL, amygdala), emotional processing (amygdala, vmPFC) and naming (ATL, IFG), as well as executive control processes, such as working memory (IFG; Adolphs 2002, 2003; Calder and Young 2005; Fairhall et al. 2014; Gobbini and Haxby 2007; Murray et al. 2012; Olson et al. 2007, 2013; Wieser and Brosch 2012). The extended system is more active when viewing familiar individuals, which has been attributed to the spontaneous retrieval of non-perceptual information (Gobbini and Haxby 2007). However, experiments that have reported activation in the extended system by familiar faces generally employ tasks requiring some form of judgement to be made about the face stimulus (e.g. identity matching, superordinate categorisation, semantic access; Fairhall et al. 2014; Gobbini et al. 2004; Leibenluft et al. 2004), and it is known that the directed retrieval of different forms of knowledge about others - social, nominal, semantic, physical or episodic memories - differentially modulates activation within components of the extended system (Aglinskas and Fairhall 2020). On the other hand, when viewing a familiar face, we effortlessly and involuntary process the individual’s sex, age, identity, emotional state and a range of other characteristics. In this way, there are two modes in which we can access our knowledge about other people when we see their faces: the automatic information we reflexively access, as well as additional information that depends on the current goals. The degree to which recruitment of components of the non-perceptual extended system depends on current needs and goals remains uncertain.

This work investigates the degree to which recruitment of the extended system is contingent on the need to purposefully extract famous face identity in order to retrieve semantic information about the individual. We compare two tasks. In the incidental-processing task, participants performed a low-level perceptual judgment where they detected a salient contrast-change of the face stimulus. In the information-extraction task, participants were required to retrieve the nationality of the famous person, a task that necessitates both identification of the individual and the subsequent retrieval of semantic information.

Here we ask two questions, 1) which regions continue to show a face-selective response during incidental face processing and 2) which, if any, are modulated when semantic biographical information must be accessed. We predict that regions reflecting reflexive, automatic, person processing will continue to show a face-selective response in the incidental-processing task, and that regions involved in the goal-directed processing of faces will show enhanced face-selectivity when identity and information extraction are required.

Materials and Methods

Participants

Thirty-three healthy volunteers (mean age 23.7, 16 males) participated in the incidental-processing study. A separate sample of 32 healthy volunteers (mean age 23.9, 10 males) participated in the information-extraction task. All participants were right-handed native Italian speakers. Participants had no history of neurological disorders and had normal or corrected-to-normal vision. All participants gave informed consent to take part in the study and were reimbursed for their time. Procedures were approved by the Ethical Committee of the University of Trento.

Stimuli

Stimuli were 54 famous people and 45 famous monuments in the incidental-processing study and 49 famous people and 49 famous monuments in the information-extraction study. Stimuli were presented inside an oval shape superimposed on a phase-scrambled background. The faces were all in frontal view with a neutral expression, and all the stimuli were matched for luminance and dimension on the screen. For the control condition, we used 24 phase-scrambled images. Stimuli (600 × 800 pixels) were presented with Psychtoolbox (Brainard 1997; Pelli 1997) running on MATLAB (Mathworks).

Experimental Paradigm

For both studies the experiment was split into 4 fMRI runs. Each run was composed of 12 blocks of famous-face stimuli, 12 blocks of famous-place stimuli and 4 blocks of scrambled images in a pseudo-randomised order (Figure 1). Stimuli were presented at 12 different ISIs (one ISI per block), ranging from 100 to 1200 msec in 100 msec steps in the incidental-processing study and from 300 to 2500 msec in 200 msec steps in the information-extraction study. Different ISIs were used to reflect the different time required to perform the two tasks. Critically, all reported analyses comparing experiments use the five ISIs that were present in both (300, 500, 700, 900 and 1100 msec). Face and place stimuli were presented for half the ISI followed by the phase-scrambled background alone for the other half. Scrambled images appeared every seventh randomised block and had a fixed ISI of 1 second. Each block lasted 12 seconds, and blocks were separated by a 2-second fixation cross.

Figure 1. Stimuli and experimental paradigms.

Figure 1

Open in a new tab

Stimuli were presented in 12-second blocks of faces, places or scrambled images (baseline). In the incidental-processing task, participants were required to detect the presence of an occasional, salient, high-contrast face. In the information-extraction task, participants had to indicate if a non-Italian famous face or a building constructed prior to 500 C.E. was presented.

In the incidental-processing study participants had to press a button whenever they saw a high-contrast face or place (see figure 1 for an example). In the information-extraction study, participants were required to press a button whenever they saw a non-Italian individual (for face stimuli) or a monument built before 500 C.E. (for place stimuli). Targets occurred, on average, 1.5 times per block. After the scanning session, participants were asked which famous people they had recognised, and which they had seen for the first time. All the participants considered in the analysis had a recognition accuracy above 95%.

fMRI data acquisition

Functional and structural data were collected with a Bruker BioSpin MedSpec 4-T scanner (Bruker BioSpin GmbH, Rheinstetten, Germany) while participants lay in the scanner and viewed the visual stimuli through a mirror system. Data collection was conducted at the Center for Mind/Brain Sciences (CIMeC), University of Trento, using a USA Instruments 8-channel phased-array head coil. Functional images were acquired using echo planar (EPI) T2*-weighted scans. Acquisition parameters were: repetition time (TR) of 2 s, an echo time (TE) of 33 ms, a flip angle (FA) of 73°, a field of view (FoV) of 192 mm, and a matrix size of 64 × 64. Total functional acquisition consisted of 796 volumes, over 4 runs, each of 34 axial slices (which covered the whole brain) with a thickness of 3 mm and gap of 33% (1 mm). High-resolution (1 × 1 × 1 mm) T1-weighted MPRAGE sequences were also collected (sagittal slice orientation, centric phase encoding, image matrix = 256 × 224, field of view = 256 × 224 mm, 176 slices with 1-mm thickness, GRAPPA acquisition with acceleration factor = 2, duration = 5.36 min, repetition time = 2700, echo time = 4.18, TI = 1020 ms, 7° flip angle).

fMRI Analysis

For the information-extraction study, data from 28 participants were included in this analysis. Data from four subjects were rejected due to within-run head motion exceeding 2 mm. For the incidental-processing study, data from 28 participants were included in this analysis. Data from three subjects were rejected due to within-run head motion exceeding 2 mm. One subject was removed due to low target-detection scores (2.5 SD below the mean). Data from one participant were deleted due to medical considerations.

For both studies, data were analysed and preprocessed with SPM12 (http://www.fil.ion.ucl.ac.uk/spm/). The first 4 volumes of each run were discarded. All images were corrected for head movement. Subject-specific parameter estimates (β weights) for each of the 24 conditions (faces and places for each ISI) were derived through a general linear model (GLM), and a more lenient implicit mask for inclusion in the GLM (.1 instead of the SPM default of .8) was coupled with an explicit grey-matter mask to maximise sensitivity in susceptibility-sensitive regions on our 4T scanner. The control condition with scrambled images formed the implicit baseline. The 6 head-motion parameters were included as additional regressors of no interest.

ROI selection

In both studies, ROIs were defined using an independent group of subjects (N=35) who performed a one-back identity-matching task on a rapid presentation of faces and places (ISI: 100-1200 msec). Participants responded when they saw the mirrored repetition of the same image (average one per block, ± 1). Data from an independent set of participants was used to define ROIs. This avoids circularity or any form of bias in the ROIs selection across tasks. ROIs were defined by a sphere of 6-mm radius around the group coordinates masked by contrast face>places (face-selective ROIs) or places>faces (place-selection ROIs) for the independent dataset, thresholded at p<.001. The centre of ROIs is indicated in figure 3. To directly compare the differences between different levels of information retrieval, data were averaged across the five ISIs which matched in the two tasks (300, 500, 700, 900, and 1100 msec).

Figure 3. Signal plots of the response to face and place stimuli as a function of task in face-selective ROIs. Bars indicate the mean ROI response and error bars indicate one standard error.

Figure 3

Open in a new tab

Results

Behavioural

Accuracy was average over the matched ISIs used in the main fMRI analyses and a mixed-measures ANOVA performed (factors: task (between), category (within)). As expected, the main effect of task (F(1,54)=63.7, p<.001) reflected higher accuracies in the incidental-processing task (93.4%; 93.1% faces, 93.7% places) than the information-extraction task (70.6%; 73.1% faces, 68.0% places). There was no main effect of stimulus category (F(1,54)=1.1, p=.31). Importantly for the interpretation of the reported fMRI results, there was no task-by-category interaction (F(1,54)=1.7 p=.197), indicating that reported fMRI results are not influenced by differences in accuracy.

As expected, reaction times (RTs) were faster in the incidental-processing task compared to the information-extraction task (F(1,54)=83.6, p<.001; 594 msec, incidental-processing, 769 msec, information-extraction). While RT did not differ as a function of category (F(1,54)=0.1), a task by category interaction was present (F(1,54)=10.3, p=.002). This effect was driven by faster responses for faces than places in the incidental-processing condition (faces: 571 msec, places: 617 msec, t(27)=5.6, p<.0001), while descriptively faster response for places than faces in the information-extraction task did not reach significance (faces: 787 msec, places: 750 msec. t(27)=1.5, p=.14).

Category-Selectivity: Whole-Brain Analysis

To assess the category-selective response in both experiments, in a preliminary analysis we contrasted faces with places, collapsing across all 12 ISIs (figure 3). In both experiments, face-selective responses (faces>places) were evident in core-system regions: bilateral OFA, FFA and pSTS, as well as the precuneus. Additional face-selective responses were seen in the information-extraction task in the vmPFC, right lateral ATL and bilateral amygdalae (figure 3B). The contrast places-versus-faces showed bilateral parahippocampal place areas (PPA), bilateral transverse occipital sulcus (TOS) and bilateral retrosplenial complex (RSC) in both experiments.

No evidence for differential tuning patterns between incidental-processing and information-extraction tasks.

Stimuli were presented across a range of ISIs (see methods) to assess differential patterns in temporal tuning (the relationship between presentation rate and fMR amplitude) between incidental-processing and information-extraction tasks. We did not observe systematic dissociations in these patterns between the two tasks, and the absence of modulation by ISI in some regions left the validity of regional comparisons uncertain. These null results are presented in Supplementary Data (https://figshare.com/s/ac55227cbca0e24140b0).

Region of Interest Comparison Between Tasks – Face-Selective Regions

Our design allowed us to directly compare regional responses between the two tasks by considering the averaged response to the shared ISIs (300, 500, 700, 900, 1100). To assess the effects of task within brain regions specialised for face processing, we employ alterations in face-selectivity (faces versus places) as the dependent measure to better control for global differences across tasks. As there were no hemisphere-by-task-by-category interactions (all f-values < 1.8, all p-values > .40), we collapsed across hemisphere. Independently defined (see methods) ROI data showing the averaged fMR response across ISI for people and place stimuli in both tasks are presented in figure 3. A face-selective response was seen in all regions in the information-extraction task (IFG, p=.042; all other regions, p<.001). In the incidental-processing task, face selectivity was evident in OFA, FFA, the precuneus and the amygdala (p<.001), vmPFC (p=.002), ATFP (p=.004) and pSTS (p=.047) but not in the IFG (p=.51) or lateral ATL (p=.12).

The effect of task on the face-selective response was assessed through a mixed-measures ANOVA. A region-by-category-by-task ANOVA indicated that the effect of task on the category selective response differed across regions (F(8,432)=2.21, p=<.021), which were assessed through task-by-category interactions within individual regions. Task led to a selective increase in regional response in the lateral ATL (F(1,54)=4.87, p<.032), which was driven by the strong selective response in the information-extraction task (t(27)=3.98, p=.0004) compared to the non-significant effect in the incidental-processing task (t(27)=1.57, p=.12). Although a significant hemisphere-by-category effect was not present, closer inspection suggests this effect was driven by the absence of a selective response in the left lateral ATL in the incidental-processing task. In this unplanned analysis, we observed that, while a face-selective response was absent in the left lateral ATL (t(27)=0.07, p<.94), a response was evident in the right lateral ATL (t(27)=2.62, p=.014). By contrast, in the information-extraction task, selective responses were evident both in left (t(27)=2.99, p=.006) and right (t(27)=3.82, p=.0007) lateral ATL. A task-by-category effect was also present in pSTS (F(1,54)=5.61, p<.021) and was driven by a stronger selective response in the information-extraction task (t(27)=4.85, p=<.00005) compared to a weak but still significant selective response in the incidental-processing task (t(27)=2.08, p=.047). These region-of-interest results are broadly consistent with the descriptive differences seen in the whole-brain analyses comparing face-selective responses within each task (figure 2). Specifically, pSTS activation can be seen to be more prominent in the information-extraction task and presence of the right ATL in information-extraction but not incidental-processing task.

Figure 2.

Figure 2

Open in a new tab

Whole-Brain analysis of face- (red) and place-selective (blue) responses in A. incidental-processing and B. information-extraction experiments. Face-selective responses are evident in core regions as well as the precuneus in both experiments and additionally, the lateral ATL, vmPFC and amygdala in the information-extraction experiment. Classic place-selective regions, PPA, TOS and RSC, are present in both experiments. See table 1 for supporting statistical and location information. Maps are thresholded at p<.001, cluster extent 25 voxels. Arrows shown in B indicate the location of the independent ROIs used to formally interrogate results in both experiments.

An unanticipated but prominent feature evident in figure 3 is the negative response in the IFG during the incidental-processing task. Here, the response to both face and place stimulus was below baseline and strongly different from the response during the information-extraction task. This difference was statistically confirmed via an unplanned main effect of task (F(1,27)=20.5, p=.0001).

Significant task-by-category effects were evident in pSTS and latATL but not in the precuneus ATFP, vmPFC or the amygdalae. However, the non-significance of effects in these latter regions does not indicate the absence of such effects. To assess the relative support of the null hypothesis, Bayes factors were calculated for the effect of task on face-selectivity (Rouder et al. 2009). While evidence favoured the null in all regions - precuneus (2.0), ATFP (2.2), vmPFC (1.4), amygdalae (1.2) - support at this level can only be considered anecdotal. For this reason, caution should be taken in the interpretation of these null effects, and it is possible that future studies with higher statistical power would detect influences of information extraction in additional regions of the extended system.

There was a significant interaction between category and task in reaction times, which may influence the response pattern observed in the pSTS, latATL and IFG. To assess whether increases in reaction time were influencing the responses to faces in these regions, the correlation between RT and the response to faces was calculated across participants. In the incidental-processing task, such a relationship was not seen, with non-significant trends towards decreased activations with longer reaction times (pSTS: r=-0.11, p=.59; latATL: r=-0.03, p=.88; IFG: r=-0.02, p=.91). In the information-extraction task, pSTS did show a numeric trend towards an increase in responsivity with RT, but this was weak and did not approach significance (r=0.003; p=.99). The opposite, non-significant, trend was seen in latATL and IFG (latATL: r=-0.06, p=.75; IFG: r=-0.28, p=.15). Collectively, these results suggest that reaction time differences do not influence responsivity in these regions.

Region-of-Interest Task Comparison – Place Selective Regions

To assess whether incidental-processing and information-extraction tasks generalise to other object categories, we performed the complementary ROI analysis in place-selective regions. A robust place-selective effect was observed across all independently defined ROIs in both tasks (p<.001; see figure 4). A mixed measures ANOVA (task x category) revealed significant differences between tasks in the selective response in PPA (F(1,54)=5.14, p=.027), driven by a larger selective response in the information-extraction task (t(27)=13.5,p<.001) compared to the incidental-processing task (t(27)=12.3,p<.001). A significant interaction was also evident in the RSC (F(1,54))=15.2, p=.0003), which was likewise driven by a larger selective response in the information-extraction task (t(27)=8.89, p<.001) compared to the incidental-processing task (t(27)=5.0, p<.001). In contrast, no effect of task was seen on the category-selective response in TOS (F(1,54)=1.67, p=.20).

Figure 4. Signal plots of the response to face and place stimuli as a function of task in place-selective ROIs. Bars indicate the mean ROI response and error bars indicate one standard error.

Figure 4

Open in a new tab

Whole-Brain Task Comparison

We complimented the primary ROI analyses with a whole-brain analysis on the effect of task on the category-selective response (figure 5). The sole region to show a significant task-by-category interaction that was not evident in the ROI analysis was a section of the right lateral fusiform gyrus. This region exhibited increased category selectivity during the incidental-processing task, which is consistent with a trend evident in the ROI analysis (figure 3). The cluster was located 6mm posteriorly and laterally to the centre of the independent right FFA ROI, with which it shared no overlapping voxels.

Figure 5. Whole-brain analysis of the effect of task on the category-selective response (task-by-category interaction).

Figure 5

Open in a new tab

Effects were driven by relative increases in place selectivity in bilateral retrosplenial cortex and the right parahippocampal gyrus in the information-extraction task. Conversely, voxels in the right lateral fusiform gyrus show increased face selectivity in the incidental-processing task. Effects sizes and error bars (+- 1 SEM) are only descriptive in nature due to the high circularity in ROI definition and interrogation and should only be used to inform the nature of the effect driving the interaction.

Other regions identified in this analysis correspond to bilateral place-selective RSC and right PPA evident in the ROI analysis (figure 4).

Discussion

In this work, we sought to determine the influence of the depth of face processing in core and extended elements of the distributed cortical network for perceiving and knowing about others. Using famous-face stimuli, we compared superficial, automatic face processing during an incidental-processing task to face processing that required the extraction of identity and semantic information in a information-extraction task. In our ROI analysis, we observed face-selective responses across regions of the core system in both tasks. Within the extended system, the precuneus, vmPFC, ATFP and the amygdala exhibited a profile consistent with the automatic processing of conspecifics, with face-selective responses being present in both tasks. In contrast, lateral ATL and IFG exhibited a face-selective response only in the information-extraction task. Moreover, we observed that increased depth of processing led to increases in selectivity in the pSTS and lateral ATL, consistent with a role in the active extraction of person-related information.

Persistent recruitment of the precuneus, vmPFC, ATFP and amygdala in incidental face processing.

Our independent ROI analysis revealed the presence of face-selective responses in the precuneus, vmPFC, ATFP and amygdala during the incidental-processing of faces. The recruitment of these regions during a low-level contrast-detection task that did not necessitate the processing of the faces can provide insight into the automaticity of the involvement of these regions in face and person perception.

The precuneus and vmPFC are part of the internalised default mode system and are removed from sensory and perceptual circuity (Cavanna and Trimble 2006; Fox et al. 2005). These regions show highly similar cognitive response profiles across access to different forms of person knowledge (Aglinskas and Fairhall 2020) and are implicated in high-level cognitive functions such as the retrieval of episodic, biographical and autobiographical memories and aspects of social cognition, such as personality traits and emotional processing (Burgess et al. 2001; Fairhall et al. 2014; Fairhall and Caramazza 2013; Frith and Frith 2007; Simmons and Martin 2009). This present work demonstrates the automaticity with which high-level, non-perceptual cognitive processes are recruited when viewing familiar members of our species and provides a putative source for the diverse social and semantic information that allows our effective interaction with others.

Unlike the precuneus and vmPFC, the ATFP is structurally and functionally connected to ventral perceptual regions including OFA and FFA (Moeller et al. 2008; O’Neil et al. 2014). This region has been implicated in the representation of identity information about faces and contains viewpoint-independent representation of individuals (Anzellotti et al., 2014; Guntupalli et al., 2017). The persistence of a response in this region during incidental face perception is consistent with the automatic extraction of face identity in the absence of explicit task demand. The amygdala is similarly connected to regions of the core system (Catani et al. 2003; Fairhall and Ishai 2007). Its involvement in face processing is known to be modulated by the emotional expression of the observed face (Breiter et al. 1996; Morris et al. 1996), even when face stimuli are presented below the threshold of perceptual awareness (Whalen et al. 1998). This response profile predicts the involvement of the amygdala during incidental face processing.

Information extraction enhances selectivity in the pSTS and lateral ATL

The region-of-interest analysis revealed that the requirement to extract identity and biographical information from famous-face stimuli increased selectivity in the pSTS and latATL. The pSTS is classically designated an element of the perceptual ‘core’ system, an attribution which would suggest that increased selectivity reflects an increasing perceptual demand associated with the identification of faces. It is true that, while the incidental-processing task did not necessitate processing of shape information or facial features, face identification (and the processing of facial features) was a necessary step in the information-extraction task. However, the pSTS is not strongly implicated in identity processing. The role of the pSTS in face perception is typically attributed to the changeable aspects of faces, such as eye-gaze or emotional expression (Haxby et al. 2000; Haxby and Gobbini 2011) or the dynamic processing of faces (Duchaine and Yovel 2015), factors that did not vary across these tasks.

An alternate explanation for the increased role of the pSTS in the information-extraction task is the region’s role in non-perceptual processes. The pSTS and neighbouring elements of the temporal parietal junction (TPJ) are implicated in social cognition, such as inferring the mental states of others (Gobbini and Haxby 2007; Saxe and Kanwisher 2003), name knowledge (Gesierich et al. 2012) and the retrieval of person-related semantic information (Fairhall and Caramazza 2013). Moreover, pSTS is adjacent to the angular gyrus, a region that is implicated in general semantic processing (Binder et al. 2009). It is possible that the role of pSTS in information extraction results from computational properties shared with these neighbouring regions.

The question of whether increased selectivity in the pSTS reflects perceptual or non-perceptual processes cannot be definitively answered by the present data. However, the dissociation between the pSTS and other elements of the perceptual core system, the paucity of existing evidence to suggest the pSTS plays a role in the identity processing of faces, and the well-documented role of this region in the representation of social and semantic knowledge about other people, support a non-perceptual role in information extraction. Lateral ATL exhibited a face-selective response during information extraction but not incidental face processing. This differential response indicates the involvement of this region in purposeful, task-relevant processing of conspecifics. The ATL has long been implicated in the representation of knowledge about unique entities (e.g. specific people or places) (Damasio et al. 2004; Grabowski et al. 2001; Ross and Olson 2012; Tempini et al. 1998) and exhibits response selectivity during the active retrieval of knowledge about person-related concepts (Fairhall and Caramazza 2013). Moreover, this region is believed to play a critical role in general semantic representation and was an early candidate for a semantic hub that links together semantic representations distributed across the cortex (Patterson et al. 2007; Tyler and Moss 2001). The relative role of the lateral ATL in person-related knowledge or general semantic knowledge, particularly relating to unique entities, is a topic of active debate (Olson et al. 2013; Simmons and Martin 2009). The interaction between task and stimulus category in the present study supports the importance of social knowledge in the function of the lateral ATL.

The results in pSTS and latATL indicate that selectivity for faces increases in these face-specialised regions. It is noteworthy that this is influenced by an attenuation in the response for the non-preferred category. This suggests a complex pattern that may result either from the relative suppression of the region when information is being extracted from place stimuli, a switch in the role of the region from a more generalised function in the incidental-processing task to a more specialised role in the information extraction task, other uncontrolled differences between these two tasks (such as cognitive load), or a combination of these factors.

A complementary whole-brain analysis indicates that task-induced modulations in category selectivity were not strongly present outside the investigated ROIs. The exception was an aspect of the right lateral fusiform gyrus that showed enhanced category selectivity during the incidental-processing task. This runs counter to the general trend of increased category selectivity during information extraction and suggests this region plays a particular role in the processing of low-level visual aspects of faces and argues against a general disengagement during the incidental-processing task.

Enhancements in selectivity observed in the pSTS and latATL were mirrored in the place-selective network, suggesting that the information-extraction related enhancement of category selectivity extends to object-domains other than people. Here we saw in both the ROI and whole-brain analyses, that PPA and RSC, but not the TOS, showed an increase in place selectivity during information extraction when compared to incidental place processing. This is consistent with the role of these regions not only in the perception of places (Epstein and Kanwisher 1998; Dilks et al. 2011) but also in higher-level spatial representations, such as the location of a building with its broader geographical context (Epstein et al. 2007) and the retrieval of geographic information about non-place objects, such as the origin of well-known food dishes (Fairhall 2020).

IFG activation is contingent on information extraction.

The IFG has long been identified as a region of the extended system, although the nature of its role remains uncertain. The role of the IFG in face processing has been attributed both to the top-down control of perception (Adolphs 2003; Calder and Young 2005; Renzi et al. 2013), as well as access to semantic and nominal knowledge (Adolphs 2002; Aglinskas and Fairhall 2020; Haxby and Gobbini 2011). The IFG represents the only region of the semantic system that is part of the executive control circuitry (Binder et al, 2009). Its involvement in selecting between multiple competing semantic responses and in making infrequent semantic associations, has led to its function being attributed to semantic control and the guided retrieval of goal-relevant semantic information (Martin and Chao 2001; Wagner et al. 2001; Thompson-Schill 2003).

The attenuation of the response in the incidental-processing condition and its recruitment during information extraction (both for faces and places), may provide insight into its role. This pattern is consistent with its role in the guided identification and extraction of person-related information. The presence or absence of IFG activation does not appear to influence the face-selective responses in OFA and FFA, which argues against a direct role in top-down modulation of ventral face processing regions. Rather, the increased response in IFG accompanies the increases in face selectivity in the lateral ATl and pSTS, which suggests that executive control circuitry is required for access to information present in these regions. Conversely, these results suggest that access to information represented in other regions, such as the precuneus and vmPFC, is not contingent on an IFG response and that access to the knowledge in these regions can proceed automatically without the need for executive control mechanisms.

In this work, we investigated the role of task demand on activation within the core and extended systems for face perception. We observed that elements of the extended system, the precuneus, vmPFC, amygdala and ATFP, continue to show a selective response during the superficial, surface-based processing of faces. Notably, this occurred in the absence of a response in IFG, a region implicated in active semantic control. The need to extract identity and biographical information produced a selective response in the IFG and was accompanied by enhanced selectivity in the ATL and pSTS, providing a potential modulatory neural mechanism for active access to stored knowledge. Collectively, these results provide insight into the diverse factors that drive recruitment and representation in the extended system and the automatic and purposeful neural systems that allow effective interaction with conspecifics.

Supplementary Material

Supplementary data

Table 1. Spherical centre of independent ROIs.

Hemisphere
Region Left Right
OFA -39 -82 -10 39 -79 -10
FFA -42 -49 -22 42 -55 -19
pSTS -51 -61 14 48 -61 11
IFG -39 20 20 42 20 20
latATL 51 -13 -13 -51 -13 -13
ATFP 33 -7 -37 -33 -10 -37
Amygdala 21 -4 -16 -21 -7 -16
Medial  
vmPFC 3 47 -22
Precuenus 3 -55 29
Open in a new tab

Table 2. Significance, extent of face- and person-selective clusters in incidental-processing and information-extraction tasks and location and t-value of within-cluster peaks.

cluster peak
Face-Selective Response   p(FWE-corr) voxels t x,y,z {mm}
Incidental Processing Right FFA <.001 504 17.0 42 -52 -19
OFA 11.8 39 -82 -10
pSTS <.001 209 6.0 51 -43 11
Left FFA <.001 408 13.0 -39 -46 -19
OFA 7.8 -48 -79 -4
pSTS 0.308 26 3.6 -51 -61 26
Medial Precuneus <.001 134 5.9 0 -55 29
Information Extraction Right FFA <.001 258 8.1 42 -49 -19
OFA 8.0 42 -79 -10
pSTS <.001 726 6.5 48 -43 14
latATL 0.080 49 4.8 54 -13 -13
Amygdala 0.222 33 4.4 24 -4 -16
Left FFA 0.023 70 7.6 -39 -43 -22
OFA <.001 314 6.1 -39 -79 -10
pSTS 5.0 -48 -61 17
Amygdala 0.086 48 4.7 -24 -4 -16
Medial Precuneus <.001 220 7.6 3 -55 32
vmPFC 0.053 56 4.7 3 53 -16
dmPFC 0.372 25 4.3 6 59 26
midCingulate 0.288 29 4.1 6 -19 44
Place-Selective Response
Incidental Processing Right PPA <.001 5089 28.1 27 -46 -7
TOS 19.4 33 -82 23
RSC 11.4 21 -55 20
Left PPA 26.4 -27 -46 -7
TOS 19.7 -33 -82 23
RSC 8.6 - 18 -52 8
Information Extraction Right PPA <.001 5342 22.2 30 -49 -7
TOS 15.1 -36 -82 14
RSC 14.1 18 -55 17
Left PPA 23.1 -27 -46 -7
TOS 15.8 33 -79 14
RSC 11.2 -18 -58 14
dlPFC 0.142 40 3.8 -48 23 38
Open in a new tab

Table 3. Significance and extent of whole-brain task-by-category interaction clusters and location and F-value of within-cluster peaks.

cluster peak
Task x Category Interaction p(FWE-corr) voxels F x,y,z {mm}
Left RSC 0.006 88 5.9 -9 -58 11
Right RSC 0.002 70 5.1 15 -55 14
Parahipp. 0.026 51 4.5 33 -43 -13
Lat. FG 3.8 36 -49 -22
Open in a new tab

Acknowledgments

The project was funded by the European Research Council (ERC) Starting Grant, ‘CRASK – Cortical Representation of Abstract Semantic Knowledge’, awarded to Scott Fairhall under the European Union’s Horizon 2020 research and innovation program (grant agreement no. 640594).

References

  1. Adolphs R. Neural systems for recognizing emotion. Curr Opin Neurobiol. 2002 doi: 10.1016/s0959-4388(02)00301-x. [DOI] [PubMed] [Google Scholar]
  2. Adolphs R. Cognitive neuroscience: Cognitive neuroscience of human social behaviour. Nat Rev Neurosci. 2003 doi: 10.1038/nrn1056. [DOI] [PubMed] [Google Scholar]
  3. Aglinskas A, Fairhall SL. Regional Specialization and Coordination Within the Network for Perceiving and Knowing About Others. Cereb Cortex. 2020 doi: 10.1093/cercor/bhz130. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Anzellotti S, Fairhall SL, Caramazza A. Decoding representations of face identity that are tolerant to rotation. Cereb Cortex. 2014;24:1988–1995. doi: 10.1093/cercor/bht046. [DOI] [PubMed] [Google Scholar]
  5. Binder JR, Desai RH, Graves WW, Conant LL. Where Is the Semantic System? A Critical Review and Meta-Analysis of 120 Functional Neuroimaging Studies. Cereb Cortex. 2009;19:2767–2796. doi: 10.1093/cercor/bhp055. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brainard DH. The Psychophysics Toolbox. Spat Vis. 1997;10:433–436. [PubMed] [Google Scholar]
  7. Breiter HC, Etcoff NL, Whalen PJ, Kennedy WA, Rauch SL, Buckner RL, Strauss MM, Hyman SE, Rosen BR. Response and habituation of the human amygdala during visual processing of facial expression. Neuron. 1996 doi: 10.1016/S0896-6273(00)80219-6. [DOI] [PubMed] [Google Scholar]
  8. Burgess PW, Quayle A, Frith CD. Brain regions involved in prospective memory as determined by positron emission tomography. Neuropsychologia. 2001;39:545–555. doi: 10.1016/s0028-3932(00)00149-4. [DOI] [PubMed] [Google Scholar]
  9. Calder AJ, Young AW. Understanding the recognition of facial identity and facial expression. Nat Rev Neurosci. 2005 doi: 10.1038/nrn1724. [DOI] [PubMed] [Google Scholar]
  10. Catani M, Jones DK, Donato R, Ffytche DH. Occipito-temporal connections in the human brain. Brain. 2003 doi: 10.1093/brain/awg203. [DOI] [PubMed] [Google Scholar]
  11. Cavanna AE, Trimble MR. The precuneus: A review of its functional anatomy and behavioural correlates. Brain. 2006 doi: 10.1093/brain/awl004. [DOI] [PubMed] [Google Scholar]
  12. Damasio H, Tranel D, Grabowski T, Adolphs R, Damasio A. Neural systems behind word and concept retrieval. Cognition. 2004;92:179–229. doi: 10.1016/j.cognition.2002.07.001. [DOI] [PubMed] [Google Scholar]
  13. Dilks DD, Julian JB, Kubilius J, Spelke ES, Kanwisher N. Mirror-Image Sensitivity and Invariance in Object and Scene Processing Pathways. J Neurosci. 2011;31:11305–11312. doi: 10.1523/JNEUROSCI.1935-11.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Duchaine B, Yovel G. A Revised Neural Framework for Face Processing. Annu Rev Vis Sci. 2015;1:393–416. doi: 10.1146/annurev-vision-082114-035518. [DOI] [PubMed] [Google Scholar]
  15. Epstein R, Kanwisher N. A cortical representation the local visual environment. Nature. 1998 doi: 10.1038/33402. [DOI] [PubMed] [Google Scholar]
  16. Epstein RA, Parker WE, Feiler AM. Where Am I Now? Distinct Roles for Parahippocampal and Retrosplenial Cortices in Place Recognition. J Neurosci. 2007;27:6141–6149. doi: 10.1523/JNEUROSCI.0799-07.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fairhall SL. Cross recruitment of domain-selective cortical representations enables flexible semantic knowledge. J Neurosci. 2020 doi: 10.1523/JNEUROSCI.2224-19.2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Fairhall SL, Anzellotti S, Ubaldi S, Caramazza A. Person- and place-selective neural substrates for entity-specific semantic access. Cereb Cortex. 2014;24:1687–1696. doi: 10.1093/cercor/bht039. [DOI] [PubMed] [Google Scholar]
  19. Fairhall SL, Caramazza A. Category-selective neural substrates for person- and place-related concepts. Cortex. 2013a;49:2748–2757. doi: 10.1016/j.cortex.2013.05.010. [DOI] [PubMed] [Google Scholar]
  20. Fairhall SL, Caramazza A. Brain regions that represent amodal conceptual knowledge. J Neurosci. 2013b;33 doi: 10.1523/JNEUROSCI.0051-13.2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Fairhall SL, Ishai A. Effective connectivity within the distributed cortical network for face perception. Cereb Cortex. 2007;17:2400–2406. doi: 10.1093/cercor/bhl148. [DOI] [PubMed] [Google Scholar]
  22. Fox MD, Snyder AZ, Vincent JL, Corbetta M, Van Essen DC, Raichle ME. The human brain is intrinsically organized into dynamic, anticorrelated functional networks. Proc Natl Acad Sci U S A. 2005 doi: 10.1073/pnas.0504136102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Frith CD, Frith U. Social Cognition in Humans. Curr Biol. 2007;17:R724–R732. doi: 10.1016/j.cub.2007.05.068. [DOI] [PubMed] [Google Scholar]
  24. Gesierich B, Jovicich J, Riello M, Adriani M, Monti A, Brentari V, Robinson SD, Wilson SM, Fairhall SL, Gorno-Tempini ML. Distinct neural substrates for semantic knowledge and naming in the temporoparietal network. Cereb Cortex. 2012a;22:2217–2226. doi: 10.1093/cercor/bhr286. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Gesierich B, Jovicich J, Riello M, Adriani M, Monti A, Brentari V, Robinson SD, Wilson SM, Fairhall SL, Gorno-Tempini ML. Distinct neural substrates for semantic knowledge and naming in the temporoparietal network. Cereb Cortex. 2012b;22 doi: 10.1093/cercor/bhr286. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Gobbini IM, Leibenluft E, Santiago N, Haxby JV. Social and emotional attachment in the neural representation of faces. Neuroimage. 2004;22:1628–1635. doi: 10.1016/j.neuroimage.2004.03.049. [DOI] [PubMed] [Google Scholar]
  27. Gobbini MI, Haxby JV. Neural systems for recognition of familiar faces. Neuropsychologia. 2007;45:32–41. doi: 10.1016/j.neuropsychologia.2006.04.015. [DOI] [PubMed] [Google Scholar]
  28. Grabowski TJ, Damasio H, Tranel D, Ponto LLB, Hichwa RD, Damasio AR. A Role for Left Temporal Pole in the Retrieval of Words for Unique Entities. 2001;212:199–212. doi: 10.1002/hbm.1033. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Guntupalli JS, Wheeler KG, Gobbini MI. Disentangling the Representation of Identity from Head View Along the Human Face Processing Pathway. Cereb Cortex. 2017;27:46–53. doi: 10.1093/cercor/bhw344. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Haxby JV, Gobbini MI. Distributed Neural Systems for Face Perception. Oxford Handb Face Percept. 2011:93–111. [Google Scholar]
  31. Haxby JV, Hoffman Ea, Gobbini MI. The distributed human neural system for face perception. Trends Cogn Sci. 2000;4:223–233. doi: 10.1016/s1364-6613(00)01482-0. [DOI] [PubMed] [Google Scholar]
  32. Leibenluft E, Gobbini MI, Harrison T, Haxby JV. Mothers’ neural activation in response to pictures of their children and other children. Biol Psychiatry. 2004 doi: 10.1016/j.biopsych.2004.05.017. [DOI] [PubMed] [Google Scholar]
  33. Moeller S, Freiwald WA, Tsao DY. Patches with links: A unified system for processing faces in the macaque temporal lobe. Science (80- ) 2008 doi: 10.1126/science.1157436. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Morris JS, Frith CD, Perrett DI, Rowland D, Young AW, Calder AJ, Dolan RJ. A differential neural response in the human amygdala to fearful and happy facial expressions. Nature. 1996 doi: 10.1038/383812a0. [DOI] [PubMed] [Google Scholar]
  35. Murray RJ, Schaer M, Debbané M. Degrees of separation: A quantitative neuroimaging meta-analysis investigating self-specificity and shared neural activation between self- and other-reflection. Neurosci Biobehav Rev. 2012 doi: 10.1016/j.neubiorev.2011.12.013. [DOI] [PubMed] [Google Scholar]
  36. O’Neil EB, Hutchison RM, McLean DA, Köhler S. Resting-state fMRI reveals functional connectivity between face-selective perirhinal cortex and the fusiform face area related to face inversion. Neuroimage. 2014 doi: 10.1016/j.neuroimage.2014.02.005. [DOI] [PubMed] [Google Scholar]
  37. Olson IR, McCoy D, Klobusicky E, Ross LA. Social cognition and the anterior temporal lobes: A review and theoretical framework. Soc Cogn Affect Neurosci. 2013 doi: 10.1093/scan/nss119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Olson IR, Plotzker A, Ezzyat Y. The Enigmatic temporal pole: A review of findings on social and emotional processing. Brain. 2007 doi: 10.1093/brain/awm052. [DOI] [PubMed] [Google Scholar]
  39. Patterson K, Nestor PJ, Rogers TT. Where do you know what you know? The representation of semantic knowledge in the human brain. Nat Rev Neurosci. 2007;8:976–987. doi: 10.1038/nrn2277. [DOI] [PubMed] [Google Scholar]
  40. Pelli DG. The VideoToolbox software for visual psychophysics: Transforming numbers into movies. Spat Vis. 1997;10:437–442. [PubMed] [Google Scholar]
  41. Renzi C, Schiavi S, Carbon CC, Vecchi T, Silvanto J, Cattaneo Z. Processing of featural and configural aspects of faces is lateralized in dorsolateral prefrontal cortex: A TMS study. Neuroimage. 2013;74:45–51. doi: 10.1016/j.neuroimage.2013.02.015. [DOI] [PubMed] [Google Scholar]
  42. Ross LA, Olson IR. What’s unique about unique entities? An fMRI investigation of the semantics of famous faces and landmarks. Cereb Cortex. 2012 doi: 10.1093/cercor/bhr274. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Saxe R, Kanwisher N. People thinking about thinking people: The role of the temporo-parietal junction in “theory of mind.”. Neuroimage. 2003 doi: 10.1016/S1053-8119(03)00230-1. [DOI] [PubMed] [Google Scholar]
  44. Simmons WK, Martin A. The anterior temporal lobes and the functional architecture of semantic memory. J Int Neuropsychol Soc. 2009;15:645–649. doi: 10.1017/S1355617709990348. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Tempini MLG, Price CJ, Josephs O, Vandenberghe R, Cappa SF, Kapur N, Frackowiak RSJ. The neural systems sustaining face and proper-name processing. Brain. 1998;121:2103–2118. doi: 10.1093/brain/121.11.2103. [DOI] [PubMed] [Google Scholar]
  46. Tyler LK, Moss HE. Towards a distributed account of conceptual knowledge. Trends Cogn Sci. 2001;5:244–252. doi: 10.1016/s1364-6613(00)01651-x. [DOI] [PubMed] [Google Scholar]
  47. Wang Y, Collins JA, Koski J, Nugiel T, Metoki A, Olson IR. Dynamic neural architecture for social knowledge retrieval. 2017:1–10. doi: 10.1073/pnas.1621234114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Whalen PJ, Rauch SL, Etcoff NL, McInerney SC, Lee Michael B, Jenike MA. Masked presentations of emotional facial expressions modulate amygdala activity without explicit knowledge. J Neurosci. 1998 doi: 10.1523/jneurosci.18-01-00411.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Wieser MJ, Brosch T. Faces in context: A review and systematization of contextual influences on affective face processing. Front Psychol. 2012 doi: 10.3389/fpsyg.2012.00471. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary data
EMS135241-supplement-Supplementary_data.pdf (756.3KB, pdf)

RESOURCES