Cognitive and neural mechanisms of social attention: studying qualitative differences between gaze and arrows

Abstract

The ability of humans to perceive and process others' gaze direction has been crucial for their survival. Gaze has, on the one hand, a great social significance that triggers the initiation of interaction with others, and on the other hand, a special status as an attentional signal, which makes us change the direction of our attention to the same location where the other is paying attention to (Birmingham & Kingstone, 2009). This interest in knowing where other people are directing their attention is what we know as Social Attention. Gaze processing requires the joint involvement of the neural mechanisms that underlie the so-called 'social brain', which includes temporo-occipital and prefrontal cortex regions. These regions participate in gaze processing together with parietal regions, reflecting the engagement of the attentional system to encode the spatial direction of the others’ gaze and the orienting of attention in that direction (Hadders-Algra, 2022). Gaze-triggered attention has been proposed to represent a unique attentional process reflecting the operation of a specialised cognitive and neural mechanism (S. Baron-Cohen, 1995). To assess this uniqueness, many studies have attempted to dissociate the attentional orienting triggered by gaze direction from that elicited by irrelevant and non-social stimuli, such as arrows, using variants of Posner's (1980) spatial orienting paradigm (Friesen & Kingstone, 1998). However, recent research shows that using such a paradigm, the attentional orientation generated by arrows and gaze is comparable, suggesting that both cues trigger the same attentional orientation as the product of a domain-general orienting mechanism (Chacón-Candia, Román-Caballero, et al., 2023). However, a different pattern of results emerges from other paradigms that seek to investigate the qualitative differences between the two cues by dissociating the attentional effects produced by the two stimuli. Recently, differences in the spatial congruency effects between arrows and gaze were observed using a spatial interference paradigm (Cañadas & Lupiáñez, 2012; Marotta et al., 2018a). These differences consist of standard congruency effects for arrow stimuli and reverse congruency effects (RCE) for gaze stimuli. The attentional mechanisms behind these differences in congruency effects have yet to be discovered. The present thesis aims to study whether the attentional mechanisms elicited by eye-gaze differ from those engaged by just symbolic non-social stimuli, such as arrows. Specifically, we sought to understand the cognitive and neural mechanisms that would underlie the differences observed in the spatial congruency effect between the two stimuli. Three studies were carried out to answer this objective, using different variants of the spatial interference paradigm, together with an extensive review of the field of study. In the first study, we sought to determine the nature of the attentional processes behind the standard congruency effects found with arrows and the RCE found with gaze stimuli. To this end, we examined whether the sequential congruency effects triggered by arrows and gaze could be generalised from one stimulus to another, using an intrablock task design, in which stimulus type was manipulated on a trial-by-trial basis. Results showed standard and reversed sequential congruency effects for the arrow and gaze stimuli, respectively, when presenting successive trials of the same stimulus. Furthermore, when presenting alternating stimuli, it was observed that the sequential congruency effects could be generalised across stimuli, regardless of the type of preceding stimulus. These results were replicated even when using other non-social stimuli, such as words indicating a direction and presenting whole faces, rather than eyes. These findings supported the co-existence of a shared spatial interference component between the arrow and gaze stimuli, which generalises sequential congruency effects between the two stimuli. The results also suggest the existence of an additional, possibly social, component involved only in the processing of gaze stimuli, which would produce the RCE. In the second study of this thesis, the eye contact hypothesis was tested as responsible for the gaze-specific mechanism that produces the reversal in congruency. For this purpose, we used an implicit task (where participants responded to the colour of the stimuli) and an explicit task (where participants responded to the direction of the stimuli). Whereas in the explicit task, opposite congruency effects were observed between the two stimuli, in the implicit task no congruency effects are observed, neither for arrows nor for gaze. This suggested that the eye contact hypothesis is not the most appropriate hypothesis to explain the RCE observed with gaze (as eye contact still occurred in the implicit task where the RCE was not observed). The same pattern of results was replicated when replacing the manual response with a verbal one, suggesting that response-related motor priming components would not be involved in congruency effects. In the third experiment, a functional magnetic resonance imaging (fMRI) design was used to explore the neural substrates involved in the attentional processes behind the standard congruency effects found with arrows and the RCE found with gaze stimuli. Shared neural mechanisms between arrow and gaze stimuli were observed in the functional connectivity between regions of the ventral attentional network and temporo-occipital regions. In addition, dissociable neural mechanisms were observed between both stimuli: 1) the functional connectivity between the right FEF and regions of the occipital cortex: this functional coupling was increased for congruent as compared to incongruent stimuli, and this result was larger for gaze than arrow stimuli. 2) The functional connectivity between the right FEF and regions of the left frontoparietal cortex was increased for incongruent as compared to congruent trials, and this result was only observed for arrows but not for gaze. These findings revealed shared neural networks between the two stimuli associated with spatial resolution and attentional orientation, as well as an additional set of regions that were differentially involved in the processing of each stimulus. Taken together, the results of the three studies indicate shared attentional and neural mechanisms between both stimuli, responsible for the standard spatial congruency effects that we propose contribute to the processing of both arrows and gaze. At the same time, a gaze-specific mechanism responsible for the RCE would coexist. This specific gaze mechanism is not related to the eye contact effect. Therefore, other hypotheses are proposed to explain the RCE that occurs with the gaze. The present doctoral thesis provides new data to help us to understand the cognitive and neural processes underlying the observed qualitative differences between social and non-social stimuli. We expect that these findings will contribute research on social attention.