Attentional mechanisms underlying social and non-social attentional orienting

Visual attention, and in particular social attention, constitutes a multi-level system that prioritizes goal-relevant information while suppressing distractors via context-dependent weighting that amplifies relevant signals and attenuates irrelevant input. Within this framework, gaze occupies a central role: it not only automatically orients the observer’s attentional focus toward potentially informative sources but also conveys intention and social relevance, integrating perceptual, attentional, and social-cognitive processes. Among the mechanisms that coordinate this dynamic, two complementary processes are central for this work: the orienting of attention toward potentially informative sources and control processes that resolve competition between visual codes at decision and response levels. From this perspective, directional stimuli such as gaze and arrows offer an important comparative case. Although both can guide attention, they differ in their perceptual composition and communicative value, providing researchers with two complementary types of stimuli to examine the general and specific mechanisms of attention. To this end, two main experimental paradigms have been employed; the cueing and the spatial interference paradigms. The first enabled the initial attempts to determine whether attentional orienting induced by gaze differed from that induced by arrows, revealing quantitatively similar effects but qualitative differences. The second offered a more sensitive framework for exploring these divergences by placing directional information in conflict with the spatial position of the stimulus. In these tasks, stimuli are presented at positions that either match or conflict with their directional information, revealing how location and direction codes interact and setting the stage for this thesis’s central puzzle: the divergent congruency patterns for gaze versus arrows. Despite important advances in spatial interference research, a central uncertainty remains: why do gaze targets frequently produce a reversed congruency effect (RCE), with faster responses on incongruent trials (direction and position in conflict), whereas arrows typically yield the standard congruency effect (SCE), with faster responses on congruent trials (direction and position match)? This divergence has sustained a debate between social accounts that grant gaze a privileged status and domain-general accounts proposing that reversal arises from conflict dynamics shaped by perceptual constraints and control. This thesis addresses this debate by examining the boundary conditions under which reversal arises and how its distributional dynamics differ for gaze versus non-social targets. It combines large-scale reanalyses with two preregistered studies using arrows designed to mimic gaze-like constraints. Studies 1–2 applied conditional accuracy functions (CAFs) and delta functions to published gaze/arrow data to isolate early response dynamic and time-dependent interference. Studies 3–4 manipulate perceptual structure in arrow targets, first by mimicking gaze-relevant features and increasing background complexity, and then by applying visual masking, to introduce face-like surrounding complexity and test whether non-social stimuli can reproduce gaze-like reversal dynamics. Within the distributional analysis, arrows showed typical non-social patterns: among the CAF fastest responses, accuracy reductions concentrated on incongruent trials, reflecting an automatic response capture mechanisms (i.e., a high proportion of responses are likely selected by spatial location rather than direction), and delta functions were positive and slightly decreasing across quantiles. Gaze diverged on both fronts: CAFs revealed accuracy reductions for both congruency conditions, with congruent trials showing even lower accuracy at short latencies, and deltas were negative and became more negative with RT, indicating an emergent, dynamic reversal. These results suggest that gaze does not merely disrupt spatial interference but reconfigures the timing and balance between automatic activation and response selection, creating conditions under which reversal becomes likely. Altering arrows to mimic iris-sclera polarity and adding complex backgrounds reduced interference and attenuated early capture but did not yield an RCE, neither at the mean level analyses nor at the distributional level. By effectively masking directional information, however, arrows shifted from SCE to a reliable RCE with gaze-like distributional profiles. Thus, non-social stimuli can elicit reversal when specific perceptual constraints interact with directional information in the same functional way that gaze naturally does. Across these findings, three implications follow. First, the reversal does not require exclusively social mechanisms. The RCE appears to emerge from the interaction between automatic response processes, potentially modulated by the elements surrounding directional information, and control processes of the cognitive system, which adjust their onset and intensity according to both their own capacities and the characteristics of the stimulus. Second, gaze may naturally instantiate these conditions: its complex morphology and communicative value likely provide multiple signals that compete with each other and are rapidly registered, reconfiguring conflict dynamics when circumstances allow. Third, arrows masked with elements resembling their own structure can approximate the visual effects produced by faces, reproducing reversal patterns similar to those of gaze and suggesting that what gaze contributes intrinsically can be recreated through non-social stimuli. These findings reframe the RCE as a domain-general phenomenon that emerges from the interaction among stimulus properties, task demands, and the dynamics of the system that processes them. In this context social stimuli such as gaze are effective in generating the reversed dynamics not because they invoke a categorically distinct mechanism, but because their perceptual and communicative complexity, when embedded in a context that disrupts typical processing, promotes the emergence of non-neutral signals that interact with directional information and modulate control mechanisms. This reframing bridges theories of social attention with a broader framework grounded in conflict dynamics. It also provides a plausible explanation for the emergence of negative delta values, for why congruent responses can become disproportionately error-prone at the fastest latencies, and for how perceptual manipulations (e.g., masking) reshape the temporal course of conflict and its inhibition until reversal occurs. In sum, the account describes boundary conditions for the RCE and yields testable predictions. It links social-attention findings to conflict-dynamics accounts and proposes concrete tests, combining behavioral and neurophysiological measures with computational models, to predict when and how the RCE should appear. It also positions gaze stimuli on a continuum of complex objects that, depending on the setting, are processed by domain-general mechanisms.