Research Area A - Seeing Others


Little is known about how the brain processes visual information about the choices and actions of others, and how this processed information is used to guide interactions with others. We will use the newly developed Dyadic Interaction Platform to investigate the neural circuits responsible for integrating sensory information about visual stimuli and their behavioural relevance with information about what others know (and do) about the same stimuli. These studies are an important step towards understanding how social context modulates perceptual experience and decision-making.


Humans are highly social beings, thus stimuli that are specifically important for successful social dynamic interaction gain selected perception and beneficial processing. This has been demonstrated particularly for stimuli that convey increased social, emotional, and motivational relevance. The temporal course, possible hierarchies, and the nature (i.e., parallel, additive, interactive) of the processing of different relevance dimensions are yet largely unknown. This project aims at a functional and temporal segregation of effects of the three different relevancies by developing paradigms with orthogonal variations of relevance dimensions and means of event-related brain potentials.


A building block of social cognition is social monitoring, i.e., attending to conspecifics and their social interactions, leading to better-informed decision-making for subsequent interactions. This field project on wild macaques in Thailand will close gaps in our understanding of selective attention by tracking head-gaze towards different spontaneously unfolding interactions of other group members, assessing hormonal correlates of inter-individual variation in attention, and determining the resulting consequences for social network position, and ultimately for individual fitness.


To understand the actions and intentions of others, primates look at others’ faces. Facial motion allows the visual system to anticipate inputs based on predictive information inherent in movement sequences. We will determine whether face processing areas in the macaque monkey’s temporal cortex utilise motion-derived predictions to optimise neural coding of visual face information. To this end, we will manipulate the predictiveness of head motion sequences while we perform fMRI-guided electrophysiology and pharmacological inactivation. This will further our understanding of the neural basis of “seeing each other” during dynamic interactions in ecologically valid settings where agents are in motion.


This project investigates how different levels of momentary predictability in the actions of one individual influence the flow of information in the cortex of another, interacting individual. Both, predictability of actions at the behavioural level as well as the flow of information at the neural level will be quantified using rigorous information theoretic measures that allow for a direct correlation between them. The presence of a significant correlation indicates the presence of predictive processing in the observing individual. The sign of the correlation is indicative of specific processing strategies. We expect a conserved strategy across species and experimental settings.


In any interaction of dyads the recent as well as the distant past plays a role, shaping the next action and the general strategy. We will investigate how and which aspects of the past behaviour of the two agents predict their subsequent actions, and how to induce strategy changes. To quantify and disentangle these past dependencies, we will employ statistical models and information theory, and to induce strategy changes in a controlled manner, we will develop a generalised dyadic foraging game. Thereby, we can investigate precisely how cognitive processes in dyadic interactions, in particular the temporal horizon over which information is integrated, predict the chosen strategy of the agent.