Neurophenomenological Phenotyping of Interpersonal Interactional Dynamics in Autism


  • Jelena Rosic University of Vienna
  • Anne Monnier University of Montreal
  • Guillaume Dumas University of Montreal


Impairments in social interaction and interpersonal dynamics are common across psychiatric conditions and require a comprehensive understanding of various aspects of a person’s life,  besides neuropsychology. Autism spectrum disorder (ASD) describes a condition with deficits in reciprocal social interactions, including features of interpersonal synchrony, sensory behavior, empathy, and emotion, among others. However, broad diagnostic categories and the heterogeneity of biological parameters in psychiatry do not capture the dimensions of a person’s lived experience in constituting psychiatric conditions and possible interventions. Moreover, the phenomenological perspective views disorders of intersubjectivity, such as autism, as rooted in the pre-reflective, embodied relationship between self and others. This stands in contrast to "mentalizing" accounts and the view that autism is a dysfunction of the Theory of Mind (ToM) modules in the brain [1].

To systematically assess parameters of subjective and intersubjective features that underlie social dynamics in autism, the present study aims to establish computational neurophenomenological modeling of embodied and situated interpersonal dynamics during dyadic encounters. Such a methodology entails non-trivial considerations of applying computational models and tools to phenomenological accounts to provide formal models of lived experience, i.e., "computational neurophenomenology" [2]. This framework is developed while attending to the problem of methodological individualism in cognitive science and social cognition [3] by implementing a second-person approach within the phenomenological method and second-person neuroscience (micro-phenomenology and two-person EEG hyper-scanning). Modeling both in-depth descriptions of experiences (synchronic) and their dynamic connections over time (diachronic), this study is set to identify what key phenomenological components underpin interaction in autistic and typical dyads and establish their points of divergence. 

A data-driven approach, on the one hand, aims to classify phenomenological categories of interaction within and between the groups of autistic and typical participants. On the other hand, a model that captures dynamic relations between the categories is developed to establish states and their transitions over time as parameters describing the interaction dynamics of the dyads. Additionally, the resulting set of interactional parameters specifying a phenomenological phenotype may predict what experiential features correlate or diverge with the underlying neurophysiology.

Characterizing psychiatric conditions beyond the brain requires systematic assessment and integration of subjective dimensions and interpersonal dynamics. Considering the overarching integrative explanatory role of computational ideas in cognitive science and further in computational psychiatry, the methodological mapping described here as neurophenomenological phenotyping aims to assess both experiential and neurobiological domains with mutually constraining levels of analysis. 


[1] T. Fuchs, “Pathologies of intersubjectivity in autism and schizophrenia,” Journal of Consciousness Studies, vol. 22, no. 1-2, pp. 191-214, 2015

[2] M. J. Ramstead et al., “From generative models to generative passages: A computational approach to (neuro) phenomenology,” Review of Philosophy and Psychology, vol. 13, no. 4, pp. 829–857, Mar. 2022. doi:10.1007/s13164-021-00604-y 

[3] G. Dumas, J. A. Kelso, and J. Nadel, “Tackling the social cognition paradox through multi-scale approaches,” Frontiers in Psychology, vol. 5, Aug. 2014. doi:10.3389/fpsyg.2014.00882