Deep computational neurophenomenology: A methodological framework for investigating the how of experience

Abstract

The context for our paper comes from the neurophenomenology research program initiated by Francisco Varela at the end of the 1990s. Varela’s working hypothesis was that, to be successful, a consciousness research program must progress by relating first-person phenomenological accounts of the structure of experience and their third-person counterparts in neuroscience through reciprocal or mutual constraints. Leveraging Bayesian mechanics, in particular deep parametric active inference, we demonstrate the potential for epistemically advantageous mutual constraints between phenomenological, computational, behavioural and physiological vocabularies. Specifically, the dual information geometry of Bayesian mechanics serves to establish, under certain conditions, generative passages between lived experience and its physiological instantiation. This paper argues for the epistemological necessity of such passages and the inclusion of trained reflective awareness in neurophenomenological empirical approaches, showcasing incremental epistemic gains that come from by shifting the focus from the contents of experience (i.e. what a subject experiences in a given experimental set-up) to the how of experience - the activities of consciousness that allow for a meaningful world to appear to us as such in lived experience. The explanatory power of the resulting framework, deep computational neurophenomenology, arises from the disciplined circulation between first and third-person perspectives enabled by the formalism of deep parametric active inference, where parametric depth refers to a property of generative models that can form beliefs about the parameters of their own modelling process. Hence, this computational formalism contributes to understanding consciousness by bridging phenomenological descriptions and physiological instantiations, whilst also highlighting the significance of trained first person investigation in experimental protocols.

Shedding light on changes in subjective experience during an intensive contemplative retreat: the Lyon Assessment of Meditation Phenomenology (LAMP) questionnaire

“…Our instrument enables investigation of causal relationships among variables, population heterogeneity and individual variability, as well as predictive factors of long-lasting changes. Such an approach could serve as a blueprint for guiding clinical and neurophenomenological investigations of mental health (34) and meditation practices (35), including neurocomputational modeling (37).…”

An overview of neurophenomenological approaches to meditation and their relevance to clinical research

“…In the fourth section, we explore how the previously mentioned crises concerning mechanism and subjectivity were initially tackled within the theoretical framework of neurophenomenology, with the aim to "naturalize phenomenology" (20,13,21,22) and how this framework has been recently updated. This updated computational framework, grounded in Bayesian mechanics (23,24), facilitates the establishment of a bridge between phenomenology and neurobiological mechanisms, enabling the interplay between first-person and third-person perspectives on understanding consciousness (discussing A4 and A5).…”

“…In this last section, we discuss a neurocomputational version of NPh that aims at "naturalizing phenomenology" (20,74), and which has seen recent developments (23,24) (table 1, levels A4-A5). This approach focuses on aspects of subjective experience, differing from other models that primarily target psychological processes of meditation (e.g.…”

“…NPh suggests that neither has epistemic primacy, implying that models must be both biologically and phenomenologically plausible (5), a double constraint solved by active inference. Active inference offers a biologically plausible framework (76,77), and computational phenomenology (23,24,40,41,79) recently proposed to use generative modeling techniques to explain characteristics of lived experience described by phenomenologists and contemplatives. Generative models in Bayesian statistics map observed data to infer hidden causes, and 'computational phenomenology' (level A4) interprets perceptual intention as an inferential process on some raw sensorial qualitative data (24), where the intended perceptual object is interpreted as the most probable inferred latent cause based on the subject's generative model (24).…”

“…Generative models in Bayesian statistics map observed data to infer hidden causes, and 'computational phenomenology' (level A4) interprets perceptual intention as an inferential process on some raw sensorial qualitative data (24), where the intended perceptual object is interpreted as the most probable inferred latent cause based on the subject's generative model (24). This approach has modeled not just the 'what' but the 'how' of experience (12,23), requiring 'deep' parametric active inference (44) (Figure 3). Parametric depth in Bayesian mechanics refers to a property of generative models that can form beliefs about the parameters of their own modeling process (23,79).…”