ART & ORGANISM

notes on HALLUCINATIONS, both everyday and extraordinary 

What are the biases that give form to our hallucinations?

PERCEPTION AND HALLUCINATION

Effects of Hallucination Proneness and Sensory Resolution on Prior Biases in Human Perceptual Inference of Time Intervals: “Bayesian models of perception posit that percepts result from the optimal integration of new sensory information and prior expectations. In turn, prominent models of perceptual disturbances in psychosis frame hallucination-like phenomena as percepts excessively biased toward perceptual prior expectations.”  Duhamel et al (2023) report:

“Current theories of psychosis posit that hallucination proneness results from excessive influence of prior expectations on perception. It is not clear whether this prior bias represents a primary top-down process related to the representation or use of prior beliefs or instead a secondary bottom-up process stemming from imprecise sensory representations at early processing stages. To address this question, we examined interval timing behaviors captured by Bayesian perceptual-inference models. Our data support the notion that excessive influence of prior expectations associated with hallucination propensity is not directly secondary to sensory imprecision and is instead more consistent with a primary top-down process. These results help refine computational theories of psychosis and may contribute to the development of improved intervention targets.

Introduction.  An intuitive and common conception of perception is that our sensory systems accurately represent the external world as it actually is, a view that likens our perception of a visual scene to a photograph of such scene. In contrast with this view, substantial evidence indicates that perception is an idiosyncratic process that does not solely rely on input from the senses but also relies heavily on the context of what is being sensed and the expectations derived from this context. (Duhamel et al. 2023).  Formalizing this notion, Bayesian models of perceptual inference portray percepts as the synthesis of an optimal integration of two information sources, that is, the sensory evidence, represented by a likelihood distribution, and context-derived predictions, represented by a prior probability distribution (Stocker and Simoncelli, 2005). Here, sensory evidence is typically thought of as a bottom-up signal reflecting sensory information associated with a given incoming stimulus, whereas the context-derived prediction is a top-down signal that conveys an expectation derived from previously experienced statistical regularities among the stimuli in a given context. Critically, Bayesian models posit that these two sources of information are weighted based on their respective reliabilities; the synthesized percept will be more biased toward the prior expectation either (1) when sensory evidence is noisier (Jazayeri and Shadlen, 2010; Ryan, 2011; Cicchini et al., 2012; Karaminis et al., 2016; Egger and Jazayeri, 2018) or (2) when context-derived predictions are more precise (Friston, 2005; Stocker and Simoncelli, 2005; Zelano et al., 2011; Adams et al., 2013; Cassidy et al., 2018), and toward the mean of the sensory evidence otherwise.