Qualitative inference rules for perceptual transparency

ABSTRACT: We propose two criteria – expressive form and deducibility – for characterizing qualitative (as distinct from quantitative) information in psychophysical contexts. We then define qualitative inference rules as predictive rules in which both the input information (on the stimulus) and the output information (on the percept) are qualitative in character. After fixing essential ideas about perceptual transparency, we illustrate the concept of qualitative inference rule by examples taken from the psychophysical literature, distinguishing rules predicting the occurrence of perceptual transparency and rules predicting properties associated with perceptual transparency. In the last section we comment on possible merits of qualitative inference rules, and notice a relationship between them and intra-perceptual dependences, these being possible items of experimental phenomenology of vision.