Representational Shifts during Category Acquisition: A Preference for Features that Provide Information for Multiple Categories

The present study was aimed at examining the factors inducing feature creation in the perceptual and semantic systems. The proposed research was guided by the hypothesis that due to the substantial load of categorization tasks, a preference exists for encoding features that are informative for the recognition of multiple categories. This claim extends theories stipulating that features are encoded for each categorization task individually and challenges theories that focus solely on encoding coincidences in input patterns. The research hypotheses were tested in a series of experiments, requiring participants to learn several new categories in controlled conditions. Experiment 1 demonstrated that while acquiring a sequence of four novel perceptual categories, features that are informative for recognizing several categories are created. This was observed despite an absence of suspicious coincidences in the input patterns. Moreover, Experiment 1 provided evidence that after acquiring novel features, the perceptual system actively reconstructs the representations of categories learned in the past so that they comply with the newly encoded features. Experiment 2 demonstrated that learning an additional, fifth, perceptual category can be significantly facilitated, if it is congruent with the previously encoded feature structure. Finally, similar results were observed in Experiments 3 and 4, that used semantic stimuli (job candidate characteristics) instead of the perceptual stimuli (color cubes configurations) used in the first two experiments. It is therefore suggested that encoding features that are informative for multiple categories may be a general principle, underlying the feature creation processes in both the perceptual and the semantic systems.