Ensemble size perception as a case study of the bounds of adaptation.

Repulsive adaptation effects are widely assumed to obtain for all perceptually represented dimensions. However, the ubiquity of adaptation effects within perception remains untested. We examined ensemble size adaptation as a case study to probe whether adaptation occurs for all perceptually encoded properties. Across four experiments, we investigated whether observers adapt to average size and/or cumulative size of dot arrays. In Experiments 1a, 1b, and 1c participants adapted to displays varying in cumulative and/or average dot size, then judged either the average dot size (1a) or the cumulative dot size (1b) of paired test displays. Results revealed robust adaptation to cumulative size but not average size, regardless of task instructions, and even when confounds with brightness were controlled for (1c). Experiment 2 tested "reverse" adaptation to displays containing smaller average and/or cumulative dots size and, again, found adaptation effects for cumulative size only. The observed lack of adaptation to average size across each of these experiments forces a reinterpretation of previous studies that have investigated size adaptation and calls into question arguments which have assumed adaptation to be universal within perception, given a large body of work that finds average size to be perceptually encoded.