On the concept of error in visual perception: An example from simultaneous lightness contrast

This work deals with the concepts of “error” and “veridicality” in visual perception studies by considering the matching paradigm often employed in empirical re-search to study simultaneous lightness contrast (SLC). Matching paradigms often employ Neutral Value Munsell scales, and there is a strong tendency in the field to consider these scales as a ruler capable of showing veridical lightness values, that is perfect transformations of reflectance values into perceptual values. If this were the case, then Munsell scales should show high constancy to critical changes inside the visual scene. We performed an experiment in which three groups of observers were asked to perform a matching task for a classic SLC display. Each group used the same Munsell scale (MS) but seen against three different backgrounds: white, black, or white-black chequered. Results showed that the background against which the MS is seen heavily influences matches for the target on the black back-ground of the SLC display. Our results support the claim that achromatic Munsell scales are not capable of showing veridical lightness values.