Four assumptions about invariance in perception.

The term invariance has become more central to current views of perception. I take this as a good trend, but the term is rooted in mathematics, and its use in perception brings with it a host of assumptions that have generally been unexamined. The purpose of this article is to state some of these assumptions and assess their validity, with the hope that we can continue to find the term useful while acknowledging its limitations. The assumptions discussed are that (a) mathematics is an appropriate descriptive language for perception, (b) mathematical truths are transportable into perception without change of meaning, (c) mathematical imports are useful in explaining perception, and (d) perceptual invariants, like their mathematical counterparts, are absolute and not subject to threshold considerations.