I have reviewed the two papers "Computational Implications of Gestalt Theory" by Steven Lehar (P&P 99-098, 99-099). In these papers Lehar develops a theory of the formation of illusory contours. Unfortunately, I must recommend that the papers be rejected. There are several reasons:
1. The papers are unnecessarily long. They could be cut in half.
2. The papers claim originality for the "perceptual modeling approach". In fact this approach is well known and widely used.
3. The papers inadequately reference the recent extensive literature on the formation of illusory contours. No published literature since 1992 is referenced at all.
4. The model developed in the papers is not adequately tested to demonstrate that it successfully accounts for the behavior of illusory contours. The evidence presented is not systematic and rigorous, and ignores all the recent literature on illusory contours.