Optical-geometrical illusions: A list of problems  

Vicario G. B.
Università di Udine, Udine, Italy

A list of problems to be solved sooner or later in investigations on optical-geometrical illusions is set forth. (a) Etymology. Optical-geometrical illusions are no "errors": illusions are surprising and unavoidable (e.g., Delboeuf, 1892*25), while errors are admitted and corrigible (e.g., Tolansky, 1964*26). (b) Naming and definition. Names and definitions are countless: optical-geometrical illusions (Oppel, 1855; Wundt, 1898; Rausch, 1952; Metzger, 1975), optical illusions (Delboeuf, 1865a; Pegrassi, 1904; Ehrenstein, 1954; Tolansky, 1964; Imai, 1984), optical paradoxa (Brentano, 1892), optical illusions of judgement (Müller-Lyer, 1889), visual illusions (Luckiesh, 1922; Robinson, 1972), illusions (Goto & Tanaka, 2005), figures trompeuses (Delboeuf, 1865b), inadequate representations (Benussi, 1906), and so on. (c) Phenomenology. Looking at some displays we feel to be in front of an illusion (e.g., Oppel, 1855*5), while looking at other ones we cannot see anything dubious (e.g., Ebbinghaus, 1908*82). (d) Nomenclature. It is at least chaotic, ranging from the name of the discoverer to the effect observed or to the supposed underlying mechanism. (e) Classification. The author is acquainted with 24 classifications of optical-geometrical illusions, but probably every scholar of the field can exhibit his own classification. (f) Delimiting the field of research. Too often optical-geometrical illusions are mixed with heterogeneous phenomena, like anomalous surfaces (Schumann, 1900a*7), masking (Metzger, 1975*79), impossible objects (Penrose & Penrose, 1958), alternating ambiguous figures (Boring, 1930, 444), coexisting ambiguous figures (Schuster, 1970*1), reversible figures (Necker, 1833*18; Schröder, 1858*12-13), brightness contrast (Kitaoka et al., 2004*6CF). (g) Pictorial perception. There is to settle the question whether the numerous optical-geometrical illusions involving two-dimensional representations of solids (to begin with Thiéry, 1895a*2, and Filehne, 1898*23) are indebted with pictorial perception (Gibson, 1954) or not. (h) Whole-parts relation. This relation is manifold and often indecipherable: sometimes parts influence the whole (Schumann, 1900b*6) and sometimes the whole influences parts (Sander, 1926b*9-I), but there are also cases by which influence is supposed but not perceivable (Vicario, 2006b*11.3) or not demonstrable (Vicario, 2006, unpublished). (i) Optical-geometrical illusions and everyday experience. There is a hýsteron próteron figure in the question: optical-geometrical illusions do not exist "also" in everyday experience (see, for instance, Höfler, 1896*2; Metzger et al., 1970*2a, Vicario, 2001*40) - they take place in everyday experience and therefore "also" in drawings or pictures, which are impoverished images of the actual environment. (j) The measure of illusions. Fisher's (1973) argument on the impossibility of obtaining a true measurement of illusions is demonstrated (Vicario, 2008, unpublished) and developed. (k) Why do optical-geometrical illusions exist? Adaptive behaviour, phylogenetically developed, should exclude their presence. A possible solution of the problem is to suppose that the evolution of perceptual systems is a matter of costs/benefits.