Assessing relations between imagery and perception.

Direct comparisons between perceptual performance and reports made under imagery instructions are uninterpretable except in cases where (a) the perceptual phenomenon is well understood and (b) experimental constraints are sufficient to establish imagery causally in the effect. Neither condition seems to be satisfied in Finke and Kurtzman's article. Furthermore, if imagery is an internal medium conceived in introspective terms, the second condition can never be met because the medium is not accessible to operational definition. Perhaps the basic problem with this research is that the instructions used to create conditions in imagery experiments do not qualify as operational definitions. Reports of successful compliance with imagery instructions are logically equivalent to introspective reports.