On the Use of Information Measures in Studies of Form Perception

Information measures have made one major contribution to the study of form perception. This is the same contribution that they have made to other areas of psychological research. They have put the old, seemingly stale problems into a new, and sometimes exciting light. In so doing, they have stimulated renewed research interests in areas, some of which have not seen such lively activity since the passing of the first generation of the "new" psychology. The advance of a science is often limited for want of new research tools. Astronomy needed the optical telescope (and later the radio telescope) to open new heavenly vistas. The biological sciences needed the optical microscope, as they more recently needed the electron microscope, to provide new visions of the minute. Between these two extremes there are many other examples of interactions between scientific advances on the one hand, and instrument inventions on the other. Some of these instruments have been "tools" in a less literal sense. Analytic geometry, the calculus, and the psychophysical methods are three examples. That "information measures" form such a research tool is the theme to be developed in the present paper. The argument is that measures such as uncertainty, redundancy, equivocation, etc., provide useful, meaningful metrics for describing certain stimulus and response conditions under which man's behavior is observed and measured. This is not to imply that information theory (or cybernetics, or any other communication theory) is necessarily offered as a paradigm for the behavior of man, anymore than are the more commonly used statistical theories so offered. Rather, it is to say that information measures provide a way of gathering and systematizing psychological data, just as the correlation and variance measures have done in the past. They may color and otherwise influence psychological theories, but this does not mean that they necessarily are the theories. In being so used, information measures have revitalized certain areas of psychological research. One of these areas is that of the visual perception