Adaptive Perspectives on Human-Technology Interaction: Methods and Models for Cognitive Engineering and Human-Computer Interaction

William M. Goldstein

Egon Brunswik (1903–1955) was a perceptual psychologist with deep interests in the history and philosophy of science. His work on perception led him to develop a general vision for psychology called probabilistic functionalism (Brunswik, 1952, 1956; Hammond, 1966; Hammond & Stewart, 2001). Brunswik's thinking had a systemic integrity with interwoven perspectives on the nature of psychology, its definitive problems, and proper methodology. Unfortunately, some of Brunswik's positions were out of step with the mainstream psychology of his day, and his work was not well received at the time (Gigerenzer, 1987; Kurz & Tweney, 1997; Leary, 1987). Even today, although a number of Brunswik's ideas are widely employed by psychologists who may be unaware of their origins, Brunswikian thinking is represented mainly by a vocal minority of researchers who study judgment and decision making (Goldstein, 2004). However, Brunswik's probabilistic functionalism has been receiving increasing attention from researchers in cognitive engineering (see chapter 1), and one purpose of this book is to announce, explain, and promote that trend. The purpose of this chapter is to acquaint readers with the basics of probabilistic functionalism so that they can see its appeal for cognitive engineering and evaluate the contributions it has facilitated in the remaining chapters of this book.

It might help to introduce Brunswik's thinking by outlining the kind of research problem that inspired it. Brunswik's early research, conducted during the 1920s and 1930s when he was a member of the Vienna Psychological Institute, concentrated on various perceptual constancies. Size constancy, for example, is demonstrated by the fact that the apparent (e.g., estimated) size of an object tends to remain more or less constant at various distances from an observer, even though the object's projection on the retina differs. Other features, such as shape and color, similarly tend to be perceived as relatively constant despite changes in the proximal information available to the observer. The issue of perceptual constancy, stated this way, contains the seeds of some enduring concerns for Brunswik: (1) the relationship between the proximal information available to the observer and the distal state of affairs, emphasizing the ambiguous or even misleading implications of individual items of proximal information taken in isolation; (2) the establishment of a stable percept despite (or perhaps because of) variability and interrelationships among the items of proximal information; and (3) the accuracy of the stabilized percept. Briefly,

-10-