The Dynamics of Ratio Scaling
Lawrence E. Marks John B. Pierce Foundation Laboratory, and Yale University
Intellectual disciplines, much like the people who engender them, ofttimes experience their own seven ages. By my reckoning, psychophysics entered its early middle age around the second to third decade or so of this century--around the time that Richardson and Ross ( 1930) devised the first, albeit primitive, version of the method of magnitude estimation. But 1930 was a bit too early for such procedures to take hold, and it was a quarter- century later that Smitty Stevens ( 1957) thwarted psychophysics's incipient mid-life crisis by reintroducing to the discipline, expanding, and expounding what he called ratio scaling methods.
Implicit descendants as we are of Kant's transcendental approach, psychophysicists have always looked in two directions--at sensory, perceptual, or cognitive systems that are elucidated by psychophysical experiments, and at the limitations imposed on our understanding of these systems, limitations imposed by processes of judgment. Depending on theoretical predilections, we tend to choose sides in the debate whether judgmental processes are postperceptual and merely tacked onto underlying sensory/ perceptual ones (as Stevens, 1975, among others, believed), or whether perception and judgment are intricately intertwined (as Helson, 1964, maintained). Much as I hate saying so, I think that this theoretical division in psychophysics mimics--though of course it precedes--a division in contemporary cognitive psychology between traditional views of information processing as largely serial and recent developments in interactive parallel processing, known as connectionism. (Cognitive psychology is now expe-