A Cognitive Revelation

Beginning in roughly 1960 the term cognitive changed from a Kantian (continental) to a Lockean (British) interpretation. This was to have an influence on the development of behaviorism, as reflected in the theorizing of Albert Bandura (1986). Trained in the Hullian tradition, Bandura added significantly to this style of S-R explanation by postulating a triadic form of causation—including a person’s (1) internal cognitions, affects, and biology; (2) external, overt behaviors; and (3) environmental pressures to perform or believe in certain ways. Each of these sources of causation influenced the other in a reciprocal manner. An important cognitive influence on a person’s performance in life’s challenging circumstances is self-efficacy. This is the sense of conviction that a person has regarding whether he or she can execute the behavior required to produce some outcome—such as passing an examination, or hosting a party. As it is a learned behavior, achieved by accomplishing challenging tasks as well as modeling the behavior of others. Bandura’s self-efficacy is seen by some as a further development of mediational theorizing. Bandura believes that agency (producing desired re-suits) can be understood through triadic causation and the development of efficacy both in persons and social groups.

Returning to the matter of how cognition is to be understood, a Kantian (phenomenological) formulation is framed in the first-person or introspective terms, considering things from the perspective of “I, me, us, we” and so on. A Lockean (empiricistic) understanding of cognition is framed in a third-person or extraspective perspective, referring to “that, it, they, them” and so on. Purposive explanation is introspective for it assigns responsibility to an identity or identities whose actions can only be understood from their own point(s) of view. Mechanistic accounts are extraspective, understanding behavior as taking place according to impersonal forces “over there.” This shift in meaning of cognition was the result of a “Cognitive Revolution” that descended upon psychology. The impetus for this revolution in psycho-logical theorizing was the perfection of electronic computer technology. The term artificial intelligence (AI) was coined to typify this approach. The general consensus is that AI as a field of study was born at an institute sponsored by the Rockefeller Foundation, held at Dartmouth College in the summer of 1956 (Newell & Simon, 1972, pp. 873-889).

It would seem that a machine, by its very nature, cannot be explained introspectively. And yet, if we want to assign purpose to some behaving organism we must see things from this organism’s (introspective) point of view. Mechanism has traditionally been understood as the very opposite of intentionality or purposeful action. But with the advent of computers a new line of argument arose, one which seemed to capture a truly cognizing, intending machine! To grasp this argument we must understand the concepts of input, output, and especially feedback. Feedback is the return of some of the machine’s signaling output as input, so that a computer is continually monitoring its ongoing actions. Positive feedback signals what is taking place and negative feedback signals whether some preselected goal or target is “on” or “off” course. If a missile—a cybernetic machine—is for some reason off course, the negative feedback will automatically redirect its firing mechanisms to put it back on target. This self-correcting capacity of negative feedback has prompted many AI enthusiasts to believe that negative feedback is the same thing as purposive behavior, intentionality, or free will (Rosenblueth, Wiener, & Bigelow, 1943).

Alan Turing (1912-1954) further captured the imagination of psychologists when he introduced a strategy in which a computer expert could write machine programs that misled people into thinking that the machines which processed these programs were alive (1950). For example, a person might engage the computer in a “conversation,” typing questions to be answered by the prearranged program (giving the illusion of talking to the machine). If the program could fool the person into believing that another person was answering these questions from another location, then the “Turing test” would have been passed. The program has simulated the cognitions of an actual, living person. If we combine this feat with the self-correcting nature of the cybernetic machine, it seems as though a completely mechanistic account of psychology is now possible. Some pretty optimistic predictions along this line have been made by those who would like to turn psychology into AI (Simon, 1985, p. 89).

The cognitive revolution was heavily underway by the 1980s but it was not without its detractors. First of all, the question has been raised by many as to whether there is anything truly revolutionary in the theoretical accounts underlying computing machinery. Clark Hull had built calculating machines to solve various problems, and many see little difference between Hullian S-R formulations and the input-output concept. Both are based on blind, efficient causation. Doubtless Hull could have found a way of explaining feedback through mediation. When framed as “information” being input and output. it is easy to think of meaningful messages taking place. Actually, the meaning of a computer is all in the program as written and logically arrayed by the programmer—a human being! The information under electronic signaling, the actual mechanism of a computer, has nothing to do with meaning (Shannon & Weaver, 1962, pp. 3, 99). Nor can the machine spontaneously negate the directions of its executive pro-gram. It cannot say no to its unquestionable assumptions in the impulsively illogical way that human beings can. The “it” here reminds us that computers are not understood introspectively. Psychologists favoring an introspective account of human behavior will doubtless continue to express dissatisfaction with the computer model.

Although the meaning of cognition has shifted to involve extraspective, mechanistic accounts, there have always been strong advocates for the original Kantian meaning. For example, from 1950 to 1960 a “New Look” in perceptual theory and research was advanced (Bruner et al.. 1957). The emphasis here was on an active role for the perceiver, who contributed to what was actually “seen” based on personal motivation. Words that had a negative meaning took longer to be recognized when flashed on a screen than words with a positive meaning. Some psychologists took this to be a perceptual defense, an intentional effort to avoid the unpleasant. In time, such work led to studies on subliminal perception, as when the word popcorn is flashed on a movie screen too rapidly to be seen but nevertheless appears to suggest the purchase of popcorn to those watching the movie. Once again, there is disagreement over whether or not such subliminal perceiving is possible.