Story of Psychology - Morton Hunt [338]
The solution, though simple, requires a seeming retreat in order to advance. The two sons get in and row across; one debarks and the other rows back and lands; the father rows across and gets out; the son on that side rows back, picks up his brother, and returns to the far shore. GPS, in devising and testing this solution, was doing something akin to human problem solving. By means of the same heuristic—a broad stratagem of exploration and evaluation—it was able to solve similar but far more difficult problems.
Two basic features of GPS and later artificial intelligence (AI) programs brought about a metamorphosis in cognitive psychology by giving psychologists a more detailed and workable conception of mental processes than any they had previously had, plus a practical way to investigate them.9
The first of those features is representation: the use of symbols to stand for other symbols or events. In GPS, numbers stand for words or relationships, and in the hardware (the actual computer) operated by GPS, groups of transistors, acting as binary switches that are either on or off, stand for those numbers. By analogy, cognitive psychologists could conceive of the images, words, and other symbols stored in the mind as representations of external events, and of the brain’s neural responses as representations of those images, symbols, and thoughts. A representation, in other words, corresponds to the thing it represents without being at all similar to it. But this was actually an old discovery in new form; Descartes and Fermat discovered long ago that algebraic equations can be represented by lines drawn on a graph.
The second feature is information processing: the transforming and manipulating of data by the program in order to achieve a goal. In the case of GPS, incoming information—the feedback of each step—was evaluated as to where it had led, used to determine the next step, stored in memory, retrieved if needed again, and so on. By analogy, cognitive psychologists could conceive of the mind as an information-processing program that transforms perceptions and other incoming data into mental representations and, step by step, evaluates them, uses them to determine what to do next in the attempt to reach its goal, adds them to memory, and retrieves them for use again as needed.
The information-processing (IP) or “computational” model of thinking has been the guiding metaphor of cognitive psychology ever since the 1960s, and has enabled researchers and theorists to explore the inner universe of the mind as never before.
One specimen of such an exploration will exemplify how the IP model enables cognitive psychologists to ascertain what takes place in the mind. In a 1967 experiment, a research team headed by Michael Posner asked its subjects to say aloud, as fast as possible, whether two letters projected on a screen had the same or different names. When the subjects saw this
AA
they almost instantly said “Same,” and when they saw this
Aa
they again almost instantly said “Same.” But the researchers, using a highly accurate timer, measured a minuscule difference. On average, subjects replied to AA in 549 milliseconds and to Aa in 623 milliseconds. A tiny difference, to be sure—but a statistically significant one.10 What could account for it?
The IP model envisions any simple cognitive process as a series of step-by-step actions performed on the data. The following simple flow diagram, typical of many drawn by cognitive psychologists, symbolizes what goes on when we see and recognize something:
FIGURE 39
A typical information-processing diagram
That accounts for the reaction-time difference in the experiment. If an image proceeds directly from the first “processing” box to “consciousness,” it does so in less time than when it must pass through two or three boxes. In order to identify