Egon Brunswik abandoned the stimulus-organism-response of the day for his environmental texture as far back as 1935. But he got little notice, especially with the advent of the computer in the 1960s. Texture could not compete with physics and the computer. But by the end of the twentieth century, there was an emphasis on ecology and environmental texture. As Hammond notes, the times have caught up with Brunswik. Of course, this emphasis goes back to Darwin’s “entangled bank”. Darwin made the point that living organisms evolve and survive in entangled relationships among other evolving organisms, all coping with the fallible indicators that are others’ behaviors.
Brunswik questioned why if psychologists demand that the logic, methods, and procedures of sampling statistics must be applied to the subjects in the experiment, why should not such logic, methods and procedures be applied to the conditions of the experiment. However, Brunswik’s representative design (sampling) still loses out to the rule of one variable. Hammond points out a recent study that took Brunswik’s ecological view by looking at the offices and bedrooms of people to determine their personalities with the finding that “individuals select and craft physical environments that reflect and reinforce who they are.”
Herbert Simon was a student of chess and also saw the computer as teaching us about our cognitive processes. Simon found a lack of coherence in human thought and this led to the idea of “satisficing”. Satisficing is searching for a better idea or move in chess until you are satisfied with it. This is in line with Simon’s other important idea, bounded rationality which admits that humans cannot search every possible move on the chess board or elsewhere as full rationality demands. Simon won the Nobel prize for his work. Hammond clearly respects Simon, but disagrees with chess being anything other than a model for a few peculiar judgment tasks. Chess demands coherence and provides a very limited environment.
Gerd Gigerenzer has taken up Simon’s bounded rationality, and wants to show that coherence is of little importance. However, he also takes up Brunswik’s ideas by calling correspondence as ecological rationality. Gigerenzer argues that “rationality can be found in the use of fast and frugal heuristics, inference mechanisms that can be simple and smart.” Hammond says that Gigerenzer tells us that the ecology–not some theory of what rationality implies–tells us “when to stop”, that is, tells us when we are right or wrong, and thus affords us a criterion for rationality.
Hammond also see value in the pursuits of the neuropsychologists like Stanislas Dehaene through the fMRI and other tools. Hammond presents one of Dehaene study’s of the so called approximate man. The left side of the participant’s brain was destroyed by a severe blow. Since this is the general situs of our calculating abilities, Dehaene wanted to see how the injury impacted that ability. He found that intuitive judgments could still be made by the intact right side of his brain, so he made reasonable quantitative approximations (2+2=3), but he could not do exact calculations. Dehaene concluded from this that: “The brain is not a logical machine, but an analog device.” Hammond suggests that means that analytical cognition must be taught.
Hammond includes Cass Sunstein, Daniel Kahneman, Amos Tversky, and Richard Thaler in this group. He seems to prefer Cass Sunstein’s work. None of these researchers supports rational choice theory as set up by the classical economists. Curiously, Hammond never mentions prospect theory. He asserts that the coherence researchers demonstrate the general failure of humans to achieve intuitively coherent judgments under uncertainty. Hammond suggests that these researchers have more or less accepted conventional rationality (and the statistical technique known as Bayes’s theorem) as a standard and then with their experiments shown that humans fail to achieve that standard. Hammond points out Robin Hogarth’s criticism of the experiments.
Researchers who study people’s decision making processes seek results that are generalizable. However, conclusions are often based on contrived experimental ‘incidents’ with little understanding as to how these samples of behavior relate to the population of situations that people encounter in their naturally occurring environments.
This again seems to be going along with Brunswik and texture of the environment.
Disagreement about Heuristics
Kahneman and the coherence researchers find that heuristics can be useful, but they lead to errors in judgment. Gigerenzer and the correspondence researchers disagree and believe that heuristics make us smart. Hammond notes that Gigerenzer acknowledges the work of the coherence researchers, but that the coherence researchers do not acknowledge the work of Gigerenzer and the correspondence researchers. I love Hammond’s mention of this. I note that Kahneman’s Thing Fast and Slow does not mention Gigerenzer in the body of the book, but does refer to him in the notes as their “most persistent critic.”
Hammond, of course, sees both sides as flawed. He points out that Gigerenzer ignores feedback. Feedback is a critical parameter of a judgment task ecology. Some ecologies provide immediate and unambiguous feedback, some delayed and ambiguous, and some no feedback at all. Tasks that offer no feedback are ones that demand coherence. On the other hand, Gigerenzer shows in great detail how difficult, time consuming it would be to perform an unboundedly rational procedure, and how irrelevant Bayes’s theorem is to any but the most simple tasks. Gigerenzer and Kahneman also tangle on the issue of human shortcomings with respect to probabilistic information. Gigerenzer is persuasive in showing that our common ecology is based on the expression of natural frequencies as opposed to probabilities.
Hammond, K.R. (2007) Beyond Rationality: The Search for Wisdom in a Troubled Time. U.S.A: Oxford University Press.