Tuesday, 10 February 2015

Can Evolution get us off the Hook?

This is the second in our series of posts on the papers published in a special issue of Consciousness and Cognition on the Costs and Benefits of Imperfect Cognitions. Here Maarten Boudry summarises his paper (co-written with Michael Vlerick and Ryan McKay) 'Can Evolution get us off the Hook? Evaluating the Ecological Defence of Human Rationality'.

In the opening lines of his essay 'An Outline of Intellectual Rubbish', Bertrand Russell wrote that 'Man is a rational animal  so at least I have been told. Throughout a long life, I have looked diligently for evidence in favor of this statement, but so far I have not had the good fortune to come across it'. Russell’s cry of despair is echoed by many writers. There is a cottage industry of books purporting to show that man is anything but a rational animal. Human reason, or so we are told, is a paltry and botched device, riddled with bias and error. Humans are foolish, obstinate, superstitious, irrational. The psychologist John Kihlstrom called it the People are Stupid School of Psychology (PASSP).

What arguments can an attorney, called in to defend Homo sapiens against these charges of irrationality, bring forward in favor of her client? In our paper, myself, Michael Vlerick, and Ryan McKay have explored one such line of defence: the Ecology Apology.

Researchers in the burgeoning field of ecological rationality have argued that, if we approach the human mind as an adaptive toolbox full of heuristics, tailored to particular ancestral environments, many of its apparent foibles and fallacies evaporate. Human rationality is not general and content-free, but is designed to work in a particular ecology, argues Gerd Gigerenzer, one of the main proponents of this school. Mind and environment, in a memorable phrase by Herbert Simon, are like two blades of a pair of scissors.

In our paper, we discern two strands in the literature on ecological rationality. One of them draws our attention to the artificial set-up of many psychology experiments. If you deliberately put people in an environment where their usual heuristics and intuitions are of no avail, it is not surprising that they perform badly. The human mind is designed to deal with the real world – not with devious psychologists bent on tripping people up.

There is a second strand in this program, however, which appeals directly to the evolutionary history of the human mind. For example, people may appear to be superstitious, but a little superstition makes sense from an adaptive point of view. In a hostile world, spotting causal patterns can be a matter of life or death. Your ancestors were those hominids who erred on the side of caution, ending up with a couple of false positives in the process. Fellow hominids may have laughed at their little superstitions, but your ancestors had the last laugh.

Though it is true that there is method in human madness, we should be wary of conflating two very different things: appraisals of rationality on the individual level, and arguments about the evolutionary 'rationale' of adaptations. The latter, as Ronald de Sousa pointed out, provides a mere simulacrum of rationality. Evolution by natural selection produces an appearance of design, to which we can fruitfully apply familiar frameworks of rationality, such as game theory and signal detection theory. But these two levels should not be conflated. Adaptive design, as it happens, is perfectly compatible with simple-mindedness, irrationality, or dumbness. Besides, your genes have different interests than you have.

Take the example of a gambler at the roulette table who bets on a number because he thinks it is 'overdue'. This is known as the gambler’s fallacy: roulette wheels have no memories, so each turn is independent from the next. Why does the gambler persist in his irrationality? There is a sensible evolutionary explanation: real life is nothing like a casino. Most events have a characteristic hazard function: the probability of each occurrence (say, a rainy day) is affected by the occurrence of earlier events (say, a streak of sunny days). In the real world, argues Steven Pinker, 'an astute observer should commit the gambler’s fallacy'.

That is true, as far as it goes, but does it get our gambler off the hook? Mindlessly paying heed to your evolved intuitive hunches does not make you rational. Rationality, as we define it in our paper, is the ability to deliberate and act on contingent and local information. Evolution works over the long haul, abstracting away from local contingencies, but human reasoners face specific decisions in specific circumstances.

If our gambler cannot be cured of his conviction that the wheel has some sort of memory, even after being given a crash course in statistics, should we not insist that he is irrational, despite the adaptive 'rationality' of his cognitive make-up? There is a rational course of action available, after all: seasoned gamblers resist the intuitive pull of their evolved pattern detection mechanisms, overruling it in light of contingent information.

In general, we applaud the efforts of ecological rationalists to correct the bleak view of human reason. In our paper, we document cases where they have successfully forced us to reconsider alleged human irrationality (e.g. the Linda problem). These examples draw our attention to artificial experiments and the rich complexities of real life, but they do not involve a locus shift between the personal and the adaptive level.

That is as it should be. If want to get human reasoners off the hook, we should do so without using evolutionary alibis.

No comments:

Post a Comment

Comments are moderated.