E Inductive Reasoning as a Default Cognitive Function
In trying to understand the irrational, as well as the rational, sides to scientific thinking, we've talked about biases and fads, and its time to revisit inductive reasoning in this context because it shares similarities with them.
One of the enduring mysteries (for me, anyway) about inductive reasoning is why anyone still believes that it is a special intellectual skill. We've seen (Chapter 1) and dismissed such profundities as “induction is how you justify the expectation that your computer will not suddenly blow up,” or that it “is the reasoning by which one determines the simple expectation that the sun will rise and then set at a set time each day.... It is a particular predictive model of reality that is verified on a daily basis.” And we know that a rationalist would counter that our scientific confidence in the sun's rising is grounded on well-tested physical theories that explain why it does not suddenly explode or go wandering away from Earth. We have no reason to think that these exhaustively corroborated theories are not true, and inductive inferences from having observed countless sunrises doesn't add anything to what we know. On the contrary, at some point in the distant future inductive reasoning will fail—our star will have entered the final stages of solar evolution and will not rise despite having done so roughly a trillion times. Long before then, however, science will have predicted its end (and, we hope, devised a back-up plan). For the most part, we need reasons to think that natural things will not continue the way they are, rather than inductive reasons to think that they will.But, if induction is not a special inferential reasoning process that leads human beings to expect regularity, why do we expect it? Or, to put it another way, does anticipating regularity in the environment qualify as “reasoning” of any kind? It is a fair question.
Expecting regularity does not, for one thing, require conscious thought; most of us never cogitate about sunrise at all: we wake up and start our days and there's the sun. Observing and responding to repetitive events in the environment is built into the human cognitive toolkit that was shaped by evolutionary changes that took place millions of years ago when our primate ancestors were eking out a living on the African veldt. Genes that made it easier to find food, shelter, and chances to procreate; genes that allowed us to notice and keep away from danger, etc., were passed on. The culmination of all this evolution is that we have a genetic predisposition to detect meaningful regularities in the environment and to act on them. This is the heart of “inductive reasoning,” and it explains why we find its conclusions so irresistible.Of course, if expecting environmental regularities is wired into our genome, there is no reason to suspect that human beings would be solely blessed with it. All animals would surely benefit by having such an ability, and indeed, they all do. It's why your dog, Snowball, runs into the kitchen when he hears you opening the dog food; it's why the ducks in the pond paddle over to you when you start opening your lunch bag, etc.
In point of fact, I may be guilty of a politically incorrect “taxonism” (an unfair preference for one phylogenetic taxon over another). It appears that bacteria and plants have claims to being inductive reasoners, as well. Bacteriologists have found that the omnipresent gut bacterium Escherichia coli (E. coli), can learn to evolve in response to regular changes in environmental conditions.40 The experimenters systematically changed the temperature in which the bacterial colonies lived from 25°C to 37°C and simultaneously lowered the oxygen content in the atmosphere from 20% to zero. After a few weeks (hundreds of generations) “the bacteria had ‘learned’ to anticipate the drop in oxygen by altering their metabolism just after the temperature change.”
Not to be outdone, botanists got into the act by showing that the shoots of garden peas will grow into the arm of a Y-maze where light is “predicted” to be found.41 Monica Gagliano and colleagues used a Pavlovian-training protocol in which a neutral cue, a breeze caused by a fan, was either paired or not paired with light during a training period.
Later, during the test phase with no light, the plants grew into the arm in which breeze had been associated with light during training. The authors concluded that “associative learning represents a universal adaptive mechanism shared by both animals and plants.” We naturally take it for granted that the human mental processes that mediate induction are vastly superior to whatever is going on in bacteria and plants, yet at a descriptive behavioral level, it is not obvious that our attitude is warranted.Actually, you might argue that the Darwinian prevalence of inductive-like behaviors is sufficiently widespread that it turns the philosophical Problem of Induction inside out. We needn’t justify our belief that nature is uniform in order to justify induction, as philosophers have always assumed. Instead, we must recognize that the actual uniformity of nature is responsible for the tendency to depend on induction in the first place. That is, the genetic advantage (i.e., continued existence) conferred on life forms having the capacity to sense and act on regularities in nature, demonstrates that nature must be fairly regular on the time spans required for evolution. In a totally chaotic world, expecting environmental regularities would be fatal because there would be none; you'd rarely find food today where you found it yesterday; the tiger would always be popping up at the water hole at unexpected times, and so forth.
Eons of correctly predicting regularity does not mean that it is always a great idea to assume nature is regular, of course. Dinosaur-extinguishing meteors or environment-reshaping human beings can forge immense changes on shorter time scales than it takes most species can adapt. Conversely, you can't make the argument that, because man-made, global catastrophes have not yet happened, they cannot happen. These illustrations show that flawed automatic inductive reasoning can play havoc with our ability to reason scientifically.
11.