Humans, it has been said since Aristotle, are rational animals. Those who scoff at the phrase misunderstand it as contrasting with irrationality. But the proper contrast is with the non-rational, or arational. Inanimate objects are arational, because it makes no sense to tax them with irrationality. Humans are rational precisely because we are capable of irrationality.
Why Think? explores some of the roots of that capacity for irrationality. It focuses, in particular, on the uneasy relationship between intuitive and analytic thought. Our unreflective emotional responses are often apt. But sometimes we need to resort to the laborious process of explicit thinking. There is much evidence to suggest that we have a "two-track mind". Track 2 is the analytic track, typically involving explicit expression and reasoning in language, including the language of mathematics. Its arsenal of tools is only a few thousand years old. By contrast, some of the resources of Track 1, the intuitive track, have been honed over hundreds of millions of years. They include hard-wired responses, shared with other animals, as well as vast repertoires of learned behaviours that have become virtually automatic. Most of what we want and do is regulated by such processes.
The environments in which many Track 1 adaptations evolved were undoubtedly very different from ours. In present circumstances, intuitive responses may be maladaptive. Not everything you need to know was acquired in kindergarten. If you want to get to Mars, you won't do it by the seat of your pants. Sometimes, the rapidity of intuition just hastens our mistakes. Conversely, analytic solutions are likely to remain slow and clumsy even when they are superior in accuracy.
But how are the two tracks related? On the evolutionary path from unicellular organism to Homo sapiens (and, for each of us, from mindless zygote to articulate human), mere detection of stimuli mutates into representions of objects, and tropisms mutate to full-fledged desires. Tropisms look like goal-directed actions; but actually they are merely effects of natural causes. In sunflowers, for example, the cells on the shady side of the stem expand under the influence of blue light, thus bending the flower's stem towards the sun. A tropism isn't very different from a physical cause. A falling stone conforms to laws of nature, but it does not compute its trajectory. Similarly a unicellular organism may act as if it sought out nutrients, but it is simply moved by causal processes. What is the essential difference that explicit thinking makes?
The answer suggested in Why Think? rests in part on the difference between digital and analog representation. Digital systems of representation are based on a fiction: every stage of a physical process is identified as belonging to one or another of a finite set of types. When a digital system is well designed, standard instances of different types are almost never confused. The alphabet is a good-though imperfect-example: regardless of font or style, an A is an A, and a D is a D. One token of a may not look very much like another, and a d might resemble an a; but nothing ever counts as anything between a and d. Thus what we describe in language is structured differently than what is represented by subtle shades of intuition along a continuum. The tension between the mind's two tracks accounts for some of the trouble we get into when we try to be rational.
In comparing the "logic" of nature with the logic of rational thinking, we encounter a puzzle. Nature has no intentions, no goals in the literal sense. Yet biology can hardly do without the teleological notion of function. Happily, in a rare instance of philosophical progress, the past half-century has seen the construction of a usable conception of teleology applicable to natural functions devoid of explicit purposes. The basic idea is that an organ's function explains the presence of the organ in terms of the adaptive effects it has had in the past. Thus the function of the heart is to circulate the blood, not to produce rhythmic sounds. For only the former, not the latter, explains why we have hearts.
So teleology is not banished from biology, but it is tied to real purposes or goals in only a metaphorical or vestigial sense. Water flows downhill. That makes it look as though it has a drive to seek lower altitudes. Similarly, when genes replicate, it looks like a "drive to reproduce". But that is no more a genuine drive, or a genuine goal, than the downward flow of water.
The actual goals and values of individual human agents obviously have roots in our biological nature; but humans are unique in both genome and experience, and can forge for themselves radically novel ranges of goals and values. The digital, abstracting tools of linguistic representation allow the individual to offer a more or less anarchic resistance to the impersonal destiny embodied in the vestigial teleology of natural selection.
In the living world, the only objective measure of "value" is the replication of genes. But why should I care that my genes replicate? I have my own plans, and I can argue about them. Argument, within and among ourselves, generates disagreement; disagreement produces a proliferation of individual human values; and thus language makes possible human rationality, irrationality, and the wondrous, chaotic multiplicity of conflicting human values.
Such is the main argument pursued in Why Think. Other aspects of our fraught aspiration to rationality are explored along the way. One question is whether rationality "scales up": Can groups, as well as individuals, be rational or irrational? If so, does group rationality automatically result from the sum of individually rational decisions? An elegant large-scale functionality sometimes emerges out of local, individual choices. Natural selection itself is an example; another is the well-known "invisible hand" of the market, providing automatic solutions to the problem of pricing. In other cases, however optimal outcomes are systematically unattainable on the basis individual rational actions. Assume that individual rationality consists in opting for what will benefit the agent regardless of others' choices; and assume that collective rationality is the attainment of the best outcome for all. Then imagine a dinner in which all diners share the cost equally. If each aims to get the best deal possible, they will try to eat and drink more than the average they must pay for. All will end up worse off, from overeating and overspending. This illustrates the problem known as the "Prisoner's Dilemma".
Our two-track mind affords us the possibility of transcending our biological nature. But it also makes us prone to systematic irrationality. Many examples point to the uneasy relation between our intuitive judgments and our capacity to manipulate explicit calculations in the currency of words. Among these are many common failures in reasoning, now increasingly well-known thanks to the work of Daniel Kahneman and others.
And as everyone knows, some failures of rationality are due to the influence of emotions, those Track 1 processes par excellence. Emotions tend to derail our estimates of risk, which are dependent on the way possible outcomes are presented. We tend to be particularly poor at the crucial task of applying probability to practical choices. These examples, among many others, bear witness to the fact that there is no prospect of defining, let alone attaining, ideal rationality.