The Problem With Trying To Be Rational - Steven Pinker
精读 · 按主题梳理核心观点、论证与案例。
§01Why anyone needs a toolkit at all
The host opens with a confession that frames the whole episode: there was a period when he was convinced that one more blog post or one more definition of a cognitive bias would solve his problems, and then discovered it did not. His question is why he needs a glossary of mental models just to function. Pinker's answer is that the equipment is not missing but narrow — we reason well about causation, logical implication and probability when the problem is one we have handled all our lives and care about deeply, and poorly when it is novel or abstract.
His prescription is not a general theory but a set of named tools that must be available at the moment of use: the names of the fallacies and how to avoid them, plus the normative models for how one ought to reason. The point is practical rather than academic — you need to recognise that a decision is a sunk-cost case while you are making it, which requires the label to be retrievable rather than merely known.
§02Can you actually become more rational?
Kahneman's reported answer to Sam Harris — that decades of studying bias had not made him more rational — gets a qualified yes from Pinker, with an important caveat about the bias blind spot: everyone believes other people are biased and they are not, so the person least equipped to detect his own biases is himself. He then offers the empirical half of the case: people who are less susceptible to classic fallacies have better life outcomes on average — fewer accidents, less job loss, fewer breakups, and markedly less vulnerability to psychics and medical charlatans.
On intelligence he is careful. The correlation with rationality is real and far from perfect, so plenty of intelligent people hold fixed beliefs and refuse to update. The exception is where beliefs are sacred values of one's tribe or political ideology: under my-side bias, smart people are "just as vulnerable as less smart people." He attributes the framing to Keith Stanovich's The Bias That Divides Us and its deliberately cheeky rationality quotient.