My review of “Thinking, Fast and Slow” a Book by Daniel Kahneman
Kahneman has put together an impressive collection of studies and carefully explains them providing a difficult case to reject, that humans are doomed as far as “academic rationality” is concerned
“[…] humans are not well described by the rational agent model.”
Fortunately, for us humans, Kahneman steps into a trap. The fallacy of self omission. A good example to explain what that is, would be the claim that knowledge is impossible to human beings. The omission in that example is that a human is claiming to know something. That “something” being known is that humans can’t hold knowledge. The claimant has disqualified himself by virtue of the conclusion. He can’t simultaneously hold knowledge about humans and say that humans can’t attain valid knowledge.
So how does Kahneman accomplish this? Unfortunately, you must wait until the very end of the book in the final conclusion chapter. Let’s look at some quotes. Before we do that I’ll note that I listened to this book on audible and had to transcribe for these quotes.
K (short for Kahneman) starts with a colloquial definition of rationality.
“In everyday speech we call people reasonable if it is possible to reason with them. If their beliefs are generally intune with reality and if their preferences are inline with their interests and their values. The word rational conveys an image of greater deliberation, more calculation and less warmth. But in common language a rational person is certainly reasonable.”.
This section is fine for its purpose of explaining how the typical person views what it means to be rational. It’s circular but we kind of know what he means. This part isn’t the part worth attacking. K is not going to use this definition of rationality as the standard for his research. He will use it when he wants to convince the general population.
“For economists and decision theorists the adjective has an altogether different meaning. The only test of rationality is not whether a person’s beliefs and preferences are reasonable but whether they are internally consistent. A rational person can believe in ghosts so long as all her other beliefs are consistent with the existence of ghosts. A rational person can prefer being hated over being loved. So long as his preferences are consistent. Rationality is logical coherence, reasonable or not. Econs are rational by this definition but there is overwhelming evidence that humans can not be[…]. The definition of rationality as coherence is impossibly restrictive. It demands adherence to rules of logic that a finite mind is not able to implement. Reasonable people cannot be rational by that definition. But they should not be branded as irrational for that reason. Irrational is a strong word that connotes impulsivity, emotionality and a stubborn resistance to reasonable argument.”
Before I go too far I’d like to say that K makes it clear he isn’t suggesting that human choices are irrational. However, K is offering a very confusing combination of statements. So far I have not shown the contradiction because I want to draw distinctions first. In the second quote I provided we see K providing a technical definition for rationality that is coming from academic circles of economics and choice theory. I don’t want to get snagged on the fact that he finishes that section by collapsing back on the colloquial definition and also assuming the definition of “reasonable”. That snag misses the point I want to make.
So what’s my point? K must accept the academic definition because his life’s work depends on it. He wouldn’t have been able to publish his results if he used only the colloquial def. K is a human after all and as such is subject to the same limits (ie. “finite mind”) that all his test subjects are limited to and that he admits can’t implement the academic standard. So what then is the status of his academic results? Which standard are we to use to judge the claims in this book? If we choose the academic, we can’t accept that his work is valid since he has a finite mind. If we choose the colloquial, we can’t justify his published result with academia and poof goes his credibility. Here lies his logical error, the fallacy of self omission. By his standard he can’t validate the claims he’s making. He’s asking the reader to apply the colloquial definition in order to accept the results he’s produced. Well what good was the academic definition?
What is K supposed to do? He could have explicitly acknowledged the contradiction he finds himself in. Nope, he didn’t do that. He could have put forth a different definition of rationality. Nope, he didn’t do that either. We are left with a muddled mess.
K might be trying to hint and wink at us. Or he might be totally unaware. Or he believes the main point of his work was to highlight the frailty of human cognition (which he only does so far as we allow him to receive the status of having a valid body of work).
Overall, K makes the point that Humans don’t do statistical thinking well. Point taken K. In the process he reveals a deeply flawed foundation in academia and his own inability to deal with that problem.
