Thinking Fast and Slow Chapter 3: The Lazy Controller

Our system 2 tends to be a lazy guy. When prompted with an intuitive answer from system 1 to a problem, the default response from system 2 is to accept it.

In a series of relevant experiments, like stating how much a ball and a bat will cost if the bat cost 1 dollar more than the bat, and together they cost 1.10 dollars, we are shown how we tend to give the inuitive -and false- answer (the ball costs 10 cents) instead the right one (the ball costs 5 cents).
The main reason behind this situation is self control depletion. To pay attention implies some cognitive costs, and we need to exert our self control to keep ourselves focuses on a problem, instead of doing something else. But we seem to have only a certain amount of self control, and if it used in one task we won’t have any more energy to do the next task, as it is shown in several experiments, like the one in which the subjects have to resist the temptation to eat some delicious chocolate and then make a poorer performance solving a problem than the test group.
This seems to be literaly an energy problem as experiments have found a close correlation between lack of glucose and lack of self control.
One of these experiments turns out to be particularly gloomy. Judges that have to decide whether to give parole to a person in jail tend to give a lot more paroles after lunch, when their glucose levels were restored. It is not a beautiful picture for rationality to be shown how such an important thing as the freedom of a person depends on whether the judge has eaten lunch or not…

One can find in this chapter a very interesting discussion of what intelligence really entails, and Kahneman argues how besides our ability to make calculations and follow an argument, intelligence also implies the ability to bring to memory relevant data as well as the rationality to realize that an intuitive answer might be false and devote some cognitive effort to solve it properly.

To me the more relevant result of the chapter is this: when system 1 already considers that it has the right answer, it is very difficult to make the mandatory mental procedures to check whether this intuition is right or not. When we accept the conclusion as true, the argument turns out to be convicing even if it just a fallacy.

So far the content is convincing me. I have, however, one difficulty with the roses experiment:

All roses are flowers
Some flowers decay
ergo: some roses decay

This silogism is incorrect, and Kahneman presents the fact that most people will say it is true as an evidence of system 2 being lazy. I have to disagree in this case. In my opinion, the problem here is that our brains do not analyze logical problems using a formal system but a material one.
Let me revise briefly the famous experiment about formal conditionals. People were invited to solve the following problem:
There are three cards on a table: one depicts a circle, the second a square and the third a triangle. Next the following rule is presented:
If there is a circle in one side, then there will be a square in the other side
And subjects are invited to rise the minimum amount of cards necesary to check if the rule is true or not.
Everybody rises the card with the circle, which is correct, but most of them also rises the card with the square, which is not needed at all. Whether there is a circle or not the rule still holds. instead, very few people rises the triangle, which is mandatory. If there were a circle in the other side the rule would have been proved false.
However, if you do the experiment with meaingful objects and rules (like say, people keeping or not keeping a promise) almost everybody solves the problem.

So, my suspicion of how the roses experiment could be pointing to another direction. The reason we fail and see the silogism as correct is because we know that all flowers decay and roses decay as well, and we treat logic from a material point of view, and we see the argument as correct.

4 thoughts on “Thinking Fast and Slow Chapter 3: The Lazy Controller

  1. Things begin to get really interesting in this chapter. A lot of interesting ideas appear in it. Among them,

    * Self control is an effort consuming task of system 2 that competes for resources with all other system 2 tasks.

    * Exertion of self control produces “ego depletion” that reduces posterior efficiency in self control.

    * A lot of thought biases are not the result of flawed rational systems but are simply the result of not turning on the rational system because of laziness.

    This is something that I have always observed with curiosity how intelligence seems to be so much domain dependent. That is, very brilliant reasoners on some issues, do really poorly on other issues. And this gets very well explained by the laziness approach.

    * Rationality, which has more to do with the systematic discipline in checking our thoughts, may be much more relevant for intellectual performance than IQ measured intelligence which reflects our potential capabilities to algorithmically solve problems.

  2. I see your point with the roses thing but I think that you have to bring it into the system 1/ system 2 point of view.

    The effect that you comment, how easily we solve logical problems when they are expressed with friendly contents, reflects the way system 1 intuitively solves in an automatic way logical problems. And it doesn’t work for abstract contents. For processing them, you need to turn on system 2 and apply hard and painful explicit logic.

    And I think that what Kahneman wants to say is precisely that, that people often do not turn on system 2 when resolving the roses problem and follow instead the answer of system 1 that happens to be wrong because of the reasons that you explain.

    • Yes, you are right Carlos. What I described can be easily reconsidered as a more general system 1 versus system 2 issue. Kahneman is developing step by step a comprehensive analysis on how we think and well it is not that flattering to us, humans

  3. My own impression is that responsens to experiments such as the rose depend not only on logic, but also on factual knowledge about the world. In a town where, say, 99% of the flowers people have ever seen are roses answers would be different than in a town were a rose is a rarity. In that sense, system 2 is not only rational, but also keeps ordered register of our knowledge… Hence itaps energy costs.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s