The Knowledge Illusion part 6

This post is on the book The Knowledge Illusion by Steven Sloman (cognitive scientist and professor at Brown University) and Phillip Fernbach (cognitive scientist and professor of marketing at Colorado's Leeds School of business).

This post is the sixth in a series of sixteen that address The Knowledge Illusion and unless otherwise noted all quotes are from The Knowledge Illusion. I recommend reading all sixteen posts in order.

I have written on numerous other books on psychology, social psychology, critical thinking, cognitive dissonance theory and related topics already but discovered this one and feel it plays a complimentary and very needed role. It helps to explain a huge number of "hows" and "whys" regarding the other subjects I mentioned, all of the subjects.


In chapter three, How We Think, the authors of The Knowledge Illusion take on that topic. Steve Sloman has a dog Cassie. Steve pointed out that he and Cassie have different levels of intelligence. When Cassie wants food she goes to her bowl, where her food is given to her. She may wait quite a while. Steve goes to his wife who gives him food.

This is the difference between having an association with the food and knowing the cause of the food being given. As a result Steve isn't sitting by a bowl hoping for food for hours, I hope.

Ivan Pavlov did research on dogs salivating at the sight of food, or smell, and the possibility that a dog could associate other stimuli like a ringing bell with food and the dogs could come to salivate when they hear the dinner bell, even with no food present.

Pavlov won a Nobel prize in 1904 and his associationist theories influenced behaviorist ideas that ruled psychology through the first half of the twentieth century.

In the 1950s psychologist John Garcia did experiments with rats and paired stimuli and found flaws in the ideas from Pavlov.

He found rats could associate some stimuli but not others. They could associate a noisy flashing light with an electric shock they can associate drinking sweetened water with a stomach ache but they could not reverse the association and associate a stomach ache with a noisy flashing light and the sweetened water with an electric shock. This showed a causal link has to be part of the association.

So even rats and dogs are causal thinkers to a degree. Associations have to make a kind of sense as being causally possible.

We are primed to think of how things work in a framework based on cause and effect. We know lighting a match makes fire. And we know how a causal relationship can be disturbed or interrupted. If the match is wet or pressed too softly or quickly it may not light and we know this.

We know hundreds of things like this. Thinking causally is first nature to us. We know yelling at someone who is too far away won't get the effect of yelling at them close enough to hear us. We know how to cause effects and why they won't occur if things are out of sequence or a part is missing or too weak.

Other kinds of reason are hard for us. Figuring out cube roots of four digit numbers, quantum mechanics and lots of other things are difficult and may go against out intuition.

We still suffer from the knowledge illusion but we are built to see causes. We think of A then B as a routine action constantly.

If I want to get in the building I need to open the door, If I want some food I need to pick something from the menu and wait in line then order it and pay then wait. If I take it after that I can eat it or take it and go.

Sometimes we get these things wrong and think that if A then always B or If A then B is true then if B then A is also true. Sometimes we get these right.

If A then B could involve some cause that always makes  an effect like if it rains the ground will be wet but the ground could be wet without rain having occurred. Water could leak from a pipe or kids could play with a hose.

Much of the time we know automatically there are multiple reasons for a result. It is often the case that we consider other causes for effects.


 "We may make inferences all the time, but those inferences are not based on textbook logic; they are based on the logic of causality.

Just as people don't think only associatively (as Pavlov thought we do), people do not reason via logical deduction. We reason by causal analysis. " (Page 56)

That is a lot to take on. Let's look at some definitions to see what is what here.

Deductive reasoning is something we can define.

Here is a few excerpts from Wikipedia on deductive reasoning.

 "Deductive reasoning, also deductive logic, is the process of reasoning from one or more statements (premises) to reach a logically certain conclusion."


Deductive reasoning goes in the same direction as that of the conditionals, and links premises with conclusions. If all premises are true, the terms are clear, and the rules of deductive logic are followed, then the conclusion reached is necessarily true.

An example of an argument using deductive reasoning:

1. All men are mortal. (First premise)
2. Socrates is a man. (Second premise)
3. Therefore, Socrates is mortal. (Conclusion)

The first premise states that all objects classified as "men" have the attribute "mortal." The second premise states that "Socrates" is classified as a "man" – a member of the set "men." The conclusion then states that "Socrates" must be "mortal" because he inherits this attribute from his classification as a "man." End quote

So deductive reasoning starts with premises, ideas assumed to true. These are general statements usually that describe a broad category at first then describe a quality of the category then move to place some one or something in the category and conclude by saying the particular person or thing has the quality.

There are lots of variations and rules for this in formal logic but that is the simplest explanation. When people talk about a detective using deduction they are talking about the detective putting things like people in the category of suspect and eliminating them with premises of when a crime likely occurred and how and eliminating suspects who have alibis that are believed or lack ability, like a suspect who is disabled and unable to stand when it is accepted that a killer stood up strangled a victim and carried the body. 

The important thing to understand is that is not how we usually think. It takes a lot of work and practice to even start to follow the rules for formal deduction. I have whole books on it and it certainly isn't first nature for anyone I know. 

Causal thinking is our regular routine. We think of what to do to get results all the time and adjust to changing circumstances. I have to sometimes adjust my route to get to work or home due to construction, accidents and bad weather. It is a breeze compared to formal logic. 

Our long term planning also is centered on causal thinking. We think of consequences way down the road. That is part of why we do many things like saving toys for grandchildren who may not be born for years or set aside money for retirement.

We even reason causally about the minds others. If your spouse isn't talking to you you might wonder if you said or did something that upset them, or forgot an anniversary or birthday or if they are mad because of something you did or said that you are not remembering.

You act like a detective and think through clues. Clues of causation.

We engage in this analysis at every social encounter. We decide why people are behaving the way they are based on the clues we collect. 

The authors noted that we reason far better from causes to effects than from effects backwards to causes. It is easier for a doctor to predict a symptom a particular illness may cause far more often than for the doctor to determine the cause of a symptom, especially a symptom that many illnesses cause. 

Perhaps only humans reason backwards from effects to causes. In reasoning forward we simulate the results of actions. We think up little movies in our minds of the results of actions. 

The authors point out a crucial error we make when using predictive reasoning. They pointed out how when we reason from cause to effect we create a mental simulation and leave no room for the possibility of alternative causes.

That is a big error. If we find an effect like a burned down house we are aware of several possibilities. Someone could have set the fire, lightning could have hit it, chemicals could have reacted, even more unlikely possibilities exist. But if we saw someone the day before playing with fireworks we might run our simulation and be sure they set the fire. 

Storytelling is our way of making sense of the world in causal terms. We have many stories that link causes to effects. If we believe in values we find stories that demonstrate the merits of those values. 

We even think about the effects of causes that don't exist. We think of stories of "what if" something unlikely or impossible happened, what would be the results ? We have science fiction and horror and fantasies of being rich or famous or handsome or other things that may never happen. 

Some people come up with brilliant ideas when imagining the impossible. Einstein did thought experiments and imagined riding light and what he would see and that helped him to develop relativity and special relativity. 

Stories are extremely important to understanding who we are and how we think. George Lakoff has extensively on how we think and understand the world in metaphor, stories on a deeply subconscious level. 

Stories create associations of the good with the good. Good people do good things and bad people do bad things in stories. Bad values have bad results and good values have good results in stories. 

The sacred is sacred because of stories and the absurd is absurd because of stories as are the disgusting and the ordinary. Stories make the world make sense and give things their meaning. 

We need to be able to think well enough to understand stories to listen to them and to tell them. Some psychologists think our cognitive development is so far past other mammals even primates for our complex social development including telling stories. An author must understand characters, the characters thoughts and feelings and actions and reactions. And understand the way to tell the reader and that telling them certain things certain ways is more likely to be enjoyed by the reader as the reader imagines the story as the author presents it. That takes a lot of causal reasoning.