Cognitive Bias vs. Logical Fallacy

People sometimes confuse cognitive biases with logical fallacies, but the two are not the same. A logical fallacy stems from an error in a logical argument, while a cognitive bias is rooted in thought processing errors often arising from problems with memory, attention, attribution, and other mental mistakes.

During a discussion (no negative connotation), when you have to concede points, it's a good idea to show the audience why the other side's argument is not important or less important to the big picture of your argument. Acknowledging counterarguments contributes to ethos and makes the arguer seem more balanced and fair

List of fallacies

Analyzing inductive reasoning

Take the following syllogism:

  • P1: If each man had a definite set of rules of conduct by which he regulated his life he would be no better than a machine.
  • P2: However, there are no such rules.
  • C: Therefore, men cannot be machines.

This is a non-sequitur for 2 reasons:

  1. Premise 1 speaks nothing about possibility, but only about opined relativity.
    • in other words, nothing is posited about whether or not men can be machines. P1 simply states that "if we had these rules, we'd be no better".
  2. P1 ignores the possibility that men could be machines without such definite rules
  3. P2 is painting "rules" as a binary construct. It is making them seem as if they exist, or do not, rather than rightfully placing them on a continuum. P2 implies, "if we do not have a strict set of rules (which is what machines have), then things without those strict set of rules cannot be a machine."

Occam's Razor

The simple explanation is more likely to be true than to be complex

  • this is because complex explanations have by definition more moving parts (think of them as variables in a function). All it takes is one variable to be wrong in either explanation, and the whole theory no longer holds water
  • take 2 explanations that explain the same phenomenon: 1 simple and 1 complex. the first has 3 moving parts (3 variables) and the second has 30. if there was 99% confidence that each variable was trustworthy, then the first explanation would have only a 3% chance of being wrong, while the second would have a 26% chance

Hanlon's Razor

the explanation most likely to be right is the one that contains the least amount of intent behind it

  • you should not attribute to malice that which is more easily explained by stupidity/ignorance/mistakes
  • Hanlon's Razor is a counter to confirmation bias

The relationship between cognition and disfluency

When things are more arcane and complex to understand (or are just organized poorly, meaning more attention has to be paid in order to progress), comprehension rises. Also, people become more deliberate and analytical with their rational thought. When things are fluent as easy to understand and reason about, we default to our lazy auto-pilot brain which kind of skims by it without registering.

  • ex. Think of how we sing a song we've heard a thousand times— we don't even realize the words we are singing. This is because knowing a song carries with it a high degree of fluency.

Problem of Induction

  • "can inductive reasoning lead to knowledge?"
  • inductive reasoning presents issues, because:
    1. you may observe a large sample of something, and still make a false conclusion about it
      • ex. the existence of black swans
    2. you may believe that sequences of events that have "always" occurred will continue to occur indefinitely (Uniformity of Nature)
      • ex. the laws of physics will always hold
        • ex. the sun will always rise

The tendency to think linearly

in a primitive environment, process and result are closely tied therefore our emotional apparatus is designed for linear causality

  • ie. given 2 causally linked variables, more input of one variable will result in more output of the other variable
  • ex. building a bridge— more efforts will result in more work getting done
  • when things are not linear, it is much easier to become demoralized, since we expect a reward for our effort (all black swans are nonlinear)

Negative Empiricism

a series of corroborative facts is not evidence.

  • ex. turkey making inductive rationales as to why life is good and will continue to be so, since he has only observed life up until his unexpected death. before that, everything would falsely indicate that everything would continue to be good however, we can know which statement is wrong, but not necessarily which is correct.
  • ex. if I see a black swan, I can verify not all swans are white.
  • ex. finding a malignant tumor proves the existence of cancer, but not finding one doesnt prove the absense of cancer therefore, we can edge toward the truth by negative instances, rather than making verifications

it is misleading to build general truths from observed facts (and our body of knowledge doesnt grow from a series of confirmatory observations (like the Turkey's).

  • therefore, observations are asymmetrical in the sense that some observations can be used to confrim facts, while others cannot
  • this shows that we may have 1000 points of data that provides less value than a single point
  • "1000 observations cannot prove you right, but 1 observation can prove you wrong"
  • takeaway: be resistant and sceptical of definitive truths
  • takeaway: when you have a theory about something, start looking for observations that would prove it wrong (this is fighting confirmation bias in a direct way)
  • there is no such thing as corroborative evidence


  1. Affirming the Consequent
  2. Begging the Question
  3. Domain Specificity
  4. Prosecutors Fallacy
  5. Reification