6. The Origin of Cognitive Biases

Towards a unifying theory

Previous chapter: The Next Paradigm Shift


Research which began in the early 1970s has uncovered a large number of universal biases in human cognition. A Wikipedia page lists almost two hundred of these ‘cognitive biases’, and the list keeps growing.

The two most discussed cognitive biases are confirmation bias — our bias towards interpreting, seeking, and remembering information in a way that confirms, or helps to confirm, what we currently believe — and availability bias — our bias towards forming judgements on the basis of whichever relevant information happens to be most readily available to our mind.

They’re the most discussed cognitive biases because they, uniquely, permeate all reasoning, whereas other cognitive biases only occur in specific types of situations, however frequent those situations may be. For example, hindsight bias is our bias towards judging, with hindsight, that a past chain of events was objectively easily predictable.

There are many theories about why each cognitive bias exists, although there are only three types of such theories.

The first type of theory states that a particular cognitive bias is the product, or by-product, of a supposed ‘cognitive heuristic’. A cognitive heuristic is a cognitive shortcut that has become either hard-wired into our cognitive processes through evolution, or ingrained in those processes through learning. Cognitive heuristics supposedly develop to counteract our cognitive limitations, which are due to our finite amount of processing power, information and time.

The second type of theory states that a particular cognitive bias is simply a direct by-product of our cognitive limitations.

The third type of theory states that a particular cognitive bias is the result of one of our universal emotions affecting our cognitive processes. However, cognitive biases are so-called because they arise from the nature of the cognitive processes which underlie our reasoning. If cognitive biases were so-called simply because they result from something affecting our cognitive processes, then any kind of bias would be a cognitive bias. And although reasoning and the emotions are closely interconnected, they’re quite distinct aspects of our psychology — that is, the brain processes underlying our emotions aren’t part of our cognitive processes. Therefore, although our emotions can indeed affect our cognitive processes, any resulting biases are actually emotional, not cognitive, biases.

Of course, it’s theoretically possible that some, or even all, apparent cognitive biases are actually emotional biases. But the next few chapters suggest that apparent cognitive biases indeed arise solely from the nature of our cognitive processes.

A popular textbook on human reasoning states: ‘We have no good reason to expect a single account to apply to all [cognitive] biases that have been discovered’. However, an unawareness of such a reason isn’t reason to expect there not to be such a unifying explanation. Also, given that all cognitive biases are obviously unified by being cognitive biases, a unifying explanation is a reasonable possibility.

A common theme in the history of science is the development of theories that can explain, and therefore unify, different phenomena that were previously supposedly explained independently of each other by distinct theories. For example, in the 19th century, the physicist James Clerk Maxwell showed that electricity and magnetism are actually different aspects of the same phenomenon — electromagnetism — and that many different forms of radiation —including light, radio, X-ray, microwave, infrared and ultraviolet — are simply different wavelengths of electromagnetic radiation. And, in the same century, the biologist Charles Darwin showed that the myriad different species of plants and animals that exist today are all the product of the same process — evolution by natural selection — and are the end points of ‘branches’ of a single ‘tree of life’.

Regarding cognitive biases, explanations involving cognitive heuristics have been dominant since the first cognitive biases were discovered in the early 1970s. However, the next few chapters show that several prominent cognitive biases — including confirmation bias and availability bias — are due to neither cognitive heuristics, nor our cognitive limitations, and that credulism (How Beliefs Form), the certainty of belief (Belief Is Certainty), and the speed of the brain, together have the potential to explain, and therefore unify, all cognitive biases.

Next chapter: The Logical Necessity of Confirmation Bias