A useful model of cognition

“A moment’s thought would have shown him. But a moment is a long time, and thought is a painful process.” - A. E. Housman

Having an intuition for the way your brain thinks can help you learn more quickly, avoid cognitive biases, and understand the reasoning of others. However, building this intuition is a challenge.

Because most of our ‘thought processes’ are subconscious, we have a limited ability to introspect about our own thinking. Thoughts and ideas float to the surface of our consciousness, but we often can’t explain how or why they emerge from the murky depths.

Despite the opaque nature of our minds, reflecting on the way our brains behave can help us use them more accurately and effectively. What can we observe about the mechanisms that power our thoughts and reasoning?

Humans can recall a stunning amount of information almost instantly. Many people can access decades worth of memories, experiences, thoughts, and associations in less than a second – all that is required is a prompt (‘remember the time when…’). The same volume of accumulated thoughts can’t be accessed sequentially. Focused recall is very fast but thoughts aren’t ordered in the mind. It seems ideas need to be retrieved one by one, or at least via a linear chain of recollection.

In contrast to ‘reading’ information from the mind (as if from a disk), ‘writing’ new information seems to be an extremely slow process. Even when applying specialized techniques, repeated exposure is usually necessary before information can be reliably recalled. Although a memory has a moment of origin, it typically must be reflected on to be available permanently.

So our brains behave like a large cache with expensive writes. What can we do with this?

Slow writes mean that both learning and original thought are inherently slow processes. A consequence of this is that often you need to wait for them to complete. Sometimes I can tell when my brain is processing information ‘below the surface’ – in these situations I can be patient knowing that progress is being made without active focus. When a new write completes, the normal quick lookup becomes available (this is what I think of when people describe inspiration ‘striking’ them).

Fast lookups create a bias – our minds almost block out the possibility of an original thought if there is an existing relevant thought or idea in the cache. This makes sense since actual reasoning is cognitively expensive. This is also why mental models are so useful: they allow us to pattern-match new ideas in a cache-friendly way that also facilitates ‘lookups’ to related information.

Recognizing that cache lookups drive our thinking also casts the behavior of others in a new light. When we say that someone is being ‘irrational’ we presume a chain of reasoning from available information to their actions when in practice such reasoning is probably not taking place. Instead, it is more practical to recognize the cached thoughts that drive their choices.