Where do emotions come from?
... our ancestors whose brain did not learn that a tiger is dangerous the very first time the new information was provided by observing a friend being eaten, most probably went extinct.
We discussed this at our “founders’ health” session based on a “hidden brain podcast” with the same title, covering the “bayesian brain theory”, featuring Lisa Feldman Barret.
The gist of the theory is the following:
The human brain creates a world model based on prior experiences. It uses this model in a predictive / generative way to reduce the cost of cognition by setting expectations about what will happen in a given context. In other words, this model provides snap predictions about outcomes. This is the same mechanism that leads into stereotyping, snap social judgements, or uncontrolled pursuit of yummy food! Evolutionarily speaking, the brain learns to follow certain patterns to improve the likelihood of finding food and mates, and avoiding danger.
Now, this model is not perfect and occasionally misfires. Those of us who have trained machine learning models know that it’s unlikely to change a whole model based on just one counter example. Brain is the same! it’s unlikely to change the first time something “surprising” happens. But what if that surprising thing is very dangerous or very beneficial. That is where emotions come in!
According to the Bayesian Brain Theory, the brain uses the error signal (difference between its prediction and the real outcome) to update its belief. Except that to increase adaptability, the chemicals and electric signals involved in emotions amplify the error signal, perhaps via increasing the neuroplasticity momentarily. In other words, our ancestors whose brain did not learn that a tiger is dangerous the very first time the new information was provided by observing a friend being eaten, most probably went extinct.
Summary: emotions intensify our ability to learn!
Of course, the above is an oversimplification of the complex processes going on in our brains and after all just a model, but nevertheless, I think, a pretty handy and interesting model. One area that is pretty unclear to me is emotions that are tied to sensory inputs, for example the feeling of calm when you hug a loved one, or when you listen to your favorite music, the delight from the bite of the favorite food. What we concluded is that most probably that’s what leads into the emotional immune system: the process that blunt happiness from too much of a good thing, or the sadness from too much of a bad thing. In other words, those kinds of emotions, perhaps, reinforce a positive behavior, but beyond a certain point, it becomes predictable for the brain and reduces its intensity in favor of encouraging us to explore other options.
The theory makes a great deal of sense to me in the social context though. If my brain thinks you’d behave respectfully but you don’t, then I’d be mad; or if my brain predicts that people like my writing, but they don’t, then I’d feel sad! Etc. This is probably the mechanism that results in social acceptance, belonging, and other in-group behaviors and feelings.
What’s the actionable insight though? Thinking about automatic cognition (intuition) vs deliberate cognition becomes very important here: acknowledging that feelings are amplified data points that are trying to help us learn from a situation faster, and mechanistically that’s all they are there to do; nothing else! In other words, it’s us, not them.
When I feel a feeling, it’s my brain that is predicting incorrectly and noticing the difference, not anything external. So, thinking deliberately about why my brain predicts in a certain way, and what it is that it’s trying to help me learn can go quite far. Let’s say someone says something that offends me. My brain predicted respect and perceived an insult instead, and therefore I’m mad. Now there are so many things that I can deliberate on; like why did it feel insulting? Why do I feel insecure about that thing? Is that a person I have to be more careful about given their general pattern of behavior?
