Questioning common sense

I had a chat on the phone with the  neuroscientist Karl Friston the other day. Not only was it illuminating, Professor Friston is a gentle, courteous and friendly man. It was an absolute pleasure and privilige to talk to him.

I had a chat on the phone with the  neuroscientist Karl Friston the other day. Not only was it illuminating, Professor Friston is a gentle, courteous and friendly man. It was an absolute pleasure and privilige to talk to him.

As I have mentioned before (although after talking to Professor Friston I now fear I did not explain his 'free energy principle' very well), he has developed an overarching theory of how perception and action work in terms of brain function.

One of his conclusions is that perceptions are not caused in us passively. Rather, the brain continually contextualises information with a view to acting upon the world. For example, if someone is speaking to you in a noisy bar your brain strains to contextualise the words as meaning something because you want to understand and respond. Thus you are actively trying to predict what the person will say, so that as Friston puts it: ‘perception is enslaved by action.’ Given the contextualising nature of perception, if the conversation had taken place in late 2008 you would have been much more likely to hear ‘credit crunch’ rather than ‘credit brunch’.

Friston’s theory also says that the brain is like an onion. On the outside are the automatic sensory layers that respond unconsciously to the inflow of sensory data. The inner layers are more concerned with learning and memory, and ‘higher’ cognitive processing. At the very inner core is the pre-frontal cortex which can think about information sent from any part of the brain.

How does this work in practice? As you walk down the street your feet feel the ground and the outer layers of your brain predict where it will be when you take your next step (perception contextualises the world according to the actions the brain is executing at the time). If the ground is not where it is predicted to be your brain automatically generates a signal telling you this. The signal is really just an alert that there is a difference between the prediction your brain has made and what has actually occurred. The outer layers of the brain feed this signal to the inner layers. This might initiate only a habitual response that is automatic – like instinctively cushioning one’s fall when there is an unexpected dip in the ground. But say the ground is not where your brain predicts it to be for several steps and you know you are in an earthquake zone. Then the inner core of the brain contextualises the signals from the outer layers and you start to think about what to do.

So according to Friston’s theory (I hope I am getting this right), at the outermost layer completely automatic sensory processing takes place. At the next layer inwards habitual learning takes place – for example, when the door you push on won’t open you pull instead, but you don’t think much about this, habit simply kicks in. And at the innermost layer consciously controlled processing - thought and deliberation - occurs. So if the door you come across opens neither inwards nor outwards you think about trying to locate a key for it.

This means that perception results from a complex active process of contextualising sensory data, all with a view to pursuing some action or other. But that means that the common sense idea that the world passively causes perceptions in us would seem to be false. Rather, the way the outer layers of the brain contextualise sensory data can greatly affect an individual’s ability to learn from perception. If these layers create feelings of fear and anger, for example, or if the information they send is in turn contextualised by ingrained habits, then the potential for processing information will be restricted accordingly.

But Friston’s theory seems to imply also that the common sense assumption that action is like turning a light switch on and off is false. Action depends on how the brain processes information through feedback loops between its inner and outer layers. Consequently an individual’s ability to perform certain actions will depend on how these layers interact. If events in the world are contextualised in terms of, for example, fearful and aggressive emotions, then this will shape the range of actions an individual can perform. But also, because perception is contextualised in terms of the actions an individual intends to perform, if someone is used to acting aggressively and out of fear then many aspects of the world will be perceived as such just for this reason.

Rejecting common sense assumptions in light of neuroscience may have far-reaching and radical effects on the way we think about ourselves in all sorts of contexts. Whether people are ready for such shifts is something the Social Brain project is interested to try to assess.