In the first four posts in this series, the relationship between memory/imagination and consciousness was the focus. The idea was that there is a structure at the heart of consciousness that is where the experiences for memory are put together, where recalled memories are housed and where imaginings are created. But there seem to be other functions that happen in this structure. One is a slight projection into the future, a prediction – just enough of a forward projection to make up for the lag caused by the work of creating the conscious moment. At the moment of awareness, the incoming sensory data is mirrored by the projection and should be the same if the projection is accurate. Any difference is an error that must be corrected.
What we have is a loop between action and perception. People often start with the perception but it is just as reasonable to start with the action. I plan an action and part of that plan is a idea of how the world changes during the action. Then I do the action and along the course of the action there is a monitor on how well it is going the comparison of the conscious ‘now’ and the sensory input. Deviations are used to fine-turn the course of the action.
Or we can have an idea of what is in the world around us. We can make a projection and compare it with our senses. If there is a discrepancy we know that our idea of the world is wrong and we should change our perception of what is in the world or react to it. Or… we can speed up the picture and it is not a loop in either direction but a meeting of the sensory half and the motor half of the brain with anything and everything tweaked until there is a perfect fit with reality. The aim is to have little tweaks but no big surprises.
Actually the perfect fit with reality is an illusion. There are many changes in the world that we just do not notice. The system is (as most biological systems are) as good as it has to be but not perfect. This is not the sort of prediction that a computer might do, with a lot of code, complicated equations and great effort. Instead it is probably done with a large number of small heuristic tricks of the sort that give us visual illusions. There is little doubt that the small (in the region of 250 msec) forward projection is done but very little knowledge of the mechanism. It is a feature of our awareness but the doing of it is hidden from us.
An advantage of having the prediction is that it would highlight conflict between to motor programs. Suppose we are running and bouncing a ball at the same time. The two motor programs are working smoothly together. Then a dog runs into our path intent on having the ball (all balls belong rightfully to every dog). Both motor programs must be quickly adjusted but what if the adjustments conflict and a particular muscle is going to be given to different signals. The theory goes that the prediction available in consciousness will avoid these conflicts. If this does happen, it would be a valuable function.
I should mention what I have taken to be a semantic argument between people (see here) who envisage the projection as a dynamic control operation by the motor side of the brain – and on the sensory side, people who envisage a prediction using cognitive calculations on a stimulation. I first encountered this when I ran across someone denying that the conscious experience is a slight prediction. On careful reading, it was the word ‘prediction’ that was the problem because the writer associated it with a particular type of prediction and used ‘prospective control’ for another type of projection into a future time. I really do not have a method that I favour but I believe that the prediction is based on both sensory and motor information and very probably is an assortment of methods.
So we can give as possible functions of consciousness, the housing in the consciousness structure of a predictive/prospective/projective ‘now’ that covers an awkward lag in awareness, monitors the accuracy of sensory perception, monitors the accuracy of motor control, and perhaps gives warning of conflicting motor commands before they are sent.