Attention and Prediction

Posted: July 8, 2013 at 2:12 pm

After watching this video, I was inspired to think about attention in relation to the Integrative Theory that has been developed during this research project.

There is something very interesting in the way that attention is so strongly rooted to task demands, and begs the question of attention in the absence of those demands. We know that our eyes do not see what we perceive, but we may still perceive even when we are not engaged in a task. Perhaps the ongoing sensory information in the absence of task demands disappears, due to its complexity (by not being engaged in a task, we are not looking for anything in particular) and habituation, and that is what causes the onset of mind-wandering. Then there is the additional question: is there attention in dreaming and imagery? Some theories of imagery state that images are constructed in early visual cortex and then perceived the same as external perception. It is clear that our perception is more spatially stable than the information from our eyes, we don’t see our saccades, so the image we perceive is not the image in our early visual cortex, which is presumably more tied to the eyes and therefore more unstable. So when we imagine, do we produce an unstable image in the early cortex that is stabilized in later visual areas? Can we imagine moving images better in our periphery, and detailed colour images better in our fovea? If imagery is in the early VC, then the fovea and the periphery should effect imagery. Kosslyn’s studies have shown how things like field or view, smooth pursuit and resolution link perception and imagery, but it has also been shown that imagery is possible without the functional use of the early VC.

The main conclusion I have come to in the Integrative Theory (of mental imagery, perception, dreaming and mind-wandering) is prediction / simulation. Our mind-wanderings and dreams are an expected sequence of events. They are an exploitation of implicit learning that teaches us what we need to pay attention to and what we can ignore in the world. We should then think about perception in relation to prediction also. Beyond task demands it’s our expectations of what may happen next that drives our attention.

I would go so far as to say that we can’t actually perceive without the ability to predict. Prediction is one of the most powerful ways we can constrain the complexity of sensory information to make sense of it. I would further say that it takes a lot of embodied sensory learning in the world for our brains to build up these predictions that allow us to perceive with increasing detail and complexity. This complexity is both a function of the complexity of sensory information in the world, and equally, from our increasingly accurate predictive mechanisms.