It is common to say 'I can't believe my eyes' when surprised by what we see. Recent scientific evidence suggests that we have a right to be sceptical and that what we see depends in no small part on what we expect to see.
It is normal to think of vision as beginning with the formation of an image on the back of the eye, which in turn stimulates a cascade of nerve impulses sending signals deep into the brain. It is in the brain's visual cortex that these signals are interpreted. Signals in the visual cortex also travel in the opposite "feedback" direction but much less in known about their function.
A recent paper in the journal Current Biology by graduate students Tamara Watson and Joel Pearson and their supervisor Dr Colin Clifford at Sydney University's School of Psychology, suggests that these feedback signals carry information about what we expect to see and that they act to constrain our interpretation of the incoming visual information.
'The separation between our eyes gives us two slightly different views of the world. Ordinarily, our brains fuse these two views to add depth to our visual world. However, if the two eyes' images are so different that they cannot be fused then we experience "binocular rivalry". During binocular rivalry, one eye's image is perceived and the other suppressed,' explained Dr Clifford.
'Every few seconds, perception switches spontaneously between the two images. While binocular rivalry is rarely encountered in the normal visual environment, it provides a useful means of probing the workings of the visual parts of our brain: although the visual stimulus is artificial, the brain is functioning in its usual way.'
Watson, Pearson and Clifford used binocular rivalry to demonstrate the importance of feedback in our interpretation of the visual image. For two images to generate binocular rivalry they have to provide conflicting evidence about what is present in any given part of the visual world. To ensure that rivalry was not generated until a late stage of visual processing, the experimenters used images of walking human figures visible only via lights placed on their joints. This sparse visual stimulus has long been known to provide a compelling impression of a walking person.
'When walking figures were presented to the two eyes, perception alternated between first one figure and then the other. When the same moving dots were scrambled in position, rivalry was abolished. Although our brain does not recognize walking figures until long after the point at which information from the two eyes is combined, these results show that once a walker is recognized it can cause dominance of signals from one eye and suppression of signals from the other,' said Dr Clifford.
The Sydney team showed there is strong evidence that the brain's expectation that there should be a walking figure between the dots is somehow transmitted to the earliest level of the visual cortex where signals from the two eyes first interact.
Dr Clifford concludes that even if our expectations can affect what we see, vision is still our most valuable asset for sensing the world around us. However, he says that 'it is as well to remember that the view it provides us might be rather more subjective than we would expect.'