Visual learning and control

[From Rupert Young (2016.07.14 13.00)]

These (rather gruesome) kitten experiments were mentioned briefly on CSG years ago, but I thought I'd bring them up again with reference to the role of control in reorganisation.

The first concerns kittens that were brought up in environments that only had either vertical or horizontal lines. The developed kittens brought up in the vertical environment can only see vertical things and are blind to horizontal things, and vice versa.

I guess in PCT terms perceptual functions have developed that could be seen as feature detectors for particular lines. However, this experiment doesn't give any indication of control being necessary. Could the perceptual systems of the kitten still reorganise, to some extent, while being a passive viewer of the environment?

The second experiment describes a set up where one kitten could roam, more or less, freely and another was in a harness connected to the first kitten and, although it could not initiate movement it moved in correspondence with the first kitten. The result was that the visual system of the immobile kitten did not develop properly; the mobile kitten was fine.

This would suggest that control plays an important role in reorganisation. But, any thoughts about the detail of what is going on here? Usually when we think of reorganisation due to error there are established reference goals, but in this case it seems that there couldn't be existing visual reference goals as the vision system starts undeveloped. What would be the references such that control brings about changes in the visual perceptual input functions?