Trained as an electrical engineer, I first approached the work of James Gibson in the late sixties, by reading his second major book "The senses considered as perceptual systems"¹. Those were the years when the extravagant hopes raised by Information Theory as a possible clue for understanding sensory processing, and the equally extravagant hopes raised by Cybernetics as a possible clue for understanding action were being exposed for what they are, just extravagant and misleading. Yet, in many quarters the notion that the connection between perception and action had to be construed as some form of algorithmic symbolic transformation implemented by dedicated hardware was still quite popular.