Yeah, I'd quite like to talk about, if it wouldn’t be a bore, you know, sort of the central theme that I’ve worked on really, really thought about, really all my academic life and that is what can be called bottom up and top down. This is a shorthand phrase but it’s quite simple. What it really means is, say, in vision, it applies to all the senses, but in vision you’ve got signals coming into the brain from the eye down the million fibres of the optic nerve, if you like, into the brain, from the real world outside you, and some theories of perception say that’s it, that’s all there is. It’s just a whole load of inputs from the senses, touch, vision, hearing, the brain analyses, responds to these sensations represented by what are called action potentials, that is, signals in the nerve, rather like Morse code, dots and dashes of Morse code, and then somehow the brain reads these dots and dashes, if you like, these little signals, but the thing is, how does it read the signals and make sense of them unless it’s got knowledge of what the signals might mean? For example, if you get a telegram and it says my- your aunt, arriving Thursday. You’ve got to know what an aunt is, you’ve got to know what arriving is, you’ve got to know what Thursday is, otherwise it’s just a load of dots and dashes of Morse code and completely meaningless. You always have to relate the information received to a load of knowledge in store, if you like. Now, I think the same is true in vision and you can call this bottom up, well, you can call this signals coming in and then also knowledge interpreting the signals so you’ve really got top down processing from knowledge, if you like, bottom up signals, coming in from the eyes. The real question is, the ratio of how much do we see things in terms of the available information from the world at the present moment? How much is it dependent on information from the past to interpret those signals and then when the knowledge is wrong or not applicable or might be out of date, how far do you see things wrongly? How far are illusions and errors due not to the signalling going wrong in the eye, if you like, but to the knowledge not being appropriate or not being used correctly? Now, ophthalmologists are really concerned, you know, with getting glasses, spectacles, for your eyes to correct the image so that you get good images but actually, you know, if the knowledge is just as important or even more important, you need to consult somebody who knows about the brain, somebody who knows about the mind, just as much as an ophthalmologist when you’ve got problems of seeing. So I’ve really spent most of my life actually on just this and when I started, and this is really quite amazing, there was no physiological knowledge that knowledge that- that top down existed in the brain. We’ve guessed that it must but there were no known fibres from the known parts of the brain which handle knowledge, rules, cognition, into the visual system. It was all thought to come from the eye into the brain directly, and then years later, only about 12 years ago, it was discovered there were actually more nerve fibres coming down into the visual system actually to relay stations just behind the eyes called lateral geniculate bodies, there are more fibres coming down from the knowledge parts of the brain, than there are from the eyes into the visual system. So it’s quite incredible, so it moved from everything is from the eyes to, golly, there’s actually more from the higher levels of the brain, the cortex, than from the eyes, the vision. Well, in a way, do you know, we guessed that from our experiments on illusions years, years before the anatomy was discovered, but there was a bit of a sort of tension here between the psychologists working on the stuff with illusions and so on, and the physiologists or anatomists looking at the brain structures and suddenly the two have come together and we now know that actually the experiments on illusions and things were actually right. They showed that there’s a huge contribution from knowledge into interpreting the signal for seeing and we now know that the anatomical pathways actually do exist for this influence. It’s really quite a nice case, and you often get this in science. You get different disciplines developing their bit and it doesn’t fit what other disciplines are saying, and then get this sort of conflict and you go to a conference and you can really get a lot of angst about this, it takes a lot of beers to sort it out, and then suddenly somebody gets a breakthrough and they find the missing link and then it reconciles apparently different views. It’s very much how science works actually. So you often get sort of competing camps, you get schools of thought, and the case for vision, there’s a very famous American psychologist called J.J. Gibson, who thought that perception is simply direct, that you just got information coming in and that’s it and I never agreed with that through having worked with illusions, and I always thought it was highly indirect, that perceptions about hypothesis just as indirect as hypothesis in science to the world outside and this sort of indirect view and the direct view have been a great controversy for years and years in thinking about perception. And, well, I think I was right, actually I still do.
