NOW YOU SEE IT

Dick Pountain/13 March 2007/12:21/Idealog 152

It's fashionable nowadays to scoff at Bishop George Berkeley who denied that material objects exist unless they're being looked at (remember, that stuff about the tree falling in the forest when no-one's watching). Berkeley had an axe to grind, to demolish materialism and prove we only exist because God sees us. His argument nevertheless contains a kernel of truth, that all the evidence we have for the existence of an outside world lies inside our own heads in the form of 'sense impressions'. Philosophers been batting this one back and forth ever since Berkeley's day, and many highly respected names like Hume and Kant arrived at similar conclusions.

Now this column has been consistently sceptical about overly-easy comparisons between the human brain and the computer, but here's one way in which we really are like computers. A computer can only interact with the world outside it if provided with sensors to sample external events in an indirect way - a digicam that captures light reflected from object, a microphone that samples sound waves generated by movement, a touch-pad or mouse that samples pressure caused by contact. The computer stores all this evidence internally as some kind of bitmap in memory. We're constantly being reminded nowadays that we live in the 'Information Age', but in a profound sense we always have done - everything we see, hear, taste, feel, think, hope, wish, remember (in fact everything except what we *do*) is just information, stored for longer or shorter times in massively-connected sheets of neurons inside our brains that you might call 'brainmaps'.

Neurophysiology has made huge strides in recent years and the way such brainmaps are encoded is slowly becoming clearer. For example the part of a rat's brain that processes signals from its whiskers is indeed a map arranged in rows and columns. However unlike a digicam where each CCD cell corresponds to one static pixel, the rat's processor is dynamic - waves of activation spread across the map and it's their phase and timing that encode the data. We're also beginning to understand the 'mirror neuron' system that enables us to guess other people's intentions - we simulate their observed movements and expressions on a model running in our own brain and work back from that to the intention. Given such progress we may fairly soon be able to resolve philosophers' disputes about how much, if any, of our knowledge is innate and how much learned, by identifying the neuronal filters and pattern-recognizers that build brainmaps out of raw sense data. Kant's categories might indeed turn out to be brain circuits.

This all implies that our knowledge of an outside world is wholly symbolic: brainmaps are symbols for things out there that provoked them, just as the letter A is a symbol for a puff of air emitted from my throat. So, going back to old George, how can we prove there's anything real out there at all? He was right, we can't. Materialists like me believe, for convenience, in stuff called matter from which other, external, beings and objects are made, but we can never experience it directly precisely because we're made from it too. My eye can't see atoms because it's made of atoms and would only see itself. But surely nowadays we can see atoms through a scanning tunnelling electron microscope? Er, 'fraid not, we still see exactly what we've always seen, light rays - in this case ones emitted by the monitor connected to the scanning tunnelling electron microscope. Atoms don't really exist except as a concept and a word, another kind of brainmap. In truth we don't even believe in atoms any more because first we decided they're 'really' bunches of protons, neutrons and electrons; then those in turn became clumps of quarks and in another decade quarks may very well be imagined as knotted strings or something. All just brainmaps.

Fortunately we don't need to experience matter directly, nor prove its existence, in order to live. Natural selection has fine-tuned our own sensors - eyes, ears, noses, skin - so well that the samples they generate *nearly* always correspond to some chunk of the indefinable, inscrutable stuff which is all that's out there. If they didn't we'd never have been able to catch food, we'd have constantly fallen down holes and we wouldn't have survived to reproduce. That 'nearly' is significant though. Perceptions, memories, thoughts and imaginings are all varieties of brainmap and although we can usually tell them apart, we don't always. Occasionally we hallucinate, or see what we expect or remember rather than what's there now - any lawyer or policeman knows that. Not all brainmaps correspond to external lumps of matter.

Language doesn't help because English words like 'be' and 'exist' don't capture the distinction. Debates between Dawkinians and religious fundamentalists are so seldom productive because God exists neither more nor less than the atom does, both are just brainmap/symbols. There is a real difference between science and religion though, because science continually extends that common sense evolution has equipped us with by testing its symbols against the world and throwing away ones that don't work. From atoms, to protons, to quarks - none of them exists but each enables us to make better predictions, good enough to make aeroplanes and paracetamol and computers.