IN TWO MINDS ABOUT MICROS (column)

Written by David Tebbutt, MicroScope 02/84 item 02 - scanned

Did you know that if the two halves of your brain were surgically separated you would effectively be left with two autonomous, independent 'minds'? Indeed, many people suffering severe epileptic seizures have voluntarily undergone precisely such an operation which has resulted in a dramatic reduction in both the severity and frequency of seizures. Not surprisingly, such 'split brain' patients have been the subject of much experimentation over the years and some interesting facts have emerged about the human brain and its operation. I can't help feeling that the results of this research have a direct bearing on the future direction of computing.

The majority of people have a very systematic 'left brain'. It does things like decoding speech, taking care of grammar, tackling logic problems and so on. It is the rational, constrained, realistic and organised side of the brain. This is the side that many of us in the computer industry use to excess. Systems analysis, system design, programming, project planning and control are all primarily left brain activities. This kind of thinking could be likened to the serial approach of Von Neumann style computers.

The right side of the brain tackles things in a far more general way. It is often, unfairly called the minor hemisphere because of its inability to communicate directly with the outside world. All words which describe the experiences of this side of the brain are created by the left brain after receipt of messages across the corpus callosum, which is a bundle of 50 million or so connections between the two halves of the brain. The right brain is responsible for the artistic, creative, intuitive aspects of our nature. Of course, it does come into play in the jobs mentioned above but it is more used by dancers, mystics, artists and musicians. It can handle masses of information at the same time but falls down on the detailed processing of that information. It is the side of the brain which can recognise a face or an object even when presented with a tiny fragment of information. In other words, the right brain has a sense of perspective.

To give a practical example of this, imagine you are riding a bicycle or driving a car. The view of your surroundings is continually changing, yet you rarely become confused. It seems improbable that your brain analyses every particle of information received by your eyes, yet you drive or ride along in perfect safety. Even allowing for the fact that the brain may process information coming in, most of which is irrelevant to the task in hand. As the part of the brain which takes the broad view, it would make sense for the right brain to be responsible for all the information coming in from our environment. But it probably does this in a very fuzzy way, by sensing significant changes between successive images received. It is reasonable to assume that changes of perspective don't count, only material changes in the content of the information. Now we still need parallel processing because our brains are fairly slow, but the task is far more manageable. Let's suppose that a car door suddenly opens ahead of you. Your right brain would sense this major deviation from its expectations and quickly trigger the analytical left brain to sort out the situation. Having subcontracted the problem to the left brain, the right brain can return to its overall monitoring role. The left brain meanwhile can analyse the significance of the car door, anticipate what might happen next and galvanise the appropriate reflexes.

Other examples of the differences between the activities of the two hemispheres abound. How many times have you played a game such as snooker, tennis or darts and done jolly well when you've been fairly relaxed and indifferent? And how many times do you really foul up when you start thinking about what you're doing? It's like running clown an escalator or catching a ball. The slow right side of the brain just cannot process this type of information fast enough to be of any use to you. The minute you start thinking about what you're doing, you trip or drop the ball. On the other hand, it's no good having a wild guess at the answer to tricky mathematical equations because the right side of the brain will have a go but usually let you down.

Now let's look at computers. For a long time now, we have stuck to serial processing methods and they have brought us a long way. Many expert systems have been built using these conventional computers. Now, parallel processing has been recognised as a feasible approach and many people are looking to this method to solve our computing problems. (Fortunately for Britain, Inmos is in the forefront of this technology with its Transputer and Occam programming language.) Some people have persuaded themselves that parallel is best or serial is best, when in fact we might benefit from a combination of the two.

It's odd, isn't it, that we've taken so long to come to two computing approaches: one which simulates the right brain and another which simulates the left? This method seems to work pretty well for brains so I can't see why it shouldn't work for computers. We could use parallel processing power to handle the masses of input from the computer's sense 'organs' and pass information regarding unexpected patterns across to a connected serial computing system for more detailed analysis. The computers will still massively outperform us intellectually because of their infallible memories and rapid access to global knowledge bases.

We humans can then perhaps return to a less hurried way of life in which our priorities revert to the more enjoyable right brain activities such as food, sex, art and music.