Читать книгу Friends and Enemies: Our Need to Love and Hate - Paula Nicolson, Dorothy Rowe, Dorothy Rowe - Страница 8

The brain, it seems, is the most complex object on this planet and exactly how it works no one knows. There are lots of ways of examining the brain – looking at brain tissue, planting electrodes in an animal’s brain, using X-rays and various scanning techniques – and lots of interesting things have been discovered, but the great problem is to construct a suitable model of the brain. Our brains think in a peculiar way. To understand anything we have to decide what it is like. We have to find something from our past experience which is similar to what we are examining. If you have to describe a pizza to someone who’s never seen one you might start by saying, ‘It’s something like a pie but without a top.’ The trouble with the brain is that there is not anything like it. It operates with various chemicals, but not in the way our heart and lungs do. It uses electrical impulses, but not like any piece of electrical equipment that has been devised. People who love computers often liken it to a computer and there are Artificial Intelligence experts who believe that if they make a sufficiently large and clever computer it will turn into a conscious brain, but such a view is based on a supreme ignorance of how human beings actually operate. Models of the brain have now become a growth industry in science and philosophy but nothing satisfactory has yet been devised.

However, one way of thinking about the brain seems to be full of possibilities. Back in 1949 the psychologist Donald Hebb suggested a model for understanding the functions of the neurons in the brain. It seemed that the more two neurons communicated with one another the easier communication became. The neurons seemed to set up a relationship which Hebb called a neural net.

This has proved to be a most useful way of thinking about the brain’s activity, especially with the development of scanning techniques like positron emission tomography (PET), nuclear magnetic resonance (NMR) and magnetencephalography (MEG). Because human subjects can tell researchers what they are doing – looking at something, daydreaming, silently doing sums or reciting a poem – while they are being scanned many interesting relationships between thinking and brain activity have been observed.

For instance, Petry and Meyer showed how when we are perceiving the world around us and when we are imagining or dreaming we are using much the same regions of our brain.¹³ Whether we are lost in imagination, or asleep and dreaming, or looking at the world around us our brain is creating a picture in our heads. We have the task of deciding whether what we are experiencing is the world around us or an image inside us. Most of the time we get it right but sometimes we do not. If something utterly unexpected happens, be it a tragedy or the best of good fortune, the world around us can take on a dreamlike quality, and we turn to others for assurance that what seems to be happening is real. If, on the other hand, we feel totally overwhelmed by events and utterly powerless to make reasonable sense of what is happening to us we can lose the ability to distinguish between the voices of our thoughts in our head and the voices of people around us. Our thoughts seem to be the voices of unseen people around us. These voices might be friendly and comforting, but often they are criticizing us and urging us to commit terrible acts. When this happens to us we need real people around us who will not tell us that we are mad but rather will assure us that our voices are simply our thoughts and help us find ways of keeping our thoughts in order. Singing can quieten the voices, as can telling them loudly and firmly to shut up. If you have to do this in public a mobile phone can be very useful.

In research on the functioning of the brain there is still a big gap between scanning an active person’s brain and looking at the firing individual neurons. The images from scanning might not be showing the mental processes themselves, but only emissions from them, somewhat analogous to showing emissions from a car’s exhaust rather than the functioning of the internal combustion engine. However, Hebb’s neural net has provided the key to finding what might be a suitable model for understanding how the brain creates meaning. Susan Greenfield defined the model of a neuronal gestalt as ‘a highly variable aggregation of neurons that is temporarily recruited around a triggering epicentre. Not all neuronal assemblies are gestalts but all gestalts are neuronal assemblies.’¹⁴ A stimulus produces an epicentre of arousal in the brain and a group of neurons firing in a particular pattern form around the epicentre. If the stimulus is repeated the pattern of the epicentre and the neurons is repeated. Further repetitions of the stimulus turn what was a transient pattern into a gestalt, a schema where the whole is more than the sum of its parts. (A common example of a gestalt is that particular positioning of an oval shape, two dots and two straight lines which to our eyes is a picture of a face.)

Neurons communicate with one another by using a chemical transmitter to empty into the synapse – that is, the gap between the axon terminals of one cell and one of the dendrites of another cell, where the transmitter binds to a target molecule. In the transmitting neuron the electrical signal is put into a chemical code and, when it crosses the synapse, it is decoded back into an electrical signal in the neuron which received it.

In the womb a baby’s brain begins working long before it is fully formed. Neuronal patterns for pleasure and pain and certain kinds of sounds are laid down before birth (babies are born knowing the sound of their mother’s voice) but even so the baby’s brain still contains some 100 billion neurons whose connections are yet to stabilize into patterns. Trillions of synapses are available, but, unless the baby’s environment calls a synapse into use, it will be eliminated. The brain, said Susan Greenfield, operates on ‘a law of the jungle – use it or lose it’. A baby who is wrapped like a parcel and left to lie except when being fed will lose many more synapses than a baby who is little restrained and talked to and shown things. The more synapses you retain the brighter you’re likely to be. The more you use your synapses the more you retain them.

Susan Greenfield reported,

Although the brain is particularly impressionable whilst it has been developing, such adaptability does not cease, but merely lessens somewhat in maturity. It is actually possible to manipulate the environment and observe long-term changes in the brain. For example, adult rats were exposed to what is referred to as an ‘enriched environment’ where they had lots of toys, wheels, ladders and so forth to play with. In contrast, other rats were kept in an ordinary cage, where they received as much food and water as they wanted: it was just that they did not have anything to play with. However, when the brains of these two groups of rats were examined, then it was found that the number of connections in the brain had increased only in the animals in the enriched environment, not in those from the ordinary cages. It appears that sheer numbers of neurons are not so important as the connections between them in the brain, and these connections are not fixed but are highly changeable, not just in development but in adulthood. Specific experiences will enhance the connectivity in highly specific neuronal circuits.¹⁵

The more synapses you have the more neuronal gestalts your brain can create. If neuronal gestalts, or patterns of neuronal gestalts, are meanings, then the more synapses you have the more meanings you can create.

Being born into a uniformly bland and uneventful environment would limit ability generally. Most of us were born into an environment which was rich in some kinds of experience and limited in others, and so we grew up with differing abilities. Babies born into households where music was constantly being played are more likely to grow up with a liking for, if not an ability in, music. Babies who are taken swimming are more likely to develop a feeling for water and movement in water than babies whose bathing is confined to a wash in a small bath. Babies born into households which ring with the sounds of people’s voices are likely to have a very different sensitivity to voices and language than those born into homes where silence reigns supreme.

Included in all these possible configurations will be the issue of whether the baby found the experience determining the configuration pleasant or unpleasant. To the baby who is sung to or talked to gently and sweetly words and music will mean something very different than they will to those babies who are continually surrounded by angry voices, by shouting, screaming and crying. Neuronal configurations can change and dissipate, but there is some evidence that those configurations laid down in experiences where the person is afraid have a greater permanence than those configurations laid down where the person is happy. There is a good reason for this. If we are to keep ourselves safe we need to remember where there could be danger, but this is a burden to carry when the danger is long in the past and not likely to recur. Our ancestors needed to remember where the wild beasts might lurk, but do we need in adulthood to remember the adults who were cruel to us when we were children?

What is clear is that each individual brain develops its own unique set of configurations. No two brains ever have the same set of configurations because no two people ever have the same experience. Your set of particular configurations and your meaning structure are intimately related and are unique.

Models of neuronal gestalts can be generated by certain computer software. Professor Igor Aleksander, head of Neural Systems Engineering at Imperial College, has developed Magnus, which stands for multi-automata-general-neural-units-structure, a piece of learning neural software. When Igor sits at his computer and clicks on the icon called Magnus his computer becomes a large neural network. This network is used ‘to see if a bunch of simulated (or artificial) neurons can carry out some of the feats that their living counterparts, the cells of human or animal brains, perform in going about their daily lives. The feats are feats which when performed by humans are described as the result of “conscious thought”.’¹⁶

The problem for computer software like Magnus is that it cannot reach out and explore the world through touch, it does not start as baby software and slowly change and grow through interaction with the world, and it does not interact with and learn from people as human beings do. Igor told me, ‘Your point about growing minds has been on my mind for some time. I think that the only way for Magnus-like-things to develop is for their minds to grow. But they do this in the sense of what in my book Impossible Minds I have called a growth of state structures (or thinking structures). I distinguish physical growth from “thinking structure growth” as the first puts the machinery in place to do the second. In babies the two are concomitant, and I have no idea what the effect of this is. All I know is that from a scientific point of view it is hard to understand.’¹⁷

Igor had been engaged in building machines which recognized patterns, but a question from a young girl at one of his public lectures set him thinking about how to model the way in which humans see. Our vision is far more complex than simply seeing patterns. – for instance, we perceive the qualities of roundness and redness, and put them together to identify a red ball. We can see when our eyes are closed. We can create images of what is not there and images of things we have not seen. Igor’s audiences never have any difficulty in imagining a blue banana with red spots.

Magnus does not have any difficulty in imagining such a banana either. To imagine something we have not seen we have to be aware, and Magnus now contains a set of software which Igor has labelled ‘Awareness Area’. It is unlikely that Magnus’s awareness resembles ours in any way, except in that it is a model based on the hypothesis put forward by Francis Crick and Christof Koch that ‘Everything of which we are aware is fully represented by the firing of some neurons.’ Francis Crick called this ‘an astonishing hypothesis’ but it seems that the only people to be astonished are those who like to think that human beings have certain powers and qualities which are not based on the functioning of our bodies. The only way to prove or disprove such a belief is to establish just what it is that the brain does. We are a long way from this goal, but meanwhile the study of the brain provides much astonishment.

The problem with models is that they can only suggest what something might be like, not what it actually is. Mathematics, like art, describes events, though in very different ways from art. To describe the activity of some object in terms of algorithms might allow for some very accurate predictions, but that does not mean that the object itself is using algorithms. Newton’s equations predict the orbit of the moon, but this does not mean that the moon itself is computing. In the same way, even if certain activities of the brain can be described in algorithms, it does not mean that the brain itself computes. Indeed, there are a number of scientists who now argue that the brain can be understood only as a dynamic system, an organ which evolves its pattern of activity rather than computes it.

The argument has been that if the brain operates like a computer then the output of each cell in the brain encodes a message. The trick was to find the ‘neural code’. Researchers tried to discover what aspect of a cell’s activity revealed the code. If brain waves were measured, could the code be found in the strength of a spike, or the average number of spikes per second, or in some numerical synchronization with the activity of other cells? Many of the results of this research looked promising, but such promises were not fulfilled. Instead, there were results which threw the whole notion of a neural code into doubt.

This research showed that the output of any individual neuron depended not just on a stimulus but on what was happening in the rest of the brain – that is, on what the brain was thinking at the time. According to John Maunsell, one of the neuroscientists carrying out this research, ‘We are coming to the end of one generation of effort. The next generation is going to have to look at the whole system [and] understand the effect that plans, decisions and actions can have on what neurons do.’¹⁸

If the brain did operate like a computer then, when nothing was happening, the brain would be a blank screen. In fact, even in a brain which seems to be doing nothing there is a steady tick-over of cell activity of at least three or four spikes a second. This might be just a leakage of current, but those scientists using a dynamic model of the brain argue this background firing maintains a certain level of tone in the brain and presumably creates some meaning. It is possible that ‘the brain stores memories as patterns of connections between cells – new experiences prompt the strengthening of old connections, or the growth of new ones. The tick-over firing echoing around the brain could be a defocused representation of everything you have ever learnt or known.’¹⁹

‘Everything you’ve ever learnt or known’ – that sounds very much like all the meanings you’ve ever created. These meanings altogether form a pattern or structure – your meaning structure.