LITMIR.BIZ

Скачать книгу

develops does not need to look very logical from the outside. One of the mistakes that brain researchers have sometimes made is to assume that the brain is designed on the basis of elegance and economy. But evolution, as we have seen several times already, does not, indeed cannot, work like that. Natural selection has to build on what is already there, it has to work with the kinds of mutations that just happen to occur, and it can only take account of the particular local conditions that happen to obtain. It can never say: ‘Hold on a minute; this brain is getting untidy. Let’s go back to the drawing board and start again.’ Paul Churchland talks in his book Matter and Consciousness of a car he used to have when he was a student. One of his friends insisted that it was not so much a car as ‘a squadron of nuts and bolts flying in loose formation’. So it is with the brain. We can ask how it has solved the problem of making sure you can quickly find the right ‘book’ in a rapidly expanding ‘library’. But we should not expect to find anything as neat as the Dewey Decimal System. We would be well advised to expect to find something that behaves like a patchwork of linked components, rather than something with a clean overall design.

      There are of course many ways of sub-dividing the work of the brain, just as there are many ways of effectively organizing one’s own kitchen. A kitchenologist could no doubt come along to each of our homes and convince us that it is irrational, or uneconomical, to keep the paper napkins in the same drawer as the bagels, and she may well be right. But the much more important consideration is: does the way I have it organized work for me? If I’m having a phase of Italian cooking, why should I not keep the tagliatelle, the tinned tomatoes, the cookbook and the oregano all together? And if I’ve gone off Italian, and am now wild about Thai, why should I not continue to keep the Italian ingredients together, if I have got used to the arrangement, and make space for the coconut cream and the freeze-dried coriander alongside them? The organization of the brain is likely to be a record of our evolving preferences and strategies for feeding, housing, protecting and perpetuating ourselves, in just the same way.

      So instead of one big ‘theory’ about how the world works, and how to shove it around to my own advantage, the brain begins to divide – functionally, not structurally – into a library of ‘minitheories’³³ Each of these sub-minds is designed to tie together a package of knowledge, skill and experience that can be brought to bear on a particular domain of life. This package specifies what it is for – the purposes to which it can be appropriately put; in what conditions it can be used – what environment it is likely to work in; and what it can do – what abilities, mental and physical, it makes available.

      Two points about these minitheories. First, nothing is needed, in the brain, to play the role of the ‘librarian’, of course. The currently active priorities make themselves known by priming their relevant minitheories to different extents (depending on the urgency of the need or desire); and the sensory receptors feed activation into those minitheories that seem to match what is going on outside. The network quite effortlessly performs its job of selecting a course of action, by waiting to see which ‘gang’ awakens first, and thereby activates its associated bundle of dispositions and expectations. And secondly, the concept of a ‘minitheory’ (or a ‘module’) is one that bridges the domains of ‘brain-language’ and ‘mind-language’. As you zoom in on a minitheory, so you can start to see its fine structure, and describe it in the language of octopus gangs and ultimately in terms of individual neurons and their properties. As you pan back from any particular minitheory (as we are in the process of doing), so you are able to see more clearly the whole mosaic mind-scape of which it is just one part.

      Keeping the Brain on Track

      One of the major technical problems which the evolving brain had to solve was that of keeping its activities precise. When an impulse was sent from A to B, it had to be sure of arriving at the right place and not being side-tracked. Marcel Kinsbourne has suggested that this problem of ‘functional insulation’ of higher level packages of capabilities, in the absence of any structural walls, is solved by the strategic use of the brain’s powers of inhibition.³⁴ Each area of the brain that is active may, almost as a matter of course, tend to create a halo of inhibition around itself. Members of an ‘awake’ octopus gang will be strongly activated enough, by mutual stimulation from each other, to resist the soporific effects of the inhibition. But other nearby units, even though they may have some connections with the active network (but which are not at the time being stimulated directly will have insufficient ‘strength’ to prevent themselves succumbing to the inhibition. As we know, this strategy for keeping the system ‘on the rails’ is by no means foolproof. There are many occasions when our minds slide off the point, and the train of thought finds itself chugging off down some apparently irrelevant siding. But no mechanism is perfect, not even in the brain.

      Mini-Worlds

      The evolution of predation, and especially of active stalking and hunting down of prey, would have acted as a stimulant for the evolution of minitheories. The hunter needs keen eyesight, hearing and sense of smell; powerful muscles for running, pouncing and killing; and greater abilities to predict the behaviour of its prey under various conditions. When the prey goes to ground, or disappears behind a clump of rocks, the more successful hunter will not immediately give up the chase and wander off, puzzled at the sudden dematerialization of its lunch.

      To keep concentrating on something you cannot see, but which you presume is still there, demands a brain that is able to keep active a representation of some bit of the world that is not, at that moment, being stimulated from outside. The minitheory for Hunting, the subroutine for Rabbit, and the running memory of The Story So Far, must be able to be stay activated, and the minitheories that relate to alternative goals must stay inhibited, at least until you have satisfied yourself that the rabbit has truly retreated to a safe place, or that it might just have been a few leaves stirred by the wind in the first place. The activity of the brain, in other words, has to become ever more capable of proceeding in an inner-directed, self-organized fashion, of ‘ignoring’ some environmental changes and fleeting body sensations, and of pursuing ‘plans’ and ‘intentions’.

      So each minitheory develops the capacity to represent one of the sub-worlds which the animal regularly inhabits. A minitheory is a functional map of the recurrent patterns which the sub-world has been found to contain. It provides a manual that enables you to ‘diagnose’ the different options that are likely (on past experience) to occur, and to know instinctively how to ‘treat’ them. The ideal minitheory is one that so comprehensively covers all the angles that it is no longer possible for this sub-world to surprise you.

      A minitheory keeps tabs on a domain – ‘mating’, ‘hunting’, ‘parenting’, whatever – in two senses. First it distils different experiences and encounters into a general set of expectations and capabilities. But secondly it also provides a continually updated record of how any particular encounter is going. As each episode unfolds, so the general expectations are continually being modified to take into account gambles that paid off and discrepancies from what was predicted. In brain-language, the levels of priming and inhibition are constantly being adjusted, on the basis of present experience, to shift the balance of priorities and dispositions that the network represents. Each bit of the brain-mind system is energized or sedated so that the system as a whole is ready to respond to whatever (on the basis of both long-term and recent influences) is the most likely and/or the most significant thing to happen next.

      Note again that, though I have used words like ‘plan’, ‘intention’ and ‘believe’, I have put them in quotes to mark the fact that these words refer to capabilities that we can infer from the way an animal behaves, without implying any role for consciousness whatsoever.

      Overcoming Internal Isolation

      Marvin Minsky has pointed out an additional benefit of a brain that is designed to keep some of its resources and processes separate. A knowledge store that is not subdivided is not capable of being very creative. The flow of activity in an undifferentiated brain-mind can

Скачать книгу