On Growth
(1974)–Willem Oltmans–
Auteursrechtelijk beschermd
[pagina 7]
| |
is considered one of the greatest minds in contemporary Britain and actively participates in Club of Rome meetings. He was born in Evesham, England, in 1905. It is important to get some idea of what computer simulations of complex situations can and cannot be expected to do in the present state of the art, and how far the MIT team, and others more recently associated with the Club of Rome, have actually got to date. The procedure in setting up a computerized model of a complex system is something like this: You start by choosing, on the basis of common sense, a certain number of major components of the system which seem likely to be of importance. The MIT study chose five main variables: population, capital expenditure, natural resources, pollution, and capital investment in agriculture. Each of these is then subdivided again into a lower level of active factors; such as for population, birth rates and death rates, or in other sectors such factors as the amount of arable land available, the capital cost of bringing new lands into cultivation, and so on. Then one has to try to put into the model quantitative estimates of the strengths of interactions between the factors. For instance, what effect does pollution have on population? These interaction effects (‘multipliers’ in the system builders' jargon) at present usually have to be guessed, since there are only a minimal number of facts on which to base the estimates. However, the computer is flexible enough to deal with ‘multipliers’ which change according to the actual values of the two things which are interacting. For instance, the MIT team suggested that pollution does not have a proportionate effect on death rates at all | |
[pagina 8]
| |
levels, but has very little effect until it reaches a fairly high level and then rapidly becomes more and more of a killer. This is a guess, based on such phenomena as, for instance, the great London smog of the winter of 1952, which was reckoned to have precipitated the death of three thousand people. The only real check - and it is a pretty feeble one - on the validity of these guesses, is to show that the whole system produces a reasonably accurate picture of what happened to all the variables for which figures are available for some period in the past, for instance, from 1900 to 1960. Any set of interaction values which fails to do that would of course be unacceptable; but quite a lot of different values would probably produce a reasonably decent fit to the existing data, particularly since some of the factors which look like becoming very important in the future, such as high levels of pollution or the exhaustion of some natural resources, did not begin to exert any important influence until quite recently. Having set up the model and shown that it does work at least for the past, the computer will work out for you what would happen if the same set of interactions continued into the future. The result the MIT model turned out is that by the end of the century the world begins to run out of natural resources and can no longer support such a large population, which gets dramatically, indeed catastrophically, reduced. The MIT team then tried the effect on the behavior of their model of altering some of the numerical values they fed into it. As a continuation of their original setup would lead to an exhaustion of natural resources, an obvious step would seem to be [to] reduce the rate at which natural resources are exploited. When they changed the values in the model to correspond with such a new policy, what came up was an eventual rise in pollution, which again led to population catastrophe. So what about both reducing the rate of exploiting natural resources and increasing the efficiency of industry in relation to pollution production? This postponed the catastrophe for a bit, but eventually led to the same type of result. In fact, the only alterations to the MIT model which gave rise to an essentially stable situation, in which the population neither rose nor fell abruptly, was one involving severe restraints on the rates of capital investment, of resource exploitation, and even of food production. To those used to everything getting continuously bigger and better, it sounds like a very cheerless prospect. | |
[pagina 9]
| |
How seriously should these results be taken?
As predictions they should not be taken seriously at all. They do not pretend and were not intended to forecast what will happen. Their importance, which is quite real, is of a different kind, or rather of three different kinds. In the first place, they do show types of catastrophic misfortunes which could take place if the world went on working according to a system which is not too drastically implausible. It will be wise for mankind to keep well in mind that catastrophes are by no means unthinkable. Secondly, they drive home the lesson that if one does something to - makes an alteration in - a complex system, the response of the system may not be at all what was expected or intended. This is a lesson which people continually have to learn over again. The behavior of complex systems is often what the MIT team call ‘counterintuitive’ - the damn thing just does not do what it ‘should’ do. An earlier study by the same group of workers at MIT was on the growth processes in cities. It has been quite a common experience, particularly in the United States, that a well-meaning authority carries out a program of slum clearance, erects a lot of new quite superior housing, and within a few years finds that the original slum area is even more overcrowded with poorer people than before. Probably one of the main factors in causing this counterintuitive behavior is that the new buildings attract a large number of people into the area, but if there are not enough jobs for them, they remain poor and let their dwellings get even more overcrowded. Understanding just why the complex systems do not behave as expected is the main purpose of trying to make models of them. A third importance of the MIT world scheme is that it is a beginning of a process of exploring a variety of models to see if one can be produced which really behaves as the world itself does.
Professor, if it is true that the treasure of life, the richness of life, is stored in the diversity of genes, of living beings, if diversity is a must, is man's behavior destroying this needed richness?
We are certainly destroying it in the natural world. We are killing off many species. We are depleting the fauna and the flora in various parts of the planet and it is very important that we stop doing this. Many | |
[pagina 10]
| |
of the genes we need for improving crop plants or for protection against disease and so on, exist in wild species and if we eliminate all these wild species, we shall have got rid of them; and it will be very difficult to get them back again. Now, within the human species, on the whole what we're doing is not so much a genetic effect, but much more a cultural effect. We are really making it possible, I think, for people to be much more diverse than they could be in the past, because we have greater wealth and greater leisure and so on. There is less pressure simply to keep themselves alive. On the whole, people are being able to be more different, and this certainly applies to the developing parts of the world where up till now everyone has simply had to work all the daylight hours to keep themselves alive. In some of the richer parts of the world, such as in Europe, possibly the richer classes could have been a bit more diverse in the past than they can now. There is a considerable pressure for uniformity. But I think nothing like the pressure for uniformity there was in the peasant medieval civilizations, when they had to spend their whole daylight hours working in the fields and had no opportunity to do anything else much.
And Skinner'sGa naar eind1 approach would lead to a form of authoritarian (world) management?
It might do, but of course, Skinner is very anxious that it shouldn't do. I think Skinner is quite right in emphasizing how much we influence each other, and how much we are programmed by our upbringing and so on. But I think it's really much too early to think we could design a system of programming to produce the optimum sort of person. We have got to experiment gradually along this way. We cannot avoid influencing each other and being to some extent programmed. But I think Skinner is overemphasizing a valuable contributary element in the situation. And I think ChomskyGa naar eind2 is overemphasizing the other side, that everything is totally spontaneous and wells up from the deeper levels of the spirit without any influence of other people and other things. I think they are both overemphasizing their particular points of view, both of which have something in them.
You would feel reality is in the middle?
Somewhere in the middle. | |
[pagina 11]
| |
What in your opinion is molecular biology, which has been called the practicing of biochemistry without a licence?
Well, a lot of the talk about genetic engineering and advanced biological manipulations is by people who are looking fairly far ahead. It's really just as well that people should start talking about these things even ten or twenty years before they are practicable. But many of these things will not even be practicable. I don't believe that you would be able to invent a lot of better genes and synthesize the DNA, and insert it. I don't think we will be able to make a radical plan for alterations to the human genes. But it is much more possible to manipulate human eggs and sperm. It is not out of the question, to produce many identical twins, for instance. In fact it has already been done with frogs. To do it with a mammal would be much more difficult, but it may well be done with livestock, with beef cattle and pigs and so on. I should say it could be done within ten years if a large research program is invented. To do it with human beings could require another great development program. To do something on human beings, it's got to get at least ninety-nine point nine percent reliable, if not more. To do it on farm animals you would not mind if one percent of them died or did not work very well. But to do it on humans it's got to be fantastically reliable and this is an enormous program. I don't know if you have ever come across an article by a man called Djerassi, who is the head of the Syntex Corporation, which discovered the source for making the steroids which are used for the contraceptive pills. He was considering what would be involved in making a real advance in a new type of contraceptive pill. And he points out that legislation in America requires extensive testing for toxicity and effects on embryos and testing on different species before you can come to doing it on experimental groups of humans. This testing takes a minimum of about fifteen years and costs a minimum of some tens of millions of dollars. He said that the chance of getting the money back from the result is small, and he maintained that it will not be done. The contraceptives in 1984, he said, using a fashionable date, would be essentially the same as they are now; nobody could afford to develop a really new one.
Unless the war in Vietnam is stopped.
Even if they stop the war in Vietnam, you could not cut down on | |
[pagina 12]
| |
this testing. But you could possibly do it in India or Brazil, or some country that has a very big population problem, and hasn't got much of a sophisticated legislation about safety. But I am mentioning this because if you think of these funny procedures like genetic engineering or producing identical twins and so on, they're going to involve the same sort of magnitude of testing and control before they can be put into human use. The only one, I think, which is quite likely to come out is the determination of the sex of your next baby. This would have some agricultural importance, though not very much. It would have enormous importance for man. I am rather doubtful whether we shall really get population control accepted everywhere, unless people can be quite certain that the two children they do have are the sexes they want. If they want boys, and had a girl at the first time, they want to be certain the next one is a boy, otherwise they would go on till they do get a boy. It would be to the advantage of population control, if you could tell people that they could choose and be sure about what sex the next baby will be. They could get it now, but only by a rather unpleasant way. You can diagnose the sex of the fetus and abort it if it is the wrong sex. But you cannot at present do this until it is fairly old. Furthermore it is rather unpleasant to abort, and by this time the mother is already rather attached to it and does not want to get rid of it, so it is not a pleasant way of doing the job.
Is work being done on this?
Well, a little. I mean, there may be a dozen scientists in the world, possibly not more than half a dozen, working on it - infinitesimal in comparison with those working on supersonic transports.
Don't you believe that it is essential that these programs be under some sort of world body's supervision, the United Nations or WHO?Ga naar eind3
Yes, I think they should be. They are bound to be under national control already, because they cost so much that they won't be done by anything but national bodies, using national funds, that is essentially taxpayers' money, with parliamentary control or whatever system of government the country has. So I don't think they can be done by private individuals. I think they should be under some sort of public control. Now, I would like to see this world public control. | |
[pagina 13]
| |
You would be in favor of it across the border, with the socialist countries, the Soviet Union, for instance.
Yes, I would. But whether the existing international societies are really adequate to do this - the experience of the Olympic Games, for instance, to give one example - suggests that the international bodies of the moment are not very world-minded. |