11.02.1999

On prediction and partial knowledge

The approach favoured by Don Hoskins of the Economic and Philosophic Science Review and Socialist News in his polemic against comrade Phil Sharpe is deeply flawed, to put it mildly (‘Historical idealism’ Weekly Worker January 21). Then again, in my opinion, there are also weakness in comrade Sharpe’s philosophical methodology.

The first comment must be that, because someone is an anti-communist, it does not mean they can never be right. But if their other beliefs are to be used as a criterion, then quoting Newton as the arbiter of truth in the same article is a bit of a joke. Newton was, by any standard, including those of the time, a religious lunatic. He thought that he had discovered the ground plan of the Temple at Jerusalem and that this was the ground plan of the universe. Newton also faked his results to prove his theories on optics, even if he did turn out to be right. On top of that he would not have been admitted to the Howard League for penal reform, given his general approach to crime and punishment. He bought the position of magistrate for all of the home counties, and had one man hung, drawn and quartered. Yet, for all this, he is still amongst the three or four of the greatest scientists the world has ever known.

An even bigger criticism of Don Hoskins is that he misunderstands Popper’s main point - one of Popper’s very few gems in a sea of rubbish - that no scientific statement can be considered, as such, unless it is possible to state the condition for its refutation. In fact, this proposition is a way of stating the precise proposition that Don Hoskins requires for science: ie, that science is predictive. So the real objection to Hoskins’ position is that he does not understand it fully.

While dealing with things not understood, Ray Hickman (Letters Weekly Worker February 4) does quite a job on that himself, because Newton made no comment on sub-atomic particles, and Newtonian laws of motion apply down to the atomic level. Newton knew that it should be possible to have a unified theory of matter, but also knew there was no possibility of him providing one. So it is not quantum physics that modify Newton. What has modified Newton’s position is non-Euclidean geometry. Moreover his physics have not been replaced - they are merely a special case of Einsteinian physics, and they are as true today as they were then. And, as it happens, Einstein also got it wrong, not just in the sense he presented a special case, but also in that he included a constant in his theory of the expansion of the universe that is not only not required, but should not be there. Thus, even if Ray’s point - that all scientific theory is only a partial truth - is correct, it is a pity he got the particular instances wrong.

Now if we turn to the roots of this debate, starting with Phil Sharpe’s article (Weekly Worker December 17 1998), it is possible to see plenty of nonsense there. Phil thinks that science is, primarily, explanatory! Well, Genesis is that, and so is a lot of mumbo jumbo. In Highgate cemetery he will find a tombstone on which it says that philosophers have explained the world. The point however is to change it. If we ask the question, “What does it mean ‘to know’?”, the answer must be that we can do something or that we can make something. The fact that we cannot do or make everything just illustrates that our knowledge is partial. We have to predict the outcomes of our action and the behaviour of the material world.

Phil maintains that the problem with all new theories is that they emerge from old theories, which are not only inadequate, but also idealist in character. It may be that this is sometimes the case, though in the sciences I am most familiar with that is certainly not the problem. It is likely to occur in sciences which directly affect people’s interests. The normal problem is lack of data, adequate instalment and mathematical technique. It is also the lack of manpower, brains and in particular money.

In the case of Newton and Darwin, that was certainly the problem. Newton was only able to make his breakthrough as a result of the invention and improvement of telescopes and lenses on the one hand, and the development of mathematics on the other. Only when the king provided the money for the Royal Observatory was he able to get the data that allowed him to predict the position of the moon.

Darwin’s problem was also lack of data as far as fossils were concerned and the general state of knowledge of the world’s fauna and flora. What is noticeable is that most of the modern advances in the theory of evolution have depended on mathematical techniques and the finding of a lot of fossils. While on the subject of Darwin, he was of course a racist and also a Tory. Darwin also could not account for the fact that favourable changes were not eliminated by outbreeding. The solution to that problem was provided by Mendel, who was a friar and eventually an archbishop in the Catholic Church. None of which makes any difference to the fact that Darwin ranks pretty close to Newton in his stature as a world scientist. Indeed, Mendel was right even though he was an idealist while Darwin was a materialist. The fact that Marx wanted to dedicate the English edition of Capital to Darwin does not prove he was not a racist nor that he was wrong in important respects.

The problem with the Bull-Sharpe-Hoskins debate is not that one or other of the writers is petty bourgeois (or whatever term of pseudo-Marxist abuse that may come to mind), but that they simply do not know enough science. Obviously they make individual points that are worthwhile or are interesting, but in order to deal with many of the questions they raise, it would be necessary to send them on science degrees in several subjects. That would be better than trying to prove a point on homosexuality on the basis of Newtonian physics, which is not only impossible, but shows that Don Hoskins has not read that much about Newton, as there is some quite weighty evidence that he may have been homosexual himself.

Science is not just predictive: it also categorises and systematises knowledge, and provides techniques and methods. It further tells us what is not true or cannot be true, but the measure of its maturity is the degree it can accurately predict.

If we look at meteorology, we see a science that is quite as advanced as any social science. Michael Fish however would soon tell us of its limitations. The problem is that, although even Don or Phil can tell us that June will be on average hotter than February, they cannot tell us what the weather will be like in 20 days time - nor can Michael Fish, and science can say why. It is simply not possible with our current computers or mathematical techniques. If all of these were improved a millionfold, we would not advance weather prediction by more than a few days. The reason for this is that the weather system is chaotic (in a mathematical sense) and therefore is calculable, but not predictable.

I noted that comrade Sharpe wrote that capitalism was anarchic. This may be true, but that is not the reason it is unpredictable. Capitalism, like every social and biological system, is chaotic: that is determined by non-linear functions. If this is true for the weather system, what does Hoskins think it is like for the capitalist system as a whole?

This does not mean we cannot make general predictions either about society or about the weather, nor that we cannot do anything to change them in a semi-predictable way.

Several things are certain: 1) the outcome of our social actions cannot in general be precisely predicted; 2) we have to have constant intervention in order to modify the feedback; 3) not even the EPSR can tell which club will win the FA Cup.

John Walsh