[To fully appreciate the following excerpt from Confessions of a Philosopher (1997) by Bryan Magee requires a little background information on ‘logical positivism.’ The school of thought known as ‘logical positivism’ originated in the Vienna between the two world wars. The Vienna Circle, as it was called, was a group of mainly scientists and mathematicians who tackled the problems of philosophy with an exclusive emphasis on logic and the methods of science. The result was the famous—or infamous—Verification Principle. According to this principle for a statement to be meaningful it had either to be analytic (eg. all quadrupeds have four legs), or it had to be empirically verifiable (eg. there are fish in the sea). From this it followed that all metaphysical statements and judgements of value were meaningless or subjective, and therefore of no philosophical significance.]
What was almost comically typical of Britain’s, and especially Oxford’s, parochialism is that a completely effective demolition of logical positivism had already been published before A. J. Ayer introduced it [in his book Language, Truth and Logic, 1936] into the English-speaking world in the first place. In 1934, in Vienna, a book had appeared called Logik der Forschung, by Karl Popper. It was not to come out in English translation until a quarter of a century later, when it was published in 1959 under the title The Logic of Scientific Discovery. In it there were many criticisms of logical positivism, including some I have mentioned already, but its central and most devastating one was that logical positivism claimed to be first and foremost a (indeed the) scientific view of the world, and yet its central tenet, the Verification Principle, wiped out the whole of science. This criticism, if clinched—and few people today would deny that Popper’s book pretty well clinched it—spelt total shipwreck for logical positivism.
Popper’s argument can be summarized as follows. From Newton until the time of the logical positivists the central task of science had been seen as the search for natural laws, these being unrestrictedly general statements about the world that were known to be invariantly true. Examples are ‘Every physical object in the universe attracts every other physical object in the universe with a force which is directly proportional to their masses and inversely proportional to the square of the distance between them’ (Newton’s general law of gravitation)...
Statements such as these laws are never analytic, and indeed, if they were they would give us no information about the world. Their truth does not follow by deductive logic from the definitions of their terms, nor would denial of them be self-contradictory. On the contrary, their discovery nearly always comes as a revelation. What they give us is empirical information which is of interest, and often of great practical value, about the way things contingently are in the world, information which observation is able only to suggest and rigourous experiment to corroborate. However, to the astonishment of most of those who understood it, what Popper demonstrated in Logik der Forschung is that scientific laws are not empirically verifiable. Nowadays people are inclined to attribute the origination of this revelation to Popper himself, but in fact, as he always acknowledged, it had been made by Hume two and a half centuries before; but although Hume’s exposition was clear and unequivocal it was so bombshell-like in its implications that only the very greatest of subsequent thinkers, such as Kant, Schopenhauer and Einstein, really took it fully on board. Nevertheless it was Hume who first pointed out, and with all his customary lucidity, that from no finite number of observations, however large, could any unrestrictedly general conclusion be drawn that would be defensible in logic. It might well be that, if every time I let go of something it falls, I conclude eventually that all unsupported objects fall, but if so the conclusion has been reached from the premises not by a logical process but by a psychological one. If you see an A, and it has the characteristic x, it does not follow logically from this that the next A you see will also have the characteristic x. Either it may or it may not—there may be some A’s with this characteristic and some without—unless of course you make the conjunction true by definition, which is to say you stipulate that something is to count as an A only if it has the characteristic x. But in that case the statement ‘All A’s are x is a tautology [a tautology is a statement that says the same thing twice, but in different words, i.e. a bachelor is an unmarried man] and conveys no empirical information.
Not only does the non-tautologous conclusion that ‘All A’s are x’ not follow from a single observation of an A, it does not follow from two such observations, nor from two thousand, nor from two billion. The best-known example that has been used in illustration of this point has to do with swans. For thousands of years before the discovery of Australia all the swans that any Westerner had ever seen had been white, and everyone seems to have taken it for granted that all swans were white—expressions like ‘swan white’ or ‘white as a swan’ were common, and the very statement ‘All swans are white’ had been made familiar by being used as a recurrent example in a standard textbook of logic that was in common use at the time of the Reformation and after. But when Europeans discovered Australia they encountered, for the first time, black swans. Now they could have reacted to this by saying that because these birds were black they were not swans, but a different sort of bird, and then given them another name. That would have been to empty the statement ‘All swans are white’ of informational content by making it true by definition. Instead they accepted that these birds were indeed swans, and that the statement ‘All swans are white’ was false. But what this means is that however many millions of people over however many thousands of years, and without one single exception in all that time being seen to be white, it had never followed that all swans were white. As Hume put it: ‘However easy this step may seem, reason would never, to all eternity, be able to make it.’ But this in turn means that unrestrictedly general statements of the form ‘All A’s have the characteristic x’ are, of their very nature, not empirically verifiable. And the disconcerting fact is that scientific laws are characteristically statements of this kind. So their unrestricted generality makes it permanently impossible to verify them empirically, by no matter how many observations—trillions, zillions, any number anyone cares to name. So, said Popper, from the Verification Principle it follows that scientific laws are meaningless statements, and are empty of informational content. The Verification Principle rules out all scientific laws, and therefore the whole of science.
Fully to understand Popper it is essential to realize that he was not a thinker who was following the same line of enquiry as the logical positivists and arriving at a different conclusion from them. He was on a different path altogether. They, it will be remembered, were in search of a criterion of meaningfulness, a criterion of demarcation between sense and non-sense. Popper always held that this, the search for a criterion of meaning, was a mistake. He pointed out that much the most useful knowledge we have, and the biggest body of it, is contained in the natural sciences, yet scientists are not given to debating the meanings of their fundamental terms, terms as widely different in kind as physics, observation, measurement, light, mass, energy—not to mention all the terms involved in the mathematics they use (what is a number?—what is mathematics?). Scientists leave this, for the most part, undiscussed, and get on with doing more science. And, said Popper, they are right. The clinging notion that if we are to have a worthwhile discussion we need first to define our terms is demonstrably self-contradictory. Every time we define a term we have to introduce at least one new term into the definition, otherwise the definition is circular. But then we are under an obligation to define our new term. And so we are launched into an infinite regress. Attempts to clarify all our terms must, and can only, result in discussions of words and meanings to which it is logically impossible that there should ever be a conclusion. So discussion, if it is to take place at all, has no alternative but to make use of undefined terms. And this is at bottom the logical justification for what scientists do in this regard. And as their example makes clear, it is no bar whatsoever to rapid, successful and continuous growth in our knowledge and understanding of the world. The only thing that discussion of the meanings of words extends our understanding of is the meanings of words: it does nothing, or next to nothing, to extend our understanding of non-linguistic reality.
For these as well as other reasons, Popper asserted from the beginning that, both in fact and in logic, for a philosopher to be centrally concerned with the meanings of words was a disastrous error. It precluded him from ever getting down to the discussion of matters of real substance. The very endlessness of the processes in which it involved its participants meant that it could not but be unproductive as regards the primary level of discussion, and in consequence of that, boring to anyone for whom it was not an end in itself. In practice it was bound to lead to interminable word-spinning, logic-chopping, and in the end scholasticism. So not only was he not in search of a criterion of meaningfulness, he perceived that those who were had waded into a quicksand from which they would never emerge unless and until they renounced the search.