The conviction that Wittgenstein was a relativist is entrenched in the literature on sociology and the history of science, although philosophers are far from united on the question (Kusch, ‘Annalisa Coliva on Wittgenstein and Epistemic Relativism’ (2013); see also Pritchard, ‘Epistemic Relativism, Epistemic Incommensurability and Wittgensteinian Epistemology’ (2010)). It seems to me to be at odds with a number of passages in which Wittgenstein expresses a quite different view of science. In a note from 1931 he wrote: ‘As simple as it sounds: the distinction between magic and science can be expressed by saying that in science there is progress, but in magic there isn’t. Magic has no tendency within itself to develop’ (Wittgenstein, ‘Remarks on Frazer’s Golden Bough’ (1993), 141). The fact that an enterprise makes progress does not necessarily mean I should adopt it: athletes run faster every year, but that is no reason why I should take up athletics. But science is a special case: if science gets better at understanding nature, gets better at prediction and control, then it is very difficult to see how I can remain indifferent in the face of such progress.
A remark from 1931 can easily be dismissed as unrepresentative, but we find essentially the same views in Wittgenstein’s last set of notes, On Certainty (1969). Consider the following passage:
131. No, experience is not the ground for our game of judging. Nor is its outstanding success.
132. Men have judged that a king can make rain; we say this contradicts all experience . . .
I take Wittgenstein to be saying that we cannot ground induction in experience, just as Hume showed that we cannot ground our notion of causation in experience; but even though we cannot ground a particular procedure in a philosophical justification, we should certainly go on using it if it is outstandingly successful. The magical claim that a king can make rain is not an ‘outstanding success’; and when we say that it ‘contradicts all experience’ what we have is a clash between their magic and our science in which our science is superior to their magic.
Compare:
170. I believe what people transmit to me in a certain manner. In this way I believe geographical, chemical, historical facts etc. That is how I learn the sciences. Of course learning is based on believing.
If you have learnt that Mont Blanc is 4000 metres high, if you have looked it up on the map, you say you know it.
And can it now be said: we accord credence in this way because it has proved to pay?
Again, the argument would seem to be that I cannot prove that Mont Blanc is 4000 metres high, but believing it, on the authority of a map, has ‘proved to pay’. In other words, the social procedures we have for establishing certain types of fact cannot be justified, but they are successful, they pay, and this is why we employ them.
And (to take one of a series of notes dealing with the idea of going to the moon: 106, 108, 111, 117, 171, 226, 238, 264, 269, 286, 327, 332, 337, 338, 661, 662, 667):
286. What we believe depends on what we learn. We all believe that it isn’t possible to get to the moon; but there might be people who believe that that is possible and that it sometimes happens. We say: these people do not know a lot that we know. And, let them be never so sure of their belief -they are wrong and we know it.
If we compare our system of knowledge with theirs then theirs is evidently the poorer one by far.
It is easy to assume that Wittgenstein’s point here is a relativist one: we say their knowledge is inferior to ours; but they say the same about us. But suppose a society which believes that one can travel to the moon by leaving one’s body, as shamans do (cf. §§106, 667), and compare it with Wittgenstein’s own world in 1950: isn’t it fair to say that the scientific knowledge of 1950, which made possible the jet engine and the atom bomb, was superior to (more successful than) the magical knowledge of a shamanistic culture? (see Child, Wittgenstein (2011), 207–12)
The same sort of point is made again:
474. This game [assuming the stability of things as the norm] proves its worth. That may be the cause of its being played, but it is not the ground.
Thus I assume that this table will continue to exist if I get up from it and leave the room. I cannot justify this belief, but believing works out well (it pays, it is successful), and this is why I continue to act as if this belief were true (this is the cause of this game’s being played).
Lastly:
617. Certain events would put me into a position in which I could not go on with the old language-game any further. In which I was torn away from the sureness of the game.
Indeed, doesn’t it seem obvious that the possibility of a language-game is conditioned by certain facts?
Take the language-game represented by Ptolemaic astronomy; that game ceased to be possible when the telescope showed that Venus has a full set of phases. Thus language-games do not simply succeed, progress, pay, or prove their worth; they can also become unsustainable if the facts change.
Taken together these passages suggest that there are some types of knowledge which are superior to others because they work, they pay, they are superior, they make progress, and they are not at odds with the known facts. We cannot provide a satisfactory philosophical justification for these types of knowledge (broadly, ‘the sciences’), but we can tell that they work, and other cultures interested in understanding, predicting, or controlling natural phenomena (and all cultures must be interested in these activities) should be able to recognize the utility of our knowledge (of our maps, or of our weather forecasts), just as indigenous Americans could recognize the advantages of horses and guns for hunting buffalo. This amounts to an anti-foundationalist but far from relativist view of science. It would follow that when scientific views are abandoned and replaced by new ones it is because the new ones are thought to be better at succeeding, paying, etc. In other words, science evolves, and it does so because theories that fail to develop, or that are unable to adapt in the face of new discoveries, are eliminated.
This is (as it happens) the view of science put forward in this present book, which, it would therefore seem, is authentically in the tradition established by Wittgenstein. But Wittgenstein’s texts are puzzling, problematic, and unfinished. They are open to more than one reading. I have no great quarrel with those who wish to read Wittgenstein as a relativist, providing they do not use this reading to justify a relativist history of science. If pointing out that Wittgenstein himself was not a relativist in his understanding of science helps persuade historians to abandon their hostility to what they (misleadingly) call ‘Whig history’ then it is worth debating what Wittgenstein really meant. For note that to say that a practice pays, succeeds, proves its worth is, of necessity, to make a retrospective judgement: we can only distinguish good science from bad science, on Wittgenstein’s account, with the benefit of hindsight.
And we cannot opt simply to ignore the distinction between good and bad science, because if we do we will miss one of science’s peculiar characteristics, that it makes progress.
The question of what Wittgenstein really thought must, in any event, be kept separate from the question of his influence: On Certainty was not published until 1969, by which point a view of Wittgenstein as an uncompromising relativist was firmly established. And so his texts played a decisive role in legitimizing the new post-Kuhnian history of science because they were wrongly read as endorsing a thoroughgoing relativism. (See also below, pp. 586–7.)