appendix: the traditional ‘AD hoc’ use of INSTRUMENTALISM

In his classic ‘Three Views Concerning Human Knowledge’ Popper tells the story of the battle between realism and instrumentalism as that be­tween the forces of light - of science, of Copernicanism, of Galileo - and the forces of darkeness - of religion, of the Church of Rome, of St.

I accept the beginning but not the end of this story. Instrumentalism is the denial of any informative force from scientific theory. It may apply to (a) science at large - by enemies of science and by friends of science - and it may apply (b) to a given defective or defunct or superceded scien­tific theory as a consolation prize by its scientific opponents. One might go to a position between (a) and (b), and declare (c) all or some of today’s science but not all science and certainly not tomorrow’s science to be a mere instrument. Or, one can (d) use the instrumentalist philosophy as a supplement to today’s unsatisfactory realist philosophy.

Consider first, then, (a) any comprehensive version of instrumentalism. Of the enemies of science I shall not speak here (see, however, appendix to Chapter 10). The friends of science who are both positivists and in­strumentalists, i.e., those who say both that only science should be heeded to, and that no scientific theory has any informative content, they are forced to see the world as a string of unconnected events. Hume found this distasteful and was driven by it to desert philosophy in favour of a game of backgammon. Mach and Wittgenstein were driven by it to mysticism.

But instrumentalism was more often used (b) to apply to your theory at the same time as realistic claims were made for mine.

Now, somehow, viewing our opponents’ theories as a consolation prize may be all right, as in the above examples, and quite objectionable, such as Bellarmine’s. Now Bellarmine was only continuing an older trend, that was started by Ptolemy and strongly rejected by Copernicus and his followers. Indeed, the arch-instrumentalist and the paradigm of instru­mentalism are Ptolemy and his system in all its transformations from his days to the days of Galileo. He, doubtlessly, viewed his own theory as a mere instrument of prediction in a mood of self-deprecation and while ascribing a realistic or philosophical statue to a theory he liked better. He had, in brief, two desiderata, and two doctrines; and each of the two doctrines answered only one desideratum: simplicity (which is a condition for truth, since nature allegedly is simple) and explanatory-cum-pre- dictive value; or, in a more traditional jargon, agreement with nature and agreement with appearances. He would have preferred a theory which agrees with both. He even accepted that in some recondite manner any theory which agrees with nature also agrees with appearances; yet agree­ment with appearances must be itself apparent. And so, Ptolemaic in­strumentalism was (c) an ad hoc, a stop-gap, the best of a bad job for the time being.

Almost the same attitude is expressed in modern times by Laplace regarding probability. It is a stop-gap for the time being, not a theory which agrees with nature: God does not play dice, as Einstein restated it. Even those who do not accept Laplace’s and Einstein’s determinism, even those who think that determinism is false and its adherents back­ward-looking, even they do not condemn Laplace and Einstein.

This is no apologia for Bohr. Popper has spotted some instrumentalist ingredients in Bohr’s views as expressed after Einstein, Podolsky, and Rosen had criticized his view of quantum theory. These ingredients are errors, and should be eliminated; but instrumentalist traces, even when presented apologetically, are not instrumentalism and not any capitula­tion to the Church of Rome.

There is, however, a difference between, say, instrumentalism which reflects (b) a somewhat tolerant attitude towards an objectionable theory, or the instrumentalism of Helmholtz which reflects (c) the demand for certain improvements of a theory before it passes from the status of a mere instrument to the status of a realistic scientific theory, and (d) Bohr’s instrumentalism that comes to exonerate quantum theory from the re­alistic assault on it by Einstein, Podolsky, and Rosen.

This is not to say that Bohr is an out-and-out anti-realist or instru­mentalist. For, the vacillation between realism and instrumentalism, be­tween nature and convention, between any two given poles, is often very reasonable on the following assumption. Suppose you have two extreme positions, each of which is untenable; you may intuitively feel that the truth is the synthesis of the best elements of both, and feel that you approximate it sufficiently closely if you alternate between the two ex­tremes, taking the best of each.

There is no better evidence that contemporary physicists are not as instrumentalist as they profess to be, than in their attitude towards applied mathematics which, almost by definition, is the home of theories which are supposed to be mere instruments, to have no depth but mere predictive value, such as classical dynamics, especially the continuum theory. The distinction between elasticity and quantum theory is made from the start from a realistic viewpoint by the same people who defend quantum duality from the apologetic instrumentalist viewpoint - thus exhibiting the desire to find the happy synthesis. Requiring them to stop doing physics until they straighten out their theory of science seems to me to be an excessive demand; trying to purge them of the vestiges of the brand of instrumentalism and/or the brand of realism they happen to know amounts to the same. I see no use for such a crusade.

It is here that one can place Bohr’s view, I feel. It is, no doubt, not an instrumentalism plain and simple: Bohr never denied the existence of sub-atomic particles. Already in his original paper on complementarity he vacillates between the two poles. There are inductivist elements in presentation - nature imposes on us conflicting images - and there is a Kantian element or an instrumentalist element - whatever is in principle beyond our horizon should be left outside science. Though his reply to Einstein, Podolsky and Rosen is a severe modification of his earlier view it is not so much of a change as to justify trying to make him out a fully fledged instrumentalist.