THE PROBLEM
Let us begin with Popper’s classic essay, ‘Three Views Concerning Human Knowledge’ ([Conjectures]), where two traditional views, of science as ultimate and unalterable knowledge, and of science as mere utility, are labelled ‘essentialism’ and ‘instrumentalism’ respectively, and where Popper endorses the third view, of science as approximation to truth in stages, which can be labelled ‘modified essentialism’.
There is a lack of symmetry here, as there is little sense to talk about modified instrumentalism. This, as Popper stresses, is because the bold idea that science is merely instrumental for technology becomes rather trite with the slightest modification: no one denies that science is also an instrument. This is not a matter of scientific instrumentalism as such, but of any instrumentalism whatsoever, aesthetic, moral, or social. In general, ‘x is also instrumental to y9 is all too often trivial, whereas ‘x is only instrumental to y9 is radical and stimulating. It is the radical view, not the trivial, which is instrumentalism proper. And it is the slight change of the radical view which lands us in the trivial.There is no point, therefore, in speaking of a modified instrumentalism - as each theory except instrumentalism itself is in a sense a modified instrumentalism; but there is a point in speaking of a modified essentialism as a theory which, like essentialism, aspires to the truth and, like instrumentalism, rejects essentialism, i.e., the idea that the last word on scientific questions is attainable, as rather naive.
This does not seem right; the feel is that, though it looks perfectly reasonable and unproblematic, we may well benefit from caution here. Let us first remember how different a feel one receives from Popper’s Logic of Scientific Discovery of 1935 from the feel one receives from his ‘Three Views’ ([Conjectures}) of 20 years later.
In his earlier work Popper contrasts not essentialism with instrumentalism, but inductivism with conventionalism. It is not easy to doubt - at least I cannot see how - that inductivism is a species of essentialism (in spite of some possible exotic exceptions in the long and varied history of inductivism, see my [‘Sensationalism9]). At least, it is not to be doubted that classical inductivists have repeatedly declared that the aim of science is the unearthing of the truth about the nature of things, that is to say, the true essences of things, and not merely convenient modes of expressing our past knowledge. Thus, Popper approvingly quotes an attack on conventionalism from the induc- tivist Joseph Black ([Logic], 82).Consequently, still intuitively and superficially, one tends to endorse the two equations,
essentialism=inductivism; instrumentalism=conventionalism.
Admittedly, both equations contain some injustice to the apriorists (see Chapter 18). But, at first blush, we may forget apriorism altogether and accept, however tentatively, the two equations above. And the intuitive superficial impression is now that Popper’s view as expressed in his Logic of Scientific Discovery is that of a modified conventionalist, not of a modified inductivist.
It is difficult to nail down impressions, and problems need not develop out of the merest uneasy feelings which accompany them. But let us wait and see. The point where Popper sounds like a modified inductivist is very important: namely, the conclusion of his Logic of Scientific Discovery, And this well accords with his ‘Three Views’. He says there, “ideas... are... our only instruments for grasping” nature (280). “Science never pursues the illusory aim of making its answer final, or even probable”, he says at the close of his earlier work. “Its advance is, rather, towards a finite yet attainable aim; that of ever discovering new, deeper, and more general problems, and of subjecting our ever tentative answers to ever renewed and ever more rigorous tests.” And, finally, Popper speaks at the end of his work of his theory as that of movement in the gmw-inductive direction, contrasting it with theories - such as Spinoza’s - of movement in the quasi-deductive direction; and he rejects movement in this latter direction, Spinoza’s or any other, on general methodological grounds, on his desire to retain science as empirical as possible, which means for him to prefer those scientific theories amongst the available ones which are as testable as possible (277-8).
This is as far as it goes.
The impression which the volume as a whole gives is that Popper is decidedly a modified conventionalist, hardly a modified inductivist - seemingly in clash with an impression he gives 20 years later. This, towards the end of his Logic of Scientific Discovery (Section 84) he suggests to drop the concept of truth altogether. In a later note he reverses this position, of course, and accepts truth as a regulative principle. This, however, seems to clash with the already quoted idea about the desideratum that the aim of science be ‘finite yet attainable’.Also, the quote about ideas being instruments for grasping nature seems not to agree with his present rejection of my own views ([Conjectures], 248n): according to my view, he says, theories are the merest instruments of exploration; my view is, he suggests, a modified instrumentalism! More conspicuously, Popper rejects inductivism as a naturalistic view of science, and endorses instead, quite unhesitatingly, the conventionalist view of science as a game played in accordance with certain rule. Though he rejects the rules offered by the conventionalists and presents rules of his own, this clearly places him more as a modified conventionalist than as a modified inductivist.
Victor Kraft ([Circle]), (127 and notes 127 and 147) has even branded Popper as a conventionalist plain and simple. What part of Popper’s early work particularly impressed Kraft as bluntly conventionalist was the idea of Popper, according to which we need a rule for accepting or rejecting factual statements or observation reports or eyewitness testimonies. The naturalist’s view will force us to acknowledge all testimonies. This openly runs counter to the desideratum, sometimes expressed even by arch inductivists (and reported by Popper, [Logic], 96), that testimonies be sometimes revisable. Nor can the discrepancy be resolved, since to the naturalist any rule is either imposed by nature and is thus not a matter of choice or makes us conventionalists who are willing to rest on arbitrariness.
This is not to endorse Kraft’s view. On the contrary, I wish to ascribe to Popper the important invention of modified conventionalism, or softened conventionalism - or partial conventionalism, to present it in Popper’s own way ([Logic], Section 27) - though somehow this was overlooked by Kraft. We may admit the presence of convention and even of sheer arbitrariness in science, provided it be confined and reducible. And, contrary to Kraft’s impression, the conventional element in testimonies is more easily reducible, particularly when the conventions offered by Popper are adopted, than the conventional element in theories, particularly if the convention offered by the conventionalists are adopted. Indeed, the new element in Popper’s conventionalism which modified it so much, is the proposed convention to be wary of the inevitable presence of the conventional element in science and to prefer at any stage moves leading to its (partial, but increasing degree of) elimination.
The question, therefore, is, shall we equate conventionalism with instrumentalism and declare Popper a modified instrumentalist? Will this not come in conflict with his modified essentialism? Is not modified instrumentalism trite?
II.