METAPHYSICAL DOCTRINES ARE OFTEN INSUFFICIENT FRAMEWORKS FOR SCIENCE
The methodology of this and the next section is a generalized Cartesian methodology, and the generalization I am offering is possible only within Popper’s framework. Descartes’s metaphysics (which was an improvement on Galileo’s), was a clockwork view of the universe.
It explained almost nothing; it was not intended to explain anything. Descartes claimed that any scientific hypothesis which he could endorse must be one which conformed to his metaphysics. He added that explanatory hypotheses conforming to his metaphysics could always be found. Boyle made the same claim concerning his own semi-Cartesian metaphysics, and so did Newton concerning his own metaphysics (in his preface and the Scholium Generate to Principia). But this repeated claim of the metaphysician is often false. It may be argued that his doctrine allows insufficient room for explanation, that it provides too narrow a framework. When this is felt to be the case, the demand for a new metaphysical framework arises. Metaphysics stagnates in scienceless (or uncritical) cultures; it is progressive in scientific ones. It progresses then because existing metaphysical doctrines are felt to be constricting frameworks, and thus unsatisfactory.Thales’s doctrine aimed at explaining (physical and chemical) diversity and change by assuming an underlying and unchanging unity. Any such approach runs the danger of being too successful and thus selfdefeating. For the assumption of an unchanging unity leads to regarding observable facts as illusory. This was the magnificent discovery of Parmenides, and it was this discovery that made him deny the existence of diversity. We have no right to despise him for having preferred his own logic to common sense, for having proclaimed appearances to be illusory; rather we should admire his dazzling logical acumen. But for him we might not have had Leucippus and Democritus.
The greatest novelty of their atomic doctrine is that it expressly allowed for both unity - of the atomic character of matter - and diversity - of the atoms’ shapes, sizes, and spatial order. To put it in quasi-ancient idiom, atomist metaphysics is a program to explain the many not by the one but by the few; it is thus more accommodating than the metaphysics of Thales.(Of Parmenides’s other great logical discovery, of the nonexistence of the void, I cannot speak here beyond saying that it was the cornerstone of the theory of space developed by Leibniz, Faraday, and Einstein. Further details of the story of Parmenides and Democritus, as well as the story of the downfall of the Pythagorean program, the demonstration of its narrowness, and its rectification by Plato, as well as the relation between Plato’s program and Euclid’s geometry, have been admirably presented by Popper in the paper already mentioned. The great role played by both Democritus’s and Plato’s programs in the seventeenth century have been beautifully told by Alexandre Koyre. The relation between Leibniz’s program and Einstein’s scientific theory of space is discussed in Einstein’s exciting preface to Max Jammer’s Concepts of Spaced)
My next example of an unsatisfactory metaphysics is Cartesian metaphysics, which contained the thesis that all (non-inertial) motion was due to push. The example for this was the suction pump whose (pull) action had been scientifically explained as due to atmospheric pressure (push). Lifting a jar seems to be pulling it upwards, but in fact it is pushing the jar upwards by the handles. Now this last example was implicitly criticized by Newton. If the jar is strong, or contains light material, it can be lifted by its handles; otherwise, pushing its handles upward hard enough will only constitute lifting the handles while leaving the jar itself on the floor. We must admit, then, that lifting it by its handles not only involves pushing the handles upward but also pulling the jar itself upward with the aid of the attractive forces which keep the jar and its handles connected - by the forces of cohesion.
This example justifies Newton’s claim (preface to Principia) that his program was in the first place more accommodating than Descartes’s, and so could be more fruitfully adopted even if ultimately we should return to Descartes’s program.But Newton’s metaphysical program, too, was so naive, that one may wonder how it was accepted for so long. Assuming, with Coulomb, that electric forces act solely between electric charges and that gross or ordinary matter is subject to the forces of gravity and cohesion only, why then does the charge remain on the charged body and pull it along when moving toward, or away from, another charge? This question (which was raised in 1800 with the discovery of electrochemistry) clearly indicates that gross matter is in some sense electrical. Yet so strongly impressed were people with Newtonianism that twenty years after Faraday had produced wonderful scientific theories which incorporate the supposition that gross matter has some electric characteristics, these theories were almost unanimously ignored (the exceptions were Kelvin and two other, rather minor, physicists). Those statements in the Encyclopaedia Britannica of the eighteen forties and fifties which appear to allude to Faraday’s theories are certainly contemptuous and derisive.
It is not accidental that Boltzmann explained in 1885 (in a letter to Nature, p. 413) the general opposition to Maxwell by the general adherence to Boscovitch’s, not to Newton’s, program. Boscovitch had modified Newton’s program to permit one material particle to dispose a variety of forces. This he did because he had discovered that otherwise the program would not accommodate any explanation of the phenomena of elastic collisions. But his program became popular only after Faraday imposed his view that gross matter had electric properties.
Incidentally, the indifference to Maxwell which worried Boltzmann shows that even Boscovitchian metaphysics may be highly dangerous; but it is a truism that any idea may become dogma.
So far I have only spoken of the requirement that metaphysical doctrines be sufficiently wide frameworks to accommodate possible future scientific theories. In the next section I shall speak of the requirement that metaphysics be inspiring and lead to the development of scientific theories.
VIII.