Modern Science and Its Philosophy

Philipp Frank


the place of logic and metaphysics in the advancement
of modern science

ONE of the most brilliant writers on intellectual history, Carl Becker, claims that the most important event in this field in modem times was the shift in the place of logic in science. According to Becker, the high esteem in which logic had been held by the scientist in the time of St. Thomas Aquinas and through all the Middle Ages declined in the period of Galileo and Newton. But at that time this decline was not yet fully understood. "The marriage of fact and reason," as Becker puts it, "proved to be somewhat irksome in the nineteenth century and was altogether dissolved in the twentieth century." The modern, twentieth‑century, physicist lives in an "atmosphere which is so saturated with the actual that we can easily do with a minimum of theoretical. . . . We have long since learned not to bother much with reason and logic." To describe the spirit of twentieth‑century physics, which emphasizes facts and minimizes reason, Becker says:

Experiments seem to show that an electron may, for reasons best known to itself, be moving in two orbits at the same time. To this point Galileo's common‑sense method of noting the behavior of things, of sticking close to the observable facts, has brought us. It has at least presented us with a fact that common sense repudiates.

I do not think that Galileo was actually less concerned with logical consistency than Aristotle, nor do I think that Einstein's theory of gravitation is more factual and less rational than Newton's. I shall not enter here into historical arguments; I shall restrict myself to investigating the nature and the bearing of that "most important event in the intellectual history of modern times" which Becker stresses so strongly.

The pre‑Galilean period, say until about 1600 A.D., adhered to a kind of organismic world view founded largely upon the philosophy of Aristotle and the medieval schoolmen. The "mechanistic" conception of nature began with Galileo and Newton. Its peak, however, was reached in the first half of the nineteenth century. According to Becker, this whole period including the Middle Ages, the Renaissance, and the eighteenth‑century enlightenment, was characterized by the belief that there is a rational picture of the world, there is a way by which man can comprehend nature by reason. But in the twentieth century this belief faded more and more. An acute observer would have already noticed this fading in the Galilean period when the confidence in the scholastic type of philosophy declined. More and more, the "climate of opinion," as Becker calls it, became less logical and rational and more factual and in some ways irrational.

We shall probably not seriously believe that science has ceased to be logical. If we use this word in its technical sense, everyone will agree that a science that repudiates logic has never existed and can never exist. Now, what did Becker really mean by his "decline of reason and logic" in the twentieth century and even in the approaches to this century? Why did he believe that the organismic medieval world picture was rational as well as the Newtonian and Laplacian mechanistic picture, while the twentieth‑century picture of physics, characterized by relativity and quantum theory, is no longer concerned with reason?

We shall understand this point better if we direct our attention to the fact that, according to Becker, this decline of the rational world view had already started in the Galilean period and is, therefore, somehow connected with the decline of scholastic philosophy. The common feature of medieval and of mechanistic physics seems to me to be that their principles seemed to have a certain plausibility by themselves. Medieval science derived all observable phenomena from the principle that they are somehow analogous to the well‑known phenomena in a living organism. Seventeenth‑ and eighteenth‑century science, in turn, preferred the analogy to simple mechanisms which are familiar to us from our everyday life experiences. Not only were the principles confirmed by demonstrating that the conclusions drawn from them were in agreement with observed facts, but also the principles themselves used to be directly confirmed by a kind of short cut. If a law of physics was in agreement with the organismic or mechanistic analogy, this agreement accounted for a certain degree of confirmation. A more careful analysis would probably lead to the result that the organismic as well as the mechanistic principles of science drew their plausibility from the fact that they seemed to reflect faithful pictures of our everyday experience. In scientific theories a long chain of intellectual and experimental work connects the principles of a theory with the protocols of observation. The organismic or mechanistic principles, however, could, it was believed, be confirmed directly by a very obvious type of experience. Everybody knew that a piece of earth that is dropped falls to the ground. It therefore seemed obvious that the earth as a whole could not remain suspended in space and circulate around the sun. It was to fall to the center of the universe like a small piece of earth. This tendency toward the center as the natural place of the heavy element earth seems to be a principle of physics with firm roots in everyday experience. In the same way, a mechanistic principle like the law of inertia seemed to be directly confirmed by the most familiar facts of our daily experience.

In this sense we can say that the organismic as well as the mechanistic physics were based on principles that could be interpreted directly in terms of everyday experience. The validity of the principles did not need any further confirmation by the development of more refined methods of observation and theoretical argument. Science, in those days, was based on principles to which an eternal validity was ascribed. They appeared to be "rational" or "reasonable" or, expressed in a loose way, "logical." The philosophers of the eighteenth‑century enlightenment still felt that theirs was a world view, provided completely by reason. The age of enlightenment shared this belief in reason with the "dark" Middle Ages. This is the historical point that Carl Becker wanted to make in his book. In the nineteenth century, according to Becker, a drive was initiated to tell reason that it had to know its place, which was considerably lower than previously; but in the twentieth century, science emancipated itself from reason altogether.

After these few remarks about the place of reason in the science of past centuries, it is obvious what the characteristic of twentieth‑century science is. With the rise of non‑Euclidean geometry, the physics of relativity and the mechanics based upon the de Broglie waves, the basic principles of physical science were no longer a direct formulation of our everyday experience; they were no longer obvious and plausible to common sense. Their only justification consisted now in their property of yielding observable facts by means of a chain of logical conclusions, The illusion disappeared that the principles of science were of eternal validity or could at least be interpreted as conclusions from such principles. Therefore, one ought not to say that twentieth‑century science has no use for logic but rather that it has no use for metaphysics. To interpret the principles of science as results of our common sense leads to the opinion that they are self‑evident and cannot be refuted by further empirical checking, This belief is the very core of the metaphysical interpretation of science.

One should rather reformulate Becker's argument as follows: In the period of Galileo and Newton the firm belief in the metaphysical foundation of science faded a little. Mechanistic physics was less imbued with metaphysical argument than was medieval, organismic, science. But the belief in Newton's laws as results of the simple experience of everyday life had been bolstered up during the eighteenth and a great part of the nineteenth centuries, although it was by no means a common opinion among Newton's contemporaries. With the new physical theories of the twentieth century—non‑Euclidean geometry, relativity and the quantum theory—the belief practically disappeared that the basic principles of physics ought to be plausible according to the criteria of common sense. The metaphysical conception of science lost ground and the logico‑empirical interpretation of the scientific method became the method that was actually used. We can see in the evolution of science from the seventeenth century to the twentieth century the gradual decline of metaphysics in favor of a positive conception, if we want to use the terminology of August Comte.

I am using the term metaphysics here with a positive and precise meaning: direct interpretation of the basic principles of science in terms of common sense or everyday experience. I think that it is not sufficient to  characterize metaphysical statements as "meaningless." There are a lot of meaningless statements that are not at all "metaphysical." By using the term in the way I suggested we can probably cover all the statements which have been made with the claim to be metaphysical. Metaphysics, according to our way of speaking, is certainly meaningless from the scientific viewpoint because the terms "true" or "false" cannot be applied to these statements. Charles Morris speaks, however, of a "metaphysical discourse." I agree with him in the sense that I regard metaphysics as a direct interpretation of scientific principles in terms of the language of everyday life experience. "Interpretation" means translation. Metaphysics attempts a translation of the basic principles of science, but not according to a strictly fixed dictionary; the univocal relation between a term and its translation has been replaced by an analogical relation. But we cannot tell by any exact criterion what is a "correct" analogy. We shall elaborate this conception of metaphysics more exactly toward the end of this paper.

If one even agreed that the place of logic and reason in science had not lost in importance in our time, a great many people, including even scientists and philosophers, would claim that the actual advance of science had been promoted not by logic and reason, but rather by intuition and metaphysics. Logic, so this argument goes, is useful only for systematizing scientific knowledge and statements that are already known; it can never be of any help in finding new statements and, still less, fundamentally new theories. This assertion has been repeated again and again, but it cannot stand a critical test. If we include in logic not only syntax of language but also the theory of meaning, semantics, one could easily make a good case for the assertion that the "experimental" theory of meaning, which has been advocated by pragmatists and logical empiricists, is the very basis of twentieth‑century physics. Not only does it provide the method of presenting this physics systematically, but one can also point out that the authors of this new physics made explicit use of this theory of meaning. This theory was one of the historical roots of the new physical theories.

It is hardly necessary to stress the fact that this theory of meaning guided Einstein in his "restricted" theory of relativity as well as in his general theory. When Einstein in 1905 introduced his interpretation of the Lorentz transformation which was the essence of the special theory of relativity, he pointed out that the statement "two events at a spatial distance take place at one and the same time" cannot be used to derive any observable fact. Therefore, this statement cannot be a part of any physical theory. Einstein clearly understood that this statement needs in addition a semantic rule or operational definition. In the development of the general theory we again have statements of the type "the rotation of this body is responsible for a centrifugal force." This statement is regarded as a legitimate statement in Newton's mechanics. But Mach pointed out that such a statement cannot be used to derive observable facts because it does not contain any rule by which one can check by observations whether a body is rotating. This remark was one root of the general theory and the theory of gravitation. This correlation between the theory of meaning and Einstein's relativity was fully appreciated by Bridgman. [1]

It is perhaps less generally known that the development of quantum mechanics also has its historical root in the empirical theory of meaning. The decisive turning point in the history of quantum mechanics was Heisenberg's paper [2] of 1925. Until this date the state of the quantum theory was characterized by Bohr's theory of the hydrogen atom, by the Keplerean orbits of the electrons around the nucleus which obeyed the Newtonian laws with certain restrictions. Bohr's theory superimposed upon Newton's laws the quantum laws, according to which only some specific orbits could be performed without giving rise to a radiation which according to classical physics would destroy the orbit. This pre‑1925 state of quantum theory can be described as a "Newtonian mechanics patched up by quantum laws." Heisenberg, in the paper mentioned, was the first to replace Newtonian mechanics, in his application to the movement of electrons, by a completely new physical theory which became known under the name “quantum mechanics." His starting point was exactly the experimental theory of meaning. He says:

In this paper I am going to attempt to find foundations for a mechanics of quantum theory. This mechanics is based exclusively on relations between quantities that are observable in principle. . . . As it is well known, a very relevant objection can be raised against the formal rules that are used in quantum theory for the computation of observable quantities (e.g., the energy of the hydrogen atom). These rules of computation contain as an essential part relations between quantities that are unobservable in principle (e.g., the position and the period of an electron). Therefore, those rules are without any intuitive foundation if one does not expect that those quantities which are at present unobservable will eventually be accessible to experimental observations. . . . Under these circumstances it seems advisable to make the attempt to build up a quantum mechanics that is analogous to the classical mechanics but in which only relations between observable quantities occur.

Accordingly, Heisenberg introduced not the position of the electron but the Fourier coefficients of the radiation that is emitted by the atom as a result of the Keplerean motions of the electron. These Fourier coefficients developed later into Heisenberg's matrices and Schrödinger's wave function.

Hence, the heuristic value of the "experimental theory of meaning" is proved by the actual history of twentieth‑century physics. It turned out later that the actual formulation of the principles of the new physics could not be achieved in this direct and simple way. It has not been sufficient to use only observable quantities as terms in these principles. One had to proceed in a more indirect and complex way. This has been the case in relativity as well as in quantum mechanics.

If we consider, as an example, Einstein's theory of gravitation, the principal part of this theory is the differential equations of the gravitational field. They contain mathematical symbols: the four general coordinates in the space‑time continuum, the ten potentials of the general gravitational field, etc. One cannot lay down practical semantic rules for these symbols as they stand in the general field equations. But by mathematical conclusions we can derive results from these field equations which can be translated by means of feasible semantic rules into descriptions of actually observable facts. If, for example, we derive the bending of light rays in the gravitational field of the sun, we obtain statements in which the general coordinates in the space‑time continuum can be connected in a clear‑cut way with the spatial and temporal distances which are measured by our traditional ways of measuring length and time intervals. But such a connection cannot be laid down in full generality. If, for example, we consider the statement “one and the same factual event can be described by different sets of values of the general coordinates in the four‑dimensional continuum," we can hardly lay down practical semantic rules by which the operational meaning of this statement can be defined in a simple way.

P. W. Bridgman maintained, therefore, that Einstein's general theory of relativity does not fulfill the requirements put forward by the experimental or operational theory of meaning. According to these requirements, one must give explicit operational definitions of all terms that occur in the general principles of a theory. While, according to Bridgman, the restricted theory of relativity was a brilliant example of the use of operational definitions, he thinks that the general theory has violated the requirement for such definitions. [3] There is no doubt that on the way from the restricted theory of relativity to the general theory the structure of a physical theory, as envisaged by Einstein, has changed in a noticeable degree. [4]

The restricted theory of relativity nearly fulfilled what has been called Mach's "positivistic" requirement, according to which all principles of physics should be formulated by using only observable quantities as terms. The general theory made use of this requirement only as a heuristic principle, as a hint of how to build up the system of fundamental principles. These principles themselves, however, fulfilled the "positivistic" requirement only in an indirect way. Einstein replaced it, consciously and deliberately, by a weaker requirement: it was merely required that from these principles mathematical conclusions could be drawn that were connected by semantic rules with statements about observable facts.

Albert Einstein, in his Herbert Spencer Lecture given at Oxford in 1933, speaks about this change in the way in which the abstract principles of physics are connected with the observable facts. This boils down to a change of the place where the semantic rules or operational definitions are attached to the abstract principles. Einstein speaks of "the ever‑widening logical gap between the basic concepts and laws on the one side and the consequences to be correlated with our experiences on the other." He insists merely on the requirement that some results or propositions in the system be connected with observational statements by means of semantic rules. In his "Remarks on Bertrand Russell's Theory of Knowledge," [5] Einstein says:

In order that thinking might not degenerate into "metaphysics" or into empty talk, it is only necessary that enough propositions of the conceptual system be firmly enough connected with sensory experiences.

According to this new conception, the sentences that have to be connected with sense observations by semantic rules are no longer the general abstract principles (such as the law of conservation of energy) but some special conclusions drawn from these principles. The "positivistic requirement" now means that there must be some consequences of the general principles which can be translated into statements about sense observations. The general principles themselves are the product of mathematical and logical imagination which has to be checked by applying the "positivistic" or "operational" requirement.

The nature of a scientific theory in this sense can be understood even more precisely if we consider the new "unified field theories" proposed by Einstein and Schrödinger. A "unified field of force” is to be constructed which contains the gravitational, the electromagnetic and the nuclear field as special cases. Schrödinger, for example, introduces sixty‑four symbols which are the components of this unified field. He does not lay down any semantic rules for these sixty‑four quantities. But if this theory is to be of any scientific value, he has to assume, as a matter of course, that some special relations can be mathematically derived from the principles which can be connected with observable facts. He hoped, for example, that it could be derived that a rotating mass which obviously produces a rotating gravitational field would entail a magnetic field. One would be able to account in this way for terrestrial magnetism.

From a psychological viewpoint Einstein describes this way of producing theories by free imagination in a letter to the French mathematician Hadamard. [6] According to this new conception, it is true that physical theories are the product of free imagination, if we take the word "free" with a grain of salt. But it must not be concluded that these theories are products of metaphysics. For these theories are subjected to the operational or experimental criterion of meaning, though in a more indirect and complex way. The criterion of truth remains ultimately with the checking by sense observations, as the older "positivists" claimed. But we know now that this checking is a more complex process than it was believed by men like Comte and Mach to be.

If we applied the name "metaphysics" to a system of statements the "truth" of which is judged according to the experimental criterion of meaning, there would be no distinction between science and metaphysics. We have, therefore, to reserve the word "metaphysics" as a characteristic of systems the truth of which is decided on other grounds.

In metaphysics a statement or a system of statements is regarded as "true" if our common sense understands the validity of the principles immediately without having to draw long chains of conclusions from these principles and without checking some of these conclusions against our observations.

Such a metaphysical interpretation of twentieth‑century physics was given, for instance, by Eddington. [7] He claimed that the validity of our physical theories can be demonstrated by what he called "epistemological arguments." This meant in his language that the principles of physics have to meet some requirements emerging from common sense. And these requirements are sufficient to determine our physics to such a degree that even the number of the electrons in the whole universe can be derived mathematically. Eddington's argument is actually "metaphysical." We could call it "epistemological" only if we regarded epistemology as a part of metaphysics. The point is that Eddington derives his system of physics from everyday experience and not from scientific experiments which are needed to check the results of a long chain of conclusions from the principles. The requirements of "common sense" actually are that these principles should be a convenient description of our everyday experience.

Certainly, men like Einstein and Schrödinger advanced their principles by following some requirements of simplicity or beauty which may also be regarded as requirements of common sense. But they would never claim that the validity of the principles could be proved without checking the conclusions drawn from these principles by physical experiments.

If we want to compare the respective places of logic and metaphysics in the actual advancement of science, we can point out two ways in which logic has been instrumental in the advance of twentieth‑century science. We have already described the heuristic value of the experimental theory of meaning which played a decisive role in the rise of relativity as well as of quantum theory. But logic has also played in another way a guiding role in the advance of twentieth‑century science. And this way is connected with formal logic, or what we may call "logical imagination." Einstein, in particular, has described repeatedly how the building of formal systems of symbols, the successive demolition, rebuilding and alteration of these symbolic structures, has had a great bearing upon the advance of new physical theories. In his letter to Hadamard, Einstein gives a psychological description of his creative work. He insists that the essential part in creative thinking is the free play with symbols. In his Herbert Spencer Lecture (1933), Einstein says:

Experience, of course, remains the sole criterion for the serviceability of mathematical construction for physics, but the truly creative principle resides in mathematics.

This means, obviously, that the creative process in theoretical physics consists, in some important ways, in the creation of symbolic or formal systems by a kind of "logical imagination." Among the systems created in this way, experience is responsible for the natural selection that determines which system is the fittest for survival and which has to be dropped.

In order to appreciate the place of logic, we have to consider the shift in the conception of a physical theory that has developed from the times of Mach to the times of Einstein. The operational definitions or semantic rules are now no longer applied to the general principles themselves but to some conclusions drawn from them. However, this distinction should not be overstressed. It would be erroneous to believe that men like Ernst Mach actually believed that all principles of physics were direct descriptions of experimental facts. In his paper on the role of comparison in physics, Mach distinguished very precisely between "direct description" and "indirect description." The latter, he said, is also called "physical theory." He gives as an example the wave theory of light. In this theory there is certainly no explicit operational definition of the "light vector."

As for the heuristic value of metaphysics, we may quote one of the most prominent contemporary advocates of metaphysics, Jacques Maritain. He says bluntly:

It is true that metaphysics brings no harvest in the field of experimental science . . . Its heuristic value, as the phrase goes, is nil . . .

It cannot be of any help in promoting scientific research. He continues:

This universe in which metaphysics issues . . . is not intelligible by dianoetic or experimental means, it is not connatural to our powers of knowledge, it is only intelligible to us by analogy.

This means in the usual language of the scientist: In metaphysics we do not use either logical (dianoetic) or experimental argument, but we interpret the principles of science in a metaphoric language. Metaphysics attempts to interpret the general principles of science in a way that is plausible to our common sense. The principles then become analogous to the laws of everyday experience; they duplicate them, but on a higher level. The law of conservation of energy becomes metaphysically plausible because it is bolstered up by the well‑known experience that objects of the physical world (stones, animals, etc.) cannot disappear. We transfer the conception of "disappearing" to an “object" called "energy" which is certainly not a physical object in the sense that a table is such an object. Therefore "disappearance of energy" can only be understood by "analogy" with the disappearance of a stone. According to Thomistic metaphysics, in an expression like “the being of God" or "the being of a spirit" the word "being" does not mean the same thing as it does in "the being of a stone." The meaning of "being" on these "higher levels" is only understandable by analogy, not by any direct operational definition or semantic rule.

I would not even go as far as Maritain in denying any heuristic value in metaphysics. The description by analogies can occasionally be of some psychologic value in setting up new principles. This was even recognized by so staunch an opponent of metaphysics as Ernst Mach in his paper on the role of comparison in physics.

But if we analyze a little further the nature of the actual metaphysical interpretations of physics, we soon notice one characteristic of them that can easily become prohibitive to any future advance of physics.

I shall not now go into an elaborate discussion of this nature of metaphysics. I shall restrict myself to showing by some examples that philosophers who went in for both metaphysics and science frequently pointed out this characteristic of metaphysics. They stressed the point that metaphysics is an attempt to interpret the general principles of physics in terms of the language of our everyday life. In this way it is possible to muster the support of our everyday experience to make those principles plausible.

A very characteristic example is the French philosopher E. le Roy. His goal was to prove on the basis of contemporary science that room is left for metaphysics and even for a metaphysical conception of religion. He especially liked to make use of the ideas that the famous French scientist, Henri Poincaré, had advanced about the logical status of science. Just at the turn of the century (1899) Le Roy published in the Revue de Métaphysique et de Morale a paper in which he says:

Science departs from common sense and does not join it in its development as science proper. Thus science by itself does not close the cycle of knowledge and does not realize the unity of knowledge. Science needs therefore a prolongation and this will be philosophy . . . in one way science itself is a prolongation of common sense.

Common sense, science, philosophy, common sense form a cycle.

The term "philosophy" refers here, of course, to the same thing as “metaphysics." In these words we see clearly the place assigned to metaphysics. Science departs from common sense by using a different conceptual scheme; words are used in a different way. Science even introduces expressions that have no meaning in the language of common sense. This becomes particularly clear if we consider the language of twentieth‑century physics. The theory of relativity uses expressions like "one and the same object has different lengths relative to different systems of reference," and quantum mechanics uses expressions like "if the position of a particle is a definite one, its velocity is always indeterminate."

According to Le Roy's cycle, there are two ways of connecting science with common sense: the direct, which we may call the "scientific" one, and a second, which connects science with common sense by means of metaphysics. In the scientific way, the statements of science are interpreted by means of operational definitions as statements about observable facts, and every observable fact in physics can be expressed in the language of common‑sense experience.

Einstein's statement that a rigid body has different lengths with respect to different systems of reference can be connected with common‑sense statements in two ways. We can describe directly the way in which the length of one and the same rigid body can be measured by putting end to end yardsticks that have different velocities relative to the body. In contrast to this scientific connection, the metaphysical interpretation would be: the statement that "one and the same length is estimated differently by different observers" reminds us of the experience of everyday life that one and the same length is estimated differently by different observers. This "subjectivity" of every judgment about length seems analogous to Einstein's contention that the length depends on the system of reference. Therefore, Einstein's statement is interpreted as claiming the "subjectivity" of human statements about length. This is again in line with some statements of idealistic philosophy.

We also find similar views in writings of other philosophers of scientific background. I mention A. N. Whitehead, C. S. Peirce, and in particular, a Thomistic philosopher, H. V. Gill. He regards the metaphysical interpretation, as many people do, as a result of "intuition." But he realizes somehow that it is actually an application of commonsense judgment to the principles of science. He noticed that the scientists (he calls them "specialists") will not do as well in finding these interpretations as the "man in the street" who relies on common sense only. Gill says: [8]

The fact that a few "specialists" call in question some intuition generally accepted by men does not furnish a valid reason for doubting its truth. The specialist is indeed perhaps the one whose view on first principles should be taken most cautiously.

The metaphysical interpretation is actually a particular kind of semantic approach; it is a translation into common‑sense language. We follow Charles Morris's excellent analysis of the "metaphysical discourse." [9] It is a "formative discourse" like the mathematical, logical, grammatical and rhetorical discourse. A criterion of truth (similar to the criterion of "scientific truth") cannot be applied. The metaphysical discourse plays a role in organizing human behavior and has, therefore, "significance." The present paper means to be more specific and to describe the language used in metaphysics as the result of an attempt to interpret the general laws of science by using common‑sense expressions. The "materialistic" and "idealistic" interpretations of science owe their appeal to the common‑sense meaning of the words "matter" and "mind" and not to their scientific meaning, which can hardly be stated precisely.

From these considerations we can easily derive a sound judgment about the role of metaphysics in the actual advance of science. What we call in a vague way "common sense" is actually an older system of science which was dropped because new discoveries demanded a new conceptual scheme, a new language of science. Therefore the attempt to interpret scientific principles by "common sense" means actually an attempt to formulate our actual science by the conceptual scheme that was adequate to an older stage of science, now abandoned.

According to these considerations, one can easily estimate what the role of metaphysics in the advance of science has been. To believe that some "metaphysical interpretation" may tell us the "truth" about the “real world" means in practice to believe that the conceptual scheme of some older stage of science is necessarily the scheme to be used for all the future. This belief is, certainly, in a way stimulating: it encourages the scientists in their attempt to stick to a unified scheme into which every new discovery has to be fitted or perhaps even squeezed. To achieve a unified scheme in all fields of physics is certainly a goal that has been occasionally of great heuristic value. The most prominent example is, I think, the attempt to interpret all physical phenomena by Newton's laws of motion. This attempt has led to great successes in optics; we owe to it the corpuscular theory of light as well as the wave theory. We owe to it almost all atomistic theories in their first stages.

But toward the end of the nineteenth century the physicists came more and more to recognize that there are phenomena which can be fitted into this "mechanistic" pattern only very artificially and incompletely. Hence the heuristic value of this "mechanistic" goal faded as time went on. Nevertheless the belief remained that only physical theories which can be derived from Newton's laws of motion satisfy human desire for "understanding" nature. In this stage this belief became a purely metaphysical creed. Some maintained, for example, that Einstein's modification of Newton's laws had to be rejected for metaphysical reasons. This actually means only that Einstein's mechanics cannot be derived from Newton's laws of motion. It is a historic fact that a great many physicists preferred to say that they rejected relativistic mechanics not for metaphysical reasons but for reasons of "common sense." Both types of reason for rejecting new physical theories, as the argument in this paper shows, are really one and the same thing expressed in two different ways.

The rationalistic metaphysician rejected new theories on the ground of "reason," the empiricist‑metaphysician rejected them on the basis of "common sense." In my view, both types of rejection have a common origin: the belief in the interpretation of new theories by using the language of older theories.

Examples are abundant. I mention only cases in which "common sense" prevented the acceptance of new physical theories, because the scientists are more easily caught by "common sense" than by avowed metaphysics.

The father of empiristic philosophy, Francis Bacon, rejected the Copernican theory for not being in agreement with common sense; the leader of nineteenth‑century British empiricism, Herbert Spencer, argued that the total mass of a system of material bodies cannot depend on their distribution in space. August Comte, the father of "positive philosophy," predicted that no mathematical theory of chemical phenomena will ever be advanced because our common sense tells us that the chemical processes are fundamentally different from physical processes. If we consider to what degree all these predictions have been refuted by the actual advance of science, we can learn two things: metaphysics has very often been an obstacle to the advance of science, and second, if we hear today that biology will never become a science in the sense that mathematical physics is, or that sociology can never use scientific methods, we shall hesitate in maintaining a smug belief in these assertions.


1  P. W. Bridgman, The Logic of Modem Physics (New York: Macmillan, 1927). [—> main text]

2  W. Heisenberg, Zeitschrift für Physik 33, 879 (1925). [—> main text]

3  P. W. Bridgman, op. cit.; The Nature of Physical Theory (Princeton University Press, 1936). [—> main text]

Albert Einstein: Philosopher‑Scientist (vol. 7 of The Library of Living Philosophers, P. A. Schilpp, ed.; Evanston and Chicago: Northwestern University, 1949). [—> main text]

The Philosophy of Bertrand Russell (vol. 5 of The Library of Living Philosophers, P. A. Schilpp, ed.; Evanston and Chicago: Northwestern University, 1944), p. 289. [—> main text]

6  J. Hadamard, An Essay on the Psychology of Invention in the Mathematical Field (Princeton University Press, 1945). [—> main text]

7  A. S. Eddington, The Philosophy of Physical Science (New York: Macmillan, 1939). [—> main text]

8  H. V. Gill, Fact and Fiction in Modern Science (Dublin: Gill, 1943), p. 21. [—> main text]

9  C. Morris, Signs, Language and Behavior (New York: Prentice‑Hall, 1946), pp. 175 ff. [—> main text]

SOURCE: Frank, Philipp. Modern Science and Its Philosophy. Cambridge, MA: Harvard University Press, 1949. Reprint: New York: George Braziller, 1955. Chapter 16, The Place of Logic and Metaphysics in the Advancement of Modern Science, pp. 286-303.

Modern Science and Its Philosophy: Contents

Vienna Circle, Karl Popper, Frankfurt School, Marxism, McCarthyism & American Philosophy: Selected Bibliography

Positivism vs Life Philosophy (Lebensphilosophie) Study Guide

Salvaging Soviet Philosophy (1)

Home Page | Site Map | What's New | Coming Attractions | Book News
Bibliography | Mini-Bibliographies | Study Guides | Special Sections
My Writings | Other Authors' Texts | Philosophical Quotations
Blogs | Images & Sounds | External Links

CONTACT Ralph Dumain

Uploaded 4 February 2006

Site ©1999-2023 Ralph Dumain