Campbell, Norman Robert (1880–1949)

views updated May 29 2018

CAMPBELL, NORMAN ROBERT
(1880–1949)

Norman Robert Campbell, the English physicist and philosopher of science, was educated at Eton. From Eton he went as a scholar to Trinity College, Cambridge, and became a fellow there in 1904. From 1903 to 1910 he also worked as a research assistant at the Cavendish Laboratory, whose director, the celebrated J. J. Thomson, became the most important inspiration of his scientific work. In 1913 he became an honorary fellow for research in physics at Leeds University, but he left this post after the war and from 1919 to 1944 was a member of the research staff of the General Electric Company.

The writers who seem to have influenced him most are Ernst Mach and Henri Poincaré, apart from classical authors such as William Whewell, John Stuart Mill, and W. S. Jevons. On the other hand, such philosophers as Bertrand Russell and Alfred North Whitehead came too late to have much effect on him; the main outlines of his thought developed during the first decade of the century, and there are only occasional references to their writings.

Campbell exhibited the very rare combination of competence in both physics and philosophy, but while he preferred to think of himself primarily as an experimental physicist, it is as a philosopher of science that he made his mark. This point is brought out in the writings of F. P. Ramsey, R. B. Braithwaite, and Ernest Nagel, although these concentrate largely on the formal parts of Campbell's doctrines and pay scant attention to the more valuable contributions that he made to certain methodological ideas, particularly that of analogy. These philosophical views, shaped by Campbell's actual experiences and ideas as a physicist and expositor of physical theories, were meant to be construed as answers to intellectual pressures and problems that confronted him in the years that saw the rise of the twentieth-century atomic theory on the one hand and relativity and quantum mechanics on the other. In philosophy of science, his most important contributions were in the fields of the logic of theory construction and (to a lesser extent) the principles of physical measurement.

Philosophy of Theory Construction

Campbell's views were stated in systematic form for the first time in a popular book, The Principles of Electricity. Thereafter they were developed, with minor changes of emphasis and greater attention to the nature of "mathematical theories," in Physics: The Elements. In contrast with the usual textbook approach, his views were deeply interwoven with, and at times even explicitly discussed in, his more formal scientific treatises.

concepts and ideas

Campbell distinguishes sharply between the laws and theories of a science. In the case of laws, the constituent terms (Campbell calls them concepts) designate entities whose magnitudes may be determined more or less directly by instrumental means; they are not unlike what later came to be called operational concepts. The explanatory part of theories, the hypotheses, involve terms that Campbell calls ideas. These lack the instrumental relations of concepts, for a variety of reasons that Campbell does not always clearly distinguish.

Sometimes the ideas refer to the unobservable infrastructure of a physical system, as in the case of the atoms and electrons of modern electrical theory or, more properly (as Campbell points out), to their adjectival aspects, such as their mass, velocity, and momentum. At other times, the ideas pertain to such interstructural devices as Michael Faraday's lines of force, or the carriers of the transfer of electrical and optical phenomena, such as light waves, light corpuscles (photons) or even the "aether," considered the substantival carrier of electromagnetic energy. (Infrastructural entities are unobservable in a sense different from interstructural ones, but the question is controversial.) A third case in which theories are said to involve unobservables is that of geological and evolutionary theories. And there is yet another case, for Campbell denominates certain notions "ideas" because they involve an amount of idealization and abstraction to which no physical entities could correspond. The most frequent and important cases are those ideas which involve infinitesimals, such as the differential coefficients in James Clerk Maxwell's equations or François Marie Charles Fourier's theory of heat.

It follows from the nature of ideas that the hypotheses in which they occur are not directly testable. Their function consists merely in systematically relating a set of corresponding laws, and, through extensions of the theory, in foreshadowing further laws and experiments. This foreshadowing is sometimes negative, for when the ideas are too narrowly framed, they demand not only extension but also the formulation of additional concepts and theories.

"dictionary" of a theory

Since the ideas of the hypotheses lack operational meaning, and since their deductive development can, in the first place, yield only statement forms containing either ideas or combinations of them, it is necessary to add certain rules (a kind of "dictionary") that will coordinate the ideas with those operational concepts which occur in the laws to be explained. Of course, not all ideas need dictionary entries. In the beta-ray theory, for instance, the velocity, v, of the hypothetical electrons means "the quantity that is defined by the relation F = e [X +(v · H )]." This expression, however, is a hypothesis in Campbell's sense because v never occurs either alone or in combination in the testable derivations at all.

"mathematical" theories

All this provided Campbell with a means of distinguishing so-called mathematical theories from nonmathematical ones. In the former, each and every idea is separately coordinated with a corresponding concept by means of a dictionary entry. It follows that whether a theory is of the mathematical type depends partly on historical accidents: Maxwell's theory became a mathematical theory only after Heinrich Rudolf Hertz's experiment had demonstrated the existence of the displacement current.

Nonetheless, ideas so far have no meaning apart from their use in hypotheses and their coordination with concepts. In the mathematical cases this is often forgotten, but in the nonmathematical cases this fact is more difficult to overlook. Because of the lack of independent significance of ideas, Campbell held that a theory is not a real explanation unless certain additional requirements are satisfied. One of his reasons for this view was that it is always possible to construct an indefinite number of hypotheses that would account for a set of laws. In the case of mathematical theories, the additional element of consolidation that Campbell suggests is the regulative feature of simplicity and aesthetic elegance—for instance, through symmetrical arrangements of the parts of a theory. (Thus, it was the introduction of Maxwell's displacement current into the original equations of André Marie Ampère and Faraday that produced a symmetrical set of equations regarding the relations between the electrical and magnetic phenomena for the case of open circuits.) Furthermore, the hypotheses are not entirely arbitrary because their ideas mirror the corresponding concepts of the laws. There is, according to Campbell, a sort of analogy between ideas and concepts (Physics: The Elements, p. 141).

analogy

Analogy plays a more central role in the case of the nonmathematical theories. As we have seen, their ideas frequently cannot be clarified at all by the concepts that occur in the laws. According to Campbell, it is an analogy of the hypotheses and their ideas with corresponding laws and concepts of some testable field of science that imparts the missing element of significance and logical strength to the theory. It follows that analogies are not merely aids to the establishment of theories; "they are an utterly essential part of theories, without which theories would be completely valueless and unworthy of the name" (ibid., p. 129).

Campbell's point is that "a theory is not a law" (ibid., p. 130); that hypotheses are, from the nature of the case, never directly testable; and, hence, that their addition to the corpus of scientific knowledge would make no difference to science at all if it were not for some additional features that make the hypotheses significant. He dismisses the fact that they supply a systematic relation between the laws of the theory on the grounds that an infinity of such hypotheses can be constructed.

Campbell's positive grounds for the necessity of analogies are of various kinds. The fundamental reason is that since hypotheses are not directly testable but are only instruments for deductive development, possessing a purely formal content, they lack the sort of meaning required for genuine explanatory power: Only analogy can supply this. Another ground of a more heuristic nature is that analogies aid in the extension of theories, especially when a new field is grafted onto the dictionary of an existing theory (as when optical conceptions were added to Maxwell's generalization of the electrical theories of Ampère and Faraday).

As mentioned, however, analogy must be supplemented by additional criteria, which are clearly needed for dealing with mathematical theories. These criteria are largely derived from Campbell's actual experience with the theories with which he had been dealing in his physical textbooks. In addition to simplicity and aesthetic elegance, there is "simplification in our physical conceptions," such as was produced by the early theories of Faraday, Thomson, and Hendrik Antoon Lorentz. Campbell insists on the importance of such regulative conceptions precisely because "scientific propositions are [not] capable of direct and irrefutable proof." An additional criterion is the "anticipative force" of a theory—for instance, the suggestiveness of Faraday's lines in the direction of the existence of electromagnetic radiation, of a motion that is displaced in time, with a given velocity, in empty space.

Finally, another regulative criterion is that of importance, or depth, of the ideas involved. This is invoked particularly in those cases where analogy is barely a relevant consideration, as in such mathematical theories as Maxwell's, or Albert Einstein's special theory of relativity.

Methodological Contributions

Campbell's clear account of the logical structure of a theory, with its hypotheses, laws, and dictionary, offers an elegant means of formalizing the place of ideas (theoretical concepts) within theories. He emphasizes also the logical gap between hypotheses and laws even in cases where its existence had previously been practically overlooked—the mathematical theories. He uses this fact to question Mach's preference for such theories (called phenomenological by Mach), on the grounds that they employ hypotheses and hypothetical ideas just like any other theory. (Whether this does sufficient justice to the difference between the two types of theories must be left an open question.) The theoretical nature of such substantival entities as atoms and electrons seems to differ from that of lines of force on the one hand and, say, from the entropy functions on the other, in deeper ways not caught by Campbell's criteria of ideas.

The fact that the systematizing power of hypotheses is an insufficient criterion of their truth or explanatory power introduces the remaining feature of Campbell's doctrine—such regulative notions as the existence of a strong analogy, of simplicity, symmetry, anticipative force, and, finally, of importance. The most interesting of these is analogy, which in the end emerges as a metaphysical device in terms of which to formulate the special aspect of those theories that involve unobservables. The "absolute necessity" for an analogy is the result of the emasculation of the semantic power of hypotheses, coupled with the consideration that this emasculation entails the introduction of a special constraint that prevents such hypotheses from being mere arbitrary formulas.

Theory of Measurement

The second part of Physics: The Elements is a detailed discussion of the principles of physical measurement; this, like most of Campbell's ideas, was already contained in embryo in The Principles of Electricity (Ch. 2). His interest in measurement is not altogether removed from his main philosophical preoccupations mentioned so far. Just as he was concerned with a clear delineation of laws from theories, he was equally firm in stating the differences as well as the relations between laws and definitions. In Measurement and Calculation Campbell defines measurement "as the assignment of numerals to present properties in accordance with … laws." Thus, every measurable property must have a definite order; the systems to be measured must be capable of "addition," but what operation is considered "addition" must be carefully specified in a given situation; and whether the resultant quantities yield consistent measurements is a matter for lawlike experience. Campbell points out that the specification in question is usually tacitly adopted ab initio and is, indeed, often suggested by theory and the relevant analogy. Hence, he believes that "no new measurable quantity has ever been introduced into physics except as the result of the suggestion of some theory" (The Principles of Electricity, p. 41).

Bibliography

Campbell's works include Modern Electrical Theory (Cambridge: Cambridge University Press, 1907; 2nd ed., 1913); The Principles of Electricity (London, 1912); Physics: The Elements (Cambridge, U.K.: Cambridge University Press, 1920), which was reprinted as Foundations of Science (New York: Dover, 1957); Measurement and Calculation (London: Longmans, Green, 1928); and Photoelectric Cells, written with Dorothy Ritchie (London: Pitman, 1929; 3rd ed., 1934).

As supplementary chapters of Modern Electrical Theory, the following monographs by Campbell were published in the Cambridge Physical Series: Series Spectra, Suppl. Ch. 15 (Cambridge, 1921); Relativity, Suppl. Ch. 16 (Cambridge, 1923); and The Structure of the Atom, Suppl. Ch. 17 (Cambridge, 1923).

For works with detailed references to Campbell, see Brian D. Ellis, Basic Concepts of Measurement (Cambridge: Cambridge University Press, 1966), Chs. 4, 5, and 8; Mary B. Hesse, Models and Analogies in Science (London, 1963), Ch. 1; and George Schlesinger, Method in the Physical Sciences (London: Routledge and Paul, 1963), Ch. 3, Sec. 5.

Gerd Buchdahl (1967)

Encyclopedia of Philosophy Buchdahl, Gerd

Campbell, Norman Robert

views updated May 21 2018

Campbell, Norman Robert

(b. Colgrain, Dumbarton, Scotland, 1880; d. Nottingham, England, 18 May 1949),

physics, philosophy of science.

Campbell was educated at Eton and became a scholar of Trinity College, Cambridge. In 1904 he became a fellow of Trinity College, where he worked mainly as a student of J. J. Thomson on the ionization of gases in closed vessels. In addition to performing successful experimental research in this field, he established, in collaboration with A. Wood, the radioactivity of potassium.

Campbell was then appointed to the Cavendish research fellowship at Leeds where, in 1913, he became honorary fellow for research in physics. While at Leeds he continued research on the ionization of gases by charged particles and on secondary radiation. After the start of World War I, he joined the research staff of the National Physical Laboratories, working on problems concerning the mechanism of spark discharge in plugs for internal combustion engines. Reports of this work were submitted to the Advisory Committee for Aeronautics.

In 1919 Campbell joined the research laboratories of the General Electric Company, Ltd. As well as continuing his work on electrical discharge in gases, he worked on photoelectric photometry and color matching, the standardization and theory of photo-electric cells, statistical problems, the adjustment of observations, and the production of “noise” in thermionic valves and circuits. He published nine books and over eighty papers. In 1912 he married Edith Sowerbutts; the couple adopted two children. The last fifteen years of his life were spent in retirement—in ill health.

Although Campbell distinguished himself as an experimental physicist, he devoted himself to careful study of both the theoretical and philosophical aspects of his science. Profoundly influenced by J. J. Thomson and the ideas of Faraday and Maxwell, he was basically a proponent of a mechanical view of physics. His reasons for this inclination were sophisticated and based not only on the “intrinsic interest” of mechanical explanations, natural enough for an associate of Thomson, but also on profound deliberations on the nature of scientific knowledge. His major work was done at a time when physical science was changing in a very radical way and his continuing examination of this process can be seen in the progressive editions of his Modern Electrical Theory.

Although Campbell saw himself as an experimental physicist, it is as a philosopher of science that he is best known. Insisting that a basic understanding of the nature of science was essential—perhaps most of all to an experimentalist—he turned to the writings of Mach, Duhem, Kelvin, Tait, Helmholtz, Stallo, and Poincaré, most of whom were physicists who concerned themselves with the broader nature of science as well as its particularities. He felt that only practitioners of science should undertake such an analysis and remarked of Mill, a nonscientist, that he “never knew a law when he saw one” (1) and that Mill’s views were often suggestive if only “because they are erroneous” (2).

The most notable of Campbell’s theses for the theory of science was his strongly urged distinction of laws and theories. He saw this as pivotal for even a beginning understanding of the nature and status of scientific propositions.

Laws, he asserted, are propositions that can be established by experiment and observation, which does not mean that there are, in the main, overly simple relations between these laws and the fundamental propositions concerning our naive observations. Rather, scientific laws depend, for both their justification and their “significance,” on other sets of laws as well as complex collections of such simple and immediate judgments of sensations. Thus, if we use a variety of different means to determine the extension in a Hooke’s law experiment—for example, optical lever, interference apparatus, micrometer screw gauge, and so forth—we depend on the assumption that certain laws hold, and in particular, that all the methods yield the same result. Indeed Campbell went further than claiming that the proof of such a law depends on the truth of other laws; the meanings of the terms involved —“extension,” “force”—require that certain of those laws hold. If we say anything about electrical resistance, Campbell insisted, we assume that Ohm’s law is true. Were it not true, “resistance” would be “without any meaning.” Terms that depend in this way for their meaning on the truth of laws are termed “concepts.” Campbell was at pains to emphasize that almost all the laws of physics state relations between such concepts, and not between judgments of simple sensations. He was not unaware that there are fundamental laws, but he was reluctant to specify exactly how they are related to fundamental judgments, for it is at this level that, almost paradoxically, the concepts in question are too familiar for ease of analysis.

With the basis of his account of the nature of scientific laws, Campbell elaborated his major thesis of the structure of theories and their distinction from laws. A theory is a connected set of propositions which fall into two different categories. The basis of the theory, and that which basically establishes its identity, is the set of propositions, termed “the hypothesis,” which concern some collection of “ideas” characteristic of that theory. In isolation, this hypothesis is incapable of either proof or disproof—it is, in a sense, merely arbitrary. The second group of the propositions that constitute a theory Campbell termed “the dictionary.” These propositions assert the relation of terms of the hypothesis (“hypothetical ideas”) with the terms of scientific laws (“concepts”) whose truth or falsity is determinate. Campbell considered a fabricated, trivial example of such a combination of hypothesis and dictionary. The hypothesis states that a and b are constants for all values of the independent variables u, v, w and that c = d, where c and d are dependent variables. The dictionary asserts for this meager hypothesis that (c² + d²)a = R, where R is the resistance of a particular piece of metal. Further, the dictionary states that cd/b = T where T is the temperature of the same piece. It is an immediate consequence of the hypothesis that (c² + d²)a/(cd/b) = 2ab, which is a constant. From this conclusion and the dictionary it may be concluded that the resistance of the metal is directly proportional to the temperature. This statement asserts the relation of observational “concepts” in the sense referred to before and is the law which the theory explains—at least in the provisional sense that it has the law as a consequence.

A paradigm example of a highly significant physical theory is the dynamical theory of gases by means of which Boyle’s law and Gay-Lussac’s law may be explained. With his thesis of the nature of theories, Campbell reconstructed in almost perfect accord this theory and the way in which it related to the physical laws mentioned. He emphasized that this theory does not exhibit the artificiality of his explicative example and offered, in contrast with it, a genuine mode of scientific explanation. The difference, for Campbell, rested on his conclusion that the propositions of the hypothesis of the dynamical theory of gases display an analogy that the corresponding propositions of the other theory do not display. This analogy is the third essential constituent of any physical theory. Although Campbell equivocated about the specific nature of the analogy, he insisted that in the case of the dynamical theory of gases statements of the hypothesis take such a form that, examining a system of particles in a box, etc., we would find that such particles obey physical laws analogous to those principles. Provided that we associate the appropriate measurable physical concepts with the various symbols in the statements of the hypothesis, we would be able observationally to establish the truth of those propositions as laws.

Campbell emphasized the role that such an analogy has in theory construction by indicating that the propositions of the dictionary are suggested by it. One particular term in the hypothesis is identified with the pressure since in what would be our law like (in Campbell’s strict sense of “law”) analogue, it would represent the average pressure on the walls of the observed box. He noted that, for the most part, philosophers of science have misunderstood the role of analogy with respect to hypothesis and represented it as an “aid,” in a plainly heuristic sense, to the construction of theories serving only a “suggestive” function (3). Campbell, however, insisted that rather than being such a merely heuristic and basically dispensable aid it is utterly essential to theories.

Indeed, theories would be without any value if such analogies were absent. Physical science is not purely logical, and physicists cannot rest content with “a set of propositions all true and logically connected but characterized by no other feature” (4). Campbell insisted that frequently the analogy is the “greatest hindrance” to the establishment of theories rather than being an aid in that process. It must be recognized that theories such as that in his explicative example are trivial not only in the sense alluded to but also in the ease with which they may be constructed. To construct, on the other hand, a theory that exhibits a significant analogy is a creative act of major importance.

Having developed his analysis of the nature of theories and having shown its applicability to at least a paradigm example based on mechanical analogy, Campbell faced the criticism of those theoreticians in the tradition of Stallo and Mach who militated against not only the essential role of analogies in theories but even against their desirability. The major problem for Campbell’s reconstruction was to give an account of a paradigm example of those “mathematical theories” which the opposing school of thinkers cited as exemplars. Fourier’s theory of heat conduction supplied the basis for Campbell’s reply.

Campbell noted two important differences between the theory of gases and Fourier’s theory. Although it was clear that both theories exhibited the hypotheses and dictionaries so basic to Campbell’s thesis, they diverged in the following respects. In the first place, it seemed that every idea in the hypothesis of Fourier’s theory was related directly by means of the dictionary to a corresponding concept; while in the theory of gases only functions of those ideas occurred in the dictionary. In the second place, it seemed to be the case that any analogy of the sort pertinent to the theory of gases was absent. Campbell’s views on this second difference were equivocal. He vacillated between countenancing the total absence of any analogy and maintaining that there was analogy but of a rather different kind from that exhibited in the theory of gases. It is clear that Campbell wished to find some basis for the claim that analogy had some essential role in theories such as Fourier’s. If it were to be conceded that there was no analogy at all, no distinction could be made—at least in Camp bellian terms of reference—between Fourier’s theory and Campbell’s trivial explicative example; and, consequently, no grounds could be given for claiming that Fourier’s theory provided a significant scientific explanation.

Campbell offered a number of views on what form the analogy pertinent to Fourier’s theory might take. First, if there is an analogy, he claimed, it is between the propositions of the hypothesis and the laws which the theory is to explain. For the theory of gases, on the other hand, the analogy is between the hypothetical propositions and a set of laws which are found to be true, but which are distinct from those to be explained. Second, regarding theories of the type of Fourier’s as “generalizations” of certain experimental laws, he asserted that it is not the case that any generalization will be adequate. The basic constraint which prohibits such license is the requirement that only the simplest generalizations be acceptable. Thus it is that “just as it is the analogy which gives its value” to the theories like the theory of gases “so it is the simplicity which gives its value” to the the ories of the mathematical type. Despite the fact that only a few pages earlier he insisted that “some analogyis essential” to Fourier’s theory, Campbell eventually concluded that “it may be true” that mathematical theories “are not characterized by the analogy and do not derive their value from it.” It is, he continued, “characterized by a feature which is as personal and arbitrary as the analogy.” It is from this element, a certain “intellectual simplicity… and ease with which laws may be brought within the same generalization,” that it derives its value.

“Value” and “explanatory power” were almost interchangeable for Campbell. This was so precisely because the primary object of theories is the explanation of laws. Explanation is, in Campbellian terms, the substitution of more satisfactory for less satisfactory ideas. The intrinsic unsatisfactoriness of ideas may be due to confusion or complexity, or, alternatively, to a lack of familiarity. One kind of explanation of laws is by laws. The concepts involved are more satisfactory in the sense of being simpler, which in turn is a function of the generality of the explaining laws. The explanation of laws by laws, although a significant scientific process in the ordering of phenomena, is not as basic as the explanation of laws by theories. It is really only with this latter kind of explanation, Campbell professed, that genuine intellectual satisfaction is achieved.

Mathematical theories like that of Fourier do indeed give intellectual satisfaction in a way that extends beyond that of explanation by laws. It should be noted, however, that the almost indefinable counterpart of analogy in mathematical theories is just that elegant synthesis of simplicity and generality which Campbell identified with the explanatory power of mere laws. It is not without reason that the role of mathematical theories “is hardly different from that played by laws.” Theories similar to the theory of gases, exhibiting analogy in the stronger sense which Campbell gave, have a value which is a function of the familiarity of the analogy. This yields a kind of explanation of laws which is distinct from the already mentioned type. Laws could not provide explanations comparable in kind.

The explanatory power of a theory is dependent on the “intrinsic interest” which it may have for a particular scientist or school of scientists, Campbell confessed that in admitting this he relativized much of the debate between the protagonists of mathematical theories and those who insist on the strong sense of analogy to considerations of a subjective kind, although not totally. Campbell viewed as mistaken the argument that mathematical theories are less likely to lead the scientist into error, on the grounds that they are, as Mach put it, “purely phenomenal,” since considerations of simplicity of the kind which Campbell identified as characteristic of mathematical theories cannot be considered as determined solely by “phenomena.” Indeed, in an important sense, mechanical theories are more “purely phenomenal” since the propositions of their hypotheses are analogous to true observational laws.

Despite Campbell’s own inclination toward mechanical theories, he was at pains to point out that one can hold theories which exhibit the strong analogy—which he found so important—without their being strictly mechanical. It is likely that the general appeal of such analogies stems from the close relation and relevance of the laws of mechanics to our voluntary actions. Although this is the case, such appeal could be displaced to other laws, say, electrical ones, for it is not the case that all changes in the world with which we may be familiar are changes of matter and motion. The important requirement, though, is still maintained for this class of theory; that is, there should be an analogy between propositions of the hypothesis and some true observational laws which have “intrinsic interest.”

One of the major stimuli for Campbell’s careful scrutiny of the distinction of theories and laws was his concern for the semantic status of theoretical propositions. Although he, once again, equivocated in his views on exactly how those propositions derived their meaning, he felt it important to distinguish between the meaning given the theories by virtue of the dictionary, which was closer to “meaning” in the sense of empirical testability, from that “significance” which the analogy provided. His basic concern was the meaning of those terms of the theoretical propositions which the dictionary only obliquely or partially made either meaningful or testable. This was clearly a most pressing problem for the kind of theory, such as the theory of gases, in which the dictionary provided relations with functions of ideas and not the ideas themselves.

Although Campbell pressed for the role of analogy in these paradigm theories, he did not believe that they provided theoretical propositions with naïve common-sensical meanings. He emphasized that “the-velocity-of-the-electron” should not be thought of as meaning the same kind of thing as “the velocity of this billiard ball,” for to think so would be to be deluded by the grammatical form. He was careful to hyphenate such phrases in order to accentuate the logical indivisibility of the notion.

As is characteristic of such an experimentalist, Campbell’s inclinations were toward a quantitative view of his science. Indeed, for him physics was the “science of measurement,” and he devoted much space in his texts to considerations of the nature of physical measurement and its relation to purer mathematics. Once again his firm commitment to the role of theories appeared in his insistence that “no new measurable quantity has ever been introduced into physics except as a result of the suggestions of some theory.”