Modern Science and Its Philosophy

Philipp Frank



1. Discussions in a Vienna Coffee House

At the time when the first chapter of this book was written (1907) I had just graduated from the University of Vienna as a doctor of philosophy in physics. But the domain of my most intensive interest was the philosophy of science. I used to associate with a group of students who assembled every Thursday night in one of the old Viennese coffee houses. We stayed until midnight and even later, discussing problems of science and philosophy. Our interest was spread widely over many fields, but we returned again and again to our central problems How can we avoid the traditional ambiguity and obscurity of philosophy? How can we bring about the closest possible rapprochement between philosophy and science? By "science" we did not mean "natural science" only, but we included always social studies and the humanities The most active and regular members of our group were, besides myself, the mathematician, Hans Hahn, and the economist, Otto Neurath.

Although all three of us were at that time actively engaged in research in our special fields, we made great efforts to absorb as much information, methodology and background from other fields as we were able to get. Our field of interest included also a great variety of political, historical, and religious problems which we discussed as scientifically as possible. Our group had at that time no particular common predilection for a certain political or religious creed. We had, however, an inclination towards empiricism on one hand and long and clear‑cut chains of logical conclusions on the other. There were quite a few occasions on which these two predilections did not mix very well.

This apparent internal discrepancy provided us, however, with a certain breadth of approach by which we were able to have helpful discussions with followers of various philosophical opinions. Among the participants in our discussions were, for instance, several advocates of Catholic philosophy. Some of them were Thomists, some were rather adherents of a romantic mysticism. Discussions about the Old and New Testaments, the Jewish Talmud, St. Augustine, and the medieval schoolmen were frequent in our group. Otto Neurath even enrolled for one year in the Divinity School of the University in order to get an adequate picture of Catholic philosophy, and won an award for the best paper on moral theology. This shows the high degree of our interest in the cultural background of philosophic theories and our belief in the necessity of an open mind which would enable us to discuss our problems with people of divergent opinions.

At that time a prominent French historian and philosopher of science, Abel Rey, published a book which later was to make a great impression upon me. At the turn of the century the decline of mechanistic physics was accompanied by a belief that the scientific method itself had failed to give us the "truth about the universe"; hence nonscientific and even antiscientific tendencies gained momentum. I quote some passages in which Rey describes this situation excellently and precisely.

Fifty years ago, he says, the explanation of nature was believed to be purely mechanical.

It was postulated that physics was nothing but a complication of mechanics, a molecular mechanics . . . Today [1907] it seems that the picture offered by the physical sciences has changed completely. The general unity is replaced by an extreme diversity, not only in the details, but in the leading and fundamental ideas . . . [This accounts for] what is called the crisis of contemporary physics. Traditional physics assumed until the middle of the nineteenth century that it had only to continue its own path to become the metaphysics of matter. It ascribed to its theories an ontologic value, and these theories were all mechanistic. Traditional mechanistic physics was supposed, above and beyond the results of experience, to be the real cognition of the material universe. This conception was not a hypothetical description of our experience; it was a dogma.

The criticism of the traditional mechanistic physics that was formulated in the second half of the nineteenth century weakened this assertion of the ontologic reality of mechanistic physics. Upon this criticism a philosophy of physics was established that became almost traditional toward the end of the nineteenth century. Science became nothing but a symbolic pattern, a frame of reference. Moreover, since this frame of reference varied according to the school of thought, it was soon discovered that actually nothing was referred that had not previously been fashioned in such a way that it could be so referred. Science became a work of art to the lover of pure science, a product of artisanship to the utilitarian, This attitude could quite rightly be interpreted as denying the possibility that science can exist. A science that has become simply a useful technique . . . no longer has the right to call itself science without distorting the meaning of the word. To say that science cannot be anything but this means to negate science in the proper sense of the word.

The failure of the traditional mechanistic science . . . entails the proposition: "Science itself has failed." . . . We can have a collection of empirical recipes, we can even systematize them for the convenience of memorizing them, but we have no cognition of the phenomena to which this system or these recipes are applied. [1]

Our group was formed during the period which was so eloquently described by Rey. His book was discussed frequently by us in the last years of my stay in Vienna (1908‑1912). The problems raised and the results obtained are reflected partly in Chapter 2 of this book. The general reaction of our group to the intellectual and cultural situation depicted by Rey can be described as follows:

We recognized the gradual decline in the belief that mechanistic science would eventually embrace all our observations. This belief had been closely connected with the belief in progress in science and in the scientific conception of the world. Therefore, this decline brought about a noticeable uneasiness. Many people lost their faith in scientific method and looked for some other method which might yield a real understanding of the world. A great many people believed, or at least wanted to believe, that the time had come to return to the medieval ideas that may be characterized as the organismic conception of the world.

In the history of science and philosophy there have always been divergent opinions about the conditions under which we may say that a scientific theory has "explained" a certain range of observations. Some authors have maintained that only an explanation by mechanical causes and by the motion of material particles can satisfy our intellectual curiosity. Others have claimed that the reduction to mechanical causes is only a superficial explanation and not a real one. Some of the opponents of the mechanistic world view have stated that all phenomena must be interpreted in terms of the evolution of an "organic whole" in order really to understand them. The decline of the belief in mechanistic science seemed to favor this organismic view, which has been attractive to many because of its religious and social implications. In this way there had arisen at the turn of the century what some called a crisis in science or, more accurately, in the scientific conception of the world. For more than two centuries the idea of progress in science and human life had been connected with the advance of the mechanistic explanation of natural phenomena. Now science itself seemed to abandon this mechanistic conception, and the paradoxical situation arose that one could fight the scientific conception of the world in the name of the advance of science.

2. The Failure of Mechanistic Science

The sixteen chapters of this book have been written over a period of almost forty years. They are all meant to be contributions to one task: to break through the wall which has separated science and philosophy for about one and one‑half centuries. The book reflects the methods by which this wall has been besieged in the twentieth century. During the last decades of the nineteenth century, a revolution in physical science started which has itself brought about a revolution in our general scientific thought. The methods that have been tried in the fight for the unity between science and philosophy have changed along with the advance of science. Two characteristic beliefs of nineteenth‑century science broke down during its last decades; these were the belief that all phenomena in nature can be reduced to the laws of mechanics, and the belief that science will eventually reveal the "truth" about the universe. In the twentieth century the revolutionary changes in science developed with ever‑increasing rapidity and intensity. It is no wonder that this rapid transformation of scientific thought has been accompanied by rapidly changing methods in the scientist's approach to philosophy and in the philosopher's views on science.

The chapters of this book have played their part in this developing and changing fight for the unity of science and philosophy. Through them we can pursue a pattern that has grown from rather tentative and naively empirical beginnings to a more and more abstract technique. In order to understand this evolution precisely, it is not sufficient to follow the gradual alterations from the purely logical angle. We must also consider carefully the historical trend and background of the arguments.

Rey had the strong feeling that the place of physics in human thought had dangerously deteriorated. He says:

The physicochemical sciences are an effort made by man to explain sensed nature or what is perceived by our senses . . . The importance of this effort has been understood in all periods of history. So early a thinker as Epicurus believed that physics was the basis of the effort to liberate the human mind from its blind instinct to believe, from its prejudices and its superstitions . . . Everyone would agree that modem positivism has been nothing but an attempt to extend to all the departments of human knowledge, without exception, the general method and conception of science, that is, science constructed according to the example of physics; the spirit of positivism and the spirit of science have become synonymous in current speech.

If these sciences, which historically have played essentially an emancipatory role, were tarnished by a crisis that leaves them no other value than that of recipes which are technically useful and that deprives them of any significance in the cognition of nature, a complete upset in the art of logic and in the history of ideas must result. Physics loses all its educational value; the positive spirit which it represents is a misleading and dangerous spirit. Reason, rational method, and experimental method must be considered in good faith as having no cognitive value. All these methods are, then, procedures of action, not means of cognition. They can be developed in order to obtain certain practical results, but we must be well aware that they have no value except in their restricted domain. The cognition of the real must be sought or be given by other means. We must guard against the dangerous illusion of rationalism and scientism. It is important to know that by this method the real is ignored and that physics leads to ignorance and not to cognition of real nature . . .

If the problem of the cognition of nature and of the possibility of the physicochemical sciences remains in the same form in which it has developed from the Renaissance to the time of positivism, the rational and positive method remains the supreme educator of the human mind, in the domain that is accessible to it, of course. To give the mind a scientific attitude in the sense that has been understood by positivism and positive physics remains the necessary and sufficient condition of intellectual sanity. Physics is the school where one learns to know things. [2]

3. Ernst Mach's Purge of Traditional Physics

In this critical situation our minds turned towards a solution that had been advanced about twenty‑five years before by our local physicist and philosopher, Ernst Mach. He maintained that "explanation" by reduction to a system of cherished conceptions is pure illusion. If all the multitude of observable phenomena are reduced to mechanical or organismic phenomena, these special types of phenomena chosen as the basis of explanation are by themselves no more understandable than the phenomena that are to be explained. Mach claimed that there is no essential difference between an "explanation" and a "description." In our everyday language the word "description" refers to a single event or phenomenon. We "describe" the fall of a specific stone at a specific time and place. If, however, we formulate Galileo's law of freely falling bodies, which tells us that all bodies fall with an equal constant acceleration, we describe the fall of a great many bodies under different circumstances. This description of a class of phenomena is called by Mach a "physical law" or an "explanation." If an explanation is nothing but a description of many cases by a short sentence, it cannot matter much whether the "explanation" is given by reduction to the laws of mechanics or to the laws of electromagnetism or even of statistics.

In this way Mach separated the conception of "scientific explanation" from "mechanical explanation." He saved the scientific world picture from going down along with the mechanistic picture. Similar ideas had been advanced by other writers in Mach's time, but none formulated them with so much lucidity and breadth of approach. No one else anchored them so firmly in the soil of science, in physics as well as in biology and psychology.

Although Mach's views were the principal background of our thoughts, they were not the most powerful stimulus to our actual work. Our group fully approved Mach's antimetaphysical tendencies, and we joined gladly in his radical empiricism as a starting point; but we felt very strongly about the primary role of mathematics and logic in the structure of science. It seemed to us that Mach had not done full justice to this aspect of science. We felt that considering the principles of science as nothing but abbreviated descriptions of sense observations did not account fully for the fact that the principles of science contain simple clear‑cut mathematical relations among a small number of concepts, whereas every description of observations contains a great number of vague connections among a great number of vague concepts. We also felt that to call the principles of science "economic descriptions of observations" was not to do justice to the predominant role of reasoning in the discovery and presentation of these principles. We were even attracted by some tenets of Kant's theory of knowledge, particularly by his "Introduction to Any Future Metaphysics that may Present Itself as Science." We saw much truth in Kant's statement that the recording of observations is not a purely passive act but that a great deal of mental activity is necessary in order to formulate general statements about sense observations.

The human mind, according to Kant, can describe natural phenomena only by using a certain pattern, certain forms of thinking that are produced by the observing mind and are not provided by the observed physical object. Since these "forms" or "patterns of experience" are provided by the human mind and not by the physical facts, they cannot be changed by the advance of scientific investigations. The material delivered by sense observation can be manufactured into a scientific system only by adding forms of experience that are determined by the nature of the human mind. Our whole group understood and fully agreed that the human mind is partly responsible for the content of scientific propositions and theories. But we also felt strongly that an orthodox belief in Kant would lead us into a new type of metaphysics, not less unscientific than medieval metaphysics. For instance, Kant, and even more, many of his followers, maintained that the axioms of Euclidean geometry or Newton's laws of motion are forms of experience that are produced by the organization of the human mind. This would imply that no advance in these fundamentals of science will be possible without changing the nature of human mind. However, about 1900, scientists became accustomed to envisaging the possibility of changes in the principles that had been traditionally regarded as self-evident axioms. Non‑Euclidean geometry was no longer considered a purely logical exercise, but was recognized as a serious system of geometry. This new conception, along with Mach's criticism of Newton's mechanics and with the new electromagnetic theory of matter, opened a path for some doubts whether Newton's laws of motion were actually the final word about the nature of the universe. A step in the same direction was the statistical interpretation of thermodynamics, which suggested the superposition of statistical laws on dynamical laws. As enthusiastic students of contemporary science our group rejected Kant's doctrine that the forms of experience provided by the human mind were unchangeable. We looked for some way to construe these forms as subject to an evolution that would be in accord with the evolution of science. We felt very strongly that there was a certain gap between the description of observations, necessarily vague and complex, and the principles of science, consisting, in physics particularly, of a small number of concepts (like force, mass, etc.) linked by statements of great simplicity. We admitted that the gap between the description of facts and the general principles of science was not fully bridged by Mach, but we could not agree with Kant, who built this bridge by forms or patterns of experience that could not change with the advance of science.

4. Henri Poincaré and a "New Positivism"

In our opinion, the man who bridged the gap successfully was the French mathematician and philosopher Henri Poincaré. For us, he was a kind of Kant freed of the remnants of medieval scholasticism and anointed with the oil of modern science.

Kant had claimed that there would never be another way to organ‑ our experience than by Euclidean geometry and Newtonian mechanics. But at the turn of the century non‑Euclidean geometry bad been established, although its importance for physical science was not completely understood. Departures from Newtonian mechanics were already in the making. We understood that Kant's erroneous conception of geometry and mechanics must have its source in his erroneous attitude toward the relation between science and philosophy.

At this time I was very much interested in the criticism launched by Nietzsche against Kant's idealistic philosophy. Kant's primary aim was to answer the question of how the human mind can make statements about facts of the external world with absolute certainty even if these assertions are not the result of experience about this world. For Kant, the propositions of Euclidean geometry were a convincing example of assertions of this kind. By claiming that the axioms of geometry are forms of experience produced by the human mind, Kant explained man's ability to produce these assertions without external experience. Nietzsche said flippantly that Kant's explanation is merely equivalent to saying that man can do it "by virtue of a virtue." Nietzsche accused him of demonstrating by sophisticated and obscure arguments that popular prejudices are right while the scientists are wrong. We appreciated in Poincaré. just what was different from Kant. We agreed with Abel Rey's characterization of Poincaré's contribution as a "new positivism" which was a definite improvement over the positivism of Comte and Mill. Rey wrote:

The new positivism certainly has its origin in the positivism of Comte, of Taine and Mill . . . It is rejuvenated but has preserved the great directive ideas from its previous stage: the relativism and empiricism of our knowledge. But it stresses . . . the idea of experimental categories which are required by science as a central necessity . . . Positivism is renewed by building a new rationalism upon the criticism of traditional rationalism in the second half of the nineteenth century . . . What was lacking in Comte's or Mill's positivism . . . was their . . . failure to have established in a new form a theory of categories. Objective experience is not something which is outside and independent of our minds. Objective experience and mind are functions of each other, imply each other, and exist by virtue of each other. To say that the relations between physical objects derive from the nature of these objects and to say that these relations are constructed by our minds are two artificial theories . . .

Our experience is a system, a relation of relations. The relation is the given. [3]

5. Mach, Poincaré and Lenin

Chapter 1 of the present book is written in the spirit of this new positivism. Poincaré claimed that the general laws of science—the law of inertia, the principle of conservation of energy, etc.—are neither statements about facts which can be checked by experiments, nor a priori statements which necessarily emanate from the organization of human mind. They are rather arbitrary conventions about bow to use some words or expressions. In this chapter Poincaré's basic tenets are applied to the law of causality.

I started from Hume's formulation of this law: When a state A of a system is followed one time by a state B, every time that A recurs it is followed again by B. I analyzed this formulation in a way that is similar to Poincaré's analysis of the general principles of science. I put the question: How do we know when the state A has recurred? There is no exact method except to investigate whether it is followed by B. Hence the law of causality is not a statement about observable physical facts but a definition of the expression, "the state A has recurred."

When I first published this paper (1907) it aroused a certain amazement among scientists. Among the comments were those of two men of world‑wide fame, although in different fields: Einstein and Lenin. Einstein's letter was my first personal contact with him. He approved the logic of my argument, but he objected that it demonstrates only that there is a conventional element in the law of causality and not that it is merely a convention or definition. He agreed with me that, whatever may happen in nature, one can never prove that a violation of the law of causality has taken place. One can always introduce by convention a terminology by which this law is saved. But it could happen that in this way our language and terminology might become highly complicated and cumbersome. What is not conventional in the law of causality is the fact that we can save this law by using a relatively simple terminology: we are sure that a state A has recurred when a small number of state variables have the same values that they bad at the start. This "simplicity of nature" is the observable fact which cannot be reduced to a convention on how to use some words. These remarks had a great influence on my thought on the future course of the philosophy of science. I realized that Poincaré's conventionalism needs qualifications. One has to distinguish between what is logically possible and what is helpful in empirical science. In other words, logic needs a drop of pragmatic oil.

Lenin's comment was rather unfavorable. In his book "Materialism and Empiriocriticism" (we would call it today "Materialism and Positivism") he maintains that, "as a Kantian," I "rejoiced" to be able to give support to Kant's idealism by the "most modern philosophy of science." From my reference to the relations between Poincaré and Kant he drew the conclusion that I tried to make use of Poincaré in the service of idealistic philosophy and that, therefore, this paper had an antimaterialistic and reactionary tendency. Lenin's book did not come to my attention before the early twenties. Then, however, it stimulated me to think over more carefully the relation between Poincaré and Kant, between positivism and idealism and, particularly, to investigate the role that metaphysical interpretations of scientific theories have played in support of political and religious philosophies.

However, during the interval (1907‑1917) between writing the first and second chapters of the present book, my interest was directed mainly toward any possible advance in the logic of science. I was convinced that the solution must be sought by starting from the ideas of such men as Mach and Poincaré.

At first glance these two authors seemed to contradict each other flagrantly. I soon realized that any advance in the philosophy of science would consist in setting up a theory in which the views of Mach and of Poincaré would be two special aspects of one more general view. To summarize these two theories in a single sentence, one might say: According to Mach the general principles of science are abbreviated economical descriptions of observed facts; according to Poincaré they are free creations of the human mind which do not tell anything about observed facts. The attempt to integrate the two concepts into one coherent system was the origin of what was later called logical empiricism.

6. D. Hilbert's New Foundations of Geometry

The traditional presentation of physical theories frequently consists of a system of statements in which descriptions of observations are mixed with mathematical considerations in such a way that sometimes one cannot distinguish clearly which is which. It is Poincaré’s great merit to have stressed that one part of every physical theory is a set of arbitrary axioms and logical conclusions drawn from these axioms. These axioms are relations between signs, which may be words or algebraic symbols; the important point is that the conclusions that we draw from these axioms are not dependent upon the meanings of these symbols. Hence this part of a theory is purely conventional in the sense of Poincaré. It does not say anything about observable facts, but only leads to hypothetical statements of the following type: "If the axioms of this system are true, then the following propositions are also true," or still more exactly speaking: "If there is a group of relations between these symbols, there are also some other relations between the same symbols." This state of affairs is often described by saying that the system of principles and conclusions describes not a content but a structure. Hence, this system is occasionally referred to as the structural system.

The simplest example is geometry. It is the first example in the history of science of a logical system that claims to make statements about facts in the observable world as well. Geometry did not proceed according to the pattern adopted by the older positivists like Hume, Comte, or Mill. It did not collect facts and draw conclusions from these observations. Instead, it built up a system of axioms which were statements containing abstract terms like "point," "straight line," "interaction." Conclusions from these axioms can actually be drawn without knowing the meaning of these terms. In the textbooks of geometry that were written in accordance with the tradition laid down by Euclid, such terms as "straight line" or "intersection" seemed to have a "meaning" in the same sense as the words "table" or "horse" have a meaning, except that the geometric concepts were supposed to be the names of "idealized" physical objects.

After the turn of the century, however, David Hilbert started a purge of the foundations of geometry and set up a clear‑cut and consistent system of axioms. He stressed that such concepts as "point" or "straight line" have no meaning besides the one defined by the axioms. The axioms are "implicit definitions" of the geometric concepts. For Hilbert, as for Poincaré, the axioms were conventions about the use of the geometric terms. In this way Hilbert made a significant contribution to the "new positivism." He restricted himself, however, to the investigation of the structural system and did not discuss, as Poincaré did, the relation of the geometric axioms to our experience or our sense observations. The propositions that are derived from the axioms cannot contain any word besides the symbols contained in the axioms. But these symbols have no meaning in the physical world. The great asset of this method is that conclusions can be drawn without being affected by the vagueness of terms that describe our actual observations, like "red," "blue," "warm," "sweet." This method of geometry became the method of the mathematical physics of the nineteenth century. Heat, electricity, and light were described by systems of principles that consisted not of observational terms but of abstract symbols. However, the symbols occurring in the axioms and propositions, of geometry as well its of mathematical physics, can be linked to observable facts in a brief and easily understandable way; for example, the straightness of a line can be checked approximately by the edge of a rigid table, or a volume or a temperature can be measured by simple physical methods.

7. The New Positivism and P. Duhem's "Thomism"

The axiomatic or structural system, including its conclusions, is merely an arbitrary convention if the purely logical viewpoint is maintained without going into the physical interpretation. It was clear to Poincaré that the structural system is logically arbitrary because it cannot be demonstrated by logical means. It is not psychologically arbitrary, however, because in practice we construct only those systems that can be interpreted in terms of physical facts and that are therefore helpful for the formulation of natural laws.

If this line of reasoning is followed, we can see, in a perfunctory way at least, how Mach’s and Poincaré’s ideas about the general principles of science can be integrated. The axiomatic system, the set of relations between symbols, is a product of our free imagination; it is arbitrary. But if the concepts occurring in it are interpreted or identified with some observational conceptions, our axiomatic system, if well chosen, becomes an economical description of observational facts.

Now the presentation of the law of causality as an arbitrary convention (Chapter 1) can be freed of its paradoxical appearance. The law of causality, as a part of an axiomatic system, is an arbitrary convention about the use of terms like "the recurrence of a state of a system," but if interpreted physically it becomes a statement about observable facts. In this way, the philosophy of Mach could be integrated into the "new positivism" of men like Henri Poincaré. Abel Rey, and Pierre Duhem. The connection between the new positivism and the old teaching of Hume and Comte is the requirement that all abstract terms of science—such as force, energy, mass—must be interpreted in terms of sense observations. Exactly speaking, every statement in which these abstract terms occur must be interpreted as a statement about observational facts. Mach formulated this requirement by saying that all scientific statements are statements about sense observations.

Mach's requirements have frequently been misinterpreted. Some authors considered them a kind of subjective idealism, meaning that the world consists of sense data only. Others considered them a kind of skepticism or agnosticism, meaning that man cannot know anything about the true or real world; man and man‑made science can tell us only about our sense observations, while the objective reality will be eternally unknown to human intelligence. In this interpretation science would be purely subjective, a collection of statements about subjective sense impressions. This misinterpretation had its source in traditional philosophy, according to which science and philosophy must find the hidden treasure of truth behind appearances. All the various systems of traditional philosophy seeking objective reality behind appearances were opposed to Mach's philosophy. Moreover, a great many scientists disliked the idea that the statements of science do not describe the real world but are only statements about our sense observations, without objective significance.

The property of the structural system of not telling us anything about the world of observable physical facts was particularly emphasized by the French scientist, philosopher, and historian, Pierre Duhem. His writings exerted a strong influence upon our group and, particularly, upon my own thinking. By studying Duhem thoroughly we gained a more subtle understanding of the relations among science, metaphysics, and religion than has been customary among empiricists. Duhem says, much as Mach had done,

A theory of physics is not an explanation; it is a system of mathematical propositions deduced from a small number of principles the aim of which is to represent as simply, as completely, and as exactly as possible, a group of experimental laws. [4]

This formulation is a great step on the way toward an integration of Mach and Poincaré. Duhem understood very well that no single proposition of a physical theory can be said to be verified by a specific experiment. The theory as a whole is verified by the whole body of experimental facts. As Duhem put it,

The experimentum crucis is impossible in physics. [5]

He says again:

The watchmaker to whom one gives a watch that does not run will take it all apart and will examine each of the pieces until he finds out which one is damaged. The physician to whom one presents a patient cannot dissect him to establish the diagnosis. He has to guess the seat of the illness by examining the effect on the whole body. The physicist resembles a doctor, not a watchmaker. [6]

And elsewhere:

The experimental verifications are not the basis of the theory, but its culmination.

One notes how far Duhem has proceeded on the way from Mach's conception of a physical theory to the conception which was later advocated by logical empiricism.

Duhem, however, was also, from another angle, a great influence upon the philosophy of our group. He believed, like Mach, that an "explanation" that would be distinct from "economical description" would require an excursion into metaphysics. He says,

If the object of physical theories is to explain experimental laws, physical theory is not an autonomous science; it is subordinated to metaphysics.

Although he was convinced that physics cannot provide an explanation of experimental laws, he did not mean to say that no explanation is possible.

In seeking to stress the distinction between physics and metaphysics I do not mean to disdain either of these sciences, and I think to facilitate their accord much better than if I had confounded the object and the method of the former with the object and the method of the latter . . . The knowledge that metaphysics gives us of things is more intimate, more profound than that which is furnished by physics; it therefore surpasses the latter in excellence. [7]

As a matter of fact, Duhem was an advocate of Aristotelian and Thomistic metaphysics and a faithful believer in the whole of Catholic theology. When I noted this merging of the most advanced type of "new positivism" with Thomistic metaphysics into one coherent system it impressed me strongly. Duhem interpreted on the basis of this hybrid philosophy the historic conflict between the Roman Church and the Copernican system.

Although our group did not follow Duhem's metaphysical predilection, his doctrine became for us a frame of reference to which we could relate all the conflicts that have raged between science and religion and, more generally, between science and political ideologies.

8. E. Mach as a Philosopher of "Enlightenment"

Despite the aversion toward Mach that had been particularly noticeable among the German scientists and philosophers, he exerted a remarkable attraction for some groups of scientists as well as of philosophers. However, the hostile reaction against him was of great intensity. Disputes about him have been repeated again and again at gatherings of scientists, the conflict of opinions has always been intense, and the discussions have frequently had an emotional flavor.

Ernst Mach died in 1917, the year in which the Soviet government seized the power in Russia. Few among the European and American scientists realized at the time that the new ruler of Russia had published a book ten years previously in which he branded "Machism" as a reactionary philosophy. This book laid the foundation for a peculiar situation in the new Russia by making Mach a permanent target of attack. He was accused of agnosticism, subjectivism, and relativism. It was alleged that Mach had denied that science could know anything about the objective world. In this way, it seemed, the door was opened to other ways of finding the truth, particularly to the ways of traditional religion. It is interesting and amazing that Lenin denounced Machism by the same argument which has been used by advocates of physical realism among European and American scientists.

In the year of Mach's death (1917), I wrote Chapter 2 of the present book. Its purpose was the investigation of the value of Mach's philosophy of science for the future development of science and of human thought in general. Mach's philosophy is presented in such a way that those features are stressed that will survive him and remain essential parts of any future philosophy of science. This means in part fouler that Mach is judged from the viewpoint of the "new positivism" of men like Poincaré and Duhem. Mach's philosophy is described and characterized with respect to its place in the history of human thought at the start of the twentieth century. It is likened to the philosophy of the Enlightenment in the eighteenth century. Mach analyzed the fundamental concepts of nineteenth‑century physics, such as mass and force, and made clear that all statements containing these words can be interpreted as statements about sense observations. Then concepts like these do not denote entities of a hidden real world behind the appearances, but are "auxiliary concepts" by which statements about observations can be expressed in a more convenient and practical way. As the auxiliary concepts of nineteenth‑century science were mainly concepts of mechanistic physics, Mach's analysis was to a certain extent a debunking of mechanistic science as a system of statements about physical reality. Nonetheless, Mach bad no special bias against the mechanistic terminology that would imbue him with a particularly antimaterialistic tendency. He tried to debunk all types of auxiliary concept in so far as they pretended to describe ontological realities or metaphysical entities. I illustrated this function of Mach's philosophy as a philosophy of enlightenment appropriate to the turn of the century by pointing out similarities between Mach and Nietzsche. The great mass of writing about Nietzsche has largely overlooked the fact that he was a philosopher of enlightenment in his acute analysis of the auxiliary concepts of contemporary idealistic philosophy.

The years after 1917 saw, as has been mentioned, the establishment of Soviet power in Russia, the end of World War I, and the founding of new democracies in Central Europe, such as the Austrian, Czechoslovakian, and Polish republics. The event that had the greatest bearing at this time on the development of the philosophy of science, however, was the new general theory of relativity advanced by Einstein after 1916. In this theory Einstein derived his laws of motion and laws of the gravitational field from very general and abstract principles, the principles of equivalence and of relativity. His principles and laws were connections between abstract symbols: the general space time coordinates and the ten potentials of the gravitational field. This theory seemed to be an excellent example of the way in which a scientific theory is built up according to the ideas of the new positivism. The symbolic or structural system is neatly developed and is sharply separated from the observational facts that are to be embraced. Then the system must be interpreted, and the prediction of facts that are observable must be made and the predictions verified by observation. There were three specific observational facts that were predicted: the bending of light rays and the red shift of spectral lines in a gravitational field, and the advance of the perihelion of Mercury.

9. A. Einstein's Now Concept of Physical Science

However, if we compare Einstein's theory with previous physical theories, we note a certain difference in structure. It is after all only a difference in degree, but it directs our attention to a considerable change in the conception of physical theory.

Whatever conclusions may be drawn from them, Einstein's fundamental laws will describe motions in terms of the general space‑time coordinates. Before the results of his theory can be verified by observation, it is necessary to know how statements about these general coordinates can be expressed in terms of observational facts. In traditional Newtonian physics, spatial coordinates and time intervals could be determined by the traditional methods of measuring length and time, by using yardsticks and clocks. However, the general coordinates in Einstein's theory are quantities that define the positions and motions of moving particles with respect to systems of reference that can possess all sorts of deformations, with variable rates of deformation at every point. No rigid and defined system of reference for space and time measurement is given as a general basis for the definition of the space and time coordinates. The methods of measurement must be developed along with the conclusions from the principles of the theory. What is the bearing of these facts upon our general conception of the structure of a scientific theory?

In nineteenth‑century physics the translations of statements that contain abstract symbols of the theory—mass, distance, time interval, and the like—into observational facts did not cause much trouble. It was taken for granted that the straightness of a line, the temperature of a body, or the velocity of a motion could be measured. At least, it was not suspected that there was any difficulty in assuming that such measurements are possible. In Einstein's general theory of relativity, however, the description of the operations by which these quantities could be measured becomes a serious and complex task; it becomes an essential part of the theory. These descriptions of the operations by which abstract symbols, such as the general space‑time coordinates, are connected with observational facts are called today "operational definitions," according to a terminology suggested by P. W. Bridgman. [8]

As early as 1905, in his restricted theory of relativity, Einstein was well aware that the "operational definitions" are an essential part of his theory. Later he described the decisive alterations that were brought about by his new physical principles in the conception of a physical theory by stressing the fact that the connection between the symbols of the theory and the observational facts following from them is much longer, much more complex, and much more difficult to deal with than the connection assumed by nineteenth‑century physics, to say nothing of the physics of the seventeenth and eighteenth centuries. The alteration brought about in the general conception of a scientific theory was a greater emphasis on the gap between the structural system and the experimental confirmation. Advancing a new theory now involved two tasks, both of which required great creative power: the invention of a structural system, and the working out of operational definitions for its symbols.

However, the great new idea in every new physical theory, according to Einstein, was the creation of the structural system. In this sense a physical theory describes "the structure of the world." This way of speaking could easily be interpreted as meaning that the symbols, which are the building stones of the structure, are also the "real building stones" of the universe and that the structure of the symbolic system is "the real structure of the world." Following Einstein's own interpretation of his conceptions, statements like "the theory describes the real structure of the world" mean that appropriate operational definitions enable us to derive from the symbolic system observational facts that check with our actual observations. Hence, the conception of physics advocated by the new positivism of Poincaré is altered by Einstein's conception in such a way that a theory remains an economical description of facts by means of a structure and operational definitions. However, the connection between the symbols and the observational fact is not so simple as was anticipated.

10. Geometry and Experience

The old problem of the relation between reasoning and experience in geometry was satisfactorily solved by Einstein's theory. Even the advocates of a new positivism, like Poincaré and Duhem, left a feeling of uncertainty concerning the actual position of geometry between the domains of logic and experience. One might even say that they left a feeling of uneasiness. Poincaré made it clear that geometry itself as a logical system says nothing about the physical world. From the viewpoint of physics, or of empirical science in general, this logical system can be judged only as a practical tool which can be used for a description of the physical world. This procedure may be easy or difficult, simple or complicated. By using Einstein's theory, a description of motions is obtained in terms of spatial and temporal coordinates that are measured by the reading of yardsticks and clocks. However, these instruments, and therefore their readings, are affected by the same gravitational field that is responsible for the motion. An important consequence is that the distances between different points in space follow the axioms and propositions of Euclidean geometry only if the field of force is almost negligible. If we have to consider "strong" fields, the distances between points no longer fit the rules of Euclidean geometry. This means, for instance, that the sum of the angles in a rectilinear triangle is no longer exactly equal to two right angles. The departure from two right angles is the greater the larger the area of the triangle and the stronger the field of gravitation. In order to give to these statements a physical meaning we assume tacitly that "straight lines" are defined by the traditional technologic procedures by which straight edges are produced on rigid bodies, such as a bar of steel.

This means in ordinary geometrical parlance that Euclidean geometry is valid only if the field of gravitation is negligible and if the areas of triangles are small. Generally the laws that determine the relation between the measured distances will be those of non‑Euclidean geometry. Poincaré and his immediate followers used to say that it is a convention whether one accepts Euclidean or non‑Euclidean geometry. Actually this statement means only that we can build up structural systems of two types: Euclidean and non‑Euclidean. As a structural or axiomatic system both are equally acceptable. If traditional methods of measurement are used—that is, if length and time are defined in terms of operations with rigid yardsticks and orthodox clocks—the fact must be considered that the instruments of measurement are affected by gravitational fields and therefore that the results of measurement, too, are affected. For every specific field of gravitation, the theory yields a specific influence on the yardsticks and clocks and therefore upon the results of measurement. By considering this influence "operational definitions" can be formulated. The statements of mechanics now become statements about the results of actual measurements or about actual observations. They are therefore no longer arbitrary. If they are checked, they will be found to be either true or false. In a gravitational field the yardsticks and clocks will be affected in such a way that the results of measurements follow non‑Euclidean rather than Euclidean geometry.

In 1921, Einstein gave a lecture at the Academy of Sciences in Berlin with the title "Geometry and Experience," [9] in which he summarized in a conclusive way the place of reasoning and experience in geometry. According to Einstein, geometry can be either a structural system with arbitrary axioms or a physical theory. In the first case the conclusions of geometry are certain but do not tell us anything about the world of experience; in the second case the propositions of geometry can be checked by experiment and are as certain or uncertain as are any statements of physics or, for that matter, of any empirical science.

By Einstein's argument it was demonstrated with great lucidity that there is no statement in geometry that is derived by reasoning without sense observations and that at the same time tells us something about the external world. Such geometric propositions, however, had been considered the most conspicuous examples of assertions about the external world that are derived from pure reasoning. If the statements of geometry do not have this property, then the scientific basis of traditional metaphysics has disappeared. The cause of positivism against metaphysics has won a major battle.

11. Neo‑Kantianism and Neo‑Thomism

When Einstein had cleared up the foundations of geometry, the believers in scientific metaphysics were put on the spot; they were deprived of their best example of the existence of metaphysical assertions in science itself.

All schools of traditional philosophy have more or less one belief in common: we can make general statements about facts of the external world with such certainty that they will never be refuted by any future advance in science. For Aristotle such statements were the principles of Greek astronomy, for Kant they were Newton's principles of mechanics. Both the Aristotelians and the Kantians of the nineteenth and twentieth centuries agreed that the validity of Euclidean geometry had been established forever. The only difference was that for one school the doctrine that was believed to prevail forever was Aristotle's physics, while the other, more modern, school permitted the pursuance of the advance of science up to Newton. Each philosophical creed petrified the state of physics that prevailed at its time.

When Einstein demonstrated the possibility or even the plausibility that Euclidean geometry might be wrong he produced a catastrophic effect upon all the schools of traditional philosophy. The metaphysical schools of the Aristotelian and the Kantian types lost their basis in science. To meet this situation two attitudes have developed within these schools. To borrow a terminology from theology, we may call them the fundamentalist and the modernist attitudes. The first group maintained bluntly and boldly that scientists were just wrong. They were specialists and not able to reason correctly because of their lack of metaphysical training.

The modernists are found in the camps of neo‑Aristotelians (mostly called neo‑Thomists) and of the neo‑Kantians. They admitted that Euclidean geometry might be wrong. This was, according to these schools, the truth, but "not the whole truth." From the viewpoint of science proper, they have accepted fully the conception of geometry and physics advanced by the new positivism. But in the background of their teaching on geometry and physics there is a hint of a more profound wisdom which is presented in terms of Aristotelian or Kantian metaphysics. As an example of neo‑Thomism we may consider the works of J. Maritain, particularly his book, "The Degrees of Knowledge." [10] An example of neo‑Kantianism is the writing of Ernst Cassirer, particularly his books on the "Theory of Relativity" [11] and on "Determinism and Indeterminism in Modern Physics." [12] The latter one is discussed in Chapter 9 of the present book. This metaphysical background, according to these authors, has no relevance for science proper; it is separated by airtight walls from the domain of scientific discourse. In this way science became autonomous with respect to metaphysics, but the validity of the metaphysical assertions in the background could not be checked by any experimental test. These assertions became more or less a tautological system of propositions, like pure mathematics or formal logic. [13] However, this did not exclude the possibility that this metaphysical background may provide some satisfaction to readers or listeners. As a matter of fact, the metaphysical discourse has been usually couched in language that has evoked a pleasant resonance in people's minds. This kind of language has been in use to present cherished types of science, ethics, or religion. If we disregard temporarily this background there is a large common ground between neo‑Thomism, neo‑Kantianism, and the new positivism. In extreme cases this common ground may be so extensive that one can read hundreds of pages of a neo‑Thomist or a neo‑Kantian without recognizing that he is not a positivist of the new type. The most outstanding example is the French physicist and philosopher, Pierre Duhem, whom we mentioned above. His writings are among the most valuable contributions of the new positivism. He was warmly recommended by Ernst Mach as a positivist. Many scientists were never aware that Duhem's background was straight Aristotelian or rather Thomistic metaphysics.

12. "New Wine into New Bottles"

The neo‑Thomist and neo‑Kantian schools reacted to the revolutionary changes that have arisen in science since the turn of the century by establishing a kind of "iron curtain" between science and philosophy. But none of these schools, and, as a matter of fact, none of the schools of traditional philosophy, of the idealistic or realistic type, were able to make a valuable contribution toward integrating the new science of the twentieth century into the general framework of human thought. From the viewpoint of intellectual history it is fair to say that the neo‑Thomist and neo‑Kantian schools have contributed, in a way, to the advance of scientific thought. They have helped to disintegrate the traditional systems to such a point that the remaining parts of the structure could easily merge with the new philosophy that would eventually arise on the basis of a new science.

Immediately after Einstein had published his general theory of relativity (1917), in which he advanced his new physics in full generality, writings appeared that did not attempt to integrate the new physics into traditional philosophy but to build up a new philosophy on the basis of. that new science. These writings did not follow the reaction of traditional philosophy—of either the fundamentalist or the modernist types—to the new science. Theirs was a radical reaction in accordance with the words of the gospel:

No man putteth new wine into old bottles; else the new wine will burst the bottles, and be spilled, and the bottles shall perish. But new wine must be put into new bottles; and both are preserved. No man also having drunk old wine straightway desireth new: for he saith, The old is better.

The old bottles were the patterns of traditional philosophy and the new wine was twentieth‑century science. A group of men went in for new bottles. They did not borrow the framework of Thomistic or Kantian metaphysics but they borrowed a pattern that had grown up in the soil of modem science, the pattern of the "new positivism." While men like Poincaré and Duhem had used this pattern for strictly domestic consumption, to clear up their own back yard, the foundations of science, the new men who emerged after 1917 ventured to build up a new philosophy that was expected to replace the traditional systems of the Aristotelian or Kantian type.

The new movement started about the time when the first world war ended (1918). New democratic republics were established in Central Europe: Austria, Czechoslovakia, Poland, and the Weimar experiment in Germany. They offered a favorable soil for the evolution of a scientific world conception. A similar situation seemed to arise in Russia after the overthrow of the Czarist regime (1917). It is interesting to note how the turn from the democratic start to the establishment of a new authoritarianism was accompanied by a turn from the philosophy of the new positivism to a philosophy which was nearer to the Aristotelian and Kantian tradition.

The first peak of the Central European movement toward a scientific world conception was reached about 1920. We can characterize it by three books: M. Schlick, "General Theory of Knowledge" [14] (1918); H. Reichenbach, "Theory of Relativity and Cognition a priori" [15] (1920); and L. Wittgenstein, Tractatus Logico‑Philosophicus (1921). [16] The link between these books and Einstein's theory is M. Schlick's small book "Space and Time in Contemporary Physics" (1917)," [17] in which the author attempts an integration of the new positivism with the ideas that have grown out of Einstein's new science.

13. M. Schlick's New Criterion of Truth

The views of the new positivism just before Schlick entered the picture are presented in Chapter 2, which was written in the same year as Schlick's "Space and Time." In this chapter a distinction is made between observational concepts--­red, warm, . . .—and auxiliary concepts—force, electric charge, . . . All statements of science that contain auxiliary concepts are translatable into statements containing observational concepts only. The expression "auxiliary concept" was often interpreted as meaning that by words like "force," electric charge, “potential," no "element of reality" is denoted. Schlick, however, stresses that the usefulness of a statement in science depends only upon whether this statement can be checked by sense observations. In other words, a scientific theory must consist of such principles that statements containing observational concepts only can be logically derived from them. It is irrelevant whether the concepts in the theory itself are "observational concepts" or "auxiliary concepts."

Schlick says:

There is no argument whatsoever to force us to state that only the intuitional elements [i.e., observational concepts], colors, tones, etc., exist in the world. We might just as well assume that elements or qualities which cannot be directly experienced also exist. These can likewise be termed “real," whether they be comparable with intuitional ones or not. For example, electric forces can just as well signify elements of reality as colors and tones. They are measurable, and there is no reason why epistemology should reject the criterion for reality which is used in physics. [18]

To say that electric forces are "measurable" means that from statements of the  form, "At this point of space acts an electric force of one hundred grams," statements can be derived about the deviation of a pointer from its coincidence with a certain mark of a scale. This means, however, statements about direct sense observation. It does not matter whether the words in a statement denote observational concepts or auxiliary concepts provided that results can be derived that contain only observational (intuitional) terms.

For otherwise no measurement would be possible. Therefore, the operational definitions"—which link auxiliary and observational concepts—are an indispensable part of every theory. This point is very important for integrating Einstein's new science into the scheme of new positivism. Now, one can, without any trouble, introduce terms like "four‑dimensional space" or "curved space" into physics. These concepts are as legitimate as three‑dimensional space and Euclidean geometry. We have only to make sure that our systems contain the rules by which, from the facts of the theory, only observational concepts can be derived.

From this characterization of scientific theories Schlick proceeds to an even more radical departure from traditional philosophy. He starts from the method that is generally used in science to check the "truth" of a theory. One considers, for example, the electric charge of an electron, which, according to the theory, should have a constant numerical value. Then one derives from the theory by a chain of conclusions the result that "the charge of the electron is 4.7 x 10-10 electrostatic units." Then one tries to find, on the basis of the same theory, a second chain of conclusions by which the numerical value of the charge of the electron can be derived. If this value is different from the first one, we say that our theory is "not true." If we obtain the same value as by the first method, we say that our theory may be true or "is confirmed." The "degree of confirmation" is the higher the more independent chains we can find that lead to the same numerical value for the charge and for other concepts of the theory.

This criterion of truth can also be formulated in a more general way. The symbols of the theory (observational or auxiliary concepts) are the state variables by means of which the course of events in the physical world—the facts of the world—can be described. Every specific state of the world is defined by specific numerical values which are assigned to these variables. Some of them, like the charge of the electron, are constants. Others, like the coordinates of an electron, are variables. As we learned by the example of the charge of the electron, we can make use of the theory to calculate these numerical values. This means that the theory indicates the facts of the world. If we obtained from two different chains of conclusions different numerical values for a symbol which, according to the theory, should be constant, then one and the same theory would indicate two incompatible facts. Such a theory would be useless.

From these remarks one will understand fairly well Schlick's criterion for the "truth" of a theory. He says:

Every theory is composed of a network of conceptions and judgments, and is correct or true if the system of judgments indicates the world of facts uniquely. [19]

If a theory yields two different values for the charge of the electron, the correspondence between the theory (concepts and judgments) and the facts would not be a unique correspondence.

On the other hand, no harm is done if different theories indicate one and the same world of facts. In the direction from the facts to the theory no uniqueness is required of "true" theory. Schlick says:

It is, however, possible to indicate identically the same set of facts by means of various systems of judgments; and consequently there can be various theories in which the criterion of truth [unique correspondence] is equally well satisfied . . . They are merely different systems of symbols, which are allocated to the same objective reality. [20]

14. M. Schlick and H. Reichenbach

From this analysis of scientific theories Schlick proceeded to the claim that every cognition, in whatever domain of knowledge, is essentially the establishment of a correspondence between the facts of the world and a system of symbols. Since between these symbols a set of relations—for example, the axioms of geometry or mechanics—is established, an arbitrary correspondence would frequently assign several worlds of facts to the same set of symbols. Then the cognition is false. According to Schlick, a cognition is "true" if the correspondence established is unique. A system of symbols indicates the facts of the world uniquely.

This conception of cognition and truth was a radical break with almost all systems of traditional philosophy, according to which cognition meant the finding of a truth that was hidden behind the appearances and could be discovered there by the power of reason, which the trained philosopher was supposed to possess. According to Schlick, however, cognition is the establishment of a correspondence; this means, primarily, building up a system of symbols with relations among them. Cognition becomes an activity, the construction of a system of symbols that has only to fulfill the requirement of uniqueness. If we start from this conception of cognition, most of the metaphysical problems that have puzzled generations of philosophers lose their insoluble aspect.

Consider, for example, the ancient puzzle whether the world is essentially mind or matter, the answer to which has been the shibboleth for distinguishing between idealists and materialists. Schlick would reformulate the problem: Are mental concepts or physical concepts better suited to build up a system of symbols that could indicate the world of facts uniquely? The problem loses its yes‑or‑no character and becomes a problem of deciding, on grounds of convenience, between two ways of describing the facts of the world.

Another author who ventured at that time to build up, on the basis of the "new positivism" and Einstein's new theories, a philosophy of science that would replace the traditional schools was Hans Reichenbach. His most significant books in this line, which to a certain extent paralleled Schlick's writings, are the "Theory of Relativity and Cognition a priori" and "Axiomatics of the Relativistic Theory of Space and Time." [21] Reichenbach went in much more for the technical discussion of physical and philosophical problems than Schlick did. He approved Schlick's conception of "true cognition," but he presented it in a way that was closer to the conception of science advanced by men like Poincaré. and Duhem. Reichenbach was perhaps the first one who formulated explicitly the requirement that every theory in which non-observational concepts appear contain relations between these abstract concepts and observational concepts. He stressed that geometry as an empirical science must contain, besides the geometric axioms (relations among geometric concepts), the description of a method whereby the straightness of a line can be tested experimentally. As the axioms of geometry and the description of measurements form a net, we can regard the whole theory as a prescription for coordinating the abstract concepts of geometry with observational facts. In this sense, Reichenbach regarded geometry as a system of "axioms of coordination.'' His criterion of truth was then similar to Schlick's, namely, that the axioms of coordination should be unique.

This uniqueness could, as he stressed, be correctly decided only by observation. On this basis, Reichenbach discussed the question whether the new philosophy was compatible with the traditional Kantian system. By Einstein's new science it was demonstrated that the Euclidean axioms of coordination may be false. Einstein described definite experiments by which one can test whether the traditional axioms of coordination lead to contradictions. Kant believed that the uniqueness of Euclid's system of geometry is founded in the organization of the human mind, while Reichenbach stressed that it has to be checked by experience. In the discussion (Secs. 3 and 4) of the relation of Poincaré to Kant I made partial use of Reichenbach's views.

15. L. Wittgenstein and R. Carnap

The work done by Schlick and Reichenbach made a strong impression upon our Viennese group (Sec. 1). In our search for a scientifically founded philosophy we were glad to find collaborators who attacked the task, I would say, from a more professional angle. At that time (after 1920) Hans Hahn was professor of mathematics at the University of Vienna, Otto Neurath started working for the City of Vienna, organizing adult education in the social sciences, and I had been since 1912 professor of theoretical physics at the University of Prague in the new Czechoslovakian Republic. Hahn had started intensive work with advanced students in the field of symbolic logic and the foundations of mathematics. In 1922, he chose as a basis of their discussions the new book by L. Wittgenstein, Tractatus Logico‑Philosophicus.

These discussions were the germ of many future developments in the philosophy of science.

Wittgenstein's book followed in some respects the same line as Schlick's and Reichenbach's. Wittgenstein, who was also of Viennese origin, was a student of Bertrand Russell, the British logician and philosopher. He added a new and important component to the integration performed by Schlick. Without making much use of the technicalities of symbolic logic, Wittgenstein showed that the new philosophy could be brought into a more perfect and coherent shape with the aid of the basic ideas of Russell's logic. Wittgenstein's formulations sounded even more straightforward and provoking than Schlick's, Reichenbach's, and even Russell's. Wittgenstein claimed bluntly that the problems of traditional philosophy are merely verbal problems. Our ordinary language, which has grown up to describe the facts of everyday life, is not adapted to the task of expressing and answering problems put to traditional philosophy. If we try to use our ordinary language in this way, we get into trouble. The real problem is to find out what one actually can say clearly. The world of facts can be described in our ordinary language; therefore, says Wittgenstein, "to understand a proposition means to know what is the case if it is true."

This conception of "meaning" and "understanding" is essentially no different from Schlick's, Reichenbach's, or, as a matter of fact, Mach's, if we understand Mach as he is interpreted in Chapter 2 of the present book. Every student of philosophy will also remember C. S. Peirce's conception of "meaning," and even William James's, who said that the meaning of a sentence is its "cash value."

Wittgenstein's merit, however, was his precise logical formulation and his cutting and striking dialectic. His line was later called, quite adequately, "therapeutical positivism." Hahn became very enthusiastic, starting a close cooperation of the new men with our Viennese group. He envisaged the appointment of M Schlick as a professor of philosophy at the University of Vienna. He met, of course, a stiff resistance among the adherents of traditional philosophy. But the interest of the scientists in the philosophical background of science has been traditionally high at the University of Vienna. Ernst Mach had owed his appointment to this predilection and Hahn succeeded in enlisting a sufficient number of scientists in a drive to carry through Schlick's appointment in 1922. In this year a close cooperation between Schlick and the old Vienna group began. This common work gained a great deal in intensity and momentum when Schlick persuaded R. Carnap to move to Vienna in 1926.

Carnap gave the new philosophy its "classical" shape. He coined many of its terms and phrases and endowed it with subtlety and simplicity. In the form created by Carnap it became a center of interest and a target of attack on a large scale.

Schlick and Reichenbach had identified "true cognition" with a system of symbols that indicated the world of facts uniquely. Carnap offered an example of such a system in his book "The Logical Structure of the World." [22] In this book the integration of Mach and Poincaré was actually performed in a coherent system of conspicuous logical simplicity. Our Viennese group saw in Carnap's work the synthesis that we had advocated for many years.

Carnap introduced as the elementary concepts of his system immediate sense impressions and the relations of similarity and diversity between them. The world is to be described by statements that may contain any symbols, provided that from them statements can be logically derived that contain nothing but assertions about similarity between sense impressions. The "meaning" of a statement in science would be the sum of all statements about similarity and diversity between sense impressions that can be derived logically from the statement in question. When I read this book it reminded me strongly of William James's pragmatic requirement, that the meaning of any statement is given by its "cash value," that is, by what it means as a direction for human behavior. I wrote immediately to Carnap, "What you advocate is pragmatism." This was as astonishing to him as it had been to me. We noticed that our group, which lived in an environment of idealistic philosophy had eventually reached conclusions by which we could find kindred spirits beyond the Atlantic in the United States.

From Carnap's presentation it was clear that a system from which one cannot derive results about similarities between sense impressions cannot be a "true cognition." The statements of traditional metaphysics, such as those about the existence or nonexistence of the external world, can obviously not be statements of the required type. For this reason Carnap said that "metaphysics is meaningless."

16. O. Neurath's "Index of Prohibited Words"

The men who had expanded the new positivism into a general logical basis of human thought—Schlick and Carnap—came now into personal contact with the original Viennese group, particularly with Hahn and Neurath, while my own contact was restricted to the time of the university vacations. As a result of the developing cooperation the new philosophy became more and more different from the traditional German philosophy, to which both Schlick and Carnap were bound to have some sentimental ties originating from their training at German universities.

They had demonstrated logically that no scientific metaphysics is possible because metaphysical statements do not fit into the pattern that statements must have in order to be called true or false. But the social scientist Neurath investigated the meaning of metaphysical statements as social phenomena. He insisted with a certain ruthlessness that no formulation should be allowed to slip in that would "give comfort to the enemy," even if it would be admissible from the purely logical viewpoint. The whole original Viennese group was convinced that the elimination of metaphysics not only was a question of a better logic but was of great relevance for the social and cultural life. They were also convinced that the elimination of metaphysics would deprive the groups that we call today totalitarian of their scientific and philosophic basis and would lay bare the fact that these groups are actually fighting for special interests of some kind.

In the long thorough and intimate discussions that the Viennese group had with Schlick and Carnap, the point was made that the new philosophy must be built up in such a way that no misinterpretation in favor of metaphysics could occur. We all knew that misinterpretations were bound to happen if and when expressions like "real," "essential,” “real building stone of the universe," were used in a loose way. Neurath even recommended, half jokingly, that an "index of prohibited words" should be set up. In a monograph [23] on the "Foundations of the Social Sciences" he avoided, as he explicitly states, words like "en­tity," "essence," "mind," "matter," "reality," "thing."

A well‑known example of misinterpretation of this type is the praise or condemnation of Mach's philosophy as a brand of idealism because his doctrine was often presented as claiming that the "world consists actually only of sensations." This was interpreted again as meaning that the "world is essentially mental." This interpretation accounts for Lenin's violent attack on Mach and for the extremely hostile attitude of the official Soviet philosophy against all doctrines that traced their origin back to Mach. This holds also for the new positivism and its generalization as achieved by Schlick, Reichenbach, and Carnap.

Perhaps the most striking effect of the cooperation of Schlick and Carnap with the old Viennese group was the shift to "physicalism" and to the "unity of science." Neurath had been particularly eager to prohibit any establishment of a metaphysical doctrine by a tactic of infiltration. He suggested that sense data should be dropped as elementary concepts of the logical structure of the world and replaced by physical things. Instead of building up the system of human knowledge upon concepts like "red spot" or "feeling of warmth," one should use elementary symbols expressing concepts like "rock" or "table," and define "redness," or "warmth" as derived concepts. As the starting point in sensation had a certain tint of idealism, so the new starting point had a tint of materialism. Carnap had in his "Logical Structure of the World" spoken of "methodical materialism" as a possible language for his system. But he had come to prefer "phenomenal language," statements in terms of sense impressions. Neurath worked out a system based on physical things as elementary concepts and called by him "physicalism." Carnap refined Neurath's physicalism to a precise logical structure and even constructed a "physicalistic language" for the field of psychology.

This transition from a quasi‑idealistic to a quasi‑materialistic language, which took place in our group about 1930, has been misunderstood by great many authors. They interpret it as a sudden jump into an opposite type of philosophy. As a matter of fact, the "jump" was an expression of our firm belief that the difference between an idealistic and a materialistic system is logically and scientifically of little importance and that there is actually only a difference of emphasis. The choice is determined largely by the emotional connotations or, in other words, by the language in which the pattern of our general culture is usually described.

17. O. Neurath and the "Unity of Science Movement"

The second reformulation suggested by Neurath was the characterization of the new movement as a work "towards the unification of science" or "for the construction of a unified science." This shift was for many people surprising too. Schlick and Carnap had stressed the point that there are no philosophic propositions but that there is a philosophic activity that consists in the clarification of the statements of the special sciences. This meant, briefly, that philosophy was to interpret the abstract and symbolic principles of science as statements about physical things. Frequently, particularly in Great Britain, the new philosophy has been distinguished from the traditional philosophy as "analytic philosophy" in contrast to “speculative philosophy." The work of men like Moore, Russell, or Wittgenstein has been described as analytic philosophy.

Our original Viennese group and particularly Neurath were not satisfied with ascribing to our new philosophical group mainly critical and analytical objectives. We knew well that man is longing for a philosophy of integration. If the new philosophy refuses to serve the cause of integration, a great many people, including even scientists, would rather return to traditional metaphysics than be restricted to a purely analytic attitude. As a matter of fact, the traditional goal of "philosophy," through thousands of years of human knowledge, has been integration.

Neurath pointed out that by Carnap's analysis the statements of all sciences, not only physics but also biology and sociology, had been reduced to statements about physical things or about sense impressions. The traditional opinion about the individual sciences has, however, been that the fundamental concepts of biology are essentially different from those of physics, the concepts of psychology different from those of biology and so on. Physics, biology, and psychology have to do with different kinds of "being" and can never be united on the level of science. Only by introducing metaphysical concepts can one achieve a unification. If we accept, however, Carnap's analysis of science, it follows that all statements of science are of only one type, that is, they are statements that can be expressed in the "thing language." Hence, it must be possible to introduce a unified language for all the sciences and to create a system of "unified science," in which the "special sciences" are merely products of the division of labor. The terminology of the special sciences is practical for restricted purposes, but no philosophic implication about unbridgeable gaps can be drawn from the differences in terminology.

The new philosophy now described its work as the building up of a unified science. With this goal we returned in some measure to the classical goal of philosophy as defined by Aristotle. As a matter of fact, August Comte, the father of positivism, said (1829):

I employ the word "philosophy" in the sense that was given to it by Aristotle, as denoting the general system of human conceptions.

Our group did not wish to stress the work on analysis in contrast to the creation of a synthesis. We never regarded the logic and analysis of science as a goal in itself; we believed strongly that this analysis is a necessary part of obtaining an unprejudiced outlook on life. The close connection between the positivistic attitude and the unification of science can be traced back to Ernst Mach himself. In Chapter 3 of the present book I analyze Mach's philosophy from the viewpoint of 1938, the one‑hundredth anniversary of his birth, and point out to what degree the further evolution of positivism was anticipated in. Mach's writings.

18. The Vienna Circle

In 1929, we had the feeling that from the cooperation that was centered in Vienna a definite new type of philosophy had emerged. As every father likes to show photographs of his baby, we were looking for means of communication. We wanted to present our brain child to the world at large, to find out its reaction, and to receive new stimulation.

We decided first to publish a monograph about our movement, next, to arrange a meeting, and eventually to get control of a philosophical journal so that we would have a way of getting the contributions of our group printed.

When we prepared the monograph we noticed that our group and our philosophy had no name. Quite a few people among us disliked the words "philosophy" and "positivism" and did not want them to appear in the title. Some disliked all "isms," foreign or domestic. Eventually we chose the name "scientific world conception." [24] Some of us, particularly Schlick, thought that every reasonable scientist would agree with our presentation of cognition. Our chosen title seemed a little dry to Neurath, and he suggested adding "The Vienna Circle," because be thought that this name would be reminiscent of the Viennese waltz, the Vienna woods, and other things on the pleasant side of life. The monograph [25] was written by Carnap, Hahn, and Neurath in close cooperation.

Two years later, A. Blumberg and H. Feigl published in the United States a paper, "Logical positivism: A new movement in European philosophy," [26] and provided the "scientific world conception" with its international trade name.

In accord with the historical and cultural inclination of the Viennese group, Neurath was eager to trace the genealogic lineage of our movement. In the monograph he recorded the following lines:

Positivism and empiricism: Hume, the philosophers of the Enlightenment, Comte, Mill, Avenarius, Mach.

Scientific method: Helmholtz, Riemann, Mach, Poincaré, Enriques, Duhem, Boltzmann, Einstein.

Symbolic logic and its application to reality: Leibniz, Peano, Frege, Schroeder, Russell, Whitehead, Wittgenstein.

Eudaemonistic ethics and positivistic sociology: Epicurus, Hume, Bentham, Mill, Comte, Feuerbach, Marx, Spencer, Mueller‑Lyer, Popper‑Lynkeus, Carl Menger.

A great many misunderstandings have been current about the original doctrine of the Vienna Circle, which became the germ of logical positivism. Again and again philosophers have sought to prove to the “positivists" that there is a "real world" and that science explores this real world. In fact, the monograph says: "Something is real if it is a part of the system of symbols that denotes the world of facts." A number of authors have tried to refute "positivism" by showing that there are also unobservable elements in science that have to be called "real," for example, the electron. The quoted passage clearly expresses the fact that "real" is not regarded as identical with "observable." We can state the meaning of the quotation also in slightly different words: Every symbol denotes something real if this symbol is a part of a that serves to describe observable facts uniquely.

19. The First Public Meeting

The arrangement of the meeting was not so easy. We wanted to reach a large audience. The ordinary regular philosophy meetings followed the traditional lines and would hardly have given us enough scope. By a happy coincidence I was just in 1929 arranging a meeting of the physicists and mathematicians from the German‑speaking regions in Central Europe. The meeting was to be held in my place of residence, Prague, the capital of Czechoslovakia. The German Physical Society, which was the official sponsor of this meeting, did not particularly like the idea of combining this serious scientific meeting with such a foolish thing as philosophy. However, I was the chairman of the local committee in Prague and they could not refuse my serious wish to attach a meeting with the topic, "Epistemology of the Exact Sciences." This meeting was to be sponsored by the Ernst Mach Association, which was the legal organization of the Vienna Circle, and the Society for Empirical Philosophy, which was organized in Berlin and followed in general the line of H. Reichenbach. In this way we provided a nucleus of interested people and hoped that quite a few mathematicians and physicists who came to Prague for their meeting would also attend our gathering.

Some scientists wanted to minimize our program and predicted that we would have no audience from the ranks of the exact scientists. As a matter of fact, our addresses had a larger audience than papers on special scientific problems. I had prepared an elaborate paper that was intended to give the scientists a kind of preview of our ideas and to prove that the new line in philosophy is the necessary result of the new trends in physics, particularly the theory of relativity and the quantum theory. I elaborated the contrast between the "school philosophy," which sought to preserve the traditional doctrines despite the new science, and the "scientific world conception," which wants to pour the new wine into new bottles. I stressed also the similarity between the new scientific world conception and the basic ideas of American pragmatism, particularly those of William James. This lecture is Chapter 4 of the present book.

Some friends cautioned me not to speak too bluntly. The audience, which consisted mostly of German scientists, knew little of philosophy, except that they had some sentimental ties to Kantianism. This doctrine was regarded in some intellectual quarters as a kind of substitute for the traditional forms of religion. My wife said to me after the lecture: "It was weird to listen. It seemed to me as if the words fell into the audience like drops into a well so deep that one cannot hear the drops striking bottom. Everything seemed to vanish without a trace."

There is no doubt that quite a few people in the audience were shocked by my blunt statements that modern science is incompatible with the traditional systems of philosophy. Probably, most of the scientists had not been accustomed to thinking of philosophy and science as one coherent system of thought. Philosophy had been for them what the Sunday sermon is for a businessman who is only interested in profit. Philosophy had been required not to be "true" but to give emotional satisfaction.

After the meeting, however, our committee received a great many letters from scientists who expressed their great satisfaction that an attempt has been made toward a coherent world conception without contradictions between science and philosophy. We even received one letter from a professor of philosophy at a German university who wanted to go on record with his conviction that the remaking of philosophy, along the lines that we followed at our meeting, is necessary.

20. M. Schlick Announces a "Turn in Philosophy"

In the years 1930‑31, there appeared the first volume of the journal Erkenntnis (Cognition), which became the main mouthpiece of our movement. The editors were R. Carnap and H. Reichenbach. The first issue began with the paper, "The Turn in Philosophy," by M. Schlick. I shall quote some lines in order to show that an optimistic belief in the new trend was the keynote of this journal. Schlick writes:

I am convinced that we are in the middle of an altogether final turn in philosophy. I am justified, on good grounds, in regarding the sterile conflict of systems as settled. Our time, so I claim, possesses already the methods by which any conflict of this kind is rendered superfluous; what matters is only to apply these methods resolutely.

In the same year (1930) Schlick published a paper, "Personal Experience, Cognition, Metaphysics," [27] in which he writes:

All cognition of the being is achieved, in principle, by the methods of the special sciences; every other kind of ontology is empty talk. Metaphysics is impossible because its goals contradict one another. If the metaphysician longs only for personal experience, his longing can be satisfied by poetry and art—or by life itself. But if he longs for a personal experience of the transcendent, he confuses life and cognition, he chases futile shadows.

For Schlick, as we know, "cognition" is the construction of a system of symbols that denotes uniquely the world of facts. It is therefore fundamentally different from personal experience.

This strong optimistic feeling is psychologically the feeling of a turn. You can ride in a car at high speed and you do not feel anything so long as the velocity remains unchanged. But if a turn or an acceleration takes place, you experience a strong reaction. Today, the movement of logical positivism is no longer so conspicuous. It had produced a turn in philosophy, which afterwards moved in a new direction and rather smoothly. I quote a passage from a philosopher who is by no means a follower of what is now called logical positivism. C. West Churchman writes: [28]

Few can doubt the healthy impact that the positivist position has had upon modes of inquiry; it has sharply distinguished the schools of thought, and has raised a standard under which the proponents of experimental methods can fight their battles against a reactionary movement. To return to a prepositivistic viewpoint is to return to a prescientific viewpoint, to become a reactionary as an advocate of the indisputable power of the sovereign in the eyes of one with a democratic outlook.

We find a similar position even in the most recent book of F. S. C. Northrop [29] who, in some sense, attempts a justification of metaphysics. He has been, however, in all his writing a very independent thinker who has given much thought to the foundations of science and particularly to the interdependence between the philosophy of science and the cultural background. He writes:

In any event, the great merit of logical positivism and its main aim is satisfied, even if one leaves the scientific concepts and their meaning just as one finds them, as prescribed by the scientists in the postulates of some specific deductively formulated theory. The important desideratum at which the logical positivists were aiming, namely operational verification, can nonetheless be obtained. There are many signs that contemporary logical positivists have now come to this position.

As a matter of fact, if one traces the history of logical positivism, one will see that this has been always the position of the "scientific world conception." This becomes clear if one considers Schlick's position in "Space and Time" (Sec. 13).

To estimate how sharp the turn was for which logical positivism was responsible, we have to compare its position with the views of the school of traditional philosophy that was the nearest to it in spirit and in time. We choose for comparison H. Vaihinger's "Philosophy of 'As If '" (1911), a very ingenious and in its time very famous book, a typical example of what I called the disintegration of traditional philosophy by neo‑Kantianism. Vaihinger tries to show that the concept of an atom (which he identifies with a mass point) in physics is a useful “fiction" although it is logically self‑contradictory. He says:

An entity without extension that is at the same time a substantial bearer of forces—this is simply a combination of words with which no definite meaning can be connected. "Simple atoms," that must yet be something material, cannot be causae verae, cannot be actual things. Since, however, the physicist does require atoms for his construction, how is this contradiction to be solved? How are we to rescue science from this dilemma? [30]

Vaihinger thinks that the method actually used in science is to

speak of atoms without really meaning to assume them . . . Unquestionably this conceptual method is the most convenient one, but this constitutes, of course, no proof of its objective‑metaphysical validity. [31]

Vaihinger's view is a clear indication of the situation in philosophy immediately before the rise of logical positivism. There was a complete lack of understanding that one must distinguish between a structural system having exact logical coherence with the world of facts, which are described with a certain vagueness, and the operational definitions, which connect both domains and participate in the preciseness of the first and the vagueness of the second.

21. P. W. Bridgman's Theory of Meaning

About the same time when the "scientific world conception" group were arranging their first public meeting, P. W. Bridgman published a book [8] in the United States in which he reacted to the same situation by which this group had been faced. In a broad sense, we can characterize his work also as an attempt to integrate Mach, Poincaré, and Einstein into a coherent picture of modern science. Bridgman’s field was not mathematics or symbolic logic but experimental physics. He has been a man of the laboratory who preferred to do things rather than set up a long chain of arguments. His approach is therefore different from that of the Central European group. It was, in a way, more similar to Mach's, who was also essentially an experimentalist. Bridgman found out what was the salient point in the integration of Mach and Poincaré, Reichenbach had explicitly pointed out that what is needed is a bridge between the symbolic system of axioms and the protocols of the laboratory. But the nature of this bridge bad been only vaguely described. Bridgman was the first who said precisely that these "relations of coordination" consist in the description of physical operations. He called them, therefore, "operational definitions." This name has been generally accepted. Bridgman was also very definite in stating that a theory which does not contain the operational definitions of its abstract terms is meaningless. In this way he arrived at a concept of "meaning" and "meaninglessness" that was similar to the concept advanced by Carnap and his group.

Bridgman, however, formulated the criterion of meaning in a much more concrete way. He did not restrict himself to the general prescription of how to investigate the meaning of a statement but investigated elaborately the operations that have to be carried out in order to define the meaning of physical terms like "length" or "thermal capacity." By these investigations he found, for example, that the operations by which one can distinguish between heat conduction and heat radiation, or between heat supplied and mechanical work supplied, cannot be performed in all cases. Only if the phenomena investigated are of a simple type are these operations feasible. Therefore, terms like "heat conduction" or "mechanical work" do not have a meaning under all circumstances.

Bridgman has contributed much to the new philosophy which has developed along with twentieth‑century science. He has advocated strongly the view that the domain of phenomena within which a word has meaning is restricted, and no word has meaning if we do not indicate the circumstances under which it is used. Bridgman has also pointed out repeatedly that this new "semantic" aspect is bound to have great repercussions upon discourse in politics and religion. [32]

In 1931 Carnap was appointed professor of natural philosophy at the University of Prague. I succeeded in bringing about this appointment, despite the strong opposition of the adherents of traditional philosophy, because of a happy coincidence. At that time the Faculty of Arts and Sciences was divided into a Faculty of Humanities and a Faculty of Science. All professors of philosophy were in the Faculty of Humanities and the Faculty of Science gave no instruction in philosophy. The president of the Czechoslovakian Republic, Thomas G. Masaryk, who had himself been a professor of philosophy, believed strongly in the educational value of philosophy. He insisted that the Faculty of Science should have a philosopher of their own. I suggested Carnap, and as there was no advocate of traditional philosophy left, the science faculty agreed. From 1931 on we had in this way a new center of "scientific world conception" at the University of Prague.

22. The Spread of Logical Empiricism

My own interest, which had been for a long time diverted from the problems offered by the philosophy of science, returned now to the object of my earlier years. The intellectual situation was now in a certain respect a similar one. The new science of quantum theory gave rise to a repetition of the crisis that had been precipitated about 1905 by the relativity theory, but with even greater intensity. Again it was maintained that scientific method had failed. The new theories do not even claim to give an "explanation" of the physical phenomena. They claim only to offer mathematical formulas from which the observed phenomena can be derived. The "explanation" is left as a field for metaphysical theories, which would claim to give the "real causes" of things. The argument went mostly that relativity as well as quantum theory give mathematical patterns without any causal justification.

Remembering our old arguments in the Vienna coffeehouses around 1907 about Abel Rey, Ernst Mach, and Henri Poincaré, I devoted some work to applying the newly developed "scientific world conception" to overcome the new crisis. I tried to show that there is not the slightest reason to see in twentieth‑century theory an argument for an idealistic or spiritualistic world conception, and that this opinion only arises from a lack of scientific formulation of the new physical theories. This lack has its source in the poor training of physicists in philosophy, which makes them often faithful believers in the metaphysical creeds imbibed in their early youth "from a nurse or a schoolmaster." [33] In Chapters 6 and 7 of the present book I attempt to give an analysis of physical theories on the basis of the new ideas; the scientific argument is carried through consistently without suddenly breaking into vague metaphysical discourse. While these chapters are more or less devoted to a general discussion of modern physics, Chapters 8 and 9 contain a special discussion of modern quantum theory from the same viewpoint, including some remarks on "determinism" that take up, in a new way, the problem of Chapter 1.

If we look from another angle upon those misinterpretations of physical theories, it is evident that they are not the result of some intellectual inability. Their real source is the urge to find support for a metaphysical creed that, for some reason, one cherishes. And this reason is, as we have already hinted, the fitness of this metaphysical creed to bolster up some political or religious creed that one believes to be indispensable for the well‑being of mankind. This sociological aspect has been for many years familiar to me from the discussions of our old Viennese group and, in particular, from the attention that I later paid to the influence of religious and political creeds upon scientific theories in two specific cases: Duhem's presentation of the action taken by the Roman Church against the Copernican theory, and Lenin's attacks on Mach's conception of physics. I have touched upon this aspect in Chapter 5 and have devoted all of Chapter 10 to it. The Copernican conflict is treated in Chapter 13, the specific nature of the Soviet philosophy (dialectical materialism) and, in particular, its relation to positivism and empiricism are described in Chapter 11.

At the time of the meeting in Prague (1929) the Vienna Circle and Reichenbach's group in Berlin were a small number of dissident people hemmed in by the vast ocean of German school philosophy, which was more or less a development of Kantian metaphysics. It was considered to be a specific "German philosophy"—­namely, "German idealism"—and philosophies of other types were regarded with a certain suspicion as something "un‑German" and "foreign." The workers for a "scientific world conception" had no hope of finding any considerable encouragement in Germany. Neurath described this cultural and historical situation in his small book, "The Development of the Vienna Circle and the Future of Logical Empiricism." [34] Neurath writes that the Kantian influence had been slight in the Universities of Vienna and Prague and that their philosophy avoided the "Kantian interlude" and passed directly from Leibniz to modern positivism. He continues, "The influences of English and French thinkers are frequent and things happen in Austria parallel to what happens in Warsaw, Cambridge, or Paris, rather than to what takes place in Berlin."

Although German science had developed along international lines and no serious scientist would have objected to any foreign influence, in philosophy things were different. There was a strong tendency to overrate "German idealism" and to minimize British, French, and American philosophical trends. We felt very soon that the future of the "scientific world conception" was to break through this wall and to make contact with "foreign" philosophies. This feeling turned out to be entirely correct and we met friendly interest on the part of American, British, and French thinkers. In the United States there was a natural common ground, the work of the American pragmatists, in particular of C. S. Peirce. Charles W. Morris had cultivated especially the ties between pragmatism and the Central European positivism. He coined for the result of the very close cooperation of these groups the name "logical empiricism," which seems to me to denote the salient point better than any other name. E. Nagel (now at Columbia) and W. V. Quine (now at Harvard) came to Vienna and Prague, as Morris (now at the University of Chicago) had done, to make personal contact with Schlick, Carnap, and the other workers in this field.

In Great Britain our natural link was L. Wittgenstein, whose own education had been half Austrian and half British, and, through him, Bertrand Russell. The brilliant young Oxford philosopher, A. J. Ayer, also came to Vienna and published the most readable book on logical empiricism that has been written in English, and perhaps the most readable book altogether. [35]

In France, the advocates of the "new positivism" were naturally interested in our movement. L. Rougier started his philosophic work on a basis similar to that of Schlick. He took his start from Poincaré, tried to integrate Einstein into the "new positivism," and wrote the best all-round criticism of the school philosophy that I know of, "The paralogisms of rationalism." [36]

Marcel Boll, an able physicist himself, saw in the Viennese movement a valuable contribution to a renewed and more vigorous positivism and a support of progressive thinking and acting. He translated writings of Carnap, Reichenbach, Schlick, and myself into French.

We encountered also a strong interest in a group of French neo-Thomists. The great influence of the neo‑Thomist, P. Duhem, upon the Viennese group repeated itself now in reverse. The French general, Vouillemin, recommended our group because we replaced the spelling "Science" modestly by "science." He also translated several papers of the Vienna Circle into French and published a small book, "The Logic of Science and the Vienna School," [37] in which he gave his interpretation of logical empiricism. The French neo‑Thomists of this group saw in logical positivism the destroyers of idealistic and materialistic metaphysics, which they regarded as the most dangerous enemies of Thomism.

To organize this international cooperation a preliminary conference was held in 1934 in Prague, at which Charles Morris and L. Rougier participated. The ground was laid for arranging international congresses "for the unity of science," which were to be held every year. They actually met in 1935 in Paris, 1936 in Copenhagen, 1937 in Prague, 1938 in Cambridge, England, and 1939 in Cambridge, Massachusetts, at Harvard University.

23. Teaching the Philosophy of Science at Harvard

In the year 1936, just while the Congress for the Unity of Science was in session at Copenhagen, Professor Schlick was assassinated near his lecture hall in the University of Vienna by a student. At the court trial the attorney for the defendant pleaded extenuating circumstances because the student was indignant about Schlick's "vicious philosophy." Everyone who knew Schlick had been full of admiration for his noble, humane and restrained personality. The political implications of the expression "vicious philosophy" were obvious. The student received a ten‑year prison term. When, however, two years later, the Nazi troops occupied Vienna and arrested a great many people, Schlick's murderer was released from prison.

The shots directed at Schlick were a dramatic indication of the dispersal of the Central European positivism that was taking place under the pressure of the advancing Nazi power. At the end of 1938 this process was completed. By far the greatest part of the Central Europeans who had worked along the lines of logical positivism had left their countries. The immediate reason was either to escape political persecution, or, in many cases, just the feeling that under the dictatorship of the Nazis there would be no place for a philosophy guided by logic and experience. The majority of the emigrants have lived since in the United States, a smaller part in Great Britain.

When I arrived in the United States in October 1938, I started a lecture tour during which I spoke at twenty‑odd universities and colleges on the philosophic interpretations of modern physics. Chapter 7 of this book presents one of these lectures.

Since the fall of 1939 I have had the privilege of teaching at Harvard University not only mathematical physics but also the philosophy of science. This teaching has been a great experience for me and has been of great influence on my philosophical writing. I started with an audience of about fifteen students. Since this was an unusual subject I did not quite know what to tell them. I began by presenting to them the logical structure of physical theories as envisaged by logical empiricism. But very soon I noticed that this was not the right thing to do. The frequent discussions that I had with the students showed me what they really wanted to know. By a process of interaction, a program was finally worked out that was a compromise between what I wanted to tell the students and what they wanted to know.

At that time Harvard University set up its "General Education Program." This was a great help to me since it was based on the deficiencies of the traditional curriculum as felt by the students, the faculty, and the general public. President J. B. Conant urged, in particular, a new approach to the teaching of science. He stressed that science teaching has to be linked up with a presentation of the historical, cultural and psychological background of the work done by the great scientists. This program was later developed in Mr. Conant's book, On Understanding Science. [38]

Stimulated by all these factors and particularly by the rapidly increasing interest of the students in the philosophical and cultural implications of science (I had later an audience of more than two hundred fifty), it became more and more clear to me how to work out a program for my students. I now put the greatest emphasis on presenting physics, and science in general, as part of our general pattern of thinking and acting. I presented it on one hand as a logical system that has to be checked by physical experiments and on the other hand as one of the means of expressing man's attitude towards the world, the small world of society and politics and the large world that is our astronomical universe. This more historical approach has been familiar to me since my student years from the meetings with my older friends.

All my papers written after 1940 follow this line. Chapter 11 of this book describes the interaction between the advance in science and the changes in metaphysical systems. Its point is that a great many metaphysical systems are merely abandoned systems of science. In Chapter 12 the Copernican conflict is discussed from the same viewpoint, particular stress being laid on the role that could be played by the intervention of political powers. Chapters 14 and 15, which are closely connected with each other, discuss directly the role that instruction in the philosophy of science is to play in the college curriculum. By the time I came to write them, I had collected a great deal of experiential material about this problem. My point was now that the philosophy of science should, on one hand, give to the science student a more profound understanding in his own field, and on the other hand, be for all students a link between the sciences and the humanities, thus filling a real gap in our educational system.

In all my writing before 1947 I had stressed the point that science gives no support to metaphysical interpretations, of whatever type. I had discussed these interpretations only as reflecting the social environment of the philosopher. However, after that time, as a result of contact with my students and fellow teachers, I became more and more interested in the question of the actual meaning of the metaphysical interpretations of science—idealistic, materialistic, relativistic, and others. For the fact that a great many scientists and philosophers advance such interpretations and cherish them is as firmly established, by our experience, as any fact of physics.

I now began a new series of investigations into the meaning of metaphysics within the framework of logico‑empirical and socio‑psychological analysis. The first preliminary result of these investigations is published in Chapter 16 of the present book.

In 1940 Professor Harlow Shapley, Director of the Harvard College Observatory, introduced me into the Conference of Science, Philosophy and Religion, which meets every year. This Conference is a group of philosophers, educators, social workers, and ministers of all denominations, and includes a sprinkling of scientists. They are interested particularly in the contributions that each of these fields can make to the understanding and supporting of the democratic way of life. The leaders in this group have been, besides Dr. Shapley, Dr. Louis Finkelstein, President of the Jewish Theological Seminary of America, and Dr. Lyman Bryson, Director of the Educational Department of the Columbia Broadcasting System. From these meetings I have learned more about the attitudes of different groups of people toward science, and what points in the philosophy of science support or are believed to support a certain way of life. The large majority in these meetings has been rather critical of the scientific outlook and its contribution to human welfare.

I addressed this group several times between 1940 and 1947. My contributions centered mostly around the question of whether the "relativism" of modern science is actually harmful to the establishment of objective values in human life. I made an argument to prove that the "relativism of science" has also penetrated every argument about human behavior. "Relativism" is not responsible for any deterioration of human conduct. What one calls "relativism" is rather the attempt to get rid of empty slogans and to formulate the goals of human life sincerely and unambiguously. My contributions to these meetings will, be published [39]  in due time, under the title "Relativity—a richer truth."


1  A. Rey, La Théorie de physique chez les physiciens contemporains (Paris, 1907), pp. 16 ff. [—> main text]

Ibid., pp. 18 ff. [—> main text]

Ibid., pp. 392 ff. [—> main text]

4  P. Duhem, Théorie physique; son objet—son structure (Paris, 1906), p. 24. [—> main text]

Ibid., p. 285. [—> main text]

6  P. Duhem, "Quelques réflexions au sujet de la physique expérimentale," Revue des questions scientifiques 36, 179 (1897). [—> main text]

7  P. Duhem, "Physique et metaphysique," Revue des questions scientifiques 36, 55 (1897). [—> main text]

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

9  An English translation of this lecture appeared in A. Einstein, Sidelights on Relativity (London: Methuen, 1922). [—> main text]

10  J. Maritain, Distinguer pour unir, ou les degrés du savoir (Paris: Brouwer, 1935); English translation by B. Wall and M. R. Adamson (New York: Scribner, 1938). [—> main text]

11  E. Cassirer, Zur Einsteinschen relativitatstheorie (Berlin: Cassirer, 1921). [—> main text]

12 E. Cassirer, Determinismus und Indeterminismus in der modernen Physik . . . (Göteborg: Elanders, 1937). [—> main text]

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

14  M. Schlick, Allgemeine Erkenntnislehre (Berlin: Springer, 1918; ed. 2, 1925). [—> main text]

15  H. Reichenbach, Relativitatstheorie und Erkenntnis apriori (Berlin: Springer, 1920). [—> main text]

16  L. Wittgenstein, Tractatus logico‑philosophicus (London: Paul, Trench, Trubner; New York: Harcourt, Brace, 1922); German and English on opposite pages. [—> main text]

17  M. Schlick, Raum und Zeit in der gegenwärtigen Physik. . . . (Berlin: Springer, ed. 3, 1920); English translation by H. L. Brose (Oxford: Clarendon Press; New York: Oxford University Press, 1920). [—> main text]

18  Ibid., p. 84. [—> main text]

19  Ibid., p. 86. [—> main text]

20  Loc. cit. [—> main text]

21  H. Reichenbach, Axiomatik der relativistischen Raum‑Zeit‑lehre (Braunschweig: Vieweg, 1924). [—> main text]

22  R. Carnap, Der logische Aufbau der Welt (Berlin: Weltkreis‑Verlag, 1928). [—> main text]

23  International Encyclopedia of Unified Science (Chicago: University of Chicago Press, 1938‑39), vol. 2, no. 1. [—> main text]

24  We chose the term "world conception" (Weltauffassung) in order to avoid the German word Weltanschauung, which seemed to us loaded with metaphysical connotations. [—> main text]

25  Wissenschaftliche Weltauffassung der Wiener Kreis (Wien: A. Wolf, 1929). [—> main text]

26  A. Blumberg and H. Feigl, Journal of Philosophy 28, 281 (1931). [—> main text]

27  M. Schlick, "Erleben, Erkennen, Metaphysik," Kantstudien 31, 146 (1926). [—> main text]

28  C. W. Churchman, Theory of Experimental Inference (New York: MacmilIan, 1948). [—> main text]

29  F. S. C. Northrop, The Logic of the Sciences and the Humanities (New York: Macmillan, 1947), pp. 113, 114. [—> main text]

30  Hans Vaihinger, Die Philosophie der "Als Ob” (1911); English translation by C K. Ogden (ed. 2, Barnes and Noble, 1935), p. 219. Incidentally, in this book the term "logical positivism" was used for the first time, although in a somewhat different sense from the one it now has. [—> main text]

31  Ibid.,  p. 222. [—> main text]

32  P. W. Bridgman, Yale Review 34, 444 (1945). [—> main text]

33  A. N. Whitehead, The Principle of Relativity (Cambridge: The University Press, 1922). [—> main text]

34  O. Neurath, Le Développement du Cercle de Vienne et l'avenir de l'empirisme logique, French translation by General Vouillemin (Paris: Hermann, 1935). [—> main text]

35  A. J. Ayer, Language, Truth, and Logic (London: Gollancz; New York: Oxford University Press, 1936). [—> main text]

36  L. Rougier, Les paralogismes du rationalism (Paris: Alcan, 1920). [—> main text]

37  Le general C. E. Vouillemin, La Logique de la science et I'école de Vienne (Paris: Hermann, 1935). [—> main text]

38  J. B. Conant, On Understanding Science (New Haven: Yale University Press, 1947). [—> main text]

39  Beacon Press, Boston. [—> main text]

SOURCE: Frank, Philipp. Modern Science and Its Philosophy. Cambridge, MA: Harvard University Press, 1949. Reprint: New York: George Braziller, 1955. Introduction: Historical Background, pp. 1-52.

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 1 February 2006

Site ©1999-2023 Ralph Dumain