Problems of Empiricism *
 in Social Origins of Modern Science
Edgar Zilsel, Diederick Raven, Wolfgang Krohn, R. S. Cohen
Kluwer Academic Publishers, 2000, 171-199.
Modern science is accustomed to decide all questions about reality by experience and experiment. This remarkable attitude is not at all self-evident. It is rather a late achievement in the history of mankind, and its import cannot be fully understood as long as that fact is not realized. We shall start our exposition, therefore, with a retrospective view of the rise of empirical thinking and the experimental method. Though at the beginning of the modem era empirical research proceeded from certain empirical achievements of antiquity, classical empiricism can be omitted in this brief survey. In antiquity the empirical sciences were considerably surpassed in intellectual influence by metaphysics and rhetoric, and empiricists always were but a small minority among ancient philosophers and scientists. It will be sufficient, therefore, to begin with the Middle Ages.
The seats of medieval civilization were not towns, which in the early centuries were rare, but monasteries and castles in the country. The castles and the cultural accomplishments of the knights have little bearing on theoretical thinking. The monasteries, being the centers of medieval scholarship, are more important for the present study. Monks, by the conditions of their lives, are not much disposed to look at the world with open eyes. Inclosed within walls, intrusted with the task of transmitting established doctrines to successors by scholastic instruction, they were compelled to indulge in abstract reasoning and to develop their sagacity. This attitude of mind was later taken over by the universites of the late medieval cities. Up to the thirteenth century the method of investigating that appears to be the most natural one to modem science was practically unknown. When medieval theorists, or theologians, intended to solve a problem, they looked first for relevant passages in the Holy Scripture, the patristic writings, and certain works of Aristotle. Then they compared affirmative and negative statements of colleagues and predecessors and, finally, drew conclusions by means of logical deduction from the premises collected. It
* [Zilsel, as in his essay ‘The Methods of Humanism’, makes use of ‘supporting evidence paragraphs’ which are given in smaller print. The original endnotes have been replaced with footnotes; Eds.]
cannot be a surprise, therefore, that the scholastic theory of syllogism is the chief contribution of medieval science.
By the end of the Middle Ages, however, a few scholars, among them Roger Bacon in England and Albertus Magnus in Germany, had begun to understand the importance of experience.  Their contemporary, the French nobleman Petrus de Maricourt, even experimented successfully upon magnets. This rise of new scientific methods in the thirteenth century is connected with a fundamental change in society. Towns had gradually grown up, the rural monasteries and castles were losing their social importance, and a new social class - the townsmen - entered upon the stage of history. Money and profit began to rule the lives of nations.
The period of transition from feudalism to early capitalism lasted until the fifteenth century. At its beginning, the craftsmen of the towns, to prevent economic competition, organized themselves in guilds which took care lest the working traditions of the past be broken. But when, with the strengthening of capitalism, competition proved stronger and destroyed the guilds, guild constraints upon working began to crumble. The craftsman who worked exactly as his master and his master’s master had done, was surpassed by less conservative competitors; one’s own spirit of enterprise, one’s own experience and inventive genius, made the successful man. The age of inventions had begun. Some great inventions of this period - the manufacturing of paper, gunpowder, and guns, the mariner’s compass, the printing press - are generally known. The invention of the blast furnace and the stamping-mill, the introduction of ventilators and hauling-engines in mining, numerous improvements in construction of weaving looms, ships, canals, and fortresses are scarcely less important. With the inventions of the fifteenth and sixteenth centuries the technology of the Middle Ages was completely revolutionized. Similar effects were produced by the great geographical discoveries. On new shores animals, plants, and things never seen before were found, which even the most acute monk would not have been able to deduce from his authorities. Authorities and syllogisms had been beaten by experience; a new, empirically minded type of man went out to conquer the world.
In virtue of their occupations these men were craftsmen and navigators belonging to the lower ranks of society. They were not esteemed too highly either by the noblemen or by the bankers and rich merchants of plebeian origin. Since the literati, the humanists, who wrote for upper-class publicity were not interested in such plebeian people as craftsmen, the biographies and even the names of most of the inventors are seldom known. There are few detailed and reliable literary reports even on the great discoverers.  The craftsmen and sailors themselves were not literary men but rather uneducated people. Their
 The empirical achievements of Bacon and Albertus are often overestimated (cf. L. Thorndike, A History of Magic and Experimental Science [New York, 1923], Vol. II).
 On the contemporary writings on inventors and discoverers cf. E. Zilsel, Die Entstehung des Geniebegriffes (Tubingen, 1926), pp. 130-143.
empiricism, very far from being science, was a matter of practical life and casual observation and, for the most part, was lacking in systematic method.
But gradually, especially in Italy, a more systematic technique of empirical research developed among certain groups of superior craftsmen whose professions required more knowledge than did their colleagues’. We are speaking of the artists, the makers of nautical and of musical instruments, and the surgeons. In the Middle Ages the painters, sculptors, and architects had not been distinguished from the whitewashers, stonedressers, and masons. With the decay of their guilds they slowly separated from handicraft and eventually, about the middle of the sixteenth century, became free artists. During the course of this evolution a remarkable group developed within their ranks. We may call them artist-engineers, for they did not only paint pictures and build cathedrals but also constructed lifting-engines, earthworks, canals and sluices, guns and fortresses, discovered new pigments, detected the geometrical laws of perspective, and invented new measuring-tools for engineering and gunnery. Many of them composed diaries and treatises on their experiences and inventions for the use of their colleagues. As they belonged neither to the Latin-speaking university-scholars nor to the humanistic literati, but were artisans, they wrote their papers in the vernacular. All of them were already accustomed to making experiments. They were the true forerunners of modem experimental science. 
Brunelleschi (1377-1446), the constructor of the cupola of the cathedral of Florence, was the first of these artist-engineers. Among his successors are the bronze founder Ghiberti (d. 1455), the architect and painter Lione Battista Alberti (1407-1472), who - as an exception - had had a classical education, the architect and military engineer Giorgio Martini (1425-1506), Leonardo da Vinci (1452-1519), and, finally, the goldsmith, sculptor, and gun constructor Cellini (1500-1571). The architect and gunnery expert Biringucci (d. 1538) may be mentioned because of his book on metallurgy, Della pirotechnia. His paper is the first treatise on chemistry based on sound experience and avoiding any alchemistic superstition. The inventors and constructors of the new clavicembali and harpsichords, and the compass-makers also belonged to the experimenting superior craftsmen. A third group was formed by the surgeons. They dissected animals and often human bodies, whereas the learned physicians seldom dared to do such untidy and sinister work. By their common interest in anatomy there were often connections between artists and surgeons.
Experiment requires manual work, and, therefore, in both antiquity and the modem era its use began in handicraft. In antiquity scientific experimentalists were extremely rare. Since rough work was generally done by slaves, contempt of manual labor formed an obstacle that only the boldest of ancient scholars dared to overcome. A similar obstacle, though by scarcity of slaves less unsurmountable, obstructed the rise of experimental science in the modem era.
 On the artist-engineers and their papers cf. Leonardo Olschki, Geschichte der neusprachlichenwissenschafllichen Literatur, I (Heidelberg, 1918), 45-447, and Zilsel, op. cit., pp. 144-157.
The educational system under early capitalism took over from the Middle Ages the distinction between liberal and mechanical arts. In the seven liberal arts (grammar, dialectic, rhetoric, arithmetic, geometry, astronomy, and music) thinking and disputing were alone required; on them alone was the education of well-bred men based. All other arts, as requiring manual work, were considered to be more or less plebeian. There are numerous instances to indicate that up to the sixteenth century even the greatest artists of the Renaissance had to fight against social prejudice. And by the same reason the two components of modern scientific method were kept apart: methodical training of intellect was reserved for university-scholars and humanistic writers who belonged, or at least addressed themselves, to the upper class; experiment, and to a certain extent observation, was left to lower-class manual workers. Even the great Leonardo, therefore, was not a true scientist. As he had never learned how to inquire systematically, his results form but a collection of isolated, though sometimes splendid, discoveries. In his diaries he several times discusses problems erroneously which he had solved correctly years before. Gradually, however, the technological revolution transformed society and thinking to such a degree that the social barrier between liberal and mechanical arts began to crumble, and the experimental techniques of the craftsmen were admitted to the ranks of the university-scholars. Rational training and manual work were united at last: experimental science was born. This was accomplished about 1600 with Galileo Galilei, Francis Bacon, and William Gilbert. One of the greatest events in the history of mankind had taken place.
The scientific work of Copernicus (d. 1543), being rational not experimental, may be omitted here.
Galileo (1564- 1642)  got his education at the University of Pisa and for more than twenty years he was professor at the universities of Pisa and Padua. His relations to handicraft and technology, however, are often underrated. During his student days there was no mathematical instruction at Pisa.  He learned mathematics privately, his tutor, Ostilio Ricci, being an architect and teacher at the Accademia del disegno which had been founded in 1562 by the painter Vasari as something between a modern academy of arts and a technical college.  Thus Galileo’s first mathematical education was directed by persons who were artist-engineers. As a young professor in Padua he lectured at the university in mathematics and astronomy and gave private instruction on engineering in his home. For his experimental studies he established working-rooms in his house and hired craftsmen as assistants.  This was the first university laboratory in history. His scientific research started with studies on pumps, the regulation of rivers, and the construction of fortresses. Ever since his student days he had liked to visit dockyards and arsenals. His first printed publication (1606) described a new measuring-tool for military
 On Galileo cf. Olschki, op. cit., III (Halle, 1927), pp. 117-467.
 Galileo, Opere (ed. nazionale), XIX, 32 fT.
 Before the foundation of the Academia del disegno, young artists had always received their education at the workshop, like all apprentices. The new school clearly shows how engineering gradually penetrated the province of academic instruction.
 Galileo, op. cit., XIX, 130 ff.
purposes. Even in his last work of 1638, which initiated modern mechanics, the setting of the dialogue is the arsenal of Venice. His greatest achievement, the discovery of the law of falling bodies, also originated in connection with the contemporary technology. Among the gunnery experts of the period there were many discussions on the shape of the trajectory. Galileo realized that the question could not be answered before the problem of falling bodies was solved. Free falling, however, was too fast to be measured exactly. In order to slow down the movement, Galileo took brass balls, made them roll down an inclined groove, and measured the spaces, times, and velocities. He succeeded in correlating his results by means of a mathematical formula and finally determined the shape of the curve of projection. In Galileo’s classical inquiry the two pillars on which modern science is based stand out: experiment and mathematical analysis. The experiments of the craftsmen, alone, would never have issued in science.
Francis Bacon (1561-1626) did not make any important discovery in the natural sciences. His writings abound with magical survivals and errors; he did not understand well the achievements of Copernicus, Galileo, Gilbert, and Harvey; the methodological prescriptions he gave in order to promote empirical research were too pedantic to be of great use to scientists. But he was the very first writer who fully realized the importance of science for human civilization. The scholastics and humanists, he explained, have only repeated sayings of the past. Only in the mechanical arts has knowledge been furthered since antiquity. Bacon, therefore, spoke of craftsmen and mariners with enthusiasm and proclaimed their works as models for the scholars. He is an enthusiastic advocate both of scientific induction and of the ideal of progress. Both ideas are closely connected: they are nothing else than the working method of early capitalist handicraft seen with the eyes of a philosopher. His whole philosophy is one great attack against the ideals of the seven liberal arts. Bacon himself performed numerous experiments: he died from a cold which he caught while stuffing a dead chicken with snow. Most of his learned contemporaries probably considered that experiment more fitting for a cook or flayer than for a former lord chancellor of England.
The first learned book of the modem era on experimental physics had appeared before Galileo’s and Bacon’s publications. It was William Gilbert’s De magnete (1600). Gilbert was physician to Queen Elizabeth I. His experimental method originated partly in contemporary metallurgy and mining, partly in the experiments of the retired mariner and compass-maker, Robert Norman. 
Why is experiment so essential to empirical science? Mere observation is a passive affair. It means but “wait and see” and often depends on chance. Experiment, on the other hand, is an active method of investigation. The experimenter does not wait until events begin, as it were, to speak for themselves; he systematically asks questions. Moreover, he uses artificial means of producing conditions such that clear answers are likely to be obtained. Such preparations are indispensable in most cases. Natural events are usually compounds of numerous effects produced by different causes, and these can hardly be separately investigated until most of them are eliminated by artificial means. There is, therefore, in all empirical sciences a distinct trend toward experimentation. Sciences in which experiment is not feasible are handicapped. They try to solve their problems by referring to other sciences in which
. Cf. “The Origins of William Gilbert’s Scientific Method” [this volume pp. 71-95, Eds.].
experiments can be performed. Thus meteorology, geology, seismology, and astrophysics make use of laboratory physics and laboratory chemistry. Sociologists and economists attempt to utilize results of psychology. A few modern geologists have even attempted to imitate formation of mountains on a diminished scale in laboratories. To a large extent the poor results of the social sciences might be explained by lack of experiments. The only substitute which, under favourable conditions, can to a certain degree compensate for the lack of laboratory experiments is the use of a great number of observations when carefully compared and worked up by means of statistical methods. Until now, however, experiment has been by far the most efficient empirical method.
It is noteworthy that one of the oldest empirical sciences, astronomy of the solar system, was highly successful without any experiment. This is due to the extraordinary fact that in our solar system superimposed effects belong to very different orders of magnitude and therefore can be separated comparatively easily. The solar system is exceptionally well isolated and the sun surpasses by far all planets in mass. Were the solar system continually bombarded by heavy meteorites or, what is the same, were it passing through a dense star-cluster, and were Jupiter’s mass of the same order of magnitude as that of the sun, Copernicus, Kepler, and Newton would not have achieved much.
The young experimental science was forced to fight hard battles with pre-scientific thinking. Primitive man does not distinguish exactly between animate and inanimate objects; he apprehends all natural events as if they were manifestations of striving, loving, and hating beings. This animistic conception of nature is predominant in all civilizations without money economy and dominated medieval thinking too. When a comet appeared in the sky or a monster was born, medieval man questioned rather the meaning, the aim, and the purpose of these events than their causes. The scholars did not think very differently either. Certainly “entelechies” and “substantial forms” of Aristotle and the Scholastics are not primitive ideas; they are complicated and highly rational constructs. Yet their animistic kernel has, as it were, only dried up; something like a soul, striving to reach its aims, still glimmers through the rational hull. The same holds of the “occult qualities” that were liked so well by the Scholastics. These could never be observed but were supposed to adhere to most objects and to produce effects by sympathy and antipathy, as if they were little ghosts. Animistic survivals like these were of no use to modern technology and had to be cleared away. Teleological explanations were gradually replaced by causal ones. Purposes of inanimate things, the meaning of natural events, and soul-like powers of physical objects cannot be ascertained by observations. On the other hand, the regular connection of cause and effect is testable by experience and experiment. Moreover, engineers are able to produce the effects they want if they know the cause. Causal explanation, therefore, became the chief aim of experimental science.
The discarding of teleological explanations may be illustrated by two well-known instances. The Scholastics, with Aristotle, explained the falling of bodies by the theory of natural places. Each body was supposed to have its correct place to which it moved when it had been brought to a wrong one. Obviously inanimate bodies were conceived as though they were cattle striving to the accustomed stable. As the theory of natural places did not give any information on the empirical details of falling, it was of no use to the artillery men of the modern era who wished to level their guns correctly. It had to be replaced by Galileo’s law of falling bodies and his calculation of the parabola of projection.
The working of suction pumps was explained in the late Middle Ages by the doctrine of horror vacui. Water was supposed to rise in pump barrels because nature had an antipathy to empty space. Since the well-diggers of the new era could not calculate from this theory how long they might make their pipes, two pupils of Galileo, Viviani and Toricelli, experimented on pipes filled with mercury and discovered and measured atmospheric pressure.
The causal mode of investigation gave rise to a basic, and previously unknown, conception. In a well-governed state there are laws which are prescribed by the government and, for the most part, observed by the citizens. Lawbreaking is rare and is punished when detected. Let us now suppose the government to be omnipotent and the police to be omniscient. In this case laws would always be observed. The seventeenth century began to compare nature with such a perfect state, ruled by an almighty and omniscient king.  Thus the recurrent associations of natural processes were named natural “laws” by the scientists who investigated them - especially if they had succeeded in expressing the regularities by mathematical formulas. The term “law” became so common that people soon began to forget that it originated in a metaphor; the idea arose that all events, without exception, were subject to natural laws. This deterministic conviction dominated philosophy and science from Descartes, Hobbes, and Spinoza, and from Galileo, Huyghens, and Newton almost up to the present time. It impelled all naturalists to look for more and more natural laws and thus proved to be extremely fruitful.
Nevertheless, we must distinguish the empirical and the metaphysical and theological components in these ideas. The statement that there are regular connections between certain events or qualities undoubtedly is an empirical one, for such connections are observable. The case is different with the assertion that events are connected not only as a matter of fact but that, moreover, some necessity subsists, forcing one event to follow the other. Necessity never is observable; it transcends the province of experience. Metaphysical and
 In fact, the metaphor is not older. In scholasticism the term “natural law” had an entirely different and merely juristical and ethical meaning. The Middle Ages could not produce the modern concept of natural law for the simple reason that the feudal state was governed not by statute but by unwritten and loose traditional law. As far as medieval princes issued regulations at all, they were for the most part privileges given to single noblemen, monasteries, and towns. They compare, consequently, rather with exceptions than with regularities in nature. Also with the ancient authors the metaphor of natural law appears but seldom and in a rather vague form. Antiquity, however, was familiar with the mythological idea of “fate”.
theological additions become the more marked when necessity is interpreted as the order of a personal deity or of impersonal nature. But even without drawing in necessity, difficulties emerge and experience is transcended when it is asserted that an observed regularity will always hold. Obviously no statement speaking of more than a finite number of objects and containing the term “all” can be completely verified by observation. All these unempirical components were ascertained and criticized by Hume. Before Hume no scientist and no philosopher conceived natural laws merely as empirically ascertained regularities, and even nowadays the idea of lawful necessity has not entirely vanished.
The logical difficulty just mentioned is repeated on a higher level in the thesis of general determinism. What exactly does it mean to say that regularities hold “everywhere” or, which is the same thing, that there is “no” fact that is not subject to some regularity? In mathematics, if any finite number of cases is given, an equation covering the given cases can always be found. Does general determinism maintain only this analytic proposition or does it maintain more? When the nineteenth-century determinists spoke of regularities and laws, they had in mind the equations of classical mechanics. Yet those equations have since then proved inadequate. Plenty of physical facts became known in the last decades which are covered neither by mechanical equations nor by equations of a similar type. Nevertheless, it is instructive to observe how well determinism turned out - up to the twentieth century at least. Even vague and dubious assertions can render good services to empirical research as a heuristic stimulus.
As we have indicated, determinism for the most part was conceived in a special form. Up to the late nineteenth century nature was interpreted as a gigantic but lawfully functioning mechanism. Since movements, pressing, pushing, and pulling are the only essential factors in mechanical devices, in nature, as well, all processes and qualities were reduced to such movements. All other qualities, though comprising the greater portion of everyday experience - as colors, sounds, and smells - were not regarded as “real” ones. They were interpreted as “illusions”, and no scientific explanation and no natural law was considered to be definitive until it was reduced to laws of mechanics.
The mechanical conception of nature began as early as Galileo. It appeared in almost all philosophers from Descartes to Kant and dominated physics from the beginning of the seventeenth to the end of the nineteenth century. As is generally known, mechanical theories of sound, heat, light, and electromagnetism were constructed in complete detail. Such theories were not completed in chemistry and remained only programs in the provinces of smell and taste.
How can the rise of this remarkable conception be explained? Man himself is, in a certain respect, a mechanical being. All actions by which he influences the world around him consist in movements, in pushing and pulling. From the days
when he learns as a baby to control his limbs, he regards that way of reacting as the only natural one. Unless he happens to be a physiologist, the more subtle chemical and electrical processes within his muscles, nerves, and bowels are unknown to him. Small wonder, therefore, that, when he looks at the external world, he takes for granted the movements, pushing, and pulling that he finds there. On the other hand, he feels that all processes differing from his own actions require further explanation and tries to reduce them to mechanical ones. An electric eel, gifted with intelligence, might behave in an analogous manner. Imagine an animal of this kind, not only defending itself and attacking by electric shocks but also attracting prey by electromagnetism, splitting and absorbing its food by electrolysis! If it were a physicist, such an animal would probably look for electromagnetic explanations everywhere. To summarize, then: the mechanical conception of nature is anthropomorphic and interprets natural processes after the pattern of human actions.
Unfortunately, more facts seem to be accounted for by this biological explanation than actually prove to be correct in history. Though men have always had a like biological organization, mechanistic theories are met with only very seldom and in very few civilizations. Mechanistic physics appeared only with ancient atomists and Epicureans and in the period we are just discussing. Obviously, some additional conditions are required if theories of this type are to develop. Apparently, the state of the contemporary technology can give the additional explanation we need. Man of the seventeenth century had outgrown primitive animism and begun to see the world with the eyes of an engineer. But since the only machines with which the period was acquainted were, without exception, mechanisms such as printing-presses, weaving-looms, pulleys, and clock-work, it was rather natural to think that the whole world was a mere mechanism as well. This prejudice was shaken only when technology had entirely changed, and nonmechanical engines had become more frequent than mechanical ones. This happened in the late nineteenth century.
The scientific value of the mechanist conception of nature cannot be overlooked. Distances and shapes, pressure and movement, can be measured comparatively easily: that is to say, can be co-ordinated to numbers without great difficulty. On the other hand, qualities such as blue and cold never, themselves, appear among the results of any calculation, and their measurement is considerably more complicated. So reduction of all qualities to mechanical ones enabled the physicists to co-ordinate numbers to qualities, and this again made it possible to grasp the physical world by mathematics. By the mathematization of physics, the building-up of deductive theories was immensely furthered. We need not point out how philosophical rationalism was stimulated by this development, as these achievements belong with an account of the rational side of knowledge. The heuristic value of the mechanist conception, however, was not slight. Everyone who knows the history of physics also knows how many new empirical facts were discovered and causally explained for the first time by means of mechanistic models. Both the rational and the empirical value of the mechanical conception of nature cannot be doubted. Yet, being based on a
prejudice, that conception had its dangers as well. Distinction between a “real” world of mechanics and an “illusory” one of qualities has produced a confusion of concepts that has interfered with the analysis of knowledge and philosophy for almost three centuries. This will be discussed in the next section.
After natural science had adopted the experimental method from the craftsmen, it developed rapidly during the seventeenth century. It was engaged in its special questions and, consequently, does not offer many problems to epistemological investigation. The case of contemporary philosophy was different. The most remarkable fact in empiricist philosophy of the classical periods might be said to be the tendency gradually to transfer its interest from objects to the subject. By this remarkable tendency it soon became entangled in pseudo-problems. Medieval animism and the Aristotelianism of the Scholastics were scarcely overcome when a new metaphysics developed, originating in the contrast of object and subject, of outer and inner world.
With Francis Bacon (1561-1626), interest in objects still prevailed; subjective components of knowledge are only mentioned in his doctrine of idols. Bacon is convinced that knowledge gives a rather accurate picture of nature if only we take care in avoiding a few errors and prejudices. These fallacies are induced partly by social, partly by individual, conditions by which the judgment of the observer is disturbed. They can, however, be eliminated if attention is called to them. As is generally known, such biases are called “idols” by Bacon and are discussed by him in an entirely empirical way without involving metaphysics. The beginnings of the new subject-object metaphysics appeared in empiricism with Hobbes (1588-1679).
Hobbes was a radical mechanist. Since in his opinion all processes consist in movement, he was faced with the task of explaining the origin of all other qualities that are perceived by us. He tried to master the difficulty by his terministic theory of sensation. In his epistemology sensations are distinguished from objective qualities. Sensations are not at all copies of the physical world but only correspond to physical qualities as symbols or terms do to their objects. Plato has already duplicated the world by opposing the realm of Ideas to the world of phenomena. Platonic Ideas, however, are extremely vague constructs, originating chiefly in certain ethical considerations and in the philosophy of mathematics; they were scarcely designed to help naturalists in interpretation of their scientific observations: the metaphysical background of the Platonic Ideas is obvious. Into natural science and empirical philosophy the two-worlds theory was introduced by Hobbes. Although his terministic theory of sensation contains valuable elements (which were not, however, developed until the time of the modern logic of relations), his distinction between the “real” world of movements and the subjective world of qualities has entangled the philosophy of
nature in pseudo-problems for more than two centuries. They are the more serious as they are inevitable implications of the mechanistic interpretation of phenomena. As long as the latter was considered the only possible philosophy of nature, the pseudo-problems connected with the contrast of external world and subjective awareness could scarcely be eliminated.
The decisive step that changed empiricist epistemology into an introspective psychological theory of knowledge was taken by Locke (1632-1704). As generally known, Locke’s theory of knowledge reduced all statements, both in everyday life and in science, either to sensation or to reflection. It belongs to the character of his philosophy that he spoke rather of ideas and sensations than of facts, observations, and statements. Modern logical empiricism restricts itself to analyzing the methods by which statements are tested and verified, whereas empiricism of the seventeenth century proceeded psychologically and investigated how ideas are obtained. Locke’s polemic against rationalism became a psychological analysis of abstract ideas, on the one hand, and an attack upon innate ideas, on the other. Since innate ideas played a rather prominent part in Descartes’s epistemology, this attack is of considerable historical importance in the further development of empiricism. But Locke soon turned to a discussion of the mind of the newborn child and revealed the concept of soul as the metaphysical background of all his psychological analyses. Mechanistic philosophy was considered so self-evident by Locke that its physical implications were not even pointed out by him; it appears only as his well-known distinction of primary and secondary qualities. The metaphysical background of his psychology of knowledge becomes most obvious in his exposition of the idea of substance. Substance to Locke was the “unknown support” of qualities. Although in his opinion that support is, and ever must stay, “unknown to us”, he did not hesitate to distinguish material and spiritual substances and to raise the problems of their interdependence. Locke, the empiricist, accepted the metaphysical dualism of Descartes without question and discussed an interdependence, never testable by experience, between constructs that could not be tested either. The outstanding scientific and historical importance of Locke’s analyses is generally known; but the metaphysical pseudo-problems, which are connected with them, must not be overlooked.
The contradiction between Locke’s empiricist principle, and his theory of substance was noticed by Berkeley (1685-1753). Berkeley adopted the principles and the introspective method from his predecessor but rejected Locke’s theory of substance. He started with a criticism of Locke’s theory of abstract ideas. Since he found by psychological introspection that he was not able to form, for instance, the idea of a color which is neither red nor blue nor green, he rejected Locke’s analysis and flatly denied the subsistence of abstract ideas. By leaving differences out of consideration and by marking common properties only, one word can obviously represent quite a group of different objects. Thus abstract ideas may and must be replaced by abstract words. Applying this analysis to the concept of substance, Berkeley concluded that substance is a mere word, void of content. The unknown “support” of qualities, which had been assumed by
Locke, although it could never be perceived, Berkeley thought could, and should, be eliminated: matter, being an abstract construct, does not exist at all. Actually there are perceivable qualities only, and these are nothing else than perceptions. Thus Berkeley’s well-known equation, esse = percipi, resulted.
The only point that strikes us in Berkeley’s remarkably consistent argumentation is that he failed to apply his analysis to the concept of mind or soul. Can souls ever be experienced? Are they not abstract supports of perceptions or qualities just as matter is? No doubt it was Bishop Berkeley’s religious attitude that prevented him from drawing this dangerous consequence. Avoiding the term “idea”, he always spoke of the “notion” of soul and would not admit that it is no less abstract than the idea of matter. Thus he built up his odd metaphysical system, constructing the world out of souls, ideas, and God. The tendency in empiricism both of turning to subjectivism and of getting involved in pseudo-problems has reached its peak in Berkeley.
Why did English empiricism tend to introspective psychology and how could it be invaded by subject-object metaphysics? Both in ancient and in medieval philosophy only slight beginnings of analogous ideas are to be found. The modern dualism of inner and outer world might be correlated with the dualism of soul and body and probably can be explained by the influence on mechanistic physics of theology. Belief in immortal souls and the mechanical conception of nature had never been united in one philosophical system before the seventeenth century. In antiquity the atomists and Epicureans were mechanists, but they did not believe in spiritual substances; Platonists and Stoics, on the other hand, distinguished souls and bodies but were not mechanists. Medieval theologians were not compelled to stress a chasm between spiritual and physical substances, since in their philosophy all physical objects were imbued with more or less soul-like powers; they could content themselves with Aristotelian entelechies which, being forms, were rather connected with than opposed to matter. Actual and radical dualism was introduced into philosophy by Descartes, mechanist and devout Catholic. It is comprehensible that, apart from direct Cartesian influence, similar tendencies invaded English empiricism as soon as it turned to mechanism. In their business of investigating phenomena, natural scientists are faced with the task of separating constant relations, on which all observers can agree, from the variable and unstable aspects which are offered under different conditions or to different observers in a different way. This is the sound basis for the distinction between objects and subjects. Bacon, for instance, did not misuse it. But this separation was misinterpreted as soon as the mechanistic philosophy of physical phenomena separated “real” qualities, such as motions, from merely “apparent” ones, such as colors. The rationalistic mechanist Descartes became guilty of the misinterpretation as did the empiristic mechanist Hobbes. Obviously, it was the rise of mechanistic physics that turned the harmless distinction between subjective and objective components of observation into a dualism of inner and outer world. And it is rather comprehensible that, under the influence of religious tradition, this dualism was more or less identified with the contrast of soul and matter. Thus analysis of experience turned with Locke to
introspective psychology - that is, to investigation of soul. Experience became a psychical duplicate of the “real” external world, and pseudo-questions arose: whether both halves of the world actually exist and how it happens that they correspond.
Locke’s and Berkeley’s principles were consistently applied to the concept of spiritual substance by Hume (1711-1776). As is generally known, he eliminated substantial souls and replaced them by heaps or bundles of impressions. Berkeley’s philosophy, which had denied the existence of physical objects but approved the existence of souls, seemed paradoxical to common sense. To critics entangled in the problems and language of the period, Hume’s position of rejecting spiritual substances as well as physical ones seemed, of course, even more paradoxical. Actually the paradox was rather diminished by him, for he treated objects and subjects in the same way and thus approached common-sense realism again. His analysis is a most important step toward overcoming the pseudo-problems of a subject-object metaphysics. In Hume’s time, however, theological ideas and the mechanistic interpretation of physical phenomena still blocked the way to a complete understanding and elimination of such pseudo-problems. Moreover, his analysis was impaired by his dealing with impressions and ideas rather than with statements. This predilection for elements of knowledge which are too minute to be very useful in analysis had originated with Locke and can be traced in the development of empiricism up to Mach and the nineteenth century. The same phenomenon occurs when Berkeley, for instance, as he often does, turns philosophical criticism to criticism of language. In this case he is discussing words rather than statements. This noteworthy but prejudicial predilection apparently also originated in the psychological attitude of classical empiricism, for statements are too logical to attract the attention of psychologists. Analysis of experience shifted from ideas to statements only when axiomatic methods in mathematics of the late nineteenth century had roused interest in statements and their concatenation, and when non-Euclidean geometry and modern symbolic logic began to influence empiricism. With rationalists and with Kant, on the other hand, statements always had played an important part.
Among the most important achievements in the history of philosophy is Hume’s analysis of the concept of causality. With surpassing clarity he showed that we speak of cause and effect if phenomena are connected regularly with one another. Since cause and effect are not linked by logical necessity, even the most common causal statements of everyday life depend entirely on past experiences and never can be inferred a priori. Induction, therefore, differs radically from deductive logic and is based psychologically on custom and belief. Even today the man in the Street is inclined to conceive cause as a thing that by its activity produces another thing. Hume’s criticism destroyed this idea of active production - that last remainder of primitive animism. Therewith it is implied that not things but processes, qualities, and relations alone are causes. And, by
stressing regularity of connection, he adapted the concept of cause to the concept of natural law, the very form in which it is used in any mature empirical science. His theory of causality would not have been possible in a period in which laws were not yet known, and it was obviously inspired by Newton’s and his followers’ successful investigation of physical laws. Moreover, it made feasible the future development of physics. Without Hume’s concept of cause, nonmechanistic physics of the late nineteenth and the twentieth centuries scarcely could have come into existence. Among later philosophers, however, his analysis has met with considerable criticism, opposition, and - even worse - complete neglect. As to induction, later philosophy has made little progress beyond Hume’s remarks.
During the eighteenth century, empiricism, by emphasizing sensations, turned in France to sensationalism and, with many philosophers, to materialism. On the other hand, empiricist philosophy of the seventeenth and eighteenth centuries developed important methods and results of psychological investigation. The laws of association were discussed again and again; contemporary reports on customs of primitive peoples were used by Locke in his polemics against innate ideas, and the anthropological method was successfully developed and applied to the investigation of religion by Hume; the psychology of people with defective senses was occasionally referred to in Berkeley’s Theory of Vision and was considerably furthered in Diderot’s Letters on the Blind (1749) and Letters on the Deaf-Mute (1751). For the first time perception of space was investigated psychologically, and the visible, tactual, and muscular components of the experience of space were more or less exactly distinguished in Berkeley’s Theory of Vision.
The last remarks have brought us to that field in the evolution of empiricism which is the most fertile one - the special sciences. Only the general expansion of the domain of empirical research can, however, be discussed here.
The empirical spirit of the modern era is entirely contrary to the spirit of medieval scholasticism and, consequently, was compelled at first to overcome the resistance of theological tradition. The rise of empiricism, therefore, is connected, though not identical, with the spreading of the Enlightenment. The beginnings of religious Enlightenment, that is, the ideas of natural religion (Herbert of Cherbury, 1624), were based rather on rational construction than on empirical investigation of the various systems. Little attention was paid by Herbert and his followers to the variety of modes of worship; priests and ceremonies were rather disliked by them. As is generally known, the main theses of natural religion stated the existence of God, immortality of the soul, and reward and punishment in the other world. They were primarily products of abstract reasoning and were based on empirical comparison only in so far as they
expressed the components common to the three great monotheistic religions - the only religions well known in this period. Obviously, the articles of natural religion indicate the common content of Judaism, Christianity, and Mohammedanism. Yet the articles were void of emotional content and do not occur, as they were formulated by Herbert and his followers, in any living religious system. The religious attitude of most of the empiricist thinkers of the seventeenth and eighteenth centuries was, however, more or less determined by these rather lean ideas.
It is remarkable that the first empirical contribution to scientific investigation of religion was made in the modern era by one of the most consistent rationalists - by Spinoza. In analyzing the worldly literature of antiquity, the humanists of the Renaissance had already created and developed the methods of philological criticism. Hobbes, in his Leviathan (1651), had advanced a few critical remarks on the composition of the Old Testament. Spinoza, however, was the first who, in his Tractatus theologico-politicus (1670), dared to apply, consistently, philological methods to the Old Testament. If painstaking philological criticism is considered an empirical achievement, Spinoza’s successful attempt to determine the time of composition of parts of the Old Testament may be reckoned with the major advances of empiricist thinking. The next step in this field was taken by Hume. Hume’s dissertation on the Natural History of Religion (1753) tried for the first time to give an empirical theory of general religious development and started comparative investigation of primitive religions. Hume, however, knew for the most part only the religious ideas of a few primitive nations, of ancient Greece and Rome, and, of course, the three great monotheistic systems. Actual knowledge of the great religious systems of India and China came to Europe not before the early nineteenth century. In the later nineteenth century the empirical science of religion rose considerably, using the philological and anthropological methods of Spinoza and Hume as well as modern psychological and sociological ones. Obviously, empirical thinking could not have spread in the field of religious and anthropological research were it not for the expansion of world-trade and the colonial system that made the white race acquainted with exotic civilizations. In Locke and Hume these connections had already become visible.
The natural sciences are more than two centuries older than the science of religion and comparative anthropology. In the field of astronomy and physics empirical research had reached its first overwhelming successes as early as the seventeenth century, in the period from Galileo to Newton. Chemistry joined the advance in the eighteenth century and later was followed by mineralogy, geology, and meteorology. From the methodological point of view success was obtained in the physical sciences by three means: they did not restrict themselves to mere observation but experimented wherever physical processes could be influenced by technological devices; they investigated the quantitative relations of phenomena; and they considered discovery of natural laws as the most
important goal of research. Application of mathematics to the empirical findings and the construction of deductive theories cannot be separated neatly from those empirical methods and scarcely are less important; but, as they belong with the rational side of science, they need not be discussed here.
The biological sciences developed considerably more slowly. Among them, medicine is the most ancient one. In the modern era medical investigation, as well as physical research, was hampered at first by the social prejudice against manual work. This prejudice was overcome in the late sixteenth century, and dissection came in use in anatomy about the same period as experiment did in physics. As early as the beginning of the seventeenth century Harvey experimented with animals in his embryological research and occasionally used quantitative methods in his investigation of the circulation of the blood. Nevertheless, the practical tasks of medicine affect the emotions of men so intensely that metaphysical, teleological, and even superstitious ideas and traditions were eliminated from medicine more slowly than from any other natural science. In zoology and botany quantitative and experimental methods did not come into general use before the nineteenth century. Up to that time, biologists restricted themselves chiefly to observation, non-causal description, and classification. Classification of the material must precede the investigation of causes and laws in all empirical sciences, as in business the stock of goods must be ordered and inventoried before actual trading starts. In the field of physics, as well, solid and liquid bodies, acoustical and optical phenomena, for instance, were distinguished before they were investigated scientifically. Whereas in physics, however, most classifications are elementary and, therefore, can be followed rather soon by causal research, the vaster variety of objects in the fields of zoology and botany is much harder to survey. Biological research, as a result, has, during almost three centuries, scarcely passed beyond classification and non-causal description. Moreover, it is far more difficult with plants and animals to produce alterations artificially than it is with physical and chemical objects. For all these reasons experimental methods and causal explanations were adopted by biology rather late. Instead of the concept of cause, prescientific teleological interpretation ruled the biological sciences until the nineteenth century.
The eighteenth century, however, was already aware of the difference between “artificial” and “natural” classification. Linnaeus, the most eminent biologist of the period, gave both artificial and natural classifications of plants and animals. Since plants are extremely numerous and varied, it is often not easy to determine the species to which some individual plant belongs. In order to facilitate that task, Linnaeus classified the plants by the number of their stamens; he was well aware, however, that this classification was merely artificial and hardly more than a technical device of nomenclature. His natural classifications, on the other hand, aimed at putting together plants or animals in one group which “actually” are connected with one another. Thus natural classifications includes the viewpoint of empirically ascertained relationships and, therefore, is the first step in the direction of causal research and the theory of evolution.
Lamarck’s, and especially Darwin’s, achievements opened entirely new ways to biological investigation. Empirical thinking was furthered by the theory of evolution in four respects. Linnaeus had still considered animal and plant species as absolutely rigid; there are as many species, he explained, as have initially been created by God. By the theory of evolution this rigid immutability was liquified, as it were: “to be” was replaced by “to become”. Certainly, the concept of temporal process is not necessarily involved in the concept of natural law. There are in physics non-temporal laws of coexistence as well as laws of temporal change, though the former might be less numerous. Moreover, the latter seem to interest men more intensely; human actions are processes themselves and always aim at influencing the future. At any rate, the precursor of the concept of law - the concept of cause - contains the element of temporal change as an essential component. For that reason interpretation of living beings as subjects of a permanent temporal process was indispensable if description and classification were to be supplied by causal explanation and later by investigation of natural laws. Second, Darwin’s theory of natural selection (1859) made the first successful, though yet incomplete, attempt to explain causally the obvious fact that organisms are well adapted to their environments. Since the decline of Aristotelian scholasticism, prescientific teleological traditions had been removed from the field of physics. With Darwin they suffered the first serious blow in biology. Third, Darwin’s exposition of the descent of man (1871) destroyed traditional anthropocentric philosophy and thus decidedly helped eliminate obstacles to the empirical investigation of mankind. In anthropology, sociology, psychology, and ethics this influence became conspicuous very soon, even if the animal ancestors of man and natural selection were not discussed at all. Darwin’s ideas throughout fitted in with the trend of modern empiricist philosophy. Already, Hobbes, Hume, and a few representatives of the French Enlightenment had, without even thinking of animal ancestors of man, consistently treated all human problems as natural, merely empirical ones. This interpretation was confirmed and enormously spread by Darwinism. And, finally, Darwin’s theory called scientific attention to a few characteristic traits which can be found in any evolution - even beyond the province of biology. As it was pointed out by Herbert Spencer, for instance, specialization, or family-tree-like ramification, appears in the historical development of occupations and sciences and in the intellectual development of the human individual as well as in the phylogenetic evolution of organisms. Thus, empirical investigation of society, civilization, and mind was also considerably furthered by Darwin’s ideas. On the other hand, however, the scientific value of the concept of evolution was sometimes overestimated. The merely descriptive statement of some “evolutionary” process in society or culture is only a preliminary step by which the solution of the scientific problem is prepared but not given. A final statement of causes and laws cannot be supplied simply by description of some one evolutionary process.
Darwin’s theory was based on the practice of cattle-breeders and on comparative observation. He did not perform experiments. This most successful and
most exact method of empirical research was applied to biological problems in the investigation of heredity and physiology in the late nineteenth century. Mendel discussed his experiments on heredity (1865) in statistical terms, whereas some physiologists of the nineteenth century transferred the ordinary methods and concepts of physics and chemistry to biology and adapted them to the new problems. Moreover, both modern genetics and modern physiology investigate quantitative relations and are seeking natural laws.
At the end of the nineteenth century teleology seemed to revive again in biology. In the final analysis Darwin’s causal explanation of the fitness and adaptation of organisms deals with a statistical phenomenon only. It discusses the survival of the fittest within large groups of plants or animals, but it is not interested in the problem as to how the processes which are characteristic of life are accomplished in individuals. Such processes as the development of the fecundated egg, regeneration of injured organs, and others, were interpreted teleologically by Driesch. Even the Aristotelian concept of entelechy was reintroduced by him - a concept which two centuries before had been eliminated from science after long and arduous intellectual conflicts. The strangest phenomenon in this resurrection is that it is connected with a considerable advance of experimental method. For the aforesaid processes were investigated by Driesch by means of experiments that proved to be most fruitful. They showed that processes in one part of an organism are often influenced not only by conditions in that part but also by other, if not all, parts of the organism. Driesch and the neovitalists assume the influence of all parts and speak, therefore, of the efficacy of “wholeness”; wholeness, however, is interpreted by them as a nonspatial, soul-like entelechy which aims at ends. This neovitalistic “explanation” is highly contestable. Non-spatial entelechies are in no way observable, and, consequently, the entelechies of a frog and a hydra cannot be distinguished. It is not clear, therefore, what help they can give in explaining, predicting, and controlling the rather different processes occurring in these different animals. Obviously, entelechies are additional metaphysical ingredients and do not contribute anything to a solution of the empirical problems. The phenomena of regeneration and formation are now actively investigated. They cannot yet be explained satisfactorily, though many partial successes have been achieved. Nevertheless, there is no reason to assume that, in this field, solution of the scientific problems will be achieved by other means than by experiments, causal explanation, and investigation of laws.
The beginnings of psychology in the modern era appear rather remote from empirical research. Descartes and Spinoza, when discussing human passions, gave theoretical constructions which were intended primarily to support the ethical ideals and theories of each philosopher. Yet it cannot be doubted that their “psychological” studies are based on a good deal of introspection into their own minds and comparative observation of fellow-men. The same methods were used more consistently by Locke, Berkeley, and Hume in their analysis of
knowledge. Most psychologists of the seventeenth and eighteenth centuries were highly impressed by the scientific success of Newtonian physics; they were, therefore, seeking psychological laws, and considered the laws of association to be analogous to, and as important as, the laws of mechanics. Such physical analogies become rather conspicuous in the psychological treatises of both the physician Hartley and the chemist Priestley. Morals, also, were again and again investigated psychologically. By application of psychological methods to the problems of ethics, certainly a major advance in empirical thinking was made. On the other hand, in the eighteenth century, psychology still dealt chiefly with knowledge and rather neglected emotions, volitions, and actions.
Psychology in the nineteenth century developed considerably, both as to its subjects and as to its methods. In the psychology of knowledge the constructs of the deductive sciences began to be investigated. It was even assumed that logic could be reduced to psychology. Quite a number of epistemologists were convinced that all problems of logic would be solved if the psychological origins of logical concepts and logical propositions were found out. By their critics their line of thought was called “psychologism”. Psychologism in the nineteenth-century epistemology distinctly mirrors the empirical tendencies of the period and even exaggerates them. It obviously committed the same error a chess player would make if he thought that knowledge of the historical and psychological origin of all chess rules could answer the question of which chess problems are solvable and which are not. On the other hand, psychologists with a leaning toward the natural sciences investigated sensations in close connection with psychological research. Thus experimental and quantitative methods were introduced into at least one branch of psychology.
With the decline of the Enlightenment, the irrational sides of the human mind also began to attract attention. In the early nineteenth century German romanticists were already highly interested in passion and emotion and, moreover, paid a great deal of attention to hypnotism, somnambulism, and insanity. Philosophers and psychologists belonging, as did Schelling and Schubert, to the Romantic current of thought, began to deal also with the irrational and abnormal components of mind. These were interpreted, both by writers and by theorists, metaphysically, and even more or less magical ideas frequently occur in the expositions. Again a new and fruitful field of research was discussed at first in prescientific terms, as often happens in the history of science. This initial stage, however, was overcome about the middle of the nineteenth century. Scientific psychology divested itself of magic and metaphysical ideas without resuming the eighteenth century’s overestimation of intellect. It turned to voluntarism and investigated emotions, appetites, and instincts empirically, frequently in close connection with biology, physiology, and animal psychology. Psychiatrical research arose, and the comparison of normal and abnormal mind contributed fruitful data to psychology. Investigation of neuroses, and of hysteria especially, added unconscious processes to the objects of physiological research.
The new conception that mind is composed of unconscious as well as conscious elements apparently is exposed to methodological objections.  Viewed empirically, mind seems to be equivalent to awareness. Unconscious mental processes, therefore, seem to be a contradiction in terms and to be metaphysical constructs which can never be tested by experience. Yet, these objections do not hold. Psychology of the nineteenth century is, for the most part, based either on introspection or on empathic interpretation of the behavior of fellow-men. In this behavior, however, speaking and intentional communications form but a small part; other, and unintentional, reactions are far more numerous. No psychologist, therefore, bases his scientific assertions only on the words by which his fellow-men describes their awareness but makes use of their physiological reactions - for instance their blushing, their actions, and their omissions of action - as well. Is there any reason to abstain from investigating processes which do produce such reactions, actions, and omissions, though they cannot be described by the individual in whom they happen? Since the individual in whom they occur is not aware of them, it might be objected that they ought to be called physiological processes. Actually, we may hope that in the future they will be explored by physiological means, and certainly physiological investigation would furnish more exact results than the psychological investigations of the present time. Nevertheless, the above procedure can be justified. When an individual acts in a certain way because of some past experience which is perfectly well remembered by him but is so disagreeable that he does not like to speak of it, and when another individual has “repressed” an extremely painful experience so that it has become entirely “unconscious”, both kinds of behavior can greatly resemble each other. The similarity becomes manifest in the fact that the observer can put himself psychologically in the place of both persons by means of empathy. The very possibility of empathy in both cases is the link by which conscious and unconscious processes are connected. This is the empirical reason why we are justified in using psychological terms and psychological methods in investigation of the unconscious. Physiological processes, on the other hand, work quite differently from unconscious ideas, wishes, and purposes and are not accessible to empathy.
We have made use of empathy as of an empirical symptom of the relationship between conscious and unconscious mental phenomena. It must be remarked, however, that it cannot be used as a definitive method of research. When we interpret the behavior of fellow-men by means of empathy, our interpretation mirrors experience which we happen to be familiar with. Empathy, therefore, is subject to major errors and always needs farther examination by means of more reliable methods. We may strongly feel that a man is angry, and yet all inferences based on our feeling may later prove to be wrong. When, on the other hand, his further behavior conforms with our expectations, then and only then our opinions of him are verified. Since empathy works much faster than careful scientific examinations, it supplies most of the psychological judgments in everyday life.
 We need not distinguish here between unconscious and subconscious processes.
Yet scientific predictions, based on empathy, may be relied on only in so far as they are confirmed by observable actions and reactions of the individuals concerned. The method of empathic interpretation may be used in scientific psychology as a preliminary heuristic tool. Certainly, it is fruitful if its results are tested later by observation of the perceivable behavior. But it is highly fallible, and the scientific content of all assertions obtained in this way consists solely in those components which can be confirmed by observation. The precariousness of empathy has led to the rise of behaviorism, which is discussed at another place in this Encyclopedia. At any rate, it is an empirical fact that man can experience empathy in certain cases and is unable or less able to experience it in other ones. The cases in which it can be experienced - that is, the conscious and the unconscious mental processes - obviously have certain empirical features in common.
To summarize: if and only if empathy is more or less feasible, may an unconscious process be named “psychological” and be reckoned among mental phenomena. Nevertheless, empathy is a symptom only. With it a certain type of functioning, that is, of causal connection, becomes manifest that, apparently, is common to conscious and unconscious mental phenomena. Or, to put in a different way: conscious and unconscious components of mind are subject to kindred laws; physiological processes, to entirely different ones. The unconscious elements of mind have been introduced into psychology in order to fill the gaps of its causal explanations and to complete the domain of validity of psychological laws. This method of completing the scientific domain is entirely legitimate and is used in the physical sciences as well. Astronomers, for example, do not hesitate to discuss multiple stars with partly bright and partly dark components. Psychology of unconscious mental phenomena is not less empirical than astronomy of invisible stars. Another point, which need not be discussed here, is that, in its present state, the method of exploring unconscious phenomena needs improvement. Psychology of the unconscious is young and fruitful; but it is as inexact as all young sciences have been initially. Greater exactness may be obtained in the future by experiments and by comparison with animal psychology. In this article, however, we are not concerned with empirical details and special questions. It has only been necessary to discuss the basic question of empirical testability.
Among empirical sciences, the social sciences are youngest. They have developed gradually from at least three different sources. Jurists, political writers, and philosophers of the seventeenth century dealt with public law and political philosophy. They were, however, more concerned with rationally establishing their political aims and theories than with careful and unbiased observation of empirical facts. Hobbes, for example, was more a rationalist than an empiricist in his theory of society. He was among the most consistent representatives of the doctrine of social contract, which dominated political philosophy until the end of the eighteenth century; and this doctrine’s disregard of experience is obvious today. Primitive man subscribes to rational agreements just as little as he believes in the rational articles of “natural religion”. With the
advance in overseas trade, savage nations gradually became better known in eighteenth-century Europe, and an increasing number of authors began to criticize, with empirical arguments, the hypothesis of a social contract.
About the same period, French writers, among them Voltaire, Raynal, and Condorcet, started investigating the history of human civilization and the development of social institutions. In political historiography, also, the advance of empirical thinking and the gradual disappearance of theological ideas might be traced. Yet, the political historians still dealt with single events and individuals and hardly belong among the predecessors of the sociologists. On the other hand, the writers who cultivated “philosophical” history, as it was called in the period of the Enlightenment, helped to prepare the way for the social sciences. They discussed general processes in the development of civilization and the interdependence of these processes. The ideas of such writers were mixed with unexamined assumptions concerning human progress and contained a good deal of wishful thinking; but, in the last analysis, their philosophy was based on empirical observation and comparison: the “philosophical” components of their expositions were but vague - and often incorrect- formulations of sociological causes and laws. Reports on exotic peoples were also published more frequently in the eighteenth century; they greatly influenced the writings of the “philosophical” historians and, therefore, may be counted among them.
A third source of the social sciences springs from the practical needs of economy. In contrast to more primitive forms of economy, capitalism requires rational regulation of economic activities. This holds true of private as well as of political economy. Since the beginning of the modern era, the princes - or rather their secretaries and counsellors - were compelled to form opinions on the question of how taxes and duties influence commerce and manufacture and are, in turn, influenced by the prosperity of the latter; their revenues depended on these questions. When, in the middle of the eighteenth century in France, problems of agriculture also became urgent, public officials and private writers began to study economies more systematically. And, finally, Adam Smith, the friend of Hume, published his comprehensive theory of economic processes (1776). He thus created the science of political economy and started a scientific development which has continued up to the present day. Political economy is the oldest among the social sciences. As it originated in the needs of an eminently rational and antitraditionalist form of economy, it was from its very beginnings based on reasoning and experience. It never went through an animistic stage and was spared controversy with Aristotelian entelechies.  Control and prediction of economic processes were among its aims from the beginning. Consequently, theoretical economists immediately began looking for causal explanations and very soon for economic laws. They obtained, and still obtain, the empirical material for their theories by comparative observation; in the nineteenth century
11 In the twentieth century, however, a few German economists attempted to introduce entelechies in political economy.
statistical methods were added and proved to be highly successful. Experiments have not yet been performed in the field of economics. The hazardous steps, in politics and business life, that are often called economic experiments are done neither for scientific purposes nor with the necessary scientific precautions.
Rational deduction and mathematics do play a large part in certain economic theories. Political economy might be the only empirical science in which, for the reasons mentioned above, the empirical elements are likely to be impaired rather by excess of rational deduction than by prescientific tradition. It is still a rather imperfect science. It is about two centuries younger than physics, and its subject matter is more complex; economic experiments are not performed, and economic research is exposed to more selfish interests, political pressure, and wishful thinking than is the case in any other science. It is comprehensible, therefore, that in political economy scientific agreement could be reached only on comparatively unimportant questions; in fact, there are separate schools which do not even recognize each other. Some of them cling to experience; the results of their inquiries are collections of material rather than theories in which facts are causally explained. Others deal with nothing but laws of economy; they investigate them by means of rational analysis of a few basic concepts and construct large deductive systems based upon scanty observations. And yet, after all, political economy might be considered the most advanced among the social sciences.
Sociology originated in the beginning of the nineteenth century. Two rather different thinkers, Hegel (1770-1831) and Comte (1798-1857), may be called the first sociologists. Hegel comes from German Romanticism. His Philosophy of History employs the dialectical method, which is claimed to be rational and forms the backbone of his whole philosophical system. This method can be traced back to Plato. The metaphysical mill of thesis, antithesis, and synthesis, however, is as empty as inexact; it could yield almost any result. In Hegel’s system it begins its work with the concept of being. It cannot be doubted that the method’s rich output has, unconsciously, been supplied by and adapted to experience, for it would be sheer magic if reasoning alone were able to derive an abundance of concrete details from a concept which is devoid of content. However, while Hegel’s exposition of philosophy of nature is rather sterile and sometimes strangely contradicts later experience, his philosophy of history, law, and religion abounds with fruitful results. Apparently, the artificial mechanism of dialectic fits, approximately, certain processes which often occur in history, society, and civilization. The exact and empirical description of those processes, and the demarcation of provinces in which they occur and in which they do not, has not yet been accomplished. Hegel discussed the general lines of development in law and ethics, in morals and political institutions, in art, religion, and philosophy. He did not describe isolated events but comprehended numerous single facts in one construct and saw relationships which no one before him had noticed. These relationships always refer to temporal succession of cultural phenomena. As they are repeated in an analogous way in various fields, they may be taken as preliminary intimations of laws of sociological succession.
Regularity, however, was always interpreted as necessity by Hegel. Fruitful empirical knowledge often appears at first in vague and metaphysical expressions; Hegel’s philosophy of history is possibly the most impressive case in point.
Comte, the disciple of Saint-Simon, stemmed from the philosophical historians of the French Enlightenment. His attitude was entirely empirical and antimetaphysical. He was well trained in the natural sciences and regarded prediction as the main goal of knowledge. For him, to know meant to foresee, and science, in his system, is identified with the investigation of laws. Like Hegel, he investigated only the main lines of historical development and comprehended numerous single facts in very general statements. Utilizing the ideas of predecessors, he affirmed the sequence of a theoretical, a metaphysical, and a positive stage in the history of civilization. This succession is supposed to appear in different fields in the same way. His theory of civilization, therefore, may be taken as a preliminary formulation of a sociological law of succession. Comte coined the names of both sociology and positivism. Compared with Hegel, his expositions contain less speculation, but their empirical results are probably poorer.
It is not necessary at this point to survey the development of sociology after Comte. And even achievements so important as the introduction of the concept of evolution by Spencer and the emphasis on economic facts and economic groups by Marx need not be discussed here. But a few methodological aspects should be pointed out. Among sociologists there are today various schools and many controversies; some schools even disregard the investigations of most of the others. It might be generally agreed that sociology is an empirical science. It is based on observation and comparison, if not yet on experiment. As it does not deal with individuals but investigates groups and mass phenomena, the general sociological statements which appear in still rather uncritical forms in Hegel and Comte must be based on careful and complete collection of material if reliable results are to be achieved. It is here that statistical and sometimes even quantitative methods were successfully introduced in sociology. They were, however, largely applied to quite elementary phenomena and their use frequently resulted in mere collections of material. Causal and comprehensive sociological theories, based on statistics, are still lacking. Apart from the difficulty of putting them into practice, there is no reason why statistical methods might not be employed in the investigation of Hegel’s and Comte’s problems as well, that is, in investigation of cultural processes. Sociological processes are usually interpreted psychologically, that is, by empathy. It has already been pointed out that empathy proves to be fruitful as a heuristic method but that its results must always be tested by observable facts.
As to sociological laws, modern sociology does not restrict itself to the investigation of laws of succession but seeks laws of coexistence as well. There are, however, sociological schools which deny the possibility of any sociological laws. These maintain that in social research causes and laws have to be replaced by “types”, by “understanding” - that is, empathy, by “wholeness”, entelechies,
and values. The preliminary methods of non-causal description and classification and, moreover, all pre-scientific and teleological concepts of the past are revived as ultimate goals of science in this rebellion against causality. No convincing reasons, however, have yet been given proving that causes and laws must be discarded. In everyday life we are wont to predict, more or less successfully, such social phenomena as overcrowdedness of railroad trains, the outcome of elections, and the trend of public opinion. Predictions like these presuppose that social phenomena are connected more or less regularly. It may be that in any society regularities practically always have exceptions and, consequently, hold only approximately. Social groups are seldom isolated and usually interact with one another; the number of their members is always comparatively small; the members are very different, and some of them exert disproportionate influence. These conditions do not favor group laws. With a gas inclosed in a vessel with permeable walls and consisting of only a million molecules, a few of them being extremely large, rather inexact gas laws could be ascertained. It is possible that in sociology, also, only very inexact regularities can be discovered. Yet, no physicist or astronomer would entirely disregard a regularity on the ground that it did not always hold.
One more point must be considered. The classical gas laws deal with the interdependence of temperature, pressure, and volume of the gas. Nevertheless, the single molecules of which the gas consists have neither temperature nor pressure; they whirl at random, have kinetic energy and impulse, and are, in classical mechanics, subject to laws entirely different from gas laws. In an analogous way sociological laws might connect quite different variables than psychological laws do, though social groups consist of human individuals. Thus we have come back once more to the question of empathy. If we look for social regularities by means of empathy, we may never find them, since ideas, wishes, and actions might not appear in them at all. That is, social regularities may belong to a type of connection entirely different from psychological ones. At any rate, if there are sociological laws, they can be discovered only by actually looking for them and not by discussing their possibility. And what methods are to be employed in this investigation? The empirical methods of causal research have, in all sciences, proved to be so fruitful that we shall not rashly give up hope of finding them successful in the field of sociology too.
We have surveyed the advance of empirical thinking in the domain of the special sciences. We have found that, though different means of inquiry were adapted to the various subjects of the various sciences, in the final analysis success was achieved by the same methods everywhere. These methods are: collection of the material, observation, and comparison; experiment wherever objects can be influenced by technological means; counting and measuring, if possible; causal investigation and investigation of laws. As to methods, there are no basic differences among empirical sciences. The empirical methods are best developed in physics, and physical patterns, consequently, have influenced the other sciences in an increasing degree, despite the remarkable countercurrents in sociology and biology of the last decades. As to methodological maturity and
scientific success, biology is today next to physics. Sociology, on the other hand, is the least mature among the empirical sciences. Whether psychology or political economy is better developed is an open question. When such questions are to be decided, the deductive theories in the various sciences and the application of mathematics to their problems must be taken into account as well; the rational methods, however, have not been discussed here. As to historical sequence, the physical sciences are the oldest. It is noteworthy that causal psychology, which originated in the middle of the eighteenth century, and even causal political economy, which originated in the late eighteenth century, are considerably older than causal biology, which originated with Darwin and the physiology of the late nineteenth century. Yet biology has since then outstripped by far its older competitors. The historical sequence and historical development of the sciences do not exactly conform to the various degrees of complexity of their problems. They are greatly influenced by economic needs and the great struggle of ideas and can be explained only in connection with the general process of history.
In the second half of the nineteenth century important physical discoveries resulted in the breakdown of the mechanical theories of light, electricity, and magnetism. As philosophy, since the period of Galileo, had been influenced by physics to a higher degree than by any other empirical science, this physical revolution also reshaped philosophical thinking and the analysis of knowledge.
The process began with Maxwell’s inquiries on electricity (1865). Using the experiments and ideas of Faraday (1791-1867), Maxwell succeeded in comprehending all laws of the spread of electromagnetic actions in two fundamental equations. Faraday, on the other hand, had interpreted electromagnetic processes mechanically and pictured electric and magnetic lines of forces as invisible ropes, acting in accordance with mechanical laws. Maxwell also obtained his equations by means of a mechanical (namely, a hydrodynamic) analogy, but the mechanical model turned out to be extremely complicated this time. Maxwell was forced to imagine a model consisting of rotating, invisible ether whirls and interspersed invisible particles pushed on by the rotations. In addition, his equations connected electrodynamics with optics, since they contained the velocity of light as an essential constant and covered all laws of propagation of light-waves as well. When H. Hertz twenty years later (1888), by his famous experiments, ascertained, for the first time, the existence of electromagnetic waves, the wave theory of light definitely became a part of Maxwell’s theory of electromagnetism. Still, electromagnetic waves were supposed to spread as a kind of movement in the ether. The ether, however, had on the one hand, to penetrate everything and, on the other, to behave mechanically as a solid body, if observable optical phenomena were to be represented correctly. Thus,
mechanical models gradually came to be regarded as tools with which science might possibly dispense. Why should one derive both electromagnetic and optical from mechanical laws by means of highly complex models if all laws of optics can be derived directly from simply constructed electromagnetic equations? Possibly mechanisms do not have any preferred place over other physical phenomena. If equations, wherever they are derived from, present all the observable facts in the simplest way possible, they may very well fulfil the task of science better than theories attempting to reveal a “real” world behind the phenomena.
At first, ideas like these ventured forth hesitatingly, and their implications were either not realized or met with considerable resistance from physicists. Even before the experiments of Hertz, Kirchhoff, the founder of modern astrophysics, advocated, in his Lectures on Mathematical Physics (1874), the opinion that physics had to “describe” rather than explain phenomena. Similar ideas were supported by Helmholtz four years later (Tatsachen der Wahrnehmung, 1879). The physical and epistemological implications of such ideas were developed, with greater radicalism, by the philosophical physicist Ernst Mach (Mechanik in ihrer Entwicklung, 1883). Mach analyzed and consistently refuted the belief in the priority of mechanics to the other branches of physics. In his conception of science, scientific explanation is equivalent to “economical description” of the observed facts; science has to represent, as he pointed out, as many facts as possible by as few concepts as possible, and it is irrelevant whether the concepts used are taken from mechanics or elsewhere. Moreover, Mach disclosed the philosophical pseudo-problems originating from the mechanical conception of nature. The opposition of an objective world of quantities and a subjective world of qualities, which had confused philosophy for three centuries, has been overcome in his analysis of scientific knowledge. Starting from different problems, the historian of physics, Pierre Duhem, and the chemist Wilhelm Ostwald helped to destroy the mechanical prejudice. Duhem (La Théorie physique, son objet, sa structure, 1906) pointed to the different styles of thinking favored by the English and the French, respectively, in their physical theories; he ascertained that the theories of the French, representing the facts solely by means of mathematical equations, were not less efficient than the English theories based on mechanical models. On the other hand, general energetics, advocated by Ostwald, was also suited to weaken the overestimation of mechanics; for energetics restricted itself to discussing transformations of energy in all fields equally and disregarded mechanical models.
Mechanism suffered its decisive defeat, however, in the theory of relativity and modern atomic physics. Influenced by Mach’s ideas, Einstein published his fundamental papers on the special theory of relativity in 1905 and on the general theory in 1916. As is generally known, the theory of relativity pointed out the connection between certain paradoxes of light-propagation and of all spatial and temporal measurements and gravitation. Those highly general connections could no longer be represented by mechanical actions of one ether. A mechanical model of all relativistic laws has not yet been mathematically constructed in
every detail. If, after the theory of relativity, a physicist still intended to be a mechanist, he would have to take refuge in a “great-” if not a “great-great-ether” behind the ether, in order to represent all relativistic facts. Evidently overcomplicated and patched up, a mechanism like this could not stand competition with the mathematical conciseness of the relativistic equations. And, finally, in 1913 Niels Bohr’s paper appeared, which succeeded in representing the hydrogen atom by a planetary system of particles of electricity. From this day modern atomic physics has, with increasing success, derived all properties of matter itself - among them pushing and pulling - from actions of particles of electricity. These actions follow entirely non-mechanical laws. It can be explained only by history why modern microphysics has been named quantum mechanics. As is generally known, Bohr’s original model of an electrical planetary system has since been given up and only equations are left. In Heisenberg’s, Dirac’s, and Schroedinger’s quantum mechanics there is, for the most part, no question of even spatial movements of particles, let alone their pushing and pulling. Thus we have come to realize that it amounts to the same thing whether attraction of electric charges is explained by the pulling of invisible cords or the pulling of cords by the behavior of electrons and protons. We accept that theory which covers the observable phenomena and which, at the same time, is the more comprehensive, the more consistent, and the simpler. The mechanical prejudice has definitely been overcome. It needs not be mentioned that the breakdown of the mechanistic conception does not mean at all a return to premechanical animistic or teleological ideas. The statistical laws of quantum mechanics are as rational and mathematical as the laws of classical mechanics and not a bit nearer to the behavior of living beings or spirit. 
The breakdown of mechanistic physics took place during a period of complete revolution in technology. The levers and pulleys of the seventeenth century had receded to the background long ago. The steam engine and, in the second half of the nineteenth century, the electromotor, the dynamo, and the internal-combustion engine had taken their place. The working of all of these is based on non-mechanical processes. Merely mechanical machines, as bicycles, typewriters, and foot-driven sewing machines, have been comparatively unimportant in the economy of the last sixty years. The textile industry, in which mechanisms played a comparatively greater part, had been the leading industry up to the early nineteenth century but was fifty years later surpassed by far in economic importance by the electrical and then by the chemical industry. Men, and even the children, of the twentieth century do not feel at all strange about the working of an electric bulb, a telephone, a photographic camera, a radio set; what is even more important, they do not feel any different toward the working of these implements than toward the working of a typewriter. The movements by which man influences the world around him have, with the technology of the twentieth century, shriveled down in many cases to turning on a switch; the elec-
 Cf. Philipp Frank, Interpretations and Misinterpretations of Modern Physics.
tromagnetic, chemical, and caloric processes which are then initiated do the rest. If man is inclined to conceive natural processes after the pattern of how he himself influences nature, it is scarcely a surprise that the priority of mechanics has completely dwindled. Certainly, modern technology and economy have not only influenced modern physical theories but have been influenced by them as well. The electrical engineer, Marconi, did succeed the theorists Hertz, Maxwell, and Mach. On the other hand, however, Mach succeeded the steam-driven factories, the dynamo, and the electromotor, and, certainly, electro-engineering and radiobroadcasting helped immensely to make the nonmechanical conception of nature less paradoxical.
The breakdown of mechanistic physics could not fail to give a new impetus to empirical thinking. With the failure of mechanistic physics, the assumption of a second world behind experience had lost its scientific support. Now the subject-object metaphysics, the pride of all philosophers, who looked down on the naïve laymen, was badly shaken; its problems began to appear as pseudo-problems. Since causes and laws were employed in the new physics as functional connections and mere regularities, the unempirical components of those concepts, already criticized philosophically by Hume, became suspect for scientists as well. All these implications were consistently developed by Mach. On the other hand, physical hypotheses and models had suddenly turned out to be unsuitable, though having proved fruitful for three centuries. Necessarily, general methodological questions arose as a result of that fact, and, for the first time in the history of modern physics, the whole internal construction of science became problematic. What part is played by simplicity in scientific theories? Which natural laws should one consider as fundamental, and which as derived, when constructing a theory? What service can be rendered to science by working hypotheses, by fictions, and by conventions? Most of these problems deal rather with the deductive side of theoretical knowledge than with its empirical components. They were raised by Mach, by fictionalism and conventionalism of the late nineteenth century, and were more or less suggested by the physical revolution. Poincaré’s conventionalism, however, was influenced by modern mathematics as well as by the new physics. In the early twentieth century those mathematical and logical influences increased, united with the empiricist tradition, and resulted finally in logical empiricism - a subject which must be reserved for later treatment.