If a historian of science were to trace backward the scientific knowledge of some selected group of related phenomena, he would find a pattern scarcely different from that exemplified here in connection with the history of physical optics. Today’s physics textbooks teach students that light is photons—that is, quantum-mechanical entities displaying both wave and particle characteristics. Research proceeds accordingly, or else according to the more refined and mathematical characterization to which this customary verbal expression leads. But it has been only about half a century since such characteristics of light were defined. Until Planck, Einstein, and others advanced them at the beginning of the twentieth century, physics textbooks taught that light was transverse wave motion, an idea based on the paradigm derived from the optical writings of Young and Fresnel in the early nineteenth century. Yet the wave theory was not the first doctrine to come to be accepted by almost all scientists in optics. During the eighteenth century, the paradigm in this field was provided by Newton’s optics, which taught that light was material corpuscles. At that time, physicists tried to find evidence for the pressure exerted by particles of light striking solid bodies—evidence that the early wave theorists had not sought.1)
These shifts of paradigm in physical optics are scientific revolutions, and the successive transition from one paradigm to another through revolution is the usual developmental pattern in a mature science. But the pattern characteristic of the period before Newton’s work is not that; it is the contrast with which we are concerned here.
From remote antiquity to the end of the seventeenth century, no single, widely accepted view of the nature of light ever appeared. Instead, there were many competing schools and sub-schools, most of them espousing one or another variant of Epicurean, Aristotelian, or Platonic theory. One group held that light was a modification of the medium situated between the object and the eye. Another group explained light through the interaction of an emanation from the eye with the medium. There were, besides, theories of every sort of combination and modification. Each of these schools gained strength by relating itself to a particular metaphysics, and each emphasized, as paradigmatic observations, the special part of optical phenomena that its own theory could best explain. Other observations were handled by special treatment or left as important problems for further research.2)
In various periods, all these schools made considerable contributions across the whole range of concepts, phenomena, and techniques, and from them Newton first drew the paradigm of physical optics that was almost universally accepted.
A definition given by scientists other than the more creative scholars of these various schools would likewise exclude their modern successors. They were scientists. Yet anyone who has examined an outline of pre-Newtonian physical optics is apt to conclude that, although the practitioners in the field were scientists, the total result of their activities was something less than science. Because there were no shared beliefs that could be taken for granted, every author on physical optics felt that he had to establish his field anew from its foundations. In doing so, he was relatively free in choosing observations and experiments, because there was no standard of methods or phenomena that every writer in optics felt obliged to adopt and explain. In such a situation, the dialogue in the resulting books was usually directed as much toward scholars of other schools as toward nature. This pattern, which is by no means unfamiliar even today in many creative fields, is not inconsistent with significant discovery and invention. Nevertheless, it is not the type of development that physical optics acquired after Newton, and with which the other natural sciences are familiar today.
The history of electrical research in the first half of the eighteenth century provides a more certain and better-known illustration of the way a science develops before it acquires its first universally acknowledged paradigm. During this period, there were as many different views on the nature of electricity as there were leading electrical experimentalists—Hauksbee, Gray, Desaguliers, Du Fay, Nollet, Watson, Franklin, and others. All their numerous concepts of electricity had something in common. Those concepts were derived in part from one or another modification of the mechanico-corpuscular philosophy that guided all scientific research of the time.
Moreover, all those concepts were components of actual scientific theories—theories partly derived from experiment and observation and partly determining the choice and explanation of the additional problems undertaken in research. But although all the experiments were concerned with electricity, and although most of the experimenters read one another’s research papers, their theories had no more than a family resemblance.3) Among the theories that followed seventeenth-century scientific activity, one early group regarded attraction and the generation of electricity by friction as the fundamental electrical phenomena. This group tended to treat electrical repulsion as a secondary effect due to some mechanical rebound, and also tried to postpone as long as possible both the discussion and the systematic study of Gray’s newly discovered effect, electrical conduction. Other “electricians” (the term they themselves used) regarded attraction and repulsion as equally fundamental actions of electricity and modified their theories and research accordingly. (In fact, this group was extremely small—even Franklin’s theory could not properly explain the mutual repulsion between two negatively charged bodies.) But except for the simplest effects of conduction, they, like the first group, struggled to give a proper simultaneous account of anything else. These effects, however, in turn provided a starting point for a third group, which sought to explain electricity as a fluid capable of flowing through conductors rather than as an “effluvium” bursting out from nonconductors. Thus this group had difficulty reconciling its theory with the various effects of attraction or repulsion. Only through the work of Franklin and his immediate successors did a theory finally emerge that could explain all these effects of electricity with roughly equal plausibility, thereby providing—and in fact did provide—the generation of “electricians” who followed with a shared paradigm for their research.
Except for fields in which the first firm paradigms go back to prehistory, as in mathematics and astronomy, and except for fields formed by the division and recombination of already mature specialties, as in biochemistry, the situations outlined above may be said to be historically typical. Although I am here repeating the inappropriate simplification of representing a broad historical event by a single, rather arbitrarily chosen name (such as Newton or Franklin), I wish to say that a similar fundamental disagreement of opinion was characteristic, for example, in the study of motion before Aristotle, the study of statics before Archimedes, the study of heat before Black, the study of chemistry before Boyle and Boerhaave, and the study of historical geology before Hutton. In some areas of biology—for example, in the study of heredity—the first universally accepted orthodox paradigms appeared quite recently.
And the question of what parts of the social sciences have acquired such paradigms, and to what extent, remains unresolved even now. History suggests that the road to firm and unwavering research consensus is extremely arduous and long.
But history also provides some reasons for the difficulties encountered on that road.
In the absence of a paradigm, or of some candidate for paradigm, all the facts that could be relevant to the development of a given science are likely to appear equally related. Thus early fact-gathering becomes an activity almost random beyond comparison with the activities familiar in later scientific development. Moreover, since there is no reason to pursue any particular form of more profound information, early fact-gathering usually results only in the accumulation of heaps of easily obtainable data. The collection of facts thus obtained includes facts readily acquired through ordinary observation and experiment, together with data of somewhat deeper significance obtainable from established crafts such as medicine, calendar making, and metallurgy. Because crafts provide a ready source of access to facts that could not be discovered merely by chance, technology has often played a decisive role in the birth of new sciences.
Yet although this kind of fact-gathering has been essential to the origins of many significant sciences, anyone who examines in detail, for example, Pliny’s encyclopedic writings or the Baconian natural histories of the seventeenth century will discover that fact-gathering brings about difficulties. Bacon’s “histories” of heat, color, wind, mining, and the like are in part filled with profound information. But they set side by side facts that would later prove real (for example, heating by mixing) and facts so complex that for some time they would remain wholly unreconciled with theory (for example, the warming of a dung heap).4) Furthermore, because any explanation cannot help being fragmentary, a typical natural history includes descriptions that later scientists would omit as merely circumstantial. None of the early “histories” of electricity, for instance, so much as mentioned that chaff attracted by a rubbed glass rod was then repelled. Such an action seemed mechanical, not electrical.5) Moreover, since the ordinary fact-gatherer has rarely possessed the tools or experience to be critical, natural histories have often set descriptions like this beside other facts, such as heating by antiperistasis (or cooling), which can no longer be confirmed.6) Only very rarely, as in ancient statics, dynamics, and geometrical optics, do facts collected with little guidance from established prior theory prove sufficiently clear to permit the birth of a first paradigm.
This is the situation that produced the various schools characteristic of the early stages in the history of scientific development. No natural history can be interpreted without at least some implicit body of intertwined theoretical and methodological beliefs that permits selection, evaluation, and criticism. If that body of beliefs is not already implicit in the collection of facts—in something more than “mere facts”6)—then it must be supplied externally, perhaps by the metaphysical perspectives of the time, by another discipline, or by personal and historical circumstances. In that case, it is not at all strange that, in the early development of any science, people confronting the same range of phenomena—though generally not the same particular phenomena—describe and interpret them in different ways. What is truly surprising, and perhaps characteristic of the fields we call sciences, is that such initial divergence gradually disappears.
Thus these differences disappear to a considerable extent, and in the end they plainly vanish altogether.
Moreover, it stems from the victory of one among the pre-paradigm schools, which, because of such preconceptions, emphasized only some particular part of an exceedingly vast and incomplete mass of information. Those electricians who regarded electricity as a fluid, and who therefore paid special attention to phenomena of conduction, provide an excellent example from this point of view. They provide the well-known example of actions of attraction and repulsion. Led on by this belief, which found it difficult to resolve the familiar plurality of attractive and repulsive actions, some among them even came to devise a way of storing the electric flow in a bottle.
The immediate fruit of their attempt was the Leyden jar, which would never have been discovered by someone staring stupidly at nature or exploring it at random, but which, in the early 1740s, came to be developed independently by at least two investigators.7) From almost the very beginning of his work on electricity, Franklin devoted particular attention to explaining this peculiar and, in the end, especially illuminating device. His success in doing so, though it could not explain all the already known cases of electrical repulsion, was able to provide the foundation for the most decisive argument that raised his theory to the status of a paradigm.8) To be accepted as a paradigm, a theory must certainly seem better than its competitors, but it need not explain all the facts it may confront; indeed, it never does.
What the fluid theory of electricity provided to the small school that believed in it was the power that Franklin’s paradigm later displayed for all electricians. It indicated which experiments were worth doing and which, because they were secondary effects of electricity or too complex in their action, were not worth doing. The paradigm performed this role very effectively, partly because the end of interschool debate brought to an end the incessant reiteration of basic principles, and partly because the confidence that the right path had been found encouraged scientists to pursue more precise, profound, and ardent forms of research.9) No longer obliged to investigate any and every electrical phenomenon, the united group of electricians could study selected phenomena in great detail, devising many special instruments for that purpose and employing them more firmly and systematically than had ever been done before. Both fact-gathering and the clarification of theory were transformed into activities with a definite direction. Accordingly, offering vivid evidence in support of a social interpretation of Bacon’s acute methodological maxim, the results and efficiency of electrical research increased: “Truth emerges more readily from error than from confusion.”10)
In the next section, we shall examine the nature of research that is thus directed, or grounded in a paradigm, but before doing so we must briefly consider what effect the birth of a paradigm has on the structure of the group that transmits the field. In the development of the natural sciences, when an individual or group first achieves a synthesis sufficient to attract the majority of the next generation’s practitioners, the older schools gradually disappear. Their decline is due in part to the conversion of their members to the new paradigm. Yet in every age there are inevitably some who cling to one or another of the older theories, and thereafter they are simply excluded from the profession that ignores their work. The new paradigm implies a new and more definite definition of the field. Those who are unwilling or unable to adapt their work to it must either continue in isolation or attach themselves to some other group.11) Historically, they have generally settled in departments of philosophy, from which many of the specialized sciences had branched off. As these observations suggest, it is sometimes precisely a group’s acceptance of a paradigm that transforms a group previously concerned only with the study of nature into a profession, or at least into a discipline. In science—“though not in fields whose principal reason for being lies in external social needs, such as medicine, technology, and law”—
the publication of specialized journals, the formation of professional societies, and claims to a special place in the curriculum have often been associated with a group’s first acceptance of a single paradigm. At least this was the case between a century and a half ago, when the institutional forms of scientific specialization first developed, and the time when the subdivided fields of specialization acquired their own prestige.
A more thorough definition of a group of scientists produces various consequences. When an individual scientist is able to take a paradigm for granted, he need not, in his major work, strive to rebuild the field from first principles and justify the use of every concept introduced. That can be left to the authors of textbooks.
Once a textbook is given, however, the creative scientist can begin his research where that book ends, and can therefore concentrate wholly on the most subtle and difficult aspects of the natural phenomena that concern his school. In doing so, his research reports will begin to change in a way whose evolution has scarcely been studied, but whose modern end result is clear to all and binding upon many. His researches will not ordinarily be embodied in works addressed to the general public that may have an interest in the subject of the field, as were Franklin’s “Experiments... on Electricity” or Darwin’s “Origin of Species.” Instead, his work will be published as concise papers directed only to professional colleagues—those presumed to possess knowledge of the shared paradigm and revealed to be the only people capable of reading the papers addressed to them.
Nowadays in the sciences, books most often take the form of textbooks or of retrospective reflections on one aspect or another of a scientific life. The scientist who writes such a book is likely to find that his professional reputation is harmed rather than enhanced. Only in the early, pre-paradigm period in the development of the various sciences did scientific books ordinarily have the same relevance to professional achievement that they still retain in other creative fields. And only in those fields where books, whether or not they contain papers, still remain the medium of research communication and transmission is the outline of specialization so vague that an ordinary person, reading the original works of the field’s specialists, can grasp what they contain. In mathematics and astronomy, the research reports of both fields had already become difficult for generally educated people to understand in antiquity. In mechanics, research similarly became esoteric in the later Middle Ages, and for a short time in the early seventeenth century, when a new paradigm replaced the one that had guided medieval research, it again became accessible to ordinary people. Before the end of the eighteenth century, research in electricity began to require translation for the layperson, while almost all other branches of the physical sciences became inaccessible to the public in the nineteenth century. During these same two centuries, similar transitions can be found in various branches of the biological sciences. In parts of the social sciences, such a transition appears to be taking place today. It is common, and surely proper, to deplore the ever-widening gulf that separates the professional scientist from his colleagues in other fields, but little attention has been paid to the essential relation between that gulf and the mechanisms inherent in scientific progress.
Since the remote beginnings of prehistory, one field of study after another has crossed the boundary that historians might call the division between the prehistoric and the historical within a discipline. These transitions toward the maturity of a scholarly tradition have rarely occurred as suddenly or as unmistakably as my inevitably schematic discussion may imply. Yet historically they have not occurred gradually and with equal breadth either, as though constituting the overall development of the fields within which they took place. Writers on electricity during the first forty years of the eighteenth century possessed far richer information about electrical phenomena than their sixteenth-century predecessors had had. But during the half century after 1740, only a few new kinds of phenomena were added to the list of electrical phenomena. Nevertheless, in important respects, the writings on electricity by Cavendish, Coulomb, and Volta during the last thirty years or so of the eighteenth century differ greatly from those of Gray, Du Fay, and even Franklin, and that difference seems far greater than the difference between the writings of those early-eighteenth-century discoverers of electricity and those of the sixteenth century.12) At some point between 1740 and 1780, electricians for the first time came to take the basic principles of their field for granted. From that point on, they pressed forward into more concrete and difficult problems, and then gradually began to present the results of their research in papers addressed to other electricians rather than in writings aimed mainly at the generally educated public. As a school, they achieved what had been achieved by the ancient astronomers, by medieval students of motion, by students of physical optics in the late seventeenth century, and by students of geological history in the early nineteenth century. In other words, they had obtained a paradigm that proved capable of directing the research of the entire group. Except with the advantage of hindsight, it is difficult to find any other criterion by which a field may be unequivocally declared a science.
“Notes”
1) Joseph Priestley, The History and Present State of Discoveries Relating to Vision, Light and Colours(London, 1772), pp.385__90
2) Vasco Ronchi, Histoire de la lumiere, trans, Jean Taton(Paris, 1956), chaps.i-iv.
3) Duane Roller H.D.Roller, The Development of Concept of Electric Charge:Electrictiy from the Greeks to Coulomb ("Havard Case Histories in Experimental Science", Case 8 ;Cambridge, Mass., 1954); and I.B.Cohen, Franklin and Newton; An Inquiry into Speculative Newtonian Experimental Science and Franklin`s Work in Electricity as an Example Thereof (Philadelphia, 1956), chaps.vii-xii. For the detailed analytical treatment of the passages appearing in the text, I am indebted to an as-yet unpublished paper by my student John L. Heilbron. While it is in press, a broad and more detailed account of the emergence of Franklin’s paradigm may be found in the following: T.S. Kuhn, "The Function of Dogma in Scientific Research", A.C.Crombie ed, "Symposium on the History of Science, University of Oxford, July 9__5, 1961", Heineman Educational Books, Ltd.
4) Compare with the outline concerning the natural history of heat in Novum Organum, Vol.VIII of The Works of Francis Bacon, ed.J.Speddiing, R.L.Ellis, and D.D. Health (New York, 1869), pp.176__203.
5) Roller and Roller, op.cit., pp.14, 22, 28, 43. Only after the research recorded at the very end of these quotations was repulsion recognized, beyond dispute, as electrical.
6) Bacon, op.cit, pp.235, 337, writes that “slightly warm water freezes more easily than very cold water.” For a partial account of the early history of this curious observation, see Marshall Clagett, Giovanni Marliani and Late Medieval Physics(New York, 1941), section iv. 7) Roller and Roller, op, cit, pp.51__54.
8) The troublesome case was precisely that negatively charged bodies repel one another; on this, see Cohen, op.cit., pp.491__94, 531__43.
9) It should be noted that accepting Franklin’s theory did not put an end to all controversy. In 1759, Robert Symmer proposed a two-fluid theory that modified Franklin’s theory, and for many years afterward electricians were divided over whether electricity was one fluid or two. But the controversy on this subject merely confirms what was explained above concerning the way in which a universally recognized achievement unifies the specialists in a field. Though electricians remained opposed on this view, they concluded that no experimental test could distinguish between the two theories, and therefore judged the two theories to be equivalent. Thereafter, both schools could, and did, account for all the advantages provided by Franklin’s theory. 10) F.Bacon, op.cit., p.210.