§ 12

Concepts, Facts, and Sedimentation in Experimental Science

Friedrich Steinle







My interest in Husserl centers on his notion of sedimentation, a notion that has proved particularly useful in my work in the history of science. Rather than offering a close interpretation of his writings, I shall approach Husserl in this chapter “from outside,” by presenting a specific case from the history of science. I shall consider some notions that appear to be appropriate to that case and relate them to Husserl’s concept of sedimentation. In the last section, I shall sketch the more general picture of scientific development that emerges from cases like the one under scrutiny and point out how the notion of sedimentation could be useful here.

Concepts: From Research to Language

Let me start with some observations of everyday and academic life. On the science pages of newspapers, we often find reports from specific research fields, addressed to a broad audience, for instance, the following:

  • Certain populations, or subpopulations, have a higher risk of contracting AIDS, and we know already that AIDS is an infectious disease, caused by a virus, HIV.
  • Modern versions of high-speed trains, such as the Eurostar, enable uninterrupted traffic over national borders by using modern thyristors, elements that make different alternating currents unipolar for high-power requirements.
  • Insights into the dynamics and kinetics of chemical reactions of carbohydrates enable the planned design of plastic materials, pigments, and pharmaceuticals.

These reports present modern achievements of science, and in doing so, they tacitly use numerous scientific concepts: concepts such as infectious disease, virus, population, alternating current, bipolarity of electricity, chemical reaction, compound, and pigment. In itself, this is nothing exciting—there is no way it could be otherwise: We cannot express anything without concepts, and the specialization of modern science and technology leads to specialized concepts. But that is not all there is to it. All of the concepts I have highlighted were once created and formed in a specific context of intense scientific research, and from there found their way into common language. At the site of their origin, they were open, tentative, instable, and flexible, while later on they appear as solidified, stable, either as “natural” categories, or in some cases just as “facts.” There is a long path from the openness of scientific research to the stability of concepts that are regarded as expressing simple facts.

This path, this development from openness to the closure of concepts has not yet received the attention it deserves. For much recent epistemological debate about such concepts and their features is conducted mainly in an analytic spirit, and there is little reference to historical or actual cases of how concepts come into being.1 As a consequence, an essential feature of concepts is not taken into consideration: the process of their formation or, in other words, their successive development from tentative proposals to fixed elements of language; that is to say, there is little regard for a possible historical dimension of concepts. But this neglect leaves essential points out of the picture, as can be illustrated by my examples above. Can we fully understand the concept of chemical reaction, for example, without considering its historical development and, in particular, the alternative conceptualizations against which it finally was favored and accepted? The path from scientific research to scientific language or sometimes even everyday language is of essential importance both to our understanding of the epistemological features of concepts and to our understanding of science and the validity of its conceptual foundation. As I understand Husserl, it was exactly this aspect he was interested in when he thought about the origin of geometry. And it is just these processes of conceptual stabilization, solidification, and, of course, sedimentation that I shall focus on in this chapter. However, whereas Husserl was dealing with geometry, my interest is in empirical or even experimental science. I shall develop these topics by examining a historical case, following the pathway of a specific concept from the complexities of its inception, via its first stabilization, to the sedimented state in which we use it today. From this case, significant perspectives open up concerning the details and meaning of processes of sedimentation, the relation of concepts and facts, and our general picture of empirical science and its development.

Charles Dufay and the Two Electricities

One of the most basic notions we associate with electricity today is its bipolarity. Electricity is positive or negative—this notion, or even fact, is so basic for us that it is usually given no extra emphasis or explication but is merely used as part of the very language employed in using electricity and explaining its use to others. Indeed, it is often taken to be a defining feature of what we call electricity. Plus and minus, positive and negative are fundamental categories we refer to when we plug in a new battery in our camera or when we need our neighbor’s help to start our old car after a frosty night. This has long been the case, at least since the mideighteenth century (when electricity meant what we call now static electricity). 2 Twenty years earlier, however, in 1730, the field was totally different. There was no thought whatsoever of two electricities. The concept was not proposed until 1734, but within slightly more than a decade, it led to a fundamental shift on the conceptual level. It is this shift that I shall focus on in the following.

The main actor is Charles Dufay, a brilliant early eighteenth-century academician, director of the Paris botanic garden, and versed in diverse domains such as botany and astronomy, dyestuffs and fire pumps, fluid dynamics and magnetism. The Jardin Royal offered ample resources for such experimental research. Dufay had, in an extended experimental series on shining minerals (Bolognese stones, as they were called), shown his interest in treating research fields in which even basic ordering was missing, and it was partly this interest that made him take up research in electricity. In the 1730s, the field was in an unstable and incoherent state.3 More than a hundred years of research throughout Europe had produced a multitude of different and puzzling phenomena, for instance:

Dealing with those questions proved difficult, even more so since the experiments were delicate, the effects tiny, and reproducibility was difficult. Dufay conducted extensive experiments, varying his procedure in many ways. He used a vast number of different materials, in individual or combined arrangements, and varied their shape, temperature, color, moisture, air pressure, and the experimental setting: two bodies in contact, in close proximity, at a large distance, connected by a third, and so on. His work led to remarkable results and bold claims, such as that all materials except metals could be electrified by rubbing, and all bodies except a flame could receive electricity by communication.

But still he was left with serious questions about when attraction and repulsion occurred and when the one sometimes suddenly switched into the other. He devoted extensive effort to this point. In a first step, he intended to clarify whether or not repulsion existed—Otto von Guericke’s report of the effect (1672) had been contested by some. By varying his experimental setups, he analyzed the conditions under which repulsion occurred. He succeeded in obtaining repulsion even when the arrangement of other bodies in the vicinity was changed drastically; thus the repulsive effect was clearly shown to exist. In a particularly delicate arrangement, Dufay was able to keep a very light leaf of gold hovering at a distance of more than one foot over an electrified glass tube for several minutes, even when he moved through the room with the tube—much after the model of Guericke, who had described and depicted a similar experiment (Guericke, 1672). When the force of the tube lessened, the leaf would lower and finally fall down to the floor.

Dufay was thereby confronted with the question of when exactly repulsion occurred in contrast to the usual attraction—after all, attraction counted as the defining feature of the electric virtue. Again, he approached the question experimentally, by varying many parameters of the arrangement:

For a long time, the results were merely puzzling, appeared instable, and did not resolve into anything resembling a regularity or correlation. Dufay could not formulate the conditions of when attraction and repulsion occurred, let alone when, as sometimes observed, a sudden switch between the two actions took place. This latter effect, however, was ultimately the first to be conceptualized in a rule. For specific arrangements Dufay was able to formulate regularities:

Dufay found this regularity, this law about the temporal sequence of attraction-contact-repulsion, to be valid without exception, and it could account for many of the effects he had already obtained. Nevertheless, it was restricted to the interaction of any pair of bodies in which the first had been electrified by communication from the second. In other cases, the regularity did not hold; they still appeared irregular. But Dufay did not stop here. He found a crucial hint in an experiment with a hovering gold leaf. When he approached a third electrified body to this leaf, strange effects ensued. If the third body was of glass, it repelled the leaf strongly, as the tube did. But when it was of copal, it was strongly attracted. Though this result was most puzzling,4 it gave an indication that the material used was crucial.

In pursuing this hint, Dufay became increasingly aware of regular behavior that could not, however, be formulated in the usual language of electricity. Sensitized by his former work, particularly in botany and luminescence, he was more aware than other researchers of how essentially such work depended on working with appropriate concepts and categories. More than others, he was ready to question and revise even the fundamental categories of the field, and this is what he indeed did in a radical way: In order to express the regularities, he proposed that, rather than speaking of electricity in general, one should more accurately speak of two electricities. The regularity then was that electrified bodies repelled all those that had the same electricity, but attracted all those that had the other. As the experiments showed, the electricities maintained their character even when communicated to other bodies. Thus the above regularity of repulsion turned out to be just a special case of the now general regularity. According to Dufay, which of the two electricities a body acquired on being rubbed depended only on its material. Thus the dichotomy of electricities induced a division of all materials into two classes. At the same time, the electricities could be characterized according to these classes—Dufay named them vitreous and resinous electricities, respectively, in reference to particular prominent representatives of the two classes of materials.

We have no explicit notes about what exactly convinced him of the power of these new concepts. Presumably their immediate “success” was central here: Dufay emphasized that, with these concepts and the regularities they allowed him to formulate, he could understand not only his own, quite numerous experiments but also those reported by others. The concepts enabled him to do exactly what previously had been impossible: to order the whole field in such a way that stable regularities could be formulated. This was indeed a fundamental achievement, and Dufay was well aware of that. He not only announced his proposal immediately at the Paris Academy (as he did usually), but he also took the unusual step of writing a letter to the Royal Society in London to make his proposal known (Dufay, 1734). This striking deviation from his usual method of communication indicates how well he was aware of the fundamental character and importance of his new proposal.

How the Two Electricities Were Received

It was a very specific level of knowledge that Dufay aspired to and achieved. He was well acquainted with the various theories that attempted to explain electric effects by air currents, by humorous components of matter, by specific electric effluvia, and so on.5 However, those theories did not appear in his research. His aim was not to consider the hidden causes of electric effects but to order them on the phenomenological level, to establish regularities, and to find the appropriate concepts that would allow him to do so. It would mean “attempting the impossible,” he emphasized, if one were to search for causes without previously having discovered the large number of phenomena and having ordered them along a “few simple principles” (Dufay, 1733b, 476). From what Dufay actually did, I interpret this as the task of establishing regularities and laws. And in such a context, there was no space for theories about hidden causes. The search for correlations and regular dependencies dominated his enterprise, and in the end he was successful. But his success depended on a step that he had not planned or envisaged at all: on the introduction not of a new theory but of something more fundamental, namely, a new conceptualization, a new way of speaking about and manipulating electricity.

Before Dufay, there was only an electric virtue, defined by the attractive action onto small bodies nearby. With Dufay’s innovation, the field looked much different. There were two species of electricity now, defined not only by their action on small bodies nearby, but essentially by a mutual interaction that could be either attraction or repulsion. Moreover, that dichotomy induced a classification of all materials according to their susceptibility to those two electricities when subjected to friction. In handling electrical effects, in designing, conducting, and evaluating experiments, new considerations became prominent. New questions were required and thus arose, questions that previously had simply been inconceivable because the very categories they used did not exist. For everyone who would adopt the new conceptualization—for the time being this was just Dufay himself—the field of electricity appeared totally transformed.

It is striking to see how the other electricians of his time reacted to Dufay’s proposal. The first historical observation is the nearly complete absence of an explicit discussion of the topic. This is all the more significant, as the proposal did not disappear. On the contrary, it soon showed up again, but in a very specific manner. Only five years later, the writer of a most important physics textbook of the period, Leyden professor Pieter van Musschenbroek presented the new concept as unproblematic. “Experience has taught us that electrical power is of a twofold manner ... ,” he told the reader.6 It is worthwhile noting that he did not say “Dufay has made this claim,” but “Experience has taught us.” Of course, this was not a denigration of the effort and achievement of Dufay. It was a significant indication of the specific level of knowledge that was addressed here. The existence of the two electricities was presented as indisputable, as a fact of nature, as we would say—and indeed, this quickly became the common view.7 Five years later, in the first textbook ever devoted exclusively to electricity, Petersburg academician Johann Gabriel Doppelmayr went even further and took the twofold nature of electricity as a very part of its definition (Doppelmayr, 1744, 1). And this was characteristic of a general attitude. In textbooks of the eighteenth century, the twofold nature of electricity was very often introduced as a defining feature, derived “from experience,” and only later in the text were the specific laws of attraction and repulsion presented (Frercks, 2004). Already here the concept was detached from the context in which it had been framed and was presented as an ontological feature, as a fact that as such was no longer subject to dispute. Such a way of “reception,” of “filtering in,” is strikingly different from the way in which theories proper are received, both by the absence of explicit discussion and by the quick introduction of the concept not as something to be problematized, but as a fact, taught by experience.

Sedimentation and the Impact of Concepts

As to the background of such a peculiar and striking way of receiving a proposal, I suggest that its peculiar character is due to the specific epistemic level in question here: It is not theories in some strict sense that are at stake in such cases, but the concepts themselves. Concepts and categories, as elements of language, cannot be true or false but are rather appropriate or inappropriate, useful or useless. They cannot be proved, confirmed, or disconfirmed, but they have to prove themselves against certain goals. What exactly the goals are, what exactly should count as “appropriate” means, is a matter of the specific historical situation. For Dufay himself, the touchstone was that any new conceptualization had to permit what the former had failed to achieve: the formulation of a coherent and general law of attraction and repulsion.

There may be other criteria, and for many others there have indeed been other criteria, though closely related ones. Perhaps the most central one had been success in manipulation. The new concepts allowed for an unprecedented stability in practically dealing with electrical effects. In 1744, Wittenberg professor Georg M. Bose redesigned the older electrical machine in a way that made a still greater stability of electrical experiments possible, and this provided the major background for the explosive spread of electrical performances throughout Europe within a short period. It is remarkable, however, that Bose explicitly made reference to Dufay and his two electricities (Bose, 1744). When, in 1746, the news of the Leyden jar shocked electricians throughout Europe and marked another starting point for what is sometimes called the “golden age” of electricity (with strong electric machines, spectacular displays at public events, and the first commercial uses of electric effects), the discourse of two electricities had become unproblematic and even “naturalized.” Students of electricity took it straight from their textbooks; users of electrical machines learned of it in the workshops where they bought the machines.

This overwhelming instrumental success is one of the most intriguing features of the golden age of electricity. It can be clearly seen now, however—and this has long gone unnoticed in the histories of electricity—that the foregoing reconceptualization of the field provided one of the essential conditions for this success. Reconceptualization may have far-reaching consequences, as this case illustrates. The new concepts were, in a sense, incorporated into the machines and instruments, for example via the choice of materials and the design of the apparatus. At the same time, it was exactly this success in dealing with electricity, in inventing and using new instruments, that provided a central support for the appropriateness of these new concepts.

Once again, the different epistemic levels involved here become visible: In contrast to the formation and quick stabilization of the concept of two electricities, the situation of explanatory theories remained incoherent and unsolved, and this state would not improve with Franklin’s theory of the 1750s. The debate over whether one should explain the bipolarity of electricity (to use a later name for what Dufay had called two electricities) by assuming one or rather two electric fluids would last for many decades, to remain unresolved and to run dry toward the end of the century. However, the very concept—or fact—of bipolarity was never questioned in this debate. It was not any clarification on the level of explanatory theories that furthered the spread of electricity since the mid-century, but the lasting success of the new conceptualization achieved by Dufay.

What we see in this historical case is how a new concept may be created in a quite specific context, by an author with definite, if not idiosyncratic interests, and in the course of broad and systematic experimental exploration.8 We see then how this concept has been stabilized and has found, in a rather brief period of time, its way into the very language of the research field. Such a process recalls what Husserl describes as the process of sedimentation, a process that results in the end in the existence of a concept that we use in ignorance of its original “meaning.”

Before I examine this analogy more closely, let me add that the case I have presented is far from unique in the history of the experimental sciences. Such processes occur again and again, in fields of research as diverse as chemistry, physiology, physics, and genetics, and in quite different periods, from the early modern period until the present day. Just to mention a few cases from the history of electricity, one may take the concept of a current circuit, covering the battery and its external wire at the same time and describing these two with the same language, a concept that Ampère framed to formulate a law of electromagnetic action (Steinle, 2002). Or one may think of the concept of magnetic and electric lines of force that Faraday developed in the course of twenty years of intense experimental research, and that allowed him finally to describe a vast realm of effects with very few laws (Steinle, 1996; 2005). In a different field, consider the concept of chemical reaction that was shaped and developed in the seventeenth century on the basis of broad experimentation, not by single prominent individual researchers, but rather within a community with a tight communication structure (Klein, 1994). Lastly, we may take the concept of infectious disease, a case to which I shall come back in a moment. Different as these cases may be, they illustrate well the fundamental character and the far-reaching impact of such processes of concept formation and stabilization.

Concepts and Facts

While these examples of generation, stabilization, and solidification of concepts resonate strongly with what Husserl calls sedimentation when discussing the case of geometry, there are also significant differences. A first indicator is the historical actor’s language. While geometricians are well aware that they deal with (man-made) concepts, in the empirical and experimental sciences the language is different: One speaks of facts (usually in contrast to theories), by which one means states of things as they are in nature, states that are not inventions of the human mind, and that are taught directly “by experience,” as Musschenbroek put it. In contrast to geometry, moreover, the awareness that basic concepts were once actively formed and stabilized tends to disappear in scientists’ own accounts and in our usual understanding of facts. Facts are taken as a given, and thus as unproblematic and undisputable starting points of theorizing. Even in philosophy of science, there is sometimes little awareness that facts might not just be given, but necessarily involve conceptual (not necessarily theoretical) activity, effort, and even creativity.9 A closer look at the practice of scientific research provides us with a more complex and differentiated picture, both historically and epistemologically.

Such alternative accounts were provided early on. Strikingly enough, it was roughly at the same time that Husserl wrote his Origin of Geometry, though independently of Husserl, that the young and unknown Polish immunologist Ludwik Fleck published a book whose very title was a provocation to many: Entstehung und Entwicklung einer wissenschaftlichen Tatsache (Origin and development of a scientific fact) (Fleck, 1935; 1980). It is still an intriguing historical question as to why exactly so many deep reflections on science were produced in the 1930s, for instance Popper’s 1934 Logik der Forschung, Fleck’s 1935 Entstehung, and Husserl’s 1936 Ursprung der Geometrie. One common background was certainly the reaction to logical empiricism, but there must be more to say. The three accounts differed drastically, both in their direction and their fate. Fleck’s book, in particular, written in German by a Jewish author, found no resonance in the political atmosphere of Nazi Germany. Even later, not the least through the dominance of the Logical Empiricist immigrants in the English-speaking world, there was no interest in an English translation. It was only Thomas Kuhn in the 1960s who realized what Fleck had to say, but it would still take until 1979 before the book was available in our modern lingua franca.

Among the three, Fleck’s approach was the only one directed primarily at experimental science, in particular at his firsthand experimental experience as an immunologist. He presents in great detail the path that led, in immunology, to identifying the agent of syphilis and to the proposing of a reliable test, a development that took place at a Kaiser-Wilhelm-Institute in Berlin, and for which its main actor, August von Wassermann (1866–1925), received worldwide honors. Fleck shows in detail how much this research was based on concepts such as infection, disease, and immunity—notions that were taken for granted by many, but which, on closer inspection, bore significant traces of their genesis, and in particular of the decisions made within that development. Fleck emphasizes how much research depends on the existence and use of concepts that are taken for granted: “Every system will then become self-evident know-how itself. We will no longer be aware of its application and its effect” (Fleck, 1980, 114).10 In order to grasp this phenomenon, he proposed his well-known notions of Denkgewohnheit and Denkstil (habit of thought and thought style).

Moreover, Fleck pointed out that when concepts disappear from the realm of explicit discussion, they are often turned into facts, that is, into a category that per se is taken as being immune to theoretical objections. Facts, in the usual understanding, present things just how they are, and we learn them from experience. But we have to use concepts and, as Fleck shows in detail from his historical case, these concepts are not just given but have been created, formed, and stabilized in a complex process, a process that lies in the past and is usually not recovered. In empirical sciences, the generation and stabilization of concepts is closely related to the genesis of facts. This is why Fleck, in his very title, speaks about the genesis, not the discovery, of facts.

As a side remark, let me note that Fleck, while he repeatedly emphasizes that these processes are inherently bound to social constellations and to communication, likewise puts emphasis on experimental outcomes and theoretical stringency. He is sometimes read as propagating a purely social-constructionist view of science, as explaining scientific development by changing sociological constellations alone, but this is clearly a misinterpretation. On the contrary, what he did was put his finger on the fact that there is no simple account of science, one focused on theory, experiment, or social setting alone, but that we have to develop a multifaceted view. It is this point that Kuhn took up and elaborated.

Sedimentation: Geometry and the Sciences

Fleck’s historical case shows striking similarities to the case of the formation and stabilization of the concept of two electricities. Both of them illustrate a process in which concepts are formed and developed in an open way and later become fixed to such a degree as to form stable and unquestioned elements of the language of the research field. It is not only a process of stabilization we see here, but also of a sort of sedimentation: The now stable elements serve as an unshakable foundation, as the unquestioned ground for further research. Of course, processes of such a type remind us immediately of Husserl’s discussion of sedimentation.

Husserl introduced this notion in discussing the origin and development of geometry.11 By attributing fixed notions or symbolic signs to geometric concepts, he says, their original meaning may disappear in the course of time. Later generations take the signs and receive them passively but no longer reactivate their original connotations and meaning. Indeed, such a process is inevitable in constructing the “Stufenbau,” the ever-growing system of geometry: Without it, everyone would have to start from the beginning and would waste her or his whole power in taking the first steps again. There is, in principle, always the possibility of “reactivating” the original meaning of geometrical concepts. As a matter of fact, however, this is simply not done. Since the early modern period at least, geometry has been in constant danger of becoming a “tradition empty of meaning,” such that “we could never even know whether geometry had or ever did have a genuine meaning, one that could really be ‘cashed in’ ” (Origin of Geometry, p. 366).

While Husserl articulates all this with respect to geometry and its development, he seems to see a fundamental contrast to the empirical sciences. There are only rare, but telling references to the sciences in his text. For the “so-called descriptive sciences,” his analysis looks totally different: At least in principle, he says, every new sentence can be “ ‘cashed in’ for self-evidence” (Origin of Geometry, p. 363). The basic idea seems to be that there is no such Stufenbau as in geometry, where the original meaning of a sentence can be veiled, but instead always a direct access to that original meaning by empirical evidence. Empirical evidence enables and guarantees, so Husserl seems to say, the authenticity and directness of our statements about the world. As a consequence, neither a process like sedimentation nor the problems connected with it play a role in empirical science.

To be sure, this is a strong interpretation, based on the few textual passages he gives in his text. And it is not clear to what extent Husserl would extend the compass of the descriptive sciences. Did he just think of natural history, or would he also embrace fields with instrument-mediated observation and even experimentation in his analysis? The text does not focus on these questions and does not provide enough material to decide them. But if we regard Husserl’s statement about descriptive science and the immediate evidence it provides as a general statement about observation, we might well take it, for a moment, as a statement about empirical science in general, as contrasted to the “deductive sciences” exemplified by geometry (Origin of Geometry, p. 365).12

Such a picture of empirical science would contrast strikingly with my reflections above concerning (empirical) concepts and facts. Rather than always recurring to direct and immediate evidence, these considerations tell us, empirical science has recourse to facts—but these facts are, no less than the concepts of geometry, the result of successive historical work, of processes of formation, stabilization, and sedimentation. And again, as in geometry, the main medium of sedimentation is language. The “seduction of language” exists no less and is no less effective in the sciences than in geometry (Origin of Geometry, p. 362). The very concepts that scientists use to express their empirical results can be subjected to an analysis similar to that developed by Husserl for geometrical concepts. Of course, Husserl’s treatment of Galilean science in the Crisis does offer the outline of such an analysis. However, even here Husserl discusses the mathematization of the world as if it were a purely philosophical or theoretical activity, and not one that transpired in large measure in the scientist’s workroom.13 What is more, he still has only those sciences in mind that have been mathematized, as if the process of sedimentation was restricted to mathematical concepts. He does not envisage the possibility that those processes might also occur in nonmathematical fields, that is, in large parts of empirical science, even in fields that he calls descriptive sciences. Cases like Fleck’s, or the one presented above, indicate that sedimentation is significantly more endemic than Husserl envisaged.

The Picture of Science

However, there is a disquieting observation: All these notions of sedimentation and solidification point to a cumulative picture of scientific development. Indeed, Husserl is often read as adhering to a picture of a science formed by successive layers, such as we have in geology, which accumulate in solid formations. At the same time, such a picture of science has been forcefully attacked by Kuhn, and with good reason. When we discuss and emphasize these processes of sedimentation, does it follow that we have to give up Kuhn’s insights, and all the others that have followed, in order to return to the older, cumulative picture?

I would argue that this is not the case. I will, once again, illustrate this with recourse to history. As I have suggested, we have numerous historical cases that can well be described as processes of conceptual sedimentation. But it is exactly this material that requires important qualifications. The picture that emerges from these cases is not the formation of a massive bedrock like sandstone, with no possible further movement, in which all gaps are filled. The emerging picture reminds one rather of a porous formation of chalk, or even better a coral reef. On such a reef, we can to a certain degree differentiate between those parts that live and develop and those that have been solidified. But the border between them is not always clear and sharp. Even those coral formations that have sedimented are not absolute and fixed. They may collapse under their own weight, cavities may not only later be filled in, but parts may break off, new vistas open, fragments may attach themselves at other locations, and so on. At many single points on the surface, at different points of time, they offer a base for further attachment; they stimulate specific directions of growth and disfavor other ones. While they offer the base for local development, their development is much too complex to be really determined by its state at a given time. While such a formation indeed develops by a process of sedimentation, it is neither absolute nor immutable. Smaller and major readjustments, even Kuhnian revolutions, are quite possible.

With such a picture I would deviate from what Husserl had in mind with sedimentation and when he talked of a Stufenbau of geometry and the deductive sciences. But Husserl is a complex writer, and still I’m not sure what a Husserl specialist might draw as “his” picture. Such a picture of a coral reef, however, certainly fits well with what Fleck describes as a Wissensgebäude, a building that has been anticipated and designed by none of its builders, and that has often been altered in unexpected ways, sometimes indeed against the clear intentions of some of its construction workers (Fleck, 1980, 91). The shift in pictures might well be related to my shift of domains, but one may doubt even this: Modern historiography of mathematics, for example, could probably well sustain such a picture even in the case of the development of mathematical research fields (Epple, 1999).

For our activity as historians and/or philosophers of science, such a picture suggests a further aspect. When we want to explain and understand the structure of any specific reef, we have to apply a peculiar type of explanation. On the one hand, we cannot do without knowing the general laws of crystal growth; the crystallization modes of the materials involved; their mechanical, chemical, thermal, and optical properties; and so on. All this general knowledge is indispensable. At the same time, however, we cannot but have recourse to the reef’s specific history, to its specific location, and to chance events: For example, some branches may have started to grow when a ship ran aground on it, destroying some of its parts, and occasioning growth in new directions. A warm ocean current may have changed direction, thus altering the fauna and flora and the process of sedimentation. General laws and the specific history are inseparably intertwined, and we shall never understand the wonderful structure of the reef without invoking such mixed explanations. Both the gross shape of the reef and its individual elements bear traces of its history and cannot be understood without it.

Analogously, we might say that we cannot understand specific scientific concepts without using a mixed method, invoking both general aspects such as appropriateness to explanatory needs and specific historical features of the process of concept generation. This is exactly what Husserl had insisted on toward the end of his Ursprung paper: We cannot understand the full meaning of geometrical concepts without looking to their history. To be sure, there are important differences in where exactly this recourse to history might lead us—Husserl’s idea of a historical a priori might no longer be shared these days, and chance might remain as an irreducible element. But the common ground is the insight that an understanding of the conceptual structure of science needs a serious look at its history. As Ian Hacking has put it, “concepts have memories” (Hacking, 2002, 37). And so do facts, we might now add, with a look at empirical sciences. Both of these statements are at this point probably more like a program of research whose general features and implications remain to be uncovered. The challenge for the future that opens up here might perhaps not be too different from the challenge Husserl wanted to pose to his contemporaries seventy years ago.