When Scientists Disagree: Carl von Heß, Karl von Frisch, and the Study of Controversies

When science studies emerged as a new field of research in the 1970s, one important starting point, besides laboratory studies, was the examination of scientific controversies.¹ Controversies represented “the very lifeblood of science,” were praised as “strategic research sites,” indicated the “interpretative flexibility of science,” and were thought to reveal “norms,” “interests,” and “rival ideals and images of science.”² In response to this “relativist empirical program,” the study of controversies also attracted attention within the history and philosophy of science (HPS).³ The “rationalist view,” as it sometimes called itself, characterized the controversy as a model case of empiricism and logic, whose results could be explained without reference to social or cultural factors.⁴ The discussion of controversies thus became partly the subject of a “meta-controversy” over the nature of science.⁵ Less militant contributions from HPS emphasized the role of controversies in scientific change, focused on the resolution of controversies, and/or examined controversies as epistemological laboratories—each in their way underlining the rationalist ratio of scientific controversies.⁶

Beginning in the 1980s, controversy also became a prominent topic in the history of science for some time. Two major innovative monographs from this period—Martin J. S. Rudwick’s The Great Devonian Controversy and Steven Shapin and Simon Schaffer’s Leviathan and the Air-Pump, both published in 1985—dealt with prominent controversies, one from the nineteenth century and another from the seventeenth, but with quite different intent. Rudwick conceived his study not least as a test case for the essential role of controversy in the development of arguments and observations. In his words, the “outcome of research is neither the unproblematic disclosure of the natural world nor a mere artifact of social negotiation.” Rather, “new knowledge” is “shaped from the materials of a real natural world” in the course of debate.⁷ Rudwick thereby took a position between the “strong program” type of social constructivism and a trivial version of philosophical realism.

Shapin and Schaffer, for their part, saw controversies as a means of applying the method of the “stranger” (that is, examining a culture or a problem with an outsider’s eye), emphasized in laboratory studies to historical material. In their view, controversies have two advantages for historians of science. First, they “often involve disagreements” over entities, problems, and practices that are “subsequently taken to be unproblematic or settled,” providing an opportunity to examine how things develop from contested to stable.⁸ Second, actors in controversies behave in some respect themselves as strangers insofar as they “attempt to deconstruct the taken-for-granted quality of their antagonists’ preferred beliefs and practices, and they do this by trying to display the artifactual and conventional status of those beliefs and practices.”⁹ Shapin and Schaffer themselves pursue this goal through their attempt to strip the experimental method of its current self-evidence as the privileged, accepted way of gaining scientific knowledge.

Rudwick’s monograph, and likewise Shapin and Schaffer’s, show that the study of controversies was constantly linked to more general ideas on the nature of science within the history of science as well. This is confirmed by the literature on the so-called Biometrician-Mendelian Controversy. At the end of the 1980s, Robert Olby attributed the growing scholarly interest in the notorious disagreement over the role of environment and life history in the inheritance of traits to the “increasing attention which has been given to the social history of science and to the sociological analysis of scientific activity.”¹⁰ Even Gregory Radick, who in his recently published monograph placed great emphasis on retelling the debate between Walter F. R. Weldon and Gregory Bateson from the ground up, continues this line of thinking when he tries to show that the advantages of Kuhnian “normal science,” rather than “social” interests, decided this controversy in favor of the Mendelian party.¹¹

Finally, historians of science have also reflected on the (ab)normality of controversies. This aspect is addressed in two ways. On the one hand, scholars emphasize that controversies contribute to a better understanding of a problem. James Secord underlined the “fruitfulness” of the disputes in nineteenth-century British geology, claiming that the “major accomplishments of Lower Palaeozoic geology in Britain prove to have been part of a debate, not something separate from it.”¹² Similarly, R. Steven Turner concluded his study of the Helmholtz-Hering controversy over the foundations of human vision with the assertion that “controversy, be it normal or anomalous, is nevertheless constitutive of deep change in science.”¹³ Although in this way the question of the normality of controversies is not so much answered as pushed aside with reference to their productivity, Naomi Oreskes reformulates the same question as an epistemological challenge. Her study of the fate of Alfred Wegener’s continental drift theory in North American earth sciences is based on a single question: “Why did distinguished scientists adamantly reject as false a claim now universally accepted as true?”¹⁴ At first glance, this question has nothing to do with the role of controversy, and in fact there is no corresponding keyword in the book’s index. On closer examination, however, Oreskes presupposes with her question that it is an event to be explained when a theory that is later regarded as correct in its basic features is originally rejected, and she thus further presupposes that a theory that is correct in its basic features is normally immediately regarded as such—without decades of controversial back and forth. In fact, the basic question of Oreskes’s monograph is not why North American geologists rejected continental drift. The basic question of her monograph is how it was possible for scientists to become fixated on rejection without appreciating the evidence presented in the ongoing debate.¹⁵

For all their profound differences and diverging interests, scholars from science studies, HPS, and history of science share a basic understanding of controversies as highly instructive instances. Trevor Pinch compares controversies to the “breaching experiments” in Harold Garfinkel’s school (that is, the deliberate violation of social norms in order to study the rules of interaction), with the only difference that the breach “in the normal social operation of science is produced by the scientists themselves.”¹⁶ For Aristides Baltas, controversies uncover the usually hidden “background ‘assumptions’” of researchers.¹⁷ Thomas Brante sees in the controversy a means of saving work because here the experts themselves disclose the weak points and blind spots of arguments.¹⁸ Shapin and Schaffer rely on the basic idea, common to all studies, that controversies provide an “opportunity to see” something that is otherwise apparently obscured.¹⁹ And Marcelo Dascal equates controversy with the “de facto nature of the workings of scientific rationality (or irrationality).”²⁰

In summary, it can be said that controversies are often treated as similar to natural experiments, which, with no planned intervention, reveal elementary attributes of scientific knowledge production. At the same time, and in some contradiction to this, controversies are understood as a special kind of disagreement. To qualify as a controversy, a conflictual event must be public and prolonged, involve more than a minor issue, and extend beyond the scientists initially involved.²¹ From this perspective, controversies do not simply happen and wait for scholars to analyze them as happens with earthquakes or sudden food scarcity. Rather, by comparing the circumstances with a list of criteria, scholars classify some conflictual events as controversies, and exclude other conflictual events as insufficiently controversial. I therefore argue that for science studies, HPS, and history of science scholars, the controversy is not primarily a research object but a conceptual tool developed to remodel current or past events in the sciences with the aim of capturing the nature of science.²² To engage with controversy, then, is to think about the properties of a tool. Central questions are: what aspects of science acquire the status of a problem through the use of the tool, what does not raise questions or seems unthinkable, and what notions of science emerge.

I will develop my observations by means of an example. In the 1910s, ophthalmologist Carl von Heß (1863–1923) and zoologist Karl von Frisch (1886–1982) came to opposite conclusions on the question of whether fish and bees perceive colors.²³ Heß proved that fish and bees perceive the spectrum of electromagnetic radiation visible to humans as gray values of specific brightness (on a scale from black to white) and concluded that both fish and bees are totally color blind. Frisch proved that fish and bees distinguish colors from shades of gray of corresponding brightness and concluded they perceive the spectrum of electromagnetic radiation visible to humans as colors of specific quality (the visible spectrum being somewhat shifted and less differentiated). Their research touched on broader theoretical issues, methodological preferences, and disciplinary features; the disagreement lasted nearly fifteen years, led to a pronounced partisanship in professional circles, and became very personal.

In other words, all the necessary elements came together in the debate between Heß and Frisch to constitute and analyze it as an instructive controversy. For example, their respective academic training and the resulting styles of reasoning and experimenting may have played a role, a point that Olby generally emphasized with respect to scientific disagreements.²⁴ It is also plausible that the dispute was a sideshow of a much older scientific conflict. For scholars interested in social aspects, the confrontation could be traced back to the specific milieu of the German university at the beginning of the twentieth century. Given its long duration, the conflict might also be a good candidate to test Rudwick’s thesis that the observations made in its course are “controversy laden,” by which Rudwick means that research is evaluated and, I would add, planned with the intention to find not only “relevant” but also “persuasive” empirical evidence.²⁵ I argue that the most important reason for the completely different positions of Heß and Frisch lies in their respective concepts of evolution. Heß followed the model of a linear ascending perfection of organisms with certain abilities assigned to each stage. For Frisch, the abilities of an organism resulted from its particular adaptation to and interaction with the environment.

In what follows, I will address some of the points just mentioned, but it is not my intention to confront the existing research with yet another, more apparently conclusive interpretation of the debate.²⁶ My intention is rather to report as much of the discussions between Heß and Frisch as is necessary to think about the controversy as a conceptual tool for exploring the nature of science. This goes hand in hand with a detailed exploration of the debate, but at the core I am interested in understanding what may happen when sociologists, philosophers, and historians of science turn events into controversies. For this purpose, I discuss four points that play a special role in studies of controversies (of course, not every point in every study carries the same weight). I start with the extensive interest in the end of controversies and the limited interest in when they begin. In the next section, I address the usual account of a controversy as a sharp confrontation between clearly demarcated positions. Then I deal with the assumption that controversies are conducted according to rules. Finally, with reflections on why controversies do or do not arise, I take up a rarely discussed aspect that nevertheless concerns our understanding of controversy at its core. With the first two points I want to show how controversy as a conceptual tool can hinder observations and insights. With the third and the fourth points, I turn to basic assumptions in the study of controversies. Finally, I will discuss which concepts of science result from the studies of controversies.

In April 1923, Karl von Frisch received the Rainer Medal of the Zoologisch-botanische Gesellschaft in Vienna. He began his lecture, entitled “Das Problem des tierischen Farbensinns [The Problem of the Animal Sense of Color],” with a reminiscence: “About a decade has passed since C. v. Heß and I came to diametrically opposed views in our research on the sense of light and color in animals. The fact that since then more than 60 studies have been published in this field is a sign of how stimulating the conflict has been. If we look back today, I hope to be able to show you that the original dispute has been decided and that other more specific problems are at the center of interest.”²⁷ In what followed, Frisch recapitulated Heß’s and his own work, dealt with objections, and added studies by other researchers in support of his position. The gesture was similar to that of a “vademecum” in which, as Ludwik Fleck has remarked, all the available results are not simply added up but in each case the author has to decide “what counts as a basic concept, what methods should be accepted, which research directions appear most promising, which scientists should be selected for prominent positions, and which should simply be consigned to oblivion.”²⁸

According to Frisch’s account, in 1923 it was no longer an open question whether fish and bees perceive colors. Heß himself had seen it quite differently the year before. In a review article on “Farbenlehre [Theory of Color],” he repeated his position that fish and invertebrates show the same behavior “as the dark-adapted normal human being who sees at reduced light intensity and as the completely color blind human being at any light intensity.”²⁹ Heß was no longer able to reply to Frisch’s most recent remarks, having passed away in late June 1923 at the age of sixty, a few days after Frisch’s Vienna lecture was published. After Heß’s death, his position quickly deteriorated. A certain distance was already expressed in the obituaries. His colleagues and former students spoke of “captivating, if controversial, aspects of the theory,” of “investigations” that were “not generally accepted,” and of Heß having “gone too far” in his conclusions.³⁰ In the following years, the importance of Heß’s contributions was greatly diminished. In a paper from late 1924, Frisch still emphasized that the “startling communication by C. v. Hess that bees and all other invertebrates (and among the vertebrates, fish) are completely color blind” had provided the impetus for his experiments. How Heß had come to this conclusion, however, was not revealed to the reader. “We do not need to discuss his arguments here,” decided Frisch, “because we know today that his thesis was based on a passionately defended fallacy. His lasting merit for the question of color sense lies in the fact that he drew attention to the inadequacy of the older ‘evidence’ for color vision in animals.”³¹ Important was Heß’s warning that a completely color-blind eye, of course, also distinguishes light of different wavelengths—but as shades of gray, not as colors.³² In the design of experiments on color vision, this circumstance had to be taken into account. Were one to believe Frisch, everything else from Heß’s studies could be safely forgotten.

In studies of controversies, special attention is paid to to their ends for two reasons. First, this should reveal the extent of the “epistemic gain” of the controversy.³³ This point is particularly relevant for those positions that conceptualize controversies as a driving force of scientific change. Second, the way in which a controversy comes to an end is meant to exemplify what characterizes science as a means of knowledge production. In the terminology of Philip Kitcher: adherents of the “rationalist model” assume that controversies “are ultimately settled by experimentation, evidence, and the exercise of reason,” while adherents of the “anti-rationalist model” (which largely corresponds to the so-called relativist position) assume instead that “context-transcendent canons of reason and evidence have no power to resolve major scientific controversies.”³⁴ From this perspective, what is at stake in the end of a controversy is what essentially defines science.

Both aspects can easily be related to the discussion between Heß and Frisch. Following Frisch’s account, we can see how the disagreement ended, in Ernan McMullin’s terminology, with a “resolution,” that is, any conflicting point was clarified by purely rational means.³⁵ If we take into account that the debate was still in flux until Heß’s untimely death, we can in turn assume that the disagreement was not resolved at all but simply ended by “abandonment.”³⁶ Finally, it was admitted recently that the contradictions in Frisch’s results discovered by Heß were systematically concealed by Frisch and his partisans.³⁷ The end of the conflict was thus the result of an arbitrary act of “closure.”³⁸ The gain of the debate is not so clear either. At first glance, it stimulated progress on a tricky problem and led to a number of other discoveries: Heß, for example, realized the reaction of bees to ultraviolet, and Frisch’s experiments en passant made him aware of the possibility of communication among bees.³⁹ Frisch, whose career was just beginning at the time, also benefited from the controversy by gaining visibility and recognition.⁴⁰ On the other hand, Heß’s reputation cannot have suffered much: otherwise the Deutsche Ophthalmologische Gesellschaft would not have awarded him the Graefe Medal, which had previously been given to only three people, including Hermann von Helmholtz and Ewald Hering.⁴¹ In the early 1950s, however, the gain was assessed much more critically. For ethologist Nikolaas Tinbergen, the whole debate proved to be an example of “sterile controversies” because, as he argued, whether bees—and the same would apply to fish—perceive colors depends on the particular experimental situation.⁴²

While studies of controversies have high expectations for their endings, they have little to say about their beginnings. According to the definition given in the introduction, the first step toward a controversy happened on June 8, 1911. On this day, Frisch gave his lecture “Über den Farbensinn der Fische [On the Sense of Color in Fish]” at the annual meeting of the Deutsche Zoologische Gesellschaft in Basel. He reported on Heß’s negative experiments before presenting his own positive experiments.⁴³ With that, a disagreement was born. The second step was taken in March 1912 when Heß discussed Frisch’s results for the first time.⁴⁴ His summary was unambiguous: he briefly stated “how little v. Frisch’s experiments are suitable to serve as support for the assumption of a sense of color in fish.”⁴⁵ Now two scientists faced each other in public. In a final step, the disagreement had to gain longevity. Decisive in this respect was probably the fact that Frisch, in a lecture in November 1912—again in opposition to Heß—extended the discussion to color vision in bees.⁴⁶ From then on, the disagreement between Heß and Frisch attracted sustained attention among zoologists and physiologists. Over thirty publications followed until Heß’s death, often harsh in tone and always at odds.

At what point a controversy begins is typically treated as a purely formal question: a controversy begins when the checklist just outlined is completed. Yet this ignores the fact that a disagreement can go back far beyond the time at which it becomes public. First, in order to argue with Heß at all, Frisch had to deal with color vision in fish. According to his autobiography, interest in the topic followed from his previous research. For his doctoral thesis, completed in 1910, Frisch had studied the nervous stimulation of pigment cells in the minnow (a freshwater fish), through which the brightness and coloration of the skin can be altered.⁴⁷ As he wrote in his autobiography, Heß’s view that fish are completely color blind therefore seemed to him “incompatible with my own observations regarding the ability of fish to adapt themselves to the colour of their background.”⁴⁸ Subsequently, Frisch began refined experiments with minnows and later with marine fish, which he conducted as a guest at the Zoological Station in Naples in spring 1911.⁴⁹ The results formed the basis for Frisch’s presentation in Basel two months later.

Following Frisch’s own account, his knowledge of the subject left him no other choice but to disagree with Heß. What Frisch did not mention in his autobiography was the circumstance that, two years before his stay at the Zoological Station, it was precisely there that Heß had begun his investigations into the sense of color in fish.⁵⁰ This may not seem particularly remarkable at first glance. The station’s countless guests in the decades before and after the turn of the century included practically every scholar engaged in biological questions. The ambience was appealing, the supply of research material ample.⁵¹ The fact that Heß had also worked there is mainly interesting because a debate had already developed about his experiments in Naples before Frisch’s work.

Victor Bauer (1881–1927), an employee of the Stazione, originally assisted Heß in his studies as a photographer.⁵² Bauer then repeated Heß’s experiments and concluded that the results apply only to “dark-adapted fish,” whereas in the case of “light adaptation” the “distinction of colors” is added to the “distinction of brightness.”⁵³ Heß responded by accusing Bauer of “giving an inaccurate picture of the factual situation through erroneous reporting and inadequate experiments.”⁵⁴ Bauer, in turn, firmly rejected Heß’s charges.⁵⁵ This was the state of affairs when Frisch arrived in Naples at the beginning of March 1911 and was received, among others, by Bauer.⁵⁶ In his Basel lecture three months later, Frisch referred to Bauer’s work and mentioned that a “polemic” had developed between Heß and Bauer.⁵⁷ Although he immediately emphasized that he “did not want to go into it,” the audience and later readers of his lecture could easily get the impression that he was on Bauer’s side.⁵⁸ In addition to the line leading from Frisch’s doctoral thesis, via Heß’s finding that fish are color blind, to the experiments reported by Frisch in his Basel lecture, a second line can now be drawn. This path starts with Heß’s investigations at the Zoological Station in the spring of 1909, leads to the furious exchange over Bauer’s deviating results in the following year, and continues through Frisch’s experiments in Naples, which, from this perspective, appear as a new contribution to an existing dispute.

Finally, there is a third possibility. For this, one must know that in 1904 Heß had an argument with Frisch’s uncle and scientific mentor, Sigmund Exner. Heß rejected Exner’s basic studies on the stimulation of the human retina as outdated, prompting Exner to dismiss Heß’s observations as marginal.⁵⁹ The larger context of the dispute was the Helmholtz-Hering controversy in which Heß, who began his career as assistant to Ewald Hering, proponent of a nativist theory of vision, was strongly engaged.⁶⁰ Exner, on the other hand, was allied with physicist and physiologist Hermann von Helmholtz, proponent of an empiricist theory of vision.⁶¹ It is therefore conceivable that Frisch had a much older quarrel in mind when he turned to Heß’s experiments on color vision in fish.

Depending on where one locates the start of the disagreement, clearly different motivations become apparent. If we accept the version from Frisch’s autobiography, we encounter a model case of the critical-rational approach. If we keep in mind the previous exchange between Heß and Bauer, we are dealing with an escalating personal confrontation. But if we think of the dispute between Heß and Exner, a long-standing controversy seems to continue into the next generation. For most studies of controversies, the beginning of a controversy might be a purely formal problem. But if one is interested in what happened before the “official” beginning, it becomes clear that very different factors could have played into the later dispute. Each of the three possibilities mentioned above has a certain plausibility, and each implies a different idea of what was going on in the background when Heß and Frisch clashed publicly.

In an article published in 1922 for Emil Abderhalden’s Handbuch der biologischen Arbeitsmethoden—a prestigious undertaking that took two decades to complete—Frisch discussed in detail the various methods used to study the sensory abilities of bees. He opened the section on the sense of sight with a presentation of what is known as the “dressage” method, using his studies of the color sense of bees as an example. Frisch first explained how to attract bees to a foraging site, how to accustom them to feeding on colored paper there—on blue paper, in his example—in the midst of a series of gray papers, what precautions must be kept in mind, and how to proceed with the actual experiment (without food) after the learning phase was completed in order to eliminate sources of error.⁶² If all the specifications were followed, “the bees rush in droves to the blue paper and search the empty dish there for the usual food, while they pay little attention to the dishes on the gray papers; in this way, they show us that they distinguish the blue paper from all shades of gray.”⁶³ Frisch could hardly have described the results more impressively: the bees preferred the dressage color, although among the gray papers, according to Frisch’s premise, there had been at least one that resembled the blue paper in brightness. Consequently, at least one characteristic of the blue paper had to be different from the gray papers. For Frisch, there was no doubt that this characteristic was its color.

Heß had little sympathy for the dressage method in studies of color vision in animals. In another article for Abderhalden’s Handbuch a year before Frisch’s contribution, dealing with the “methods for the investigation of the sense of light and color,” Heß asserted: “In higher animals, the method can provide information within certain limits if used properly. Uncritical use of the method in lower animals has often led astray.”⁶⁴ The latter remark was aimed directly at Frisch’s experiments. The fact that Frisch took over the method from Heß, who had introduced it in the investigation of the color sense of fish, did not change Heß’s skeptical position.⁶⁵ Not only did Heß originally insist that feeding fish colored food and testing whether they could distinguish that color from any shade of gray was not dressage, further, such experiments were a mere makeshift alternative for Heß when he could not use his preferred method of the brightness equations.⁶⁶

Brightness equations were applicable only to phototactic animals like juvenile fish and bees, not to adult fish. Heß started from the observation of the Hering school that the relative brightness of various parts of the visible spectrum characteristically differ between a completely color-blind human eye and a normal human eye.⁶⁷ Based on this, he developed two key experiments.⁶⁸ In the first, a spectrum was projected onto the long side of a glass tank, and it was determined at which point the animals, previously kept in the dark, spontaneously gathered. For positive phototactic animals, this location denoted the point that appeared brightest to them; for negative phototactic animals, the location denoted the point that appeared darkest to them. The results were further refined by the second type of experiment, the actual brightness equations. For this purpose, the long side of the glass tank was divided into two halves of equal size, either with one half illuminated by a monochromatic light (for example, with a blue light of a certain wavelength) and the other by white light, or with both halves illuminated by monochromatic light (for example, one blue and one red). The intensity of the light sources was then changed until the fish gathered completely in one half of the glass tank or were evenly distributed throughout the space with no preference. From these experiments, Heß deduced the relative brightness of each color within the spectrum for the species under study. Then he compared the results with those for the normal and color-blind human eye and concluded that fish and bees were color blind.

The research of Heß and Frisch also differed in terms of their theoretical settings. Starting with his first studies on the color sense in birds, Heß referred to the characteristics of the human eye to determine whether or in what way an animal perceives colors.⁶⁹ Frisch, on the other hand, linked his results to ecological arguments: in the case of fish, he referred to the often striking colorations during the spawning season, in the case of bees to the relationship between flower colors and pollination by insects.⁷⁰ Heß and Frisch also integrated their results into contrasting evolutionary frameworks. Heß followed a linear model of evolution according to which color vision emerged only “with the transition from life [in water] to air” and here only with vertebrates.⁷¹ Accordingly, Heß understood color blindness in humans “as stagnation at a lower stage of development.”⁷² Frisch, in turn, saw in the fact that bees perceive colors, although “so different in their organization from a vertebrate,” an indication “that the foundations of the sense of color are the same here as there, and that the beginnings of its development may lie far back, in common ancestors of vertebrates and insects.”⁷³ Thus, in contrast to Heß, Frisch understood evolution not as the incremental development of ever more abilities along the “evolutionary tree” but as the varied unfolding of an early common disposition.

In their publications, Heß and Frisch almost never commented upon the methodological preferences and theoretical assumptions of their counterpart. Of course, they pointed out errors in the design and realization of the experiments and disputed the validity of the results, but only on one point did they enter more fundamental terrain. This point concerned the way in which Heß drew parallels between his studies of fish and bees and studies of the human eye. Frisch acknowledged the results of the brightness equations but rejected Heß’s conclusions: “That animals with such a sense of brightness as v. Hess found in fish and invertebrates must be totally color blind is a generalization of an empirical proposition gained from humans—a generalization whose justification is not proven by anything.”⁷⁴ Heß countered:

Even in this case, however, the argument remained on a technical level. Heß and Frisch argued about the criteria of a well-founded analogy, not about whether analogies are meaningful at all.

Ariane Dröscher’s account creates a different image of the discussion between Heß and Frisch:

Each of the juxtapositions has a foundation in the statements and activities of Heß and Frisch, but overall Dröscher’s account gives these statements and activities a new status: divergent positions and procedures reappear as “profoundly different approaches.” On one side the “experimentalist” Heß, on the other the “naturalist” Frisch; here “conditions required to display the desired reactions,” there “an environment that was as natural as possible”; here “measuring reactions,” there “biologically meaningful questions.”

It is not inconsequential when disagreements are treated as “manifestations or condensations of deeper contradictions of a theoretical, social, or other type.”⁷⁷ Such accounts harden actors’ ways of thinking and acting in certain situations into fixed attitudes that explain what happens. As such, Frisch’s preference for the dressage method and Heß’s preference for the determination of brightness equations each takes on the character of an epistemological credo. Accordingly, it must seem contradictory that Heß, although he considered the dressage method to be worthless for fish and bees, could claim to have made it “scientifically useful.”⁷⁸ Likewise, Frisch would have to reject Heß’s methods entirely. But in his handbook article, he described them as “very suitable” for the “investigation of the brightness sense of bees.”⁷⁹ Only for the investigation of color vision did he consider them unsuitable. Likewise, Heß and Frisch did not follow completely antagonistic research programs. As Kelle Dhein has pointed out, the difference was not “that one pursued reductionist explanations while the other did not; both Hess and Frisch sought such explanations. The key difference between Hess’s and Frisch’s experiments is rather the degree to which those experiments employed a holistic perspective on animal capacities as phenomena exhibited in particular contexts for adaptive purposes.”⁸⁰ On the whole, the confrontation between Heß and Frisch was probably not as clear-cut as one might expect after reading Dröscher’s interpretation. And indeed, one of the most intriguing points that emerged in the context of the debate between Heß and Frisch escapes precisely her account.

In his review article on the “Farbenlehre [Theory of Color],” Heß claimed that in his brightness equations animals “free of any constraint would exclusively follow their innate inclination to brightness,” while in the dressage experiments the animals would first have to learn something new—implying that learning was a less original and thus less natural situation in his eyes.⁸¹ With this idea, Heß silently responded to the criticism that the phototactic reactions of his experimental animals were merely a “‘laboratory product’” that did not occur “in these animals in a normal environment.”⁸² The distinction between artificial and natural is set in motion here such that laboratory situations can be thought of as more natural in the sense of being closer to an animal’s basic behavior than situations in which animals show learned responses in environments in which the animals can move (more) freely. Interestingly, Frisch’s contribution to Abderhalden’s Handbuch contains a parallel passage to Heß’s remark on the advantage of the brightness equations. After discussing the dressage method for bees and Heß’s methods in detail, the presentation takes a surprising turn. Up to this point, Frisch had expressed reservations about the validity of the methods only with regard to their application for certain research questions, but now he claimed quite fundamentally: “In the experimental setups discussed so far, the bees were observed under more or less abnormal conditions. One could be concerned whether, for example, the results of the dressage experiments allow conclusions to be drawn about the natural conditions during flower visitation, which are of primary interest in this context. One could demand experiments for control, which take better account of the natural conditions.”⁸³ Such experiments had already been carried out, but it proved difficult to isolate the relevant factors.⁸⁴ Dressage experiments therefore offered more explanatory power than experiments that came closer to “natural conditions.” Finally, at the end of the section, Frisch even distinguished between “natural dressage,” in which non-trained bees prefer the flower of a particular plant on their own, and “artificial dressage” as he practiced it.⁸⁵ As a result, a number of distinctions among artificial, normal, natural, and learned can be observed that escape the plain confrontation of the “experimentalist” with the “naturalist.” This is not to say that this confrontation is entirely wrong. But it narrows the discussion between Heß and Frisch to such an extent that we lose track of what are perhaps not the least interesting of the actors’ considerations.

In December 1922, Frisch, by then professor in Rostock, sent Christmas greetings to the Munich zoologist Richard Hertwig (1850–1937). Frisch had been Hertwig’s assistant between 1912 and 1920 and was soon to become Hertwig’s successor in 1925. In his letter, he reported on his research, his family, his children, and his holiday plans. Most of it, however, was about a well-known affair. “Recently,” Frisch wrote, “Heß published a paper in the Ergebnisse der Physiologie in which he almost surpasses the tone in which he attacked me (and others) earlier. I was particularly displeased by his way of exposing his opponent in a cheap way by misrepresenting his statements.”⁸⁶ Together with a number of other colleagues, Frisch planned to publish a statement in order to “take a firm stand against a procedure that is quite incompatible with the demands of literary loyalty.”⁸⁷ This was not the first time he complained about how Heß handled his work. In a 1914 article, the reader learns: “He [Heß] juxtaposes sentences from different works and from different places of one and the same work that now, however, seem to contradict each other.”⁸⁸ A year later, Frisch noted in his article on color vision in bees: “Where he [Heß] accuses me of ‘experimental errors,’ these are either incorrect assertions on his part or details that are completely irrelevant to the essence of the experiments. He declares all my decisive experiments to be incorrect. In this way, he uses negative results, which he has obtained with an essentially different experimental arrangement from mine, as ‘proof’ for the ‘incorrectness’ of my positive results.”⁸⁹

When Heß rejected the results of Frisch’s bee studies in 1918, Frisch responded with a detailed rejoinder, arguing that a reader who did not compare his work “page by page” with Heß’s account might get the impression “that I had made gross experimental errors and that my conclusions were not justified.”⁹⁰ The following year, Frisch returned to Heß’s style of argumentation. This time he resumed: “As much as I would like to respond to objective critique, I regret being confronted with a critic who disregards the main things with general phrases and distorts the facts by putting unessential things in the foreground and concealing the essential things.”⁹¹ Heß never responded to Frisch’s comments on his style of argumentation, and he only occasionally made similar accusations.⁹² Instead, Heß was bothered by something different. In July 1913, he handed over to Hertwig, from Institutsdirektor to Institutsdirektor, a memorandum concerning his dispute with Frisch. In order to understand the content, one must know that Heß, coming from Würzburg, had become director of the Munich Eye Clinic the year before.⁹³ Until then institutionally and spatially relatively far apart, Heß and Frisch now found themselves working at the same university. The new proximity was accompanied by considerable social and academic distance. The Ordinarius Heß, who had arrived at the peak of his career, was confronted with the Privatdozent Frisch, who had been habilitated only a few months earlier with his work on color vision in fish.⁹⁴ The memorandum consisted of a compilation of what Heß considered to be offensive passages from Frisch’s article “Sind die Fische farbenblind? [Are Fish Color Blind?]” published in November 1912, and a report of two encounters with Frisch at his institute during this period. About the second, which took place shortly after the publication of Frisch’s article, Heß provided a detailed account:

Heß insisted not on rules of scientific procedure, as Frisch did, but on rules of academic conduct. For him, scholarly exchange was synonymous with decent behavior, and this included that Frisch should disagree with the views of an established professor only with restraint, and preferably not at all in public. Heß thus framed the matter as a question of etiquette, which, because Frisch had continually disregarded it in his eyes, had to lead to the breaking off of “personal relations.”

For scholars concerned with controversies, these kinds of accusations point to a more or less informal set of rules by which controversies between scientists proceed: “Controversialists tend to annoy their opponents by evading their arguments, by writing incomprehensibly, by intentionally misunderstanding them, by insulting them, and by committing all kinds of fallacies. These complaints or accusations indicate a background of principles or maxims which seem to be accepted in a given scientific community, even if they are frequently violated. Such principles form an important part of what one could call the normative element of an implicit theory of controversy that people apply in their practice.”⁹⁶ The importance of rules is especially emphasized by those scholars who understand controversies to be not only “opposition of standpoints” but also a “dynamic linguistic exchange and a case of social interaction.”⁹⁷ From this perspective, a controversy does not proceed through new research alone. Also demanded is “general obedience to the norms of communicative cooperation and of the scientific search for truth, however differently interpreted by the contenders.”⁹⁸ With regard to the debate between Heß and Frisch, however, one can hardly say that they merely interpreted the norms differently. Of course, there was no lack of references to broken rules, but they belonged to different norm systems. Frisch had in mind the norms of good scientific practice in today’s terms. Heß, in turn, relied on the kind of masculine honor code reconstructed by Robert Nye.⁹⁹ In fact, Heß and Frisch spoke, as Heß had remarked, “two quite different languages.” Nevertheless, the exchange between them continued until Heß’s death. With this in mind, one might wonder whether shared rules are really a prerequisite for controversy. I am interested, however, in what scholars exclude when they hold shared rules to be indispensable.

The dispute between Heß and Frisch was more than heated. Heß once let his readers know that Frisch “did not recognize at all” what “was the subject of my investigations here.”¹⁰⁰ For him, Frisch’s research was generally characterized by a “lack of care and criticism.”¹⁰¹ Frisch’s argument that the striking colors of fish during the spawning season are indicative of color vision went, according to Heß, in an “unscientific direction.”¹⁰² Regarding Frisch’s findings confirmed by Hertwig, Heß remarked that their “incorrectness is easily and vividly noticeable to the attentive observer.”¹⁰³ Going further, Heß blamed all zoologists who positioned themselves against him for not having the “desire” to “familiarize themselves with even the rudiments” of physiological optics.¹⁰⁴ And to give a last example: Heß denigrated the already well-established view that the formation of flower colors is related to pollination by insects as “an instructive example of the suggestive effect inherent in traditional lines of thought even when they are demonstrably devoid of any scientific justification.”¹⁰⁵

For Frisch’s part, the best impression of the sarcasm with which he regarded his counterpart is given by his own account of the meeting with Heß in November 1912. In a letter to his mother, he reports:

Overall, what was generally noted about the debates between Heß and various further counterparts also applied to the interactions between Heß and Frisch: “There is no doubt that these mutual discussions sometimes took forms that all impartial parties found unpleasant and unnecessary.”¹⁰⁷

This kind of behavior must disturb scholars who, despite all the excitement, see controversies as rule-based deliberations. There is something uncontrolled and uncalculated, or in the words quoted above, “unpleasant” and “unnecessary” about these attacks. In contrast, the study of controversies assumes that a controversy can be fierce and even unfair but not imprudent and rampant: “controversies,” declares Marcelo Dascal, “are not anarchical.”¹⁰⁸ In fact, if historians, sociologists, and philosophers of science assume, depending on their point of view, that controversies are about advancing science, clarifying differences, or strengthening one’s own point of view, then they must expect these actors to behave efficiently and rationally according to their respective intentions. Consequently, they must rule out the possibility of scientists becoming entangled in the conflict in a wild, ill-considered, and “anarchic” way, while the issue at stake recedes into the background.

In September 1925, Frisch reported that in freshwater fish, analogous to the human eye, color perception disappears in the transition from day to twilight vision.¹⁰⁹ Fourteen years after the Basel lecture and two years after Heß’s death, this was the last contribution directly related to the questions discussed between Heß and Frisch. The intervening period was characterized by intensive research activity, with Heß and Frisch always struggling with each other. For example, when Frisch, together with his colleague Hans Kupelwieser (1879–1939), suggested color vision in Daphnia, small water crabs, Heß not only followed up with a critique of the work but presented corresponding negative studies by one of his students.¹¹⁰ Or when, after the war, Heß questioned Frisch’s entire experiments on color vision in bees because he discovered that bees perceive ultraviolet, Frisch showed that the training also worked when all ultraviolet was filtered out of the colors presented to the bees.¹¹¹

A similar dynamic characterized the means Heß and Frisch used to strengthen their arguments. Heß was the first to add photographs to his publications.¹¹² Frisch adopted this for his bee experiments,¹¹³ but Heß was already one step ahead and relied on cinematographic images,¹¹⁴ which Frisch, for his part, perfected after the war.¹¹⁵ When Frisch included excerpts from the experimental records in an article, Heß immediately followed his example.¹¹⁶ After Heß informed the readership that he had asked colleagues to make statements about the coloration of fish without prior instruction, reports of witnessing and blind judgment immediately appeared in Frisch’s publications as well.¹¹⁷ Overall, it can be said that despite the irreconcilable differences in content, the parties were inspired by each other’s approaches. A fierce debate can obviously go hand in hand with the development of converging strategies.¹¹⁸

The interested public was also eagerly courted. Most famous is Frisch’s demonstration at the 1914 meeting of the Deutsche Zoologische Gesellschaft in Freiburg.¹¹⁹ According to Franz Doflein (1873–1924), a former colleague of Frisch in Munich, everyone who saw the experiments went home with the impression that the discussion about color vision in fish and bees had been positively decided.¹²⁰ Probably attracting the most attention was a series of articles that appeared between 1914 and 1917 in the new flagship journal Die Naturwissenschaften.¹²¹ Finally, other researchers also began to address the issue of color vision in fish and invertebrates.¹²² Among the research in support of Frisch, the most important was the work of Alfred Kühn (1885–1968), who in the early 1920s achieved experimentally purer bee dressages by using spectral light sources instead of colored paper.¹²³

After an extremely condensed tour through the debate, one gets the impression of a constant back and forth driven by new experiments, new evidence, new arguments, performed epistemic virtues, and contributions from others. In the literature, we encounter such a sequence of events as a kind of standard model.¹²⁴ On this model, the controversy would resemble a game of ping-pong, where every ball is steadily returned until one player is defeated or both players lose interest. From this point of view, it is not necessary to explain why every action is followed by a counteraction: it is inherent in the matter, a reflex. And also left out is how the first action arises. Both elisions indicate that scholars understand controversies as basically inevitable.

Similar to the question of when a controversy begins, the question of why a controversy arises (not to be confounded with the constantly asked question of why scientists disagree) has received comparatively little scholarly attention. Strikingly often, historical accounts of controversies, be they from a so-called relativist or a so-called rationalist position, begin with the immediate introduction of actors and viewpoints, without addressing the circumstances under which the controversy originated.¹²⁵ In other words, the controversy is there from the start. More often, but rarely in more detail, articles dealing with controversies in general discuss the question of why a controversy arises. One idea is that a conflict must be taken sufficiently “seriously” by the research community for it to become a controversy.¹²⁶ Or the conflict must concern a matter “that is not resolvable by standard means of the discipline involved.”¹²⁷ Such considerations have the character of conditions, the fulfillment of which apparently must lead to controversy. Helga Nowotny has chosen a different path. In an article published in 1975, when interest in controversies was starting to grow, she noted:

This observation resembles Fleck’s remarks on the “tenacity of systems of opinion.”¹²⁹ A system of opinion—that is, an established scientific view—actively defends itself against “anything that contradicts it” by making contradiction seem “unthinkable,” by leaving everything that does not fit “unseen,” by concealing contradictory facts, by making “laborious efforts” to explain contradictory facts as non-contradictory, or by describing and presenting facts that correspond to the “current views” despite “the legitimate claims of contradictory views.”¹³⁰

With Fleck and Nowotny in mind, it seems as likely that divergent views on a scientific question will remain ignored as that they will lead to open conflict. For Nowotny, “it is only under exceptional circumstances that they become relevant and that the difference matters.”¹³¹ And even then, controversies can fail to appear. As Nowotny points out, incompatible views may be visible in retrospect or to third parties but are not expressed—“articulated”—or put into action—“activized”—by the involved scientists.¹³² The latter situation is the most interesting. Although scientists are aware of the incompatibility or, to put it more cautiously, the differences, they do not play them off against each other: “The lines of fission that separate the two sides can be as great as those that become visible during a controversy, yet neither side is willing or bothers to commit itself and to engage fully into a controversy. The difference does not matter.”¹³³ In these cases, Nowotny argues, the procedures of knowledge production are too different for a controversy to arise despite all the disagreement.¹³⁴ Scientists share the object of research (if we define “the object of research” broadly) but follow different aims and standards and should therefore not interfere with each other.

Nowotny’s point considerably sharpens the question of why controversy arises. To answer it, it is no longer sufficient to refer to fundamental incompatibilities and the importance of the matter in question. Instead of assuming that controversy arises automatically, it is now necessary to justify why an open conflict arises at all. Or, put the other way around, in need of justification is why conflicting views do not remain “non-controversies.”¹³⁵ This turn is not unproblematic because it leads into the field of counterfactual argumentation. Nevertheless, without speculation, indications suggest that a controversy is a possibility but not an inevitability.

If we return to the Basel lecture in June 1911, it is notable that Frisch initially took a modest position compared to the later course of events. He did not criticize Heß’s experiments; rather, he spoke of “finely thought out, precisely carried out comparative investigations.”¹³⁶ At the end of the lecture, Frisch also restricted the validity of his own observations to the species he had studied and pointed out that Heß had experimented primarily with juvenile and dark-adapted specimens. Such circumstances could explain discrepancies—even if Frisch already considered this “improbable.”¹³⁷ For him, the solution was rather that “fish are thought to have a sense of color that differs from that of humans in terms of the brightness in which the spectral colors appear.”¹³⁸ Heß himself had raised this possibility—albeit with little conviction—in his early paper on color vision in fish.¹³⁹ Later, in his first discussion of Frisch’s experiments, he distanced himself from this possibility, but his overall tone was less bluntly dismissive than top-down schoolmasterly. At one point Heß even remarked that “v. Frisch’s data should not be doubted”: possibly “individual differences” and “different behavior” at different times of the year would come into play with the minnows used.¹⁴⁰ And indeed, as it turned out, the capacity of minnows to adapt to the color of the ground depended on the origin of the specimens.¹⁴¹

From the fall of 1912, the relationship between Heß and Frisch shifted significantly. As we know, one reason could have been their very different expectations of how to behave in a situation of disagreement. A second reason would flash up years later in a popular lecture by Frisch in the winter of 1918. After introducing the studies of Heß, Frisch remarked: “If v. Hess had declared only worms or snails or millipedes to be totally color blind, his doctrine might have been accepted without lively opposition. But the assertion that also all insects and that among the vertebrates the fish are totally color blind had to arouse doubts with everyone who is familiar with the biology of these animals.”¹⁴² On the one hand, this reflection confirms the assumption that the path to controversy runs through the perceived weight of the issue under discussion. In the case of insects and fish, Frisch apparently felt that more evolutionary certainties were at stake than in the case of worms and so forth. On the other hand, it becomes clear that for Frisch the possibility of a non-controversy, to use Nowotny’s apt term, was by no means excluded.

Why it came to an open conflict between Heß and Frisch need not be explained here. My argument concerns the fact that in studies of controversies, the question of why, similar to the question of when, is essentially conceived as a technical one that can only teach us the level of urgency (however defined) at which scientists make dissenting positions explicit. Nowotny shows instead there is much more to the question of why. The idea that a controversy need not necessarily arise demands a detailed account of the circumstances that nevertheless does lead to it. Beyond that, a more open discussion of why controversies arise brings into focus all those Fleckian tenacities that in studies of controversies appear at best negatively as a retarding factor of the discussion. To investigate those tenacities would be one object of non-controversy research. Its starting point is the hypothesis that divergent views are often ignored by the participants or included in conventional neutralizing forms of contradiction such as the formulation “we have come to different results.”

Today, studies of controversies have lost their splendor. To be sure, accounts of controversies are still produced in history of science, HPS, and science studies. But for one thing, some controversies appear in a different light since we learned about “manufactured” controversies.¹⁴³ On the other hand, exhaustion is spreading among the ranks. A recent overview comes to the conclusion: “the ‘royal’ entry through the controversy has become so banal and routine that it is tiring due to its exuberant obviousness.”¹⁴⁴ Nevertheless, the study of controversies has not lost its prominent status, as demonstrated by the fact that through accounts of controversies students are supposed to learn how science works.¹⁴⁵ Transferred to didactics, the question arises all the more about which understanding of science is connected with controversy when it is used as a conceptual tool.

The modeling of conflicting views as controversies is accompanied by preferences and assumptions that largely determine what appears to be characteristic about science. Studying disagreements in this way means it is not the variety of initial motivations that is of interest but the end as the moment in which the nature of scientific thought and action and the gain of the debate should become clear. It means that positions are transformed into confrontations, with the result that observations that do not fit the scheme are filtered out. It means that even the most violent quarrel always takes place in a calculated way—denying that scientists, like everyone else, can quarrel simply for “unworthy” reasons. And it means that overt conflict must be inevitable once some conditions are met, rather than recognizing that tolerated or ignored inconsistencies are just as likely.

That controversy, like any conceptual tool, has certain properties that frame and limit what can be experienced is trivial. Yet this truism is rarely made explicit in studies of controversy, which tend to emphasize the advantages of the tool while overlooking its problems. To understand the prominent status of such studies, one must go back to the beginnings of extensive research into controversies in the 1970s. Scholars who engaged in the reconstruction of controversies emphasized that disputes between scientists were considered insignificant, if not disreputable, at the time. In traditional philosophy of science, controversies were seemingly treated as “something irrational, an embarrassment that really should not occur at all,” and an event that has “no significant effect on the end product of knowledge.”¹⁴⁶ The sociology of science in the wake of Robert Merton, on the other hand, is said to have understood disputes as the result of ideological clashes or struggles over scarce resources thought of as external to the actual scientific work.¹⁴⁷ Seen in this light, the career of the controversy as conceptual tool begins with a double rejection: controversies are not irrational, and they are not merely a contamination of science. On the contrary, they are rational in that they are rule-governed, calculated, and goal-oriented (only the rules and goals differ depending on the approach). And controversies are defined as the “‘natural state’ of science”¹⁴⁸ or, in the words of Bruno Latour, as the “most general situation.”¹⁴⁹ A supposedly marginal phenomenon was ennobled into a worthy object of reflection. Where scholars had previously suspected an indecent mess, the most understandable order actually prevailed; and where scholars had previously seen useless friction, they now encountered a helpful revelation. As a result, disagreements became domesticated: as controversies, they were stripped of everything idiosyncratic and wild.

Ultimately, the study of controversy has remained a faithful child of that philosophy and sociology of science whose ignorance of the value of controversies scholars once sought to correct. For studies of controversies, too, every conflict remains uncanny as long as it has not been “cleaned up”—keep in mind: “controversies are not anarchic.” Whether this applies in the case of Heß and Frisch, or whether the excessive mutual attacks did not perhaps even become the actual topic, could be assessed only if more informal sources on Heß’s views and his perception of the debate were available. With the exception of the complaint addressed to Hertwig, which is kept in Frisch’s papers, no such sources are known to date.¹⁵⁰ This circumstance also limits all other conclusions about the discussion between Heß and Frisch.

The question remains of how to think about divergent views and open disagreements beyond the controversy as a conceptual tool. A first step would be to analyze disagreements not according to whether they are rational or irrational, or whether they show the dominance of interests or the purity of scientific method, or whether they embody the nature of scientific thought and action or, on the contrary, represent an exception, or whether they generate or retard progress in knowledge. Having abandoned the established dichotomies, we can consider why some disagreements remain non-controversies for whatever reasons and why others become manifest. With regard to Heß and Frisch, one can then consider whether Frisch’s lack of respect alone provoked Heß into an open dispute and whether Frisch was motivated to deal with the matter so intensively solely by the fact that Heß did not stick to worms and snails in his research—or which other aspects were decisive for the fact that the divergences were not ignored. And one step further, when we discuss disagreements outside the conceptual framework of controversies, the focus then can be on how tenaciously and how fluidly the work of science unfolds, how strong the forces of persistence are, and how strong the forces of transformation. In the case of Heß and Frisch, we could then ask what element of their thought style in particular prevented them from accepting that fish and bees apparently behave as if colorblind in some situations and apparently perceive colors in others. In other words, if, as Kelle Dhein remarks, in “hindsight, the dispute largely resulted from both researchers’ failure to contextualize their findings,” that is, from the failure to realize that their results were sensitive to the respective experimental situation, we could try to explore under what conditions this “failure” was an expected, correct behavior for the actors at the time.¹⁵¹ I cannot deny that I hope studies of this kind can also help us better understand how science works.

First, I would like to give my deepest thanks to the von Frisch family for granting me exemplary access to Karl von Frisch’s papers. I am especially grateful to Prof. Dr. Herbert Zenker for helpful information on Carl von Heß and his personal and scientific papers. Kris Decker, Sandra Gratwohl, Verena Halsmayer, Eric Hounshell, two reviewers, and the editors of HSNS provided helpful comments on versions of the article at various stages of completion. In particular, I would like to thank Eric for editing the final English version. Early ideas for this article were discussed at the retreat of the Chair of Science Studies in September 2020. A preliminary version was written in April and May 2022 during a stay at the Villa Garbald in Castasegna. My cordial thanks go to Siska Willaert and Arnout Hostens for their warm hospitality on this and many other occasions. This article is dedicated in memory of Siska Willaert.