Towards a Negative History of Science: The Unknown, Errors, Ignorance, and the “Pseudosciences”

Abstract

This article outlines elements of a negative history of science. For historians wishing to get a fuller picture of scientific practice both internally and externally, there is a lot to be gained by considering the dialectical constitution of scientific knowledge. To fully comprehend this relationality, historians should, therefore, trace the negative relations science maintains. Through oppositions, such as known/unknown; success/error; consideration/ignorance; and inclusion/exclusion, scientific knowledge emerges and disappears, and the social position of scientific practice is both established and contested. To exemplify our argument, we present four areas: the unknown, errors, ignorance, and the “pseudosciences”. Taken together, this approach allows us to understand how science constitutes itself epistemically and socially across different locations and historical periods.

1. Introduction

Do the boundaries of the history of science mirror the boundaries of science itself? Or does the former always transcend the latter?1 Questions such as these have been a part of the history of science since its inception as an independent discipline (Daston 2017). In what follows, we want to contribute to this ongoing discussion with the following hypothesis: to get a better picture of both the internal as well as the external workings of scientific practice, the history of science may gain a lot by considering the dialectical constitution of scientific knowledge. In calling this approach a “negative history of science”, we are not arguing that historians of science should position themselves negatively against their own discipline. Rather, we are offering a new perspective that looks at the negative relations scientists maintain and how they constitute their practice through them.

Drawing on different lines of research in the history of science, we will map out this “other” side of science. Four areas will be presented below: the unknown, errors, ignorance, and the “pseudosciences”. On the one hand, we shall direct our attention to social practices that interrupt or are perceived by scientists as dangers and setbacks to scientific knowledge production. On the other hand, we aim to show how science only functions in and through its relation to the unknown or errors.

In what follows, we argue that all these four areas refer to phenomena from which scientists or the scientific community at once seek to distinguish themselves while remaining closely bound up with them. The analysis of the unknown and errors draw our attention to localized laboratory or visual practices, such as map making, while looking at the production of ignorance and “pseudoscience” debates takes us to the interface between the scientific community and sources of power, such as institutional funding and public authority. Although scientists usually relate to these areas in negative terms, as something that has to be eliminated or excluded, this relation also carries important social and epistemic functions. Taken together, this approach reveals a dialectic movement that is central to the establishment of scientific practice across different locations and historical periods.2

2. Tackling the Unknown

At first glance, it seems obvious to identify the unknown with the opposite of knowledge.3 This may lead us to assume that what we do not know is necessarily located outside of the established scientific order. From an epistemological perspective, however, such an assumption is problematic because it obscures the ways in which the unknown has been and continues to be part of scientific knowledge production—rhetorically as well as practically.

For example, in Francis Bacon’s Novum Organum, the unknown figured as the frontier that knowledge was to conquer—famously visualized in the original frontispiece where a ship sets out to the great sea (Shapin 1996, p. 20). The colonial underpinnings of such a view of scientific advance are quite striking, and Bacon’s imbrications with colonialism have been the subject of some research (Zeitlin 2021). The historian of science Carolyn Merchant developed this point into a feminist and ecological critique of Bacon’s “new science” (Merchant 1980, 2008). As she argues, Bacon’s new experimental method that aimed to extract the “secrets of nature” was closely tied to the desire of extracting the “secrets of women” (Merchant 2008, p. 151).

Such a territorial understanding of the unknown as a terra incognita yet to be explored and conquered leads back to the history of early modern cartography and the problem of empty spaces (Relaño 2002; Laboulais-Lesage 2004; Hiatt 2008). However, according to Cornel Zwierlein, the unknown was not only limited to the empty spaces on maps that cartographers had to designate and visualize, but the period in general was characterized by a new and evolving drive to “define the unknown across the four fields of politics and economics, religion, general knowledge and history, and science” (Zwierlein 2016b, p. 2). Zwierlein’s approach thus shows how orders of knowledge, such as enlightened empires, were built on the processing of the unknown in different epistemic fields (Zwierlein 2016a).

While the frontier suggested a potentially limitless expansion of scientific knowledge, 19th century scientists increasingly underscored the epistemological limits of their practice. In 1872, the German physiologist Emil du Bois-Reymond famously proclaimed that certain “riddles” of the universe—the nature of matter and the relation of the brain to consciousness—would never be solved (du Bois-Reymond 1874). To ensure science’s continued progress, he further argued scientific practice had to be restricted to controlled measurement and calculation. This in turn entailed recognizing an absolute limit of knowledge, which science would never be able to surpass (Finkelstein 2013; Anacker and Moro 2016).4

With respect to 20th century scientific practice, the historian of science Hans-Jörg Rheinberger argues that the unknown is not just an external point of reference, but lies at the very heart of what he calls “experimental systems” (Knorr-Cetina 1991; Rheinberger 2005, 2010a, 2010b, 2021). Studying the historical emergence of the concept of the gene as an “epistemic object”, Rheinberger has shown that the productivity of this concept resulted from the fact that it organized experimental access to the organism and thus allowed acting in and through the unknown. Experimenting with the gene presupposed not knowing what it was. Once it became known, it was no longer an epistemic object but a technical one (Rheinberger 2000). To drive the main point home, experimental practice precedes knowledge. Rather than starting from the basis of the known to then build up ever more knowledge, experimental systems instead integrate the unknown “in such a way that it becomes the source of knowledge” (Rheinberger 2005, p. 81). For Rheinberger, therefore, the whole material practice of experimenting is centered around generating and increasing the possibility of unforeseen events rather than eliminating the unknown (Rheinberger 2010a, p. 149; Rheinberger 2021, p. 200).

By underlining the epistemic functions of the unknown, however, we should not fail to recognize that scientists may also engage in covering up its existence. Such an observation can be made concerning a well-known topic in the history of science, namely gene mapping. Genetic researchers—not least in order to gain financial resources—had to present genetic maps by arguing for their completeness, while, in truth, their mapping method always left gaps open. Consequently, the unknown was rejected or transformed into an empty formula that suggested knowledge where knowledge was actually missing. As the historian of technology David Gugerli noted, “[m]aps ask to be completed” (Gugerli 2004, p. 215).

Whether understood as a frontier, an absolute limit, or as an experimental tool for the production of scientific knowledge, scientists continuously relate to the unknown in different ways. In the examples we portrayed, the unknown did not figure as a threat or an obstacle to knowledge but rather served as the condition for science’s continued existence and historical development. By forming an epistemic horizon, the unknown gives science its direction. This mobilization of the unknown can have rhetorical dimensions, as in the case of Francis Bacon, as well as practical ones, as in the case of Rheinberger’s conceptualization of experimental systems. There is much to be gained, therefore, by taking a closer look at the ways in which the unknown forms the substructure of epistemic orders.

3. Dealing with Errors

The history of science is littered with past errors. As science evolves, sometimes through revolutionary means, what was once considered scientifically valid can turn out to be based on a cascade of errors; an observation that is equally true with respect to our current knowledge. Historians of science have directed their attention to errors in varying degrees. In the history of astronomy and mathematics, for example, errors have been a somewhat recurring topic of research for historians interested in observational errors and the mathematical solutions proposed in response to them (Sheynin 1977; Proverbio 1988).

However, errors are not only hindrances to knowledge and looking at them should not lead to the adoption of a whiggish perspective. As some examples drawn from the history of physics show, they are also integral to the process of scientific knowledge production. This is particularly visible within the research tradition of “New Experimentalism”. In 1983, the philosopher of science Ian Hacking described one of the great scientific breakthroughs of the 20th century, the discovery of cosmic microwave radiation, as resulting from the analysis of possible sources of instrumental error. Against this backdrop, Hacking then famously argued that experiment “may for long have a life of its own” (Hacking 1983, p. 160). In 1987, Peter Galison urged historians and philosophers of science to get rid of their theoretical aloofness and follow the historical actors into the material depths of their experimental labs (Galison 1987, p. 19).5 To understand the particular moment when experimenters considered the reality of an investigated phenomenon to have been convincingly established, historians needed to take account of the role dealing with errors or background noise played in this process: “In the laboratory the scientist wants to find local methods to eliminate or at least quantify backgrounds, to understand where the signal is being lost, and to correct systematic errors” (Galison 1987, p. 245).

The sociologist of science Karin Knorr-Cetina sees errors as part of the larger category of “liminal phenomena”. Instead of treating liminal phenomena, which also include uncertainties and imperfections, as marginal, Knorr-Cetina argues, the high energy physicists she studied at the CERN put them at the center of their attention, thus cultivating a form of “negative knowledge” which is “not nonknowledge, but knowledge of the limits of knowing, of the mistakes we make in trying to know, of the things that interfere with our knowing” (Knorr-Cetina 1999, p. 64). Because high energy physicists never directly see the objects they are studying—a fact Knorr-Cetina calls the “loss of the empirical”—they spend a lot of time figuring out if what their apparatuses are showing them refers to something really existing or if it is not simply the result of an instrumental artifact with no basis in reality (Knorr-Cetina 1999, p. 46).

Instead of reducing the experiment to an abstract practice of theory confirmation or refutation, looking at how scientists deal with errors reveals something about experimental procedures as a concrete historical practice. The history of experimental practice in particular can be written as a history of practices concerned with identifying and solving errors, an area which has been of particular interest to scholars working at the intersection of historical and philosophical approaches to the study of science (Mayo 1996; Elliott 2004; Schickore 2005; Buchwald and Franklin 2005; Boumans et al. 2016). More recently, a systematic approach to this topic was taken up in an edited volume that mapped out the epistemological functions of errors within the broader category of science “going amiss” (Hon et al. 2008).6

However, the study of errors need not be restricted to the study of experimental practice. According to the historian of science Lorraine Daston, the emergence of modern science can be explained by a shared “fear of errors” and the search for possibilities of counteracting them (Daston 2005, p. 4).7 To this day, the scientific community continues devising mechanisms that supposedly shield it from erroneous knowledge. The peer-review process is the most well-known example in this regard. Yet, while many consider this to be science’s self-correcting mechanism par excellence, it often fails to stop research from being published that later turns out to be faulty or difficult to reproduce.8 An awareness of this problem exposes that science’s error-correcting processes do not work correctly, and it has forced the scientific community to rethink its standards of publication and quality assurance.

4. Producing Ignorance

The role of ignorance in science can be approached from different angles (Wehling 2021). Ludwick Fleck and Thomas Kuhn pointed out that ignoring, selecting, and even rejecting certain facts were important to establishing the coherency and productivity of a given group of scientists (Fleck 1979; Kuhn 1962).9 In “Against Method”, Paul Feyerabend even argued that “ignorance […], far from impeding the forward march of knowledge may actually aid it” (Feyerabend 1993, p. 197). Studying how these types of ignorance, integral to scientific practice, are socially constructed was further developed in the 1980s by sociologists of science Michael Smithson and Jerome Ravetz (Smithson 1985, 1993; Ravetz 1986).10 While the just mentioned examples stressed the non-intentionality of ignorance as a necessary byproduct of scientific practice, we should also focus on intentional scientific ignorance. For example, scientists may block certain data for career reasons, choose to ignore scientific knowledge that may cause public harm or ignore data that contradicts one’s own argument.

In the past three decades, intentional forms of ignorance production were mainly investigated in connection with practices that strategically aim to hamper and weaken the status of scientific knowledge in society. The main approach to study such practices has come to be known as “agnotology”, a term developed in 1995 by the historians of science Robert N. Proctor and Iain Bol. In its most concise formulation, Proctor defined “agnotology” as “the study of ignorance making, lost and forgotten”. (Proctor and Schiebinger 2008, p. I)

Investigating how ignorance is made and sustained strategically was then taken up by a range of historians (Proctor and Schiebinger 2008; Proctor 1996; Schiebinger 2004; Galison 2004; Oreskes and Conway 2008, 2010; Kleinman and Suryanarayanan 2013; Henry 2017; McGoey 2019). In “Merchants of Doubt”, the most prominent work to have emerged from this field, Naomi Oreskes and Eric Conway outlined how four distinguished scientists adopted strategies that the tobacco industry had used since the 1960s to make people doubt the dangers of smoking. Backed by major conservative think tanks and the Reagan administration, these scientists also produced doubt regarding global warming, acid rain, and the ozone hole. By fostering such a false culture of doubt, they constructed themselves as defenders of open scientific discourse—they were “using science against science” (Oreskes and Conway 2010, p. 13)—and thus transformed doubt, which is normally considered as an important epistemic virtue, into its opposite, an epistemic vice (Cassam 2021).

Oreskes’ and Conway’s study reveals the often-hidden mechanisms of power that can shape science’s position and influence on public discourse and it also challenged some of the popular misconceptions of science. We are often led to assume that science provides us with absolute certainty. Yet, precisely this assumption allowed these so-called “merchants of doubt” to continuously question the scientific consensus surrounding the hazards of tobacco consumption. According to Oreskes and Conway, we should adopt a more nuanced image of science that acknowledges that science does not produce certainty, while it nonetheless produces consensus driven by standards and norms shared by the scientific community. We would then be in a better position to question which kind of doubt is based on genuine scientific interest and which one is driven by covert political and economic ones (Oreskes and Conway 2010, pp. 266–75).

Janet Kourany and Martin Carrier synthesized forms of scientific ignorance as well as the production of doubt under the more general header of ”Agnoepistemology” (Carrier and Kourany 2020b). In their understanding, ignorance encompasses both intentional as well as non-intentional and “passive” constructions of ignorance that scientists perform on a regular basis (Carrier and Kourany 2020a, p. 4). Arguing from a slightly more historical angle, Lukas Verburgt and Peter Burke recently proposed using ignorance as an umbrella term that points to the fact that the “knowledge society” may already be a thing of the past (Verburgt and Burke 2021; Verburgt 2020). Verburgt urged us to consider phenomena such as ignorance of the potentially negative effects of techno-scientific progress and of large-scale interventions into our natural environment as part of a joint history of ignorance and the Anthropocene. He then coined the term “agnotocene” to describe this current predicament. Although this approach may be read as a simple extension of Ulrich Beck’s conception of “reflexive modernity”, Verburgt departed from Beck’s modernist convictions, according to which ignorance and uncertainties can be turned into a “knowable risk” (Verburgt 2021, pp. 6–9; Beck 1986; Bauman 1991).

We agree that ignorance should not be considered simply as the absence of knowledge but using it as an overarching concept would not adequately capture the dialectical dynamic between science and its others that our contribution underscores. We also agree that it is important to devote attention to all practices of ignorance, including the selection and rejection of knowledge.11 However, it is equally important to maintain some differences. This concerns, for example, the deliberate manufacture of doubt. To be sure, doubt can also cause ignorance. However, what seems to distinguish the production of doubt from the other kinds of ignorance making, such as the rejection or outright censorship of knowledge, is that the former proceeds by producing a surplus of (dis)information that destabilizes the perception of scientific consensus.

Despite such reservations, looking at intentional as well as non-intentional practices of ignorance in science and the wider context of the (non-)knowledge society allows a precise description of the moments in which the production of scientific knowledge and its political, economic, and cultural impact is interrupted. Considering recent political upheavals, such as the Trump presidency; the rise of ultra-right-wing conservatism across Europe; as well as the COVID-19 pandemic, studying the intentional production of ignorance offers an important research perspective. In the age of “fake news” and the growing impact of conspiracy theories—consider the “doubts” raised against the safety of vaccines—the critique of politically and economically motivated doubt is a necessary political task.

5. Debating “Pseudosciences”

Calling an entire domain of research “pseudoscientific” results in a particularly strong form of doubt, for what has been labelled as “pseudoscientific” can be safely excluded from the scientific community. Looking at “pseudoscience” debates, therefore, offers an important extension to the concept of ignorance because it brings the question of the boundaries separating science from non-science into clearer focus. Studies exist on a diverse array of sciences that the majority of today’s scientific community considers “pseudoscientific” (Pigliucci and Boudry 2013). These include, among others, astrology (von Stuckrad 2007), phrenology (Cooter 1984; Poskett 2019), eugenics (Adams 1990; Levine 2017), creationism (Numbers 2006), and parapsychology (Mauskopf and McVaugh 1980; Collins and Pinch 1982; Oppenheim 1985; Wolffram 2009; Noakes 2019). However, how were these areas of research distinguished from accepted science? For what reasons were they excluded from the scientific community? What does this distinction do for scientists and historians of science?

Paralleling the emergence of the modern concept of “science”, the term “pseudoscience” began to attract wider currency in the 19th century (Thurs and Numbers 2011). In 1822, the French physiologist Francois Magendie used it to discredit the claims of phrenology to scientific status (Poskett 2019, p. 10). Although phrenology enjoyed quite a strong degree of popularity throughout the 19th century, it never achieved a firm place within the ranks of official science. Yet treating this episode merely as a pseudoscientific blunder, risks losing sight of the way in which phrenology captured many 19th century aspirations and sensibilities, such as a desire for a strict empiricism and the search for a material basis of racial, gender, and cultural inequalities—some of which continue to this day.12 A similar thing can be said about eugenics, a field of research which was imbricated and largely compatible with the scientific and political aspirations prevalent at the time of its emergence (Levine 2017, p. 23; Adams 1990, p. 219).

Should we therefore refrain from using the term “pseudoscience” altogether? If not, who exactly is in charge of labeling a doctrine a pseudoscientific one? One way of answering these questions has been to shift the burden onto the historical actors themselves. The historian, in other words, should not use the term “pseudoscience” as an explanatory category but rather as an object of historical analysis (Adams 1990, p. 220). In his history of creationism, the historian of science Ronald L. Numbers, argued that, instead of exposing the defenders of creationism as “pseudoscientists”, the task of the historian should be to trace “how persons and parties used ‘science’ and ‘pseudoscience’ to further their ends rather than in judging whether they employed these labels appropriately by the standards of the 1990s” (Numbers 2006, p. 14).

The more science consolidated itself into an “orthodoxy”, the more it became necessary to distinguish itself from rival practices by denouncing them as pseudoscientific (Thurs and Numbers 2011, p. 284). Against this backdrop, a number of historians have recently taken to understanding “pseudoscience” as a particular instance of “boundary-work”, a concept first developed in 1983 by the sociologist of science Thomas Gieryn to account for the ways in which scientists practically engage in demarcating their own work from other professional activities.13 “Pseudoscience” comes into play in specific historical circumstances when individual or groups of scientists perceive a doctrine as a threat to their own social status. This threat is especially felt when the doctrine and the proponents thereof aspire to achieve access to the same resources, such as state funding or official university recognition (Rupnow et al. 2008; Hagner 2008, p. 25). The historian of science, Michael Gordin, who specializes in the history of the “pseudosciences”, pointed out that studying “pseudoscience” debates reveals how “scientists thought about their standards, their position in society and their future”. (Gordin 2012, p. 3). Most recently, Gordin has, therefore, contended that the existence of “pseudosciences” is inevitable and the sign of a healthy and dynamic scientific culture. Understanding how the boundaries between science and “pseudosciences” are established and disputed can help us come to grips with those “pseudosciences” that cause actual harm (Gordin 2021, p. 101).

In this light, the history of the “pseudosciences” is inseparable from the history of science. Science continuously relies on the construction of another to sharpen its own boundaries. The study of pseudoscience debates reveals how scientists continuously struggle to retain their place within the economic, political, and cultural structures of power at work in society. Looked at from the opposite side of the fence, these debates allow the study of how scientists deal with social trials in which their authority is questioned.

6. Conclusions

Our contribution has presented the unknown, errors, ignorance, and the “pseudosciences” as possible areas of interest for what we have called a negative history of science.14 Apart from these there are others, such as precarious and threatened knowledge, which reveal similar dialectical dynamics. Nevertheless, we have chosen the above four because they provide us with a cross section through the history of science—from scientists exploring the unknown in the imperial age to scientists defending knowledge from being ignored in the age of disinformation. This cross section is a heuristic that is able to show both the tense relationship of science and its other and the historical dynamics of this field.

To different degrees, and at different moments in history, certain areas may become more important than others. Although we treated them separately for reasons of clarity, interrelations of course do exist. They bleed into each other and thus transcend a clear-cut distinction between an internalist or externalist understanding of science. For instance, errors or the existence of blank spaces of knowledge may just as much be ignored instead of becoming the center of attention. Yet scientists who choose to ignore or reject certain errors, which the majority of the scientific community considers as such, risk sliding off into the realm of “pseudoscience”.

We would like to close our contribution with a systematic consideration drawn from the above analysis. Our article has suggested that these four areas, which previous research has considered separately, can be brought into a closer relationship with each other. All fields refer to phenomena from which scientists at once seek to distinguish themselves while simultaneously maintaining a close connection to them. As long as scientists carry out their work, the unknown will be tackled, errors will be dealt with, ignorance will be produced and the “pseudosciences” debated. Taken together, this has revealed a dialectic movement lying at the core of science’s epistemic and social formation. Through oppositions, such as known/unknown; success/error; consideration/ignorance; and inclusion/exclusion, scientific knowledge emerges and disappears, and the social position of scientific practice is both established and contested. It is precisely this historical dimension of scientific practice that a negative history of science aims to capture. In this manner, science itself may provide some answers to the question of what it is—and it does so negatively, by confronting areas where it is not, cannot be, or does not want to be.