Evidence—long regarded as an important, but relatively mundane and uncontroversial ingredient in scientific knowledge production—has lately become a political rallying cry. When in 2017 and 2018, large crowds of citizens took to the streets in numerous places around the world to express their discontent with the new US administration led by Donald Trump, the political response to climate change, and the role of science in public discourse more generally, they were holding up signs with slogans such as “evidence, not arrogance,” “trust the evidence,” or “evidence trumps opinion.” Probably the most memorable chant heard during these Marches for Science doubled down on the procedural detail: “What do we want? Evidence-based policy! When do we want it? After peer review!”¹ Similarly, during the early days of the COVID-19 pandemic, “follow the evidence” was not only the main advice offered by the American Medical Association (AMA) and other medical professionals, but became almost a sort of mantra on the part of those advocating for more stringent public health policies.² However, while (scientific) evidence seems ubiquitous as a slogan, it appears more unclear than ever what actually counts as such, and how and by whom this is determined. News reports, evening talk shows, and not least social media are awash with alleged “evidence” of wildly varying kinds and quality: on the effectiveness (or harmfulness) of face masks or vaccines, the reality of human-made climate change, the legitimacy of elections and the occurrence of voter fraud, or even the existence of various kinds of global (and sometimes extraterrestrial) conspiracies.

Fierce arguments about the conclusiveness of specific pieces of (scientific) evidence are of course not a new phenomenon. In this regard, historical controversies on issues such as the smallpox vaccination in the late nineteenth century, the carcinogenic effects of cigarette smoke and certain chemicals in the 1950s and 1960s, or the use of atomic energy in the 1970s and 1980s show many similarities to—and direct continuities with—the recent “climate wars” over the extent and causes of global warming, or the conflicts surrounding COVID-19 policies.³ Given the ever-increasing importance of technoscientific expertise in contemporary “knowledge societies,”⁴ it is not surprising that the validity of scientific knowledge has itself become a central battleground in a growing number of cases. In the words of the sociologist Peter Weingart, the advancing “scientification of society” came at the price of a “politicization of science.” The intensified use of scientific knowledge as a legitimatory resource in political discourse tended to destabilize the impression of scientific neutrality and objectivity.⁵

This dynamic was one of the driving factors behind the perceived loss of public trust in scientific authorities, which numerous observers have deplored since the 1970s.⁶ With regard to the position of evidence, this had a twofold effect. On the one hand, the reference to verifiable evidence became increasingly important not only for the validation of knowledge, but also for the self-legitimation of experts and institutions, who attempted to replace waning public confidence with increased transparency and better mechanisms of control. As the “‘trust me’ society” started to turn into a “‘show me’ society”—as one prominent chief executive claimed—that which could be shown acquired additional prominence.⁷ On the other hand, the increased attention also served to highlight growing doubts about the concept of evidence itself. A variety of factors such as the growing complexity and hyperspecialization of contemporary technoscience, insights into the (socially) constructed nature of scientific knowledge, as well as the rise of an anti-intellectualist and populist political discourse insisting on “alternative facts” have all made the idea of one discernible scientific “truth”—and the notion of evidence as a seemingly objective arbiter—increasingly problematic and contentious.⁸

Both parts of this dynamic are visible in the rise of the so-called “evidence-based practice” movement, arguably the main driver behind the sudden ubiquity of the term “evidence.”⁹ Its roots lay in the appearance of evidence-based medicine (EBM) in the early 1990s, a self-proclaimed “new paradigm” for clinical practice that aimed to improve the efficiency of therapeutic interventions by linking them directly and systematically to the newest and most reliable research findings.¹⁰ Although it also inspired considerable resistance among clinical practitioners, the idea quickly spread to a large number of other professional fields. This resulted in the emergence of an astonishing number of new evidence-related subfields, such as evidence-based policy, evidence-based research, evidence-based education and social work, evidence-based management, evidence-based policing and prosecution, evidence-based design and architecture, evidence-based conservation and philanthropy, and many more.¹¹ In spite of some considerable differences, they all share the goal of greater efficiency and transparency, a propensity for quantitative and statistical measurements, the widespread use of formal procedures and guidelines, and the introduction of a “pyramid of evidence” formally categorizing research findings by quality and reliability. On the other hand, critics have often pointed to what they see as a tendency toward inflexibility and formalism, a failure to take practical and personal experiences into account, as well as a scientistic adherence to outdated ideas of a pure and objective science clearly separable from the realm of politics and application.¹² Nevertheless, evidence-based practice has significantly transformed both professional practice and scientific research in a number of fields—thanks not least to its complementarity with wider societal trends toward (financial) efficiency, transparency, risk management, and data-driven rationality.¹³

There seems to be a need, therefore, for a broader historical, sociological, and cultural approach to the phenomenon of evidence. This volume is an attempt to provide one such, by means of drawing together new perspectives and original research on the topic. Growing out of an international conference held in Munich in February 2020 organized by the DFG research group “Practicing Evidence—Evidencing Practice,”¹⁴ it reunites contributors from a wide variety of disciplines, working in very different fields of knowledge, on different historical periods, and in different geographical contexts. As they explore intersecting lines of inquiry, the chapters below are all concerned, in one way or another, with what counts (or does not count) as evidence in a given context. How do certain bodies of knowledge and certain practices of validating it become accepted in specific scientific fields or disciplines? In what ways are these practices influenced by their use as evidence—within the scientific community, but also in other contexts such as law, politics, or the media? How are scientific knowledge and scientific practices distinguished from other forms of knowledge production, and what connects them? What happens to practices of evidence when established knowledge is challenged?

While we cannot offer an easy solution to the problem of knowledge validation in the so-called “post-truth” society, nor an encompassing new epistemological model, we do believe that this book can not only generate new insights into the historical and contemporary dynamics of scientific knowledge production, but also insert some much-needed perspective into current political and academic debates. For this to work, however, we will first have to turn to a deceptively simple question: What do we actually mean when we talk about “evidence”?

In spite of its undisputed centrality for almost all scientific and scholarly endeavors, evidence is a shifting and elusive concept, which cannot easily be reduced to one agreed-upon definition or theory. In legal theory and epistemological philosophy—the two fields that have arguably devoted the most systematic attention to defining the term—evidence is usually understood along the lines of “a good reason to believe.”¹⁵ In legal cases, presenting evidence to the court plays an important role in reaching a verdict, while in science, empirical evidence confirms or invalidates scientific hypotheses. Within this basic conception, one can find distinctions between a number of approaches. Philosophers disagree, for instance, whether evidence is best understood in an objectual manner (as a material thing), or a phenomenal manner (as an experience); whether it should be seen as a mental state (of belief), or whether it consists of facts and propositions.¹⁶ Equally varied are the roles that evidence is thought to fulfill. Most authors emphasize that evidence not only acts as a “guide to truth” in the thought of a rational individual coming to a conclusion, but also holds an important intersubjective function as a “neutral arbiter,” to which one can appeal when trying to persuade others and settle disputes over conflicting claims.¹⁷

However, the notion of evidence as proof is not without its own inherent contradictions. Indeed, the Latin meaning of evidentia originally referred to “clearness” and “obviousness,” a quality possessed by something that is known to be true without requiring further proof.¹⁸ Later, this idea became bound to scholarly authority: in early scholastic writings, as Ian Hacking has argued, “things could count as evidence only insofar as they resembled the witness of observers and the authority of book.”¹⁹ In the Early Modern era, scientific truths started to be seen less as derived from the authorities of their authors, but from things itself; more precisely, from impersonal scientific instrumentation and material devices.²⁰ Evidence therefore became “a matter of inferring one thing from another thing,” defined as “one thing pointing beyond itself.”²¹ At least in its English-language form, this relational conception—in which evidence works as proof for or against something else—has come to dominate contemporary usage, while the notion of self-referential obviousness split off into the concept of “self-evidence.” Nevertheless, the older association lingers on in the often normative use of the term, causing a certain amount of conceptual confusion—particularly since the vocabulary split does not necessarily exist in other languages.²²

Given the newfound prominence of evidence as a sociopolitical issue today, in addition to its long-recognized centrality to (scientific) epistemology in general, one might assume that the phenomenon would also figure prominently in historical, sociological, and culturally informed studies of science. Curiously, this has so far not been the case. As late as the early 1990s, a prominent volume on the topic still found it “extraordinary how little direct attention” evidence had received in the history and theory of science.²³ This has only relatively recently started to change.²⁴ Most of the newer relevant work can be attributed to three major approaches and directions. First, as a highly specialized field of study, the study of evidence-based practice has been a growth area over the past decade. However, although topics such as random control trials in evidence-based medicine or mathematic models in evidence-based policy have attracted profound interest from scholars, they have rarely been examined from an interdisciplinary perspective.²⁵

Secondly, scholars from science and technology studies (STS) and media studies have highlighted how evidence practices as well as scientific evidence itself are inextricably linked to the medium in which they are communicated. Focusing on the epistemological functions of models, charts, tables, and images in science, as well as more broadly on narratives and numbers, has added an important dimension to the study of scientific “facts” and their evidentiary modes.²⁶ Contrary to the representational and purely documentary modes that these media are often assumed to operate through, this strand of research has illuminated how specific aesthetic and material qualities bestow credibility and authority beyond that granted through their content. While, for example, different forms of visualization mobilize certain ideas about validity, narratives work to create coherence between a variety of different pieces of evidence from different sources that otherwise do not appear to fit together.²⁷

Thirdly, the debate over trust in science and its importance for societal cohesion is having direct implications for the question of evidence. Political decisions not only have to be legitimized by scientific expertise; scientific research is also judged by its relevance to politics and society. Given the spread of scientific methods in the social, political, and economic fields, the related debate on scientification has shown the changing role of expert knowledge in science and society.²⁸ Moreover, as for example recent investigations into forensic cultures have shown, epistemic practices are deeply embedded in cultural contexts and political regimes. And—as encapsulated by the term “CSI effect”—they are often influenced by their own public image.²⁹ Several studies have explored the role of scientific expertise in politics and society with special attention given to experts, populism, and democracy. Many of these recently published works share an interest in how societies cope with decreasing trust in science and explore participatory formats of knowledge production.³⁰ A recent attempt to “democratize science” can be seen in citizen science, in which knowledge is coproduced. This again has important implications for the question of evidence.³¹

Notions of evidence are highly dependent on different chronological frames, cultural contexts, fields of application, and disciplinary cultures. Moreover, with its centrality to narratives of continuous scientific progress, the rise of evidence-based practice, and a polarized discourse of science denial and conspiracy theories, the term “evidence” has become normatively loaded and difficult to disentangle from political and cultural disputes. For these reasons, we propose to adopt a more ...