Tensions in agnotology: Normativity in the studies of commercially driven ignorance

Introduction

As scientific research moves increasingly to the private sector, the social organization of science undergoes important transformations. By describing different ways in which the commercialization of scientific research has encouraged the production of ignorance, agnotology has challenged the epistemic adequacy of such transformations. However, agnotology is not done in a normative vacuum. Scholars working on agnotology seem to hold implicit normative commitments that are in tension with their descriptive accounts of ignorance-constructive practices. By highlighting tensions in agnotology, this article begins an exploration into normative aspects of the studies of ignorance.

Strictly speaking, agnotology is the study of ignorance broadly conceived. Nevertheless, agnotology’s innovative contribution stems from its treatment of ignorance as a social construction, as opposed to the traditional treatment of ignorance as a natural vacuum. As Proctor (2008) claims, ‘The point is to question the naturalness of ignorance, its causes and its distribution’ (p. 3). Uncovering the pervasive effects of ignorance-constructive practices in science today, agnotology serves as a complement to studies of science that tend to focus on the production of scientific knowledge.

The study of the social construction of ignorance is constructivist in that it understands ignorance as a human-made product, emerging either passively from political, cultural or institutional frameworks, or actively manufactured to deceive others (Proctor, 2008). It is social in that it is not concerned with the individual knower or ignoramus and the conditions that contribute to his or her holding certain beliefs, but with the social arrangements that lead to a particular state of non-knowledge. Studying the social construction of ignorance thus contributes to social epistemology by uncovering the mechanisms that obstruct the production of knowledge within the scientific community, and the dissemination of knowledge in society at large.

Scholars from a diversity of disciplinary backgrounds have contributed to defining agnotology’s conceptual framework. Historians of science Proctor and Schiebinger (2008) have made a central contribution to this endeavor in their programmatic volume Agnotology: The Making and Unmaking of Ignorance. Earlier contributions to this framework include Smithson’s (1989) social theory of ignorance, Mills’ (1997) and Sullivan and Tuana’s (2007) epistemologies of ignorance, and Stocking and Holstein’s (1993) study of ignorance in the media. More recently, McGarity and Wagner (2008) have provided a detailed account of ignorance-constructive strategies in public health research.

Case studies of the social construction of ignorance can be divided into two broad categories: studies on the systemic production of ignorance – what Proctor (2008) calls ‘ignorance as a passive construct or selective choice’ (p. 6) – and studies on the deliberate manufacture of ignorance – or ‘ignorance as an active construct or strategic ploy’ (p. 8). The first category tends to focus on the institutional mechanisms that support the production of certain types of knowledge while suppressing others. For instance, Frickel et al. (2010) examine the role of social and civil movements in uncovering areas of research that have been left incomplete, unfunded, or ignored, focusing on the concept of undone science. Kleinman and Suryanarayanan (2012) study the case of Colony Collapse Disorder (CCD) in bees due to insecticide use, showing that the institutionalization of certain epistemic forms has led to different types of ignorance. Elliott (2013) examines an international report on agricultural research for the assessment of genetically modified crops and uncovers the production of ignorance through selective choices within the research process.

The second category, studies of the deliberate production of ignorance, is similarly various. For example, some studies focus on the identification of the actors involved and the strategies implemented in cases of ignorance construction. Prominent examples are studies on governmental and military secrecy. For instance, Galison (2010) uncovers the vastness of the US classification system, showing the shortcomings of governmental secrecy. Challenging the traditional view of secrecy as mere suppression of information, Balmer (2012) examines the history of chemical and biological warfare research in the United Kingdom during the Cold War, documenting how secrecy actively altered lines of research and policy making. Rappert (2012) uncovers the ignorance-constructive strategies implemented also by the UK government in counting civilian deaths during the 2003 invasion of Iraq. Kuchinskaya (2014) provides a detailed analysis of the formal representation of Chernobyl’s environmental hazards, showing how official government standards shaped the scope of environmental risks and made public hazards invisible.

This article is mainly concerned with studies of the ways in which private interests deliberately promote ignorance, in the context of the recent commercialization of scientific research. As scientific research moves increasingly into the private sector (National Science Board (NSB), 2014: 4–15), the threat that the social construction of ignorance poses for the production of scientific knowledge and the development of appropriate public policy increases as well. Here agnotology has proven to be an illuminating perspective into the social and epistemic consequences of current trends in science organization. Proctor’s (1995, 1999, 2011) work on the tobacco industry has been central to this endeavor, as has been Oreskes and Conway’s (2010) work on global warming. Epidemiologist Michaels (2008) also has provided an agnotological analysis of the chemical industry, and more recent agnotological studies in the history of the pharmaceutical industry (Nik-Khah, 2014) and the history of the 2008 economic crisis (Mirowski, 2013; Mirowski and Nik-Khah, 2013) make further contributions in this respect. ¹

Agnotology has contributed to achieving a more complex understanding of the mechanisms of ignorance-production, or agnogenesis, that operate in current commercialized science. The emphasis on scientific uncertainty, the creation and support of research organizations, the search for scientific allies, the public advertisement of industry-friendly research, and the attack on unfavorable scientific research are just some of the main ignorance-constructive strategies that private industry has implemented in a variety of cases and that scholars working on agnotology have helped to identify (e.g. Michaels, 2008; Oreskes and Conway, 2010; Proctor, 2011).

Agnotology opposes and complements epistemology, taking ignorance, instead of knowledge, as its main object of study. Ignorance, however, is not taken solely as an absence of knowledge. Instead, the study of the social construction of ignorance challenges the understanding of ‘ignorance’ and ‘knowledge’ as mutually exclusive, and encompasses, among other socially constructed types of ignorance, ignorance produced in the pursuit of certain kinds of knowledge (Sullivan, 2007), localized knowledge that contributes to increasing general ignorance (Proctor, 2011), and knowledge turned into ignorance as a result of political pressure and power imbalances (Kuchinskaya, 2014).

Despite encouraging a more complex understanding of ignorance and its relation to knowledge, most works on agnotology lack an explicit conceptual analysis of the terms ‘ignorance’ and ‘knowledge’. ² Instead, as I aim to show, agnotologists implicitly impose their own normative assessments of ‘ignorance’ and ‘knowledge’ in their interpretation of cases.

The main aim of this article is to examine implicit normative commitments in agnotology, focusing on studies of the social construction of ignorance in commercialized science. I examine recent works by Oreskes and Conway (2010), Michaels (2008), Proctor (2011) and Mirowski (2013). I show that these authors hold implicit normative commitments regarding what science is and, accordingly, how to distinguish ‘good’ from ‘bad’ science (i.e. how to separate scientific research contributing to agnogenesis from ‘good’ or legitimate scientific research). I argue that these normative commitments are in tension with their proponents’ descriptive accounts of ignorance-constructive practices.

Oreskes and Conway: trust the experts

In Merchants of Doubt, Oreskes and Conway (2010) examine the scientific research behind global warming as a case of agnogenesis. The authors show how the American public remained skeptical about the reality of anthropogenic climate change, even after its endorsement by the International Panel on Climate Change (IPCC) in 1995 (IPCC, 1995) and its further ratification by the scientific community (Oreskes and Conway, 2010: 169). They attribute this skepticism to the strategic use of ignorance-constructive tactics by a handful of scientists – Robert Jastrow, William Nierenberg, Frederick Seitz, and Fred Singer, among them – who put in place similar strategies to deceive the general public and to obstruct the production of scientific knowledge in a variety of public policy debates, ranging from tobacco to global warming, always arguing on behalf of free markets and against government regulation.

Merchants of Doubt is a book for a broad audience, and aims at explaining to the general public ‘why scientists would not only participate in challenging established scientific conclusions, but actively initiate it’ (Conway and Oreskes, 2011: 24). Accordingly, one should not expect an engagement with the philosophical debate regarding normative commitments. Despite this fact, the careful reader can also trace a series of normative claims throughout the book that articulate an underlying conception of science and an implicit criterion for separating ‘good’ from ‘bad’ research.

First of all, Oreskes and Conway (2010) locate the evaluation process of the scientific community, what they call ‘peer review’, at the core of the scientific enterprise: ‘It’s a foundational ethic of scientific work: no claim can be considered valid – not even potentially valid – until it has passed peer review’ (p. 3). Although they recognize that scientific knowledge should be understood in procedural rather than static terms (p. 16), the authors also present the approval from peers not only as the culmination of the whole process but also as the criterion for identifying ‘good’ research:

A conclusion becomes established not when a clever person proposes it, or even a group of people begin to discuss it, but when the jury of peers – the community of researchers – reviews the evidence and concludes that it is sufficient to accept the claim. (Oreskes and Conway, 2010: 33)

The idea is simple: no scientific claim can be considered legitimate until it has undergone critical scrutiny of other experts (Oreskes and Conway, 2010: 154).

In fact, Oreskes and Conway (2010) consider the peer review process as fundamental to science as experimental evidence: ‘Whatever the body of evidence is, both the idea and evidence to support it must be judged by a jury of one’s scientific peers. Until a claim passes that judgment – that peer review – it is only that, just a claim’ (p. 269).

Thanks to the constant evaluation of claims and counterclaims within the community of experts, scientists can and do achieve consensus on different topics, which for the authors is tantamount to achieving scientific knowledge: ‘Research produces evidence, which in time may settle the question… After that point, there are no “sides” … there is simply the consensus of expert opinion on that particular matter. That is what scientific knowledge is’ (Oreskes and Conway, 2010: 268). ³ In sum, Oreskes and Conway endorse a view of science in which the process of peer review, broadly conceived so as to include critical assessment by scientific peers and consensus formation within the scientific community, is central to scientific research. This is deemed to guarantee the proper evaluation of evidence and counterclaims, give legitimacy to scientific claims, and establish agreed-upon scientific knowledge.

Second, Oreskes and Conway (2010) suggest that scientists don’t have a real problem identifying ‘bad’ or dubious research: ‘Scientists are confident they know bad science when they see it. It’s science that is obviously fraudulent – when data have been invented, fudged, or manipulated’ (p. 153). They characterize many of the merchants of doubt’s efforts to counter scientific knowledge as cases of scientific fraud (such as manipulating evidence, hiding unfavorable data, cherry-picking favorable evidence) or cases of science imitation (such as creating industry-friendly peer-reviewed journals, funding industry-friendly conferences, and using non-expert scientists to support their claims). Oreskes and Conway compare the whole situation with the erection of a Potemkin village ‘populated, in only a few cases, with actual scientists’ (p. 245).

Despite the distinction they want to make between ‘good’ and ‘bad’ science, Oreskes and Conway (2010) acknowledge that ‘good’ science can be used illegitimately. For instance, when talking about R.J. Reynolds’ funding of scientific research, they argue: ‘All of the chosen studies addressed legitimate scientific questions … All the investigators were credentialed researchers at respected institutions. Some of the work they were doing was path breaking …’. Despite the apparent good quality of research, they identify a central problem: ‘But was the purpose simply to advance science? Not exactly’ (p. 12). As the authors claim, the idea of ‘fighting science with science’ was central to the tobacco industry: ‘Like the magician who waves his right hand to distract attention from what he is doing with his left, the tobacco industry would fund distracting research’ (p. 13). Focusing on ‘good’ scientific research in some areas undermined the advancement of science with regard to a pressing issue: understanding the health hazards of tobacco smoke.

The idea that perfectly adequate scientific research, that is, research that accurately follows the standards imposed by the scientific community, can be used as a distraction to other ends is central to the tension I want to highlight here. If what scientists consider ‘good’ science actually threatens scientific knowledge, broadly speaking, are they really capable of distinguishing legitimate from illegitimate scientific research? In other words, if scientific standards are not adequate to the task of identifying undesirable or problematic research, it seems that scientists do not have good enough parameters to evaluate their own research.

On the one hand, Oreskes and Conway (2010) are projecting a view of scientists as capable, given their expertise, of identifying and rejecting scientific malpractice. In fact, the authors state this view in their conclusion: ‘So it comes to this: we must trust our scientific experts on matters of science, because there isn’t a workable alternative’ (p. 272). On the other hand, they also acknowledge that there are cases in which what scientists consider ‘good’ research has served as a distraction for industry to obstruct the production of scientific knowledge. In other words, Oreskes and Conway are committed to the internal distinction between ‘good’ and ‘bad’ research, even though they have shown that ‘good’ scientific research can also contribute to illegitimate ends.

Oreskes and Conway (2010) do not solve this tension in the book, in part because they seek a broader impact. As they acknowledge, part of the reason ignorance-constructive practices are in place is that the general public has inadequate views of science and hence has unrealistic expectations about scientific knowledge (p. 34). This misinformation contributes immensely to the fostering of ignorance. By uncovering ignorance-constructive practices, Merchants of Doubt helps counter public misinformation about science.

In addition to the previous tension, Oreskes and Conway also seem to advocate a view of science in which political constraints should never divert the research process. ⁴ This is most obvious in the way they attempt to answer one of the fundamental questions of their book: why would renowned scientists, such as Jastrow or Nierenberg, undermine science?

Oreskes and Conway (2010) identify a specific commitment as the main factor behind the actions of their protagonists: ‘The issue was not free speech; it was free markets. It was the appropriate role of government in monitoring the marketplace. It was regulation’ (p. 248). They point to ‘free market fundamentalism’, borrowing the term from George Soros, as the political ideology that has corrupted the scientific process from the outside, turning against science when scientific results seem to counter its political goals: ‘the doubt-mongering campaigns we have followed were not about science. They were about the proper role of government, particularly in redressing market failures … The enemies of government regulation of the marketplace became the enemies of science’ (p. 262).

While not necessarily endorsing a value-free view of science, Oreskes and Conway portray science as independent from politics, in the particular sense of science being an institution that works best when isolated from the concerns of politicians and public policy issues. This is not to deny that political values can and many times do play a role in the scientific process, through the commitments of the scientists. ⁵ However, the scientific process should not be obstructed by political concerns from the outside. Thus, the underlying moral of Merchants of Doubt seems to be the following: let the scientists do their science, trust their scientific expertise, and reject any outside interference.

If my interpretation is correct, Oreskes and Conway face another tension here. On the one hand, they trust the scientific process and they criticize the input of political values from the outside. On the other hand, they have documented a number of examples in which ‘free market fundamentalism’ has pervaded the scientific process in ways that are difficult to identify (e.g. funding high-quality academic research, guiding favorable publications through peer review, creating research scholarships and fellowships, etc.). And so different questions arise: how is science supposed to remain isolated from any external political influence if, for instance, free market fundamentalism has already pervaded the scientific process? Is it really tenable to speak of the scientific process as ideal and to argue for its isolation from political issues, when ignorance-constructive practices are already in place?

In sum, Oreskes and Conway seem confident in the established mechanisms of scientific knowledge production (viz. peer review and scientific consensus), and they provide many compelling examples of the damage done by tampering with and obstructing those mechanisms. But let us keep the two tensions in mind: (T1) the tension between trust in the scientific process and cases in which ‘good’ science is used against itself, and (T2) the tension between isolating the scientific process and uncovering the pervasiveness of ignorance-constructive practices.

Michaels: science should be independent

An epidemiologist by training and Assistant Secretary of Labor for the US Occupational Safety and Health Administration (OSHA), Michaels (2008) has been directly involved in some of the public policy controversies that he addresses in Doubt is Their Product. In his book, he studies mechanisms of ignorance construction in different industrial settings, such as the tobacco industry, the chemical industry, and the pharmaceutical industry. His scientific and science policy background enriches the historical perspective common in agnotology.

Like Oreskes and Conway, Michaels (2008) aims to underline public misunderstandings of science that ignorance-constructive practices have created and maintained. In particular, Michaels attacks an outdated conception of science that the industry has promoted again and again, which is based on the idea that science provides the truth about the world and that it gives us certainty. Instead, ‘Scientists do not have the truth; we seek the truth. We deal not in absolute certainties but in the “weight of the evidence”’ (p. 60). Thus, Michaels considers uncertainty to be characteristic of science: ‘Absolute certainty in science is rarely an option; uncertainty is the norm, not the exception … Uncertainty does not mean the science is flawed’ (p. 165). Scientists achieve legitimate results despite the uncertainty that they constantly encounter in scientific practice: ‘It is entirely possible to draw a sound conclusion despite flaws or limitations in each and every test and study that constitute the evidence for that conclusion’ (p. 163).

Besides identifying uncertainty and the weight of evidence as the fundamental characteristics of scientific practice, Michaels (2008) also distinguishes between legitimate and illegitimate scientific research, according to the conflict of interests that scientists have or lack. ‘Good’ scientific research is, according to Michaels, research ‘carried out with true independence’ (p. 247). Central to his argument is a well-studied phenomenon concerning conflicts of interest, referred to as the ‘funding effect’: corporate funding of scientific research increases the likelihood of results favoring those interests (p. 143).

Part of the difficulty of understanding the mechanisms of the funding effect stems from the fact that most scientists hired by the industry are not committing overt scientific fraud: ‘close examination of the manufacturers’ studies showed that their quality was usually at least as good as, and often better than, studies that were not funded by Big Pharma’ (Michaels, 2008: 144). The funding effect thus has a broader scope than initially thought: even scientists following the appropriate scientific standards of their disciplines seem to be influenced by their corporate funding, affecting in turn the results of their research. Given the pervasiveness of this corporate bias, Michaels (2008) argues for a radical change:

I am convinced that conflict of interest cannot be ‘managed’. It must be eliminated. Too much is at stake. Data interpretation requires independent judgment; the public needs assurance that the opinions expressed in these settings are unbiased by commercial interests. (p. 256)

Given his experience in public policy, Michaels (2008) is well aware of the close relation that industry and universities have developed in recent decades, and how it has affected science (p. 243). In order to eliminate conflict of interest, thus preserving quality science, a series of governmental regulations should be instituted. Michaels (2008) dedicates one chapter of his book to explaining twelve suggestions to improve the regulatory system (pp. 241–65). In this sense, Michaels does not believe that the dynamics of science are enough to control for the funding effect and other negative influences from commercialization; not because he doubts that most scientists follow scientific standards appropriately, but because studies have shown that even scientists with the best intentions and qualifications skewed their research in ways favorable to their sponsors in almost imperceptible ways (e.g. Lundh et al., 2012). ⁶

Michaels’ analysis is less prone to T1, the first tension previously identified in the case of Merchants of Doubt, since he rejects the idea that the scientific process is enough to achieve legitimate scientific knowledge. He considers uncertainty to be fundamental to science and he praises scientists’ ‘weight of evidence’ approach. However, he knows that scientists can follow the standards of their disciplines – that is, they can do ‘good’ research – and still not produce legitimate results. In the end, what Michaels (2008) considers fundamental to the process is the independent characters of the researchers: ‘the best available science is science done by independent investigators’ (p. 247). Hence, Michaels argues for the scientific research process, but he does not grant this process the autonomy and self-corrective capacity that Oreskes and Conway grant to it. In this sense, his views escape T1.

In contrast, T2 is still present in Michaels’ work, because of his commitment to independent science, and thus the isolation of science from other social spheres. Although Michaels claims that science cannot operate independently without explicit regulatory measures, he seems confident that, once appropriate regulations are in place, science will no longer be vulnerable to ignorance-constructive practices. Regulation thus would prevent undesirable political influences in science. However, given the characteristics of commercialized science and the different ways in which the mechanisms of agnogenesis have pervaded the scientific process – ways that Michaels himself has helped uncover – one should remain at least skeptical about the impact of regulations over conflict of interest, and their capacity to reverse the political, academic, and social structures that are already in place (e.g. university and industry ties, strong intellectual property rights, fellowship systems, etc.). Removing conflict of interest through regulation might deal with specific cases of ‘good’ science being manipulated, but it is not clear how it will get rid of more entrenched institutional structures. Thus, the second tension remains present in Michaels’ work.

Both Doubt is Their Product and Merchants of Doubt point to a problem that arises when industrial investment in scientific research – which in many cases encourages ignorance-constructive practices – not only affects research outcomes and the public understanding of science, but also corrupts science from within. For Oreskes and Conway, the problem arises when political concerns try to shape science from the outside, whereas for Michaels the problem lies in the funding itself. In a way, these authors are concerned with science being independent. Although they consider the lack of independence (in terms of money, political values, or other factors) to be detrimental to science, they do not explain what true independent science would look like. Michaels (2008) suggests that science would be independent if regulations impeded the industry’s ability to affect the scientific process (p. 247), while Oreskes and Conway trust scientific consensus, granting the scientific process enough independence, and instead warn us about the ways in which this consensus is manipulated to shape the public understanding of science.

Proctor: science is political

Robert Proctor (1995, 1996, 1999, 2008, 2011) has dedicated much of his academic career to studying the tobacco industry’s campaign against the health hazards of smoking. The ‘tobacco strategy’ as it has been called, was initially designed and developed by the public relations firm Hill & Knowlton as a campaign for obstructing the production and dissemination of scientific research linking smoking to cancer and delaying regulation. As Proctor has documented, a big part of the strategy consisted in the tobacco industry’s financial and institutional support of industry-friendly scientific research. After more than 60 years of industry ties, tobacco companies are not only using research to their benefit, but shaping the path for the advancement of scientific knowledge:

Tobacco industry sponsorship has been corrosive of honest intellectual inquiry on a scale that is difficult to comprehend. Tobacco expertise for many years was dominated by the industry, as it still is in many parts of the world. Entire scientific societies have been formed to sabotage the science showing harms of one kind or another, and several scientific journals owe their existence to the industry’s need for ‘friendly research’ (Tobacco Science but also more scholarly organs like Indoor Air). The industry is not just corrupting academia; they are also creating it. (Proctor, 2011: 458)

The tobacco strategy proved to be highly successful: the industry did not lose any court cases until the 1990s, and the Food and Drug Administration (FDA) was not able to regulate tobacco as an addictive drug until 2009 (Oreskes and Conway, 2010: 33). In addition, the strategy was also used to counter scientific consensus in other debates regarding, for example, the ozone hole, acid rain, and global warming. The fact that the same strategy has been applied successfully to defy scientific findings in different fields makes the tobacco strategy crucial for the history of scientific ignorance in recent decades. The tobacco industry continues to be profitable at the cost of human health, and the practice of spreading ignorance has been key to achieving this goal.

Proctor is also more explicit than other authors about the philosophy of science behind his agnotology, even though he has not developed a full-fledged theoretical account. For instance, Proctor (1991) has given a historical account of the thesis of the value-freedom of science, one of the central normative questions regarding the nature of scientific knowledge. From the beginning, Proctor diverges from the more traditional accounts of science that authors such as Michaels or Oreskes and Conway seem to hold. He does so by explicitly advocating a broader view of science, one that understands science as fundamentally political:

Science is political in the sense that science has become a power to be reckoned with in the military and industrial strength of nations. Science is also political in the sense that there are always alternatives to the form and substance of its production and practice … Most importantly, science is political whenever the objects under investigation are matters of vital human interest – problems of health, security, and the various forms of privilege and exclusion which societies enjoy or from which they suffer … So long as the priorities of science are shaped by larger social priorities, the ideal of the ‘neutrality’ of science confronts the reality of the politics of knowledge. (Proctor, 1991: 267)

A political conception of science also comes along with a rejection of the value-free ideal that science should be neutral or impartial with respect to social and political value commitments. ⁷ Proctor (1991) considers the value-free ideal to have a particular political motivation in the historical context in which it developed: ‘The neutrality of science is not the consequence of a logical gulf between fact and value … It is a reaction to larger political movements …, attempts to isolate science from the sensitive questions of the day’ (p. 267). Moreover, Proctor highlights the problems of adopting the value-free ideal:

[Neutrality] serves also as well as a mask, disguising whatever interests may lie behind the origins and maintenance of research priorities. The ideal of value-neutrality may have consequences for the kinds of sciences actually pursued: a science that does not question the ends it serves (or potential sources of bias) may miss certain fruitful areas of research that do not serve those interests. (Proctor, 1991: 230)

As expected, Proctor’s rejection of the value-free ideal and his adoption of a political understanding of science raise the problem of the objectivity of science. If scientists must pay attention to the social and political goals of their research, if they have to keep in mind the political interests behind their research (corporate, environmental, governmental, etc.), how can they maintain their objectivity? In other words, how can they remain impartial enough not to skew the results of their work? Proctor (1991) has a tentative answer to this question: ‘Advocacy need not compromise objectivity. Objectivity is a matter of method; advocacy is a matter of commitment’ (p. 226). But of course, if Proctor wants to maintain objectivity as a fundamental value of scientific practice, he will need not only a better account of the distinction between methods and commitments, but he also needs to develop a more precise notion of scientific objectivity. As it will soon become clear, Proctor’s own research on agnotology shows that value commitments and scientific methodology are intertwined, creating a tension between his methodological view of objectivity and his research on ignorance-production.

Proctor’s agnotological concerns and his political understanding of science lead to an innovative criterion for distinguishing legitimate science. The cases of agnogenesis examined thus far uncover a crucial problem when identifying genuine research and differentiating it from illegitimate types: in trying to influence the public understanding of science, the industry’s ignorance-constructive practices have pervaded the scientific process, manipulating what we would consider ‘good’ science for their own interests:

The terrain here is different from what we sometimes imagine when we think of corrupt science. The key is not so much that the companies suppressed science (which they certainly did) … The genius of the industry was rather in using even ‘good’ science, narrowly defined, as a distraction. (Proctor, 2011: 255)

As this quote suggests, adopting a ‘narrow’ understanding of science obscures the problem at stake. Focusing on the particular methodological or epistemological standards of a scientific discipline would contribute to the evaluation of cases of scientific fraud and malpractice as ‘bad’ science. However, it would not contribute to understanding how scientific research that follows the standards – for example, goes through peer review – can also be contributing to the manufacture of ignorance. It is only when adopting the broader political conception of science that one is able to see the manipulation of ‘good’ science at the macro-level, and thus to question whether this ‘good’ science is legitimate after all. Even though Michaels and Oreskes and Conway have hinted at similar issues, Proctor (1995) addresses the conflict explicitly:

Bias in trade association science, though, is sometimes subtle. Bias in this context does not necessarily mean the science is phony: support even of ‘good science’ may assist an industry in deflecting attention from the hazards of a product … The net effect is to jam the scientific airwaves with true but trivial work, distracting from what is going on more fundamentally. Bias in such cases lies typically not in the falsification or misrepresentation of research (though both of these occur, as we have seen) but rather in the diversion of attention from one problem to another … (p. 131)

In Golden Holocaust, Proctor (2011) suggests a distinction between micro- and macro-biases. While industry-sponsored research might seem perfectly unbiased at the micro-level, when we look from the macro-level, the same research might turn out to be skewed in favor of particular interests: ‘It is not so much micro- as macro-bias – a bias visible not in any one study but rather in the aggregate collective body of research’ (p. 271). In commercialized science, where ignorance-constructive practices seem pervasive, it is no longer enough to appeal to scientific standards for distinguishing legitimate from illegitimate science. Or in other words, we need criteria operating at both levels. Proctor, however, does not tell us exactly what these criteria at the macro-level would be.

In addition to the acknowledgement of macro-biases, the political perspective allows Proctor (2011) to uncover another important obstacle stemming from agnogenesis. After decades of funding and supporting favorable research, the tobacco industry has influenced the scientific process in ways much more pervasive and damaging than is normally granted. As quoted earlier, ‘[t]he industry is not just corrupting academia; they are also creating it’ (p. 458).

At least two things should be noticed here. First, Proctor shows that industry funding has been epistemically detrimental to science: it has corrupted ‘honest intellectual inquiry’ and it has obstructed ‘popular and scientific knowledge’. Second and even more importantly, Proctor warns us about the industry’s capacity to change science from within. We are now watching the results of decades of industry involvement in cancer research: a reconstruction of what cancer research is from the inside, in terms of how and where it is done, who are the scientists involved, which professional societies and journals support it, which are the best schools, and so on. If the mechanisms of agnogenesis are as entrenched in the scientific process as Proctor indicates, then we cannot resort to the scientific process, as Oreskes and Conway suggest, as a normative guide to distinguish legitimate from illegitimate research.

Proctor’s analysis of the normative implications of agnotology contributes to clarifying some of the central problems that concern us. First, due to agnogenesis, distinguishing between ‘good’ and ‘bad’ science is not enough, but a criterion for distinguishing between legitimate and illegitimate research at the macro-level is also needed. And second, Proctor acknowledges that this problem arises from the industry’s influence in the internal aspects of the organization of science procured by the recent commercialization of scientific research. In this sense, commercialization becomes an unavoidable (political and economic) aspect of current science.

Proctor’s concerns here are also closely related to the tensions found in Oreskes and Conway. Proctor makes the distinction between micro- and macro-biases as an attempt to clarify and solve the tension that arises from the manipulation of perfectly ‘good’ research (T1). Instead of focusing exclusively on internal epistemic standards, we need to zoom out. In fact, research that complies with internal epistemic standards might contribute to creating epistemic obstacles at a broader scale, a tactic well known to the tobacco industry, which sponsored good-quality research at top universities on lines of research that diverted attention from the causal link between smoking and cancer. Here, Proctor also seems to distance himself from Michaels’ defense of independent science: adopting the political perspective entails that science is inevitably embedded in a broader socio-political and economic context, from which no true independence is possible.

Proctor’s analysis does not suffer from T2 either. Remember that T2 arises between a general trust in the scientific process and the pervasiveness of ignorance-constructive strategies. Of course, Proctor (1991) does not have an idealized conception of the scientific process: he rejects the value-free ideal as he advocates a political view of science. However, he aims to overcome the problem of preserving objectivity while rejecting the value-free ideal, by claiming that objectivity is connected only with methods, not value commitments (p. 226). However, Proctor’s research on the tobacco industry shows that ignorance-constructive practices have altered scientific standards, documenting that commercial interests have pervaded the most internal aspects of scientific research (e.g. experimental design, data collection and analysis, data interpretation, etc.). Thus, he has shown that methodological concerns are not separable from value commitments, for values in the form of commercial interests have successfully altered scientific research. In this way, Proctor has not completely overcome the traditional problem of preserving objectivity while acknowledging that science is value-laden. As a result, a tension arises (T3) between his view of objectivity as a matter of method, not commitment, and his research on agnotology, which exposes the consequences of ignorance-production in scientific methodology.

Mirowski: a historical approach to science

In August 2012, the Justice Department closed the criminal investigation of Goldman Sachs regarding its role in the 2008 financial crisis. As the public saw this financial giant walk away from prosecution, it became reasonable to expect that no top executives of any major US bank would face criminal charges for the financial collapse. Not just executives, but also economists appeared to have been caught by surprise by the crisis, and seemed totally unprepared to explain it (Krugman, 2009).

In Never Let a Serious Crisis go to Waste, Phil Mirowski (2013) presents an agnotological analysis of the 2008 financial crisis, uncovering some of the reasons why there has not been a serious effort by the economics profession to explain the crisis in theoretical terms, which in turn has made any attempt at prosecution pointless. In particular, he argues that different strategies to confuse the general public and delay any serious interpretation of the event have been implemented in the aftermath of the crisis:

The most important part of the history of the crisis that has been neglected … is that there have surfaced in the crisis some relatively systematic attempts to pump doubt and confusion into public discourse; in other words, some ‘explanations’ of manifestations of the crisis and its aftermath have been launched as trial balloons not expressly for purposes of further test and elaboration by sanctioned professional economists, but rather as calculated interventions in public discourse in order to buy time and frustrate any shared impressions of a few sharply delineated positions on a contentious issue. (Mirowski, 2013: 226)

Mirowski understands scientific knowledge and the problem of demarcating between ‘good’ and ‘bad’ scientific research as fundamentally historical. In this, his approach complements Proctor’s: of course science is political, but its present political landscape is the product of historical changes. Mirowski (2011) has given an account of these changes distinguishing three regimes of science organization in the 20th century (pp. 87–138). Within each regime, different conceptions of science and of the constitution of good scientific research have developed: ‘our claim is that alternative forms of funding and organization have shaped both conduct and content of science throughout its history, characterized by shifting alliances among the commercial corporation, the state, and the university’ (Mirowski and Sent, 2008: 673). Since the early 1980s, we have seen the emergence of a new regime of science organization, which Mirowski (2011) calls the globalized privatization regime, characterized by a series of structural transformations, such as stronger intellectual property rights, weaker antitrust laws, outsourcing and off-shoring of research, cheaper communication technologies, and so on (p. 25). Central to this new form of science organization is the politico-economic doctrine of neoliberalism or, as Mirowski calls it, the ‘neoliberal thought collective’. ⁸

So far, I have examined some of the normative commitments of scholars working on agnotology, but I have barely hinted at the normative commitments of the actors involved in cases of agnogenesis. ⁹ Mirowski’s historical analysis and his critique of the neoliberal thought collective will bring to the front important tensions between scholars exposing cases of agnogenesis, on the one hand, and the actors behind these cases, on the other hand.

At the core of the neoliberal thought collective is the idea that the market functions as an ‘ideal information processor’, making ‘every successful economy a knowledge economy’ (Mirowski, 2011: 29). ¹⁰ Building upon the ideas of Friedrich Hayek (1945), the neoliberal thought collective replaces the problem of resource allocation in the market (the central problem for Neoclassical economics) with the problem of knowledge or information distribution. ¹¹ The neoliberal idea of a knowledge economy begins with an understanding of knowledge as a commodity, tradable in the market. The quality of the knowledge is, however, left aside (ignored for theoretical purposes), leaving us with the idea that any accumulation of knowledge or information is always positive: ‘just as there are no “negative” prices, there is putatively no such thing as negative information’ (Mirowski, 2011: 318). Hence, the more information the market acquires, the better. There is no room for understanding a qualitative change in the acquisition of information.

It is precisely in this re-conceptualization of the market in epistemic terms, with the idea of knowledge as a commodity at the core, where Mirowski (2011) sees the main problem for the neoliberal doctrine: ‘The Achilles heel of neoliberalism is that it gets the functions of markets in society all wrong: Markets are not only limited and intermittently unreliable information processors; they can equally well be deployed to produce ignorance’ (p. 318). In this way, Mirowski understands the cases of ignorance construction uncovered by agnotology and fostered by the commercialization and privatization of science (and in particular by neoliberal trends like strong intellectual property rights) as important counterexamples to the neoliberal understanding of ‘the market’ as an information processor.

Even though neoliberals praise knowledge as ‘the chief good’, they also have an idealized view of the market as knowing better than any particular individual: ‘central to neoliberalism is a core conviction that the Market really does know better than any one of us what is good for ourselves and society, and that includes the optimal allocation of ignorance within the populace’ (Mirowski, 2011: 324). In this view, science, too, is subordinated to the market, and ignorance-constructive strategies stop looking detrimental to science because they can be understood as contributing to increasing knowledge in the market:

As long as scientists can be cajoled to defer to the Market to decide how knowledge will be subsidized, sorted, winnowed, and allocated, their resulting personal ignorance can only be eventually conducive to the public good, because it makes the system work more smoothly. (Mirowski, 2011: 327)

However, it is precisely this conception of knowledge as relative to the market that prevents neoliberalism from acknowledging ignorance-constructive practices that might be detrimental to scientific knowledge. If any piece of information is considered inherently ‘good’, then the cases of agnogenesis previously examined, such as the claims of climate skeptics, would be considered just another piece of information for the market to process. As long as the amount of available information keeps increasing, the ‘knowledge economy’ keeps growing. But now the mistake becomes obvious. When ignorance-constructive practices pervade the research process, as agnotology has shown, the ‘knowledge economy’ is not growing (unless one simply equates a priori the increase of knowledge with the growth of the market). Accordingly, Mirowski (2011) claims,

If the neoliberal reengineering of science has resulted in a vast ramping up of the production of ignorance … then it immediately follows that more expenditure on science does not necessarily result in more scientific output. Agnotology destroys the correlation. (p. 331)

I have shown two important aspects of Mirowski’s analysis. First, he identifies neoliberalism as the core doctrine behind agnogenesis in current commercialized science. This clarifies some of the normative commitments that the actors of agnogenesis (scientists such as Seitz and Singer) endorse, such as a relativist view of knowledge and an idealized view of the market. Second, Mirowski also provides a critique of neoliberalism that focuses on how neoliberalism fosters ignorance-production: understanding the market as an ideal information processor and the economy as a knowledge economy does not leave space for understanding the phenomenon of ignorance construction as detrimental to science, which is precisely what agnotology has shown us to be the case.

Because he is committed to a historical understanding of science, Mirowski builds his critique of neoliberalism from within; starting from neoliberal expectations about the economy and showing that they are self-defeating: ‘the fundamental crux of the modern economics of science is that … the current configuration of the commercialization of science is neither stable nor viable’ (Mirowski and Sent, 2008: 672). The historical perspective allows Mirowski to develop a critique of the neoliberals’ normative commitments: granting the normative commitments of the current regime, Mirowski is able to offer reasonable insights about its problems. If someone then objects to the underlying arrangements of commercialized science (together with the ignorance-constructive practices that it fosters) on the basis of normative commitments belonging to past understandings of science (e.g. that science should be apolitical, isolated, or independent from society, which are normative commitments proper to the Cold War regime of science organization according to Mirowski), he or she will be drawing an ahistorical and therefore irrelevant critique. For Mirowski, then, the historical character of normative commitments should be taken seriously if we want to understand their scope and limitations in a particular period.

Finally, Mirowski’s historical account leads to a new tension (T4) between the normative commitments of agnotologists examined so far and the normative commitments of the people involved in cases of agnogenesis. Granting Mirowski’s historical interpretation of the changes that science organization has undergone in the past century, and his claim that the current regime of commercialized science is characterized by neoliberal commitments concerning what ‘good’ science is, how should one understand Oreskes and Conway’s plea for a science free of political intervention or Michaels’ defense of independent science? Are they giving historically appropriate reasons to fight agnogenesis or are they just committed to normative views of the organization of science that are no longer current? Mirowski’s historical understanding of science and ignorance-productive practices seems in tension with the ahistorical normative views that agnotologists endorse. Despite this, Mirowski favors the agnotological critique of commercialized science. In this sense, Mirowski’s support of agnotology and agnotology’s account of the problems of current commercialized science are in tension with his historical view of science’s normative commitments.

Summary

Agnotologists are primarily committed to uncovering the ways in which ignorance-constructive practices have contributed to the public misunderstanding of science. The study of the social construction of ignorance has shown that it is mainly with the aim of shifting public discourse and public policy that industry advocates have engaged in these practices. But in that process, these studies also present a much deeper problem: in trying to confuse the public and delay legislation, the mechanisms of agnogenesis that the industry has put in place have changed in non-trivial ways the organization of science itself, affecting in turn the process of knowledge production.

If ignorance-constructive practices have transformed the scientific process, then how can one evaluate whether these changes are contributing to the production of scientific knowledge or whether they have been detrimental to this process? I have shown that different authors working in agnotology have implicit normative commitments that guide them in answering such question, and that unresolved tensions emerge from these commitments. In Oreskes and Conway, there is (T1) a trust in the scientific process and, at the same time, the acknowledgment that ‘good’ science is a tool for agnogenesis. In Oreskes and Conway, and also in Michaels, there is (T2) the claim that science should be conducted by independent investigators, but also the claim that ignorance-constructive strategies are pervasive. In Proctor, there is (T3) the view of objectivity as a matter of method, not commitment, and his research on agnotology, which exposes the consequences of ignorance-production in scientific methodology. And, finally, in Mirowski, there is (T4) agreement with agnotology’s denunciation of ignorance-constructive practices, alongside a historical account of science organization in which current scientific research coheres with a neoliberal understanding of scientific knowledge.

In general, these emerging tensions in agnotology show a broader tension between the authors’ descriptive aims – that is, to provide accurate descriptions of these practices – and their normative aims – that is, to argue that agnogenesis is detrimental to science. The latter presupposes a normative framework that agnotologists have not yet elaborated.

Conclusion

As agnotology has shown, the social construction of ignorance has achieved dangerous proportions in the current state of commercialized science. On the one hand, it threatens science’s epistemic goals by obstructing the mechanisms of scientific knowledge production (even by triggering personal attacks on scientists doing legitimate work), and also by channeling scientific research toward industry-friendly lines of research (e.g. by flooding conferences and academic programs with industry-friendly research), which in many cases don’t allow a proper understanding of the relevant phenomena. On the other hand, and most importantly, agnogenesis also obstructs a pertinent response to issues with a broad social, economic and environmental impact that affect us all. In fact, scholars working on cases of agnogenesis warn us against the social costs of these practices, emphasizing the dramatic consequences that they can bring, for example, massive starvation and migration due to global warming, life-threatening health risks, and unprecedented economic collapse affecting a majority of the population worldwide.

Despite its importance for agnotology, none of the authors provides a careful philosophical account of the normative commitments that seem crucial to counter ignorance-constructive practices. The present analysis also reveals an important lack of consensus among scholars regarding agnotology’s normative aspects. Different authors work with different normative views. While Oreskes and Conway and, to a certain extent, Michaels seem closer to the traditional normative understanding of science, according to which science should remain independent from social and political interests, Proctor and Mirowski offer us a more complex understanding of science as a social institution embedded in particular political and historical contexts. In addition, these implicit normative commitments create tensions with the authors’ descriptive aims.

As a result, agnotology poses a serious challenge. It uncovers significant epistemic and social problems emerging in current commercialized science, simultaneously making evident and failing to fulfill the need for a well-articulated normative approach capable of identifying and evaluating these problems.

Making explicit agnotology’s normative aims opens the discussion about the social and epistemic goals of science and their relation to agnotology. As the analysis in this article has shown, the identification of ignorance-constructive practices as detrimental is connected to the failure of scientific research to meet certain expected social or epistemic goals. For example, Oreskes and Conway’s critique of climate skeptics is tied to the obstruction of climate science’s social goals – in this case, informing public policy on climate change. Similarly, Proctor suggests that there should be standards for the evaluation of scientific research at the macro-level, in addition to the traditional epistemic standards at the micro-level, because the manipulation of scientific knowledge at the macro-level is obstructing science’s goals – in the tobacco case, not rendering the appropriate knowledge regarding the causal link between smoking and lung cancer. Accordingly, denunciation of these practices by agnotology would require a discussion about the epistemic and social goals of science, against which practices of ignorance construction can be measured. In this sense, agnotology could also expand its contributions to discussions regarding the proper roles of science in society (Collins and Evans, 2002; Kitcher, 2011; Kourany, 2010; Shrader-Frechette, 2007), the appropriateness of science for public policy (Elliott, 2011; Haller and Gerrie, 2007; Maasen and Weingart, 2005; Pielke, 2007), and debates regarding science policy (Biddle, 2014; Guston and Sarewitz, 2006). ¹²

To conclude, the specification of agnotology’s normative aims would not only contribute to building a stronger rationale for the studies of ignorance, but it would also strengthen agnotology’s contributions to other debates in the social studies of science.