What Exactly is Presupposed by Agnotology? The Challenge of Intentions

ABSTRACTThe paper seeks to contribute to clarifying agnotology as an ‘epistemic strategy’, conceived as ‘epistemically damaging and hurt[ing] the production of knowledge’. My general claim is that the grammar of intentions ‘embedded’ in agnotological arguments is often not considered accurately. I use considerations from the philosophy of action as a theoretical framework to make more explicit what is implied in agnogenetic manoeuvres. Agnotology, as a ‘theory’ about epistemic states, in particular knowledge and ignorance, would be seriously incomplete without that component. The following can thus be read as a contribution to an analysis of the presuppositions of the strategic variant of Agnotology. My first claim is that the more common objections to the introduction of intentions are in no way definitive. My second, more specific, claim is that we need a room, in our conceptual toolbox, for ‘anti-epistemic intentions’, which play a key role in agnotological arguments.KEYWORDS: Normatively inappropriate dissentagnotologyintentionsaction (philosophy of) AcknowledgementsI am extremely grateful to Martin Carrier for his generous feedback on an earlier version of this paper.Disclosure StatementNo potential conflict of interest was reported by the author(s).Notes1 Proctor and Schiebinger (Citation2008), Introduction. The very first introduction of the term by Proctor referred to the ‘politics of knowledge’: ‘Ignorance […] has a distinct and changing political geography that is often an excellent indicator of the politics of knowledge. We need a political agnotology to complement our political epistemologies’(Proctor Citation1995, 8n.). In order to avoid ambiguities, I shall use agnotology and its variants to refer to theories about and explanations of the production of ignorance, and agnogenesis/agnogenetic to refer to the processes themselves.2 The term ‘Agnotology’, even to refer to intentional productions of ignorance, is not always used by the authors describing these issues.3 One could perfectly resist this claim. That was Popper’s stance on the ‘conspiracy theory of ignorance’, which was for him a consequence of the misleading view that truth was ‘manifest’ and that all exceptions to its manifestation had to be explained (Popper [Citation1963] Citation2002, 4).4 A temptation would be to say that ‘publicly available knowledge’ can work as a proxy, but it will not be enough. We can have agnogenetic manoeuvres even when the larger public is unaware of what is at stake. See Horel (Citation2015) for the fierce battle around the criteria used to assess endocrine disruptors in Europe.5 I have tried to provide some elements in Girel (Citation2022).6 For a more recent version of the argument, published as the present paper was under review, see De Melo Martin, in this volume.7 The discussion of ‘bad faith dissent’ belongs to a general examination of the possibility of finding necessary and sufficient criteria to identify NIDs, together with ‘Failing to Play by the Rules’, dealing with ‘shared standards, uptake and expertise’ (Chapter 4), and ‘Imposing Unfair Risks’, which is a discussion of the ‘inductive risk account’ and in particular the views offered by Biddle and Leuschner (Citation2015) (Chapter 5). The authors conclude that none of these resources will allow us to reliably identify NIDs and proceed to a study of ‘trust’ as an alternative (in particular in chapters 7, 8, and 9).8 On this, see Longino (Citation1990).9 Before publishing Proctor (Citation1995) and Proctor (Citation2011), Proctor gave a historical account of the genesis of the ideal of ‘value-free science’ and cannot be naive on this point (Proctor Citation1991).10 ‘Epistemology asks how knowledge can be uncovered and secured. Antiepistemology asks how knowledge can be covered and obscured’ (Galison Citation2004, 237).11 On this problem and the willingness of major scientists to address bad science as a kind of public service, see De Morgan (Citation1872).12 It would be interesting to explore the opposite notion of NID, Normatively Inappropriate Assent (or Acceptance), which is not thoroughly investigated in the book, if only for the sake of symmetry. The authors briefly envision this possibility in the notes, see De Melo-Martín and Intemann (Citation2018, 33 n.4). See also Carrier (Citation2019, 64) (‘Groundless strife and premature unanimity are symmetrical in their deceiving impact’).13 For a discussion of the possible consequences of applying too rigidly criteria, see Girel (Citation2020).14 Shakespeare, Hamlet, V, 2 (‘That I have shot my arrow o’er the house / And hurt my brother’)15 In reference to an episode where attorneys working for the Tobacco Industry tried to convince epidemiologists, for the sake of ‘Good Epidemiological Practices’, that relative risks of less than 2 may be artifactual for secondhand smoke studies; if epidemiologists had done so, passive smoking would have been dismissed as a cause of concern. This example is developed in Ong and Glantz (Citation2001).16 I thank an anonymous reviewer for his suggestions on that point and for the phrasing of this tentative definition.17 ‘Inductive risk, a term first used by Hempel (Citation1965), is the chance that one will be wrong in accepting (or rejecting) a scientific hypothesis’ Douglas (Citation2000).18 It is addressed in Chapter 5 of De Melo Martin and Intemann, but I would have the same kind of reservations I gave in the first section.19 Biddle and Leuschner (Citation2015, 277).20 Such studies did in fact violate standards, at least in cases that were widely discussed in France, see Foucart (Citation2010).21 One strength of this approach, compared to the impact-centred one, is that the quality of the motives, good or bad, will not change the agnotological nature of the process: ‘In the impact-centered approach, benefiting the public by breaking a methodological rule never qualifies as agnotological, whereas it may do so in the false-advertising account’ (Carrier Citation2018, 164).22 Perhaps even these examples could count as ‘false advertising’ since the actual author of the paper is not the one publishing and endorsing it.23 For an interesting collective volume, see Coady (Citation2006).24 I use conspiracy theories in the plain sense, as simply theories that assert the existence of a conspiracy (i.e., a group secretly pursuing a secret goal). In this sense, the official account for 9/11 is also a conspiracy theory, as already noted by Coady (Citation2006). Very often, though, ‘conspiracy theory’ has a richer meaning where the use of the term is associated with the insinuation of a paranoid element, and/or the assertion that proponents of the theory make circular judgements and immunise their theory against falsification. As a result, one might be deterred from using the notion of conspiracy in its plain sense in order to avoid ambiguities. My claim here is that even in the plain sense, the lessons that we can draw from actual conspiracies can help in defending a robust rationalism and in promoting some degree of epistemic vigilance. I have discussed this in Girel (Citation2017).25 See, for behaviour design, Fogg (Citation2003). In political theory and economics, see Thaler and Sunstein (Citation2008).