PRINCIPLES OF DISAGREEMENT, THE PRACTICAL CASE FOR EPISTEMIC SELF-TRUST, AND WHY THE TWO DON'T GET ALONG.

• Author: Barker, Simon

1. Introduction

   This paper explores tensions that arise between principles that require one to revise one's beliefs in the face of disagreement and the role of epistemic self-trust in our everyday intellectual lives.

   Broadly speaking, epistemologists working on disagreement are concerned with questions about how we should respond to disagreement. In particular, we might consider whether there are any substantive epistemic principles of the kind that require one to revise one's beliefs in certain general conditions or classes of disagreement. On a meta-level, we might consider also whether such principles conform to any more general normative structure. In sections 2.1 and 2.2, I survey the general terrain of the epistemology of disagreement, and especially the literature on peer disagreement, to address just this question. I argue that whatever else their content, all such principles are committed to proscribing, in the relevant circumstances, against one's continued reliance on the practices via which one came to hold the beliefs under dispute.

   In section 3, I turn to the topic of epistemic self-trust. Broadly speaking, epistemic self-trust can be understood as the combination of reliance on one's epistemic practices and some conscious positive attitude towards those practices. In 3.1, I develop what I call the 'practical case for self-trust', according to which practical limitations upon the possibility of establishing the reliability of one's epistemic practices push us to acknowledge the existence of a mode of epistemic self-trust without belief. In 3.2, I draw on the wider literature on trust to explore the idea that epistemic self-trust without belief is affective self-trust, characterised as one's having an attitude of optimism about one's reliance on one's own epistemic practices. I wrap up this discussion by considering what kinds of consideration will determine and influence the psychological availability of affective self-trust.

   In section 4, I put the discussions of the epistemology of disagreement and epistemic self-trust into contact to highlight significant tensions between revisionary principles of the kind discussed in section 2 and the psychological availability of affective self-trust. These tensions, I suggest, present a serious problem for accounts of disagreement committed to such principles.

2. Principles of disagreement

   2.1. Questions

   Disagreements are a ubiquitous feature of our social lives to which we can and do respond in a variety of ways: Sometimes, our response to disagreement is conciliatory--we lower our confidence in the disputed beliefs, suspend judgement on the issue, or even accede to the views of our interlocutors. Other times, we respond in more steadfast fashion, sticking to our beliefs despite realising that others have come to believe differently. Given the different ways in which we can and do respond to disagreement, the question arises: how should we respond to disagreement?

   Given the ubiquity of disagreement, never to revise one's beliefs in the face of disagreement would be tantamount to a thoroughgoing dogmatism. Presuming that one ought not to be dogmatic, then, the real question is not whether one ought ever to revise one's beliefs in the face of disagreement, but whether those cases that do generate such normative demands conform to any substantive epistemic principles of the kind that track general conditions and classes of disagreement.

   It is this latter question with which I engage in the first section. Rather than try to identify specific principles, however, I want to come at the question from above and consider whether such principles conform to any more general normative structure. As I shall argue, substantive revisionary principles may differ in terms of the conditions in which they apply, and the degree and method of revision prescribed. However, all such principles will be committed to proscribing, in the relevant conditions, against continued reliance on the epistemic practices via which the agent in question came to hold the beliefs under dispute.

   As the reader will no doubt be aware, epistemologists have, for the most part, considered questions about the normative significance of disagreement in respect to the class of disagreements between epistemic peers. Likewise, it is in discussion of peer disagreements that epistemologists have most explicitly posited substantive revisionary principles of disagreement. For that reason, this discussion will also be a useful starting point to consider questions about the general structure of such principles.

   2.2. From peers to principles

   To illustrate the idea of peer disagreement, consider the following example:

   CASE-1: Kazimira is a highly competent journalist working at a respected newspaper. She is currently investigating the possibility of electoral fraud in the recent general election. Throughout her investigation Kazimira has been in contact with her colleague Tom, who is also looking into the issue. Tom is also a highly competent journalist and Kazimira recognizes that. Throughout their inquiries, each has shared all their research with the other. At the next editorial meeting, Kazimira presents her research and her conclusion that there was electoral fraud. To her surprise, Tom--who clearly seems to be in full command of his cognitive faculties at the time--expresses his disagreement. In Tom's opinion, the evidence that he and Kazimira have collected does not support the conclusion that there was electoral fraud. As it happens, the evidence Tom and Kazimira collected supports Kazimira's conclusion, not Tom's, and Kazimira competently assessed the bearing of that evidence on the possibility of electoral fraud, Tom did not. Consider this case from Kazimira's perspective. Given the stipulations about how each of the pair has performed in their assessment of the shared evidence, there is no question here that, prior to Tom's exclamation of disagreement, Kazimira is rationally permitted to believe there was electoral fraud and Tom ought not to deny this. Despite that asymmetry, however, Kazimira enjoys no clear epistemic advantage of the kind that would allow her to settle the disagreement in her favour independently of the substance of the disagreement itself. Nor, for that matter, does Tom have any such advantage that would allow her to settle the disagreement in his favour. Moreover, this symmetry holds whether one's view is that the normative features of the case are determined by the objective facts about what the case-evidence supports and Tom and Kazimira's competence when it comes to assessing that evidence (as may be so on externalist accounts of disagreement) (1); the evidence that Kazimira has about their comparative competence and access to the case-evidence (as per evidentialist accounts) (2); or Kazimira's beliefs about the case-evidence and their comparative competence (as per accounts that lean on subjective conceptions of rationality). (3) Taken together, these features of the case have the result that Kazimira cannot permissibly both take a position on the question of who is more likely to be mistaken about the way in which the election was conducted and bracket from consideration of that issue the fact that she and Tom disagree.

   Whilst the literature offers a number of more precise characterizations of the peerhood relationship, I would suggest that, generally speaking, these can be understood as encapsulations of this feature of Case 1 cashed out in terms of the author's own preferred normative framework. (4) For the purpose of the current discussion, then, let's say that Case 1 is a case of peer disagreement, and, give a catch-all definition of epistemic peerhood as follows:

   Epistemic peers If S believes p, S realises that R disagrees with S about p and, independently of the substance of the disagreement, S ought not to believe that she has the relative advantage over R vis-a-vis p, or that R has the relative advantage over S vis-a-vis p, then R is S's epistemic peer. Accounts of the normative significance of disagreement under such conditions can be divided into three camps. Conciliationists argue that S should always revise her beliefs in the direction of her peers' (e.g. Christensen 2009, 2011, 2016, Elga 2007, Feldman 2007, 2009, Matheson 2009, 2015). Defenders of steadfast-ness argue that, under such conditions, S can be permitted to retain her original beliefs (e.g. Enoch 2010, Schafer 2015, Wedgwood 2010). (5) Those who advocate non-uniform views maintain that the appropriate response to disagreement under such conditions can vary between cases (e.g. Faulkner 2016, Feldman 2009, Kelly 2010, 2013, Lackey 2010,2013).

   Discussing the provenance of this divide, David Christensen keenly observes that:

   [i] All parties hold that the proper response to learning of another's disagreement depends on one's epistemic evaluation of that person. [...] [ii] The camps differ, though, on this question: In evaluating the epistemic credentials of an opinion expressed by someone who disagrees with me about a particular issue, may I make use of my own reasoning about this very issue? (Christensen 2011: 1. Numbering additional). Ithink Christensen's assessment of where the literature stands on [i] is quite correct. Since the motivation for this is obvious enough, let's label this idea credentials and move on to [ii]. Call the question Christensen raises here permissibility. I presume Christensen's reference to 'reasoning' in permissibility is shorthand for a much broader range of ways via which we might come to hold beliefs. After all, we do not arrive at all our beliefs via practices of reasoning. We also form beliefs via perception, intuition, testimony and so on. Nonetheless, I take it to be a truism that, for any belief one might come to hold, there will be at least one such practice via which one came to hold that belief (let's call such practices 'epistemic practices'). With that...