An Inductive Risk Account of the Ethics of Belief

Ðàçìåùåíî íà http://allbest.ru

An Inductive Risk Account of the Ethics of Belief

Guy Axtell

1. BRINGING RISK-MANAGEMENT FOR THE FORE: EPISTEMOLOGY AS THEORY OF INQUIRY

My approach to the ethics of belief from a focus on responsible risk management, where doxastic responsibility is understood in terms of the degree of the riskiness of agents' doxastic strategies, which is, in turn, most objectively measured through accordance or violation of inductive norms. With respect to beliefs, a person's rights and obligations, use deontological language, or virtues and vices, to use virtue theoretic language, are attributable to agents on the basis of how epistemically risky was the process or strategies of inquiry salient in the etiology of their belief or in the strategies by which they maintain or revise a belief already held.

From what norms does the ethics of belief derive its virtues and vices, its permissions and censure? Since pragmatists understand epistemology as the theory of inquiry, the paper will try to explain what the aims and tasks are for the ethics of belief, or project of guidance, which best fits with this understanding of epistemology. And it will try to support it with the work of William James and several contemporary pragmatists. Making the “doxastic strategies” of individual and collective agents central to the project of epistemic assessment results in a significantly different account than either the standard epistemological externalists focus on “processes” in the objectively reliable etiology of belief, or than the standard evidentialist focus on an agent's reflectively available “reasons” which lend the agent a certain kind of rationality, and subjective justification for her belief. This alternative would seem to require a more robust social epistemology than externalism or internalism, since strategies are valued as intelligent responses to problems for inquiry, and so focuses on agents' actual, intersubjective practices in pursuit of epistemic ends of true belief, knowledge, and understanding. While that would be far too large a project for the present occasion, we can at least try to explain how these two kinds of normativity, the projects of present assessment (the state or propositional attitude has toward a target proposition, and the standing of that attitude as warranted or meeting conditions for knowledge, etc.) and forward-looking guidance (praise and censure, improvement, etc.), are related.

This paper will continue the development of the claims that an ethics of belief for the present age, and for the future, needs to be risk-aware. It will develop the claim that the defense of permissivism is mightily improved by incorporating a principled account of the limits of responsible faith ventures. But it will develop especially the further claim that these limits come into view when we attend to the doxastic strategies which agents employ, and to agents' sensitivity or lack of sensitivity to the inductive risk they incur. This will show that the right to remain “steadfast” as the polar opposite of the equal-weight view is described as, is not the upshot of risk-aware permissivism, either. The defense of an agent's intellectual right to a blanket steadfast response seems to be what phenomenological conservativism, like religious apologetics, aims for: a situation where the higher-order evidences of peer disagreement and even of bias-mirroring in the acquisitions of my beliefs, can be safely set aside and thereafter ignored. Attention to inductive risk, I hope to show, challenges this policy as well.

The next two sections discuss psychological work on narrative aspects of personal identity. Section 2 discusses how doxastic risk-taking is often involved in the formation of identity. It finds much of this normal and healthy, but also motivates the need for risk-aware social epistemology, and therefore a permissivism able to offer a principled account of the limits of responsible faith ventures. Section 3 reviews and further develops my earlier argument (Axtell, 2011) that the evidentialist ethics of belief as articulated by Feldman and Conee has weak roots and sour fruits. It buttresses these claims by harnessing the work of Susanna Rinard and others on how pragmatic factors can serve as reasons for belief, and how we should understand the relationship between trait-dependent factors in the etiology of a belief -- factors such as pragmatic reasons, temperament, emotion and affect -- and epistemic normativity. Following this general ground clearing to make way for a more inquiry-focused foundation for the ethics of belief, sections 4-6 hone in on inductive risk, the risk of “getting it wrong” in an inductive context of inquiry. They describe some of the specific, empirically-informed markers / measures of inductive risk, and how they answer to our concern to determine limits to a person's reasonableness in their doxastic ventures.

2. IDENTITY FORMATION AND THE EMBRACE OF RISK

The value of identity construction might function as a defense of an opportunistic approach to risk-taking. This view recognizes that the value of doxastic risk-taking for the individual is bound up with self-experimentation, autonomy, personal identity, and group identity. For the project of epistemic assessment (and the “project of analysis”), epistemic risk and epistemic luck are closely connected. But moral risk, which the broader literature on inductive risk in policymaking directly involves, is not so nearly synonymous with moral luck. What this means for my approach is that while it makes the doxastic strategies of individual and collective agents central to the projects both of epistemic assessment and of guidance, it does not make guidance mirror or follow from assessment nearly as closely as it does for these other approaches. Some agents' beliefs or attitudes towards disagreeing may deserve censure for their riskiness, but especially in domains of controversial views, riskiness is not properly measured by the evidentialist notion of synchronic evidential fit, but rather by objective reliability or ability according to inductive norms. So, if indeed we want to make a fresh start we should avoid reducing the primarily diachronic norms that inform the ethics of belief from to the primarily synchronic norms that evidentialism touts; we must also avoid reducing guidance to staying in or returning to synchronic rationality: S having no attitude towards a target proposition p, or just that attitude or level of credence “supported by S evidence at time t”.

We should take note of two basic perspectives of risk, both one-sided, so ill-fit to rule alone. The one characterizes risk as an irresponsible choice; the other as an opportunity to be seized. The William Clifford -- William James debate reflects this contrast to some degree, so one place to look for it is this classic debate over the merits of the evidentialist principle, and over which domains of discourse, all, none, or some, this principle should be conceded as being normative for. Clifford asserted a social-moral duty to honestly “earn” one's right to their beliefs and / or credence, by pursuing rather than avoiding inquiry and evidence, and in attending to inquiry and evidence, by never believing anything more firmly than their present evidence logically supports. This principle exhibits a risk-averse attitude towards the beliefs and / or credences / acceptances to which one acquiesces: it sees doxastic irresponsibility as risk-taking, and as potential moral harm to others besides oneself who might be affected by it, and it makes no exceptions for domains where the censure is not equally fitting. James countered William Clifford's moral evidentialism primarily by contrasting. He countered it with a private right of each individual to be a “chooser of their own risk”, at least with respect to their faith ventures or those worldview beliefs he described as fulfilling the conditions of a “genuine option”. It is true that “No man is an island”, but it is equally true that humans neither do nor to be reasonable always ought to “Wait for the bell” of logically sufficient evidence on all matters of philosophical or religious metaphysics that matter to them.

James conceded to Clifford that by and large we humans need caution in our doxastic strategies more than we need “courage”. The council of courage he concedes leads to much credulity, and it would be question-begging to treat “chooser of their own risk” (italics added) as undercutting the impact of our beliefs upon others, which was basic to how Clifford motivates moral evidentialist argument. In his famous early paper “The Will to Believe” which answers Clifford directly, James tried to properly circumscribe the Lockean-Humean norm of proportioning one's beliefs to the evidence one has for it, by an exception case for choices that meet conditions of a “genuine option”. But this take of excusing conditions does not speak to dogmatism and fanaticism and the conflict and intolerance they breed, which were, of course, a shared concern of Locke, Hume, and Clifford. For this reason Jame' early way of circumscribing this right to be chooser (or manager) of risk through appeal to genuine options as exception cases, has not proved very convincing1.

In subsequent works, James does address credulity and intolerance, and in one of his last writing, “Faith and the Right to Believe”, James wrote much more directly that faith “must remain a practical and not a dogmatic attitude. It must go with toleration of other faiths, with the search for the most probable, and with the full consciousness of responsibilities and risks” (James, 1979: 13). This essay is still a robust defense of fideistic faith ventures and a criticism of both skeptical and religious evidentialism. But it offers a more agent-centered way of framing responsibilities, and a more balanced conception of personal intellectual right and “that spirit of inner freedom” with personal virtue and responsibility. Recent work on social us / them biases and on how we sometimes commit epistemic injustices towards others through the attitudes and beliefs we hold towards or about them only accentuates the need for such a properly qualified or circumscribed permissivist ethics of belief. Now more than ever we can see how agents are not always in the best position to judge their own rationality.

So, James' psychological insights and his ethics of belief provide good resources for the development of permissivism in regard to religious and philosophical “over-beliefs”. In recent papers and in Problems of Religious Luck: Assessing the Limits of Reasonable Religious Disagreement (Axtell, 2019). I have tried to develop the resources in James for a robust permissivism, and for a Dialogue model of science and religion, in contrast to Independence and Conflict models that portray science and religion either as complete strangers, or as enemies See Henry Jackman commutativityfor an argument that the exception case approach is inconsistent with and less strong than the more naturalized role with pragmatist philosophers ascribe to prudential reasoning within our cognitive economy (Jackman, 1999). I thank an anonymous Referee for suggesting clarification of the first draft's thesis, and several further, more specific suggestions which I try to incorporate below. See Axtell, 2013 for an explication of James' Dialogue model, and how it supports permissivism and “friendly” theist / atheist, etc. relations in contrast to polarized relations. At the outset of “Reflex Action and Theism” James proposes that “Among all the healthy symptoms that characterize this age, I know no sounder one than the eagerness which theologians show to assimilate results of science”. Interestingly, Nottelmann & Fessenbecker, 2019 seem to interpret Clifford as a critic of the Independence model, or at least Arnold's version of it which describes religious language as morality touched with emotion. Permissivism recognizes that the sources of persistent diversity over what John Rawls (1995) referred to a comprehensive conception of the good, are plentiful and often unavoidable. They include symbolic aspects of cultural identity, conditions of robust evidential ambiguity, the role of temperament, will, values and practical reasons, and the need for inference to the best explanation (the holistic nature of the evidence for worldviews and ideologies), what the other “burdens of judgment”I find it highly instructive that Rawl' holism agrees both with James and with theologians like Basil Mitchell, but that in explaining when disagreement can be reasonable, evidentialists largely ignore holistic evaluation along with temperament's influence.. But I have also aimed for what I call risk-aware permissivism, or “permissivism with teeth”, to contrasts its account of the epistemology of disagreement with that of toothless dogmatism or phenomenal conservativism. Recent work on social us / them biases and on how we sometimes commit epistemic injustices towards others through the attitudes and beliefs we hold towards or about them, only accentuates how proper caution and intellectual humility itself takes courage. The concern with hermeneutical and testimonial injustice does not come to an end because one's attitudes or beliefs are claimed to have a religious or theological or scriptural sanction. The study of epistemic injustice must pierce the veil of insulation from criticism for beliefs which “mirror” known social biases, or show other markers of high inductive risk. If epistemic injustices can be committed by people taking attitudes that are sexist or racist, then it stands to reason that epistemic injustices can be committed by people taking attitudes that are theologically absolutist or who engage in polarized and polemical religious apologetics.

James famously held that a person's “over-beliefs” guiding their practical judgments are the most interesting thing about a person, and that they embody the risks of different experiments of living. Jamesian meliorists might heartily agree with psychologists like Cynthia Lightfoot: “Risks are actively sought for their capacity to challenge, excite, and transform oneself and one's relationships with others. In this regard, risks are speculative, experimental, and oriented toward some uncertain and wished-for future” (Lightfoot, 1997: 2,³6ç)With more reference to James, Jennifer Welchman also writes in defense of an “ethics of self-experimentation” (Welchman, 2006). Psychological studies, especially those that notice the narrative dimensions of identity construction, support shared risk-taking as often promoting social bonding, and other healthy aspects of development. Kierkegaard could not agree more: “Without risk there is no faith”, Kierkegaard insists: “To venture causes anxiety, but not to venture is to lose one's self [...] The most common form of despair is not being who you are”. ethics doxastic epistemic faith

The two basic attitudes towards risk-taking, as irresponsible choice and as seizing the opportunity, are more obvious when speaking of actions than of beliefs. But they do have application in regard to theoretical as well as practical judgment, which we can best see if we don't get too stuck on the concept of “belief”, but consider it as standing in for a broader range of propositional attitudes or affirmations, especially in domains of controversial view: politics, religion, morals, and philosophy. Theologians who want to justify religions-specific “knowing” by faith, or to reduce faith to assent to a creed, and skeptical evidentialists like Feldman and Conee who endorse the “triad”model which reduces propositional attitudes to belief, disbelief, or suspension, each find a reason to reduce disagreement to dispute about belief. In both the religious and the skeptical case, disagreement now implies falsehood and the irrationality of someone (someone not me!). But epistemic assessment is not always about the standings (warranted, justified, etc.) of propositional attitudes assumed to be beliefs. The states themselves are attributed by assessors according to criteria, not just by what that agent or assessor wants to comfortably treat them as. An agent's mental state may exhibit tensions that suggest a self-deception on the part of the agent, or a sub-doxastic rather than doxastic state, or it may require refraining from ascribing or denying belief. If we recognize that our professed beliefs may be out of accord with our actions, or that there is a gulf between occurrent judgment and dispositional belief, then the attributional practices of epistemologists need to be sensitive to belief-behavior mismatches and to match in other ways their epistemology to the phenomenaSee Schwitzgebel, 2010, Leeuwen, 2017 for discussion..

Philosophers and social scientists have different kinds of interest in the concepts of risk. It seems to have an abundance of descriptive, explanatory, and normative roles to play in philosophy and science. For example, the riskiness of practical judgment, such as the decision to x, may be part of an attempt to objectively describe that judgment within the context in which agents individually or collectively make that decision. Riskiness now seen as an objective feature of an environment or domain, or as a characteristic of a choice to be made or a decision already made, might then also play a role in the explanation. The riskiness of a choice may be salient in explaining why the agent made it, or how she came to it. For instance, in Rebel Without a Cause, Jim (the James Dean character) and another young man decide to play chicken by racing their two cars to the sea cliff's edge. That “You've got to do something” is about all the explicit motivation it is given, but the riskiness of the action and the social context of male rivalry and female admirers, pride, and boredom seem all to be salient to its explanation. Risk tolerance and pride are certainly among factors one might site why these boys raced while others didn't, and for why Jim roles out of his car to save his life, rather than simply breaking while there was still time. His rival's misjudging of risk is of course also salient in his subsequent death.

Now it is vitally important to properly distinguish descriptive or psychological questions from normative ones. For example, especially with beliefs in domains of controversial views -- moral, philosophical, political -- we recognize “conviction” and “certitude” as terms that may be descriptive of a person's psychological state, and we distinguish such convictions from “certainty” as a normative concept in the epistemic sense of drawing upon norms of evidence and inference, i. e. , the certainty of a proof in mathematics, or of an observable event in the world for which we have abundant and incontrovertible evidence. Risk and luck, closely connected concepts, play important roles in epistemic assessment. The norms of epistemic assessment address such things as whether an agent has grounds for her belief, and whether the reasons she has are the actual reasons for which she believes. They address problems of luckily true belief, and preclude beliefs that are only luckily true, in one or another sense of malign epistemic luck, from qualifying as instances of knowledge. They address problems that an agent's justification, or the etiology of her belief, is not “connected” in a normatively appropriate way, with the truth of that belief. Safety and ability conditions on knowing, for example, are some of the ways epistemologists have sought to expunge beliefs acquired in a lucky or risky way, from counting as instances of knowledge. Epistemic luck is for these reasons a prime topic in the project of analysis of knowledge. Risk is less often the invoked term, but is here taken as closely overlapping.

3. WEAK ROOTS AND SOUR FRUITS, REVISITED

In a chapter of Trent Dougherty's collection Evidentialism and its Discontents (Evidentialism and its Discontents, 2014), focusing on Feldman and Conee's articulation of evidentialism, I argued that their ethics of belief has “weak roots and sour fruits”. The weak roots involve the defense of RUT, the Rational Uniqueness Thesis, and what domains it might be normative for. Thomas Kelly has argued, I think persuasively, that those who appeal to the principle in discussing the epistemological significance of disagreement frequently equivocate between a stronger and weaker claim, between two kinds of epistemic “slack”, interpersonal and intrapersonal. Sliding from the more moderate intrapersonal to the interpersonal gives RUT the appearance of being normative over peer disagreement. The stronger thesis needs to be made for evidentialism to enjoy the tight connection that they see between synchronic rationality and all-things-considered-guidance. But among Kelly's arguments for this is his construal of what is correct in the Jamesian view that belief does not have a single aim, truth, but rather a more pluralistic axiology where acquiring true beliefs and avoiding false ones are different epistemic values that potentially come apart. My own argument also focused on the inadequacy of the epistemic value monism or Feldman and Conee. I focused on their most explicit argument, where evidentialism as an ethics of belief is represented by this thesis (Conee & Feldman, 2004: 185, 258).

O2 An agent ought to always have just exactly that attitude towards a proposition supported by his /her evidence at that particular moment.

The support for O2 which they offer in their “Ethics of Belief” article and others is:

V3 Being synchronically rational at every moment is uniformly what it is to constitute epistemic success and to maximize epistemic value.

Certainly an ethics of belief should issue guidance that isn't just for an ideal-qua-atemporal agent; there is a place for more and less ideal-agent norms, but should not treat the actual agent in such an idealized way that the agent's ecological rationality is ignored in the way I take it to be in his influential evidentialist accountSometimes contingency or variability arguments are described as arguments from eviden-tially irrelevant causes of belief, or simply as debunking arguments, on assumption that they threaten to explain the etiology of these beliefs in a way that has nothing to do with their likelihood of being true. I doubt this approach, since I doubt the rational-social dichotomy on which it is based.. I think of an ethics of belief as all-things-considered guidance, but that is because I do not divide the diachronic and the pragmatic off from epistemology in the way that evi- dentialists do. So, discussing the relationship between epistemic assessment and guidance, as we are doing now, is an appropriate topic for us. To summarize the argument of Axtell (Axtell, 2011), V3 is false, and so does not support O2. That is aimed to undercut the tight connection of RUT to the ethics of belief that obscures the centrality of diachronic norms. This criticism of the centrality of synchronic rationality to the ethics of belief applies whether one accepts their “Triad view” that there are only the three propositional attitudes of belief, disbelief, or suspension, or whether one understood “attitude towards a proposition” in terms of credence and degrees of assent. In the same volume Keith DeRose challenges O2 with cases he presents as counter-examples, including this case: Henry has evidence e, and e supports p, and he believes p based on e. But if he had inquired responsibly, he would have had c that supports not-p.

Why does the sense in which Henry ought to believe p counts as epistemic, while the sense in which Henry ought not to believe p, does not? I argued that V3 is false by the epistemic value monism that it presupposes, and looking back at this argument I think that De Rose's, Kelly's and my own argument support each other's rejection of the evidentialist's reduction of epistemic obligation to synchronic evidential rationalityThis extends to Jason Baehr's article in the same collection, which also concerns the responsibility of the agent-inquirer, and the exclusion on it from epistemic relevance on the Feldman and Conee's account (Baehr, 2011). On criticism of O2, thanks to Kraig Martin for sending me his 2011 conference paper, “Can Epistemic Obligation be reduced to Synchronic Evidential Justification?” Trent Dougherty's exchanges with Martin and myself focus on this question, and on the claim that diachronic considerations are always moral-pragmatic and not epistemic. I call that one of the “dogmas” of internalist evidentialism, the other being that the ethics of belief issues guidance of the same synchronic sort. Pragmatists reject the rational-social dichotomy that informs this view, but through that rejection, the rational intelligibility of epistemic norms is actually greatly improved.. We have discussed the close connection between taking an externalist turn in epistemology and allowing those diachronic considerations are epistemically relevant, rather than necessarily pragmatic-moral, as they are for the evidentialist. How can they not be when the etiology of belief is itself a diachronic affair? To look at reliable or unreliable etiology is the same time to look back on a process or strategy that culminates over time in the actual agent's coming to hold a belief. In withholding epistemic status to diachronic concerns, evidentialists are only signaling that they have no way to integrate internalism and externalism. But this is due to inadequacies of their own approach.

But I also argued that synchronic rationality, as the focus of the assessment of the status of beliefs among internalist evidentialists, misses the externalist turn in epistemology and confuses its own project with the project of analysis. Really the evidentialist concern synchronic fit is a concern with a theory of rationality a kind of rationality that supplies personal, but not directly epistemic justification. There is a confounding of what are really two projects, and only on the basis of this conflation do Feldman and Connie evidentialism as an epistemological theory in the first place. Propositional and doxastic justification might be understood in different ways, but the externalist turn is a turn towards the reliable etiology of belief, and. So evidentialism has weak roots in this stronger sense as well, that the ex-ante approach that it takes by focusing on abstract propositions that might be held, rather than actually held beliefs and the processes or strategies that gave rise to them, makes their approach far less central to epistemic assessment than they have taken it to beSee Pritchard on the contours of a “two project” response to internalist evidentialism, and Richards Foley's various papers critiquing the “Unfortunate Assumption” as one well-argued support of separating analysis of knowledge and the internalist theory of justified belief. The problems for RUT and for evidentialism as a theory of knowledge are worse when the further distinction between instrumentalist and intrinsicalist defenses of evidentialism are added to discussion. Conee & Feldman, 2004 and others seem committed to an evidentialist principle being “truth's own ethics” but this intrinsicalist argument is again a hard one to make, and like Meylan, 2019 I do not think they make it..

4. AGAINST THE FACTS: DENYING PRAGMATIC FACTORS AS REASONS FOR BELIEF

Some evidentialists argue that if permissivism about evidence in any domain is possible, then non-evidential believing must be possible, yet it is not. Susanna Rinard and Miriam McCormick are contemporary pragmatists who argue that this modus tollens argument is unsound (Rinard, 2018a, McCormick, 2020). I agree with Rinard that people can believe for pragmatic reasons and that this assumption is crucial for properly distinguishing between and relating norms for guidance-giving and for epistemic assessment. Conscious non-evidential believing is not impossible; it happens all the time. I do not see how this is can be denied when (as I argue, every religious conception of faith has a fideistic minimum, and the more fideistic, the more that non-evidential belief is prescribed, and taken as paradigmatic of genuine religious faith. There are paradoxes in what I call “prescribed certitude” and of course in believing something as historical on a non-historical basis, but the impact of fideism needs acknowledging. But it is a difficult argument that faith-based belief is philosophically incoherent. Pragmatists want faith venturers and religious philosophers to acknowledge epistemic risk, not philosophers to treat self-consciously risky believing as impossible. Contrasting the evidentialist principle with the more pragmatism-friendly one, Rinard argues the norm of action / practical reasoning and of assertion having a relevantly similar structure, and so an “equal treatment” that acknowledges this resolves issues about the relationship between ethics and epistemology, in ethics of belief. “Insofar as we have control over these beliefs (be it direct or indirect), Equal Treatment acknowledges the moral dimension as highly relevant to the question of what we should believe” (Rinard, 2oig)See also Gerken, 2017; McCormick, 2020; Rinard, 2oi8a,b. I would argue that Susan Haack's “Overlap Model” (Haack, 1997) elaborating the best way to conceive the relationship between moral and epistemic factors in the ethics of belief is similar to Rinard's, and a good approach for the kind of risk-aware permissivism I develop..This is not in tension with but supports an approach to normativity which builds off of her call for a more careful distinction between the “guidance-giving” and the “epistemic” senses of “should”. Rinard shows that “Evidentialism about the guidance-giving” should is incompatible with Equal Treatments“But,” she still asks, “what about evidentialism about the epistemic should--or, for that matter, other theses about the epistemic should?” (Rinard, 2019). ““Some prominent evidentialists argue that practical considerations cannot be normative reasons for belief because they can't be motivating reasons for belief. I propose a new strategy for the pragmatist [response to evidentialism]. This involves conceding that belief in the absence of evidence is impossible. We then argue that evidence can ... play a role in bringing about belief without being a motivating reason for belief, thereby leaving room for practical considerations to serve as motivating reasons. These arguments push the debate between the evidentialist and the pragmatist into new territory. It is no longer enough for an evidentialist to insist that belief is impossible without evidence” (ibid.). She also develops advantages that Equal Treatment has over evidentialist alternatives, and develops an account on which non-evidential considerations sometimes serve as reasons for which one believes11.

Still another suggestion to counter tunnel-vision on synchronic evidential rationality would be to distinguish among epistemology's first, second, and third-personal tasks, and then to use these distinctions to help us more carefully explain the relation between epistemic assessment (the project of analysis of normative standings of knowing, justification, etc.) and guidance (what the agent all things considered ought to do or not do in their inquiry when the standing of a belief is substandard, or when their belief about others may entail an epistemic injustice). What thus becomes clear is that the ethics of belief has less to do with the third and first personal norms that epistemic assessment or analysis of knowing draws upon, but much more to do with second-personal normativity.

Second personal normativity, largely ignored in the ex-ante approach where an individual agent, a proposition, and evidence bearing on that proposition are basically all that matter, becomes more relevant when we instead take an ex-post approach where actual belief, and then explore its safe or unsafe etiology. Second personal normativity as it is discussed in metaethics concerns the intersubjective, rather than the subjective or the objective. It concerns the demands we make upon one another, how these demands get motivated, and how they are negotiated. These include demands for reasons-responsiveness, and people's commitment to abide by assertability conditions appropriate to the type of claim they make. On my proposed picture, epistemic assessment is a matter of meeting certain third-personal constraints on the riskiness of their belief-forming cognitive strategy, and certain first-personal constraints as well. One first-personal constraint that most epistemologists would agree must be met for knowledge possession is the agent consciously “basing” of her belief on the good reasons that she has available to her. This is the internalist demand that an agent who knows that x does not just “have” sufficiently good epistemic reasons for some target proposition x, but that the agent's belief is based on those reasons. Recently, this has been articulated as a demand to preclude another malign form of epistemic luck, what Bondy and Pritchard term propositional epistemic luck (Bondy & Pritchard, 2018).See also Pritchard, 2016. I take an initial foray into the impact of propositional religious luck on religious epistemics. If I believe that x and have sufficient reasons supporting x but base my belief not on them but for some reasons or out of some other irrelevant causal factors, what I believe, even though true, will not qualify as an instance of knowing. I will be deluded about my actual reasons for belief, or aware of them but not of their rational insufficiency to ground my belief. So, there seems to be a special connection between believing for pragmatic reasons and propositional luck I hope to explore in future papers, a connection especially problematic where agents convince themselves post hoc of the logical sufficiency of their evidences. The third-personal constraints on knowledge are seen in different ways, but basically refer us to an externalist condition (typically an aretaic and /or a modal safety condition), aimed at assuring the reliable etiology of the agent's belief. Staying with the language of epistemic luck, knowledge requires that the etiology of one's belief is not an etiology impacted by malign epistemic luck, as we find for instance, in Gettier cases: this externalist kind of malign luck is termed veritic luck in the literature. There is nothing new in the foregoing, which might just be taken as simply describing a “mixed” analysis of knowledge that has both an Internalist (doxastic justification condition or anti-propositional luck condition) and an externalist condition (a condition to preclude Gettierization). But perhaps it opens us up to see that epistemic assessment has little to do with second-personal constraints, while agent rationality and the ethics of belief draw far more heavily from it than they do from the first or third personal concerns. Axtell (Axtell, 2019), chapter 1 treats propositional religious luck as one of the six types in my taxonomy, and suggests why its recognition might have strong implications for religious epistemics.

We have seen that following Feldman and Conee's V3 to their O2 conflates the ethics of belief with what the evidentialist takes as epistemic assessment, pragmatism and virtue epistemology do not have this problem. This of course is why they not only have stronger epistemological roots but also better practical fruits: why they are able to as evidentialists are not to defend the possibility of reasonable disagreements, and to support Rawlsian reasonable pluralism.

What about modified versions of evidential such as Scott Akin and Robert Talisse, or the principled agnosticism that J. Adam Carter argues for? The evidentialists who accept the obvious fact that we acquire our evidences chronologically, seem driven to “split the difference” with each new disagreeing agent one comes across. But this is a criticism of the literature on the equal weight view though and the evidentialism that motivates it, not on Aikin or Feldman specificallyInterestingly, Aiken and Talisse have made an excellent critique of the equal weight view, leading to absurdities. The problem is that it's not clear in their work what other alternatives there are consistent with their belief in the evidentialist principle as being normative over the ethics of belief. The alternative to recognizing the naturally heavier emphasis one puts on first -- acquired evidence, is to accept an evidentialism rooted in something like the commutativity principle. I am sure this ideal agent principle has its uses in decision theory, but to have it be a source of blame I would argue again asks too much for the kind of doxastic responsibility that virtue requires. Even if one accepts commutativity as guiding evidential fit for the purposes of epistemic assessment, what comes out as fitting is whether an agent “should” believe to be synchronically rational will depend on further assumptions the attributor makes unless evidence is all of just the same kind. This clouds the picture for the reason that it brings us back to the need for holistic valuation, and not by the agent, but by the attributor of knowledge as well. But that holistic reasoning, the need for which arises for inquirers especially under conditions of local or chronic underdetermination, will necessarily bring temperament and legitimate practical interests back into the picture. And this is what, to protect the purity of epistemology, most evidentialists seem unwilling to abide. This is why often say that the issues of pragmatic encroachment are somewhat misstated. Often the deeper issue is diachronic encroachment on to an artificial purity where only synchronic evidential rationality is properly epistemic. But does Rawlsian religious pluralism accept “lawful and interminable inconsistenc”, as the moral evidentialist Scott Aikin thinks is true of permissivist ethics of belief more generally? For discussion of this argument and response, see Axtell, 2019.. They have themselves been critical of the equal weight view, at least on my reading it's not at all clear what positive alternative they offer, consistent with their view that the evidentialist principle should be normative over the ethics of beliefAiken (Aikin, 2014), Feldman and Conee seem to advocate the intrinsical in contrast to the instrumentalist account of the evidentialist principle's normativity. See Anne Meylan (Meylan, 2019) for an argument that the view that believing in accordance with one's evidence is intrinsically right is untenable.. One who rejects equal weight seems driven to the normativity of a commutativity of evidence principle, a principle which tells us that each piece is to be weighed equally, was given its proper due, irrespective of the actual agent's actual earlier or later acquisition of those pieces. This I argue raises a different but equally absurdly high bar for reasonableness.

Finally, reasonable pluralism thought of as supported by this “diachronic turn” in the ethics of belief, may well be consistent with something like a Peircean eschatological convergence of views. The latter view is not obviously inconsistent with a rejection of the Uniqueness ThesisThanks again to an anonymous Referee for suggesting this last response to the evidentialist and proponent of UT.. If this is correct, there needs to be nothing lawful or interminable about the kind of inconsistency pragmatism permits and accounts for. An initially robustly ambiguous evidential situation can also change in positive ways, reducing the underdetermination of our conflicting beliefs, and favoring one or the other of them, or some other previously un-envisioned belief instead of either. So, the “interminable” charge doesn't seem correct of the “lawful” (reasonable) disagreement which permissivism defends. And although the intellectualism of evidentialist internalism tries to cut off the influences of individual temperament and social milieu over our worldview beliefs, treating these as non-rational influences that might just be cast aside by agent's modeled as evidentialism models them, this seems misguided also. Due to our intellectual engagements with one another our temperaments themselves might also moderate, leading to convergence down the road willThanks to an anonymous referee for this suggestion. For a sharp rebuttal of evidentialism based on its ill-fit with the manner in which temperament deeply affects many beliefs associated with personal identity, see Kidd, 2013..

Much has been written between synchronically-focused “equal-weight” and “dogmatic” responses to peer disagreement. Carter's controversial view agnosticism offers many improvements to the debate. Primarily, it recognizes as most evidentialists do not, the need to an epistemology of domains of controversial views apart from every day or straightforwardly empirical beliefs. While I admire his many valid criticisms of Feldman's assumptions, the normative upshot for J. Adam Carter in his “On Behalf of Controversial View Agnosticism” his controversial view agnosticism, is still a form of impermissivism. But principled agnosticism is equally as given to RUT and to a synchronic conception of guidance as is the “equal-weight” view. The “liveability” of either sort of guidance is challenged not just by religious philosophers but by secular ones like myself who find the idealized conception of mind and voluntarism presupposed in their guidance a legacy of internalism's truck with philosophical rationalism. Carter claims that the epistemological significance of disagreement is such that “we are rationally obligated to withhold judgment about a large portion of our beliefs in controversial subject areas, such as philosophy, religion, morality, and politics”. While he recognizes that a thorough-going agnostic suspension might be charged with “spinelessness”, and impracticability-- the un-liveability objection--he qualifies Feldman and Conee's evidentialism to acknowledge these concerns. Carter qualifies Feldman's “tacit commitment to the Triad View, with its deontological categories of belief, suspension of belief, and disbelief”, by arguing that it “has the effect of artificially restricting the range of reasonable attitudes we might take up in controversial areas...” (15). The abandonment of this commitment opens up other possible sub-dox- astic attitudes besides “suspension”. Carter expands the connotation of “agnosticism” to include the sub-doxastic attitude of “suspecting that”. By then introducing a right to “suspecting that” but not “believing that”, when conditions are right, Carter thinks he held onto the guiding principle of conformism while rendering it more “livable”. But the account remains impermissivist: there is still assumed a single right response to revealed peer disagreement among controversial views: agnosticism. The principled agnostic's categories of doxastic attitudes are still essentially treated deontologically since they line up with epistemic duties or entitlements.

These are things denied by permissivists, who tend to favor assertability rather than propositionally-focused treatments of well-founded belief, or treatments which cohere better with the ecological rationality of agents, and the mixed valuative / epistemic character of controversial views. I argue that the what an inductive risk approach brings to the table to distinguish motivated from unmotivated etiological challenges, also shows how overgeneralized are the prescriptions that conformists and steadfasters each take to be the normative upshot of genuine peer disagreement. They accord well also with Susan Haack's work of the ethics of belief developing the advantages of an “Overlap” model according to which there is, not an invariable correlation, but partial overlap, where positive / negative epistemic appraisal is associated with positive / negative ethical appraisals (129). Haack's model allows her to identify the scope and proper limits of such doxastic responsibility by (a) restricting the domain in which instances of believing may be judged on ethical as well as epistemic grounds; by (b) distinguishing role-specific responsibilities from those that are more generally appropriate; and by (c) identifying circumstances that serve to exonerate individuals from unfortunate epistemic failures I here paraphrase R.A. Christian's (Christian, 2009: 468-469) useful elaboration of Haack, ¹997..

In this connection we might pause to note that there are some interpretations of Clifford's thought which puts him on the side of the Mil- lian-Jamesian emphasis on character types, instead of opposed to it. This opens paths for mediation between Clifford and James, and more generally between impermissivism and permissivism in the ethics of belief. Niko- laj Nottelmann and Patrick Fessenbecker (2019) argue that “in Clifford's eyes the shipowner's relevant failure is not so much volitional as charac- terological” (Nottelmann & Fessenbecker, 2019: 3). Darwin was a strong influence for Clifford, and if we pay attention to his “implicit Darwinism” we find a considerably less voluntaristic Clifford than the shipowner case is usually taken to present. These authors, like myself, greatly prefer the Clifford who invite attribution theory, the Clifford who wrote that an agent's “mental character” determines his course of action, something that only requires “conscious consultation of [one's] past history”Clifford, The Ethics of Belief, 52, as discussed in Nottelmann & Fessenbecker, 2019.. Whether of Clifford himself or of the ethics of belief debate, I would certainly endorse the author's proposal to dismiss the volitionist interpretation associated with the shipowner case, “replacing it with an attributionist interpretation more naturally covering his full range of salient cases”ibid.. These authors' ability to utilize Haack's mediating approach to the ethics of belief is most amenable conciliation. I argue that her “overlap” model offers the best account of the relationship between moral and epistemic concerns. I still see Clifford's “sole and supreme allegiance of conscience to the community” as unhelpful, a kind of social duty so strong as to be a kind of dulce et decorum est. While no moral duty could be as strong as Clifford, or the equal-weight view, or principled agnosticism demands, the moral approach to evidentialist constraints I do hold as far superior to that of Feldman and Conee's epistemological account, which purposely eschews moral evidentialism..

We have explained why there is a need for a more principled and fine-grained application of bias studies to well-motivated etiological challenges. But what, more specifically, has inductive risk got to do with this? How do inductive norms and the study of the epistemic risk entailed by their violation help us to motivate the censure of an agent as unreasonable? We can now turn to these questions more directly.

5. VIOLATIONS OF INDUCTIVE NORMS AS PROPER GROUNDS OF CENSURE

The project of guidance, we have said, is a quite different project than analysis of knowledge and shouldn't be confused with it. But guidance clearly does connect with censure, and when it does, its force is in the direction of humility or increased deference to higher-order evidence. But this can manifest in different ways, not just synchronically in a lowered degree of credence or a shift from belief to suspension of belief. It can manifest diachronically in habituating oneself to general virtues of doxastic responsibility, and in revising commitments so as to minimize or eliminate the manners in which one's faith ventures put other people at moral risk.

These different ways of responding to strong etiological challenges to our beliefs are not incommensurable. Either is better than the “method of tenacity” and the “method of authority” that C. S. Peirce found so prevalent in our all-too-human ways of settling belief.

But given how unsettling the preference of many philosophers for principled agnosticism may be for testimonial faith traditions, and the important positive connections between risk-taking and identity, permissivists should likely set themselves apart by allowing for guidance that make demands less drastic, and more focused on forward-looking inquiry. We have to start with where we are and learn to recognize and resist biases; no one can simply put them aside and then recalculate the weight of evidence as an ideal agent for whom the temporal order of an agent's coming to hold evidence bearing on a belief is completely discountedEvidentialism and its thesis O² are committed to the Principle of Commutativity, where the order of the acquisition of evidence should be irrelevant to its weight in a person's deliberations. For a discussion of the commutativity principle, see Georgi Gardner (Gardiner, 2014).. Nothing could be further from the psychological truth for human beings, whose “nurtured” beliefs coming to them from an early age and make up a large part of their developing identity.

Belief revision and responses to higher-order evidence are perhaps the more direct focus of the ethics of belief than belief acquisition, at least where agents appear to have more control over revising their beliefs than over how they originally come to hold them.