Responsibilist Evidentialism

Responsibilist Evidentialism Christopher Cloos University of California, Santa Barbara forthcoming in Philosophical Studies; please cite the published version available at: http://dx.doi.org/10.1007/s11098-015-0454-9 1 Introduction Epistemic justification is a central concern in epistemology. Those concerned with giving a fully general account of epistemic justification must address the question: When is a subject S justified in believing a proposition p at a specific time t?1 For example, what conditions must be met for Marie to be justified in believing at t that the Big Bang occurred 13.8 billion years ago? Some theorists think all that is needed for Marie to be justified in believing that proposition is for Marie to have evidence that, on balance, supports believing the proposition or lending it substantial credence. I think this is not the whole story. What’s missing? Marie must not only have on balance evidential support for the proposition in order to be justified in believing it, she In this paper, I follow Earl Conee and Richard Feldman’s usage of “epistemic justification” as synonymous with “propositional justification.” Such justification contrast with “doxastic justification,” or, in Conee and Feldman’s terminology, “well-foundedness.” Propositional justification involves having adequate grounds for a proposition in order to be justified in believing that proposition. One can have propositional justification for believing a proposition even if (i) one does not actually believe the proposition, or (ii) one does believe the proposition but not on the basis of the adequate grounds that one has for the proposition. This contrasts with doxastic justification, which requires: propositional justification, actually believing the proposition, and properly basing one’s belief on the adequate grounds that one has for the proposition. This paper is an attempt to add an objective requirement to propositional justification (i.e., whether or not one has in fact responsibly acquired and sustained one’s evidence relevant to the proposition). Whether an epistemic agent has the evidence he ought to have need not be represented mentally by the agent or be introspectively accessible to the agent upon reflection in order for it to impact propositional justification. Doxastic justification, which involves proper basing, may require reflective access to the fact that one has the right grounds for one’s belief. The basing relation is a vexed notion that will not be addressed in this paper. For a good tour through views on the basing relation see Evans (2013). 1 must also have responsibly acquired and sustained the evidence that supports the proposition. She must have the “right evidence”—the evidence she ought to have. If she does, then she is responsible in believing the proposition in light of that evidence and justified in doing so. In slogan form, past responsibility impacts present justification. History matters when it comes to epistemic justification.2 To motivate the view just sketched, I will bring out a deficiency in a prominent theory of epistemic justification. Evidentialism, as defended by Earl Conee and Richard Feldman, holds that the justification of a doxastic attitude (i.e., believing, disbelieving, or suspending judgment) is completely determined by the evidence that a subject possesses at a particular point in time. How the subject came to possess the evidence that she possesses is irrelevant when it comes to justification.3 According to Evidentialism, if a subject neglected to gather relevant evidence for a proposition, evidence that it is reasonable to think she could and should have gathered, this fact does not impact whether she is presently justified in believing the target proposition. Synchronic (i.e., at-a-time) considerations are all that matter. To challenge the evidentialist thinking just glossed, philosophers have put forward cases that intuitively suggest that diachronic (i.e., overtime) considerations impact epistemic justification.4 These cases, plus other evidentialist commitments, generate what I call the “insufficiency objection” to Evidentialism—evidential support is insufficient for securing propositional justification. When a subject makes poor use of her epistemic agency in gathering and evaluating evidence across-time, and such epistemic irresponsibly results in her possessing the evidence that she has at-a-time, a subject can fail to be justified in believing a proposition given that evidence, even when that evidence on balance 2 This idea finds its contemporary provenance in Kornblith (1983), Greco (1990), Greco (2005), and Foley (2005). For a clear statement of this commitment see Feldman (2004: 89). 4 Examples of such cases can be found in Kornblith (1983), Greco (2005), DeRose (2011), and Baehr (2011). Responses to such cases are found in Conee and Feldman (2004: 89-90, 186-89) and Feldman (2003: 47-48). 3 RESPONSIBILIST EVIDENTIALISM | 2 supports the proposition. Diachronic considerations matter. Conee and Feldman’s brand of Evidentialism cannot properly account for such considerations. Here is how this paper will proceed. In section 2, I will present a case involving epistemic irresponsibility. Using this case, along with an evidentialist commitment regarding epistemic obligations, I will establish the “insufficiency objection” to Evidentialism. Section 3 addresses a response to the insufficiency objection by Earl Conee. I will argue that Conee’s response does not succeed and the insufficiency objection still stands. To overcome the insufficiency of Evidentialism I will introduce, in section 4, a modified version of Evidentialism. I call this amended version of Evidentialism “Responsibilist Evidentialism” because it adds an epistemic responsibility requirement to epistemic justification. For Responsibilist Evidentialism, being justified in believing a proposition at-a-time involves having on balance evidential support for the proposition as a result of responsible evidential conduct over-time. Responsibilist Evidentialism adequately handles the case involving epistemic irresponsibility. Lastly, in section 5, I will continue developing Responsibilist Evidentialism in response to objections. 2 The Insufficiency Objection As championed by Conee and Feldman (2008: 83), Evidentialism (E) endorses the following principle of epistemic justification: (E) S is justified in believing p at t iff S’s evidence at t on balance supports p. Jason Baehr (2011) proposes a counterexample to (E). This case involves epistemic irresponsibility in the form of negligence regarding evidence gathering. The Oblivious George case runs as follows: 3 | RESPONSIBILIST EVIDENTIALISM Oblivious George. George epitomizes the vices of intellectual laziness, apathy, and obliviousness. He goes about his daily routine focusing only on the most immediate and practical of concerns: feeding himself, getting to work on time, doing his job in a minimally satisfactory way, paying the bills, etc. He lacks any natural curiosity and is almost entirely tuned out to the news of the day. Unsurprisingly, George has many beliefs he should not and fails to believe many things he should. In the former category is George’s belief that exposure to secondhand smoke poses no significant health risks. Given his extremely narrow and practical focus, George is oblivious to all of the wellpublicized research indicating the hazards of secondhand smoke. In fact, George actually has positive evidence in support of his belief. He recalls having learned from a reliable source some years ago that a considerable amount of research had been conducted concerning the effects of exposure to secondhand smoke and that this research had failed to establish any correlation between such exposure and any serious health problems. And as far as George knows, the research on this topic has not changed. Nor, we may suppose, does he have any reason to think that it might have changed.5 George has at time t1 evidence e1—a memorial experience as of “having learned from a reliable source some years ago” that there is no correlation between exposure to secondhand smoke and serious health problems.6 George’s evidence e1 supports the following proposition: (p1) Exposure to secondhand smoke poses no significant health risks. George has no reason to doubt his memorial experience, so George’s total evidence E at t1 on balance supports p1. According to evidentialist principle (E), George is justified in believing p1 at t1. Contrary to (E), why is George not justified in believing p1 at t1?7 5 Baehr (2011: 90). Conee and Feldman (2008: 87-88) think evidence is non-propositional. It ultimately consists in experiences. Against Conee and Feldman’s stance on evidence being non-propositional see Dougherty (2011). I have doubts that the constitution of evidence is non-propositional, but, for the purposes of this paper, I endorse my interlocutors’ conception of the nature of evidence as experiential. 6 RESPONSIBILIST EVIDENTIALISM | 4 George is not justified in believing p1 at t1 because George is culpably ignorant of propositions relevant to p1 due to his epistemically blameworthy evidential inquiry over an earlier period of time (i.e., t0,…,t0-n). Experiences with the propositions George is culpably ignorant of as their content are defeaters regarding believing p1.8 Over that earlier period of time t0,…,t0-n George exhibited a pattern of negligently omitting to gather the readily available defeaters concerning believing p1, as evidenced by the following fact of the case: (f1) Due to his intellectual laziness, apathy, and obliviousness, George is entirely oblivious to all of the well-publicized research indicating the hazards of secondhand smoke.9 George should have at t1 evidence e1 plus experiences as of well-publicized research having established that there is a positive correlation between exposure to secondhand smoke and serious health problems. George is culpably ignorant of propositions indicating that p1 is false. His past epistemic irresponsibility affords the following epistemic ought (O) claim: (O1) George ought not to possess only e1 as evidence concerning p1 at t1. At t1 George ought to have evidence e1 plus evidence e2,…,en, i.e., experiences as of learning about the up-to-date research on the link between secondhand smoke and health issues. The expanded body of evidence E* (i.e., e1 & e2,…,en) that George should possess at t1 would not on balance support p1. So George’s evidence at t1 supports p1, in part, as a result of his irresponsible evidential conduct at an earlier time. This explains why, as the case indicates, George’s belief Answering this question establishes the “insufficiency objection” to Evidentialism, which shows that principle (E) is false—the left-hand side of the biconditional is false while the right-hand side of the biconditional is true. 8 A defeater is something that can undermine or diminish the justification of a doxastic attitude toward a proposition in relation to a body of evidence. When a defeater is added to a body of evidence it can undermine or override such justification. 9 Baehr (2011: 90). 7 5 | RESPONSIBILIST EVIDENTIALISM that p1 falls into the category of a belief that George should not have. These considerations support the second epistemic ought (O) claim: (O2) George ought not to believe that p1 at t1.10 If George ought not to believe that p1 at t1, as O2 indicates, then George is not justified in believing that p1 at t1. This stems from an evidentialist commitment regarding epistemic obligations. As Conee and Feldman (2004: 88) indicate, “We think that being epistemically obligatory is equivalent to being epistemically justified.”11 Conee and Feldman think there is an equivalence relation that holds between what one ought to believe—what it is rational to believe—and what one is justified in believing. Given that believing p1 at t1 is not the attitude that George ought to have, it follows that: (NJ) George is not justified in believing that p1 at t1. This Not Justified (NJ) claim establishes that evidentialist principle (E) is false: George’s evidence at t1 on balance supports p1, but George is not justified in believing p1 at t1. Having a doxastic attitude that “fits” one’s evidence, when that evidence on balance supports a proposition, is insufficient for securing epistemic justification and epistemic obligation concerning that proposition—mere on balance evidential support is not enough. As Feldman acknowledges: 10 For a related argument against Evidentialism using epistemic ought claims see DeRose (2011). This commitment to the equivalence of epistemic rationality, reasonableness, and justification is reflected in the linguistic practices in the literature. For instance, Feldman (2006: 220) addresses the use of the term ‘reasonable’ with regard to the epistemology of disagreement and says, “I will sometimes substitute the words ‘rational’ or ‘justified’.” Conee and Feldman think reasonable beliefs are rational beliefs, epistemic obligations go with rational beliefs, and rational beliefs are justified beliefs. Stewart Cohen (1984: 283) exhibits a similar commitment when he says that, “‘reasonable’ and ‘rational’ are virtual synonyms for ‘justified’.” 11 RESPONSIBILIST EVIDENTIALISM | 6 Some philosophers who are generally sympathetic to evidentialism may think that merely having one’s evidence on balance support a proposition is insufficient for it to be true that one epistemically ought to believe the proposition. (from Conee and Feldman 2004: 177) This establishes the insufficiency objection to Evidentialism. Given that O1 supports O2, and evidentialists think that there is an equivalence relation between epistemic obligation and epistemic justification, NJ follows. The Oblivious George case undermines Evidentialism. 3 An Evidentialist Reply In this section, I will address an evidentialist response to the insufficiency objection. First, it is helpful to set aside a flat-footed evidentialist response. An evidentialist might say that the reason there is an equivalence relation between epistemic obligation and epistemic justification is because both notions rely on the same analysis, namely, having the doxastic attitude that fits one’s evidence. So, O2 is false because George ought to have the doxastic attitude that fits his evidence at t1—George ought to believe that p1 at t1. As a result, NJ is false—George is justified in believing that p1 at t1. This response is flat-footed because it simply reiterates the very analysis the skeptic of Evidentialism is calling into question. Simply reasserting the analysis at stake is not reestablishing the truth of the analysis. This is because O2 is supported by O1. A better evidentialist strategy involves showing: if O1 supports O2, NJ does not follow. In fact, this is the strategy Conee pursues in response to the insufficiency objection.12 Conee (2014) argues against the insufficiency objection using a distinction between subjunctive and indicative readings of epistemic ought claims O1 and O2. What makes O1 true is 12 In his comments on an earlier version of this paper presented at the 2014 Pacific Division Meeting of the American Philosophical Association (APA). Comments included with permission. 7 | RESPONSIBILIST EVIDENTIALISM a subjunctive statement (i.e., a statement contrary to what’s actually the case). The subjunctive reading of O1 (i.e., � � ) is: (� � ) George ought not to possess only e1 as evidence concerning p1 at t1 because it would have been epistemically best for George to have at t1 additional evidence e2,…,en concerning p1. O1 only supports O2 if O2, like O1, is given a subjunctive reading: (� � ) George ought not to believe that p1 at t1 because it would not have been epistemically best for George to believe that p1 at t1. Regarding p1, � � holds because it would have been epistemically best for George to have conducted his evidential inquiry in a responsible manner. He would have acquired additional evidence concerning p1, evidence indicating that ¬p1 is the case. � � is supported by � � because given the evidential conduct that would have been best at an earlier time it would not have been best for George to believe that p1 at t1. However, as Conee (2014) says, “what would have been epistemically best does not decide the actual justificational status of p1 for George. That is determined by how epistemic things actually are, not how they would have been at their best.” The reading of O1 that supports O2 does not, in turn, support NJ. NJ concerns, “the actual justificational status of p1 for George,” whereas the subjunctive readings of O1 and O2 concern a counterfactual epistemic situation for George regarding “what would have been epistemically best.” Despite this, there is a reading of O2 that directly concerns George’s actual epistemic situation at t1. It concerns an indicative statement (i.e., a statement concerning what’s actually the case). The indicative reading of O2 (i.e., �� ) is: RESPONSIBILIST EVIDENTIALISM | 8 (�� ) George ought not to believe that p1 at t1 because it is epistemically best for George not to believe that p1 at t1. Conee (2014) argues that �� is false, so it does not support NJ: George’s rational grounds at t1 concerning the truth of p1 consist in his memory of a reliable source reporting extensive research that found no correlation of secondhand smoke with health problems. This supports p1. George has no reason at t1 to mistrust that support. Once again, we are assuming that he has no indication that his intellectual vices might have led him astray about p1. Under these conditions, it is epistemically best for George to heed the support of the reasons that he does have. So it is epistemically best for George to believe p1. On this reading, O2 is untrue. It thus does not help to defend NJ.13 In response, I will argue that the subjunctive reading of O1 supports the indicative reading of O2. What would have been epistemically best for George to have at t1 concerning p1—additional evidence e2,…,en—impacts the attitude toward p1 it is epistemically best for George to take at t1. Conee (2014) claims that, “what attitude is actually epistemically best for George to take is the attitude that it is actually rational for George to take.” So, the question becomes: What attitude is actually rational for George to take toward p1 at t1?14 13 For a similar response strategy to a case involving epistemic irresponsibility see Feldman (2003: 48). In this section, I provide a negative answer to this question by indicating which attitude is not actually rational for George to take toward p1. This negative answer implies a positive answer to the question. Given that George is going to take an attitude toward p1 at t1, he ought not to (i.e., is obligated not to) believe that p1 at t1. Believing p1 is rationally impermissible; yet, either disbelieving p1 (i.e., believing ¬p1) or suspending judgment concerning p1 (i.e., neither believing nor disbelieving p1) is a rationally permissible attitude for George to take toward p1. Responsibilist Evidentialism is not committed to saying that one of the attitudes is uniquely rationally permissible or that one attitude is more permissible than the other, though this is possibly the case. Instead it is plausible that George’s epistemic irresponsibility has eventuated an epistemic situation whereby no attitude is uniquely rational toward p1, given George’s irresponsibly sustained body of evidence. Such situations are epistemically permissive, allowing for more than one attitude toward a proposition to be epistemically rational given one’s total evidence. This runs counter to the “uniqueness thesis,” which holds that, “Given one’s total evidence, there is a unique rational doxastic attitude that one can take to any proposition” (White 2005: 445). For arguments against the uniqueness thesis, and in support of permissiveness, see Brueckner and Bundy (2012) and Peels and Booth (2014). Is disbelieving p1 or 14 9 | RESPONSIBILIST EVIDENTIALISM It is something of a truism that rational thinkers respect their evidence. Thomas Kelly (2006) explains this truism as follows, “Insofar as one is rational, one is disposed to respond appropriately to one’s evidence: at any given time, one’s views accurately reflect the character of one’s evidence at that time.” What are the distinctive features that comprise the “character” of one’s evidence at a time? There are, at least, two relations a body of evidence might bear to a proposition. These relations comprise the character of a subject’s evidence at a point in time. The first relation is the epistemic support relation. This relation holds when a subject’s evidence confers rational support on a proposition. For Conee and Feldman, evidence consists in experiences.15 It is nonpropositional. Yet, for Conee and Feldman, experience has propositional content and an experience rationally supports a proposition when that proposition is part of the best explanation of that experience.16 That a subject’s body of evidence supports a proposition is part of the character of that body of evidence, but it is not the whole character of that evidence. The second relation that comprises the character of a subject’s body of evidence is the normative support relation. This is an enabling condition on rational support.17 An enabling condition ensures that nothing undermines rational support. It allows rational support to have its suspending judgment concerning p1 a justified doxastic attitude for George? According to Responsibilist Evidentialism, neither attitude is an epistemically justified attitude for George to take toward p1 at t1. Though such attitudes are epistemically rational, they are not epistemically justified because George’s total evidence at t1 does not on balance support suspending judgment or disbelieving. This result reinforces the epistemic costs of epistemic irresponsibility. In specific cases, it results in epistemic rationality and epistemic justification coming apart; it results in the absence of a justified doxastic attitude for the agent to take toward the proposition, given the epistemic situation the agent has gotten himself into. Justification goes with obligation, and there is no attitude George is epistemically obligated to take toward p1 at t1. 15 Refer to Conee and Feldman (2008: 87-88). 16 See Conee and Feldman (2008: 98). This is an explanationist construal of the epistemic support relation. For an argument against explanationism refer to Byerly (2013). For defense of an explanationist construal of the epistemic support relation see McCain (2013) and McCain (2014). 17 Another enabling condition is the “no defeater condition.” For an instructive discussion of normative adequacy and enabling and conferring conditions refer to Kvanvig (2014). RESPONSIBILIST EVIDENTIALISM | 10 effect and secure epistemic rationality, obligation, and justification.18 The normative support relation holds between a subject’s body of evidence and a proposition when that evidence is the evidence the subject ought to have concerning the proposition. Phrased negatively, the relation holds in the absence of the subject possessing the wrong evidence. Determining whether the normative support relation holds requires looking at what would have been epistemically best, with regard to the subject’s evidential conduct, for the subject to have been a responsible epistemic agent. What is a “responsible epistemic agent?” As John Heil (1983: 362) explains, The ‘responsible’ epistemic agent is, roughly, one who goes about the activity of information gathering in a suitable fashion….We enjoin one another (and of course ourselves) to ‘act responsibly’ in this regard, I think, because acting in this way tends to make our beliefs more dependable, more reliable, more likely to be true. It is our status as intelligent, belief-acquiring instruments that constitutes our status as epistemic agents. We do not choose what beliefs we acquire, but we have a hand in determining how we shall go about the business of gathering those beliefs. And this is a business that may be conducted responsibly or casually.19 We can ask: Did the subject act responsibly in her role as an epistemic agent in gathering evidence at an earlier time? If the subject did not, the normative support relation does not hold. 18 There is a normative enabling condition on epistemic justification because epistemic justification is a normative notion. As Lehrer and Cohen (1983: 193) indicate, “Justification is a normative concept. It is an evaluation of how well one has pursued one’s epistemic goals.” For more elaboration on this point see Alston (1989: 115), Plantinga (1993: 14), and Pryor (2001: 109-12). 19 Feldman, in Conee and Feldman (2004: 175), endorses a similar line of reasoning regarding responsibly fulfilling the role of being an epistemic agent (i.e., a believer), “There are oughts that result from playing a certain role or having a certain position…They are based on what’s a good performance. Furthermore, it is plausible to say that the role of a believer is not one that we have any real choice about taking on. It is our plight to be believers. We ought to do it right. It doesn’t matter that in some cases we are unable to do so…Even in cases in which a believer has no control at all, it makes sense to speak of what he ought to believe and ought not believe.” I concur with both Heil and Feldman that deontological judgments regarding epistemic agency can be true even when the agent in question lacks direct voluntary control over what they believe (i.e., even when doxastic voluntarism is false). Given George’s evidence, he may not be able to not believe p1. Yet, we can still judge that George ought not to believe p1 by reference to what would be a good performance in the role of a believer in George’s epistemic situation. 11 | RESPONSIBILIST EVIDENTIALISM The evidence that supports the proposition is not the evidence that should support the proposition. It is not the “right” evidence, the evidence the subject ought to have. What is the character of George’s total evidence E at t1? George’s evidence at t1 stands in an epistemic support relation to p1. George’s evidence on balance supports p1. Does George’s evidence stand in a normative support relation to p1? Baehr (2011: 92) thinks not, as he says, “The evidence possessed by George…is in a certain substantial way defective or contaminated— and for reasons that trace back to [his] own epistemic wrongdoing.” George’s blameworthy evidential inquiry over a time period prior to t1 (i.e., t0,…,t0-n) has infected his evidence with an undesirable quality, namely, George’s evidence “ought to be other than it is.”20 In their response to Baehr (2011), Conee and Feldman (2011: 313) agree that George should have gotten himself different evidence and that George’s evidential conduct was unjustified.21 As a result, George’s evidence at t1 is deficient, and his evidence does not stand in a normative support relation to p1. It would have been epistemically best for George to have at t1 additional evidence e2,…,en concerning p1, had George performed well as an epistemic agent at a time prior to t1. So, as � � indicates, George ought not to possess only e1 as evidence concerning p1 at t1. George’s total evidence E at t1 is not the evidence that George ought to have. The subjunctive reading of O1 (i.e., � � ) supports the indicative reading of O2 (i.e., �� ). Given the character of George’s actual evidence, George ought not to believe that p1 at t1. As a rational thinker, George is not respecting the actual character of his actual evidence by believing 20 Baehr (2011: 91). However, they do not agree that George’s evidence “ought to be other than it is” because George’s evidence is a memory and his memory could not have provided different evidence than it did (Conee and Feldman 2011: 313). This is a confused point. For Conee and Feldman, experiences are evidence. George remembering a proposition about secondhand smoke and health is evidence that the proposition is true. His memorial experience does not simply provide evidence—it is evidence. Additionally, Baehr’s point is about George’s total evidence. Baehr is not implying that George’s memory should have been different or had different propositional contents. Rather, Baehr is indicating that George should have had additional experiences in his total evidence at time t1. 21 RESPONSIBILIST EVIDENTIALISM | 12 that p1. George’s actual evidence is deficient. Given that George’s evidence does not stand in a normative support relation to p1, George believing that p1 is not epistemically best. So what would have been epistemically best regarding George’s past evidential conduct impacts what is epistemically best. George ought not to believe that p1 at t1. This is what �� indicates. Lastly, given that � � supports �� , NJ follows. George is not justified in believing that p1 at t1. Epistemic justification is fundamentally a normative concept,22 and the normative support relation enables the epistemic support relation to secure epistemic justification. George’s evidence not being what it ought to be prevents believing p1 from being the attitude George ought to take. Given the tight connection between epistemic obligation and epistemic justification, and that George ought not to believe that p1 at t1, it follows that George is not justified in believing that p1 at t1. NJ is true. 4 Responsibilist Evidentialism To remedy the insufficiency of Evidentialism I will add a responsibility condition to evidentialist principle (E). This amended version of Evidentialism is called “Responsibilist Evidentialism.”23 It embraces the main intuition behind the insufficiency objection, which is that the etiology of a subject’s body of evidence—how it was acquired and sustained—impacts whether a doxastic attitude toward a proposition given that evidence is justified. Ultimately a subject is 22 As previously mentioned in footnote 18. Though, one might wonder: if epistemic justification is fundamentally a concept regarding epistemic permissions and obligations, how can it be an externalist notion? Isn’t a deontological conception of epistemic justification an internalist conception such that the elements that factor into justification are internal to the subject or cognitively accessible upon reflection? Wouldn’t this be a mark against my proposal that combines a deontological approach with an externalist/objectivist condition on justification? Discussion of whether justification is internalist or externalist goes beyond the scope of this paper, but there is reason to think that a deontological account of epistemic justification is not necessarily internalist. For arguments supporting the possibility of combining epistemic deontology with externalism see Brueckner (1996) and Littlejohn (2012). 23 Responsibilist Evidentialism is a more robustly deontological theory of epistemic justification than Conee and Feldman’s Evidentialism. I agree with Baehr that there is no conflict, in principle, with combining deontological elements with evidentialist elements, as he says, “I see no reason to think of ‘evidentialism’ as necessarily excluding deontological accounts of justification” (Baehr 2011: 92 n.14). 13 | RESPONSIBILIST EVIDENTIALISM epistemically responsible in believing a proposition when the subject’s body of evidence that supports the proposition was responsibly acquired and sustained. If it wasn’t, the person is not justified in believing the proposition. I will unpack Responsibilist Evidentialism in a stepwise fashion. The most general formulation of Responsibilist Evidentialism (RE) is an emendation of evidentialist principle (E): (RE) S is justified in believing p at t iff S’s evidence at t on balance supports p and S is epistemically responsible in believing p at t. For (RE), evidential support and epistemic responsibility are individually necessary and jointly sufficient for a subject to be justified in believing a proposition. The notion of a subject being “epistemically responsible in believing” can be cashed out in terms of epistemic blamelessness.24 As John Greco (2005: 261) indicates, [J]udgments concerning whether a person is epistemically responsible with respect to some belief b are often equivalent to judgments about whether the person is epistemically blameworthy with respect to b.…[W]hether a person is epistemically blameworthy for holding some belief is partly a function of the person’s prior behavior: if S’s reasons for believing b are the result of prior negligence, then S is not now blameless in believing b. Along these lines, the condition of epistemic responsibility in (RE) is further articulated in the Responsible Belief (RB) principle: 24 For a thorough treatment of epistemic blameworthiness and its applicability to belief see Nottelmann (2007) and Russell (2001). RESPONSIBILIST EVIDENTIALISM | 14 (RB) S is epistemically responsible in believing p at t iff S is epistemically blameless in believing p at t.25 (RB) indicates that responsible belief is blameless belief. Blamelessness attaches to belief in a derivative way.26 Believing a proposition is blameless when a subject is not worthy of blame for the evidence that the subject possesses for the proposition. This idea is articulated in the Blameless Belief (BB) principle: (BB) S is epistemically blameless in believing p at t iff it is not the case that S ought not to have only the evidence S has for p at t because S is culpably ignorant at t of propositions relevant to p due to S’s epistemically blameworthy evidential inquiry at a prior time (i.e., t0,…,t0-n). According to (BB), epistemic blamelessness involves the absence of culpable ignorance of relevant propositions, ignorance that stems from evidential misconduct at an earlier time. Now I will explain the notions in (BB) of “culpable ignorance” and “epistemically blameworthy evidential inquiry” and relate these notions back to the Oblivious George case. What is epistemic ignorance? Rik Peels (2010: 63) provides a helpful definition of Epistemic Ignorance (EI): 25 (RB) is modeled after principle (DDB) in Booth and Peels (2010: 257). Weatherson (2008) holds that responsible belief is praiseworthy belief, not blameless belief. (RB) is translatable into a principle involving praiseworthiness. But, to avoid confusion, I consistently formulate responsible belief in terms of blamelessness. 26 Belief-formation is not under direct voluntary control, so blameworthiness does not initially target belief. Instead, blameworthiness initially targets activities under voluntary control (e.g., gathering evidence, cultivating intellectual virtues, etc.). Such activities indirectly influence belief-formation. Thus, in relation to blame, belief derives its status from the status attached to such voluntary activities. For more on the “indirect voluntary influence” view regarding belief-formation and epistemic deontology see Alston (1989: 136-42). 15 | RESPONSIBILIST EVIDENTIALISM (EI) Some cognitive subject S is ignorant of some proposition p iff (i) p is true and (ii) S disbelieves p, S suspends judgment on p, or S has never considered p, but would disbelieve p or suspend judgment on p upon considering p. (EI) indicates that a subject is ignorant of a proposition just in case the proposition is true and she has never considered the proposition, but the subject would disbelieve it or suspend judgment on it were she to consider it. In the Oblivious George case, George is ignorant of the following proposition: (p2) Secondhand smoke is hazardous to health. Proposition p2 is a true proposition that is relevant to p1. Had George had an experience with p2 as its propositional content, such as reading about the “well-publicized research” in a reliable publication, George would have acquired a defeater for believing p1. Such evidence with p2 as its content is: (e2) A reading experience as of learning from a reliable publication that research has established that secondhand smoke is hazardous to health. Evidence e2 would be a rebutting defeater for p1 because it would be an epistemic reason for believing ¬p1.27 George should have e2 as part of his total evidence for p1 at t1. Is George culpably ignorant of p2? In general, a subject S is culpably ignorant of a true proposition q, which is relevant to p, when the reason why S is ignorant of q is because of S’s blameworthy evidential inquiry at a The distinction between “rebutting” (AKA, “overriding”) and “undercutting” (AKA, “undermining”) defeaters traces back to Pollock (1986: 38-39). Rebutting defeaters are epistemic reasons (i.e., evidence) for believing the negation of the proposition, whereas undercutting defeaters are epistemic reasons that attack one’s evidence base supporting the proposition. Rebutting defeaters directly undermine the proposition, and undercutting defeaters indirectly undermine the proposition by undermining one’s evidential support for the proposition. 27 RESPONSIBILIST EVIDENTIALISM | 16 prior time.28 Epistemically blameworthy evidential inquiry occurs when a subject poorly evaluates, investigates, or handles evidence. It is when a subject makes poor use of her epistemic agency in relation to readily available, highly relevant evidence. 29 George’s negligence regarding evidential inquiry resulted in him not gathering evidence e2 at some point prior to t1 (i.e., at some point during t0,…,t0-n). Consequently, George is culpably ignorant of p2 at t1. Now I will use Responsibilist Evidentialism to evaluate the Oblivious George case. The facts of the case make the right-hand side of the biconditional in principle (BB) false. It is the case that George ought not to have only e1 as evidence concerning p1 at t1 because George is culpably ignorant at t1 of p2, which is relevant to p1, due to George’s negligence regarding evidential inquiry at a prior time. This means that George is not epistemically blameless in believing p1 at t1. The left-hand side of the biconditional in (BB) is false. Given this, the righthand side of the biconditional in (RB) is false, and the left-hand side of (RB) is false as well. George is not epistemically responsible in believing p1 at t1. The right-hand side of principle (RE) involves a conjunction of evidential support and epistemic responsibility. Because George is not epistemically responsible in believing p1 at t1 the right-hand side of the biconditional in (RE) is false and the left-hand side is false as well. George is not justified in believing p1 at t1. Thus, Responsibilist Evidentialism is not impugned by the Oblivious George case. The view returns the correct result—mere evidential support is insufficient for securing propositional justification. 5 Objections In this section, I continue developing Responsibilist Evidentialism by responding to objections. 28 29 For more on the notion of culpable ignorance refer to Peels (2011) and Smith (2011). See Engel (2013), Reed (2013), and Setiya (2013) for important issues regarding epistemic agency. 17 | RESPONSIBILIST EVIDENTIALISM 5.1 An Analogical Argument Conee (2014) poses a challenge for Responsibilist Evidentialism involving an analogical argument. The analogical argument is based on another situation involving George: Observant George. Suppose that George also recalls his having seen on a friend’s birth certificate that she was born in Winnetka, Illinois. He did not employ any of his lazy narrow obliviousness in acquiring or sustaining this memory. He just happened to notice the birth location when she showed him her birth certificate to win a bet about her birth date, and his memory of the location has remained firm. He has received no reason to doubt it. Its retention has involved no negligence. This memory is George’s basis for believing that his friend was born in Winnetka. In the Observant George case George is justified in believing the following proposition about his friend, whom we can call “Anne”: (p3) Anne was born in Winnetka. In contrast to George’s belief that p1, George’s belief that p3 does not involve epistemic irresponsibility. George responsibly acquired and sustained his evidence concerning Anne’s birthplace. There is no relevant evidence that George should have gathered but failed to gather. George’s Winnetka belief and his secondhand smoke belief have in common epistemic properties that provide evidential support. Both beliefs are supported by a memory that George has no reason to doubt. So, if George’s Winnetka belief is justified, then George’s secondhand smoke belief is justified. The manner in which George’s evidence was acquired and sustained is irrelevant to whether George is now justified in believing a proposition given such evidence. I reconstruct Conee’s analogical argument as follows: RESPONSIBILIST EVIDENTIALISM | 18 (1) George’s Winnetka belief and George’s secondhand smoke belief have the following epistemic properties in common: (i) relying on a clear memory that supports the belief, and (ii) relying on a memory that George has no reason to doubt. (2) George’s Winnetka belief has the epistemic property of being justified. Thus, (3) George’s secondhand smoke belief has the epistemic property of being justified. Analogical arguments are inductive arguments. An inductive argument is strong if, given that all of the premises are true, it is improbable that the conclusion is false. Conee thinks the analogical argument is a strong inductive argument, and he thinks the truth of the premises make the conclusion highly probable. This follows from his explanation of the force of the argument, as Conee (2014) urges, “the beliefs are equal in propositional justification,” and “it is no more rational for George to believe the one than the other.” The beliefs share the additional epistemic property of being justified, and they have that property to the exact same degree. As a result, countering this argument requires weakening it by showing that the beliefs are unequal in justification. To do this, I will highlight a disanalogy between the beliefs.30 This disanalogy is relevant to the justification of the two beliefs. It concerns the content of the evidence supporting the two beliefs and the epistemic obligations generated by that content. The propositional content of George’s evidence for his Winnetka belief concerns information on a birth certificate, whereas the content of George’s evidence for his secondhand smoke belief concerns scientific results. Birth certificate information doesn’t typically change and evolve, whereas scientific research does typically change and evolve. This generates an asymmetry in epistemic obligations. Given the nature of scientific research there is a 30 I thank Patrick Ryan for helpful suggestions regarding this disanalogy. 19 | RESPONSIBILIST EVIDENTIALISM presumption in favor of—an obligation regarding—staying apprised of the research relevant to one’s beliefs.31 This obligation holds in the absence of a reason to suspect that the research might have changed. On the contrary, an obligation to gather evidence regarding a birth certificate only holds in the presence of a reason to suspect that the information might have changed. How does this asymmetry in obligations generate an asymmetry in justification? One way to capture the asymmetry in justification involves what Jennifer Lackey (2005) calls “normative defeaters.” A normative defeater is: [A] proposition D that S ought to believe to be true, yet [D] indicates that S’s belief that p is either false or unreliably formed or sustained. Defeaters in this sense function by virtue of being propositions that S should believe (whether or not S does believe them) given the presence of certain available evidence.32 As scientific research evolves it frequently generates normative defeaters. This happens in George’s case regarding his secondhand smoke belief. When he first forms his belief there are no normative defeaters of his belief, as it reflects the state of the art. His belief is epistemically justified. As the years pass, his information regarding secondhand smoke and health becomes The duty to stay apprised of research is an imperfect duty—there’s a lot of leeway in how one can fulfill this diachronic duty. One does not need to neurotically pour over scientific journals or magazines to satisfy the obligation. However, if one has exhibited a pattern of neglecting to make easy efforts to avail one’s self of readily available updates on the state of research relevant to one’s beliefs, then one has been negligent and not done one’s due diligence with regard to the scientific data. One has breached one’s epistemic duty. The Oblivious George case is such a case. More difficult cases are what we might call “one-shot” cases. Perhaps the data was only accessible, in principle, at one point in time, and a reasonable epistemic agent in the situation would have gathered the data, but one missed gathering the data. Is one epistemically blameworthy for failing to gather that data? Another question to pursue in the future: Are people who support their beliefs using old scientific research epistemically blameworthy when there is recent research that corroborate the old data? Relatedly, how important is it that new data is adverse to one’s evidence or beliefs in terms of generating obligations to gather that data? 32 Lackey (2005: 638-39). These defeaters are related to the issue of whether unpossessed evidence can defeat justification for believing a proposition. For more on this issue see Lycan (1977) and Harman (1980). 31 RESPONSIBILIST EVIDENTIALISM | 20 outdated and normative defeaters crop up given the presence of the “well-publicized research.”33 Proposition p2 is just such a normative defeater. It is a true proposition that George should believe given the availability of evidence e2. Also, p2 indicates that George’s belief concerning p1 is false. Further, at time t1, George is culpably ignorant of p2. George’s negligence regarding availing himself of the readily available research relevant to his belief concerning p1 results in him being epistemically blameworthy in believing p1 at t1. Responsibilist Evidentialism explains how this impacts the justification of George’s secondhand smoke belief. George is not epistemically blameless in believing p1 at t1, so George is not epistemically responsible in believing p1 at t1, and, as a result, George is not justified in believing p1 at t1. By contrast, in the Observant George case, George is not culpably ignorant of any undefeated normative defeaters regarding his Winnetka belief; George does not have any unfulfilled epistemic obligations regarding the birth certificate information. This makes it the case that George is blameless, responsible, and justified in believing p3—that Anne was born in Winnetka. George’s secondhand smoke belief and his Winnetka belief are disanalogous regarding epistemic justification; the two beliefs are not “equal in propositional justification,” as Conee (2014) urged. Thus, I have weakened the analogical argument against Responsibilist Evidentialism and prevented it from working as a counterexample to the view. 5.2 On Blamelessness This objection drives a wedge between epistemic responsibility (i.e., epistemic blamelessness) and epistemic justification. It is possible for a subject to be blameless in believing a proposition This is another feature of normative defeaters, “a subject’s relation to normative defeaters can change over time as a result of changes in the external environment” (Lackey 2005: 641). 33 21 | RESPONSIBILIST EVIDENTIALISM yet for the subject not to be justified in believing the proposition. Conee and Feldman (2004: 63) describe such a situation: Someone who innocently held a belief as a result of external manipulation or psychological compulsion would be blameless in holding this belief. Nevertheless, if the belief is induced directly and the person has good reason to deny the belief and no reason to believe it, then it is clearly not a justified belief. Responsibilist Evidentialism adequately handles this scenario. In light of principle (RB), epistemic responsibility is cashed out in terms of epistemic blamelessness. Making the requisite substitution in principle (RE) yields the following: (RE*) S is justified in believing p at t iff S’s evidence at t on balance supports p and S is epistemically blameless in believing p at t. In Conee and Feldman’s scenario a subject comes to possess her belief through no fault of her own. This makes her blameless in believing the proposition, but it does not make her justified in believing the proposition. Why? She has no evidence for the proposition. In fact, she has evidence against the proposition. Conjoined with the facts of the case, (RE*) entails that she is not justified in believing the proposition. Given that her evidence for the proposition does not on balance support the proposition, one of the conjuncts in the right-hand side of the biconditional in (RE*) is false. This makes the right-hand side of (RE*) false, and it makes the left-hand side false as well. The subject is not justified in believing the proposition, as the case indicates. Another objection to Responsibilist Evidentialism claims that blamelessness is not necessary for epistemic justification. Such a counterexample consists in the subject being justified in RESPONSIBILIST EVIDENTIALISM | 22 believing the proposition, having on balance evidential support for the proposition, and being blameworthy in believing the proposition. Conee and Feldman describe (2004: 63) such a case: Someone who, for barely adequate epistemic reasons, accepts that a friend has been malicious is epistemically justified in that belief. But such a person may be blameworthy for having insufficient trust in the friend. This case does not work against Responsibilist Evidentialism. The blameworthiness in the case concerns a lack of trust that one’s friend would not intend to do harm to one’s self or others. Such trust need not be proportionate to one’s evidence, as being a good friend might involve trust in the goodness of one’s friend despite the (perhaps misleading) evidence to the contrary. Assuming that “barely adequate evidence” is evidence that makes it slightly more likely than not that one’s friend has been malicious, Conee and Feldman’s case suggests that one is blameworthy for not over-believing in the good character of one’s friend—as one’s evidence on balance supports the proposition that one’s friend has been malicious. Blameworthiness for lack of trust in a friend is not the type of epistemic blameworthiness at issue concerning Responsibilist Evidentialism. According to principle (BB), a subject is blameworthy when it is the case that the subject should not have only the evidence the subject has because the subject is culpably ignorant of propositions relevant to the truth of what she believes, due to blameworthy evidential misconduct in the past. In the case at issue, the subject is blameworthy for not having trust in the good character of her friend, but it is unclear if this holds because the subject has neglected to gather evidence she should have gathered regarding her friend’s good character. So, it is unclear whether the subject is blameworthy in the sense relevant to Responsibilist Evidentialism. 23 | RESPONSIBILIST EVIDENTIALISM Now imagine that we alter the case so that the friend is blameworthy in the sense relevant to Responsibilist Evidentialism. Does this show that blamelessness is not necessary for epistemic justification? If one is blameworthy for failing to have the evidence that one ought to have regarding the good character of one’s friend, yet one’s evidence on balance supports the proposition that one’s friend has been malicious, then this is the sort of case Responsibilist Evidentialism was designed to handle. The blameworthiness of one’s past evidential inquiry undermines one’s justification for believing that one’s friend has been malicious. This holds even though one’s evidence makes it slightly more likely than not that one’s friend has been malicious. Responsibilist Evidentialism is not undermined by such a case. Blamelessness is still necessary for epistemic justification. Conclusion In this paper, I argued for a hybrid theory of epistemic justification. It combined Evidentialism with Responsibilism. Being justified in believing a proposition involved having adequate grounds for a proposition as a result of a history of responsible evidential conduct whereby one fulfilled one’s epistemic obligations. I motivated this view by way of an objection to Evidentialism, and I defended the objection against an evidentialist reply. Then I articulated an amended version of Evidentialism capable of overcoming the objection, which was called “Responsibilist Evidentialism.” Lastly, I continued developing Responsibilist Evidentialism in response to objections to the view. Acknowledgements I am deeply grateful for the inspiration and guidance I received from my late advisor, Anthony Brueckner, concerning my exploration of Evidentialism. Thanks to Aaron Zimmerman, Rick Stoody, Patrick Ryan, Earl Conee, and Todd Long for helpful comments and conversations regarding this paper. I also thank audiences at RESPONSIBILIST EVIDENTIALISM | 24 the 2014 Pacific APA conference and the UCSB Graduate Student Colloquium (Spring 2014) for instructive feedback on an earlier version of this paper. References Alston, William. 1989. Epistemic Justification: Essays in the Theory of knowledge: Cornell University Press Ithaca. Baehr, Jason. 2011. Evidentialism, Vice, and Virtue. In Evidentialism and its Discontents, ed. T. Dougherty, 88-102: Oxford University Press. Booth, Anthony Robert and Rik Peels. 2010. Why Responsible Belief is Blameless Belief. Journal of Philosophy 107: 257-65. Brueckner, Anthony. 1996. Deontologism and Internalism in Epistemology. Noûs 30: 527-36. Brueckner, Anthony and Alex Bundy. 2012. On “Epistemic Permissiveness”. Synthese 188: 165-77. Byerly, T. Ryan. 2013. Explanationism and Justified Beliefs about the Future. Erkenntnis 78: 229-43. Cohen, Stewart. 1984. Justification and Truth. Philosophical Studies 46: 279-95. Conee, Earl. 2014. Comments on “Responsibilist Evidentialism: Overcoming the Insufficiency Objection to Evidentialism”. Paper read at the Pacific Division Meeting of the American Philosophical Association (APA), at San Diego, California. Conee, Earl and Richard Feldman. 2004. Evidentialism: Essays in Epistemology: Oxford University Press. ———. 2008. Evidence. In Epistemology: New Essays, ed. Q. Smith: Oxford University Press. ———. 2011. Response to Baehr. In Evidentialism and its Discontents, ed. T. Dougherty, 310-14: Oxford University Press. DeRose, Keith. 2011. Questioning Evidentialism. In Evidentialism and its Discontents, ed. T. Dougherty: Oxford University Press. Dougherty, Trent. 2011. In Defense of Propositionalism about Evidence. In Evidentialism and its Discontents, ed. T. Dougherty, 226-32: Oxford University Press. Engel, Pascal. 2013. Is Epistemic Agency Possible? Philosophical Issues 23: 158-78. Evans, Ian. 2013. The Problem of the Basing Relation. Synthese 190: 2943-57. Feldman, Richard. 2003. Epistemology: Prentice Hall. ———. 2006. Epistemological Puzzles about Disagreement. In Epistemology Futures, ed. S. Hetherington, 216-36: Oxford University Press. Foley, Richard. 2005. Justified Belief as Responsible Belief. In Contemporary Debates in Epistemology, ed. M. Steup and E. Sosa, 313-26: Blackwell Publishing. 25 | RESPONSIBILIST EVIDENTIALISM Greco, John. 1990. Internalism and Epistemically Responsible Belief. Synthese 85: 245-77. ———. 2005. Justification is not Internal. In Contemporary Debates in Epistemology, ed. M. Steup and E. Sosa, 257-70: Blackwell. Harman, Gilbert. 1980. Reasoning and Evidence One Does Not Possess. Midwest Studies In Philosophy 5: 163-82. Heil, John. 1983. Doxastic Agency. Philosophical Studies 43: 355-64. Kelly, Thomas. 2006. Evidence. Stanford Encyclopedia of Philosophy, http://plato.stanford.edu/entries/evidence/. Kornblith, Hilary. 1983. Justified Belief and Epistemically Responsible Action. The Philosophical Review 92: 33-48. Kvanvig, Jonathan L. 2014. Epistemic Normativity. In Epistemic Norms: New Essays on Action, Belief, and Assertion, ed. C. Littlejohn and J. Turri, 115-34: Oxford University Press. Lackey, Jennifer. 2005. Memory as a Generative Epistemic Source. Philosophy and Phenomenological Research 70: 636-58. Lehrer, Keith and Stewart Cohen. 1983. Justification, Truth, and Coherence. Synthese 55: 191-207. Littlejohn, Clayton. 2012. Justification and the Truth-connection: Cambridge University Press. Lycan, William G. 1977. Evidence one does not possess. Australasian Journal of Philosophy 55: 114-26. McCain, Kevin. 2013. Explanationist Evidentialism. Episteme 10: 299-315. ———. 2014. Evidentialism, Explanationism, and Beliefs About the Future. Erkenntnis 79: 99-109. Nottelmann, Nikolaj. 2007. Blameworthy Belief: A Study in Epistemic Deontologism: Springer Netherlands. Peels, Rik. 2010. What is Ignorance? Philosophia 38: 57-67. ———. 2011. Tracing Culpable Ignorance. Logos and Episteme 2: 575-82. Peels, Rik and Anthony Booth. 2014. Why Responsible Belief Is Permissible Belief. Analytic Philosophy 55: 75-88. Plantinga, Alvin. 1993. Warrant and proper function: Oxford University Press. Pollock, John L. 1986. Contemporary Theories of Knowledge: Rowman & Littlefield. Pryor, James. 2001. Highlights of Recent Epistemology. The British Journal for the Philosophy of Science 52: 95-124. Reed, Baron. 2013. Fallibilism, Epistemic Possibility, and Epistemic Agency. Philosophical Issues 23: 40-69. Russell, Bruce. 2001. Epistemic and Moral Duty. In Knowledge, Truth, and Duty: Essays on Epistemic Justification, Responsibility, and Virtue, ed. M. Steup, 34-48: Oxford University Press. Setiya, Kieran. 2013. Epistemic Agency: Some Doubts. Philosophical Issues 23: 179-98. RESPONSIBILIST EVIDENTIALISM | 26 Smith, Holly M. 2011. Non-Tracing Cases of Culpable Ignorance. Criminal Law and Philosophy 5: 11546. Weatherson, Brian. 2008. Deontology and Descartes's Demon. Journal of Philosophy 105: 540-69. White, Roger. 2005. Epistemic Permissiveness. Philosophical Perspectives 19: 445-59. 27 | RESPONSIBILIST EVIDENTIALISM