Debunking and fully apt belief

One of the contentious philosophical issues surrounding the cognitive science of religion (CSR) is whether well-confirmed CSR theories would debunk religious beliefs. These debates have been contentious in part because of criticisms of epistemic principles used in debunking arguments. In this paper I use Ernest Sosa’s respected theory of knowledge as fully apt belief—which avoids objections that have been leveled against sensitivity and safety principles often used in debunking arguments—to construct a plausible debunking argument for religious belief on the assumption that religious belief is formed simply through processes theorized by CSR. But, in fact, most believers also rely on arguments of various sorts, and their beliefs are not debunked.

One of the interesting philosophical issues raised by the cognitive science of religion (CSR) is whether we l-confirmed CSR theories would debunk religious beliefs.What exactly it means to debunk a belief wi l be discussed shortly, but as a first pass: to debunk a belief is to give good reason to think that the belief is poorly formed/unrelia le/incompetent.There has been some controversy about whether we l-confirmed CSR theories would debunk religious beliefs; some scholars say 'yes' (Bering, 2011;Bloom, 2009;Bu bulia, 2013;Leben, 2014;Wi kins and Griffiths, 2012), others say 'no' (Bar ett, 2007;Leech and Visala, 2011;Mur ay, 2009;Mur ay and Schloss, 2013;Thurow, 2013;2014;van Inwagen, 2009).This whole debate is mir ored by a debate in metaethics over whether evolutionary explanations of moral beliefs debunk such beliefs-with just as much controversy (see e.g.Joyce, 2006;Shafer-Landau, 2012;Street, 2006;Vavova, 2014Vavova, , 2015;;Wielenberg, 2016).
Much of the controversy rests ultimately on how to understand debunking ar uments and on whether there is a plausi le epistemic principle undergirding such ar uments.In this paper I want to explore whether a compe ling debunking ar ument can be built using Ernest Sosa's epistemological framework, which prizes fu ly apt belief as a particularly valua le human form of knowledge.Sosa's framework may help to advance discussion on our central question because (i) his framework has gained widespread acceptance, unlike some of the principles used in previous versions of debunking ar uments, and (ii) Sosa's central notion of apt belief has various advantages over the notions of sensitivity and safety, which have featured prominently in the epistemological principles used in previous debunking ar uments.
I sha l begin by presenting a general schema for debunking ar uments and then discussing the cha lenges that face some of the prominent epistemological principles that have been plu ged into this schema.In the second section I briefly present Sosa's epistemological framework and in the third section I use that framework to construct and evaluate a few debunking ar uments using epistemological principles derived from Sosa's framework.I ar ue that a plausi le epistemological principle can be drawn from Sosa's framework that can be used to construct a plausi le debunking ar ument for religious belief on the assumption that religious belief is formed simply through processes theorized by CSR.But, in fact, most believers also rely on ar uments of various sorts, and their beliefs are not debunked.

Debunking arguments
To debunk a belief is to cha lenge that belief in a certain sort of way.Familiar claims of bias are examples of debunking ar uments.Joe says to Jane, "you just believe the cops are innocent of murder because you are fearful of lack people." Joe thus cha lenges Jane's belief that the cops are innocent.Notice that the cha lenge doesn't attempt to provide evidence against Jane's belief-Joe does not in this statement attempt to show that her belief is false or doubtful in itself.Rather, he cha lenges her belief by charging that it was formed improperly-using some method or under some influence or in some circumstances that are risky.In this case, he charges that her belief is formed under the influence of fear rather than through any competence at assessing the uilt or innocence of the cops.Fear is assumed to be an improper or risky ground for this belief.Why? Different versions of the debunking arument, assuming different epistemic principles, wi l answer this question differently.Some wi l say that fear doesn't track the truth, others that it is an unrelia le uide, or that it is an insensitive or unsafe ground for belief.
Many debunking ar uments fit the fo lowing schema: (1) My class of beliefs C are influenced by factor F.
(2) When C beliefs are influenced the way they are by F, they do not satisfy epistemic condition R. (3) Once I become justified in believing that a set of beliefs do not satisfy epistemic condition R, then I lack epistemic status E in continuing to hold those beliefs.
[=Epistemic Debunking Principle] ------C.I lack epistemic status E in continuing to hold the beliefs in C.
Premise (3), the Epistemic Debunking Principle, is the lone philosophical premise in this ar ument.Different substitutions for R and E wi l yield different epistemic debunking principles.Possi le instances of R include being relia le, sensitive, or safe.Typica ly, the instances of E wi l be either knowledge or justification.
To i lustrate one way to fi l in the debunking ar ument schema, return to the example of Joe and Jane.Joe claims that Jane's belief that the cops are innocent is influenced by fear of lack people.A common way to fi l out the debunking ar ument is as fo lows: believing that one person is innocent on the basis of fear of others is an improper way of forming a belief because if the cop were uilty, Jane would sti l believe he was innocent because her fear of lack people would sti l be present.In other words, beliefs of this sort formed in fear are not sensitive to the truth.And once we realize a belief is formed insensitively, we are thereby unjustified in continuing to hold that belief.This debunking ar ument rests on the following epistemic debunking principle: Sensiti ity EDP: once I become justified in believing that a set of beliefs are insensitive to the truth given the way that they are formed, then I lack justification in continuing to hold those beliefs.
This principle is but one of many possi le principles one might use as an epistemic debunking principle.However, sensitivity debunking principles like this one are frequently appealed to-implicitly or explicitly-in debunking ar ument literature (e.g.Be ke, 2014;Thurow, 2013;White, 2010;Wi kins and Griffiths, 2012).In the remainder of this section I wi l briefly describe the pro lems that face this and one other popular principle.We wi l then be left with a question: is there a more plausi le epistemic principle we can use in debunking ar uments, and can a plausi le CSR-based debunking ar ument be built on it?Our subsequent investigation of Sosa's epistemology wi l begin to answer these questions.
Many have found the sensitivity principle plausi le because it captures an intuitive idea: that we want our belief-forming methods to track the truth in such a way that we are not led astray when we use them.Sensitivity is one way of having a tight connection between beliefs formed by our methods and the truth.A sensitive method M won't lead us astray because if it is sensitive, then the fo lowing is true: if p were false, then I wouldn't believe p using method M. Despite the intuitive attra iveness of this principle, it has faced many objections-most derived from classic objections to the truth-tracking theory of knowledge developed by Robert Nozick (1981), which employed a sensitivity requirement.(A) Every inductively-infer ed belief is insensitive, even when the inductive inference is very strong.To take an example inspired by Jonathan Vogel (1987): Knowing I am inebriated and a poor shot, I decide to attempt to hit the bu lseye with a dart by throwing the dart backwards over my shoulder while looking away from the dartboard.I am very unlikely to succeed.I am justified in believing that I wi l miss.But my belief is insensitive because if my belief were false-i.e. if I were to hit the bu lseye-I would sti l believe that I was going to miss.But even knowing my belief is thus insensitive, I seem justified in maintaining my belief that I wi l miss.(B) The principle wi l imply that justification is not closed under recognized entailment.(C) It implies that I can never be justified in believing that my belief that p ar ived at using M is not incor ect, because if that belief were false-i.e. if my belief that p ar ived at using M were incor ect-I would sti l believe otherwise (because, trivia ly, I would sti l believe that p and so sti l believe that that belief is not incor ect).This ap ears absurd.
Just as sensitivity principles in the theory of knowledge were superseded by safety principles, some have responded to these wor ies by employing safety epistemic debunking principles.Tomas Bogardus (2016) considers a debunking ar ument that implicitly employs the fo lowing principle:

Safety
K EDP: if a set of beliefs are unsafe given the way they are formed, then these beliefs do not count as knowledge, where a belief is safe given it is formed using method M just when if I were to believe p using M, then p would be true.
He then goes on to critique this principle by pointing out that knowledge need not be safe: imagine Jones has the world's most accurate atomic clock in his office which is accurate due to a sophisticated radiation sensor.That sensor, however, can be made to malfunction if a nearby radioactive isotope were to decay.Earlier this morning, someone left a sample of radioactive isotope near the clock in Jones's office.Unlikely enough, the isotope has not decayed.Jones is unaware of the isotope's presence.He wa ks into his office, checks his clock, which is functioning properly and accurately reads "8:22" and he thus believes it is 8:22.Bogardus contends that Jones clearly knows it is 8:22 despite his belief being unsafe, for in many nearby worlds he would have believed the time the clock read, but the clock would have been incor ect because it would have been made to malfunction by the isotope. 2f course, there are other safety-related epistemic principles one might consider using, for instance:

Safety
K2 EDP: once I become justified in believing that a set of beliefs are unsafe given the way they are formed, then these beliefs no longer count as knowledge for me, Safety J EDP: once I become justified in believing that a set of beliefs are unsafe given the way they are formed, then I lack justification in continuing to hold these beliefs.I won't discuss these now, but some of what I have to say later about Sosa's epistemological framework wi l be useful for evaluating these principles as we l.
So, there is some controversy about which epistemic principle(s) can plausi ly be used in debunking ar uments.Can Sosa's sophisticated epistemological framework sup ly such a principle?Before answering this question, we first need to lay out his framework.

Sosa and fully apt belief
According to Sosa's virtue epistemological theory of knowledge, knowledge is a certain sort of quality performance of belief (or affirmation, as he puts it).Performances in general can be evaluated for their accuracy (i.e.whether they achieve their aim), their adroitness (whether they were performed competently), and for their aptness (whether they were accurate because adroit).Beliefs are accurate when they are true, they are adroit when they are formed competently, and they are apt when they are accurate because they were formed competently-or, as he sometimes puts it, the accuracy of the belief manifests competence.Knowledge is apt belief.
Aptness differs from safety.A belief can be apt though unsafe-as i lustrated by the fo lowing example from Sosa (2017, p. 215-217).I can competently form a belief that my car wi l start a moment after I turn the key and that competence wi l then lead me to be cor ect, since my car actua ly does start.My belief is thus apt and constitutes knowledge.A l of this can remain true even if there is a mad bomber who wi l detonate an atomic bomb right when I turn my key if the coin he flips turns up heads (in fact, lucky me, it comes up tails).In these circumstances my belief ar ived at using my competent method is not safe because in half the nearby worlds my belief turns out to be false (the half in which he detonates the bomb, destroying me and my car and so my car, contrary to my belief, does not start).
The notion of a competence is central to Sosa's system.For him a competence is "a disposition to succeed when one aims at a given objective, in certain (favora le enough) conditions while in (good enough) shape" (Sosa, 2017, p. 193).A competence has a SSS structure-that is, a person's competence is composed of a ski l, the shape one is in, and one's situation.One's ski l is whatever it is inside one that grounds one's disposition to succeed in a certain range of conditions and situations.'Shape' refers to other personal chara eristics relevant to performance-being alert, sober, with one's eyes open, and the like.'Situation' refers to a ects of the environment that must be present for competent performance-for instance, if the competence is playing basketba l, the situation might include things like playing with a properly inflated ba l, on an indoor surface that is dry, with rims that aren't crooked.To be competent in a performance you must use your ski l in an ap ropriate shape and situation.
So, for Sosa, a belief is formed competently when it is formed using a belief-forming ski l, while in an ap ropriate shape and situation.Perception is one example of a belief-forming ski l.The ski l is the inner ability to form beliefs based on certain perceptual inputs.That ski l operates properly when in a certain range of shape and situation-e.g. while being sober and alert, in normal lighting and in normal atmospheric conditions (if one couldn't breathe properly one's ski l wouldn't operate properly and one's beliefs wouldn't likely be true).One has perceptual competence in forming a belief provided that one's perceptual ski l, when operating within the ap ropriate range of shape and situation, produces beliefs that are relia le (that is, safe for the most part).So, a belief can issue from a ski l even if the ski l that produces the belief is unrelia le-because, for instance, in fact one only uses that ski l in inap ropriate shapes or situations.And, again, a belief can issue from a competence and yet be unsafe (unrelia le)3 because, even though it does issue from a ski l in an ap ropriate shape and situation, it could have very easily been the case that your shape or situation was inap ropriate, resulting in a false belief.A competence (with a certain SSS structure) has to be relia le (safe) only in this sense: beliefs produced with this particular ski l would mostly be true in ap ropriate shape and situation pairs.Sosa accepts that there are levels of knowledge.Apt belief is what he ca ls animal knowledge, but humans often find good reason to reflect on whether their beliefs are apt.An apt recognition that a belief is apt is what he ca ls reflective knowledge.More recently, Sosa has made reflective knowledge subsidiary to a more important kind of knowledge-what he ca ls knowing fu l we l.One knows p fu l we l when reflective knowledge uides animal knowledge-in other words when an apt recognition of the aptness of belief-formation methods leads one to employ an apt method in coming to believe p (see Sosa 2017, p. 95-101;2011, p. 1-14).
Knowing fu l we l is important, according to Sosa.He points out that it is an instance of a valua le sort of performance-one that is not only we l done but also we l-chosen.In a dition, (A) the distinction helps explain why Gettier cases like the 10 coins case are clear cases of a lack of knowledge (because the agent in the case lacks both animal knowledge and knowledge fu l we l), whereas barn façade cases are more controversial (because the agent in the case has animal knowledge but lacks knowledge fu l we l).(B) Humans judge propositions and judgment is a state in which they consider which first-order attitude to take (belief or disbelief ).Judgement thus essentia ly involves considering whether a belief that p is apt and aims to come to an apt belief about whether the belief would be apt.(C) Humans re ularly have reason to take a perective on their beliefs and how they are formed, and their knowledge should reflect the reasons provided from this per ective.Knowing fu l we l is the knowledge that can result from taking this per ective.(D) The belief we come to at the reflective level should take precedence over the belief we come to at the animal level.If, at the reflective level, I competently judge that I should su end judgment about whether a belief about p would be apt (or if I competently judge that forming a belief about p would be inapt), then I should su end judgment about p.As Sosa states, "it would seem deplora ly stu born to sustain and endorse a belief in the teeth of total availa le evidence strongly against it.And this is so even if that belief is cor ect through diachronic first-order competence that is supremely relia le… once a belief is under scrutiny, only such reflective knowledge qualifies as a proper basis for conscious reasoning, pra ical or theoretical" (Sosa, 2017, p. 164-165).
Lastly, we come to Sosa's view on justification.He treats justification as necessary for knowledge.He does this by identifying justified belief with a ski lful attempt to get things right.If you employ a cognitive ski l in attempting to get the truth and come to a certain belief, then that belief is justified even if you fail to get the truth, and even if you weren't competent because you didn't use your ski l in a proper shape or situation.With this notion of justification he can explain why people who are being deceived by an evil demon have justified beliefs: they are genuinely using cognitive ski ls (perceptual faculties) to come to beliefs.They just hap en to be in a situation that is not ap ropriate for those ski ls, and so their beliefs are not competent.

CSR debunking in Sosa's epistemology
Debunking ar uments criticize the way a belief was formed, so they dwe l at the reflective level.Sosa's epistemology thus already provides a place for such ar uments and acknowledges their potential significance for our first-order beliefs.If, at the reflective level, we can acquire good reason to believe that our belief-forming method for a first-order belief that p is in some way i l-formed, then we shouldn't continue to believe that p.
We' l proceed by extra ing a series of epistemic debunking principles from Sosa and then discussing whether a successful CSR-based debunking ar ument for religious belief can be constructed using these principles.The centrality of competence to Sosa's system implies the fo lowing principle: Comp 1 EDP: if I come to know that my belief that p wasn't formed competently, then I lack animal knowledge, reflective knowledge, knowledge fu l we l, and justified belief that p.On Sosa's view knowledge, which is apt belief, requires competence.So, if I know my belief wasn't competent, I know I don't have animal knowledge of it.Thus, I also can't have reflective knowledge or knowledge fu l we l of it.I also won't have a justified belief because if I know my belief was formed incompetently, then I know that I did not form it in a way (using a ski l given my shape and situation) that was likely to uide me to the truth.And if I know that, then I won't be justified in continuing to hold on to the belief, for given what I know the likelihood of getting the truth would ap ear to be too low.
Sometimes it might be hard to know, although one has good justification for believing, that a belief wasn't formed competently.Comp 1 EDP would not be useful on those occasions.But there is a very similar principle that would be ap lica le: Comp 2 EDP: if I become justified in believing that my belief that p wasn't formed competently, then I lack reflective knowledge, knowledge fu l we l, and justified belief that p.
If I am justified in believing that my belief that p wasn't formed properly, then I do not aptly believe that p was formed aptly, and so I lack reflective knowledge and knowledge fu l we l.I wi l also lack justification for reasons similar to those given regarding Comp 1 EDP.However, I might sti l have animal knowledge-for I might have a justified false belief that my belief that p was formed incompetently.But this wi l be cold comfort because Sosa ar ues (as noted above) that our beliefs should align with our reflective-level assessment when we have such an assessment.So, when the antecedent of Comp 2 EDP is satisfied, I might sti l have animal knowledge that p, but I nevertheless shouldn't rationa ly continue to believe p.I should su end judgment even if I might we l be a le to anima ly know that p.
Our next question is this: do CSR theories, assuming belief in them is we l-justified, give us good reason to think that religious beliefs are formed incompetently?To answer this question, we first need to chara erize the religious belief-forming process, as theorized in CSR.Then we wi l need to examine whether that process constitutes a ski l and whether we are in the right shape and situation for the ski l to operate relia ly.
CSR theorists have developed several different theories of why humans have religious beliefs.Byproduct theories hold that religious beliefs commonly result as a non-selected byproduct of the operation of cognitive processes that have been selected-for (Bar ett, 2004;Boyer, 2001).Adaptive the-ories hold that religious beliefs are individua ly or co lectively adaptive and so have been cultura ly selected-for (Wilson, 2002;Bering, 2011).And then there are exaptation theories which hold that religious belief entered the human scene as a byproduct, but was later selected-for (Norenzayan, 2013;Bu bulia, 2007).I'm not going to ar ue for one of these views.Fortunately, I don't need to; the philosophical points I make wi l ap ly mutatis mutandis to whichever theory you like.So, I'm just going to pick the exaptationist theory, in part because it combines the explanatory hypotheses of the other two kinds of theories.It is thus particularly useful for my purposes because it ena les me to, in a way, be most generous to CSR-a lowing the debunker to assume a robust, richly developed, multifarious explanation of religious belief.
Here then, to briefly and bruta ly summarize a lot of interesting work, is the theory we sha l accept of why humans have religious beliefs.Humans have a variety of cognitive mechanisms that incline them towards belief in invisile agents: a highly sensitive agency detection module that is prone to fire at the slightest indication of agency (Bar ett, 2004); a preference for spreading and considering minima ly counterintuitive concepts-which include concepts of invisi le, nonphysical agents (Boyer, 2001); a mechanism for attributing and eculating about the mental states of others that operates independently of mechanisms for thinking about the behavior of bodies-and this mechanism tends early in life to attribute a lot of knowledge to agents (Bering, 2011;Bar ett, 2012); and an inclination toward teleological reasoning about the features of the physical world (Kelemen, 1999(Kelemen, , 2003(Kelemen, , 2004)).A l of these mechanisms make it so that we can very easily think about invisi le agents, their minds, and their desires, and we are prone to ta k about them and to consider them as explanations for various events and a ects of the physical world.Certain of these concepts are socially adaptive-namely concepts of powerful, knowledgea le, nonphysical agents who care about what we do.If we believe in these beings-gods, we' l ca l them-we wi l feel watched, which wi l encourage pro-social behavior and discourage free-riding.This is genera ly adaptive.But then, in a dition, we wi l begin to provide signals of being particularly devoted to the gods so that we can find and cooperate with those people who are most likely to cooperate with us, thus further benefiting us.A community that has a decent number of people who believe in such gods wi l thus thrive, and the belief in these gods wi l spread (Norenzayan, 2013).
Given this theory, how should we chara erize the human religious belief-forming process?Notice that this theory is a theory of why religious beliefs, particularly beliefs in invisi le agents like gods and ance or spirits, are widespread in humanity.But it doesn't propose to explain why any particular human holds religious beliefs.We sha l draw further attention to this point later.But given the explanation ap eals to what are proposed as roughly universal human chara eristics and situations, if the theory is true, then proba ly many if not most humans' religious beliefs are influenced by the various factors the theory mentions.This process ap ears to work as fo lows: it takes as input (i) various experiences of the physical world that seem to involve agency where there isn't a clear physical agent, and (ii) testimony about possi le invisi le agents and then, through the mechanisms described above, produces or maintains a belief in an invisi le agent.The shape one must be in for the process is just whatever shape a lows these mechanisms to function-so having normal use of your perceptual faculties, being alert, not suffering from various disabilities.The situation for the proper operation of this process would ap ear to be the situation in relation to which these mechanisms evolved: an environment with a wide range of agents and obstacles that make cooperation beneficial.
Does this process count as a ski l-that is, is it a disposition to produce beliefs that are mostly true in the proper shape and situation?The process proba ly is relia le at producing beliefs about agents in general-when and where there are agents in the environment and what those agents might intend and believe.But we're intere ed in a more ecific set of beliefs-beliefs about invisi le agents.A process could produce relia le beliefs of type B, but also produce unrelia le beliefs of type C (even if C is a subset of B).
This process produces lots of different beliefs about invisi le agents.Some people believe in ance or spirits, sprites, ghosts, others in gods of various sorts, and some in one supreme god.Almost surely not a l of these things exist.But even if many of them exist, the process as described doesn't track which ones exist; it' l produce a belief involving whatever kind of invisi le agent concept hap ens to be floating around one's community (provided it is minima ly counterintuitive).So, it would be pretty easy for this process to ar ive at a false belief, thus it ap ears not to be a ski l.
But what if we consider the more general belief, "there is some sort of invisi le agent"?Might the process produce beliefs like this that are mostly true?We l, it is hard to answer this question without some independent way to evaluate whether there are invisi le agents and, by hypothesis, we're imagining that the only way we have is the CSR belief-forming process.One might thus ar ue that when we move to the reflective level we can rely on a default entitlement to fo low our basic belief-forming methods, which confirm themselves, and remain untrou led by this sort of circularity (as Sosa has long ar ued; see e ecia ly Sosa, 2009).Thus, at the reflective level we are not justified in thinking that religious beliefs are formed incompetently through the CSR process, and so we can't employ Comp 2 EDP in an ar ument to debunk religious beliefs.In reply, one might note that the belief that there is some sort of invisi le agent is presuma ly infer ed from another belief in some particular invisi le agent.Since that latter belief is not relia ly formed, any belief logica ly infer ed from that can't be any epistemica ly better than the epistemica ly flawed belief it is based upon.By way of rejoinder, however, one might su gest that perhaps the CSR mechanism sometimes directly produces more general beliefs like, "some sort of invisi le agent is at work here." And if that is so, then we are brought right back to the ar ument of the previous paragraph.In a dition, one might ar ue that if a god like the Judeo-Christian God exists, then he would proba ly intend for humans to form beliefs through this process that there are invisi le agents.Thus, part of the situation of the process would include God's existence and intention.But once again, by hypothesis, we don't have any independent means of determining whether that situation obtains, and so we are again brought back to the ar ument of the previous paragraph.Debunking ar uments based on Comp 2 EDP ap ear sunk.
Sosa's epistemology gives us another debunking principle that may we l be more successful, however.The principle goes as fo lows: Apt EDP: if I become justified in believing that my belief that p wasn't formed aptly because I might easily have formed it incompetently, then I lack reflective knowledge, knowledge fu l we l, and justified belief that p.
Return to Sosa's example of my belief that my car wi l start, ar ived at while the mad bomber flips his coin.Let's modify the example slightly: now I know about the mad bomber and his plan (but not how the coin flip comes out).In this modified example I have good reason to doubt whether my belief that my car wi l start is apt because I have reason to doubt that the background conditions of the situation wi l continue to hold.My belief might turn out to be competent and apt, and thus amount to animal knowledge (if the bomber's coin flip te ls him not to detonate), but it might not and the risk of the latter is pretty high.So, at the reflective level, I can't te l that my belief is apt, thus I lack reflective knowledge, knowledge fu l we l, and justification (because there is a sufficiently high chance of forming the belief incompetently). 4 In short, once we rise to the reflective level we want good reason to think that we are going to employ a first-order process that would likely enough be apt.If we lack such a reason, then (for the reasons given above) reflective knowledge, knowledge fu l we l, and justification are absent.Let me give one more example to i lustrate this point.In the film Total Recall, Douglas Quaid purchases a "virtual vacation" to Mars through the Reka l company. 5The vacation is virtual because the company merely implants in the "vacationer" a bunch of false memories as if the vacationer had gone on vacation.However, while Quaid is hooked up to the machine that is about to deliver him false memories, he tries to escape-believing that he is a secret agent whose cover wi l be lown.He proceeds, without undergoing the false memory implantation-or so it seems-to go on an adventure that takes him to Mars.At the end of the film he wonders whether a l the experiences he has had on his adventure rea ly hap ened, or whether they were themselves implanted vacation memories.
Take one of Quaid's beliefs ar ived at during his adventure-"I have visited Mars." This belief of Quaid's is not one that he reflectively knows, fu l we l knows, or justifia ly believes.He has good reason to doubt that he formed it competently for he might we l have had the procedure at the Reka l company and if he did, then his memory isn't competent for this belief.But he might rea ly have escaped the company before having the procedure (as he seems to remember), in which case his memory is competent and he could have animal knowledge that he visited Mars.But from what he can te l there is a good chance of either.So, there is a good chance he formed the belief incompetently, thus unrelia ly in the circumstances.These grounds for doubting whether his memory is competent indicate that he rationa ly shouldn't trust his memory, and so he should su end judgment (again even if he in fact didn't have the procedure and thus is a le to have animal knowledge).
This example of Randy Quaid is useful because we can give a para lel sort of ar ument regarding religious belief (on the assumption that the CSR process is what produces humans' religious beliefs).We have some independent reason to doubt that God exists-the pro lem of evil, for instance.Now, I realize that my belief that God exists issues from the CSR process.I'm like Randy Quaid-if God exists, as the process itself te ls me, then my process is competent (for the reasons given in my discussion of Comp 2 EDP) and can easily be apt (just as Quaid's belief can be competent and apt if he rea ly did escape the company and get to Mars, for his memory would then competently deliver him a true belief that he had escaped).But if God doesn't exist, then the CSR process is incompetent (again, for the reasons given above).Given the independent reason to doubt God's existence, it (epistemica ly) could turn out either way.I thus have good reason to doubt that the CSR process that issues in my belief in God is competent, so I shouldn't trust it.Belief in God, as we l as other religious beliefs issuing from the CSR process, are thus debunked.And this ar ument rests on Apt EDP.
This debunking ar ument relies on the claim that there is at least some good reason to doubt that God exists.But it also essentia ly relies on the description of the CSR process.Other sorts of belief-forming processes wouldn't be debunked in this way-for instance reliance on some classic ar ument for God' s existence.The fact that we have reason to doubt that God exists doesn't give us any reason to believe that we are incompetent at assessing ar uments for God' s existence.So, if we had such ar uments that we found plausi le, they would count as good reason to believe and then would just need to be balanced off against the reason to doubt God' s existence in order to determine whether belief in God was justified. 6his brings me to the final move in my examination of the debunking ar ument.We've seen that Sosa's epistemology, via Apt EDP, does ground a plausi le debunking arument for religious belief on the assumption that religious belief is formed via the CSR process described earlier.But I don't think that the CSR process completely captures most believers' belief-forming process because most believers in a god have fairly ecific beliefs about that god and those ecific beliefs are sup orted by various ar uments that believers typica ly find plausi le-e.g.cosmological ar uments, design ar uments, ar uments from miracles and religious experience, and the like and reliance on testimony from those who are familiar with these ar uments.I'm not saying that every religious believer has carefu ly thought through these ar uments; and I don't think one needs to do so to be reasona ly persuaded by an ar ument.Often with ar uments for God's existence there is some phenomenon that is taken to be a sign of God's existence.There are many ways of trying to ar ue from that sign to God's existence-many bad ways, possi ly some good ways.I think that as long as there are some good ways and the phenomenon rea ly is a sign of God's existence, it can be rational to believe on the basis of that phenomenon even if one hasn't carefu ly thought through ecific versions of theistic ar uments.7Many people who believe in God do believe at least in part on the basis of the phenomena that are taken to be signs of God's existence.It is only in this sense that I say that most believers believe at least in part on the basis of ar uments.And as I just ar ued in the previous paragraph, the debunking ar ument does not succeed in debunking our ability to evaluate these ar uments.And so the rationality of believing in a god in this way-perhaps primed by the CSR process, but boosted by ar uments-isn't debunked. 8