Testimony and the selection of belief-forming practices

The first problem with focusing on conscious awareness is that religious practitioners report conscious awareness of God. The reason this is a problem for Schellenberg is that testimony is an established socially accepted practice (which Schellenberg (Reference Schellenberg2011, 170 ff.) accepts), and in many contexts can confer justification and knowledge. To assert that it does not confer justification in the religious case is to prejudge the truth of such claims. To treat religious claims prejudicially threatens non-resistance. To be fair, Schellenberg's reason for discounting the testimony of others is based on his argument for discounting the legitimacy of the religious believer's first-person experiences. Aside from the problems I note in the next sub-section, this claim puts Schellenberg in the odd position of telling others which of their seemings are justified and which ones are not.

The first stage of the argument aimed to defeat the veridicality of religious experience is to consider which belief-forming processes we should trust. In general, we should trust only those sources of belief formation that are innocent until proven guilty. Epistemic processes that include religious practices, such as prayer, are not innocent for Schellenberg. He states, ‘if we really are would-be investigators concerned for the truth and seeking understanding, then we will ascribe epistemic innocence . . . only where we have to’ (ibid., 170). And we only have to rely on sensation, introspection, memory, reasoning, and intuition. On Schellenberg's view, religious belief-forming practices do not make the cut ‘since the outputs of religious experiential belief-forming practices conflict, and thus not all such practices can be reliable’ (ibid., 171). Because some religious practices are epistemically unreliable, religious practices fail Schellenberg's universality requirement, which states that we choose belief-forming practices that are universally accepted as not unreliable (we cannot know they are reliable without circularity (Alston (Reference Alston1993), ch. 3)).

On this reasoning, however, we should not trust our moral reasoning since our moral belief-forming practices yield incompatible beliefs. We should not trust those epistemic practices and faculties that yield philosophical beliefs either since they too conflict. The scientific method might not be a faculty, but it is a belief-forming process that yields different empirically testable conclusions. We should not cease using it though. We should distinguish between whether a faculty or practice is overall reliable, and whether we have a choice in using it. It can be the case that a cognitive faculty, such as our moral faculty, is not impressively reliable, but we must use it just the same when and always when we engage in moral enquiry.

Schellenberg is correct that disagreement can exert epistemic effects, specifically, peer disagreement provides metacognitive evidence that either I or my peer have made a processing error. It does not follow, however, that I should not use that same faculty in my enquiry. If I discover that someone comes to a different position on, for example, progressive taxation, I should enquire further into that issue using the very same faculties that I have been using. I just use those faculties more carefully and comprehensively. So, neither supposed unreliability nor peer disagreement justify not using religious belief-forming practices.

Schellenberg is also correct to point out that in the setting of conflicting religious beliefs, all that follows is that not all the religious belief-forming practices are reliable. Here again, it does not follow from this claim that religious belief-forming practices should not be used at all.Footnote ¹⁰

There are, of course, reasons for using religious belief-forming practices when the domain of enquiry concerns personal knowledge of God. That argument is this: we tailor our epistemic practices to the nature of the domain of enquiry. If my enquiry concerns what we morally ought to do, I will use epistemic practices ordered to sharpen my intuitions on one's moral obligations. If my enquiry concerns what people's moral beliefs are, I will use epistemic practices ordered to survey what their beliefs are. If the domain of enquiry is a wholly other divine being whose nature is non-physical, Personal, and the ground of all being Who has purportedly revealed Himself through certain religious traditions, then I would conform my belief-forming practices to what these traditions recommend. This is not a random choice of epistemic methods. Religious practitioners are submitting to what God purports to have established. And we have every reason to think that God would prescribe certain practices as means to bequeath reverence and humility. James Cutsinger explains that union with God is certainly possible, but we should expect that such a union is achieved through a framework willed by God, and not by our own initiatives ‘for our liberation from the ego and the world can take place only by means of what enters into that world from above, and not by the necessarily individual initiatives of the ego itself’ (Cutsinger (Reference Cutsinger1994), 193). And most religious traditions recommend prayer, participation in liturgical practices, etc.

So, the preclusion of religious belief-forming practices is biased from the start. The argument for such preclusion does not follow from the arguments Schellenberg gives. In general, it is hard to argue prior to one's enquiry that whole belief-forming practices should be precluded without assuming that the beliefs they produce are false (and all belief-forming practices cannot be proven reliable without circularity, as argued by Alston (Reference Alston1993)). The next section descends closer to the core critique I wish to offer according to which either non-resistant non-belief does not exist, or if it does, it is not of the sort that is incompatible with God.

Religious experience and circularity

The problem highlighted in the previous sub-section is that Schellenberg's selection of belief- forming practices is biased or entails counter-intuitive results. The problem highlighted in this sub-section is even more severe, namely, his argument against the epistemic worth of religious experience must assume a naturalistic explanation, but assuming a naturalistic explanation is incongruent with non-resistance.

Schellenberg does not consider the evidential force of testimony from religious practitioners remarkable because he thinks that first-person experiences have little epistemic standing. The argument begins by making clear the distinction between first-person and third person justification. ‘Third-person justification are arguments to the effect that the occurrence of religious experience throughout the world is best explained by reference to a divine source’ (Schellenberg (Reference Schellenberg2011), 160). Of course, explaining the distribution of religious experience throughout different traditions is not the only way to infer that God exists. One could accept on trust or testimony from a religious practitioner that God exists. Accepting testimony does not require arguments with an explanatory structure (Coady (Reference Coady1992)). In any case, first-person justification is the justification one has given one's own experiences. Even for first-person justification, however, ‘[r]eligious experience affords no immunity to religious skepticism’ (Schellenberg (Reference Schellenberg2011), 162). The reason is that it is ‘always appropriate . . . to remember the possibility of misperception’ (ibid., 165). And it is the possibility of this misperception that serves as a challenge according to which: ‘For the experience to be taken as undefeated evidence, that challenge [misperception] must be turned aside’ (ibid.).

This initial attempt to file off the bite of an argument from religious experience lands Schellenberg in incoherence. Arguably, conscious apprehension of God is a token instance of a religious experience. But if our religious experiences are so epistemically fragile as he suggests they are, (i.e. one must turn aside the possibility of misperception to be justified), then why would a good and loving God desire conscious apprehension? To introduce a common motif in this sub-section, if non-belief is a function of there being no conscious apprehension of God and such an apprehension of God is epistemically fragile, then such non-belief fails to count as evidence against God's existence. God would not desire to bequeath such experiences as they are not, on Schellenberg's own argument, true epistemic goods, that is, they are not a source of justification.

Maybe there is a good reason that misperception is always a possibility, or that one must defeat all other non-theistic explanations for a religious experience before taking such experiences as veridical. I resist interpreting Schellenberg here as assuming that religious experiences are unveridical from the start of his enquiry. That would be incongruent with non-resistance, a feature of which requires that the seeker is open to belief in God, and in some places, is described as wanting to believe in God (quotations above). It is better to understand Schellenberg as having a hermeneutic of suspicion towards religious experiences, not that they are unveridical.

To justify this suspicion, Schellenberg focuses on (i) neurological/naturalistic explanations, (ii) religious diversity, and (iii) peer disagreement. Neurological explanations claim that certain brain states are associated with mystical type experiences. Schellenberg endorses Evan Fales' explanation according to which, mystical experiences ‘are associated with micro-seizures of the temporal lobes of the brain . . . But mild seizures, which can be artificially induced during brain surgery can result in powerful mystical experiences’ (quoted from Schellenberg (Reference Schellenberg2011), 186). Schellenberg infers from this data that ‘if [mystical experiences are associated with mild brain seizures] is correct, then it is clearly possible for mystical experiences to occur even when they are unveridical’ (ibid.).

What is supposed to follow from such observations? It does not follow that such experiences are unveridical. First, even if it is possible that mystical experiences can be invoked by mild seizures in the brain, it does not follow that they all are. Even if they all are, it does not follow that a religious experience is unveridical. Given the current evidence in neuroscience, almost every experience is associated with a brain state defined in high enough resolution of neural activity. Suppose I am under an fMRI and I am asked to think about a priori truths; certain areas of my brain will become activated. Those truths are still truths, even if they are associated with brain states (Plantinga (Reference Plantinga2000), 145).

Furthermore, there is reason to suppose that God could choose, according to the mode of the knower, that a veridical experience of God is mediated by seizure like activity in the temporal lobe. If God made us, neurological architecture and all, God would be expected to disclose Himself by way of our embodied existence; God would reveal Himself by activating certain centres of the brain. In any case, there is no reason to suppose that religious experiences can be veridical only if they are experienced apart from any neurological activity.

If these seizures are distributed as widely as Fales indicates they are, that could just as well be an argument that God has distributed this feature in the temporal lobe to experience more easily mystical like experiences. This could count as data that God has distributed mechanisms by which to apprehend God widely throughout the population. If so, the evidence in favour of the hiddenness argument significantly truncates. The point is that a religious experience (or any experience) can remain veridical even when mediated by brain areas.

The argument so far is that we do not have a purely naturalistic explanation of a religious experience (or any experience) by simply noting that it is associated with a brain state. God's self-disclosure could be by means of brain states, and the association between an experience and a brain state is compatible with the experience being veridical. One would have to argue that such experiences were ‘nothing but’ or ‘merely’ neurological states. What would such an argument have to look like? Quite obviously it won't do to presuppose that the experiences are unveridical and infer from that assumption that the neurological state associated with the experience is merely neurologically induced.

Is this how Schellenberg's reasoning for (i) must go? Running on from the quotation from Fales, Schellenberg says the following: ‘If this is correct, then it is clearly possible for mystical experiences to occur even when they are unveridical. Indeed, given the possibility of deliberate induction, we would have to say. . . that mystical experiences do occur even when they are unveridical’ (Schellenberg (Reference Schellenberg2011), 186). The reference to deliberate induction refers to the apparent fact that such mystical experiences can be induced during brain surgery. Notably, other experiences can be induced during brain surgery such as seeing a light in the top corner of one's visual field. That ‘deliberate induction’ does not render one's visual experiences of light outside the surgical suite unveridical. In any case, Schellenberg concludes as follows,

Schellenberg wants to know what truly causes religious experiences. I have already noted that it is fallacious to infer from the claim that certain brain states are associated with mystical experiences, to the conclusion that such experiences are unveridical. I am more interested here with Schellenberg's inference that if such explanations are possible, the epistemic state of doubt is a reasonable response. More generally, under what conditions does an agent have a defeater to the putative veridicality of her experiences?

The chief problem with (i) in this context is that non-resistance is unlikely to be compatible with assuming a naturalistic explanationFootnote ¹¹ for religious experience. The basic idea with this riposte is captured by C. S. Lewis, who, in autobiographical mode, relates that he has known only one person who claimed to have seen a ghost. It happened to be someone who did not believe in the immortality of the soul. Though the experience felt veridical, she interpreted it as an elaborate hallucination. Here is a person whose perception of a putative ghost appeared veridical, and yet it failed to invite belief. Lewis observes that whatever ‘experiences we may have, we shall not regard them as miraculous if we already hold a philosophy which excludes the supernatural . . . We can always say we have been the victims of an illusion’ (Lewis (Reference Lewis1970), 8). So too with naturalistic explanations of religious experience. If we assume naturalistic causes, those experiences have no epistemic standing. Privileging naturalistic explanations, however, is incongruent with non-resistance.

Lewis' reflections introduce a theme according to which it is rather easy to defeat the putative justification of a religious experience, that is, simply assume a naturalistic explanation. Whether this is congruent with non-resistance requires discerning when a belief suffers defeat. Broadly speaking, a defeater D for a belief p is any proposition that renders p unjustified. Of course, to function as a defeater, the agent must hold, believe, or consider D. If an agent thinks that D is implausible, or at least, less plausible than p itself, then D cannot function as a defeater to p – not for that agent anyway. Creationism (understood crudely) is a defeater to certain aspects of modern evolutionary theory, but the former position does not function as a defeater to evolutionary theory for most biologists. So, if a naturalistic explanation (hereafter Ne) for one's putative veridical experience is to function as a defeater to its veridicality, one must believe (Ne) to an equal or greater certainty that one has or had a veridical religious experience (hereafter Ve). If Ne is merely possible, it cannot function as a defeater.

Assessing a defeater as of ‘equal or greater certainty’ is a delicate matter. From some of what Schellenberg says, already quoted, it appears that merely having the attitude of considering a defeater to Ve renders Ve unjustified for that agent. But I might consider any number of sceptical hypotheses for my experiences and everyday beliefs ranging from global hypotheses such as Cartesian evil demon scenarios, to more specific beliefs, such as flat Earth hypotheses or Marxist hypotheses for my everyday moral beliefs, such as that pederasty is immoral. In each case, I do not abandon or think that my target beliefs lose any of their justification. I do not, for instance, think that my belief that pederasty is immoral is epistemically compromised by entertaining Marxist suspicions of traditional morality. Again, the mere presence of possible alternative explanations for my beliefs does not function as a defeater for those beliefs. Moreover, there is no reason that I should consider an experience with the depth and felt veridicality that typifies some religious experiences as unveridical until proven innocent. I do not treat my other perceptual beliefs with such a hermeneutic of suspicion. Singling out religious experiences for greater scrutiny is incongruent with non-resistance.Footnote ¹²

So, we might distinguish between an agent considering a sceptical hypothesis in which the intellectual goal is understanding the contours of that hypothesis; and in which the goal is something like checking my beliefs against others to achieve wide reflective equilibrium (Rawls (Reference Rawls1971), 20ff.) according to which belief revision is a distinct possibility. When the former goal is in mind, Ne does not function as a defeater to Ve; it may if the latter goal is in mind.

What may follow is that a defeater D exerts its sceptical effects only if D is believed or considered (with the goal of achieving reflective equilibrium), to be equal to or more likely than the target belief. If Ne is held with equal plausibility, then evidential parity exists, and this is compatible with God supplying the necessary conditions for belief. As yet, there is no violation of what God would do on Schellenberg's terms. If Ne is held to greater plausibility than Ve, then either this greater plausibility is assumed, or it is based on independent evidence. If it is simply assumed, the seeker is likely to be resistant. Again, non-resistance is characterized as being open to belief in God and in which the agent would ‘love’ (Schellenberg (Reference Schellenberg2011), 205) to believe and would put herself in a place ‘to be convincingly . . . met by evidence of God’ (Schellenberg (Reference Schellenberg2015), 56). Such epistemic desires are incongruent with assuming Ne for any putative conscious apprehension of God.

What follows if the greater plausibility is a function of independent evidence? Let's assume for simplicity and concreteness that this independent evidence is the result of adjudicating the project of natural theology. If Ne enjoys greater epistemic probability because of this adjudication, the hiddenness argument is redundant – the existence of a Personal God is adjudicated on other grounds. Furthermore, if Ne enjoys greater plausibility in the agent's noetic system than Ve – on the basis of evidence other than the hiddenness argument – then Lewis's reflections above apply; to wit, the experience could be veridical, namely, an actual conscious apprehension of God occurs, and yet the agent draws the wrong conclusions, or fails to interpret the experience correctly. What else could a good and loving God do? Hiddenness, on this disjunction, is a self-fulfilling prophecy in which any putative Ve can be reinterpreted with reference to some Ne. This argument does not entail that atheists are self-deceived. What follows on this disjunct is that the hiddenness argument does not function as an independent argument for them.

Either one of two conclusions follow. Plausibly, holding any Ne to a greater epistemic probability than Ve is resistant.Footnote ¹³ So, the failure of a seemingly veridical religious experience to deliver the epistemic goods is not a case of non-resistant non-belief. However, quibbles about the borderlines of resistance can be intractable. So, another plausible conclusion of the argument is to assume that such non-belief counts as non-resistant. It is not, however, incongruent with what a good and loving God would allow (or could prevent). What else could God do if both the agent holds Ne to a greater plausibility than Ve and God aims to provide merely necessary conditions for belief? God does exactly what Schellenberg recommends, namely, God provides a religious experience, but if that agent thinks that a possible Ne defeats Ve, as Schellenberg recommends, the agent will never interpret or understand that experience as being of God. The agent is like Lewis's friend, ‘there must be a naturalistic explanation’. Either option deflates the force of the hiddenness argument. Either the non-belief is not non-resistant, or it is non-resistant but not the sort that is incompatible with what God would allow or could prevent.

Does this mean that anyone who is suspicious of one's putative religious experience (either one's own or others) must be culpably resisting theistic belief? The answer is ‘yes’ if one is privileging Ne as an assumption. The answer is ‘no’ if one already thinks that Ne is more plausible on independent grounds. But for the latter option, no amount of a dazzling religious experience could offset Ne, per Lewis's reflections. Thus, God could not do otherwise, and such non-resistant non-belief is not incongruent with what God can be expected to do.

The above disjunctive arguments are, however, not the strongest point that can be made – even though I think they are valid and sound. Quibbles may occur over the inference, if an agent S assumes a naturalistic explanation for any Ve, then S evinces resistant non-belief. What is incontrovertible, and essential for my argument, is that if Schellenberg's argument against the evidential force of religious experience is correct, such experience is no longer the epistemic good that we should expect God to bequeath. If for any given conscious apprehension of God, it is defeated simply in virtue of considering possible naturalistic explanations, then it cannot function as the epistemic good Schellenberg supposes it to be in his hiddenness argument. So, if Schellenberg's argument against the evidential force of religious experience is correct, the hiddenness argument fails – a loving God would not be expected to bequeath conscious experiences. If his argument against religious experience is incorrect or weak, then the reports of religious practitioners and their epistemic practices should be recognized as bequeathing epistemic goods, which can be enjoyed by trusting their testimony or by engaging in such practices oneself.

As for (ii), religious diversity is an odd objection to the truth of religious claims. We do not discount the occurrence of certain events in history even if historians disagree about the specifics of those events, for example the one gunman versus multiple gunman theories concerning the Kennedy assassination. No one infers that Kennedy was not assassinated. If one person reports that the blue car hit the red car first, and another person reports that the yellow car hit the red car first, we do not conclude that an accident did not occur. On such matters as religious diversity, one should recall the Thomistic dictum according to which knowledge is in the knower according to the mode by which the knower knows. Armed with this epistemic wisdom, it is expected that religious believers in different sociocultural contexts, in widely variant epochs, with distinct mores, customs, meaningful symbols, and dominant virtues, and with radically different intellectual heritages, will experience one and the same God quite differently. But even so, there are incredible similarities in religious experience (see Otto (Reference Otto1958)), and there is a good argument to the effect that the diversity of religious experience is willed by God to achieve a transcendental unity starting from terrestrial plurality (see Schuon (Reference Schuon1984)).

Turning finally to (iii), namely peer disagreement, Schellenberg supposes that one must ‘turn aside’ defeaters partly because there are those who disagree. Schellenberg seems to think that the hypothesis of misperception is strong enough to require attention because there are those whose intellectual abilities are similar to a religious practitioner's but they do not have such experiences. Again, peer disagreement exerts an epistemic effect that supposedly undercuts the justification a religious practitioner has in her religious experiences. Here is Plantinga responding to a similar point.

An epistemic peer is partly defined as one who is engaged in the same belief-forming practices as you, and those practices comport with the kind of enquiry in question. As pointed out by Plantinga, if someone else is not praying, participating in liturgical practices, informing oneself of religious insights, then there is no reason why their absence of religious conviction exerts any sceptical effects on one who is engaged in such practices.

One may retort by observing that religious experiences are astonishing, or unusual. The idea might be that when I tell someone that I saw an accident on a busy street, no one is suspicious; when I relate that I saw phosphorescent wheels slightly below the surface of the sea while sailing (Coady (Reference Coady1992), 190ff.), many would be suspicious. Religious experience, Schellenberg may say, is closer to the latter than the former. But notice, to assess religious experience as astonishing, one must already think that the objects of the experience probably do not exist. Such an assessment cannot itself be justified by divine hiddenness considerations without vicious circularity. It does not help to assert that the default position for interpreting religious experience is a naturalistic one. Again, to assume naturalistic interpretations is to resist, even on Schellenberg's understanding of non-resistance, which requires openness to believing. To think that they should be privileged because religious experience is not veridical is logically circular.

There is a tension here within our own intuitions. On the one hand, the more dazzling the experience may be, the more I will think that it is unusual and I would be more inclined to treat it sceptically. Knowing this, God is justified in not disclosing Himself in such dazzling ways – and Schellenberg rightly recognizes this point. On the other hand, the more humble, mundane, or terrestrial the experiences are, the less I will think they are religious. Consider an everyday example such as entertaining an objective moral truth. C. Stephen Evans points out,

Of course, one might not see a moral truth as an expression of God's will for our lives. Judging from the ripostes to the moral argument for God's existence, there is no reason why one should not since they all exploit some version of the claim that the epistemic authority of moral truths can be explained naturalistically. But refusing to see moral truths as an expression of God's will for my flourishing fails non-resistance if one doubts because of a possible naturalistic explanation.

To summarize, here is how to think about what is fundamentally problematic with Schellenberg's riposte to the argument from religious experience. If we pair Schellenberg's articulation of the hiddenness argument, i.e. its focus on the absence of conscious awareness, and his rejoinder to the epistemic standing of religious experience we get the following pastiche of claims. Schellenberg argues that conscious awareness of God is something to be expected in a universe in which a good and loving God exists, but he also thinks that such an experience holds little epistemic weight. This poses a dilemma. Either conscious awareness of God is epistemically important enough for a good and loving God to make available to all non-resistant seekers or it is not. If Schellenberg thinks that religious experience has no epistemic standing, then there is no argument for thinking that a good and loving God should make that experience available to non-resistant seekers. Conversely, if religious experience of either the dazzling or humble sort has epistemic standing, then there is no reason for not trusting the testimony of religious practitioners for they are readily at hand (Davis (Reference Davis1989)). One would not be non-resistant by disbelieving their reports. One might think that such experiences should be distributed equally. But for such a riposte to function as an argument against God's existence it must endorse something like the following inference ‘if I have not experienced God, God does not exist’. My understanding of Schellenberg's work, in particular his reflections on scepticism (Schellenberg (Reference Schellenberg2011), 15–49), suggest that he would not endorse this inference on grounds that it is epistemically narcissistic.Footnote ¹⁴