Against Quasi-Fideism ^†

Abstract

In a series of papers, Duncan Pritchard has argued that Wittgenstein’s remarks in On Certainty (OC) provide the foundation for a quasi-fideistic account of religious belief. This account seeks to navigate a middle path between evidentialism—the view that religious belief is rational only if supported by epistemic reasons—and fideism, which holds that religious belief cannot be rationally justified precisely because it is not grounded in epistemic reasons. Central to Pritchard’s quasi-fideism is a parity argument, which asserts that religious and non-religious beliefs are on equal footing. For the rationality of both depends on their connection to a background of a-rational foundational commitments, or “hinges”. In this paper, I challenge Pritchard’s account on two fronts. First, I argue that his quasi-fideism relies on a contentious interpretation of OC, and of Wittgenstein’s views about religious discourse, making it questionable whether it can genuinely be considered an application of Wittgenstein’s views to religious epistemology. Second, I contend that quasi-fideism, as developed by Pritchard, is inherently unstable, as it risks either collapsing into fideism or permitting relativism, contrary to Pritchard’s intentions. I then explore the potential for quasi-fideism—or, at any rate, a position meant to steer a middle path between evidentialism and fideism—under an alternative interpretation of OC. While this alternative framework may offer better prospects than Pritchard’s original formulation for avoiding the pitfalls of fideism and relativism, while sidestepping evidentialism, I argue that it ultimately fails to support a parity argument between religious and non-religious beliefs. Consequently, it cannot sustain a religious epistemology that avoids both evidentialism and fideism.

In a series of papers, Duncan Pritchard (2000, 2011, 2015, 2017, 2024) argued that Wittgenstein’s (1969) remarks in On Certainty (OC) provide the foundation for a quasi-fideistic account of religious belief. This account seeks to navigate a middle path between evidentialism—the view that religious belief is rational only if supported by epistemic reasons—and fideism, which holds that religious belief cannot be rationally justified precisely because it is not grounded in epistemic reasons. Central to Pritchard’s quasi-fideism is a parity argument, which asserts that religious and non-religious beliefs are on equal footing. More specifically, the rationality of both depends on their connection to a background of a-rational foundational commitments, or “hinges”.

In this paper, I challenge Pritchard’s account on two fronts. First, I argue that his quasi-fideism relies on a contentious interpretation of OC, and of Wittgenstein’s views about religious discourse,1 making it questionable whether it can genuinely be considered an application of Wittgenstein’s views to religious epistemology. Second, I contend that quasi-fideism, as developed by Pritchard, is inherently unstable, as it risks either collapsing into fideism (Section 1 and Section 2) or permitting relativism (Section 3), contrary to Pritchard’s intentions.

I then explore the potential for quasi-fideism—or, at any rate, a position meant to steer a middle path between evidentialism and fideism—under an alternative interpretation of OC, which draws on Coliva (2025d) On this reading, non-religious hinges can be categorized as either de jure or de facto. This distinction allows for a nuanced understanding of hinges as rational in distinct ways: either because they are constitutive of epistemic rationality, despite lacking evidential justification (de jure), or because they are so much supported by evidence that they now function as norms of inquiry and are no longer subject to doubt (de facto) (Section 4 and Section 5). While this alternative framework may offer better prospects than Pritchard’s original formulation for avoiding the pitfalls of fideism and relativism, while sidestepping evidentialism, I argue that it ultimately fails to support a parity argument between religious and non-religious beliefs. Consequently, it cannot sustain a religious epistemology that avoids both evidentialism and fideism.

1. Pritchard’s Parity Argument and the Threat of Fideism

No doubt, the central theme of OC is the idea that rational evaluation is essentially local. That is, any rational evaluation proceeds by taking for granted certain presuppositions, which are not themselves evidentially justifiable or known (to the extent that knowledge requires justification), yet are exempt from doubt. As Wittgenstein famously put it (OC 341–343):

The questions that we raise and our doubts depend on the fact that some propositions are exempt from doubt, are as it were like hinges on which those turn.

That is to say, it belongs to the logic of our scientific investigations that certain things are in deed not doubted.

But it isn’t that the situation is like this: We just can’t investigate everything, and for that reason we are forced to rest content with assumption. If I want the door to turn, the hinges must stay put.

The idea that all empirical inquiries depend on taking for granted some propositions which, while looking like empirical ones, are in fact exempt from investigation, and can be likened to “hinges”, which must stay put for ordinary empirical propositions to be justified or rationally doubted, is the central tenet of contemporary hinge epistemology in its many guises.2

The status of hinges as well as what counts as a hinge are hotly debated topics. For present purposes, we can think of “I have hands”, or “The Earth has existed for a very long time”, or even “There are physical objects” (understood as mind-independent entities, that exist even when not perceived) as bona fide hinges. That is, as propositions we take for granted and that allow us to bring various kinds of evidence to bear on the epistemic assessment of ordinary empirical beliefs, such as “I will burn myself if I take the pan out of the oven without potholders”, or “There were dinosaurs on the Earth before human beings existed”, or “The car keys must be in the drawer where I left them yesterday”.

A distinctive aspect of Pritchard’s version of hinge epistemology is his characterization of hinges as a-rational, visceral commitments. According to Pritchard, hinges possess propositional content that is truth-apt,3 but they are held with a unique kind of attitude—the attitude of hinge commitment—that is not epistemic in nature. This attitude is not governed by evidential or epistemic norms and is not subject to rational evaluation. That is, it is not correct just in case its content is true, or justified, or even known.

The passages in OC which support, at least prima facie, Pritchard’s interpretation are the ones in which Wittgenstein famously claims the following (OC 358–359, my italics):

Now I would like to regard this certainty, not as something akin to hastiness or superficiality, but as a form of life. (That is very badly expressed and probably badly thought as well.)

But that means I want to conceive it as something that lies beyond being justified or unjustified; as it were, as something animal.

In these passages Wittgenstein seems to put forward the view that all hinges are held with an attitude—he calls it “our certainty”4—that is neither justified nor unjustified and, therefore, neither rational nor irrational, and is thus beyond rational appraisal altogether.5

According to Pritchard, then, we ought to appreciate the force of a parity argument to the effect “that religious belief [should] be held to no more demanding an epistemic standard than we would apply to non-religious belief” (Pritchard 2015, p. 8).6

In fact, since, according to Pritchard’s reading of OC, non-religious belief can be assessed in terms of epistemic rationality only against the backdrop of a-rational hinges, religious belief too can be deemed rational only against a background of a-rational religious hinges, such as, presumably, “God exists”, or other suitable candidates. Thus, to redeem the rationality of faith, there is no need to provide evidential justification—a priori or a posteriori—for religious hinges. All we need is to viscerally hold on to them. The rational justification of “ordinary” religious beliefs—e.g., for Christians, that Jesus was resurrected from the dead after three days—will stem from them.7

Let us consider these claims in more detail. According to Pritchard’s reading of OC, ordinary empirical beliefs can only be epistemically justified by taking for granted several hinges to which one is a-rationally committed. Such a commitment, in turn, will depend, presumably, on one’s upbringing within a community that shares them. These hinge commitments, in turn, are manifestations of a more general über-hinge commitment to the fact that we cannot possibly be massively mistaken.

Here is a way of unpacking this idea.8 If I had to give up the hinge commitment that (luckily) I have hands, I could no longer trust my eyes which manifest the presence of my hands right now as I am typing on the keyboard of this computer. If I had to give up the hinge commitment that I live in such-and-such a place, then I could no longer trust my memory that attests otherwise. By dint of such hinge commitments, I am then justified in taking the deliverances of my senses and memory at face value to epistemically justify “ordinary” empirical beliefs formed on the basis of observation or memory, such as “Today, my hands are puffy” or “I left the car key in the apartment”, etc.

What exactly the hinge commitments are is a matter of contention, but whichever they are, they must be such that if I had to relinquish them, my entire epistemic worldview would collapse with them, according to Pritchard, because—presumably—I could no longer trust the cognitive faculties and epistemic methods normally used to form justified or knowledgeable empirical beliefs.

Now, if a parity argument holds between religious and non-religious beliefs, then the former too will have to be rationally evaluable only against the backdrop of a-rational, visceral religious hinge-commitments. Once again, it is not exactly clear what the relevant hinges might be in this area of discourse.9 Perhaps “God exists” would be a good candidate, even though this would also speak in favor of considering “There are physical objects”, a hinge of empirical discourse (see Section 4)—something that Pritchard repudiates.10

Or perhaps the relevant religious hinge is something like “God loves me”, which may well be part of a Christian worldview (although it might not be a hinge in other religious worldviews, where God is taken to exist but is quite often indifferent to human beings). Surely, if one were mistaken about that, one’s entire worldview would collapse, in a sense. Yet, in a very different sense than in the non-religious case—as a moment of reflection shows.

For, clearly, in the non-religious case, if it turned out that we cannot hold on to our hinges, this would open the door to forms of radical skepticism which would drag with them all our epistemic methods by means of which we form epistemically rational beliefs. By contrast, in the religious case, we would certainly receive an existential blow, but nothing detrimental to the proper exercise of our rational faculties. That is, one might lose hope or faith in the meaningfulness of life, or in the possibility of being reunited with one’s loved ones in an afterlife. Yet, one would not lose the ability to form evidentially justified beliefs and, with it, a grip on epistemic rationality altogether.

As mentioned, Pritchard’s views seem to have evolved regarding what should count as a religious hinge. Be that as it may, suppose the relevant hinge is “God exists”. If a Christian, say, lost faith, they could no longer consider the Bible the revealed word of God. Many religious beliefs based on the teachings of the Bible would therefore have to be abandoned. Yet, one could still form justified beliefs about specific physical objects, or revise them, based on empirical evidence, including perception, memory, testimony, induction, etc. Thus, one could still engage in the epistemic practices in which epistemic rationality unfolds.

If, in contrast, the relevant hinge were “God loves me”, which, as noted, is very specific to a Christian worldview, and one lost faith in it, this would no doubt have existential consequences but would not drag with it a loss of faith tout court, at least not necessarily. Still more, it would not affect one’s ability to participate in said epistemic practices.

Thus, there is no real “parity argument” on offer between religious and non-religious belief capable of salvaging the epistemic rationality of the former, which we may evince from Pritchard’s interpretation of Wittgenstein’s OC. More specifically, Pritchard’s own way of construing quasi-fideism is problematic because, on closer examination, nothing capable of salvaging the epistemic good standing of religious beliefs has been proposed in the first place. For questioning one’s religious hinges does not imply that any of our epistemic faculties or methods are unreliable.

For such a connection to be established, one would need to posit something like the existence of a sensus divinitatis—a mental faculty designed to perceive God’s presence, in the vein of reformed theology and of reformed epistemology (Plantinga 2000, who follows Calvin in thinking that there is a sensus divinatitis). According to this view, the proper functioning of the sensus divinitatis would be reflected in one’s acceptance of religious hinges. Consequently, abandoning these hinges would entail rejecting the idea that this faculty was ever functioning correctly or aligned with its intended purpose. Whatever the merits of this view, I seriously doubt that Pritchard or indeed Wittgenstein would be inclined to posit such a special faculty.

In particular, Wittgenstein was never drawn to the idea of multiplying mental faculties to explain why we have certain beliefs or why they would turn out to be reliable and therefore justified. In OC, he is very clear that his views have nothing in common with reliabilism. In particular, he maintains that hinges are not held fixed because they have been reliably formed. Rather, he holds that it is only by taking hinges for granted that we can then inquire and determine which epistemic methods are reliable and such as to give rise to justified beliefs, when properly utilized (OC 508–509; 514–515, 94, 205).

In the case of religious belief, he was more inclined to see it as a worldview—that is, a specific way of looking at human life and the Universe, leaving room for questions regarding their existential meaningfulness. In general, he was attracted to the idea that a religious outlook would be a powerful counter against scientism (not science).11 This perspective asserts that the only legitimate questions and answers are those framed within the scope of scientific inquiry, specifically addressing immanent efficient causes that can potentially be explained or subsumed under scientific laws.

Concerning OC, moreover, I have already hinted at the fact that the interpretation Pritchard puts forward is problematic. Indeed, if we read the famous passages in OC 358–359 in context, it is not at all clear that Wittgenstein’s definitive view is that certainty belongs, for him, to the animal/natural order, or to put it with Pritchard, that it is something “visceral”.

Let us go through them, to appreciate the misgivings Wittgenstein has with respect to the appeal to the “animal” (OC 358–361, my emphasis).

Now I would like to regard this certainty, not as something akin to hastiness or superficiality, but as a form of life. (That is very badly expressed and probably badly thought as well.)

But that means I want to conceive it as something that lies beyond being justified or unjustified; as it were, as something animal.

I know that this is my foot. I could not accept any experience as proof to the contrary—That may be an exclamation; but what follows from it? At least that I shall act with a certainty that knows no doubt, in accordance with my belief.

But I might also say: It has been revealed to me by God that it is so. God has taught me that this is my foot. And therefore if anything happened that seemed to conflict with this knowledge I should have to regard that as deception. (my italics)

Wittgenstein’s formulations are clearly very guarded and do not wholeheartedly support the appeal to the “animal” or the “visceral”. In my view, OC 361 is key here, for no matter where our certainty comes from—either from our animal nature (OC 359), or from a form of life (OC 358), or even from God (OC 361)—in our actions we will proceed “with a certainty that knows no doubt”. If the origin of this certainty, which often (not always) manifests itself in action, is irrelevant, then Wittgenstein is quite uncommitted about the source of that certainty. Rather, Wittgenstein’s point is a normative one, as the end of OC 361 makes clear, and it is about the very content of our certainty as an attitude. That is, we hold fast to certain propositions of the form of empirical ones (yet with a different role within our language and linguistic practices). That is, if putative evidence against them came up, we would either dismiss it or explain it away and would not revise those propositions. Just as we do with “3 + 2 = 5”: if putatively contrary evidence came up, we would ignore it or explain it away and would not revise that arithmetical proposition. Indeed, for Wittgenstein, this is part of the “logic” (OC 341) or the “method” (OC 151) of our investigations. That is, it is constitutive of them.12 To anticipate a point I will dwell on in the following: taking for granted these hinges is constitutive of the epistemic practices within which epistemic rationality unfolds. Epistemic rationality, therefore, comprises not only the “moves” we make within those practices—the justifications we possess by dint of empirical evidence –, but also the norms that determine which moves are permitted or forbidden—that is, the hinges that together with said evidence allow us to acquire justifications for or against ordinary empirical propositions. Thus, it is only by taking for granted that there are physical objects, broadly manifest in perception, that one’s current perceptual experience—i.e., one’s evidence—with a given content that p amounts to a (defeasible) justification for the corresponding belief.

If this view is correct, then the visceral dimension is integral to the phenomenology of our certainty. However, this does not imply that the contents of our visceral certainty—namely, the hinges—lie outside the scope of epistemic rationality. On the contrary, hinges are not merely enabling conditions for epistemic rationality that exist beyond its boundaries. Rather, they are intrinsic to epistemic rationality itself, much like the rules of a game are intrinsic to the game as a whole.

Contrary to Pritchard’s interpretation, then, Wittgenstein is concerned more with the contents of our “certainty” than with that very attitude and its “visceral” phenomenology.13 Furthermore, he is not committed to the view that said hinges fall outside the boundaries of rationality altogether. Even if they are not evidentially justified, they are constitutive conditions of epistemic rationality.

2. Further Problems with Pritchard’s Parity Argument

Key to the quasi-fideistic reading is (1) the parity argument between the requirements of rationality imposed onto religious and non-religious beliefs, and (2) the claim that both classes of beliefs are rational only against the backdrop of a-rational hinges, characteristic of each domain.

Now, if this view is meant as a response to external world skepticism of Humean descent, it has several problems. For how could hinges, which are outside the bounds of (epistemic) rationality, ground epistemically rational beliefs about physical objects in our surroundings?14 It would seem like magic if they could. In particular, how could taking for granted that I have hands be an a-rational commitment and ground the rationality of believing, for instance, that today my hands are puffy, based on observing them?15

There are two problems with this view: first, we would need a myriad of hinges to similarly ground all specific empirical beliefs, such as, for example, “The credenza I am seeing in front of me is brown”, “The chair I am sitting on is black”, etc. For, recall that Pritchard does not want to include the general hinge “There are physical objects”, or indeed broadly methodological ones, such as “My sense organs work mostly reliably”. Second, even if he were to admit these latter kinds of hinge, if they lie outside the boundaries of epistemic rationality altogether, it would be mysterious how they could give rise, in tandem with the exercise of one’s perceptual faculties, to epistemically justified empirical beliefs.16

As a response, one might want to say that hinges would not serve as foundations for ordinary empirical beliefs, in the way of propositions from which the justification of ordinary beliefs would somehow derive. All they would do is simply to entitle us to use our faculties—i.e., perception, memory, reasoning—and other epistemic methods—e.g., testimony –, which we normally use to form justified and knowledgeable beliefs.

As already mentioned (Section 1), I have serious doubts that this would be in keeping with Wittgenstein’s OC. Moreover, it would leave open why we would need to rely on hinges such as “I have hands”, before being allowed to take for granted the results of our various epistemic faculties and methods. After all, if “I have hands” is a hinge, it is so because I have always seen and proprioceptively felt my hands throughout my life. This kind of experience has “inculcated” that hinge in me, just like attending school has inculcated in me the hinge that water boils at 100 °C, or that the earth rotates around the sun, and that it is (roughly) spherical. Thus, it is thanks to perception and proprioception that I can so much as be viscerally certain and committed to “I have hands”. It is therefore difficult to see how I should be committed to such a hinge to be entitled to take the deliverances of perception and proprioception for granted in the first place.

Conversely, if I can blindly trust these faculties, then whatever belief (in a generic sense) they maximally sustain would turn out to be rationally grounded based on the evidence provided by the deliverances of these very faculties. It is then difficult to see how such beliefs could turn out to be a-rational. They would rather amount to j/k-apt beliefs—to use Pritchard’s own terminology.17

Be that as it may, we also saw how insisting on the a-rational acceptance of hinges such as “God exists” would not serve the goal of avoiding fideism. Even though “ordinary” religious beliefs, like “Jesus Christ resurrected from death after three days”, would turn out to be rational thanks to a-rational religious hinges, it would remain that religious hinges themselves would not be rationally held. This would be tantamount to giving in to fideism about religious hinges, and would be disastrous, by the lights of an anti-fideist. For the very core commitment on which the rest of a believer’s faith would be built would not be rationally held.18

3. The Issue of Epistemic Relativism

A further concern with Pritchard’s parity argument is that if any (system of) visceral, a-rational commitments were sufficient to generate epistemically rational beliefs, it would pave the way for radical forms of epistemic relativism. For if (epistemic) rationality is not absolute but dependent on a system of a-rational hinges, and if different, potentially incompatible, systems of a-rational hinges could exist, then ordinary beliefs based on them would also be rational, despite their incompatibility.

Consider a believer in astrology, like an ancient Roman, who, in virtue of their upbringing within that culture, was viscerally committed to the fact that certain astronomical patterns deeply influence human affairs, and someone who is not such a believer. For the former the outcome of a war, say, may be predicted based on the observation of such patterns; for the latter, it cannot, and their respective predictions may diverge. Indeed, even if their predictions did converge, they would be justified on the basis of entirely different and incompatible reasons.

In the case of religious discourse, this would make it hard to avoid seeing the worldviews of a religious believer and of a non-believer as epistemically on par.19 While the issue of Wittgenstein’s relativism—in general—is moot, he sometimes leaves open such a possibility, by equating our non-religious world-picture to a “mythology” (OC 95, 97).20

Yet, Pritchard aims to oppose epistemic relativism. Here is what he has to say about the issue (Pritchard 2022, pp. 8–9):

There is also a more specific notion of epistemic relativism, however, which maintains not just that there are distinct, but equally legitimate, epistemic systems, but also that these epistemic systems are epistemically incommensurate in the sense that rational resolution of deep disagreements between these epistemic systems is in principle impossible. Notice that this incommensurability claim isn’t entailed by epistemic relativism alone, as there could be distinct, but equally legitimate, epistemic systems which do in principle allow for rational resolution of disagreement. Indeed, hinge epistemology offers a case in point. While any divergence in hinge commitments would entail epistemic relativism in the general sense, it wouldn’t entail epistemic incommensurability so long as the divergence in hinge commitments is peripheral. Provided that there is a wide-ranging overlap in the hinge commitments held by both parties, then resolving deep disagreements, while it may face practical hurdles, is not in principle impossible. Call any form of epistemic relativism that entails epistemic incommensurability strong epistemic relativism, and any form of epistemic relativism that doesn’t entail epistemic incommensurability weak epistemic relativism. The point of the foregoing is that hinge epistemology in itself only seems to entail weak epistemic relativism, as while it might entail the possibility of divergent hinge commitments, this is compatible with such a divergence being peripheral rather than substantial.

Now, Pritchard seems to equate relativism—however weak it might be—with simple disagreement. You and I may disagree, occasionally, on how much we owe if we want to split the bill evenly. Still, we have enough background beliefs and epistemic methods in common to rationally resolve the disagreement.

Furthermore, it is difficult to see how a disagreement about hinges, which are viscerally held, and that are ways of encoding the über-hinge that one’s worldview is not fundamentally mistaken, could ever be “peripheral” in an epistemic system. Hence, it is difficult to see how weak epistemic relativism could be rationally resolved—that is, to what kind of common hinges one might appeal to resolve the issue in a way which does not beg the question against the other party’s hinges.

Be that as it may, and focusing just on religious discourse, one might think that it is somewhat more peripheral to one’s epistemic life than other areas of discourse.21 For, if one’s religious hinge commitments turned out to be false, this would not show that one is massively mistaken about everything else one believes outside the religious sphere. I think this is one way of potentially interpreting the gist of Pritchard’s (2022) lengthy discussion, where he seems to suggest that we can “compartmentalize” (Pritchard 2022, p. 18) religious hinge commitments. That is, we can detach them from the rest of our hinge commitments and epistemic practices. Of course, this would also mean that the interplay between religious hinges and the über-hinge would be different than in the non-religious case. This, in turn, and as suggested in (Section 1), would signal that there is no real parity argument between the religious and the non-religious case, in the first place. Still, it may be a way of confining religious disagreement—as typified by the believer and the non-believer—to the periphery of one’s belief system and of insulating it in such a way that it would not infect the rest of the system. Thus, one may grant that in this case there is a form of relativism but also hold that it does not spread over to the rest of one’s epistemic system. In other words, it would be a form of local relativism.

However, Pritchard seems to equate strong relativism, which he takes to entail incommensurability, with global relativism—that is, relativism in all areas of discourse. As discussed in Baghramian and Coliva (2019, chap. 2), nowadays almost no one holds global relativism. That is, the view, often associated with Protagoras, that all areas of discourse—factual, ethical, religious, etc.—are such that any judgement is true/justified/known only relative to a conceptual/epistemic/evaluative system, and that there are multiple, incompatible such systems, which would return opposite verdicts with respect to the truth/justification/knowledge of the relevant judgements. Thus, the issue of the kind of relativism in question—based on disagreement, say, or else, on incommensurability—and of its width—local contra global—should be treated separately.

If so, then, I think a deep divergence with Wittgenstein’s position emerges, relative to religious discourse. The point is not that, for Wittgenstein, the believer and the non-believer would not understand one another due, presumably, to an alleged incommensurability between their conceptual schemes. For, clearly, by and large they do share the same conceptual scheme, and may even share the same concept of God, while taking opposite stances regarding its instantiation. Instead, they would rely on two distinct types of evidence to uphold their opposing views: non-believers would appeal to purely immanent empirical evidence, while believers would turn to existential evidence, such as what Pascal referred to as “the evidence of the heart”.

As previously noted, Wittgenstein tended to interpret the religious worldview as a particular perspective on human existence and the Universe, which allows for the exploration of questions concerning their ultimate meaningfulness, as well as for their explanation, not (just) in terms of immanent efficient causes, but (also) in terms of transcendental final ones.

Thus, Wittgenstein is very sensitive to what he perceives as a “gulf” between the systems of evidence to which the believer and the non-believer would appeal. In other words, he makes room for what, following Bernard Williams (1981), we may call a “relativism of distance” between their respective worldviews. Hence, for Wittgenstein, in the case of the believer and the non-believer there is indeed a form of incommensurability between the kinds of questions they admit as pertinent and meaningful and the kinds of evidence they consider relevant to answering them. For instance, a believer may concur with a non-believer about the efficient causes that brought about the existence of the Universe, or more specific events, while still raising questions about the significance of either, which cannot be addressed by appealing to efficient causes. After all, some seasoned scientists are believers. Still, they would see the Universe as created in the sense of conforming to God’s intentions or design. In a sense, they cannot accept that what there is is what it is just out of chance (together with some initial conditions and given physical laws). Conversely, they would accuse the non-believer of—in a sense—surrendering to chance. Questions regarding the ultimate meaningfulness of one’s life, or of the Universe, that are central to the worldview of a believer, simply do not find a place within a non-religious worldview.

On top of that, the believer thinks that certain existential experiences—such as feeling secure and loved, even in the middle of a world-war, or thinking of oneself not just as imperfect but as irremediably sinful, no matter how virtuous one’s conduct is22—are manifestations of God’s existence. A non-believer, in contrast, would not take them as such and would either impute them to some psychological cause, or explain them away as illusory. Surely, this does not mean that epistemic relativism holds in all areas of discourse, for Wittgenstein.23 Yet, it means that it does hold—in a “strong” vein, according to Pritchard’s taxonomy—entailing epistemic incommensurability in the case of religious discourse.

4. “God Exists” as a De Jure Religious Hinge

Let us now explore the potential for quasi-fideism—or, at any rate, for a position meant to steer a middle path between evidentialism and fideism—under an alternative interpretation of OC, which draws on Coliva (2025d). On this reading, non-religious hinges can be categorized as either de jure or de facto.

De jure hinges, such as “There are physical objects”, are those hinges that are presupposed by entire areas of discourse and that let us transcend our “cognitive locality”, as they allow us to bring mind-dependent evidence to bear on the truth and justification of beliefs about mind-independent entities. Given their role with respect to empirical justification, they can neither be empirically proved nor disproved, for that would presuppose them either way. Furthermore, they cannot be proved by a priori arguments (see Coliva 2015, chap. 2). Yet, as I claimed in Coliva (2015, chap. 4), they are rational insofar as they are constitutive of epistemic rationality. As we saw, the taking for granted of these hinges is constitutive of the epistemic practices within which epistemic rationality unfolds. As anticipated in (Section 1), epistemic rationality, therefore, comprises not only the “moves” we make within those practices but also the norms that determine which moves are permitted or forbidden. De jure hinges, then, play a constitutive role for epistemic rationality. As mentioned, they are not just its enablers, falling as such outside its boundaries. Rather, they belong to epistemic rationality just like the rules of a game, and not only the moves within it belong to it. In fact, it is only thanks to them that specific actions can be characterized as moves of the game and normatively assessed as permitted or prohibited. To exemplify: kicking a ball between two poles does not count as a goal as such, but only within a game, like soccer, for instance, with its constitutive rules. Furthermore, it is only thanks to further rules in soccer that it will be deemed either valid or invalid, or a goal as opposed to an own goal, etc. Similarly, a visual experience as of a cup in front of me counts as a perception of the cup, only thanks to taking for granted that it has been caused by the interaction with physical objects, broadly manifest through sense perception; and only once it is rightfully considered a perception of a cup, instead of an hallucination of it, can it count as a justification for the corresponding belief.

Now, in addressing the problem of epistemic relativism, de jure hinges function as the fundamental boundaries of epistemic rationality. Consequently, they cannot be questioned without, by definition, lapsing into irrationality. For this reason, de jure hinges are both immutable and universal. If the proposition “God exists” were considered a de jure hinge, it too would be immune to that challenge.

One might then think that, much like “There are physical objects”, “God exists” is a hinge that (i) cannot be proved by empirical evidence, since God cannot be perceived, and (ii) that it cannot be proved by a priori arguments, since these arguments ultimately presuppose God’s existence, like in the case of the ontological argument, or fall short of establishing that God is needed to explain the existence of the universe. Just as “There are physical objects” allows us to bring mind-dependent evidence to bear onto the truth and the justification of beliefs about mind-independent objects, so (iii) “God exists” allows us to bring biblical testimony to bear onto specific religious matters. For it is only if the Bible is the word of God, that it can be taken to be authoritative with respect to specific religious beliefs such as the belief in the Holy Trinity, God’s incarnation in Jesus Christ, the Last Judgement, etc.

Yet, it is apparent that “God exists” is not itself a hinge of epistemic rationality. For one can form justified beliefs about material objects, without holding such a hinge. Its abandonment has no effect on those practices of ours that are constitutive of epistemic rationality, and it would not plunge us into irrationality altogether. To repeat: it may plunge us into existential upheaval, but it would have no consequence on the rest of our epistemic lives. Hence, the epistemic rationality of “God exists” cannot be redeemed by arguing that such a hinge is constitutive of epistemic rationality and is therefore mandated by it.

Thus, there is no viable parity argument which could rescue such a hinge from the charge of being held merely on faith and therefore of rescuing a believer in God’s existence from the accusation of fideism. That is, a parity argument against fideism (and relativism24) must do more than merely highlight a formal similarity between religious and empirical discourse. It must show, rather, that both rely on epistemically valid hinges, leading to beliefs that can be rationally evaluated.

5. “God Exists” as a De Facto Religious Hinge

De facto hinges, like “I have hands”, are those hinges, […] that, having been verified endless times, are excluded from further empirical investigation, and now play a rule-like role. After repeating endless times one’s full name and date of birth, “My name is AC” and “My date of birth is such-and-so” are no longer subject to empirical investigation. If I were to be mistaken or uncertain about them, this would likely be a sign of being cognitively impaired. Moreover, some de facto hinges, e.g., “The Earth is a globe”, are inherited from our community. Playing a rule-like role, putative counterevidence will be ignored or explained away. Moreover, they may be used as a rule of testing (OC 98), e.g., to test if one’s sense-organs, memory, and other cognitive faculties are working correctly, or if one has been acculturated as we have, etc. (cf. Coliva 2025d).

Furthermore, in the case of de facto hinges, it is possible to challenge them if new, previously unavailable evidence comes to light. In such instances, they would be reclassified as empirical propositions and evaluated based on their evidential merit. Unlike de jure hinges, de facto hinges allow for disagreement and variability, yet this does not lead to epistemic relativism. After a car accident, “I have hands” may no longer be a hinge, and would likely be reclassified as an ordinary empirical proposition, subject to verification. Throughout the passage from the Ptolemaic to the Copernican system, “The sun rotates around the earth” ceased to be a hinge, was reclassified as an empirical proposition, and eventually found to be false.

Now, if the proposition “God exists” were a de facto hinge, instead of a de jure one, it would be shielded from objections, as any purported evidence against it could be epistemically addressed and countered. And, if not, “God exists” would be reclassified as an empirical proposition and would be evaluated as such. In such an event, the disagreement between religious believers and non-believers would be epistemically comparable to the kind of dispute that might arise between epistemic peers over the correct amount owed when dividing a bill equally. In other words, rather than giving rise to incompatible yet equally correct judgements, their disagreement would admit rational resolution, at least in principle.

Furthermore, one may notice that, like many de facto hinges, “God exists” is passed down through our upbringing in a community, where that hinge is taken for granted. That is, much like “The Earth is a globe”, “God exists” is a hinge for many believers because they have been brought up in a community that shares that conviction. Moreover, that hinge shapes their lives, their rituals, and their overall worldview.

Yet, two points are worth making in this connection. First, that Wittgenstein was very much opposed to considering the dispute between believers and non-believers as a factual one, capable, in principle, of being decided based on empirical evidence.

Second, and more importantly, that there is a glaring difference between “God exists” taken as a de facto religious hinge and non-religious de facto hinges, like “The earth is a globe” or “I have hands”. For the latter are not sustained just by tradition. Rather, they are held fixed because all the available empirical evidence speaks in their favor. That is why they are allowed freely to circulate within the epistemic community, without being subject to further checks.

Thus, the epistemic rationality of “God exists” cannot be redeemed by comparing it to non-religious de facto hinges, which freely circulate without further scrutiny due to their long-established epistemic credentials. Therefore, no argument can equate “God exists” with these hinges to defend it from being based solely on faith, or to shield believers from the charge of fideism.

6. Conclusions

Contrary to Pritchard’s claims, this paper has shown that no epistemology of religious belief capable of steering a middle path between fideism and evidentialism, while also avoiding the dangers of relativism, can be evinced from OC, as read by Pritchard, who insists on the a-rational nature of hinges and on their eschewing epistemic relativism. I then argued that the project fares no better under a different interpretation of OC, based on Coliva (2025d), which distinguishes between de jure and de facto non-religious hinges. These hinges are considered rational either because they are constitutive of epistemic rationality or else because they are so much supported by empirical evidence that they have been removed from inquiry and have become like yardsticks. However, as we have seen, religious hinges are neither constitutive of epistemic rationality nor supported by overwhelming empirical evidence. Thus, this interpretation too cannot be used to defend an intermediate position, capable of steering away from evidentialism and fideism. Finally, we observed that Wittgenstein was inclined toward a form of epistemic relativism, in the form of a “relativism of distance” (Williams 1981), regarding the differing worldviews of religious believers and nonbelievers. These worldviews are characterized by distinct frameworks of questioning and responding, where certain inquiries and their corresponding answers are either inapplicable or cannot be adequately addressed within the epistemic system of the other group, thus leading to a form of incommensurability between them. In light of the foregoing, no matter whether quasi-fideism is developed along Pritchard’s lines or mine, it would permit such an outcome; pace Pritchard.