Introduction
Defenders of theism often maintain that belief in the existence of God and other core tenets of theistic religions is rationally permissible even in the absence of objective evidence. While different authors employ a wide range of strategies to defend the acceptability of non-evidential theistic beliefs, they all appeal in some form to the epistemic presumption of innocence (EPI).Footnote 1 According to the EPI, a subject S is justified in believing what seems true to S if no defeaters are present. Under this presumption, we routinely accept many basic beliefs about reality that cannot be proven in a non-circular way, such as the existence of the external world, the past, other minds and objective moral values, or the general reliability of our perceptual faculties. It has been argued that, to avoid a double standard, the EPI should likewise be extended to religious experiences, internal ‘seemings’ and intellectual intuitions underlying theistic beliefs. Therefore, if it seems to you or you have a feeling that God exists or Christ is our Saviour you are within your rights to believe that this is the case until proven otherwise. This idea, which can be traced back to Broad (Reference Broad1939) and Chisholm (Reference Chisholm1989, 64), has been further developed in the principle of credulity (Swinburne Reference Swinburne2004, 303-324), the critical trust approach (Kwan Reference Kwan, Craig and Moreland2009) and Michael Huemer’s phenomenal conservatism as applied to the epistemology of religion (Burns Reference Burns2017; Tucker Reference Tucker, Clark and VanArragon2011). Alston (Reference Alston1991), in a similar vein, maintains that religious experience associated with Christian mystical practice yields prima facie justified beliefs. Finally, Plantinga (Reference Plantinga2000) argues that theistic beliefs when taken in the basic way – that is immediately, not by way of an argument, but like perceptual or memory beliefs – can be justified if they are free from de facto objections.Footnote 2
One of the most popular arguments against the EPI is that it is dragging us into the ‘swamp of gullibilism’ (Martin Reference Martin1986). If religious experiences and intuitions should be regarded as prima facie justified, then not only belief in a theistic God, but also belief in fairies, ghosts, pagan gods and countless other supernatural beings would be rationally acceptable and permissible in many epistemic situations. Dawkins (Reference Dawkins2006, 53) states that belief in God is on an equal footing with Zeus, Apollo, Amon Ra, Mithras, Baal, Thor, Wotan, the Golden Calf, and the Flying Spaghetti Monster. Martin (Reference Martin1990, 272) and DeRose (Reference DeRose1999) have pointed out that Voodoo epistemologists could follow in Plantinga’s footsteps to portray Voodoo beliefs as properly basic and hence justified. It is not hard to see that Zeus’ worshipers could also use the EPI to their benefit. Indeed, Swinburne’s principle of credulity and phenomenal conservatism, previously used by defenders of theism, has also been employed for justifying animistic, spirit and polytheistic beliefs (van Eyghen Reference van Eyghen2023; Dumsday Reference Dumsday2024, 63–86). However, what exactly is the problem with the idea that belief in God and other parts of the theistic creed might be epistemically on a par with non-theistic beliefs? Why should theists be concerned about beliefs in non-God supernatural entities that are defensible on the same grounds as belief in God by those who have corresponding religious experiences and intuitions?
Theism and theistic-unfriendly beliefs
To answer this question, we should distinguish mere non-theistic beliefs from theistic-unfriendly ones (see Kwan Reference Kwan, Craig and Moreland2009). Mere non-theistic beliefs, such as beliefs in minor supernatural agents or forces that are not viewed as eternal or self-sustaining, could be easily accommodated in theistic worldview. If I believe that an elephant is in front of me because it appears to me that there is an elephant, and another person elsewhere believes that a mouse is in front of her, our beliefs are not in conflict and could be easily reconciled with each other. Likewise, if someone has an experience of seeing an elf in the forest or an intuition about the invisible, intangible, inaudible unicorn on a distant planet, beliefs formed in this way do not contradict belief in a theistic God, provided that elves and the unicorn have been created by God and nothing in experience of them excludes such a possibility. Historically, many theists have perceived many other supernatural beings alongside God, such as angels, demons, jinns, and the saints, coming from the heaven. In principle, belief in an invisible unicorn is no more incompatible with the theistic worldview than belief in St. Michael the Archangel. So the fact that some people justifiably hold mere non-theistic beliefs under the EPI poses no problem for theism.
At this point, one might object that the EPI is too permissive, so we are in danger of finding ourselves in a universe populated by strange things of all stripes (Martin Reference Martin1986). However, the rejection of the EPI pushes us too far in the opposite direction, making us too sceptical and disbelieving even some elementary truths, such as the existence of other minds or the reality as it appears to our senses. Determining what exactly is ‘too much’ for us and what constitutes a reasonable balance between excessive credulity and excessive scepticism is up to our intellectual intuitions and common sense, but to trust them we need the EPI, so it is indispensable in any case.
It also could be argued that believing in numerous supernatural entities under the EPI violates the principle of simplicity, but it is not clear why this principle should be prioritized over the EPI in each and every epistemological context. Even if the principle of simplicity is a useful heuristic device in empirical sciences, it does not follow that it should regulate all domains of cognition. Moreover, even in the course of the history of science more complex hypotheses sometimes replace simpler ones. Therefore, it is doubtful that the principle of simplicity is truth-conductive on its own (Fawkes and Smythe Reference Fawkes and Smythe1996; Lataster and Philipse Reference Lataster and Philipse2017). Given this, why not say that experiencing the world as populated by various invisible powers and agents could be closer to reality than a dry positivistic attitude? So the fact that by considering non-evidential belief in God as permissible we also open the door to belief in many other supernatural entities does not, by itself, constitute a valid objection to this epistemic move.
The real challenge for theism comes from theistic-unfriendly beliefs, which are incompatible with belief in a big-G God – that is, an all-powerful and all-wise personal Being who made all things from nothing and is in full command of them. Theistic-unfriendly beliefs include various forms of monism that identify God with the universe and beliefs in supernatural agents or forces with characteristics that do not fit into the theistic worldview. Here are several instances of theistic-unfriendly beliefs backed by corresponding religious experiences and intuitions: ‘it feels like there is a Manichean evil force coeternal with the Creator’, ‘I have an impression that Poseidon is the master of the sea’, ‘it seems that the world was created by Odin and his fellow gods from the flesh of the slaughtered frost giant Ymir’. No doubt such theistic-unfriendly beliefs could be reinterpreted within a theistic framework – for example, the Early Church Fathers viewed pagan gods as demons (i.e., fallen angels created by God). However, theistic beliefs can also be interpreted the other way around in light of theistic-unfriendly experience, so that Yahweh would be viewed as a manifestation of Zeus and Christ as an incarnation of Osiris.Footnote 3 Why should we consider one interpretation more rationally acceptable than the other?
It is doubtful that the arguments discussed in the current literature provide sufficient reason to prefer theism over alternative belief systems. For example, as Hume already recognized in his Dialogues Concerning Natural Religion, teleological arguments, such as the argument from the fine-tuning of the universe, do not necessarily point to an all-powerful, all-wise Creator. The same features that suggest design could just as plausibly be explained by the hypothesis that the universe was created by a team of limited deities.Footnote 4 In fact, this explanation might even be more convincing, especially given the mess and huge amounts of evil present in the world (Palmqvist Reference Palmqvist2023). Metaphysical considerations that supposedly favour theism over its rivals, such as a principle of simplicity, dubious as it is, also do not seem to be very persuasive, since the theistic omni-God is anything but a simple and easily imaginable object (Fawkes and Smythe Reference Fawkes and Smythe1996; Palmqvist Reference Palmqvist2023). Finally, belief in many gods is more difficult to refute than belief in the theistic God. Unlike the God of theism, non-God agents are not assumed to be fully in charge of what is going on in the world. They are also not expected to act according to the highest standards of morality and love. On the contrary, mischievous and envious gods are a common feature of polytheist pantheons. As a result, many theistic-unfriendly belief systems are much less vulnerable than classical theism to such objections as the argument from pointless suffering and the argument from divine hiddenness (Dumsday Reference Dumsday2024, 86–91).
The true contours of the problem of religious diversity for theism
Clearly, the fact that there are many incompatible religious belief-systems equally justified from the perspective of their adherents does not necessarily imply that all of them untrue or unacceptable. To show this, let us consider G. Chesterton’s helpful analogy between religious diversity and horse racing.
These are certainly solemn convictions [in hearts of derby-bettors]; men risk ruin for them. The man who puts his shirt on Potosi must believe in that animal, and each of the other men putting their last garments upon other quadrupeds must believe in them quite as sincerely. They are all serious, and most of them are wrong. But one of them is right (Chesterton Reference Chesterton1912, 311).
In the same way, it is possible that people who believe in a particular theistic religion will win while others will lose. Unlike horse racing which you can simply ignore, you cannot refuse to bet when it comes to religion, because agnosticism and atheism are also bets. And if there are no clear favourites, it is absolutely rational to listen to your gut feeling in making a decision. If, in absence of other evidences, you bet on a particular theistic ‘horse’ simply because a little voice in your head suggests that it will win, then your choice is justified under the EPI. Being an exclusivist about your own religion is a tenable position even if you do not have any conclusive arguments capable to convincing those who hold opposing religious views (Plantinga Reference Plantinga and Senor1995).Footnote 5
However, these considerations do not fully neutralize the problem of religious diversity as it is faced by theism. As Plantinga remarks, ‘for at least some Christian believers, an awareness of the enormous variety of human religious responses does seem to reduce the level of confidence in their own Christian belief’ (Plantinga Reference Plantinga2000, 456). Alston (Reference Alston1991, 275) expresses similar concerns: ‘it can hardly be denied that the fact of religious diversity reduces the rationality of engaging in Christian mystical practice (for one who is aware of the diversity) below what it would be if this problem did not exist’. The reason why this happens is evident from the horse racing metaphor. The more equally qualified contenders run the derby, the less confident you become that your chosen horse will finish first. Certainly, no matter how many horses join the race, in the absence of other considerations it would still be justified for you to bet on a particular horse simply because it seems the most appealing to you. However, as the number of competitors with equal chances of winning increases, you become less and less sure about your choice. If the number of contenders grows indefinitely, your belief in the desired outcome would become exceedingly weak. The same goes for religious beliefs. If you become aware of the followers of rival religions who are epistemically on par with yourself you still remain justified in accepting your religious experience as veridical (since no other clues are available). However, the strength of your belief would decrease in proportion to the number of rival options.
As is well known, in addition to a handful of major world religions, including Hinduism, which actually is an umbrella term for a wide variety of different religious traditions, there exist thousands of indigenous religions, new religious movements, including various strains of neopaganism, and small religions like mormonism. To this list should be added a bunch of religions of the past and new ones that will likely emerge in the future. Therefore, religious experiences and intuitions could potentially justify a vast, if not indefinite, number of mutually exclusive belief systems. The only conceivable limit on the diversity of experience-backed religious beliefs is the cognitive constraints of the human mind. For example, religious concepts that fail to meet the standard of minimal counterintuitiveness have low chances of being memorized (Boyer Reference Boyer2001). However, even these cognitive constraints permit a very broad range of equally justified yet contradictory religious options. A theistic believer, therefore, is interested in minimizing the number of rationally acceptable alternatives to theism. The fewer justified rival beliefs there are, the stronger his confidence in his own theistic faith. It follows that, in order to make theism more tenable, its defenders must demonstrate that a substantial portion – the larger, the better – of theistic-unfriendly religious experiences and beliefs is unjustified, unlikely to be veridical, and cannot be considered a rational choice.
Naturalistic argument for theism
Those who reject the rational acceptability of theism under the EPI such as Dawkins and Martin often adhere to a naturalistic outlook. My contention is that naturalism could become an unexpected ally of theism in tackling the problem of religious diversity. The term ‘naturalism’ has various meanings in the literature, but here I define it minimally as a conjunction of the causal closure principle (CCP) and the supervenience thesis (ST). Naturalism in this minimal form is fully consistent with scientific data and the practices of science.
CCP: According to the CCP, the observable universe is causally closed in the sense that non-physical entities cannot have any causal effect on the physical realm. In other words, all physical effects have physical causes. The CCP does not rule out the existence of transcendent supernatural agents, moral facts, mathematical objects, and other entities outside the space-time world but treats them as causally inert (Kim Reference Kim2003; Papineau Reference Papineau, Gillett and Loewer2001).
ST: The ST states that mental states are either identical to brain states or supervene on them, so it is impossible to change beliefs, feelings and other conscious content of a person without changing the physical properties of her brain.
Both the CCP and the ST are inductive generalizations based on contemporary neuroscience, the law of conservation of energy and the absence of empirical support for the existence of vital and mental forces irreducible to elementary physical and chemical processes (Harbecke Reference Harbecke2014; Papineau Reference Papineau, Gillett and Loewer2001). However, these empirical constraints are inapplicable to the initial act of creation, through which God is believed to have brought the observable universe into being. In initiating creation, God could not have been subject to the law of conservation of energy or to any other laws of nature, since He had not yet established them. Theoretically, God could have created a world that is not causally closed at the physical level – a world in which, for instance, people could move objects with their minds, and magic operated as it does in the Harry Potter books. If theism is true, then the CCP and the ST hold only because God chose to create this world in a particular way. Therefore, there is no inherent contradiction between naturalism (in its minimal form) and the idea of a non-physical agent creating a physical universe that appears causally closed to internal observers. To incorporate into naturalism the claim that there is no supernatural creator or sustainer of the observable world is to presuppose the falsity of theism and thus to beg the question.
If we accept the CCP and the ST, it is no longer possible to believe that supernatural agents can directly produce religious experiences in humans. The ST implies that Voodoo spirits, Poseidon, or God have no direct access to our consciousness to inform us of their presence, bypassing our nervous system. To generate religious experience in humans, supernatural agents would first have to activate neurons in our brains or to produce changes in our immediate environment such as table-turning or making noise. However, under the CCP they cannot do so due to their non-physical nature.Footnote 6 In light of this, we should dismiss suggestions by Swinburne (Reference Swinburne2004, 320), Plantinga (Reference Plantinga2000, 285) and Alston (Reference Alston1991, 59–60) that some experiences of God could be supernaturally induced. However, the same authors see no problem with assuming that God could also implant beliefs in humans through natural causes like Freudian mechanisms (Swinburne Reference Swinburne2004, 320; Alston Reference Alston1991, 233; Plantinga Reference Plantinga2000, 197), and naturalism leaves this option intact. Indeed, the causal closure of the physical does not pose a challenge for an almighty and omniscient Creator. Even if the physical universe is built to be impregnable for supernatural interventions, at the moment of creation God could have designed causal chains in such a way that they would produce true religious experience in people chosen by him ‘before the foundation of the world’ (Eph. 1:4). For example, consider the extravagant hypothesis that St. Paul’s conversion on the road to Damascus was caused by a bright falling meteor (Hartmann Reference Hartmann2015). If the Biblical God exists, he could have set up natural laws and the initial conditions of the universe so that a piece of cosmic debris would fall to Earth at the exact time and place necessary to make Paul aware of God’s will.
Certainly, an omniscient God could guarantee the desired outcomes of belief-forming processes without further interventions only if the universe is deterministic. As Jonathan Edwards suggested, had God created the universe with one more or one less atom, this could have had dramatic consequences on human beliefs.
The influence of the least particle may, for aught we know, have such effect on something in the constitution of some human body, as to cause another thought to arise in the mind at a certain time, than otherwise would have been; which, in length of time (yea, and that not very great), might occasion a vast alteration through the whole world of mankind (Edwards Reference Edwards1851, 153).
This may sound like old-fashioned Laplacian determinism no longer tenable in the age of quantum mechanics. But although universal determinism may be false, the stochasticity of quantum systems is not necessarily inconsistent with macro-level determinism. Individual quantum effects average out at a large scale, so that macro-level events, including neural events in the brain, could still be described by classical physics and would remain fully predictable for an omniscient God.
By contrast, fairies, Voodoo spirits, ghosts, and local deities are not omniscient and have no role in settings the initial conditions of the universe. As a result, they cannot communicate religious experience to humans through a predetermined series of physical events. We would have true justified beliefs about non-God supernatural beings in a causally closed universe only if God foreordained belief-producing processes for this purpose. Under naturalism, therefore, any religious experience incompatible with the existence of God as the sovereign Creator of all things is self-undermining. If there is no God guiding our belief-forming processes, yet the world is full of fairies and Voodoo spirits, we would have no knowledge of them because they are causally inert. Even if someone formed a belief about a fairy while walking through the forest and by happy coincidence the fairy indeed inhabits that particular forest, the truth of this belief would be a matter of sheer luck, so it cannot be regarded as knowledge in any sense. Therefore, given naturalism, theistic-unfriendly experience is unjustified even if true whereas theistic beliefs could result from reliable truth-aimed processes if there is a God who authored the universe and cares about what we (or at least some of us) believe.
It is important to note that God could also provide us with true beliefs about minor supernatural entities created by Him, such as angels, demons, or forest spirits. For example, if the angel Gabriel intended to announce to Mary that she would conceive and bear Jesus, God could have arranged the natural course of events in such a way that, due to the biochemical properties of her brain, Mary experienced a vision of Gabriel at the appropriate moment.Footnote 7 Thus, the naturalistic constraint on religious belief does not rule out all belief in non-God supernatural beings as unjustified, but only those non-theistic beliefs that are inconsistent with the notion of an almighty God exercising full control over created reality. This means that naturalistic argument for theism targets precisely those theistic-unfriendly belief systems that pose a problem of religious diversity for believers in God.
One might object that a similar argument could be construed to maintain the justifiability of belief in higher creating deities of polytheism. Is it not conceivable that Zeus, Marduk or Vishnu could have arranged physical events from the very beginning in such a way that humans would become aware of them at a later point of history?Footnote 8 However, creating deities as they perceived in polytheistic religions lack absolute sovereignty over creation unlike a theistic God. First, they do not create ex nihilo but from the pre-existing chaos or the primordial matter, and this significantly limits their ability to impose their decisions. The creation process is often depicted in polytheistic myths as a violent struggle such as Marduk’s slaying of Tiamat or the war between Zeus and the Titans. Even those sympathetic to polytheism acknowledge that ‘the numerous gods posited by polytheistic proponents could hinder each other and cause conflict in the universe’ (Lataster and Philipse Reference Lataster and Philipse2017). The same applies to the dualistic religions like Zoroastrianism. Second, polytheistic deities are typically conceived of as subject to impersonal forces beyond their control such as Ananke (destiny) which shapes the future of both humans and gods in the Greek mythology or the law of Dharma which Hindu deities must obey when recreating the world in each new cycle. Considering this, it is unlikely that such deities could have arranged causal networks once and for all from the very start to ensure the reliable production of true religious experiences. If Zeus had fought for power with his father Cronos, it is doubtful that he exercised sufficient control over cosmogony to decide in advance that this particular lightning strike would produce a true sense of his presence in someone. Given the causal closure of the physical, this event is more likely a mere coincidence, even if Zeus does exist. By contrast, an experience of theistic God caused by the same lighting strike would have a higher chance of being justified because if such a God exists he would be capable of prearranging it in a non-accidental way.
To illustrate this point, let us imagine that you live in a hard-to-reach area in the mountains. Once you find two letters at your doorstep. One letter is purportedly written by a tribal chief from a remote island in the middle of Pacific Ocean untouched by civilization. You begin to wonder how he could have known about you, obtained your address and managed to deliver the letter. The only conceivable scenario is that he placed the letter in a bottle, cast it into the ocean and left the rest to chance. However, a far more probable explanation is that the letter was fabricated by a local prankster. The second letter is allegedly sent by the powerful president of a developed country, who is in charge of special services capable of tracking almost every individual. This president is also believed to have hundreds of couriers equipped with off-road vehicles and helicopters, ready to deliver his messages to every corner of the planet. It is certainly possible that the president’s letter is also a hoax, as there is no independent evidence for the existence of either the president or the tribal chief – only unverified rumours about them circulating among your neighbours. Nevertheless, it is more rational to put greater confidence in the authenticity of the president’s letter, because if it were genuine it is at least clear how it could have reached you. Likewise, experience of God has a small but important epistemic advantage over its theistic-unfriendly counterpart. The rough and rugged landscape that divides the sender and the receiver in this example should be understood as a complex tangle of causes and effects that separates a religious believer from the moment of creation of a causally closed universe.
As is clear from the above discussion, the privileged epistemic status of theistic beliefs hinges on the notion of God’s all-encompassing providence and absolute sovereignty, as expressed in the Bible: ‘As I have planned, so shall it be, and as I have purposed, so shall it stand’ (Isaiah 14:24). However, if we reject the traditional view of God’s omnipotence, as process theologians and other theological revisionists do, this undermines the epistemological advantages of theism. Process theologians hold that creatural reality is coeternal with God and inherently free rather than created out of nothing and wholly dependent on God’s will. They maintain that the world does not passively obey God’s decrees, but responds to God in an unpredictable manner: ‘each divine creative impulse into the world is adventurous, in that God does not know what the result will be’ (Cobb and Griffin Reference Cobb and Griffin1976, 57). Consequently, God cannot ensure the desired outcomes of natural processes, including human history and biological evolution: ‘since God cannot control the evolutionary process, there is no reason even to assume that God was aiming that process specifically at us’ (Mesle and Cobb Reference Mesle and Cobb1993, 62). If evolution was not specifically aimed at producing humans, then our theistic beliefs (even if true) are merely a matter of chance and thus unjustified, which means that belief in a ‘weak’ God carries no more epistemic weight than belief in Zeus or Quetzalcoatl.
Theological implications of naturalism
Would adopting a naturalistic position require a substantial revision of traditional theistic doctrines? Although a full discussion of the theological implications of naturalism lies beyond the scope of this paper, it can be said that some of these implications are not as drastic as they might seem at first glance. For example, the possibility that God operates through natural processes in granting salvation has long been acknowledged by the Church fathers. St. Augustine suggested that the election of grace can be realized through the secondary causes, like the circumstances which allow a newborn baby to be baptized or an unbeliever to be exposed to the truths of the gospel. Those who, in Augustine’s words, ‘are not by the predestination of grace separated from the mass of perdition’, would die before their parents could bring them to church for baptism or would never encounter the divine words or deeds ‘by which they might believe if they only heard or saw such things’ (Augustine St 1876, 203). To put another way, if God elected a person for salvation before the foundation of the world He would grant her with appropriate psychological dispositions through genetic inheritance and other predetermined natural processes and place her in life circumstances in which she would inevitably come to believe the doctrines of true religion.
It is clear, then, that the Christian doctrine of predestination prominent in some Protestant churches and occasionally also present within the Catholic tradition (notably in the Jansenist movement), is fully in line with determinism implied by the idea that God produces true religious beliefs through preordained causal processes. The same applies to the Islamic doctrine of qadar, particularly as interpreted by Jabarites and Ash’arites, according to whom Allah has determined everything that will happen in the universe, so the fate of all humans was inscribed before the creation in al-Lauh al-Mahfūz (The Preserved tablet), which can never be changed (Watt Reference Watt1948).
However, the concept of miracles traditionally embraced by many theists is notably more difficult to reconcile with naturalism.Footnote 9 While some miracles are relatively easy to dispense with, such as the Old Testament plagues of Egypt which might have occurred through natural mechanisms, others, like the bodily resurrection of Jesus, are sine qua non of theistic creeds. Still, it is possible to argue that, after the crucifixion, Jesus entered a new transcendent mode of existence in a glorified body, thereby opening the path to salvation. Thus, the disciples’ visions of the resurrected Jesus, though potentially brought about by purely natural causes, could nevertheless reflect an important truth about the divine dimension of reality (if God prearranged them for this purpose). It is also possible that God made belief in the resurrected Jesus, however understood, necessary for union with Him in the afterlife. In short, adopting naturalism would not necessarily affect the core beliefs that one must confess to receive eternal life, provided that God’s truth is indeed revealed in a particular form of theism. After all, the creation of the world by God, once regarded as miraculous, is now widely understood in naturalistic terms as a process of evolution, yet this does not exempt Christians from confessing belief in ‘the Father almighty, maker of heaven and earth, of all things visible and invisible’, as prescribed in the Nicene Creed.
Naturalistic argument for theism in internalist and externalist contexts
Non-inferential religious beliefs can be considered prima facie justified within both internalist and externalist epistemologies. However, naturalistic argument for theism works differently in internalist and externalist contexts. If naturalism is true, then from the externalist standpoint theistic-unfriendly beliefs are unjustified for all individuals as they result from unreliable chance-ridden processes. If a person is unaware of the causal closure of the physical her theistic-unfriendly experience would still lack justification – just as a hallucinatory belief remains unjustified even if a person holding it is unaware that her brain is damaged and functioning improperly. By contrast, from the internalist perspective theistic-unfriendly experience is justified until the epistemic subject gains a cognitive access to naturalism. Once believers become aware of the CCP, the ST and the scientific evidence behind them they acquire an undercutting defeater for their theistic-unfriendly beliefs. However, people who are not familiar with science and contemporary naturalism remain fully justified in believing spirits and polytheistic gods if it appears to them that these supernatural agents really exist and could interfere with the physical events.
The epistemic situation was especially favourable to theistic-unfriendly experiences and beliefs in the pre-modern age, when the boundary between mind and world was perceived as ‘porous’, making phenomena such as the evil eye or demonic incursions into one’s thoughts widely accepted as a part of the natural order of things (Taylor Reference Taylor2007, 29–41). At the same time, the problem of religious diversity was far less pressing for the pre-modern people than it is today. For instance, the average Medieval European Christian was hardly aware of Hindu polytheism and other diverse theistic-unfriendly religions across the globe. As a result, he had no need for the naturalistic argument for theism to reinforce his Christian belief.
By contrast, the contemporary epistemic landscape is very different for the most people in developed countries. They are well-informed about religious diversity through mass education and mass communication, which may lead to reduced confidence in their theistic faith. However, while the modern age has challenged theistic belief by exacerbating the problem of religious diversity, it has also provided some relief through naturalistic worldview centred on science. To use Taylor’s expression, modern people live in a ‘disenchanted’ universe and view their minds as ‘buffered’, not open to an invisible world of supernatural agents and powers, which makes the naturalistic argument for theism more plausible for them. Therefore, this argument remains relevant within internalist epistemology, even though it does not deny the justification of non-inferential theistic-unfriendly beliefs for those unaware of the causal closure of the physical realm and the supervenience of consciousness on the brain.
Dawkins’s dilemma
If a naturalistic argument for theism gets off the ground, then a person cannot consistently hold the following three propositions simultaneously:
1. The epistemic equivalence thesis (EET): Without further evidence, theism, when held under the EPI, is no more justified or more likely to be true than any other religious beliefs held in the same way.
2. The CCP
3. The ST
To remain consistent, an opponent of theism must reject either the EET or the CCP & ST. Let us call this Dawkins’s dilemma, because Richard Dawkins teases Christians by lumping belief in God together with beliefs in fairies and pagan gods, yet he fails to recognize that these beliefs do not share the same epistemic status within his own naturalistic framework (Dawkins Reference Dawkins2006, 53).
Both horns of the Dawkins’s dilemma favour the theistic worldview. If an opponent of theism chooses to give up the CCP &ST, thereby conceding that non-physical agents can intervene in our consciousness or the physical world, he must also accept the possibility of divine miracles and revelation as traditionally understood in theistic religions. In this case, it would no longer be possible for him to outright reject Plantinga’s suggestion (Plantinga Reference Plantinga2000, 285) that ‘belief in the main lines of the gospel is produced in Christians by a special work of the Holy Spirit, not by the belief-producing faculties and processes with which we were originally created’. Clearly, one cannot consistently allow for such interventions by a theistic God, while denying them to other supernatural entities. Consider, for example, Voodoo spirits, who are not full-fledged creators of the universe and thus could not have designed human cognitive faculties to produce true Voodoo beliefs. Nevertheless, nothing prevents Voodoo spirits, if they are really there, from instilling true Voodoo beliefs in humans through the internal instigation and other supernatural means just as the Holy Spirit does according to Plantinga. The same goes for wide spectrum of pagan gods worshipped in theistic-unfriendly religions. Therefore, both theism and theistic-unfriendly belief systems could be the result of reliable truth-conducive mechanisms and thus equally justified. This outcome, though it might lead to theism losing its privileged epistemic status relative to other religions, could still be welcomed by some theists. After all, it opens the door to the reality of miracles, which many believers in God have long yearned for and revered.
Alternatively, if an opponent of theism prefers to retain the CCP &ST, he would have to reject the EET and concede that under naturalistic constrains theism has a much higher chance of being true than many alternative religious belief systems. If the CCP &ST hold, theism and theistic-compatible beliefs might result from the truth-aimed reliable belief-producing processes planned before the beginning of the physical universe and thus constitute the religious option acceptable to rational persons. By contrast, many other religious belief systems are fortuitous and cannot be considered justified even if by an unlikely coincidence they are true, and are therefore not rationally acceptable under naturalism.
To be sure, the problem of religious diversity would persist even if all theistic-unfriendly beliefs were dismissed as unjustified. For instance, the Muslim experience of Allah of the Quran is clearly at odds with the Christian experience of the Trinitarian God of the Bible. Nevertheless, if naturalism takes a considerable part of theistic-unfriendly beliefs off the table, it would significantly narrow the range of rationally available religious options, thereby boosting the confidence with which any form of theism – such as Christianity, Islam, and others is held. By ruling out a substantial portion of theistic-unfriendly alternatives as unjustified, the naturalistic argument for theism also appears to mitigate the many-gods objection to Pascal’s wager (Saka Reference Saka2001).
In conclusion, it is important to emphasize that the CCP and the ST are not ad hoc hypotheses introduced to favour theism over rival religious belief systems. Instead, the CCP and the ST are typically defended by non-theists for reasons unrelated to theism such as avoiding troubles with mental causation and maintaining the universality of the law of conservation of energy. Considering this, it is remarkable that naturalism, irreligious as it is, brings grist to the mill of theism.
Acknowledgements
I am deeply grateful for the extensive and constructive comments of the three anonymous reviewers, which helped me to improve this paper significantly. I am also indebted to Matvey Sysoev, Vladimir Shokhin, and Igor Gasparov for their helpful discussions of the argument developed herein.
Competing interests
The author declares none.