Hostname: page-component-cd9895bd7-dk4vv Total loading time: 0 Render date: 2024-12-26T17:33:16.779Z Has data issue: false hasContentIssue false

Atheistic modal realism

Published online by Cambridge University Press:  24 October 2022

Eric Steinhart*
Affiliation:
Department of Philosophy, William Paterson University, Wayne, NJ 07470, USA
Rights & Permissions [Opens in a new window]

Abstract

Atheistic modal realism asserts roughly that there are many concrete possible worlds and that the actual world is entirely godless. Here I will refine this position using the modal realism of David Lewis. For Lewis, all gods (including the Christian God) are contingent superhuman persons, who inhabit non-actual worlds. Although gods are concrete world-bound particulars, atheistic modal realism has room for impersonal absolutes and ultimates (which are not gods). Since no gods are actual, atheism is true. Yet there are infinitely many non-actual gods. Non-actual gods and worlds provide resources for analysing religious beliefs and practices. Lewisian theology provides a powerful new way for atheists to understand religion.

Type
Original Article
Copyright
Copyright © The Author(s), 2022. Published by Cambridge University Press

Introduction

Atheistic modal realism asserts roughly that there are many concrete possible worlds and that the actual world is entirely godless. Of course, there are many ways to be a modal realist. Since the modal realism of David Lewis has received the most discussion (and criticism), and since Lewis himself was an atheist, here I will develop atheistic modal realism in Lewisian terms. According to Lewis (Reference Lewis1986, 1), a world is a spatio-temporally maximal whole. There are many worlds. The class of worlds is closed under maximal variational principles: every variation of a world is also a world (Lewis (Reference Lewis1986), ch. 1.8; Wilson (Reference Wilson, Beebee and Fisher2022)). I will not defend either atheism or modal realism here; I only aim to show how Lewisian modal realism offers some advantages for atheists.

There are at least six reasons why atheists should be interested in atheistic modal realism. The first reason is that atheists probably need the full power of modal realism to reply to the fine-tuning argument for God. Many atheists have already turned to something like modal realism in their responses to that argument: they posit a vast plurality of universes plus the anthropic principle (e.g. Dennett (Reference Dennett1995), ch. 7.2; Dawkins (Reference Dawkins2008), 169–176; Stenger (Reference Stenger2009), ch. 4). These universes come from various cosmologies (Kragh (Reference Kragh2009)). But these universes are merely physically connected cosmic domains inside a single world (see Kraay (Reference Kraay2010), sec. 4). Since they are tightly constrained by the laws of that world, it has been argued that these cosmic domains are not sufficiently variable to counter the fine-tuning argument (Mawson (Reference Mawson, Bullivant and Ruse2013); Soler Gil and Alfonseca (Reference Soler Gil and Alfonseca2013)). To neutralize the fine-tuning argument, atheists require the maximal variability of modal realism. If this is right, then atheists have a good reason to embrace modal realism.

The second reason comes from the metaphysical narrowness of contemporary atheism. Atheism is often thought to entail one-world naturalism: all that exists is a single spatio-temporal-causal system composed of physical things (Armstrong (Reference Armstrong1978), 261; Draper (Reference Draper and Wainwright2005), 278). This entailment makes atheism vulnerable to attack: to refute atheism, just refute one-world naturalism. One-world naturalism is vulnerable to philosophical attacks. Its philosophical critics have long argued that it has fatal problems with morality, mathematics, mentality, and modality. It is also vulnerable to attacks from science. The cosmologist Max Tegmark (Reference Tegmark1998; Reference Tegmark2008) has recently argued that every mathematical structure exists physically. One does not have to agree with Tegmark to see that science itself might refute one-world naturalism. While one line of atheistic defence strives to protect one-world naturalism, another line drops it. Atheists Gray (Reference Gray2019) and Law (Reference Law2020) correctly point out that atheism does not entail one-world naturalism. Atheism is strengthened by developing atheistic alternatives to one-world naturalism.

The third reason comes from the axiology of theism (Kraay (Reference Kraay and Kraay2018); Lougheed (Reference Lougheed2021)). Axiologists investigate the comparative values of gods and their worlds. Modal realism helps by supplying the required worlds. An anti-theistic atheist may thus argue that worlds without God are better than worlds with God. The fourth reason comes from the study of the cosmological conditions in which gods exist or fail to exist. Atheists say the actual world contains no gods. Oppy (Reference Oppy2018, ch. 2.3) argues further that gods do not exist in any worlds that share laws and initial histories with the actual world. Some theists, following Tipler (Reference Tipler1995), might argue that worlds with certain non-actual physical laws contain gods. Both theists and atheists gain clarity by mapping out the modal boundaries of their doctrines. Modal realism supplies the logical space for this mapping.

The fifth reason comes from religious discourse. At least two logical pressures push atheists towards fictionalism. First, to debate with (say) a Christian, an atheist has to recognize that the Christian assigns true to ‘God has a son’. An atheist can do this by affirming that, in the Christian fiction, God has a son. Second, it would be absurd for an atheist to declare that, since all religious sentences are false, all religions agree. An atheist can avoid this absurdity by affirming that different theological fictions assign truth-values differently. Le Poidevin (Reference Le Poidevin, Buckareff and Nagasawa2016, 178) expresses theological fictionalism like this: ‘any given theological statement p is true if and only if it is true in the theological fiction that p’. But what are theological fictions? The modal realist answers that they are non-empty sets of non-actual worlds. Unlike non-realist fictionalisms, atheistic modal realism is not an error theory. Religious discourse involves silent modal operators; with them in place, it truthfully describes existing things. The sixth reason is that atheistic modal realism provides a novel analysis of theistic belief and behaviour. Further reasons can be given, but I hope these six suffice to show that atheists stand to benefit from modal realism.

Lewisian ontology and theology

I begin with gods. Since I am talking about all possible gods in all possible worlds, I need a maximally general definition of gods. For maximal generality, it is plausible to say that a god is a superhuman person. I take it that a human is a biologically defined kind of animal. Human biology entails that, for any kind of excellence, there is some finite degree of that excellence such that no possible human exceeds that degree. There is some finite speed which cannot be surpassed by the fastest possible humans, some finite degree of cognitive excellence which cannot be surpassed by the smartest possible humans, and so on. A superhuman person exceeds every possible human on every excellence. Whenever Lewis speaks about gods (or God, and so on) in the context of his modal realism, he always treats them as superhuman persons. Swinburne (Reference Swinburne1968) would add that gods must be immaterial. While Lewisian gods may be immaterial persons (Lewis (Reference Lewis1983a), 343, 362–363; Idem (Reference Lewis1986), 1, 73), they may also be material persons (Lewis (Reference Lewis1979b), 520–521; Idem (Reference Lewis2020a), Ltr. 205). Hence a Lewisian will not require immateriality. Tuggy (Reference Tuggy2017) and Milem (Reference Milem2019) say gods transcend the laws of nature (e.g. they can work miracles). Since Lewis denies that anything violates natural laws (Lewis (Reference Lewis1979a), 468–469; Idem (Reference Lewis1981), 114; see Idem (Reference Lewis1973), 75–77), a Lewisian will not add this requirement. For generality, I say gods are superhuman persons.

On this definition, the deities of indigenous paganisms are gods (Athena, Odin, Isis, Quetzalcoatl, and so on). The observable and intelligible gods of Neoplatonism are gods. The Yoruba orisha are gods. The Hindu and Buddhist deities are gods. If the earth, the sun, or the universe were intelligent living organisms, they would be gods. If nation-states really are persons, then they are gods. Superhuman extraterrestrial aliens are gods. Genetically engineered superhuman organisms are gods, and superhuman robots are also gods. The God of flesh and bone described by Mormon theology is a god; the personal God described by many Abrahamic theologies is a god.

Of course, many theologies declare that God is not a person: God is an impersonal cosmic force, or the impersonal abstract ground of being, and so on. Nevertheless, if the concept of gods is so variable as to include personal gods, impersonal gods, concrete gods, abstract gods, and so forth, then it is likely to be meaningless (Lewis (Reference Lewis1986), 140). Since the vast majority of human religions treat gods as persons, it is fair to say that gods are persons. Other terms exist for allegedly impersonal or abstract gods or Gods: the One, the Absolute, the Ground of Being, and so on. And while I will focus on personal gods, I will point to places for impersonal absolutes in atheistic modal realism.

All gods are superhuman persons. But how should they be categorized? Following Lewis (Reference Lewis1983b, 39–40), atheistic modal realism recognizes three general categories. The first category contains possible individuals. Every possible individual exists in exactly one world. It is a concrete world-bound particular. The second category contains impossible individuals. They are not parts of worlds. They are concrete trans-world wholes, that is, wholes whose parts are distributed over many worlds. The third category contains non-individuals. These are classes (Lewis (Reference Lewis1983a), 343). The pure classes are ‘not made out of the parts of the worlds’ (Lewis (Reference Lewis1986), 111–112). They are abstract objects that do not exist in worlds, but exist from the standpoints of all worlds. The impure classes contain some concrete Urelemente (that is, possible individuals). They do not exist in worlds either. Impure classes exist from the standpoints of the worlds containing their Urelemente. Since Lewis affirms that gods exist, they must go into one of these categories.

Consider the category of possible individuals. There are four arguments that Lewisian gods go into this category. The first argument comes from physical relations: Gods participate in causal relations (ibid., 3). And in spatio-temporal relations (Lewis (Reference Lewis1983b), 138–141). But the things that participate in such relations are concrete world-bound particulars. So gods are concrete world-bound particulars. The second argument comes from the Lewisian principle that worlds are isolated wholes sharing no physical parts: Lewis (Reference Lewis2020a, Ltr. 205) says the principle of isolation applies to gods: ‘No god is contained in more than one world.’ If isolation applies to any objects, then those objects are world-bound particulars. Therefore, gods are concrete world-bound particulars. The third argument comes from the Lewisian principle that duplicates of parts of worlds can be patched together to make further worlds. This duplicating and patching together is recombination. The set of worlds is closed under recombination. Recombination applies to all gods (ibid.). If recombination applies to any objects, then those objects are world-bound particulars. Therefore, all gods are concrete world-bound particulars.

The fourth argument comes from the way Lewis deploys quantifiers to analyse the omni-properties of gods. It goes like this: Lewis often talks about the omni-properties of gods (omniscience, omnipresence, and omnipotence). When he talks about them in the context of his modal realism, he says their quantifiers are restricted to a single world. A theistic modal realist can believe in a God who is ‘a maker of all things; but he would have to say this under a tacit restriction of quantifiers and mean a maker of all things that are part of the same world’ (ibid., Ltr. 181, his italics). When it comes to omnipotence, he says he can tell a Christian ‘to restrict his quantifiers and say “allmighty” and just mean “all-His-own-worldmates-mighty”’ (Reference Lewisibid., Ltr. 198). Thus ‘there are many gods in many worlds . . . none is (speaking unrestrictedly) omniscient or omnipotent’ (ibid., his italics). Likewise an omnipresent spirit is present only at all the places in a single world (Lewis (Reference Lewis1986), 73). So Lewis (Reference Lewis2020a, Ltr. 198) restricts the quantifiers in these omni-properties to a single world. If gods were not concrete world-bound particulars, then Lewis would not have defined their omni-properties via restricted quantifiers. Assuming Lewis means what he says, Lewisian gods are concrete world-bound particulars.

Consider the impossible individuals. These trans-world individuals exist (Lewis (Reference Lewis1986), ch. 4.3). Their parts come from many worlds. But here is an argument that they are not gods: Lewis says the trans-world individuals are not things that ‘we ordinarily name, or classify under predicates, or quantify over’ (Reference Lewisibid., 210). But gods are things that we ordinarily name, classify under predicates, and quantify over. Therefore, Lewisian gods are not trans-world individuals. Nevertheless, Nagasawa (Reference Nagasawa, Buckareff and Nagasawa2016) has advanced a modal panentheism which identifies ‘God’ with the totality of possible worlds. Since Lewis (Reference Lewis1986, 212) recognizes absolutely unrestricted composition, this totality exists: it is a trans-world individual. An atheistic modal realist can affirm its existence too. However, since it is not a person, nor even personal, it is not a god or God. It is an impersonal Absolute. Moreover, as a trans-world individual, it stands outside Lewisian model-theoretic semantics, and therefore has its properties and relations only in an analogical sense.

Consider non-individuals. These are abstract objects that exist from the standpoints of some worlds. Lewis says the non-individuals are classes (Reference Lewis1983b, 40). Examples include numbers and properties (Lewis (Reference Lewis1973), 39; Idem (Reference Lewis2020a), Ltr. 136; etc.). However, Lewis (Reference Lewis1991, 8–9) includes God among the ‘remarkable non-classes’ and says such things are world-bound. Lewis never treats God (or any god) as an object existing from the standpoint of any world. Therefore, no gods exist from the standpoints of any worlds. Mathematical objects like pure classes are necessary beings in the sense that they exist from the standpoints of all worlds. But these things are not gods, and using standpoints is not a way to fit some necessary God into Lewisian ontology. Lewis (Reference Lewis2020a, Ltrs. 138, 181, 192) objects to theologies that make God an exception to his ontology. The necessary God of analytic theism has no place in his modal realism (e.g. Sheehy (Reference Sheehy2006); Pigden and Entwisle (Reference Pigden, Entwisle and Maclaurin2012); Lewis (Reference Lewis, Loewer and Schaffer2015); Vance (Reference Vance2016); Collier (Reference Collier2019)). Lewis frequently and explicitly says that that necessary God does not exist at all (Reference Lewis2020a, Ltrs. 181, 194, 198, 205).

What about universals? Lewis was neutral about universals (e.g. Lewis (Reference Lewis1983a), 343; Idem (Reference Lewis1986), 67, 205). Since Lewis was neutral, atheistic modal realists are free to deny or affirm them. An atheistic modal realist might therefore say that Existence-Itself is the ultimate universal, and, as such, the source of the being of all other things in Lewisian ontology. The Plotinian name for Existence-Itself is ‘the One’. Tillich refers to it as ‘the Ground of Being’; he says God is the Ground of Being. Johnston (Reference Johnston2009) calls it ‘the Highest One’; he says God is the Highest One. To avoid confusion, atheistic modal realists do not use ‘God’ for any impersonal absolute or ultimate. In any case, even if universals exist, Lewis never classifies gods as universals. Atheistic modal realists say gods are not universals.

To summarize: (1) Any god (including God) is either a possible individual, an impossible individual, a non-individual, or a universal. (2) Lewis always classifies gods as possible individuals. He never places gods into any other categories. He objects to theologies that make exceptions for God. (3) Therefore, all Lewisian gods are possible individuals. Every god is a concrete world-bound particular. Since gods are possible individuals, the only way they can be associated with worlds is as their (concrete) parts. Lewis persistently argues that worlds share no parts (Lewis (Reference Lewis1968), P2; Idem (Reference Lewis1986), chs 1.6, 4.2; etc.). Again, no god is a part of many worlds. That is Lewis's own theology, at least as I see it. I have no doubt that there are reasonable ways to modify Lewis to yield other theologies (e.g. Cameron (Reference Cameron2009); Almeida (Reference Almeida2017); Collier (Reference Collier2021); Maia (Reference Maia2021)). But here I focus exclusively on what I take to be Lewis's own theology.

Gods exist in non-actual worlds

Lewis (Reference Lewis2020a, Ltr. 205) gives an argument from recombination to possible gods. It can be filled out in more detail like this: our actual world contains organisms and machines. These agents have varying degrees of excellence (power, intelligence, virtue, and so on). Transhumanists like Harari (Reference Harari2015) describe super-organisms far more excellent than humans. Technologists like Moravec (Reference Moravec1988) and Kurzweil (Reference Kurzweil2005) describe super-machines far more excellent than humans. These super-agents (organisms or machines) are made by recombination. Although they are not (as far as we know) actual, they are physically possible. Since these super-agents are Lewisian gods (Lewis (Reference Lewis2020a), Ltr. 205), they are gods by recombination. Lewis says Thor is a god by recombination (ibid.). Moravec presents his cosmic super-machines as equivalent to the biblical God. If he is right, then the biblical God is also a god by recombination. For any god made by recombination, there is some non-actual world in which it exists. Therefore, some worlds contain these gods by recombination. It is possible that gods exist. Moreover, pantheists like Buckareff (Reference Buckareff2019) suggest that the parts of the actual world combine into a physical whole which is a Divine Mind (that is, a god). Even if the actual world is not a pantheistic god, it is at least possible that other worlds are such gods.

Some gods require alternative physics. The Olympian deities (as described by the poets) have physical bodies. They have superhuman powers that cannot be realized by combinations of actual physical particles. So the Olympians cannot be generated merely by recombination of duplicates of actual particles. They are made of alien particles governed by alien laws. Lewis (Reference Lewis1973, 88) says ‘there are worlds where physics is different from the physics of our world’. If there are no worlds containing these Olympian particles and the Olympian bodies made from them, then there are gaps in logical space. Lewisian plenitude rules out such gaps. Therefore, there are worlds in which Olympian deities exist. The deities of other polytheistic pantheons are also made from alien physics. Tipler (Reference Tipler1995) describes an entirely physical infinite God. He takes great care to precisely define its physics as well as to argue for its equivalence to the biblical God. Tipler's God is possible in worlds with slightly non-actual physics. For the modal realist, such worlds exist. Via alien physics, it is possible that gods exist.

Plenitude and alien properties yield immaterial gods: There are non-physical auras, entelechies, spirits, spooks, and deities (Lewis (Reference Lewis1983a), 343; Idem (Reference Lewis1986), 1, 73). If there are no worlds containing these immaterial gods, then there are gaps in logical space. But plenitude rules out such gaps. Therefore, immaterial gods inhabit some possible worlds (Lewis (Reference Lewis1983a), 362–363). However, since these gods cannot be generated by physical recombination, they are generated through alien properties. Either through recombination or alien properties, possible gods exist (Lewis (Reference Lewis1973), 99; Idem (Reference Lewis1979b); Idem (Reference Lewis1983b), xi, fn. 4; Idem (Reference Lewis1986), 136; Idem (Reference Lewis2020a), Ltrs. 181, 198, 205, 206, etc.). A world that contains a god is a godly world.

Lewis often says our actual world contains no gods (Lewis (Reference Lewis1983b), xi, fn. 4; Idem (Reference Lewis1986), 1, 132, 136; Idem (Reference Lewis1993), 150; etc.). He says ‘I think of this world we live in as entirely godless’ (Lewis (Reference Lewis1983b), xi, fn. 4). And: ‘I believe that we inhabit a godless world (and in this sense am an atheist – I hold that among my worldmates there is no god)’ (Lewis (Reference Lewis2020a), Ltr. 181). Since all the traditional arguments for atheism are compatible with modal realism, they can all be marshalled against any allegedly actual gods. Among those arguments, Lewis endorses the argument from evil against the Christian God (Reference Lewis1993, Reference Lewis1997; Lewis and Kitcher (Reference Lewis, Kitcher and Anthony2007)).Footnote 1 Lewis (Reference Lewis1986, 1) appears to give an argument from physicality: our actual world is composed entirely of physical things. If our world is composed entirely of physical things, then it does not contain any non-physical gods. Therefore, our actual world does not contain any non-physical gods. But this argument doesn't rule out all gods. Lewis (Reference Lewis2020a, Ltr. 205) regards Thor as a physical god. The gods that sit on mountains are physical (Lewis (Reference Lewis1979b), 520–521). So perhaps physical gods exist elsewhere in our world.

Another argument against actual gods comes from miracles. Many gods of many traditional religions are portrayed as performing miracles. A miracle is a deed which would violate actual physical laws if it were done in the actual world. For example, Thor is portrayed as violating actual physical laws (he miraculously resurrects his goats). But Lewis says the laws of physics at any world are never violated in that world (Lewis (Reference Lewis1979a), 468–469; Idem (Reference Lewis1981), 114; see Idem (Reference Lewis1973), 75–77). Therefore, our actual world contains none of the traditional miracle-working gods. It follows that miracle-working gods (like Thor) are non-actual. Of course, in their own non-actual worlds, the amazing deeds of gods are not miracles but conform to non-actual physical laws. While this argument against miracle-working gods entails that many miracle-working gods of traditional religions are non-actual, it still permits the actuality of many physical gods (e.g. those of the transhumanists, technologists, and pantheists). For all we know, they actually exist.

Atheistic modal realism is therefore falsifiable (this is a virtue). If transhumanist, technological, or extraterrestrial gods ever do appear, or if our world really is a pantheistic god, then atheistic modal realism is false. It will need to retreat to some fall-back position. One such position says atheism (as a-theism) rules out only theistic gods (Cliteur (Reference Cliteur2009)). Since theists favour non-physical gods, or gods who work miracles, this fall-back merely says those theistic gods are non-actual. But these are almost all the gods of our religions. This fall-back permits actual non-theistic gods (gods consistent with actual physics). Of course, since ignorance cuts both ways, for all we know, such physical gods do not exist. So I will proceed on the Lewisian thesis that our world is ‘entirely godless’.

Religious theists and religious atheists

Atheistic modal realism provides a new way of thinking about religious behaviours. It starts with Atran and Norenzayan (Reference Atran and Norenzayan2004, 713), who stipulate that religious behaviours are ‘passionate communal displays of costly commitments to counterintuitive worlds governed by supernatural agents’. To develop his atheistic approach to religion, Oppy (Reference Oppy2018, ch. 3) adapts Atran and Norenzayan. Here I will also adapt Atran and Norenzayan: religious behaviours are passionate communal displays of costly commitments to worlds governed by gods. To be religiously oriented to some godly world is to participate in those commitments to it. To be religious is to be religiously oriented to some godly world. Since there are degrees of commitment, there are degrees of religiosity (and irreligiosity). Religion and irreligion cut across theism and atheism in complex ways. An atheist believes our actual world is godless, while a theist believes our actual world is godly (it has at least one god). Atheistic modal realism says atheists are correct, while theists are incorrect. The belief in actual gods is a theistic illusion. However, since some worlds are godly, atheistic modal realism implies that it is theism but not religion that is delusional.

Atheistic modal realism permits both theistic and atheistic religion. On the one hand, theistic religion consists of passionate communal displays of costly commitments to non-actual godly worlds which are incorrectly believed to be actual. On the other hand, atheistic religion consists of passionate communal displays of costly commitments to non-actual godly worlds which are correctly believed to be non-actual. One way to be a religious atheist (to participate in an atheistic religion) treats gods as ideals. Lewis says idealizations of actual objects inhabit other worlds (Reference Lewis1986, 26–27). Some (but not all) gods are morally ideal persons. An atheist can and should strive to be like those gods. Plato says we ought to strive to become as godlike as possible (Theaetetus, 176a5–b2). Plotinus says our goal is to live ‘the life of the gods: for it is to them . . . that we are to be made like’ (Enneads, 1.2.7.25–32). The modern pagan writer Beckett (Reference Beckett2017, 144–146) says that we ought to try to become gods. The superhuman persons posited by the transhumanists are also non-actual gods. By engaging in collective projects to build those gods, or to transform humans into them, these transhumanists are engaged in passionate communal displays of costly commitment to their godly worlds. Thus a religious atheist is religiously oriented towards an ideal god which they affirm to be non-actual.

Atheistic modal realists are interested in the moral characters of gods and their worlds. On the one hand, the Homeric Olympians often act immorally; they are not worthy of imitation. On the other hand, Plato argued that there are morally ideal versions of those gods; they are worthy of imitation. Zeus embodies ideal justice; Athena has ideal wisdom. Lewis is interested in the axiology of gods. He believes any god who eternally tortures people in hell is evil. Hence ‘God as portrayed by most Christians, is horrendously evil’ (Lewis (Reference Lewis2020b), Ltr. 641). There are horrendously evil worlds governed by such Gods; we should be glad we are not in them. Lewis argues that it is evil to be religiously oriented towards an evil god (Lewis and Kitcher (Reference Lewis, Kitcher and Anthony2007); Lewis (Reference Lewis, Loewer and Schaffer2015), 213). Yet theistic worlds exist at which much of the Bible is false, and at which hells do not exist. Perhaps the Gods at those worlds implement Hick's progressive soteriology (Hick (Reference Hick1976), chs 15, 20, 22). Again, atheists can be religiously oriented towards such non-actual gods as ideals. Since an ideal spectator is non-actual, an atheist can adopt an ideal spectator theory of morality.

Lewis (Reference Lewis1978) uses worlds to distinguish between fact and fiction. A text recited in our world is factual iff it is recited truthfully here; it is fictional iff it is recited truthfully only in non-actual worlds. Since religious stories (stories involving gods) are recited truthfully only at non-actual worlds, all religious stories are fictional. Atheistic modal realism thus predicts that religious texts closely resemble literary fictions. And they do (Mackendrick (Reference Mackendrick2012); Petersen (Reference Petersen2016); Gericke (Reference Gericke2017)). Atheistic modal realism predicts that texts of abandoned religions will be treated as fictional literature. The texts of ancient religions are now treated as myths. It predicts that atheists will regard the texts of current religions as fictions. And they do. It likewise predicts that literary fictions will give birth to new religions. And there are many new literary religions (Cusack (Reference Cusack2010); Davidsen (Reference Davidsen2014)). Examples include the Church of All Worlds, the Church of the SubGenius, Pastafarianism, the Tolkienist and Lovecraftian religions, Jediism, and Dudeism.

No gods exist necessarily

Atheism provides a direct argument against divine necessity: All gods are concrete world-bound particulars; they are possible individuals. If any possible individual exists necessarily, then it has a counterpart in every world. The counterparts of gods are gods. So if any god exists necessarily, then it has a counterpart god at every world. No gods are parts of our actual world. So no god has a counterpart god in our world. Therefore, no god exists necessarily. Recombination yields a second argument against necessary gods: Again, if any god exists necessarily, then it has a counterpart god in every world. Speaking of concrete things, Lewis says (Reference Lewis1986, 88) ‘anything can fail to coexist with anything else’. He applies this failure to gods: ‘if there are two distinct things, say a god and his creation, then at some world there's a duplicate of the second unaccompanied by any duplicate of the first’ (Lewis (Reference Lewis2020a), Ltr. 181). Consequently, for any god in any world, recombination makes worlds that lack counterparts of that god. No god has a counterpart in every world. All concrete things, including all gods, are contingent. No god exists necessarily.

Here is a third Lewisian argument against divine necessity: the only arguments that entail divine necessity (in terms of worlds) are ontological. But Anselm's ontological argument is unsound (Lewis (Reference Lewis1970)). And Plantinga's modal ontological argument is unsound. Lewis (Reference Lewis2020a, Ltr. 194) says ‘I deny that “God exists necessarily” is true at any possible world’. Similar objections render all ontological arguments unsound. Hence there are no arguments for divine necessity. And while there are no arguments for divine necessity, there are arguments against it. Consequently, Lewis says ‘there are many gods in many worlds, . . . but none of them is a necessary being’ (Reference Lewisibid., Ltr. 198). And he says ‘I think it's inconsistent to say He [God] exists necessarily’ (Reference Lewisibid., Ltr. 205). But this applies to all gods: no gods are necessary; all gods are contingent.

A fourth Lewisian argument against divine necessity comes from generalizing arguments involving the moral character of the Christian God: Christians say God is maximally good. However, our actual world contains too much evil to be compatible with maximal goodness (Lewis (Reference Lewis1993), Idem (Reference Lewis1997); Lewis and Kitcher (Reference Lewis, Kitcher and Anthony2007)). Therefore, the Christian God does not actually exist. The argument from evil ‘succeeds conclusively’ against a maximally good God (Lewis and Kitcher (Reference Lewis, Kitcher and Anthony2007), 231). However, since Lewis (Reference Lewis2020b, Ltr. 641) thinks the Christian God is ‘horrendously evil’, perhaps he would run a converse argument: our world contains too much goodness to contain God. These arguments involve conflicts between the essential character of some god and its world (Schlesinger (Reference Schlesinger1997)). And so they generalize. Any god has some essential character. For any essential character, some worlds are incompatible with its instantiation. Therefore, for any god, there are some worlds at which it lacks counterparts. No god exists necessarily.

Divine contingency has religious implications. For every possible god, there exists exactly one world in which it exists. The Christian God is no exception. He exists in exactly one world. Again, when the Christian says God has omni-properties, Lewis (Reference Lewis2020a, Ltrs. 181, 198, etc.) says their quantifiers are restricted to the single world inhabited by that God. Following Atran and Norenzayan, the Christian religion consists of passionate communal displays of costly commitments to some non-actual world governed by the Christian God. But the Christian God governs only one world among the infinitely many worlds in logical space. Hence Lewis says ‘The thesis of plurality of worlds belittles god – belittles any god, the god of any world that has a god – by portraying him as only one among equals, a minute part of reality’ (Reference Lewisibid., Ltr. 181). And he says ‘It strikes me that the many godlings of the many worlds are not what the [the analytic theist] had in mind’ (Reference Lewisibid., Ltr. 198). It follows that, even if analytic theists believe they are thinking about a necessary God, they are in fact only thinking about a non-actual contingent God. The necessity of God is another theistic illusion. Of course, the Christian God has many counterparts at some (but not all) other worlds. They too have their omni-properties only via restricted quantification. These counterparts all share the essential features of the Christian God (as far as those features can be represented in modal realism). Yet they will vary in their accidental features. Through this variation, they may be picked out more precisely by the varying theologies of different Christian sects.

Since the Lewisian gods are small contingent parts of reality, and are ontologically equal to each other, Lewis says they are not worthy of worship (Reference Lewisibid., Ltr. 181). Atheistic modal realists believe in non-actual gods but do not worship them. If idolatry is worshipping something that shouldn't be worshipped, then all worship is idolatry. As idolatry, worship is always religiously inappropriate. The illusory necessity of the Christian God gives rise to an illusory obligation to worship that God. But atheistic modal realists agree with Findlay (Reference Findlay1948, 182–183) when he argues that all possible gods reject worship. Lewis (Reference Lewis2020a, Ltr. 181) also says the gods are worthy of fear, loyalty, veneration, and amazement. Again, religious atheists can venerate (good) gods and strive to be like them. On this point, religious atheists resemble those contemporary pagans who venerate but do not worship their gods (Sylvan (Reference Sylvan2003), 14). An old Wiccan maxim goes like this: when anyone bows down to the Goddess, the Goddess says ‘Rise!’ (Reference Lewisibid., 31). The pagan writer Beckett (Reference Beckett2017, 84) says: ‘We have sovereignty before the gods, even if we are not their equals. Render honor and respect, not submission.’ Since no gods are necessary, no religions are necessary. Many religions are possible.

Different religions are about different worlds

Religionists tell stories about their gods. Yet their stories often contradict each other. The atheist Sam Harris (Reference Harris2014, 19) says ‘It is impossible for any faith, no matter how elastic, to fully honor the truth claims of another.’ For example, Christians say ‘God has a son’ while Muslims say ‘God has no son’. For atheistic modal realists, religious contradictions are resolved in two steps. First, religious truth-claims are treated as fictions: in the Christian fiction, God has a son; in the Islamic fiction, God has no son. Thus atheistic modal realists agree with Harrison (Reference Harrison2010, 52) when she says ‘Different religious discourses no longer need to be perceived as competitors for truth; but can instead be regarded as different fictions.’ Second, these religious fictions are non-empty classes of non-actual worlds: at Christian worlds, God has a son; at Islamic worlds, God has no son.

However, atheistic modal realism differs from the fictionalisms of Eshleman (Reference Eshleman2005) and Harrison (Reference Harrison2010). Unlike their fictionalisms, atheistic modal realism agrees with theists that their religious sentences make truth-claims about existing objects. The Christian God and His Son exist in some Christian world; the Islamic God with no son exists in some Islamic world. These Gods are distinct Gods. Counterparts of those Gods exist in many other worlds with or without counterpart sons. Although all religions are false at the actual world, they are true at their own godly worlds. Thus ‘God has a son’ is true at Christian worlds, but not at Islamic worlds. Vice versa for ‘God has no son’. Actual religious discourse involves silent modal operators (Scott and Malcolm (Reference Scott and Malcolm2018)). These are set in place by linguistic conventions adopted in religious contexts (Lewis (Reference Lewis1969)).

Atheistic modal realists say different religions are about different non-overlapping classes of non-actual godly worlds. To paraphrase Lewis (Reference Lewis1978, 44), these are ‘the collective belief worlds of the [religious] community, comprising exactly those worlds where the overt beliefs [of the religion] all come true’. There are Olympian worlds whose only gods are Olympians. There are Yoruba worlds whose only gods are the orisha. Likewise for Norse worlds, Hindu worlds, Christian worlds, Islamic worlds, Jewish worlds, Shinto worlds, and so on. These worlds also contain humans who perform religious activities like prayer, sacrifice, meditation, divination, and so on. By assigning different religions to different worlds, atheistic modal realists affirm genuine religious diversity. Unlike the pluralism of John Hick, or the perennialism of Huston Smith, atheistic modal realists do not try to reduce religious diversity to some hidden unity. Religions are neither portraits of the same noumenal elephant nor paths up the same sacred mountain. The many earthly religions are not vague or approximate (and thus inaccurate) descriptions of our one actual world; they are accurate descriptions of distinct non-actual worlds.

The religious region of logical space (religious space) is just the class of godly worlds. Lewisian plenitude entails that there are no gaps in religious space. The traditional religions define a basis set of godly worlds. So long as the Lewisian provisos on recombination are satisfied, duplicate parts of distinct worlds can be combined to form other worlds. Alien properties can also be used. Although some aspects of some religions are contradictory, many aspects of many religions are compossible. For many sets of godly worlds, there are ways to copy-paste their parts together to make hybrid godly worlds. There are worlds containing some Olympians along with some Aztec deities. There are hybrid Norse-Shinto worlds, hybrid Catholic-Yoruba worlds, and so on. Consequently, recombination and alien properties ensure that there are no gaps in religious space. The number of syncretic godly worlds vastly exceeds the unmixed traditional godly worlds. Syncretic religions involve passionate commitments to syncretic godly worlds.

How actual people participate in godly worlds

When they discuss religious behaviours, Atran and Norenzayan (Reference Atran and Norenzayan2004) focus on socially coordinated ritual behaviours (chanting, singing, swaying, dancing, etc.). They say these ‘communal rituals rhythmically coordinate emotional validation of, and commitment to, moral truths in worlds governed by supernatural agents’ (Reference Lewisibid., 714). But the role of worlds in their analysis of religious behaviours is far from clear. Their detailed discussion of religious behaviours (Reference Lewisibid., sec. 1.5) does not mention worlds at all. It is easy to display commitments to your worldmates (to your family, your country, and so on). Yet Atran and Norenzayan talk about ‘costly commitments to counterintuitive worlds’. What does it even mean to display a commitment to some non-actual world? Or to display some commitment to a god who is not even one of your worldmates?

Religionists display their commitments to their gods through devotional behaviours like prayer (or sacrifice, thanks-giving, worship, etc.). To perform the ritual acts of prayer to a god is to orisonate. If somebody orisonates to a god in any world that contains that god, then they pray to that god. However, no gods actually exist. Therefore, if any actual religionist orisonates to some god, they do not pray to that god. Since atheists (whether or not they are religious) believe that no gods are actual, they do not pray to gods (or sacrifice, give thanks, worship, etc.). Since only theists (who falsely believe their gods are actual) engage in these behaviours, they are theistic behaviours. Following Lewis, I discuss two ways that theists display commitments to non-actual gods and their worlds.

The first way involves simulated behaviour. Although many actual theists often orisonate to their gods, no actual person ever prays to any god. But plenitude entails that anybody who actually orisonates to some god has a counterpart who performs a type-identical orisonation in a world that contains the intended god. By orisonating in those worlds, those counterparts do pray to their gods – they are praying counterparts. Lewis (Reference Lewis1973, 39) says we have some properties vicariously in other worlds. A person who is actually dishonest ‘is vicariously honest through his honest counterparts.’ He says ‘something vicariously satisfies ϕx at a world i if and only if it has some counterpart at i that satisfies ϕx at i’ (Reference Lewisibid., 40). Thus an actual theist vicariously prays to some god if and only if they are actually orisonating and they have some praying counterpart in their godly world. Their praying counterpart displays commitment to their god. By vicariously praying, an actual theist displays vicarious commitment to their non-actual god.

When an actual theist vicariously prays, they simulate one of their non-actual praying counterparts. More generally, when any actual theist performs any theistic behaviour, they are simulating some counterpart in some non-actual godly world. If a theistic simulation is the simulation of some non-actual godly world in some actual brain, then theists are mentally living in a theistic simulation. Within our actual world, ‘I am praying to my god’ is always false; however, within a theistic simulation, ‘I am praying to my god’ appears to be true. As long as they are mentally living in some theistic simulation, the theist falsely believes they are living in some non-actual godly world. Lewis (Reference Lewis1979b; Reference Lewis1983c) talks about people who make errors of modal self-location. They make these errors by falsely self-ascribing properties. The orisonating theist falsely ascribes to their actual body the property x-is-praying-to-God. When a person mentally lives in a simulation, their mental state is usually classified as dissociative. Atheistic modal realism therefore predicts that theistic behaviours involve dissociation. And such behaviours are central to theistic religions (Schumaker (Reference Schumaker1995); Bronkhorst (Reference Bronkhorst2017); Lifshitz et al. (Reference Lifshitz, Van Elk and Luhrmann2019); Luhrmann (Reference Luhrmann2020)).

The second way that theists display commitments to non-actual gods and their worlds involves simulated identity. Some theistic rituals involve the erroneous self-ascription of identities. Cuneo (Reference Cuneo2016, 97–102) describes a ritual in which you identify yourself with the Prodigal Son from the Bible. He says this is literal identification: ‘the claim is not that I am like the Prodigal in some respect or other but that I am the Prodigal’ (Reference Cuneo2016, 102; his italics). Lewis often talks about people who erroneously ascribe to themselves the property of being someone else. Mad Heimson incorrectly believes he is Hume (Reference Lewis1979b, sec. VI). He falsely self-ascribes the property x-is-identical-with-Hume. Analogously, the theist who successfully performs the Prodigal Son ritual falsely self-ascribes the property x-is-identical-with-the-Prodigal. They are exactly like mad Heimson.

Lewis (Reference Lewis1983c) links erroneous identification with hallucination. When Macbeth mistakenly identifies with some non-actual counterpart, he hallucinates his dagger. And Lewis (Reference Lewis1980) talks about veridical hallucinations. Godly worlds are often extensively empirically indiscernible from our actual world. Mountains and trilobites look the same to both the atheist and the young earth creationist. Atheistic modal realists say the theist who identifies with some religious counterpart in a godly world is living in a veridical hallucination. Atheistic modal realists say theists generally identify with non-actual religious people who are intensely devoted to their gods (gods who are their worldmates). By identifying with some otherworldly devout person, an actual theist displays hallucinatory commitment to their otherworldly god and its godly world.

Erroneous identification is extremely dissociative: it resembles dissociative identity disorders. Atheistic modal realism predicts that theistic religions will involve rituals to induce dissociative trances (e.g. possession trances) in which people channel religiously significant persons (gods, goddesses, angels, demons, etc.). However, since dissociative identity is more extreme than dissociative behaviour, possession trances will be less common. The evidence agrees (During et al. (Reference During, Elahi, Taieb, Moro and Baubet2011); Ventriglio et al. (Reference Ventriglio, Bonfitto, Ricci, Cuoco and Bhavsar2018)). At the most extreme, dissociative trances include dissociative fugues. When travelling in the Holy Lands, some theists suffer fugues known as Jerusalem syndrome: they psychotically identify with some non-actual person from some biblical world.

A person in a dissociative state has beliefs about some non-actual world, beliefs which are quarantined in some relatively isolated mental compartment. Lewis endorses the existence of multiple mental compartments (Reference Lewis1979b, Reference Lewis1983c, Reference Lewis1986, 34–36; etc.). Thus beliefs about other worlds (like godly worlds) are processed in some relatively isolated mental possible world box (Nichols and Stich (Reference Nichols and Stich2000)). When theistic beliefs are quarantined in possible world boxes, this isolation decreases their interactions with beliefs about our actual world. Atheistic modal realism therefore predicts that theistic beliefs will tend to be immune to actual counter-evidence. The data agrees with this prediction (Zamulinski (Reference Zamulinski2003); Talmont-Kaminski (Reference Talmont-Kaminski2014); Van Leeuwen (Reference Van Leeuwen2017)).

Conclusion

Many atheists endorse one-world naturalism: gods do not actually exist; if they do not actually exist, then they do not exist at all; hence gods do not exist at all. But atheism does not require one-world naturalism. An atheist can be a modal realist who affirms that gods exist in non-actual worlds. Atheistic modal realism permits an extremely wide range of possible gods. On the one hand, since no gods are actual, atheism is the correct orientation to our actual world. On the other hand, atheists can believe in and be practically oriented towards possible gods, without thereby committing the theistic illusion of believing that those gods are actual. Thus atheists can be religious. Since no gods are necessary, atheistic modal realism says that all worship is idolatry. It is inappropriate to worship gods, yet it can be appropriate to venerate them and strive to be like them. Atheistic modal realism permits all religions to be true, while preserving their diversity and refusing to reduce them to any occult unity. Atheistic modal realism accounts for the interest of atheists in new fictional religions, and for the emergence of new syncretic religions. It explains how theistic behaviours (like prayer) enable theists to mislocate themselves in logical space. It explains the relevance of dissociation to theism, and argues that theists are often living in a kind of veridical hallucination. Atheistic modal realism is a novel atheistic philosophy of religion, which atheists can put to work in many ways.

Conflict of interest

None.

Footnotes

1. I recognize that Christians have conceived of the Christian God in many different ways. For example, Thomists, Calvinists, process theists, and open theists, among others, may all claim to be Christian, yet conceive of their God very differently.

References

Almeida, M (2017) Theistic modal realism I. Philosophy Compass 12, e12419.10.1111/phc3.12419CrossRefGoogle Scholar
Armstrong, DM (1978) Naturalism, materialism, and first philosophy. Philosophia 8, 261276.10.1007/BF02379243CrossRefGoogle Scholar
Atran, S and Norenzayan, A (2004) Religion's evolutionary landscape. Behavioral and Brain Sciences 27, 713770.10.1017/S0140525X04000172CrossRefGoogle ScholarPubMed
Beckett, J (2017) The Path of Paganism: An Experience-Based Guide to Modern Pagan Practice. Woodbury, MN: Llewellyn.Google Scholar
Bronkhorst, J (2017) Can religion be explained? The role of absorption in various religious phenomena. Method and Theory in Religion 29, 130.10.1163/15700682-12341375CrossRefGoogle Scholar
Buckareff, A (2019) Unity, ontology, and the divine mind. International Journal for the Philosophy of Religion 85, 319333.10.1007/s11153-018-9679-9CrossRefGoogle Scholar
Cameron, R (2009) God exists at every (modal realist) world: response to Sheehy. Religious Studies 45, 95100.10.1017/S0034412508009827CrossRefGoogle Scholar
Cliteur, P (2009) The definition of atheism. Journal of Religion and Society 11, 123.Google Scholar
Collier, M (2019) God's necessity on Anselmian theistic genuine modal realism. Sophia 58, 331348.10.1007/s11841-018-0659-4CrossRefGoogle Scholar
Collier, M (2021) God's place in the world. International Journal for Philosophy of Religion 89, 4365.10.1007/s11153-020-09764-wCrossRefGoogle Scholar
Cuneo, T (2016) Ritualized Faith: Essays on the Philosophy of Liturgy. New York: Oxford University Press.10.1093/acprof:oso/9780198757757.001.0001CrossRefGoogle Scholar
Cusack, C (2010) Invented Religions: Imagination, Fiction and Faith. Burlington, VT: Ashgate.Google Scholar
Davidsen, M (2014) The Spiritual Tolkien Milieu (Dissertation). Leiden, Netherlands: University of Leiden.Google Scholar
Dawkins, R (2008) The God Delusion. New York: Houghton-Mifflin.Google Scholar
Dennett, D (1995) Darwin's Dangerous Idea: Evolution and the Meanings of Life. New York: Simon and Schuster.Google Scholar
Draper, P (2005) God, science, and naturalism. In Wainwright, W (ed.), The Oxford Handbook of Philosophy of Religion. New York: Oxford University Press, pp. 272303.10.1093/0195138090.003.0012CrossRefGoogle Scholar
During, E, Elahi, F, Taieb, O, Moro, M-R, and Baubet, T (2011) A critical review of dissociative trance and possession disorders. The Canadian Journal of Psychiatry 56, 235242.10.1177/070674371105600407CrossRefGoogle ScholarPubMed
Eshleman, A (2005) Can an atheist believe in God? Religious Studies 41, 183199.10.1017/S0034412505007602CrossRefGoogle Scholar
Findlay, JN (1948) Can God's existence be disproved? Mind: A Quarterly Review of Psychology and Philosophy 57, 176183.10.1093/mind/LVII.226.176CrossRefGoogle Scholar
Gericke, J (2017) Metaphysical perspectives on YHWH as a fictional entity in the Hebrew Bible. HTS Teologiese Studies/Theological Studies 73, 4566.10.4102/hts.v73i3.4566CrossRefGoogle Scholar
Gray, J (2019) Seven Types of Atheism. New York: Picador.Google Scholar
Harari, Y (2015) Homo Deus: A Brief History of Tomorrow. New York: Vintage.Google Scholar
Harris, S (2014) Waking Up: A Guide to Spirituality without Religion. New York: Simon and Schuster.Google Scholar
Harrison, V (2010) Philosophy of religion, fictionalism, and religious diversity. International Journal for Philosophy of Religion 68, 4358.10.1007/s11153-010-9251-8CrossRefGoogle Scholar
Hick, J (1976) Death and Eternal Life. New York: Harper and Row.Google Scholar
Johnston, M (2009) Saving God: Religion after Idolatry. Princeton, NJ: Princeton University Press.10.1515/9781400830442CrossRefGoogle Scholar
Kraay, K (2010) Theism, possible worlds, and the multiverse. Philosophical Studies 147, 355368.10.1007/s11098-008-9289-yCrossRefGoogle Scholar
Kraay, K (2018) Invitation to the axiology of theism. In Kraay, K (ed.), Does God Matter? Essays on the Axiological Consequences of Theism. New York: Routledge, pp. 135.Google Scholar
Kragh, H (2009) Contemporary history of cosmology and the controversy over the multiverse. Annals of Science 66, 529551.10.1080/00033790903047725CrossRefGoogle Scholar
Kurzweil, R (2005) The Singularity is Near: When Humans Transcend Biology. New York: Viking.Google Scholar
Law, S (2020) Naturalism versus theism is a false dilemma. Think 56, 103107.10.1017/S1477175620000251CrossRefGoogle Scholar
Le Poidevin, R (2016) Playing the God game. In Buckareff, A and Nagasawa, Y (eds), Alternative Conceptions of God. New York: Oxford University Press, pp. 178192.10.1093/acprof:oso/9780198722250.003.0011CrossRefGoogle Scholar
Lewis, D (1968) Counterpart theory and quantified modal logic. Journal of Philosophy 65, 113126.10.2307/2024555CrossRefGoogle Scholar
Lewis, D (1969) Convention. Cambridge, MA: Harvard University Press.Google Scholar
Lewis, D (1970) Anselm and actuality. Nous 4, 175188.10.2307/2214320CrossRefGoogle Scholar
Lewis, D (1973) Counterfactuals. Cambridge, MA: Harvard University Press.Google Scholar
Lewis, D (1978) Truth in fiction. American Philosophical Quarterly 15, 3746.Google Scholar
Lewis, D (1979a) Counterfactual dependence and time's arrow. Nous 13, 455476.10.2307/2215339CrossRefGoogle Scholar
Lewis, D (1979b) Attitudes de dicto and de se. The Philosophical Review 88, 513543.10.2307/2184843CrossRefGoogle Scholar
Lewis, D (1980) Veridical hallucination and prosthetic vision. Australasian Journal of Philosophy 58, 239249.10.1080/00048408012341251CrossRefGoogle Scholar
Lewis, D (1981) Are we free to break the laws? Theoria 47, 112121.Google Scholar
Lewis, D (1983a) New work for a theory of universals. Australasian Journal of Philosophy 61, 343377.10.1080/00048408312341131CrossRefGoogle Scholar
Lewis, D (1983b) Philosophical Papers, I. New York: Oxford University Press.Google Scholar
Lewis, D (1983c) Individuation by acquaintance and by stipulation. The Philosophical Review 92, 332.10.2307/2184519CrossRefGoogle Scholar
Lewis, D (1986) On the Plurality of Worlds. Cambridge, MA: Blackwell.Google Scholar
Lewis, D (1991) Parts of Classes. Cambridge, MA: Blackwell.Google Scholar
Lewis, D (1993) Evil for freedom's sake? Philosophical Papers 22, 149172.10.1080/05568649309506401CrossRefGoogle Scholar
Lewis, D (1997) Do we believe in penal substitution? Philosophical Papers 26, 203209.10.1080/05568649709506566CrossRefGoogle Scholar
Lewis, S (2015) Where (in logical space) is God?. In Loewer, B and Schaffer, J (eds), A Companion to David Lewis. New York: John Wiley and Sons, pp. 206219.10.1002/9781118398593.ch13CrossRefGoogle Scholar
Lewis, D (2020a) Philosophical Letters of David K. Lewis, I: Causation, Modality, Ontology. Beebee H and Fisher ARJ (eds). New York: Oxford University Press.Google Scholar
Lewis, D (2020b) Philosophical Letters of David K. Lewis, II: Mind, Language, Epistemology. Beebee H and Fisher ARJ (eds). New York: Oxford University Press.Google Scholar
Lewis, D and Kitcher, P (2007) Divine evil. In Anthony, L (ed.), Philosophers without Gods. New York: Oxford University Press, pp. 231242.10.1093/oso/9780195173079.003.0018CrossRefGoogle Scholar
Lifshitz, M, Van Elk, M and Luhrmann, TM (2019) Absorption and spiritual experience: a review of evidence and potential mechanisms. Consciousness and Cognition 73, 102760.10.1016/j.concog.2019.05.008CrossRefGoogle ScholarPubMed
Lougheed, K (2021) The Axiological Status of Theism and Other Worldviews. New York: Springer.Google Scholar
Luhrmann, TM (2020) How God Becomes Real: Kindling the Presence of Invisible Others. Princeton, NJ: Princeton University Press.Google Scholar
Mackendrick, K (2012) We have an imaginary friend in Jesus: what can imaginary companions teach us about religion? Implicit Religion 15, 6179.10.1558/imre.v15i1.61CrossRefGoogle Scholar
Maia, N (2021) Theistic modal realism and causal modal collapse. Religious Studies 57, 120135.10.1017/S003441252000027XCrossRefGoogle Scholar
Mawson, TJ (2013) The case against atheism. In Bullivant, S and Ruse, M (eds), The Oxford Handbook of Atheism. New York: Oxford University Press, pp. 2237.Google Scholar
Milem, B (2019) Defining atheism, theism, and god. International Journal for Philosophy of Religion 85, 335346.10.1007/s11153-019-09702-5CrossRefGoogle Scholar
Moravec, H (1988) Mind Children: The Future of Robot and Human Intelligence. Cambridge, MA: Harvard University Press.Google Scholar
Nagasawa, Y (2016) Modal panentheism. In Buckareff, A and Nagasawa, Y (eds), Alternative Conceptions of God. New York: Oxford University Press, pp. 91105.10.1093/acprof:oso/9780198722250.003.0006CrossRefGoogle Scholar
Nichols, S and Stich, S (2000) A cognitive theory of pretense. Cognition 74, 115147.10.1016/S0010-0277(99)00070-0CrossRefGoogle ScholarPubMed
Oppy, G (2018) Naturalism and Religion. New York: Routledge.10.4324/9780429487866CrossRefGoogle Scholar
Petersen, A (2016) The difference between religious narratives and fictional literature: a matter of degree only. Religion 46, 500520.10.1080/0048721X.2016.1221670CrossRefGoogle Scholar
Pigden, C and Entwisle, R (2012) Spread worlds, plenitude and modal realism: a problem for David Lewis. In Maclaurin, J (ed.), Rationis Defensor: Essays in Honour of Colin Cheyne. New York: Springer, pp. 155176.10.1007/978-94-007-3983-3_12CrossRefGoogle Scholar
Schlesinger, G (1997) Theological necessity. Religious Studies 33, 5565.10.1017/S003441259600368XCrossRefGoogle Scholar
Schumaker, J (1995) The Corruption of Reality: A Unified Theory of Religion, Hypnosis, and Psychopathology. Amherst, NY: Prometheus Books.Google Scholar
Scott, M and Malcolm, F (2018) Religious fictionalism. Philosophy Compass 13, e12474.10.1111/phc3.12474CrossRefGoogle Scholar
Sheehy, P (2006) Theism and modal realism. Religious Studies 42, 315328.10.1017/S0034412506008419CrossRefGoogle Scholar
Soler Gil, F and Alfonseca, M (2013) Fine tuning explained? Multiverses and cellular automata. Journal of General Philosophy of Science 44, 153172.10.1007/s10838-013-9215-7CrossRefGoogle Scholar
Stenger, V (2009) The New Atheism: Taking a Stand for Science and Reason. Amherst, NY: Prometheus Books.Google Scholar
Swinburne, RG (1968) Miracles. The Philosophical Quarterly 18, 320328.10.2307/2217793CrossRefGoogle Scholar
Sylvan, D (2003) The Circle Within: Creating a Wiccan Spiritual Tradition. Woodbury, MN: Llewellyn.Google Scholar
Talmont-Kaminski, K (2014) Religion as Magical Ideology: How the Supernatural Reflects Rationality. New York: Routledge.10.4324/9781315729046CrossRefGoogle Scholar
Tegmark, M (1998) Is ‘the theory of everything’ merely the ultimate ensemble theory? Annals of Physics 270, 151.10.1006/aphy.1998.5855CrossRefGoogle Scholar
Tegmark, M (2008) The mathematical universe. Foundations of Physics 38, 101150.10.1007/s10701-007-9186-9CrossRefGoogle Scholar
Tipler, F (1995) The Physics of Immortality: Modern Cosmology, God and the Resurrection of the Dead. New York: Anchor Books.Google Scholar
Tuggy, D (2017) On counting gods. TheoLogica: An International Journal for Philosophy of Religion and Philosophical Theology 1, 188213.10.14428/thl.v1i1.153CrossRefGoogle Scholar
Vance, C (2016) Classical theism and modal realism are incompatible. Religious Studies 52, 561572.10.1017/S003441251600010XCrossRefGoogle Scholar
Van Leeuwen, N (2017) Do religious ‘beliefs’ respond to evidence? Philosophical Explorations 20, 5272.10.1080/13869795.2017.1287294CrossRefGoogle Scholar
Ventriglio, A, Bonfitto, I, Ricci, F, Cuoco, F, and Bhavsar, V (2018) Delusion, possession, and religion. Nordic Journal of Psychiatry 72, S13S15.10.1080/08039488.2018.1525639CrossRefGoogle ScholarPubMed
Wilson, A (2022) Plenitude and recombination. In Beebee, H and Fisher, A (eds), Perspectives on the Philosophy of David K. Lewis. New York: Oxford University Press, pp. 4165.10.1093/oso/9780192845443.003.0004CrossRefGoogle Scholar
Zamulinski, B (2003) Religion and the pursuit of truth. Religious Studies 39, 4360.10.1017/S0034412502006339CrossRefGoogle Scholar