Why Open Theism Is Natural and Classical Theism Is Not

: The cognitive science of religion (CSR) indicates that belief in supernatural agents, or “gods”, is underpinned by maturationally natural cognitive biases and systems (Natural Religion). It is unclear, however, whether theism is natural. Does the god concept that our cognitive biases and systems give rise to approximate theism? In other words, is Natural Religion “theism-tracking”? As Christian theologians have different views of what God is like, we argue that the answer depends partly on one’s model of God. We discuss two models: classical theism and open theism. We argue that classical theism is far from being natural. The classical divine attributes are very hard to comprehend. Moreover, people naturally conceptualize God as a special sort of person, but the classical God strongly deviates from our cognitive expectations about persons. Open theism is much more natural. However, recent ﬁndings in CSR challenge the suggestion that Natural Religion tracks open theism. The possibility that we are “born idolaters” rather than “born believers” might undermine the Christian doctrine of general revelation and attempts to make CSR compatible with theology.


Introduction
Many scholars working on the cognitive science of religion (CSR) have argued that religion is "natural" (e.g., Barrett 2012a;Bloom 2007;cf. Launonen 2018). In most eras and environments, human minds tend to give rise to roughly similar religious beliefs, for example, about supernatural agents or "gods". 1 Some treat this finding as evidence that belief in God, or gods, is irrational (see van Eyghen 2020). However, others argue that CSR might help clarify, and even support, traditional Christian claims about the natural knowledge of God (Green 2013), or what John Calvin called the sensus divinitatis (Clark and Barrett 2010). Clark and Barrett (2011) have suggested that God may have guided the evolutionary process in order to give rise to the cognitive biases and systems now unearthed by CSR. These suggestions point to an idea known as the doctrine of general revelation-the claim that God has made his existence and basic attributes apparent to all people (see Demarest 1982).
The problem with such theological interpretations of CSR is that, while belief in human-like finite supernatural agents is natural, it is far from clear whether theism (belief in the "God of the philosophers and theologians") is natural. However, if such a God exists, has created humans, and has fashioned the human mind so that it could become aware of his existence and attributes, we should expect the "arriving at the correct god" concept to be cognitively easy. Hence, attempts to make CSR compatible with theological claims about general revelation face the following question: Do the cognitive biases and systems point people to God-or do they only produce beliefs in finite supernatural agents, such as angels, demons, ancestor spirits, ghosts, and goblins? In other words, is natural cognition "theism-tracking"?
In this paper, we argue that one's answer depends partly on one's model of God. Christian theologians hold to very different views of God. There are several models on the theological market: classical theism, neoclassical or modified classical theism, open theism, panentheism, and process theism (Diller and Kasher 2013;Mullins 2016). Different models map differently onto our natural evolved intuitions about supernatural agency.
We begin by explaining what CSR scholars mean by what Barrett calls "Natural Religion". After this, we compare Natural Religion with classical theism and conclude that this model of God is far from being cognitively natural. The classical attributes of timelessness, immutability, impassibility, and simplicity make God maximally counterintuitive. Recent work in CSR suggests that people draw on their core knowledge of persons to conceive what divinity is like (Heiphetz et al. 2016). The classical God does not fit together with our cognitive expectations about persons. If classical theism is true, our cognitive tools are not theism-tracking. The God of open theism, however, resembles a human person in relevant respects. However, even if we take open theism to be true, it is not clear whether Natural Religion is theism-tracking. As classical theists, open theists also view God as omnipresent, omniscient, and incorporeal/disembodied, but these attributes may not be cognitively natural.

Natural Religion
According to Justin Barrett and Aku Visala, religion is cognitively natural in the following sense: [T]here is something about our minds that dispose it to catch religious ideas . . .
[O]ur belief-forming mechanisms would be biased in such a way as to create a tendency or a disposition to acquire, think, and transmit religious ideas instead of some other kinds of ideas. (Barrett and Visala 2018, p. 69) In his book, Why Religion Is Natural but Science Is Not, Robert McCauley (2011, p. 37) gives four marks for what he calls maturationally natural cognitive systems:
Most (not all) natural systems begin functioning early in life; 3.
They are designed for the "fundamental cognitive challenges" that humans historically have faced; 4.
The operations of these systems "do not depend on anything that is culturally distinctive-not on instruction, or on structured preparations, or on artifacts".
According to McCauley, recurrent religious beliefs and behaviors are typically expressions of maturationally natural cognitive systems. The first criterion has to do with intuition. Cognitive science describes two basic types of cognitive operations: intuitive and analytic/reflective thinking (Evans 2003;Kahneman 2011). Intuitions, also known as implicit beliefs, are products of System 1. System 1 operates automatically, quickly, and typically outside of our awareness. Explicit beliefs, however, are products of System 2: the conscious, effortful, and reflective operations of our mind. The question "do you believe in God" is usually a question about someone's explicit beliefs, about what they consciously hold to be true.
A central idea in cognitive science is that explicit beliefs are informed and constrained by implicit beliefs (Barrett 2004;White 2021). While "intuitive" and "natural" are often used interchangeably in CSR (as well as in this paper to an extent), not all scholars do. Boyer (2001) has argued that recurrent religious ideas are typically minimally counterintuitive. This makes them interesting and catchy without ridding them of plausibility. However, highly counterintuitive ideas, while certainly not nonexistent in religions (Sterelny 2018), are never cognitively natural. Such ideas need so-called "cultural scaffolding" to survive. That is, unlike the products of maturationally natural systems, they depend on what is culturally and historically distinctive, for example, on instruction, on structured preparations, or on artifacts. Some theological ideas, such as the doctrine of the Trinity, are highly counterintuitive. They depend upon the cultural scaffolding of the Christian tradition. Only by studying and utilizing System 2 can one begin to understand such ideas. As surveys on Christians' theological beliefs show, the everyday thinking of ordinary believers, pastors, and even professional theologians, often contradicts official theology. In CSR, this phenomenon is known as "theological incorrectness" (Barrett and Keil 1996;Slone 2004).
Justin Barrett (2012aBarrett ( , 2012b coins the term "Natural Religion" to describe cognitively natural religious beliefs. The following excerpt brings together a number of CSR theories regarding our intuitions about supernatural agency: (A) Elements of the natural world, such as rocks, trees, mountains, and animals are purposefully and intentionally designed by someone(s) who must, therefore, have superhuman power; (B) These agents are not human or animal (D) Moral norms are unchangeable-even by gods; (E) Immoral behavior leads to misfortune; moral behavior to fortune; (H) Gods exist with thoughts, wants, perspectives, and the free will to act; (I) Gods may be invisible and immortal, but they are not outside of space and time; (J) Gods can and do interact with the natural world and people, perhaps especially those that are ancestors of the living and, hence, have an interest in the living. This interaction with the world accounts for perceived agency and purpose in the world that cannot be accounted for by human or animal activity; (K) Gods generally know things that humans do not (they can be superknowing or superperceiving, or both), perhaps particularly things that are important for human relations; (L) Gods, because of their access to relevant information and special powers, may be responsible for instances of fortune and misfortune; they can reward or punish human actions. (Barrett 2012b, p. 322) Matthew Braddock summarizes the list into one sentence: "humans are disposed to believe in non-human, invisible, disembodied, immortal, super-powerful, super-knowing, superperceiving, infallible, morally interested, punishing/loving, causally active, and minded agents (with beliefs, desires, intentions, character, and free-will) who possess creator or designer status" (Braddock 2018, p. 178;italics original).
What are the theological implications of Natural Religion? According to Barrett, "CSR provides evidence that humans do have natural propensities toward believing in some kind of god, and perhaps particularly a super powerful, immortal, creator" (Barrett 2012b, p. 324). He also suggests that CSR supports the Apostle Paul's ideas about the natural knowledge of God (Rom 1:18-20), and John Calvin's idea of the sensus divinitatis (the sense of deity) (cf. Clark and Barrett 2010). According to T. J. Mawson (2014, p. 164), "the findings of CSR . . . may be taken by theists as reason to suppose that they are right-God made us (or at least some of us) for himself, so our brains are restless until their find their rest in Him". Alvin Plantinga likewise writes: God has created us in such a way that we can know and be in fellowship with him. He could have done this in many ways; for example, he could have brought it about that our cognitive faculties evolve by natural selection, and evolve in such a way that it is natural for us to form beliefs about the supernatural in general and God himself in particular. (Plantinga 2011, p. 140) These claims indicate that the doctrine of general revelation is a good fit with CSR. Barrett makes clear, however, that Natural Religion deviates from Christian theology in several respects. For instance, it may not be natural to believe in one god only. Moreover, cognitively natural gods are highly anthropomorphic beings. They are supernatural persons (Boyer 2001). While most theologians also say God is a person, our core knowledge of human persons often interferes with our god concept in a way that is theologically problematic. This raises an interesting question for our purposes.
Despite theological incorrectness, are our natural cognitive biases and systems reliable enough? Do they generate belief in God in different eras and environments? Does CSR tend to support or undermine the doctrine of general revelation? On the one hand, while the history of religion is populated with finite supernatural agents, widespread belief in moralizing high gods (such as the Abrahamic Yahweh/Allah) seems like a recent phenomenon (Norenzayan et al. 2016). High gods are also lacking from the religions of many contemporary hunter-gatherer tribes (Peoples et al. 2016). Jason Marsh (2013) argues that prehistoric people lacked the concept of a high god altogether. He cites Barrett who writes that, "Arguably the oldest and most widespread form of god concepts is the ancestor spirit or ghost" (Barrett 2007, p. 775). Since ancestor spirits and ghosts are a far cry from theism, Marsh argues, it was impossible for the earliest humans to believe in God. 2 The theologically correct god concept was simply unavailable. Presumably, in his view, any people without the necessary cultural scaffolding just cannot know God. On the other hand, Braddock (Forthcoming) argues that Barrett's list shows Natural Religion to be "theistic-like". Our cognitive defaults fall near the mark. He also points out that we cannot conclude that prehistoric people did not believe in high gods just because ethnographic or archaeological evidence for them is missing. In this paper, we focus on the psychological evidence regarding natural intuitions. Our question is whether Natural Religion is theism-tracking: Theism-tracking: natural cognitive biases and systems are theism-tracking if the god concept they give rise to approximates the theologically correct model of God.
A few clarifications. First, a god concept is one's cognitively implicit blueprint of the basic attributes of a supernatural agent, be they ontological (e.g., incorporeality), mental (e.g., omniscience), or moral (e.g., benevolence) attributes. Second, models of God are theological theories of the divine attributes of what God is like. Such models are cognitively explicit. Since there are many versions of theism, we need to be clear which one we are talking about when considering whether Natural Religion is theism-tracking. Third, when does an implicit god concept "approximate" a theological model of God? Moreover, why do we employ such a criterion anyway? Would it not be simpler to just ask whether theism is natural? The reason is that the correct model of God does not need to have a perfect fit with Natural Religion for the doctrine of general revelation to be true. However, a model of God should not be too counterintuitive either. Some counterintuitiveness is allowed. After all, scholars, such as Boyer, argue that religious ideas are natural exactly because they are minimally counterintuitive.
Fourth, it is important to recognize that god concepts never come about in a cultural vacuum. Religion is never purely natural, and Natural Religion is not a real religion. Religion takes different forms in different cultures. CSR scholars often compare the naturalness of religion with the naturalness of learning a language. Learning a language always means learning some particular language, such as Finnish or Chinese. Likewise, religious beliefs and practices need human communities to develop. For this reason, it cannot be expected that human cognition gives rise to theistic-like beliefs without any cultural support. What is important is that theism does not require a particular cultural environment to develop.
Fifth, natural cognition can be theism-tracking, even if the belief in nontheistic gods, ghosts, and goblins is also natural. Christians have traditionally believed in the reality of angels, demons, and immaterial souls. From this viewpoint, natural cognition is not completely off-track in supporting belief in finite and creaturely supernatural agents, even if it gives rise to false god-beliefs as well. The question is whether theism is likewise natural, or close to natural.
Finally, as a reminder, what makes this question interesting is that the answer can either support or undermine the Christian claim about general revelation and, consequently, the attempts to make CSR compatible with theology. Having a relatively correct concept of god has been traditionally viewed as a precondition for theistic belief, and the generation of theistic belief has been viewed as God's primary purpose for general revelation. 3 If theistic belief were counterintuitive, unnatural, and only showed up around twelve thousand years ago, it would speak against the claim that God has created all humans for a fellowship with him. Even the existence of God itself may be on the line. According to Marsh (2013), the naturalness of nonbelief serves as evidence against God. According to Braddock (2018), however, our "supernatural disposition" is more surprising given naturalism, but less surprising given theism. In our view, the question cannot be settled without comparing our implicit concept of god with particular theological models of God. In the next section, we consider whether Natural Religion is theism-tracking if classical theism is the correct model.

Classical Theism Is Not Natural
Classical theism is said to be the doctrine of God that one finds in historical Christian thinkers, such as Augustine, Anselm, and Thomas Aquinas, and that is defended by the Catholic and Protestant scholastics. Classical theists (as well as theists of various other stripes) affirm that God is a necessary being who is omnipotent, omniscient, morally perfect, and perfectly free. They also affirm the doctrine of creation ex nihilo. By looking at Barrett's list, one can see that many of these attributes seem to fit well together with our natural intuitions about gods as superpowerful, superknowing, and so on. Yet those attributes are not unique to classical theism. What distinguishes classical theism from its rivals is its commitment to a God who is timeless, immutable, simple, and impassible (Mullins 2021a). These attributes turn out to be highly counterintuitive. We give a short overview of each before discussing their cognitive fit.

Four Classical Attributes
According to classical theism, God is timeless if, and only if, God necessarily exists without beginning, without end, without succession, and without temporal location. This means that God does not do one action and then another action. God does not experience one event and then experience another event. Instead, classical theists say that all of God's acts and experiences occur at a single timeless moment, or the eternal now. This timeless moment does not stand in any kind of temporal relation to the world. God is not temporally before or after any event in time. God is not simultaneous with any moment of time. God does not exist right now since the present is a temporal location. Timelessness is systematically connected to divine immutability. As immutable, God cannot undergo any kind of change, be it intrinsic or extrinsic. God cannot change in relationship to other things since that would involve God undergoing succession from one moment to another moment of time.
Divine impassibility makes three claims. First, it is impossible for God to suffer. Second, it is impossible for God to be caused, moved, or influenced by anything outside of himself. God cannot be influenced by anything external to think, feel, act, or be in any particular way. God is completely and utterly uninfluenceable. Third, God lacks so-called passions. This is a bit difficult to grasp because the English term emotion covers a wide range of affective states that include things that classical theists wish to affirm and deny of the impassible God. This is why contemporary work on impassibility has identified three inconsistency criteria used throughout the classical tradition for discerning which emotions the impassible God can have. The claim is that it is impossible for God to have any emotion that is inconsistent with God's perfect moral goodness, perfect rationality, and perfect happiness. An impassible God is in a state of pure undisturbed bliss or happiness that is grounded entirely in himself. Nothing can move or influence God to feel anything other than pure happiness. In fact, nothing can move or influence God to feel anything at all since it is impossible to influence the impassible God. Because God cannot be moved or influenced by anything external to himself, the impassible God does not have any empathy or compassion. (Davies 2006, p. 234;Mullins 2021b, pp. 19-22.) Divine simplicity says that God is not made up of parts. This might seem like a trivial statement, but the classical theist is quite permissible in what counts as a part. The classical theist says that distinct attributes, actions, thoughts, feelings, and so on are all parts. Therefore, in reality, God does not have any distinct attributes, actions, thoughts, or feelings. Instead, all of God's actions are identical to each other such that there is only one act. This act is identical to God's existence. Furthermore, all of God's so-called attributes, thoughts, and feelings are identical to each other and identical to God's existence. God is just a simple indivisible substance, without any distinctions at all.

Natural Religion and Classical Theism
These attributes deviate from Natural Religion. As Jonathan Jong, Christopher Kavanagh, and Aku Visala have argued, this is why CSR says virtually nothing about belief in the classical God (Jong et al. 2015). Such a being is totally different from cognitively natural supernatural agents. For example, as Item (I) shows, cognitively natural gods "are not outside of space and time". Thus, they are not omnipresent or timeless. 4 Consider also items (B) and (L): "Things happen in the world that unseen agents cause" and gods "may be responsible for [some] instances of fortune and misfortunes". While the classical God is likewise causally active, he is also the prime cause of absolutely every event. This idea does not seem like it is cognitively natural: [T]here is a sense that the God of classical theism does everything, such that it is impossible to point at particular events that God causes as if to distinguish them from those that God does not cause. It is impossible to point at a particular bush that God rustles, a particular gust of wind in which God is present; God rustles all bushes and is present in all gusts of wind. Thus, insofar as the cognitive mechanisms described above-considered by some to be a "god faculty"-detect God in this bush but not that, this gust of wind but not that, then it is not a particularly good detector of God, who is everywhere always acting in all things, causing them to be. (Jong et al. 2015, p. 256) Next, consider item (H): Gods exist with thoughts, wants, perspectives, and the free will to act. God's free actions become rather odd on classical theism. The classical theist, Kathrin Rogers, says that an action is something that a person does, not something that a person is. Yet, Rogers also says that a simple God is his act (Rogers 1996, pp. 172-73). As Rogers admits, that is counterintuitive. The idea of a timeless action is even more difficult to grasp. Typically, a free action is conceptualized as occurring over a series of moments of time. Consider a human person named Sally. At this present moment, Sally is sitting. It is too late for her to make a choice about sitting or not sitting. Any choice she makes will be about actions that she performs at subsequent moments of time. She has several options about what she will do at the next moment. She could sit up or remain seated. Whatever she decides to do at the next moment is up to her. However, this does not capture the actions of a timeless person. A timeless God can only do whatever he is timelessly doing. A timeless God cannot transition from not performing an action to performing an action. Nor can a timeless God transition from not making a choice to making a choice. This makes it very difficult to understand how God is freely doing anything. It seems like God is just acting, and never has the opportunity to make a choice about what that act is. Many religious people will speak of what God has done in the past and will say that God will do certain things in the future. Given classical theism, this is incorrect.
Item (H) also includes the notion that gods have wants. This seems to contradict divine impassibility. A being in a state of undisturbed bliss or happiness, uninfluenced by anything outside himself, does not have wants or desires (Mullins 2021b, pp. 44-45). Moreover, because God is impassible, he cannot really "interact with the natural world and people" (J) in any normal sense of the term "interact". The classical theist, James Dolezal (2017, p. 2), resists a popular view of God he calls "theistic mutualism", the idea that "God is involved in a genuine give-and-take relationship with his creatures". However, of course, this is exactly what normal believers take to be true. They believe that God can influence them and that, through prayer, they can influence God. Impassibility might also make it difficult to hold on to the idea that God really rewards or punishes human actions (L), since we typically conceive of such actions as reactions to human behavior (Mullins 2021b, pp. 52-53).

God's Personhood
Such observations may cause one to question whether the classical God is actually a person. Some contemporary classical theists do say that God is not a person (e.g., Davies 2006, p. 61). However, this is contrary to what the actual classical Christian tradition says. Personhood is classically taken to be a perfection that one must predicate on God (Wiertz 2016, p. 45;Marschler 2016, p. 85). The classical understanding of a person is an individual substance of a rational nature. Sometimes, philosophers state that a person is a mental substance with the capacity for consciousness, feelings, and intentional actions (Swinburne 2016, p. 105). It is difficult to say that God is omniscient, omnipotent, and has free will without saying that God is conscious and performs intentional actions. Furthermore, the God of classical theism is in a state of pure happiness. Therefore, the God of classical theism clearly has the marks of personhood: consciousness, feelings, and intentional actions.
Nevertheless, the classical attributes run counter to our natural expectations about persons and minds. There is evidence that humans employ their core knowledge of human persons/minds to understand God's mind and personhood. For example, even though gods are generally viewed as having supernatural mental capacities, children start off by viewing divine minds as similarly limited, such as human minds (Heiphetz et al. 2016). Moreover, the theory of mind is a cognitive mechanism that is instrumental in our understanding of other people's mental states-and God's (Schjoedt et al. 2009;Bering 2002). Mentalizing, or "mindreading", is easy for humans because of this mechanism. Mentalizing skills also seem to foster belief in a personal God (Norenzayan et al. 2012). Therefore, the finding that "gods exist with thoughts, wants, perspectives, and free will to act" (H) implies that cognitively natural gods are essentially persons, not unlike human persons. However, the classical God is unlike a human person in many respects.
For example, we know intuitively that the mental states or acts of persons are not one and the same for all eternity. Rather, they are distinct, and one follows another. All the mental states or acts of persons are certainly not identical to one another, nor are they identical to the existence of persons themselves. Moreover, most persons we know have a wide range of emotions. They interact by responding and reacting positively, neutrally, or negatively to things and events around them. We never encounter a person who is in a state of pure undisturbed happiness. Something seems perverse about a person that is completely undisturbed by what happens in the world. Even the classical theist, Thomas Aquinas, argues that a virtuous person must be disturbed by witnessing a truly tragic event. If a person is not disturbed by witnessing a tragic event, Aquinas says that this person must lack moral knowledge or have a vicious moral character. Perhaps this person is undisturbed because she doesn't understand all the relevant moral facts of the situation. Or maybe she is just a terrible person and does not care about the tragic event (Aquinas, Summa Theologiae 2.Q39.a2). We find Aquinas's claims to be intuitive-a morally good person is upset by witnessing a tragic event. Yet, we find Aquinas's affirmation of impassibility counterintuitive. This is because it is impossible for an impassible God to be disturbed by anything outside of himself. It is impossible for an impassible God to experience anything other than pure happiness. This cannot mean that the classical God lacks moral knowledge or has a dubious moral character, for he is omniscient and perfectly good. So there seems to be a conflict between the notion of impassibility and what we normally take to be a good person (Wolterstorff 2010, pp. 223-38). At the very least, this shows that the classical God's personhood is very different from ours.

Implications for the Doctrine of General Revelation
If classical theism is true, and yet our cognitive biases lead us to conceptualize God as a person not unlike human persons, there is something to the suggestion that we are "born idolaters" rather than "born believers". 5 Our cognitive mechanisms are not theism-tracking. Most believers fashion God in the image of man, the Creator in the image of a creature. In fact, classical theists may agree. The classical tradition itself admits that the classical attributes are difficult to grasp. For instance, divine simplicity is often systematically connected to ineffability, or the unknowability of God (Hick 2000). Many defenses of divine simplicity often appeal to the ineffability of God because they see simplicity and ineffability as going hand-in-hand (Feser 2017, pp. 224-26;Dolezal 2011, pp. 206-12). Classical theologians likewise acknowledge that the Bible itself is permeated with anthropomorphisms. Some of them argue that God does not reveal himself as timeless or impassible in scripture because humans are not mature enough to understand these deep truths about God (Helm 2001, pp. 44-47;Dolezal 2019, p. 33). For example, it is too difficult for us to understand that God does not have any compassion, so God must reveal himself as having compassion in order to draw us closer to himself (Helm 1990, pp. 133-34).
This suggests that neither general revelation, nor special revelation (the Bible), adequately help humans understand God as he truly is. However, as Stephen T. Davis points out, Christianity is based on the assumption that God has revealed himself as who he really is (Davis 2017, p. 566). Here, we are concerned especially about the idea of general revelation, or the natural knowledge of God. According to this doctrine, "man at large knows both that there is a God and in broad outline what he is like" (Demarest 1982, p. 14). However, if classical theism is true, then true knowledge of God seems dependent on a particular cultural setting. Classical theologians, such as Augustine and Aquinas, took Greek philosophy to be a prime example of how God can be known without access to special revelation. As Adam Green (2013) notes, philosophy and theology are rare cultural achievements in human history. Their products, including classical theism, are not cognitively natural: Philosophy and theology as cultural phenomena are likely dependent on literacy and on sufficient prosperity to allow for the cultivation of expertise in the construction of abstract systems of ideas. The religious impulse of human beings is much more ancient and more widespread. Furthermore, from a CSR perspective, religion is thoroughly enmeshed with the social mind being engaged in a pragmatic mode. Though doctrines and practices can be built on the foundations of natural religion and this superstructure of orthodoxy can circumscribe or even contradict natural religion, doing so will require a great deal of cultural scaffolding. (Green 2013, p. 410) If classical theism is the correct model of God, only a few theologians and philosophers have ever attained true knowledge of God. Now, the classical theist might respond by saying that the theologically incorrect intuitions are the result of the cognitive consequence of sin (see Peels et al. 2018). For example, because of the "noetic effects of sin", Calvin did not imagine sensus divinitatis to produce belief in the one true God, even if this was its original purpose (Helm 1998). He believed that the only path to knowing God in a postlapsarian world is through the Bible. However, since reading the scriptures rarely makes people classical theists, this response will not help the classical theist much. Moreover, many Christians themselves display a theologically incorrect concept of God. Given the Christian doctrines on regeneration and sanctification, we would expect the Holy Spirit to be gradually healing the noetic effects of sin and, as a result, believers' understanding of God to be gradually becoming more and more correct (Launonen 2021).

Open Theism Is Cognitively Natural
Open theism is a model of God that has a better fit with Natural Religion than classical theism. 6 As noted above, most models of God affirm that God is a necessarily existent being, with attributes such as omniscience, omnipotence, omnibenevolence, and perfect freedom. Open theism is like classical theism in affirming these divine attributes. Yet, open theism presents God as a relational being, a person with a give-and-take relationship with humans. The open theist, Clark Pinnock, writes: "We are not dealing with an unapproachable deity but with God who has a human face and who is not indifferent to us but is deeply involved with us in our need" (Pinnock 1994, p. 102;italics ours). This model of God also rejects the four classical attributes. The open God is temporal, mutable, passible, and unified. 7 Hence, when the open theist says that God exists "with thoughts, wants, perspectives, and free will to act" (H), he uses these terms in a very normal sense.

Basic Characteristics of Open Theism
David Basinger (1994, p. 156) lists five basic characteristics of open theism:

1.
God not only created this world ex nihilo, but can (and at times does) intervene unilaterally in earthly affairs; 2.
God chose to create us with incompatibilistic (libertarian) freedom-freedom over which he cannot exercise total control; 3.
God so values freedom-the moral integrity of free creatures and a world in which such integrity is possible-that he does not normally override such freedom, even if he sees that it is producing undesirable results; 4.
God always desires our highest good, both individually and corporately, and, thus, is affected by what happens in our lives; 5.
God does not possess exhaustive knowledge of exactly how we will utilize our freedom, although he may well, at times, be able to predict, with great accuracy, the choices we will freely make.
The first three claims address God's power. While both classical theists and open theists affirm divine omnipotence, in practice, open theists see God's power as more limited. They maintain that, in creating the world, God has voluntarily limited the use of his power in order to make room for human free will. Barrett's list indicates that people naturally think of gods as having "superhuman power", (A). Now, this could indicate that omnipotence is also cognitively natural. What theologians usually mean by omnipotence is that God can do anything that is logically possible. However, humans may not intuitively think of God as all-powerful in this sense. Perhaps the idea that God's power is limited is more natural. Moreover, according to open theism, God limits his power because of the "moral integrity of free creatures". This coheres with item (D): "Moral norms are unchangeable-even by gods".
As the fourth claim shows, the open God is not impassible, but has desires, (H), and can be affected by humans. Open theists also embrace the idea that prayer can influence God. According to Basinger (1994, p. 156), "[W]ithin most Christian traditions it is quite clearly held, to use the words of David Mason, that believers 'are to ask God for things' and that God 'hears, is affected by our importunities, and responds adequately to them'". This coheres with the cognitively natural idea that "Gods can and do interact with the natural world and people" and "have an interest in the living" (J). Moreover, the causal power of humans is more real on open theism than on classical theism. God is not the cause of every single event, but the created reality has a considerable amount of autonomy. From this perspective, it makes more sense to say that "things happen in the world that unseen agents cause" (B).
The fifth claim is what open theism is probably best-known for: God does not possess exhaustive foreknowledge of the future. Open theists agree with classical theists that God is omniscient. This means that God knows of the truth-values of all propositions, or God knows all of the facts of reality. Classical theists and open theists agree that, prior to creation, there are many possible future timelines, or ways, that history could evolve (Rhoda 2011). What they disagree over is what is included in the total set of facts. The open theist says that if God creates a universe with free creatures, then history could evolve in several possible ways. God knows all of the possible ways that history could evolve, and God knows the objective probability of history evolving in any particular way. Yet, there is no specific timeline, no single way, that history will in fact unfold.
Do people naturally conceptualize God's knowledge like this? In our experience, ordinary believers typically think of God as having perfect knowledge of the future. What they do not often think about is that libertarian freedom (the cognitively natural concept of free will) may conflict with God's exhaustive foreknowledge of the future, as open theists claim. However, this aspect of open theism is clearly the conclusion of a philosophical argument (System 2). Thus, it does not seem to be cognitively natural (System 1). However, as we will see, perfect omniscience may not be natural either.

A Modified Account of Natural Religion
Is Natural Religion theism-tracking if open theism is true? So far, it may seem so. However, there is evidence that makes Barrett's description of Natural Religion somewhat outdated. Just like classical theists, open theists also believe that God is transcendent, omnipresent (not restricted by place), that God does not have any false beliefs, that God is incorporeal/disembodied, and that God is a moralizing deity who rewards the good and punishes the evil (e.g., Swinburne 2016). There are reasons to question whether these attributes are cognitively natural. As we will see, even professing Christians display theologically incorrect intuitions regarding them. According to the so-called coexistence model, acquired theological information about God does not replace the cognitive tendency to draw from our core knowledge of persons when thinking about God (White 2021, pp. 123-28, 134-37). Our implicit beliefs often conflict with our explicit beliefs-yet both manage to live together in our heads.
First, consider again Item (I) on Barrett's list: gods are not outside of space and time. In everyday life, Christians tend to think of God as limited by time and space to some extent. In a study, Barrett and Keil (1996) had participants-some of the Christians-read some version of a short story. In one version, a boy gets his foot caught between rocks while swimming. The boy prays to God and God saves the boy, despite answering another prayer on another side of the world at the same time. The participants were then presented with recall items about the story they had to recognize as either correct or incorrect. Some of the incorrect claims presented God as limited, even though the original story had not explicitly done so.
[The] subjects seemed to characterize God as having to be near something to receive sensory information from it, not being able to attend differentially to competing sensory stimuli, performing tasks sequentially and not in parallel, having a single or limited focus of attention, moving from place to place, and sometimes standing or walking. (Barrett and Keil 1996, pp. 229-30) Many incorrect recall items were recognized as correct if they depicted God as somehow human-like. 8 Importantly, these characterizations conflicted with the subjects' own explicit, culturally acquired, theologically correct description of God. In answering a survey about God's attributes, almost all agreed that God is everywhere, can perform multiple mental activities at once, is everywhere/nowhere/many places at once, and that God need not be near anything to see, hear, smell, taste, or touch it. Therefore, even if God is naturally conceptualized as superknowing and superperceiving (item K), this does not mean omni-perception/attention is intuitive. Barrett sees "no reason to believe that children or adults find the property of unlimited attention the least bit natural or intuitive" (Barrett 2012a, p. 143).
Is this a problem for the naturalness of open theism? Earlier, we noted that open theists affirm that God is temporal, and, therefore, not outside of time. In open theism, God exists in the present, just like creatures do. This fits nicely with Barrett's understanding of Natural Religion. When it comes to space, however, things appear to be different. The open theist, William Hasker, maintains that "God must, somehow, transcend space" (Hasker 1989, p. 178). In standard definitions of omnipresence, God is said to be present to all spatial locations in virtue of the exercise of his power and knowledge. God is present everywhere because God is causally sustaining everything in existence, and God knows what is presently happening at every location. In his discussion of omnipresence, the open theist, Richard Swinburne, says the disembodied God is present to all spatial locations because God can cause effects at every place, and knows what is happening at every place. Furthermore, God's knowledge of what is happening at any given location is noninferential and does not rely on sensory input from some physical organism (Swinburne 2016, p. 113). This seems to conflict with the natural intuition that gods are not outside of space.
Second, consider the open theist claim that God is omniscient. Open theists are adamant that God infallibly knows all the facts of reality. Is this naturally intuitive? Experiments with children and adults suggest that people have intuitions about God having false beliefs. Michael Barlev et al. (2017) examined whether core intuitions about persons interfered with people's acquired theological beliefs. Christian adult participants were presented with four kinds of claims about God: two types of consistent claims (i.e., theologically and intuitively true (e.g., "God has beliefs that are true"), or theologically and intuitively false (e.g., "All beliefs God has are false")) and two kinds of inconsistent claims (i.e., true on intuition but false theologically (e.g., "God has beliefs that are false), or false on intuition but true theologically (e.g., "All beliefs God has are true")) claims. According to the authors, "Participants were less accurate and slower to respond to inconsistent versus consistent statements, suggesting that the core intuitions both coexisted alongside and interfered with the acquired beliefs (Experiments 1 and 2)" (Barlev et al. 2017, p. 425). In the third experiment, participants were made to respond under time pressure. In this case they were far more likely to make errors on inconsistent versus consistent statements. This suggests that System 1 intuitions about persons kick in while System 2 (with access to correct theological information) is still starting its engine. In other words, perfect omniscience, even about present facts, does not seem to be perfectly intuitive.
Third, virtually all theists claim that God is a disembodied spirit. By using a similar methodology, Barlev et al. (2019) have also found evidence that Christian adults naturally conceptualize God as embodied rather than disembodied. This contradicts the popular idea that people are, by nature, Cartesian dualists and view other persons essentially as minds rather than mind-body composites (Hodge 2008). On the one hand, the evidence is unsurprising given the depictions in the Hebrew Bible of God as walking in the Garden of Eden, sitting on his throne, and engaging in other physical activities. Countless numbers of artists have also represented God as embodied (it is difficult to paint a purely spiritual being, after all). On the other hand, it seems clear that people explicitly know such representations of God to be anthropomorphisms. Nevertheless, apparently Christians have to think through their answer a few seconds longer when they are presented with a question that is theologically incorrect, but in line with their core knowledge of persons (e.g., "God is at my church when he is not at other churches"). Just as in the case of God's fallible beliefs, intuitions about human persons coexist and interfere with theological conceptualizations of God.
Fourth, consider the open theist's claim that God is morally invested in the universe. It is not clear whether cognitively natural gods are morally interested (E). The so-called "big gods" account of the cultural evolution of religion suggests that moralizing deities only emerged around twelve thousand years ago (Norenzayan et al. 2016). When agriculture was invented and large groups of people began living and working together, free riding (reaping the benefits of cooperation but not paying any costs) also became a problem. A heavenly big brother was needed to weed out free riding. "The gods of small-scale societies", however, "are typically cognitively constrained and have limited or no concern with human affairs or moral transgressions" (Norenzayan et al. 2016, p. 7).
Braddock's summary of Barrett's account of Natural Religion may thus be revised accordingly: humans are disposed to believe in nonhuman, disembodied or embodied, immortal, superpowerful, superperceiving, but not perfectly perceiving, superknowing, but sometimes erring, punishing/rewarding 9 , or morally disinterested, causally active, and minded agents (with beliefs, desires, intentions, character, and free will) who possess creator or designer status and are restricted by space and time. Natural Religion begins to seem less theistic-like, even on open theism. This modified account seems to support Marsh's claims that prehistoric humans did not have the concept of a high god. It also seems to undermine the doctrine of general revelation. Nevertheless, perhaps there is a way to salvage the theism-tracking thesis by making Natural Religion and open theism somewhat compatible.

Salvaging the Theism-Tracking Thesis
Consider the following responses to each of the four findings discussed above. First, we return to the issue of omnipresence and Item (I) on Barrett's list: natural gods are not outside of space. Contemporary philosophical discussion on the attribute of omnipresence is more complicated than it might at first appear. The current literature on omnipresence is divided on all manner of issues (cf. Arcadi 2017). There is even debate as to whether or not omnipresence actually entails God being spaceless or outside of space. Earlier, we noted that many theologians affirm that God is spaceless (e.g., Helm 2010, p. 41). The medieval scholar, Robert Pasnau, has called this into question. According to Pasnau, Although it is now commonly supposed that God exists outside of space, this was not the standard conception among earlier theologians. Medieval Christian authors, despite being generally misread on this point, are in complete agreement that God is literally present, spatially, throughout the universe. One simply does not find anyone wanting to remove God from space, all the way through to the end of the seventeenth century. (Pasnau 2011, p. 19) If Pasnau is correct, this will provide some help in making omnipresence more intuitive. Yet, there is still the issue of God being "more" present in certain locations than others. Current discussion on omnipresence might help with this issue as well. "Omnipresence", writes the self-proclaimed modified classical theist John Peckham, "does not entail uniformity of presence (as sometimes assumed)." Consistently holding that God is omnipresent, but not uniformly present, merely requires that in addition to God's general omnipresence, God may be specially present to particular locations in some way(s) he is not present to other locations. This may be so whether the manner of divine omnipresence is understood as derivative or nonderivative. For example, God might be derivatively omnipresent while also specially present in specific locations in some derivative fashion. Perhaps God is omnipresent in that God's power extends everywhere, sustaining all things (Heb. 1:3), while instances of special divine presence are special manifestations of divine power in creation (i.e., instances of special divine action) that do not involve God being spatially present. In this way, general omnipresence and special divine presence could involve the same kinds of presence while being distinguished in terms of degree. (Peckham 2021, p. 88) Our point is that theologians who are passionate about describing God correctly sometimes say that God is not present at all places at the same time in a similar fashion. This suggests that the participants in Barrett and Keil's (1996) study may not have been so theologically incorrect after all.
Second, Christian theology does not reject the claim of God as embodied. The doctrine of incarnation means that God (specifically, the second person of the Triune God) once assumed a human body-and remains embodied forever! Many theologians also believe God appeared to his people many times in a bodily form already before Christ. For instance, the patriarch Jacob is often believed to have wrestled with Yahweh himself (Gen 32:22-32) (e.g., Peckham 2021, p. 91). While most theologians would say God is essentially incorporeal, conceptualizing God as embodied may not be theologically incorrect. In fact, some even say that standard definitions of God's omnipresence entail that God is embodied in the universe (Mawson 2006).
Third, as we saw, while open theists say God is omniscient, they don't believe God infallibly knows the future. While God knows the objective probability of history evolving in any particular way, God's expectations about the future can, nevertheless, turn out to be false since he cannot predict the future free acts of humans with perfect accuracy. Regarding divine guidance, Basinger (1994, p. 165) writes that, "It is always possible that even that which God in his unparalleled wisdom believes to be the best course of action at any given time may not produce the anticipated results in the long run". Even though, technically, this does not mean that God has false beliefs, open theism comes close to suggesting that God is susceptible to epistemic error. To be sure, the theologically incorrect intuitions people displayed in the study of Barlev et al. (2017) were not specifically about God's beliefs about the future. Nevertheless, the idea that God has false beliefs is not so far from the suggestion that what God "believes to be the best course of action" may not turn out to be the best after all.
Fourth, not all CSR scholars agree with Norenzayan et al. (2016) that moralizing big gods depend on cultural scaffolding, namely, on large-scale cooperation that creates the need for supernatural surveillance. Dominic Johnson (2016) argues that the fear of supernatural punishment is cognitively natural. Evolutionary, neuroscientific, and psychological evidence indicates that we are "wired for punishment" (Johnson 2016, p. 29). The concept of supernatural punishment is also prevalent among culturally isolated groups living in Hawaii, Micronesia, and Sub-Saharan Africa, for examples. Johnson also cites anthropologists, such as Harvey Whitehouse (2008), who list supernatural punishments and rewards among the "twelve characteristics that tend to be found among all religions, irrespective of period, continent, or culture" (Johnson 2016, p. 58).
One more general point. The path from intuitive/implicit beliefs to explicit beliefs is not straightforward (see Vainio 2016). On the one hand, many theological beliefs prevalent in the West depend on particular cultural scaffolding, on schools and churches, pastors and theologians, books and blogs. On the other hand, people have always engaged in analytic thinking. The countless number of counterintuitive ideas in indigenous religions indicate that Natural Religion does not exactly correspond to the explicit religious beliefs of any culture (cf. Sterelny 2018). Therefore, even if our natural intuitions do not perfectly correspond to any particular model of God, perhaps it takes only a bit of analytic thinking, or a minimal amount of cultural scaffolding, to arrive at a theistic-like concept of god. For example, even if people display intuitions about embodied gods when forced to respond quickly, is it not obvious that most people conceptualize supernatural agents as essentially spiritual beings. The open theist could, thus, argue that God has created people with sufficient theism-tracking cognitive systems, such that it only takes a minimal amount of reflection and general cultural scaffolding to arrive at a theistic-like god concept.

Concluding Remarks
We have argued that the answer to the question of whether Natural Religion tracks theism partly depends on one's model of God. Classical theism makes Natural Religion seem too far off the mark. While open theism is much more natural than classical theism, even here it is not clear whether Natural Religion is theism-tracking. More work is needed in order to properly answer this question. Archaeological, ethnographic, and other kinds of data must be taken into account. Moreover, the implications of our conclusions for the doctrine of general revelation need further scrutiny. One could argue that all the doctrine indicates is that people do not need any special revelation to know that there exists a powerful creator who rewards and punishes (a creator and a judge). These two ideas can be defended as cognitively natural.
Author Contributions: All authors contributed equally to this paper. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Conflicts of Interest:
The authors declare no conflict of interest.
Notes 1 In all that follows, "God" with a capital G stands for the God of Christian theism while "god" stands for any supernatural agent. 2 According to Marsh, this fact bolsters the argument from divine hiddenness. This argument claims that God does not exist because "nonresistant nonbelief" does (Schellenberg 2015). Nonresistant nonbelief is lack of belief in God that is not due to the person resisting a relationship with God. According to the argument, any being worthy of the name God would be perfectly loving. A perfectly loving God would always be open for a relationship with creatures as long as they are also open. 3 Alternatively, one could view a correct concept of God as a precondition for a relationship with God. Some claim that one can be in a relationship with God without believing in God (see Schellenberg 2017). 4 It is worth noting that there is a debate over the exact nature of omnipresence. The classical theist Paul Helm says that the omnipresent God is spaceless (Helm 2010, p. 41). Yet the medieval scholar Robert Pasnau says that the classical Christian tradition did not affirm that God Is spaceless until the end of the 17th Century (Pasnau 2011, p. 19). 5 See the titles of Jong et al. (2015) and Barrett (2012a). 6 There may be also other models of God that seem similarly cognitively natural, such as some versions of neoclassical or "modified" classical theism (e.g., Peckham 2021). However, open theism is a clearly defined model of God and thus serves as an example. 7 God is unified in that all of God's essential attributes are co-extensive and compossible or coherent. 8 Regarding the methodology in Barrett and Keil's study, it has been pointed out that the stories themselves include anthropomorphic cues (Shtulman 2008). However, for the sake of argument, we will not appeal to this criticism here.