Introduction
Let's say that idolatry is religious activity centred on a real or imagined entity other than the God of classical theism.Footnote 1 While idolatry has often been discussed as an accusation lodged by one religion against another, we introduce, develop, and defend two irreligious arguments, originating with David Hume, utilizing idolatry, which we call the ‘argument from idolatry’ and the ‘argument from indifference’. Both suppose that if God desired for humans not to be idolatrous, the human cognitive apparatus would not gratuitously incline humans to idolatry. The two arguments differ on whether classical theism entails that God desires for us not to be idolatrous. The argument from idolatry supposes the god of classical theism would desire for us to avoid idolatry; in that case, our gratuitous natural inclination to idolatry entails that no such god exists. In contrast, according to the argument from indifference, classical theism does not entail that God would desire for us to avoid idolatry; the argument concludes that God does not care whether we avoid idolatry.
In the next two sections, we offer the arguments from idolatry and indifference. Then, we begin our discussion of Hume's psychology of religion and relate Hume's theory of action to two general categories of religious activity: superstition and enthusiasm. The following section offers an interpretation of superstition, while updating and vindicating Hume's account with work in the social scientific study of religion. Likewise, the subsequent section offers an interpretation of enthusiasm and, again, updates Hume's account. We reinterpret superstition and enthusiasm as two widespread, significant, natural tendencies (a superstitious tendency and an enthusiastic tendency) and suggest that both produce (or are) gratuitous inclinations to idolatry.
The argument from idolatry
The elements comprising the arguments from idolatry and from indifference are scattered throughout Hume's corpus. We do not claim either argument is explicitly offered in any particular text. Nonetheless, the two arguments have a distinctly Humean flavour; previous authors have identified similar themes (Siebert (Reference Siebert1984); Holden (Reference Holden2010); De Cruz (Reference De Cruz2015) ). Some of Hume's discussion of religion (e.g. in his Natural History of Religion and ‘Of Superstition and Enthusiasm’) resembles early modern authors engaged in internecine religious conflicts (e.g. Pocock (Reference Pocock1997) ). But, unlike his orthodox contemporaries, Hume leaves little room for non-idolatrous religion. This is for good reason: for Hume, non-idolatrous religion was not possible. Two possible consequences can be drawn: either God does not exist (the argument from idolatry) or is indifferent to our religious activity (the argument from indifference). We begin with the argument from idolatry:
(1) If the God of classical theism exists, then God wants us to avoid idolatry.
(2) If God wants us to avoid idolatry, then our cognitive apparatus does not gratuitously incline us to idolatry.
(3) Our cognitive apparatus gratuitously inclines us to idolatry.
(4) So, God does not want us to avoid idolatry (modus tollens from 2 and 3).
(5) Therefore, the God of classical theism does not exist (modus tollens from 1 and 4).
Premise (3) is our most controversial and will again be utilized in the argument from indifference. Therefore, we spend most of this article defending (3). Likewise, as we will describe, Hume devoted his attention to (3). While we provide a more complete discussion of Hume's psychology of religion in subsequent sections, we briefly summarize the argument here.
Hume argued that actions originate from a combination of passions, sensations, and ideas (T 2.3.3; 3.1.1). Ideas and sensations are motivationally inert, but determine the conditions under which our passions arise. In turn, ideas aid our evaluations of how to deliver on our ends, as provided by our passions. So, to have actions motivated towards God, one must have an impression or an idea of God. God is ‘no object of the senses or the imagination, and very little of the understanding’. Moreover, God is infinitely remote and incomprehensible; one may ‘feel no Affection towards him’ (WC). We conceive of God as remote due to a combination of cognitive and cultural dispositions. Though psychologically disposed to religion, humans elevate their deities until their deities can no longer generate passions. Finally, humans reorient their religious activity around non-transcendent entities (NHR, especially section VIII). Thus, human psychology inclines people to idolatry because their psychology inclines them to worship real or imagined entities importantly unlike God. Having summarized the argument for (3) – which we will momentarily return to in section IV – something needs to be said with respect to (1) and (2). Hume does not devote much time to defending either, so we embark on a long digression.
While traditional theology has defended (1) in a variety of ways,Footnote 2 the most convincing argument for (1) appeals to the nature of love. We pump the relevant intuitions through a thought experiment. Suppose Pam tells us that Roy loves her. Subsequently, we learn:
(i) Roy avoids Pam.
(ii) Pam spends her time with Jim, whom she thinks is Roy.
(iii) Roy knows Pam thinks Jim is Roy.
(iv) Roy has expended effort in strengthening Pam's belief that Jim is Roy.
From (i)–(iv), we should infer that, ceteris paribus, Pam is mistaken about Roy's feelings. Likewise, if God loves us, ceteris paribus, God would prefer for us to engage God instead of surrogates in our religious activity and God would not cause us to be disposed to mistake surrogates for God.
One reason (i)–(iv) are incompatible with Roy loving Pam is that Roy has deceived Pam so that she is unable to respond appropriately to Jim or Roy. Likewise, idolaters are unable to respond appropriately to God. Through relationship with God, we could come to learn the Good and lead lives through God's teachings. To grow in relationship with God could involve moral maturation. Those who have a mistaken view of who God is, what God desires for us, and who are disabled from moral maturation can wreak incredible moral evil. (Note that the appeal to moral error does not necessarily suggest a version of the problem of evil. Here, our claim is that, ceteris paribus, God desires that we avoid moral error and that we strive towards moral maturity. This may be so even if moral error is consistent with God's goodness.)
Perhaps these reasons for accepting (1) – or the others we list in footnote 2 – are not convincing, and perhaps no other convincing reasons can be provided. We will consider the consequences of rejecting (1) when we turn to the argument from indifference. For now, suppose (1) is true. Let's turn to (2) (if God wants us to avoid idolatry, then our cognitive apparatus does not gratuitously incline us to idolatry). Here, we've modified Hume's view and added the notion of a gratuitous inclination. Whereas Hume believed that religious activity oriented around the classical theistic God is not possible, we maintain the weaker thesis that religious activity oriented around God is more difficult (in a sense to be specified) than expected on classical theism.
God may have countervailing desires overriding God's desire for us to avoid idolatry and may therefore allow some inclination to idolatry. For example, God may recognize our free will as a great good. In order to allow us to have free-will, God might have created us with fallible access to theological truths, and that fallible access may incline some to idolatry. Those who love do not coerce the beloved into relationship (or, if they do, this is a flaw in the way that they love; God lacks flaws, so would not coerce us). Likewise, we find it plausible that, ceteris paribus, God would not coerce us into mistaking other entities (real or imagined) for God. Consider again our thought experiment concerning Pam. Suppose that one evening, while Pam is sleeping, Roy performs brain surgery on Pam, so that Pam will be much more inclined to mistake Jim for Roy than she would otherwise be. This would be yet stronger evidence that Roy does not love Pam; Roy would have produced in Pam a gratuitous inclination to mistake Jim for Roy.Footnote 3
A gratuitous inclination to idolatry is an inclination fulfilling at least one of two conditions. An agent A is gratuitously inclined to idolatry if:
(v) due to A's cognitive (and possibly social or economic) situation C, A is more inclined to idolatry than is necessary for the attainment of some overriding divine purpose or
(vi) in the case that God exists and created A, God would antecedently know that if God created A such that C differed, the world would certainly or probably be a better fit for God's purposes than the world would otherwise be.
Here, we've adapted Kraay (Reference Kraay2016)’s discussion of gratuitous evil in order to develop our notion of a gratuitous inclination to idolatry.Footnote 4 Nonetheless, gratuitous inclinations to idolatry need not be gratuitous evils and our argument should not be understood as an instance of an evidential argument from evil. This is so for two reasons. First, God's purposes in allowing or disallowing the degree to which A is inclined to idolatry may not involve moral considerations and so may not involve considerations about goods or evils at all. The argument from idolatry might simply involve considerations about, for example, God's love.Footnote 5 Or consider van Inwagen (Reference van Inwagen, Howard-Snyder and Moser2002), where van Inwagen considers possible worlds where (a) there are horrendous evils but God's presence is known to all who are non-resistant and (b) there are no horrendous evils but there are non-resistant non-believers.Footnote 6 If van Inwagen is right, it's hard to see why we couldn't similarly imagine worlds where there are (a*) gratuitous evils but no gratuitous inclination to idolatry and (b*) no gratuitous evils but there is gratuitous inclination to idolatry. If such worlds are conceptually possible, then the gratuitous inclination to idolatry is not necessarily an instance of gratuitous evil, and the incompatibility between classical theism and gratuitous idolatry is not a concern about God's distinctively moral properties. Second, although idolatry may be bad for a variety of reasons (as surveyed in note 2), it is unclear whether an inclination to some bad activity is itself bad; it is also unclear whether the badness is gratuitous whenever the inclination is gratuitous.
Insofar as God is instrumentally rational, God's design and providence should reflect God's will. Therefore, while there may be some tendency to idolatry even in the case that God desires for us not to be idolatrous, the idolatry would not be gratuitous. In this article, like Hume, we've focused on how people are contingently cognitively arranged.Footnote 7 Given other cognitive arrangements, humans would have been less inclined to idolatry than they are, and, according to the argument from idolatry, this is reason to think the classical theistic God does not exist.
The argument from indifference
As discussed previously, God might not desire for us to avoid idolatry; perhaps (1) is false. Supposing (1) is false, (2) and (3) can be used to construct the argument from indifference:
(2) If God wants us to avoid idolatry, then our cognitive apparatus does not gratuitously incline us to idolatry.
(3) Our cognitive apparatus gratuitously inclines us to idolatry.
(4) Therefore, God does not want us to avoid idolatry.
Serious theological implications follow from (4). Suppose classical theism is true. The God of classical theism is perfectly good and perfectly loving. If a perfectly loving being would not want us to avoid idolatry, the concept of perfect love must not entail that one would not incline one's beloved to mistake others for oneself. We've apparently gone wrong in how we've conceived of loving relationships.
Moreover, as a perfectly good and loving being, ceteris paribus, God would desire for Her creatures to experience good lives. So, if having theologically correct beliefs offered some unique route to important goods, ceteris paribus, God would desire for us to have theologically correct beliefs. If God does not desire for us to avoid idolatry, then God does not desire for us to have theologically correct beliefs; by implication, having theologically correct beliefs does not offer some unique route to important goods. Contrary to doctrines found in Christianity, Islam, and Judaism, theologically correct beliefs would not be a unique route to, e.g., salvation, afterlife in God's presence, a meaningful, purposeful, or ethical life, and so on (Megill & Linford (Reference Megill and Linford2017)).
Remember Pam, Jim, and Roy. One reason Roy's deception was taken to be incompatible with Roy's love was that Pam would have been disabled from responding appropriately to Roy. This example seemed to suggest idolatry inhibits proper relationship with God, inhibiting moral growth. But if a gratuitous inclination to idolatry does not inhibit moral growth, as suggested if God is indifferent to theological correctness, then moral growth is available regardless of one's religious beliefs. The argument from indifference may provide new grounds for religious pluralism; if having theologically correct beliefs is not a unique route to important goods, then membership in any particular religion is not important for obtaining those goods.
Hume and our inclination to idolatry
Our arguments from idolatry and from indifference are intended as further developments of arguments originating with Hume's psychology of religion. For the remainder of our article, we turn to explicating Hume's defence of (3) – that our cognitive apparatus inclines us to idolatry – and updating Hume's defence in light of the contemporary social scientific study of religion. Hume discusses two ‘corruptions of true religion’, ‘superstition’ and ‘enthusiasm’, resulting from our cognitive inability to hold God as a proper object of the passions (SE 75). We are thereby inclined to idolatry. Hume left little room for other kinds of religious activity. While Hume's Natural History of Religion traces religion's ‘origin in human nature’ (NHR Intro.1), the focus is on an anthropology and history of religion. Hume's psychology of religion, especially its connection to Hume's theory of human action, needs to be constructed from what Hume says elsewhere. For Hume, only the passions (and not reason) motivate action (T 2.3.3; 3.1.1); therefore, to interpret Hume's psychology of religion, we need to examine those passions to which Hume attributed religious activity. The theme throughout is that there is no passion whose object is a transcendent God.
Hume's 1743 letter to William Mure of Caldwell (WC) is crucial for understanding the relationship between Hume's action theory and account of religion. For Hume, our emotions are more aroused by someone closer than by someone further away. For example, we feel more sympathy for a neighbour than for a distant stranger (T 2.3.7; 3.2.1). A remote and distant ancestor who left us ‘Estates & Honours, acquired with Virtue’ might be a ‘great Benefactor’, but we cannot possibly ‘bear him any Affection’ due to his remoteness. We have never encountered, through sense experience, anything like God. Neither the ‘Passions’ nor the ‘calm Affections’ can operate ‘without the Assistance of the Senses, & Imagination, or at least a more complete Knowledge of the Object than we have of the Deity’. God is an infinite, incomprehensible being, radically beyond the finite experience or cognitive abilities of human beings; God cannot be ‘the natural Object of any Passion or Affection’, because God is not an object of the ‘Senses or Imagination’, and enters ‘very little of the Understanding’. Therefore, a fortiori, our passions cannot be aroused by God (WC).Footnote 8 Because a remote, incomprehensible God cannot be the object of religious activity, Hume argues that religious activity can originate in two ways: either religious individuals invent subordinate beings who can be the objects of religious activity (as in superstition) or the object of religious activity is the self (as in enthusiasm). We discuss these in turn.
Superstition
For Hume, superstitious religion is a response to an internal theological tension. The Abrahamic religions teach that religious activity should be oriented around a radically transcendent God. Simultaneously, the greater God's transcendence, the less God can be an object for the passions. Superstitions are ‘false religions’ originating in weakness, fear, melancholy, and ignorance, and are immoral (SE 75).Footnote 9 Of these, only fear and melancholy (despair) are passions.Footnote 10 As Cohon interprets, the objects of direct passions – of which fear and melancholy are examples – is the pleasure or pain we may experience (Cohon (Reference Cohon and Radcliffe2011), 171).Footnote 11 Due to our weakness, the world can harm us in innumerable ways. Due to our ignorance, we postulate and fear powerful, unseen invisible agents responsible for various harms, real and imagined. To avoid experiencing harm, the superstitious develop methods to appease the unseen forces; these unseen forces can explain the natural phenomena that might do harm to us (SE 75; NHR III.1). To placate the unseen agents, the unseen agents are praised and through praise elevated until they vastly exceed the cognitive abilities of human beings. The many unseen agents become a singular being of the ‘utmost bounds of perfection’ (NHR VIII.2), whose properties – e.g. unity, simplicity – are categorically distinct from anything in the material realm.
A radically elevated conception of God is unstable and unfit for ‘vulgar comprehension’. In order to orient religious activity around the incomprehensible, new subordinate entities (angels, demigods, demons, saints, etc.) were introduced (NHR VIII.2); but the subordinate entities are idols. Alternatively, some religions slide into worshipping artwork. This explains doctrinal prohibitions on artwork.Footnote 12 And so our cognitive architecture leads practitioners to replace a remote, transcendent God with more immediately accessible surrogates.
The empirical evidence
We defend three claims crucial for Hume's account of superstition: (i) religious belief frequently originates with existential fears, (ii) humans have cognitive limitations in conceiving of a transcendent being, and (iii) due to our cognitive limitations, humans tend towards idolatry. We turn first to the empirical evidence for claims (ii) and (iii); afterwards, we turn to (i). We do not follow Hume in identifying specific religions as monolithically superstitious. Instead, we maintain that the phenomenon Hume identifies as superstition is a significant, widespread, and natural psychological tendency, which may be found across a variety of religions, and that gratuitously inclines the folk to idolatry.
For all we say here, careful academic theological study may aid people in reasoning about God and so reduce the degree to which they are inclined to idolatry. However, most theists have not undergone careful theological study and most do not have the time or resources to do so. We are specifically interested in the religious beliefs cognitively or contextually available for the folk. We can distinguish two contexts of reasoning: theoretical reasoning – as employed in academic theology – and the application of reason to concrete instances in ordinary life – which we will call ‘practical reasoning’. Scientists studying religion now generally accept, ‘as an empirically established fact’, that the folk use anthropomorphic God concepts, differing from ‘theologically correct’ conceptions, in practical reasoning (Westh (Reference Westh and Geertz2014) ).Footnote 13 Helen de Cruz has previously written on the relevance of Hume to the scientific study of religion. As she describes, Hume was right to think theological correctness is difficult to maintain. And yet folk religious beliefs are ‘easy to acquire’ (de Cruz (Reference De Cruz2015), 6). The result is what Pascal Boyer terms the ‘tragedy of the theologian’ (Boyer (Reference Boyer2001), 283–285).Footnote 14 Theologians might attempt to maintain official doctrines, but their work does not provide substance to which religiously oriented passions coincide. Social scientists report that the folk endorse a kind of ‘folk religion’, constructed in the gap between the pulpit and the pew, and that exists ‘apart from and along-side the strictly theological and liturgical forms of the official religion’ (Yoder (Reference Yoder1974), 14).
Among the most cognitively appealing beliefs, common in folk religions across cultures, are those which involve anthropomorphic supernatural agents, such as angels, protector spirits, or otherwise non-transcendent divine entities, who intervene in otherwise threatening situations. For example, in the United States, widespread belief in, and reports of visitation by, angels transcends institutional religious boundaries and is found among those who have rejected traditional religions in favour of New Age practices (Draper & Baker (Reference Draper and Baker2011) ). The cognitive appeal of such agents makes ‘such beliefs likely to be adopted, as they are “immediately relevant for human concerns” ’ (ibid., 627).
Similarly, the folk are inclined to an anthropomorphic conception of God, especially in the context of cognitive load, even when they abstractly endorse classical theism. According to cognitive scientists Barrett and Keil (Reference Barrett and Keil1996), the folk reason inconsistently about the divine between abstract and practical reasoning. Abstractly, the folk often say God possesses transcendent properties as described by their theological traditions (including those properties included in classical theism).Footnote 15 Barrett and Keil found unanimous agreement among twenty-two undergraduate and graduate students that ‘God can read minds, knows everything, can perform multiple mental activities at the same time, does not need to be near something to receive sensory information about it, and has non-natural spatial properties’ (ibid., 228). To examine how the students cognized God in a practical context, and under conditions of cognitive load, they were provided with a narrative recall task.Footnote 16 In the task, the subjects were read short vignettes concerning God's involvement in particular situations. The vignettes are too long to commit completely to memory. Subjects were later asked to recall the vignette. Differences between the original vignette and the recalled vignette reveal how subjects processed the vignettes and committed them to memory. For example, one vignette concerns God answering two prayers, one the prayer of a boy caught in a river (ibid., 229–230). The story does not indicate whether God answered the other prayer before the boy's prayer. Someone who conceives of God anthropomorphically might say God finished answering the first prayer before attending to the boy. Someone who consistently reasons about God in non-anthropomorphic terms – as without limitations – may say God helped the boy while simultaneously attending to the first prayer. When asked to explain the vignette, subjects typically indicated that God answered one prayer at a time.
In explaining other vignettes, subjects utilized an anthropomorphic conception of God at odds with their theological traditions (and at odds with classical theism). God is conceived as a special ‘Big Person’: ‘there was a strong tendency for subjects to think of God [as] exhibiting human limitations’ (ibid.). Barrett and Keil summarize:
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. (ibid., 229–230)
Later work reproduced the effect among Hindu and Muslim participants, from India, with diverse gender, educational, and economic categories (Barrett (Reference Barrett1998) ).
While Barrett and Keil have their critics, the critics agree that anthropomorphism is a common feature of religion. Moreover, theological incorrectness can take forms other than anthropomorphism. For example, (Westh (Reference Westh and Geertz2014) ) provides cross-cultural examples supporting Hume's psychology of religion, in that the folk either anthropomorphize God or conceptualize God in mundane, earthly terms. The Maasai, for instance, ‘represent their god EnkAi in radically different ways in different contexts’ (ibid., 409). Mythically, EnkAi is ‘fully anthropomorphic, even living among men in primeval times’. As Westh goes on to explain, ‘in traditional hymns of prayer, [EnkAi] is identified variously with the sky, the earth and with other elements of nature’ (ibid.; also see Olsson (Reference Olsson, Sand and Sørensen1999), 83). While the Maasai's traditional hymns differ from the mythical representation, in that the hymns do not anthropomorphize God, they still identify EnkAi with mundane experiences in common life (the sky, earth, and ‘other elements of nature’). Westh appeals to Olsson's ethnography to conclude that when the Maasai turn to serious theological reflection, neither mythical nor hymnal representations ‘are considered to be literally true’ (Westh (Reference Westh and Geertz2014), 410). Olsson explains that ‘the Maasai conclude from myth . . . that primeval man somehow had a direct experience of God and from this knew his nature, but what this experience really meant is not known any more’. Consequently, ‘no human being . . . can know what God used to be like’ (Olsson (Reference Olsson, Sand and Sørensen1999), 79).
Similarly, the Babylonian god Girra was ritualistically identified with fire, but understood to be anthropomorphic in ‘narrative texts’ (Westh (Reference Westh and Geertz2014), 410). While non-anthropomorphic, fire does not transcend common life, so the identification is consistent with Hume. ‘[I]n the context of the official [Babylonian/Assyrian] temple cult most gods lived in their temples, where they received daily food offerings, while in other contexts they were represented as immanent in natural phenomena, or as anthropomorphic beings living in heaven’ (ibid.). Similar conclusions can be reached for ancient Egyptian, Abrahamic religions, and the ancient Greeks (Olsson (Reference Olsson, Sand and Sørensen1999), 87–91).
Further instances of religion oriented around this-worldly items can be offered and compared to Hume. For Atran and Sperber, supernatural entities need to be related to mundane experience for belief retention and cultural transmission; religious beliefs are predictably counter-intuitive, but only in specific and minimal ways.Footnote 17 Minimally counter-intuitive beliefs about x depict x as violating only a few expectations given by the ontological category in which x appears (Boyer (Reference Boyer2001), 64–65). Importantly, ‘only to the degree that the [minimally counter-intuitive concepts] remain bridged to the everyday world can information be stored and evoked in plausible grades’ (Atran & Sperber (Reference Atran, Sperber and Landsmann1991), 52). Justice and Lambert (Reference Justice and Lambert1986) and Birky and Ball (Reference Birky and Ball1988) suggest that people's conceptions of God are partly determined by the personalities of their parents. Barrett and Van Orman (Reference Barrett and Van Orman1996) suggest that those who use images of God in worship anthropomorphize God to a greater degree than those who do not. Many more examples could be given (Guthrie (Reference Guthrie1980) ). Theologians have sometimes maintained a similar point, e.g. Marcus Borg's discussion of fathers as improper (patriarchal) models of God and then parents as appropriate models of God (Borg (Reference Borg1997), 64–71, 73).
Fear and the social scientific study of religion
We've established that humans have cognitive limitations in conceiving of a transcendent being (claim ii) and that due to those limitations, humans tend towards idolatry (claim iii). We turn to addressing claim (i): frequently, religious belief originates with existential fears. For Hume, superstition originates with existential fears in a dangerous and uncertain world. Because of those fears, humans elevate the initial objects of their religious activity beyond their cognitive capacities. At that point, humans turn to surrogates. Intuitively, we think love of God is an appropriate motivation for devotion to and worship of God. For Hume, love of God cannot motivate religious activity. The object of love of God is God, but Hume thought no passion could have God as the object. Fear provides an alternative to love of God as a motivation for devotional and worshipful activity. Recall the argument from our conception of love for the view that God would not gratuitously incline us to idolatry. If Roy loves Pam, then Roy desires that Pam is motivated to relationship with Roy out of love and not out of fear. A fortiori, if we learn Roy rewired Pam's brain so that Pam is more disposed than she would otherwise be to enter relationship with Roy out of fear as opposed to love, we have reason to think Roy does not love Pam. On our modified Humean account, while we allow for the possibility that some people love God, a gratuitous disposition to enter relationship with God out of fear is incompatible with classical theism, and this is doubly so if the disposition to relate inappropriately to God results in theologically incorrect beliefs.
The social scientific study of religion has long associated fear (or, more generally, anxiety) with religious belief. As Hume thought, people fear the harms that powerful, unseen forces might do to them; people fear most that they will be killed. To cope with their anxieties around death, many cultures have constructed elaborate systems of rituals and venerations to appease powerful, unseen forces, or to ensure their passage into an afterlife.Footnote 18 As we argue in this section, devotional and worshipful activity is often motivated by fear and oriented around accessible, anthropomorphic conceptions of the divine.
Morton-Williams's comments on Yoruba religion are helpful in illustrating some implications of Hume's psychology of religion. Yoruba cosmology divides the world into three realms: Earth, Sky, and ile aiye, the house of the World, which we inhabit. Earth and Sky are populated by unseen, powerful forces the Yoruba either work with or against. Morton-Williams writes: ‘Life in the third cosmic realm, ila aiye . . ., is good only when good relationships are maintained with the gods and spirits of the other two’ (Morton-Williams (Reference Morton-Williams1964), 246). Elsewhere, Morton-Williams describes the Yoruba as ‘constantly on guard against [death]’. The Yoruba ‘make sacrifices to the gods to gain their protection and avoid their wrath’. While the Yoruba endorse several hundred gods, individuals worship the few ‘concerned with him’ and ‘proper to [their] lineage’. Other deities are sometimes addressed in response to other anxieties: ‘The gods not only shield their worshippers, they also give them the blessings Yoruba most value: long life, increasing prosperity, and, above all, children’ (Morton-Williams (Reference Morton-Williams1960), 34–35). The Yoruba acknowledge a hierarchy of gods, overseen by the transcendent creator deity Olurun (literally ‘Sky-Owner’, translated by Morton-Williams as ‘God’). Yet Olurun has no cult on Earth, because – as Hume predicts – religious activity is primarily focused on interactions with the subordinate deities (ibid., 40).
Consistent with Hume's account, Yoruba religious activity is typically motivated by anxieties associated with the lesser deities. The lesser deities can be unjust or vengeful, but can be sated or placated through rituals or tattoos. If one finds favour with the minor deities, the deities can provide the ‘blessing of children, health, and, if their God-ordained fate permits it, wealth’ (Morton-Williams (Reference Morton-Williams1964), 246). The gods can be predictable and ‘the terrified apprehension of death . . . is eased by carefully watching for signs of their demands’ (Morton-Williams (Reference Morton-Williams1960), 35). Death is explained in a variety of ways, each involving powerful, unseen entities. Witchcraft explains the death of newborns (‘The Yoruba admit that they fear witchcraft more than anything else’). Sango, the god of thunder, causes lightning strikes. Sopona delivers small-pox epidemics (ibid.).
Contemporary cognitive science of religion similarly links anxiety or fear of death to religious belief. We offer some representative studies.Footnote 19 Harding et al. (Reference Harding, Flannelly, Weaver and Costa2005) used standard psychological instruments to measure death anxiety, death acceptance, and religiosity (together with relevant controls) among 130 individuals from an Episcopalian church in New York City. Death anxiety and death acceptance were negatively correlated; subjects who reported more anxiety about death also reported that they were less accepting of the inevitability of death. Belief in God, the afterlife, and gender were the only variables that explained the variance in the death anxiety scores; only belief in God and the afterlife predicted the death acceptance scores (ibid., 256). People who reported more certainty about God's existence reported less anxiety and more acceptance of death.Footnote 20
Harding et al. did not record longitudinal data and cannot be used to establish conclusively that religious belief lowers death anxiety and increases death acceptance, or that pre-existent death anxiety brings about heightened religious belief. Nonetheless, the correlation identified by Harding et al. provides some evidence that religious belief lowers anxiety, at least for those who already believe. Plausibly, anxious individuals pursue activities they have found to be useful in the past for relieving anxiety or which they have observed to be useful for others. Therefore, theists who are anxious about death might be motivated to pursue a deeper commitment to their faith. Some communities might make use of the effect. Whitehouse (Reference Whitehouse1996) provides ethnographic data suggesting Melanesian initiation cults put their initiates through terrifying rituals (‘Rites of Terror’) after which the initiates have a greater commitment to the group.
Norenzayan and Hansen (Reference Norenzayan and Hansen2006) performed four experiments providing evidence that fear of death motivates religious activity. According to Terror Management Theory (TMT), ‘cultural worldviews, along with the desire to live up to the standards of one's culture (i.e., self-esteem), serve as a primary psychological buffer against the uniquely human awareness of death. If so, heightening such terror (mortality salience) should increase the need to bolster these worldviews’ (ibid., 175). TMT predicts that subjects who have religious beliefs will report greater commitment to those beliefs if exposed to a mortality salience stimulus, and will simultaneously report a stronger rejection of foreign religious notions. However, according to other theorists (and, for our purposes, Hume in particular), subjects who are primed to think about death will report a strengthened commitment to the existence of supernatural agents in general and will thus be more open to culturally foreign supernatural agents. The four experiments can be used to distinguish empirically between TMT and Hume's account, even though Norenzayan and Hansen do not utilize their experiments for that purpose.
The first two experiments were consistent with both TMT and Hume, while the latter two support Hume over TMT. In the first experiment, subjects reported a statistically higher commitment both to their personal religion and to their belief in God, as compared to a control group, when primed to think about death. In the second experiment, subjects who were exposed to a mortality salience stimulus explained an unusual phenomenon (the efficacy of prayer) with general affirmation of God's existence and specific affirmation of God's role in the phenomenon (God answers prayers). Subjects were more likely to report a theistic explanation for the efficacy of prayer after having been primed to think about death than they were when primed to think about religion (ibid., 178–179).
Experiments three and four repeated experiment two, but replaced God with culturally foreign supernatural agents: in experiment three, Buddha and, in experiment four, supernatural agents that do not belong to any world religion (‘ancestral Shamanic spirits’; ibid., 180–183). The results were consistent with Hume but inconsistent with TMT: mortality salience stimulus increases the propensity to accept supernatural agents generally – even culturally foreign supernatural agents – for individuals with a pre-existing proclivity to religious belief. Fear of death motivated religious belief and not, as one might think, love of God. The result is a gratuitous inclination to idolatry, in part, comprised by theologically inappropriate reasons for relating to God and theologically incorrect beliefs.
Enthusiasm
Superstition is not the only form of idolatry that Hume discusses. Like superstition, enthusiasm involves passions for which the object is something other than the radically transcendent God. According to Pocock, ‘The Greek enthousiasmos carries the Latinate meanings of infusion and inspiration: the in-pouring or in-breathing of the divine, which comes to inhabit the person possessed’ (Pocock (Reference Pocock1997), 9). In Christian theology, one may experience the indwelling of the Holy Spirit; thereby, one might be (figuratively) infused with the divine. Most early modern authors reserved ‘enthusiasm’ for radical protestant sects for whom the (supposed) experience of the Holy Spirit was deemed politically disruptive (ibid., 10). On Hume's account, ‘enthusiasm’ came to be identified with a more general social or psychological phenomenon for explaining tendencies in Protestant churches.
For Hume, positive emotions – such as personal success, health, confidence, boldness, and elation (Yandell (Reference Yandell1990), 281–282) – can result in what Pocock describes as the ‘self-deification of the auto-intoxicated mind’ (Pocock (Reference Pocock1997), 22). As Hume describes, the mind can be ‘subject to’ an elevation and a presumption ‘arising from prosperous spirits, luxurious health, strong spirits,’ or a ‘bold and confident disposition’. Hume remarks the ‘imagination swells with great, but confused conceptions, to which no sublunary beauties or enjoyments can correspond’ (SE 75).Footnote 21 One turns away from the Earthly realm, partaking in ‘raptures’, ‘transports’, and ‘surprising flights of fancy’. Since these feelings go beyond what the enthusiast understands as mundane sensations, they are attributed to God (‘who [they think] is the object of devotion’; SE 75). In the process, one comes to see oneself as the special creation of God. Thereby, one becomes prideful.
Hume's account of enthusiasm begins with the observation of ‘external [social] events’, e.g. ‘prosperous success’, ‘luxuriant health’, but also ‘strong spirits’ and a ‘bold and confident disposition’ (Immerwahr (Reference Immerwahr1996), 327–328). Enthusiasm is associated with hope, pride, presumption, ‘warm imagination’, and ignorance. Of these, hope and pride are the only passions. Hope is a direct passion while pride is an indirect passion. The object of hope is the pleasure we may have from the thing hoped for (Cohon (Reference Cohon and Radcliffe2011), 167). The object of pride is the self (or perhaps an object related to the self, e.g. one's child or a friend). For Hume, the locution ‘self-love’ is confused; one cannot truly love the self, but one can be prideful. Pride and love follow from a coupling of ideas and sensations. Love follows from a pleasant sensation coupled with the idea of another person. Pride follows from a pleasant sensation coupled with the self (T 2.2.1.2; 2.2.1.3). Instead of worshiping a statue, a relic, or a lesser deity, enthusiasts are motivated by pride to self-worship.Footnote 22 Enthusiastic religion, then, is developed through a general ignorance of our emotions and the source of our successes, and through the influence of hope, pride, confidence, and anthropomorphism.
The empirical evidence
As with superstition, we do not follow Hume in his thought that specific religions are monolithically enthusiastic. Instead, we understand the phenomenon Hume identifies as enthusiasm as a widespread, significant, and general psychological tendency (or disposition) gratuitously inclining people to idolatry. The theological and popular religious literatures frequently claim that self-worship results from a common variety of religious error. Religious authors have argued that apostates chose to worship themselves in place of God (Plantinga (Reference Plantinga2000), 208; MacArthur (Reference MacArthur and MacArthur2005), 192) or that secular society is morally degraded because people have placed themselves before God. (For a particularly vivid example of the latter, see Scalia (Reference Scalia2013), ch. 2.) Religious scriptures are replete with prohibitions against self-worship and with instructions to abdicate oneself before God.
Could the common prohibition against prideful self-worship be a consequence of the ease with which religious individuals slip into prideful self-worship? Communities infrequently prohibit actions that do not occur among their members.Footnote 23 The common occurrence of prohibitions against prideful self-worship suggests that members of religious communities easily slip into self-worship in acts of religious veneration. When interviewed, sociologists found that some Christians attribute to atheists a self-interested egoism into which the Christians fear the Christians might slip if they lost their faith (Edgell et al. (Reference Edgell, Gerteis and Hartmann2006), 227–228). Psychologists have shown that self-esteem correlates with one's conception of God. Benson and Spilka (Reference Benson and Spilka1973) studied 128 Catholics; those with high self-esteem tended to posit a loving and accepting God. As Hume writes, one's ‘bold and confident disposition’ may play a causal role in explaining one's conception of God. Perhaps the object of the emotions inclining some to religious activity is the self. And if all that is right, many are gratuitously inclined to idolatry.
Conclusion
Hume discusses two forms of idolatry (superstition and enthusiasm) generating the arguments from idolatry and indifference. Both result from our cognitive inability for God to be the object of our passions. Only passions can motivate, so religious activity must originate elsewhere than God. The superstitious worship non-transcendent, anthropomorphic entities because a transcendent being far exceeds our cognitive abilities. The enthusiasts worship themselves in disguise. Some are moved to religious activity by fear, whose object is pain, and others worship idols or their passions are stirred by the self. Fear and pride are inappropriate motivations for religious activity and result in idolatry.
There are reasons to reject aspects of Hume's psychology, epistemology, action theory, his taxonomy of emotions (or passions), claims made about particular religions, and so on. But if superstition and enthusiasm are reinterpreted as two widespread, significant, general tendencies gratuitously inclining people to idolatry, core Humean claims are supported by empirical work in the scientific study of religion; these core claims allowed us to construct the arguments from idolatry and indifference. First, due to their psychologies, the folk hold crude God concepts falling far short of classical theism. Much religion is oriented around either lesser divine beings or crude anthropomorphisms. Second, Hume suggests humans are cognitively wired for idolatry. We endorse a weaker version of this claim: for all we know, some humans avoid idolatry, but humans have a gratuitous inclination to idolatry explained by their cognitive architecture. Third, the human inclination to idolatry is partly explained by our fears and prides, motivations inappropriate for devotional and worshipful behaviour. Our disposition to such behaviour is contrary to intuitive notions of love, either suggesting that no loving God exists or that our conception of love requires radical revision.
