Holding doctrinal belief as an artefact

Abstract

In this article, I consider how empirical research on religious belief poses a puzzle for the relationship between religious doctrines and lived religion, and develop one solution to that puzzle. The empirical evidence shows that believers are commonly incorrect about the content of doctrinal statements, and apparently not much interested in that content. I argue that this finding calls for a new understanding of the role of doctrinal commitments in religious life. I propose that in many cases believers relate to doctrinal statements in rather the way that they relate to sacred artefacts, that is, by way of an attitude of reverence and strong adherence, and independently of any attempt to discern the content of doctrinal claims. I note how this account avoids some of the difficulties of alternative solutions, which may undermine the claim of religious beliefs to count as genuine beliefs.

In this article I propose a novel solution to a puzzle posed by some empirical studies of religious belief, which suggest that such beliefs commonly diverge from officially sanctioned church teachings. These studies have led some cognitive scientists to suggest that religious beliefs should not be considered as beliefs in the strict sense at all. I shall develop an alternative solution to this puzzle, which has the merit of preserving the status of religious beliefs as genuine beliefs. I will propose that doctrinal statements are often treated by believers as sacred artefacts, on a par with material artefacts such as icons and relics. Let us begin by looking at three features of religious beliefs.

Discrepancies between doctrinal statements and actual beliefs

It is an established practice in philosophy of religion to characterize religion through what people believe, and to identify particular religious traditions primarily by differences in their credence and doctrines (Lopez (1998), 21). The believer's apprehension of the propositional content of relevant doctrinal statements is commonly taken to motivate their practice: Paul prays to God because Paul believes that God exists, is omnipotent and wholly good – therefore, is likely to respond to Paul's prayer. However, studies in the cognitive science of religion suggest that officially sanctioned accounts of the content of doctrinal statements do not match the real convictions of religious believers, a phenomenon which Justin Barrett has dubbed ‘theological incorrectness’. In an experiment conducted by Barrett, Hindu and Christian believers were first asked whether certain properties, such as omniscience and timelessness, were to be ascribed to God. The majority of the subjects assented to theologically correct beliefs about these matters. Next, the experimenters told the subjects a short story about God, and asked them to repeat the story. Here is the story:

A boy was swimming alone in a swift and rocky river. The boy got his left leg caught between two large, gray rocks and couldn't get out. Branches of trees kept bumping into him as they hurried past. He thought he was going to drown and so he began to struggle and pray. Though God was answering another prayer in another part of the world when the boy started praying, before long God responded by pushing one of the rocks so the boy could get his leg out. The boy struggled to the river bank and fell over exhausted. (Barrett (1998), 613)

While the believers in the study were inclined to assent to the theologically correct claims of divine omniscience and timelessness, most tended to retell this story as if God had to finish answering another prayer before responding to the boy's prayer. Thus they misreported what was said in the story, by supposing that God is subject to spatial and temporal constraints. Barrett (1998) concluded that while Hindu and Christian believers, when asked, assent to theologically correct doctrinal statements regarding, for example, God's omnipresence, in practice they tend to relate to God as if God were a human-like agent, bound by limitations of time and space. Recent research corroborated these results, showing that theologically correct doctrinal beliefs somehow coexist with intuitive ideas about persons (Barlev et al. (2017) ). Barlev's experiments shown that it takes a longer time for religious believers to make logical inferences from theologically correct but counterintuitive beliefs (such as ‘‘All beliefs God has are true’’), and under time pressure believers were disproportionately more likely to make errors regarding the latter (ibid., 425).

There are other experiments and data which point to the same phenomenon: in practice, humans often relate to God as if God were a human-like agent. According to research by Luhrmann (2012), Evangelical Christians attribute person-like mental states to God because, as they argue, it allows them to experience God more closely and intimately, even though, when pressured, they admit that this attribution might deviate from theological views of their traditions. In prayer believers look upwards and direct their prayers to heaven, even though they assent to the doctrine of divine omnipresence. (For a well-known example, see Titian's Penitent Magdalene.) Summarizing such cases, Pascal Boyer has concluded: ‘experimental tests show that people's actual religious concepts often diverge from what they believe they believe. This is why theologies, explicit dogmas, scholarly interpretations of religion cannot be taken as a reliable description of either the contents or the causes of people's belief’ (Boyer (2003), 119).

Even though I agree to some extent with Boyer's second claim, I cannot accept his first claim that people's actual religious beliefs diverge from the doctrines they assent to. This radical conclusion is not the only possible explanation of the experimental data, and in this article I am going to propose an alternative account. Barrett himself has offered another explanation of the same phenomenon. He argues that in tasks requiring quick and rich inferences, we rely on intuitive knowledge, where our intuitive knowledge includes the idea that agents are constrained by time and space limitations, and the idea that they have to complete one action before starting another one. But given more time for thought, and when asked to reflect on their theological knowledge, religious believers can draw correct logical conclusions. So outside time-pressured conditions, they use intuition-violating theological concepts and compose narratives more appropriate to doctrinal statements about God (Barrett (1999) ). Thus from Barrett's perspective, the propositional content of doctrines can be apprehended under the right conditions, notwithstanding theological incorrectness.

Nevertheless, the proposal that doctrinal beliefs are often not properly embedded in ‘operative knowledge’ – i.e. knowledge which is used for on-the-fly problem solving – and do not sufficiently guide the believer's action remains. Even though in special settings believers can draw proper logical entailments from doctrinal statements, still in their religious practice – including practices which are definitive of religious behaviour, such as prayer – humans tend to relate to God as if God was located up in Heaven. Even though they assent to doctrinal statements, their behaviour is often not guided by those doctrines, and in this respect religious beliefs appear to differ from everyday beliefs. No matter which way we interpret the results of Barrett's experiments, they show that some doctrinal beliefs (such as beliefs in divine timelessness, omnipotence, and omnipresence) do not play an action-guiding role, or at least their contribution to religious action is not mediated by the correct interpretation of their propositional content. My position will differ from both Boyer's and Barrett's: from the first as I shall claim that believers do in fact believe the doctrines to which they assent, and from the second as I shall argue that doctrines do guide practice, only not by means of logical inferences from their propositional content.

Doctrinal ignorance

There is a further respect in which the relationship between theological doctrine and religious practice proves to be puzzling: as just discussed, believers appear to assent to doctrines they do not fully understand, but in addition, it seems that they are commonly unaware of the doctrines themselves. The interviews conducted by the sociologist Meredith McGuire reveal that the adherents of various religions assent to verbal expressions of belief that are diverse and commonly divergent from institutionally sanctioned doctrinal formulae. She writes:

Our scholarly theories about religious socialization, conversion, and religiously plural societies have long depicted religion at the level of the individual in terms of commitment to the relatively coherent beliefs and practices of a single, received faith tradition (as identified through an organized religion such as Catholicism or Judaism). What if this picture of the historical norm is completely mistaken? (McGuire (2008), 186–187)

In lived religion believers seem to be surprisingly uninterested in the doctrinal statements of their religion. Instead of sticking to clear and systematic doctrines, religious adherents frequently hold to vague, piecemeal, and even theologically wrong expressions of belief. For instance, according to a survey by Life World Research from 2016, 56 per cent of Americans believe that the Holy Spirit is a force, and not a person (Smietana (2016) ). Pew Research Center (2010) revealed that 45 per cent of American Catholics do not know that according to Catholic doctrine the bread and wine used in Communion do not merely symbolize but actually become the body and blood of Christ. In sum, ordinary believers not only misunderstand the theological entailments of officially sanctioned teaching, and show a degree of carelessness about these entailments, but they also do not accord doctrinal formulae the central role in religion that they are commonly assigned by scholars, as McGuire notes. As the surveys show, it is often the case that even the most fervent believers are rather ignorant about the doctrines of their religion and do not consider their lack of doctrinal knowledge to undermine their religious commitment in any way.

Nevertheless, it would be incorrect to claim that religious believers are indifferent to doctrines: assenting to the creed is, of course, still very important for many traditions (Seul (1999) ). This importance is emphasized by religious authorities, as Hugh Nicholson (2016, 1–3) shows. Believers studied by Barrett (1999) and Barlev et al. (2017) assent to doctrines, and do not merely claim that they are aware of the content of doctrines. Apparently these believers treat doctrines as a part of their tradition that is to be accepted and assented to. Finally, as anyone who has ever witnessed inter-religious dialogue knows, attacks on a doctrine can easily elicit a highly defensive stance in believers, even if they barely know the doctrine.

Exempt from doxastic norms

Our everyday beliefs are supposed to track observation and to conform to basic logical norms, whereas the formation, refinement, and abandonment of religious beliefs seem to be exempt from standard doxastic norms. By contrast with many other kinds of belief, doctrinal beliefs can be stubborn and sometimes impervious to argument, logic, and evidence. But one might ask, what can be counted as evidence and counter-evidence when it comes to religious doctrines? To develop this point I am going to use Van Leeuwen's definition of evidential vulnerability of belief.

According to Van Leeuwen (2017), a cognitive state is vulnerable to evidence and can therefore be called belief:

1. I. If it is involuntarily prone to being extinguished if (a) it conflicts with perceptual states or if (b) it is realized to lead to a contradiction.

2. II. If it is involuntarily prone to being extinguished if it contradicts or does not cohere with other evidentially vulnerable states.

To put it simply, to be called evidentially vulnerable, doctrinal beliefs should be subject to change when they (a) lead to contradiction or (b) conflict with perception and other evidentially vulnerable beliefs. In such cases doxastic norms demand resolution of the contradiction or the abandoning of a conflicting belief. A well-known alleged contradiction is posed by the problem of evil: the omnipotence and omnibenevolence of God, postulated by theistic doctrine, are said to be inconsistent with the existence of all kinds of evil in the world. However, as there are multiple theological arguments that strive to resolve this contradiction or to show that it is only illusory, the problem of evil cannot be said to demonstrate conclusively the evidential invulnerability of doctrinal beliefs. If we now consider cases of type (b), here evidential vulnerability demands the involuntary abandonment of a belief conflicting with perception. For example, if I open a fridge and do not find milk there, my belief that there is milk in the fridge disappears. At first glance, it is quite hard to imagine a similar case with religious doctrines: what kind of perceptual state could possibly disprove, for example, that God is a Trinity? Most religious doctrines describe states of affairs beyond human perceptual capacities, such as the Trinitarian nature of God.

Nevertheless, we can still find examples of conflict between doctrinal belief and perception in Doomsday cults and other kinds of religious cults based on more specific doctrines than a doctrine of the Trinity. However, even here, belief seems resistant to counter-evidence. For instance, according to a recent study conducted by the sociologist Sauvayre (2017), factual evidence against cult doctrines, such as a failure of the core prophecy in a Doomsday cult, minimally (if at all) affects the beliefs of the cult adherents. Doomsday cults often continue even after their prophesied dates for the end of the world have passed (Festinger et al. (1956) ).¹ Even when the cultists studied by Festinger witnessed three consecutive factual disproofs of their doctrines over a four-day period, it did not give rise to an abandonment of doctrine. We might suppose that Christians of the first century, who expected Doomsday to happen and Jesus to return within a generation of his disciples, underwent similar processes, as failure of this prophecy is not known to bring about any significant crisis of belief in Christianity. Furthermore, insensitivity to factual evidence is not limited to Doomsday cults. Sauvayre gives other examples, such as that of a cult based on a doctrine that the laying on of hands miraculously heals cancer and other incurable diseases. A cult member Laurianne becomes seriously ill, and despite all her fellow believers’ efforts, prayers and laying on of hands, her condition only gets worse. Nevertheless, when in the end her life is saved thanks to medical assistance, rejected by the cult, it is still an insufficient reason for her to abandon her beliefs and disengage from the cult. In fact, it does not even raise in her any significant doubts regarding the truthfulness of the doctrine of the miraculous healing power of the laying on of hands (Sauvayre (2017) ).

Extensive interviews conducted by Sauvayre (2011) show that in 71 per cent of cases it was not factual evidence against the doctrinal teachings of the cult, but a conflict over values, such as the immoral conduct of a cult leader, which triggered departure from the cult. And as other empirical investigations show, this pattern is common for the adherents of traditional religions as well. From Uecker et al.’s 2007 study of the abandonment of religious belief in early adulthood in the US, it seems that, contrary to expectations, college education – that is, becoming more familiar with a scientific understand of the world, and acquiring more sophisticated reasoning skills – is not a contributory factor in diminishing religiosity. Indeed, the biggest religious decline was seen in people without any college education. What in fact accelerated diminished religiosity and loss of faith was engaging in the actions prohibited or denounced by religion as immoral, such as extra-marital sex, drug-taking, and alcohol use. What in most cases makes people believe or disbelieve and eventually abandon their religious adherence is, it appears, not argument or factual evidence against religious doctrines. Van Leeuwen (2014) thus argues for the evidential invulnerability of religious beliefs: they are not liable to be abandoned even if they conflict with perception or lead to a contradiction. So it is not, it seems, the details of the propositional content of religious doctrines or their failure to fit with the evidence that motivate people to give up membership of a religious tradition, and getting that content right does not appear to be vital for participation in such traditions.

A puzzle

From the three kinds of consideration we have just rehearsed, we can formulate the puzzle to be addressed:

The puzzle of religious beliefs: (1) religious believers do not understand the entailments of their doctrinal commitments; (2) they are apparently unaware of many of the doctrines that are deemed important by religious authorities; and (3) they hold to doctrinal claims independently of counter-evidence and apparent contradictions. Given this empirical data, how can we connect doctrinal claims to lived religion?

If the facts presented above are true, then an understanding of doctrine cannot provide the basis for participation in religious tradition. Lived religion shows that even believers who firmly assent to doctrinal propositions, such as those who were studied by Barrett, do not hold them as fully comprehended propositions. The fact that people make elementary mistakes in their reasoning from these doctrines shows that the propositional content of the doctrines is not crucial for participation in religious tradition. So what is the role of doctrine in lived religion?

Some possible solutions to our puzzle are:

1. 1. Doctrinal beliefs lack factual content

   1. a. If religious beliefs have no factual content, that would explain why religious actions are not guided by the content of doctrines.

2. 2. Doctrinal beliefs do have factual content. Here, we have three further possibilities:

   1. a. Religious beliefs, including doctrinal ones, have factual content, but are irrational, and therefore no coherence of belief and action can be expected.

   2. b. Religious beliefs are factual and action-guiding by virtue of their proposal content. But the empirical data we have been reviewing makes this an unpersuasive option.

   3. c. Doctrinal beliefs have factual content, and guide action without the mediation of propositional content. This is the position I am proposing to defend.

Let us think about some of these responses a little more closely. The unresponsiveness of religious beliefs to evidence has led some commentators to conclude that they are irrational – this is a conclusion drawn by the New Atheists. If believers do not grasp the entailments of their doctrinal commitments, that might seem like a further reason to conclude that they are irrational. However, claiming that religion is just irrational means the end of inquiry and hardly fits the self-understanding of religious believers. There are, however, other ways of solving the puzzle. A possible explanation of the failure of believers to respond to evidence would be to claim that religious beliefs lack factual content. In that case, there is indeed no need for such beliefs to be adapted to the evidence. And if they lack factual content, then it would be unsurprising if they fail to guide action. This is the response of fictionalism. Cognitive scientist Neil Van Leeuwen (2014) goes even further and argues that religious credences have aetiologies and behavioural and cognitive effects quite different from those of factual beliefs and are, therefore, not beliefs at all. Thus, Leeuwen concludes, what we call religious belief is an attitude more similar to such cognitive states as fictional imagining, and assumption for the sake of argument. I agree with Leeuwen in his claim that religious beliefs have some properties distinct from those of factual beliefs. However, I think we should not equate religious beliefs in general and doctrinal beliefs in particular with imaginings and other similar states, as religious beliefs, unlike imaginings, are remarkably stable and possess high value for believers, which no ‘assumption for the sake of argument’ can have.

Another fictionalist solution to the puzzle maintains that the strong adherence of common believers to doctrines is to be explained in terms of social identity: assent to doctrine serves as a form of self-identification for group members, whereby they can distinguish themselves from other groups, as Jeffrey Seul (1999) has argued. Similarly, Van Leeuwen (2017) calls beliefs in religious doctrine ‘identity-constituting attitudes’, which are activated in religious or identity-testing contexts. And from a philosophical vantage point, Hugh Nicholson (2016) has argued that it was the need for a strong social identity that motivated the development of sophisticated and counterintuitive religious doctrines, such as the doctrine of no-self in Buddhism and of the Trinity in Christianity. Social identity interpretations tend to treat religious doctrines as non-cognitive: they are to be affirmed, but their propositional content has no action-guiding significance.

While I agree that assent to religious doctrines can serve as a form of group-identification, I do not think that their identity-constituting function exhausts the role of doctrinal beliefs in lived religion. In the next phase of this discussion, I will argue that apparent discrepancies between religious and factual beliefs can be explained without rendering religious beliefs irrational or ‘identity-constituting attitudes’ that cannot be called beliefs at all. I shall also endorse the idea that doctrinal beliefs play an action-guiding role for their adherents, albeit one that is not mediated by the apprehension of their intellectual content.

Towards a solution of the puzzle

Before presenting my own solution to our puzzle, I will first consider in some detail another important aspect of religion – now not doctrines, but sacred artefacts and their contribution to lived religion.

Émile Durkheim proposes the following definition of sacred artefacts: they are ‘things set apart and surrounded by prohibitions’, while religion is ‘a unified system of beliefs and practices relative to sacred things’ (Durkheim (2001), 46). In a religious context, it is religious authorities who set sacred things apart and ensure that the prohibitions surrounding the artefact are respected. These prohibitions can vary, but in general they proscribe irreverent treatment of the artefact. So we can note two primary properties of sacred artefacts. They are: (1) set apart and (2) protected by prohibitions. Next, we can distinguish the following key functions of such artefacts:

1. 1. Holding. The sacred artefact is held as set apart from its profane counterparts and irreverent treatment of the object is forbidden.

2. 2. Acceptance. The origins of the artefact and the authority of the person transmitting it have a key role in securing its sacred role.

3. 3. Formation. Artefacts are created, refined or shaped by religious authorities only.

4. 4. Role in practice. Artefacts are to be treated with reverence and elicit activities such as prayer, liturgical behaviour, and meditation.

Let us consider these functions in further detail by taking an example of a particular religious artefact: the icon. In the Orthodox tradition the icon is an ‘immediate presence of the invisible Holy in historical time’ (Stylianos (1989), 5). Thus the icon presents religious states of affairs, since it shares in them, and demands the same reverent treatment as the objects it presents. Icons are believed to be ‘an expression of a theological reality’, as contemporary Orthodox theologian Archbishop Stylianos writes:

For icons, in general, presuppose the prototype which is reflected in different ways and in varying degrees throughout the entire course of divine economy, so that the divine plan of salvation is served accordingly from the very first steps of revelation to the final stage of parousia. The very concepts of ‘revelation’ and ‘parousia’ presuppose and clearly indicate the visual joy of the icon as the most characteristic way of understanding salvation. (ibid.)

Meanwhile, there is no doubt that the icon is an artefact: icons are created by human beings as vehicles of revelatory religious experience.²

The holding of icons in contrast to other objects in the home shows their property of being ‘set apart’. In the Orthodox tradition icons are widely used not only in churches but also at home, where they are placed in a special corner or on a wall. To a disinterested bystander, an icon is just another tempera-on-wood painting. It is made of the same kind of matter and occupies a space similar to that of other paintings. But for a religious person, the icon is distinct from other paintings and cannot be treated in the same way. If the homeowner feels like it, the interior design of the house can be restyled, for instance by changing the colour scheme of the walls, which in turn requires a matching change of paintings, curtains, furniture, and so on. But the icon is excluded from all these changes. Its role in the domestic interior is completely different from that played by other parts of the interior. Even if the modern style of an interior renders an icon glaringly obsolete, this will not force the religious person to remove the icon or exchange it for a modern one, fitting the interior better. If this were to happen, then this icon would not be treated as a religious object, but merely as an object of art or interior design. Nevertheless, this does not mean that once hung, the icon will stay in its place forever. If the owner were to convert to Buddhism, it would not be surprising if the icon were to be removed from the wall. Thus the holding of a religious artefact as sacred is not arbitrary or irrational, but governed by rules and norms, different from those governing the treatment of mundane things.

Another particularity of religious artefacts concerns their acceptance as sacred. The norms pertaining to acceptance of religious artefacts are: (1) the proper origin of the artefact matters most for its acceptance; (2) artefacts are to be accepted only with the sanction of religious authorities (or the sanction of those who have the endorsement of such authorities). Icons and other artefacts are produced by religious authorities and transmitted to believers in a ready-made form. They are accepted and valued as a part of divine reality, as a presentation of that reality. The role of religious authorities is to guarantee the relation of the artefact to religious reality.

The formation or development of religious artefacts is also distinctive. We can differentiate between two kinds of sacred artefact:

1. (a) created artefacts, such as sculptures, temples, icons, and so on.

2. (b) natural artefacts, such as sacred groves, or artefacts of supposedly miraculous, non-human origin, such as the Black Stone, which is believed to be brought from Heaven. (Newby (2005), 5050)

The production and elaboration of created artefacts involves the artistic, imaginative response to religious states of affairs and is surrounded by strict prohibitions. For example, only those who have received the requisite special education are permitted to make icons, and the icon is to be signed with the name of the divine person it depicts and sanctified in a church. Neither is one allowed to change the image on an icon, by adding or removing anything. Believers are required to accept the icon in its entirety, in the form produced by the relevant religious authorities.

When it comes to religious artefacts of natural or supposedly supernatural origin, they are framed, delimited, and properly separated from profane objects by religious authorities. Sometimes artefacts of type (b) undergo significant elaboration and expansion. This development is partly guided by contingent historical circumstances and partly by the logic and demands of worship. We can consider as an example the Black Stone, one of the most revered Islamic relics. Since the original single fragmented into several pieces, its fragments were placed in a silver frame, and subsequently fastened by silver nails to the Kaaba, a square-shaped building, which later was surrounded by the Masjid al-Haram mosque (ibid., 5049–5050). The elaboration and expansion of the initial artefact over centuries was extensive: by itself, the original Black Stone is estimated to have dimensions of about 25 × 20 × 20 centimetres (Thomsen (1980), 87), while the size of the Kaaba is about 13 × 12 × 10 metres (Newby (2005), 5049), and the surrounding mosque is now being expanded to cover about 1.47 million square metres (Fahy (2015) ). The silver frame, the granite building, and the mosque were not parts of the initial artefact. However, through their association with the stone, they have themselves acquired a sacred significance, and play an important part in worship.

What matters most for the role in practice of sacred artefacts is the adoption of a proper attitude towards them. Artefacts are to be treated with reverence, which is expressed in looking, touching, in positioning one's body in relation to the artefact in the requisite way, and in the tone of voice with which one speaks when in the presence of the artefact, and so on. In the Christian Orthodox tradition, for example, upon entering a church, one is required to make the sign of the cross twice, then kiss an icon, make the sign of the cross again, make a bow, and then move away. It is deemed inappropriate and irreverent to stand gaping at the icons during the liturgy, as if one is in a museum. Thus the proper treatment of icons and other artefacts is expressed primarily in the person's bodily demeanour towards them.

Artefacts have a special cognitive function as well. For instance, in the Orthodox tradition, by focusing upon an icon in prayer, the believer can draw closer to God, who is presented by the icon, and understand more about God.³ The icon is not, then, an arbitrary token representing some states of affairs, as propositional statements are often deemed to be. Instead, the icon presents religious states of affairs, by being a part of them. Following Tillich's (1959) distinction between sign and symbol we can state that the icon does not merely point at something beyond itself, but also is a part of the reality it is depicting. As a part of religious reality, and not merely the realm of human society, the icon should not be acted upon or understood as simply an object of art or a coloured piece of wood. Like other religious artefacts, the icon affords or elicits certain actions in religious contexts, thereby enabling the believer to unfold and encounter the reality of which it is part.

We have been considering the icon as an example of the general category of sacred artefact. Next, I want to explore the parallels between icons and another item that can be held as a sacred artefact – namely, the religious book. Here it is worth recalling that icons since patristic times are often called ‘Scripture for the illiterate’ or ‘books for the illiterate’ (Payton (1996), 175), as they express the same religious content as books, but in a different form. Like an icon, a book both has a particular content and can be treated as a religious artefact. And the reason for treating the book with respect is that it contains a certain propositional content, the sacred status of which is affirmed by religious authorities. To treat the book this way one does not need to be able to rehearse its content. The Bible or Quran can be revered by a person who has only a vague idea of its propositional content, but knows that this text is sacred to his or her religious tradition.

Finally, we can in turn parallel a book with a doctrine. Like a book, a doctrine has a propositional content, which is given in the words used to articulate the doctrine. And when that content takes on the quality of being sacred, then the doctrine as the container or vehicle for the articulation of that content should be treated in the same sort of way. Furthermore, religious doctrine is believed both by theologians and by common believers to contain much more than the human mind can possibly comprehend, since these doctrines propound divine mysteries, revealed to human beings by God. Thus when one accepts a doctrine, what one actually accepts is not only the comprehended part of the doctrine, but also those divine mysteries which are believed to be encapsulated in the doctrine. In this sense, religious doctrines serve as a container with a highly valuable and sacred propositional content. And just as in the case of sacred books, the container itself takes on something of the value and sacred quality which pertains to its contents. In these ways, a religious doctrine is to be treated differently from everyday kinds of affirmation, which do not have any sacred content, and which are fully comprehensible.

To take an example, in Islam, the Quran is believed to contain the very words of God, and accordingly only the Arabic text is deemed to contain the divine revelation in full. Great value is attached to reciting the Quran in Arabic: teachers of the Quran often emphasize correct pronunciation of the text, and learning the words by heart, even if children or others in the class do not speak Arabic, so do not understand the meaning of the Arabic text (Berglund (2010), 70). Those who can recite the entire Quran by heart are given an honourable title: hafiz, or hafiza in the female form. However, to be hafiz one does not need to understand what one recites: it is enough to be able to recite the Quran in Arabic; knowledge of Arabic – and therefore understanding the propositional content of what one recites – is not required. We can suppose that similar considerations, and an attachment to the sacred status of the verbal container of a sacred meaning, underlie the resistance in many Orthodox churches to translations of the liturgy into more readily comprehensible, everyday language. The insistence on using ecclesiastical Latin in certain contexts in the Roman Catholic Church can be interpreted similarly. And in general, it is common for religious traditions to use a special ‘holy’ language for professing creeds and engaging in rituals and prayers. Typically, this will be the language in which the sacred texts of the tradition were written: Pali in Theravada Buddhism, Sanskrit in Mahayana Buddhism, Church Slavonic in the Eastern Orthodox churches, and so on.⁴ In these cases, the sacred quality of propositional content confers a sacred significance upon the verbal container of that content. Sometimes this affects simply the verbal formulation of a doctrine, where that formulation is open to translation into another language, but in other cases it can involve the very words in which the doctrine is expressed.

Accordingly, I hypothesize that these otherwise puzzling phenomena can be explained by supposing that doctrines are treated by believers not only as expressions of ideas, but as the verbal containers of a sacred propositional content, with the result that the doctrine, and commonly the very words in which it is cast, share in some measure in the sacred significance of that content. And this is to say that, in these respects, doctrines can take on the status of sacred artefacts. I shall argue next that this account of the nature of doctrines can help to solve what I earlier called ‘the puzzle of religious belief’. This solution, I shall argue, has some significant advantages: it recognizes the cognitive significance of doctrine, it is more faithful to the self-understanding of believers than are the solutions proposed by fictionalists and social identity theorists, and at the same time it allows us to see how doctrines can play an action-guiding role in religious life, albeit one that is not mediated by the apprehension of their intellectual content.

Religious doctrines as sacred artefacts

Let us briefly recall the puzzle with which we began: (1) religious believers do not grasp the entailments of their doctrinal commitments; (2) they are apparently unaware of many of the doctrines that are deemed important by religious authorities; and (3) they hold to doctrinal claims independently of counter-evidence and apparent contradictions. We can see immediately that by thinking of religious doctrines as sacred artefacts, we can understand why doctrinal beliefs differ from beliefs about tables and mice. As religious artefacts, doctrines are treated as sacred objects that are set apart from ordinary or profane objects, and from the processes of everyday belief formation. They are treated with the same respect and reverence as other sacred artefacts. This account provides a ready understanding of the attachment to (3), and believers’ unswerving attachment to doctrines. Moreover, in this way, we can also understand why many religious believers are in some degree uninterested in religious doctrines: religion proposes a wide variety of artefacts, including rosary beads, icons, and so on. And the believer can choose which artefacts are most compelling for his or her own religious practice. Doctrines are just one of a multitude of possible sacred objects, and for some people, they may not be the most accessible and compelling kind of artefact. In these terms, we can explain (2) from our original puzzle. Finally, we can also understand in these terms the action-guiding role of doctrinal beliefs: doctrines share the capacity of other sacred artefacts to elicit and guide action, but this action takes the form of a practical reverencing the artefact, rather than requiring sophisticated engagement with its intellectual import. In this way, we can also throw light on (1) from our original puzzle.

Now let us set out this general approach in a little more detail. Following the scheme we introduced earlier, we can note these similarities between artefacts and doctrines as they feature in lived religion:

1. 1. Holding. Religious doctrines are held as set apart from profane counterparts and irreverent treatment of such doctrines is forbidden.

2. 2. Acceptance. The origins of a religious doctrine and the authority of the person transmitting it have a key role in securing its sacred role.

3. 3. Formation. Religious doctrines are created, refined, or shaped by religious authorities only.⁵

4. 4. Role in practice. Religious doctrines are to be treated with reverence and play a role in eliciting activities such as prayer, liturgical behaviour, and meditation.

Given these parallels, we can understand how religious actions involving doctrinal statements, such as prayer, meditation, and the profession of creeds, need not require any engagement with the propositional content of the doctrines. For example, one can reverentially repeat the words of a teaching such as ‘God is Love’ or ‘There is no self’, and in this way use the doctrine as an aid to meditation or prayer, without attempting to apprehend its propositional content in intellectual terms. An example of how particular doctrine can be acted upon in religious practice might be helpful to illustrate my point here. A friend of mine, a highly intelligent and well-educated Orthodox Christian, wrote to me in 2002 how he struggled to accept a doctrine of Mary the mother of God. He could neither understand nor intellectually accept the idea of virgin birth and perpetual virginity, finding it a challenge to rationality. Thus he followed the advice of his priest and started to pray to the Virgin Mary, repeating the verse from Bible: ‘Hail, thou that art highly favoured, the Lord is with thee: blessed art thou among women’ (Lk 1:28), as a poetic expression of the doctrine. He prayed this way until eventually the doctrine of Mary settled down in his heart. Thereby he found himself in peace with this doctrine, accepted it, and never tried to challenge it again.

So doctrines, just like icons, are interwoven into religious practice, where they obtain their meaning in particular religious actions as prayer, meditation, and contemplation. Doctrinal statements or sacred texts can even be written down and worn on the body in small boxes (tefillin) or placed on a doorway at home in small containers, as mezuzahs in Jewish tradition. We can see remarkable parallels in the practical and cognitive relation to doctrines that is achieved through the embodied reverencing of icons in Christianity and the reverencing of mezuzah and, particularly, tefillin in Judaism. The inscriptions in tefillin and mezuzah are made by a specially trained and religiously observant scribe (sofer). But these inscriptions are not for reading or reciting: the little black boxes of tefillin are sealed, and the mezuzah container is supposed to be opened only once in seven years to check that the inscriptions have not faded. Instead the mezuzah should be touched whenever passing through a doorway, as a simpler way of saying prayer, while tefillin are to be worn on the hand and on the head. According to a tradition cited in Mekhilta Derabbi Yishma'el: ‘someone who wears tefillin is like someone who reads Torah’ (Cohn (2008), 150). This equation of wearing a sacred artefact and reading Torah is particularly remarkable.

Thus the main way of relating to the sacred content of tefillin in practice is to wear them on the body, thereby securing an embodied relation to the doctrine. Tradition puts a particular emphasis on the correct wearing of tefillin, even calling doing so in the wrong way a heresy, not merely a wrong practice: ‘One who makes his tefillah circular; that is a danger and there is no mitsvah to it. If he placed it on his forehead or on his palm; that is the way of heresy (derekh haminut)’ (ibid., 130). There is also a clear distinction between tefillin and other written excerpts of sacred texts: according to the tractate Tosefta Shabbat, tefillin are supposed to be saved from a fire on Shabbat as sacred writings, while other Torah excerpts are to be left to burn (ibid., 150). Therefore, what distinguishes tefillin from mundane objects is not the doctrinal inscriptions per se contained in them, but their being turned into sacred artefacts, produced by religious authorities in accordance with the strict rules of the tradition, and treated reverentially. Thereby their relation to the whole domain of the divine, and not merely to the propositional content of inscriptions, is secured. Thus we can conclude that not only icons, but other artefacts, such as tefillin, can afford this particular embodied way of reaching the divine mysteries without having first to apprehend the propositional content of the doctrines. The way these artefacts possess and convey religious meaning is not as simple as representing the states of affairs contained in a proposition. Religious artefacts, including doctrines, function as a sort of scaffolding for religious actions and religious thinking, where their meaning is reached via bodily practice.

We can now view the empirical data which I noted in the first part of the article from this new angle. Although doctrinal propositions are not material objects, religious believers can treat the words in which they are expressed in the same Durkheimian way as they treat other sacred artefacts, that is, as set apart from everyday affirmations of belief, and as meriting an attitude of reverence and religious respect. Again, this attitude reflects the status of the words as containers for, or vehicles for the expression of, the relevant propositional content. This understanding of the significance in religious life of religious doctrines helps to explain why believers are not always concerned with the logical entailments of those doctrines, and how they may even come to adopt apparently contradictory doctrines.⁶ Similarly, like artefacts, doctrines are surrounded by prohibitions, including prohibitions to make changes to them, and in these terms we can understand their invulnerability to counter-evidence and seeming contradictions. In these ways, the doctrine-as-artefact account provides new insight into the different cognitive roles of doctrines and everyday beliefs.

It might be objected that, in one important respect, doctrines differ from standard religious artefacts, since they are supposed to be rooted in fundamental truths, and so cannot be simply created. For instance, the doctrine of the Trinity is taken to record fundamental truths concerning God's eternal nature. However, on this point, we can compare doctrines with the second kind of artefact I distinguished in the earlier discussion – that is, artefacts of natural or supposedly supernatural origin. Moreover, as we have seen, an artefact which is taken to have a supernatural origin can then be elaborated upon, and those elaborations can come to share in the sacred significance of the original, divinely instituted artefact. Again, we could take the case of the Kaaba as an illustration of this sort of process. And we might understand in similar terms later refinements in the verbal formulation of the doctrine of the Trinity.

Let us now return to the two alternative solutions to our puzzle that I mentioned at the beginning of this discussion, namely, irrationalism and fictionalism. The solution we have been developing offers a response to the charge that doctrinal beliefs are simply irrational, by showing how phenomena such as theological incorrectness, and believers’ lack of interest in the propositional content of doctrines, reflect the artefactual status of doctrines. If we allow that religious doctrines are held as sacred artefacts, then the particularities of their formation, acceptance, and abandonment cease to seem irrational: in these respects, the believer's relationship to doctrines is no more vulnerable to a charge of irrationality than is their relationship to a sacred artefact, such as an icon. The same kinds of norm apply in the two cases. The formation and adjustment of everyday beliefs are governed by a different set of norms, for the reasons we have discussed. For instance, by contrast with everyday beliefs, doctrinal beliefs can only be produced by authorized religious specialists: like icons, doctrines are developed by religious authorities and supplied to the believer in a ready-made form. Accordingly, the believer has a reason to give up a doctrinal belief if those authorities are discredited, as may happen if they engage in immoral behaviour, as in the cases studied by Sauvayre (2011). In these and other ways, doctrinal beliefs, as they feature in lived religion, often follow the norms pertaining to the formation and acceptance of artefacts.

It might be wondered whether the account we have been developing downgrades the truth-stating role of doctrine, and thereby implies a fictionalist kind of perspective. I have indeed suggested that the believer's attitude to doctrines is not so much one of intellectual engagement, as of practical reverencing. But this is not because the propositional content of the doctrine, and the truthfulness of the doctrine, are of no consequence. It is rather because a doctrine is taken to record the fundamental nature of things that its vehicle, the doctrine as verbally formulated, is thought to merit reverence. So the doctrine-as-artefact account does not require a fictionalist reading of religious doctrines, and on the contrary, it seems to imply the falsity of that reading. In brief, the action-guiding role of doctrine, as it has been understood here, does not depend on an apprehension of the propositional content of the doctrine, but does still seem to depend on a commitment to the truth in fact of that propositional content. For this reason, we should also say that the role of doctrinal beliefs in lived religion is not simply a matter of make-belief cognitive attitude activated in religious or identity-testing situations, as Van Leeuwen (2014; 2017) has proposed.

Saving religious belief

In these ways, our solution to the ‘puzzle of belief’ is more sympathetic to the believer than are various alternative solutions. Notably, the approach we have been defending allows religious beliefs to count as genuine beliefs, and at the same time saves those beliefs from the charge of irrationality. This approach also enables us to understand the action-guiding role of religious doctrines. A full account of that role should recognize the significance for the status of a doctrine of its propositional content. But this is not to say that the action-guiding role of the doctrine depends on the believer's comprehension of that content. Instead, it may be better to understand that role in terms of a practical reverencing of the doctrine, where this reverencing is to be understood in much the same way as the reverencing of religious artefacts such as icons. If all of this is so, then phenomena such as theological incorrectness should not be seen as unfortunate aberrations, which reflect simply the ignorance of some believers, but rather as a clue to the kind of contribution that doctrines can make to the lived practice of religion.⁷