Understanding religious fictionalism

Religious expressivism ‘emphasizes the non-cognitive role of … religious discourse in expressing emotions, or in prescribing values, as well as perhaps one's intention to act in accordance with the latter’ (187). For example, saying ‘God is just’ could be a way of expressing how important the speaker takes justice to be, how the speaker thinks that everyone ought to be just, and how the speaker herself is committed to the cause of justice. It does not describe God. Religious instrumentalism takes ‘religious discourse … to be understood as useful fiction, a powerful vehicle through which we might realize fuller and less ego-centred lives’ (188). Because of the benefits it brings, one lives as if a religion were true. Religious fictionalism combines both ideas. By pretending that there is a just God one not only expresses certain attitudes about justice, one adopts a frame of mind from which it is easier to contribute to making a better world. The atheist can thereby reconcile her metaphysics with her participation in religious practices and use of religious discourse. Eshleman is an error-theorist regarding standard religious discourse: it aims at giving true descriptions and fails.

The most compelling reason to be a fictionalist is that it connects one to a religious community and ‘[i]t is a powerful thing to gather with people who embrace similar ideals, who are willing to devote time to their elaboration and celebration, and who share a commitment to pursuing their realization’ (191). Statements like ‘God is great’, ‘God is love’, and ‘God is just’ can serve as communal affirmations of common values, attitudes, and commitments and as parts of a shared narrative through which to engage the world. One can accomplish more in terms of self-transformation and world-transformation by being a part of a like-minded group than alone.Footnote ²

Religious fictionalism is not a novel theory of the meaning of religious language but a call for reform. Religious language is no longer to be used to describe and explain reality but solely to edify. The way Eshleman would put it, fictionalism is not a theory of the meaning of religious discourse as used but a reinterpretation. This distinguishes his position from other forms of religious anti-realism that either deny a fact-stating role to ordinary religious discourse or are based in more global anti-realist positions.Footnote ³Qua error theorist, Eshleman thinks that the actual use of typical religious language is problematic – believers, among other things, attempt to describe reality accurately and fail. Eshleman proposes that this use of religious discourse can be replaced with a fictionalist use.

Eshleman repeatedly describes the fictionalist as advocating the reinterpretation of religious language (184, 186, passim). I think it would be better to say that the fictionalist is re-appropriating religious discourse. On the basis of her interpretation of Exodus, a realist might hold that certain wondrous events happened in the history of Egypt. The fictionalist rejects this and Eshleman describes this as a difference in interpretation. However, according to the fictionalist, what's wrong with the realist is not necessarily her interpretation of Exodus but that she thinks what it says is really true. The realist and the fictionalist can agree about what Exodus says; the realist, however, takes Exodus to speak truth while the fictionalist does not. Even so, it can still be used constructively. That's re-appropriation, not reinterpretation. Compare Eshleman with Julian Huxley who, simplifying a bit, proposes that new meanings be assigned to theological terms so that they refer to aspects of humanity and nature.Footnote ⁴ Both think that religious language has been problematic because users have attempted to describe a reality correctly that doesn't exist; however, Huxley proposes that with new meaning assignments to terms, religious language can still be used to speak the truth, whereas Eshleman proposes that religious terms keep their original meaning but be put to a new use. Huxley reforms by reinterpreting whereas Eshleman reforms by re-appropriating.

Fictionalism and varieties of atheism

Eshleman sets out to answer the question ‘Can an atheist believe in God?’ (183), where ‘believe in’ involves holding that ‘it is reasonable … to pursue a form of life shaped by full engagement with theistic religious language and practice’ (184). Whether or not such an engagement is reasonable depends on what the grounds of atheism are. These are diverse. So, in order to know if an atheist can coherently and reasonably accept religious fictionalism, it is necessary to consider reasons for atheism.Footnote ⁵ Some of these reasons tell against adopting fictionalism.

Some atheists are such because they view religion as an inherently oppressive social institution. A Marxist, for example, would understand religion primarily as a tool for class warfare and would not adopt religious language because that language is oppressive. Another sort of atheist may well think that atheism itself has political or moral import. For example, a number of atheists adopt egoism and would not see anything positive in using religious language expressively or instrumentally. Rand, in particular, would accuse the fictionalist of lacking integrity. Eshleman writes that for ‘the instrumentalist, inhabiting the time-tested world of religious narrative and imagery is a valuable means of structuring one's life around a conception of the good’ (188). However, some atheists would say that insofar as this is the very same time-tested narrative and imagery that has given us opposition to scientific and medical advances, opposition to the expansion of human rights, and has provided a rationale for all sorts of violence and exploitation, the general usefulness of such narratives and imagery for furthering the good is questionable. Moreover, there is nothing to block an atheist from appropriating piecemeal whatever wisdom, inspiration, or solace is there without being a fictionalist.

Many atheists, insofar as they have a positive image of humanity, have reason to reject religious fictionalism. Theistic religions glorify God by contrasting God with humanity and they assert that humans profoundly need God. In order to present an image of God as holy, religious orators and writers often portray humanity in extremely negative terms. They also do this to show that humans need God. This is not mere happenstance; it is longstanding practice and is almost dictated by the logic of the situation. Supposing that there is no God, the only way to represent God is by means of things that are real. Insofar as God is represented by means of predicates that in our experience only apply to human beings (e.g. moral and psychological predicates) and insofar as God is represented as meeting a human need, the only way to portray God is to contrast God with human beings. One limns God's greatness by, for example, exaggerating human weakness. Positive terms used to describe God are borrowed from terms used to describe humanity and then God is exalted far above humanity by giving an extravagantly negative depiction of the human condition.Footnote ⁶ Since then there is such a close connection between an account of God and a depiction of humanity, for many atheists, rejecting theism is united with rejecting certain depictions of the human condition. Many would also say that humans simply don't need God. Atheists thus have reason to reject fictionalism insofar as fictionalism would sustain depictions of the human condition that they reject.

There is something deeply problematic about the way the fictionalist proposes appropriating the claims of religious traditions. The fictionalist holds that religious traditions are not reliable guides to truth but are reliable guides to what is good and how it can be achieved. The project seems to suppose that we can form radically separate judgements about a religion's factual claims and moral judgements. However, moral values and beliefs about reality are not separable. For instance, the moral advice given by Jesus was deeply impacted by his belief that the end of the world was at hand. His pacifism was rational given this belief. The problem here is endemic: religious believers valued what they did and advised what they did precisely because they believed that the end of the world was near, or that there would be a final judgement, or that God made the first woman from the rib of a man, etc. Someone who rejects what religious believers regard as the highest truth should have a healthy scepticism regarding what those very believers valued and advised.

A number of atheists, if they have good reasons to be atheists, also have good reasons to be sceptical of fictionalism. That said, some atheists may be atheists for purely evidential reasons that are unconnected with either a moral or political critique of religious practice itself and they may also find the values and depictions of humanity within a given religion to be reasonable. There are, after all, many atheists and many, many religions.

Fictionalism and meaning what one says

The fictionalist says the very same sorts of things as a realist in the same types of situations as realists but allegedly means them differently. While the conventional meaning of the fictionalist's terms is the same as the believer's, the fictionalist does not use them to describe reality but rather to inhabit a fictionality and/or give expression to certain attitudes and feelings. Qua error-theorist, the fictionalist holds that what the ordinary believer says is false, and that this is a failure since she intends to speak truthfully; the fictionalist also regards what she herself says as false, but this is not supposed to be a failure because she did not intend to speak truthfully. Given that the fictionalist's discourse takes place in a public context in which religious language has a standard truth-telling purpose, the fictionalist might be accused of dissembling. A more fundamental issue has to do with meaning: insofar as the fictionalist means her discourse differently than the realist, how does she manage to mean it differently? The difference in meaning cannot derive from the overt behaviour of the speech act itself since these are the same. How then does this difference originate?

Maybe some inner activity accompanying the speech-act gives it its non-standard function. Perhaps the act is accompanied by a kind of inner speech such as: ‘Now, I mean this instrumentally’, or ‘Now, I mean this expressively’, or ‘I don't mean to be describing reality when I say this’. While I have no theory to contribute as to how utterances get their meaning, the inner-speech model is deeply problematic. Inner speech might be related to outer speech as a kind of rehearsal or as a kind of inner commentary through which one tracks one's place within the conversation. As such, it serves a critical-reflective role, not a meaning-giving one. It couldn't possibly serve a meaning-giving role because having certain inner speech is not a necessary condition for meaning something in a certain way. A person can use language in many ways without thinking to themselves anything at all, let alone anything about the purpose of their speech. It would be too peculiar if the fictionalist use of language required a certain kind of inner activity when others did not.

Maybe the key to the difference in meaning between the fictionalist and the realist is not in inner speech but in their speech dispositions. If the speaker is asked what she meant when she said ‘God is just’, her answer could provide the key to correct interpretation. However, when one elaborates, one doesn't merely repeat what was said in different words, what is said is developed and expounded on. Elaboration is a dynamic, forward-looking process aimed at achieving mutual understanding. Thus, it could be that when elaborating, one switches from an instrumental to an expressivist use of language or a realist use to an instrumentalist use. In elaborating one might modify or qualify what was originally said – elaboration can both clarify and correct. Moreover, a fictionalist might elaborate on what she meant from within the fictional narrative she is living – elaboration doesn't seem to require one to, as it were, step out of character.

Let us suppose though that the fictionalist is always able to give a fictionalist self-interpretation of her own utterances – at any moment she is able to step out of character and explain herself. We might even suppose that she has a constant self-interpreting inner monologue in which she distinguishes between her real beliefs and her in-character beliefs. Her utterances are supposed to mean the same as her self-interpretations. Self-interpretation, however, is quite fallible. Someone might misspeak. When asked to elaborate, she may well pronounce what she had really intended to say. The actual meaning of what she originally said need not be the same as her self-interpretation. As she elaborates, she corrects herself. What she said did not mean what she intended it to.

There are cases here other than ordinary misspeaking. Tyler Burge gives an example of someone who says he has arthritis in his muscles.Footnote ⁷ This is not a straightforward case of misspeaking because what the speaker says is what he intended. A speaker's disposition to use the word ‘arthritis’ to refer to a muscle condition does not make it do so. Nor does anything in the speaker's inner speech. Nor does the account the speaker might give of the word. Instead, the word ‘arthritis’ is governed by larger social conventions and practices that give it its meaning and reference.

There is as much a problem assigning a private use to one's language as there is assigning private meaning to one's terms. This will become clearer if we reflect on the fact that what the fictionalist says is said as a contribution to the discursive life of a community. Just as larger social practices govern the meaning of one's terms, so too do they govern what one is doing when one says something.

The way the fictionalist is trying to use religious language is much like the way an actor playing a character uses language. Like the actor, the fictionalist appears to be describing and explaining events, but is really just playing a role. Two actors performing a dialogue are not really conversing, they are co-operating to construct and dwell within a fictional world. With this parallel in mind, suppose a fictionalist is conversing with a number of realists (maybe they are testifying about what God has done for them). The realists are all trying to describe and explain reality and are communicating with each other. For the fictionalist, what the realists say is merely an occasion to perform a certain role. This is similar to what the actor does, except that the actor does so in a special context. Qua conversational speech-acts, the speech-acts of the fictionalist fail to be a performance. The context is simply wrong. While the fictionalist does not intend to be describing and explaining events she is because she is taking part in a conversation in which the group is doing this. The larger conversation in which she utters them plays a role giving meaning to what she says. Although the fictionalist does not wish to be taken as trying to correctly describe reality, the social conventions governing the use of religious speech often presuppose a commitment to truth-telling. Given this, it is correct for her interlocutors to take her as speaking truthfully. Her intention not to be taken this way, as a purely private intention, fails. In other words, while she may really have been trying to use language in a non-descriptive way, the larger conventions of use indicate that she actually was using language in a descriptive way.

Truth-speaking is a longstanding and widespread aim of religious discourse as a social activity and the fictionalist is not at liberty privately to opt out of this. Since fictionalism is supposed to be a reform proposal, we should expect it to be at odds with standard practice. However, a reform cannot be carried out within the private consciousness or practice of a single individual. One is simply not at liberty to use one's own religious language in a privately determined way. In order for the fictionalist to mean what she intends, she must convince others to play along, and this can put her at odds with her community.

Fictionalism and hearing what one wants

In addition to problems with meaning, fictionalism gives rise to problems with interpreting. A fictionalist can interpret a religious claim in one of three ways. A claim might be interpreted purely expressively, it might be judged to be false but incorporated into one's fictionality, or it might be judged to be false but not so incorporated. The religious claim itself might derive from a sacred text (or tradition) or it might come from a fellow community member. These two sources generate distinct problems. Before discussing those, it is worthwhile to note that just as giving meaning to an utterance is not a private affair, neither is interpreting it. There will be problems with interpretation that parallel those discussed above with meaning.

Interpreting members of one's own community is different from interpreting a sacred text or tradition. Maybe one can view a religious text as a living document and selectively ignore what its original authors meant. But, one can't ignore what others in one's community mean. The interpreter has to be true to the person being interpreted. This means that the fictionalist cannot interpret realists as if they were fictionalists. Doing so would violate their independence. It follows that the fictionalist cannot view herself as essentially in agreement with the realist and is not really a fully engaged member of the community.

As Eshleman indicates, one of the purposes of religious discourse is to produce and affirm a common understanding and project. However, since religious discourse is ordinarily used descriptively, the fictionalist rejects the common understanding. Verbally, there may be a harmony between the fictionalist and others, but there is no common understanding of what the shared project is because the realist believes that the goal of religion is to realize the will of God and the fictionalist believes there is no such thing. If it were known that the fictionalist is an atheist, the realists in the community would probably not regard the fictionalist as a full member of the community in good standing.Footnote ⁸ Were the fictionalist to explain and justify her position to members of the community, she would give them reasons to think that God is a fiction and hence reasons not to believe and therefore, given that they are realists, reasons not to be members. Insofar as the fictionalist places a great value on the unity of the community, she has a strong motive to lie. However, she would not feel pressured to lie if she felt at home in the community as a full member in good standing. Further, if the fictionalist thinks members of the community might be led into despair or vice if they are atheists, she could be strongly motivated to excessive paternalism. Eshleman's discussion of the noble lie suggests as much (196–197). If, on the contrary, she recognizes a need to reform the community along fictionalist lines, then she will see the need to persuade believers to change their minds. This would put her at odds with the community. I conclude that the verbal harmony the fictionalist is able to achieve is not sufficient to count as successfully being a full member in good standing.

Regarding the interpretation of sacred texts and traditions, the fictionalist must distinguish the acceptable from the unacceptable. Unless one is going to be an uncritical conformist, some standard or guide is needed. One may not want to accept that God, say, ordered the killing of all the residents of Jericho. Eshleman writes that ‘[t]he image of Yahweh in the Hebrew Bible’ is one ‘unrelenting in love and the pursuit of justice’ (191). While Hebrew scriptures do sometimes describe God in this way, the image of God they present is often quite different. Something is needed to distinguish the good from the bad.

Realists use truth as a standard: we should accept whatever religious claims are true and reject those that are false. Some truth may be inconvenient or difficult, but all of it is to be accepted. Likewise, some false claims could be constructive, but still they are to be rejected. Fundamentalists accept a literal interpretation of Genesis even though doing so is inconvenient and liberals reject it because it is false. Truth is an ideal all realists acknowledge. Since the fictionalist is an error theorist, the fictionalist cannot use this ideal without excluding everything.

Whatever standard the fictionalist uses, it is prima facie implausible that everything the fictionalist thinks should be incorporated into her fictionality is the same as everything realists in her community believe to be true. Very likely she will use a secular moral standard to mark the bounds of her fictionality – fictionally, God is taken to be a perfect utilitarian or a perfect deontologist, etc. However, a Kantian or utilitarian would probably judge some of the actions ascribed to God by religious texts and traditions to be immoral. The realist believer will probably not. Moreover, as Hume and Kant pointed out long ago, religious rituals and practices tend to function as an ersatz morality – religions often present devoutness as an alternative ideal to moral goodness.Footnote ⁹Qua atheist, the fictionalist should reject this ersatz morality. The stories that are alive and authoritative for the realist members of the community will be different from those a fictionalist would choose and the differences are unlikely to be few or minor. In general, the only way for the fictionalist to accept what the community at large does is to be satisfied with something she regards to be far less than ideal.

Fictionalism and belonging as one wishes

The problems of being a fully engaged member of the community and interpreting the religious tradition in a way that one regards as constructive come to a head when we consider that discussing the self-understanding of the tradition is a vital part of the life of the community. In order to be an active participant, the fictionalist must be able to make normative collective belief statements of the form ‘We should believe X’. Such claims might be made to express solidarity or to contribute to the ongoing evolution of the religious tradition's self-understanding. Explaining to members of her own community why she thinks their representation of reality is right allows her to demonstrate her shared understanding and show solidarity. As disputes arise and as the religion changes over time, she'll need to explain why the community should go one way rather than another. In some group endeavours it is enough that participants act co-operatively even though their individual reasons may be wildly diverse or even conflicting. But, insofar as religion is primarily a spiritual endeavour, it requires a considerable common understanding (though not perfect agreement). The fictionalist will thus have to be able to make and justify collective belief statements. This is problematic.

One problem has to do simply with the expression of collective belief. When expressing solidarity, the word ‘believe’ should be spoken univocally with other members of the community. The paternalist fictionalist, for example, is not really expressing solidarity because she means ‘I should believe-as-if X while you should literally believe X’. Qua paternalist, such a fictionalist understands herself to be above the community rather than within it. Another problem has to do with the fictionalist's ability to justify her collective-belief statements to members of her own community. The difficulty is that she must, in order to impact the conversation, give reasons that aim at being intersubjectively valid, but must also give reasons, if she is to really participate rather than simply pretend to participate, that she herself accepts. Qua error theorist, she cannot say something like ‘We should believe this because it is true’.

Once truth is eschewed as a reason to represent reality in a certain way, any replacement becomes problematic. Given a set of conflicting narratives, at most one will be true. So truth, as an ideal, allows one to pursue an objectively valid narrative in order to establish a common mindset. Setting aside truth, no other reasons, even if they aim at intersubjective validity (like moral reasons), could, in principle, select a unique narrative from a set of conflicting narratives.Footnote ¹⁰Many conflicting religious narratives can be constructed to be compatible with deontology or utilitarianism. So, whatever narrative one favours for use within the community must ultimately be favoured for subjective reasons. Someone might favour a particular narrative because she finds it especially moving or because it is more traditional and she is of a conservative bent. Moreover, by eschewing truth as a reason, the fictionalist rejects the very basis others give for their beliefs and thereby, according to the self-understanding of the community, implicitly condemns those beliefs.

Contrast the fictionalist's predicament with scientific instrumentalism. In many cases, a scientific instrumentalist can give intersubjectively valid reasons to accept a certain theory rather than its competitors. In science, the field of competition is relatively small and considerations like empirical adequacy, simplicity, and technological utility can be decisive. When they aren't, the scientific instrumentalist is not disturbed because instrumentalism is essentially pluralistic. There is no need to settle on a single account of how things are or even to pursue such an account – what counts as a useful theory can vary from one situation to another and need not be unique even in a given situation. Scientific instrumentalists do not commit to a single account and treat it as if it were true. In contrast, theistic religion pursues a common, enduring mindset – commitment is essential. This pursuit looks suspiciously like the pursuit of truth and, in any case, it is difficult to understand how to pursue such a mindset without pursuing truth.

The fictionalist cannot fully participate in conversations about how the community should represent reality because she does not really believe the representation, the reasons she can give for collective belief cannot, in principle, be decisive, and she rejects the main reason others give (i.e. ‘because it is true’). Fictionalists do not have the resources to draw a distinction between what the community should accept and reject and are thus dependent on realists for deciding this despite the fact that whatever it is the realists settle on is unlikely to resemble what a fictionalist would think ideal. In order to belong to the community, the fictionalist accepts what others accept even though she thinks the others do not give good reasons to accept what they do as true. As a result, she does not have real ownership of her own religious representation of reality.

Fictionalism and the justification of theism

Eshleman presents fictionalism as analogous to scientific instrumentalism and all along I have been assuming that scientific instrumentalism is not problematic. Although there are instances of selective instrumentalism today, as a philosophy of science, most scientists appear to me to be thoroughgoing realists. Many philosophers are too. Through the work of thinkers like W. V. O. Quine, J. L. Austin, W. Sellars, and others it has turned out quite impossible and unnecessary to draw a systematic distinction between observable entities and theoretical entities. Kripke, Putnam, and others have attacked descriptivist theories of reference and their subsequent work on rigid designation suggests that we can refer to the entities posited by our scientific theories if those entities are real and even if our scientific theories do not accurately describe them. The philosophical ground for wholesale instrumentalism in science has been washed away. As a result, good reasons to posit a theoretical entity as a useful fiction are likely to be good reasons to take that entity to be real. Fictionalism faces a similar problem.

Good reasons to pretend to believe are probably good reasons to believe. It seems to me that the fictionalist wants the world to be the way religion describes it to be or thinks the world ought to be as religion describes it to be. If pretending to believe makes a person better and improves the world, then actually believing should do the same or more. Given this, it seems to me that if one has good reasons to pretend to believe, one has good reasons actually to believe. The fictionalist does not reject theism for moral or political reasons. Instead, the fictionalist rejects theism because of some purely evidential consideration like the problem of evil. Assuming a distinction can be drawn between evidential beliefs (beliefs about the evidence) and other beliefs, evidential beliefs underdetermine other beliefs. The spectacular failure of logical positivism demonstrated this: the movement failed not simply because one can't give a systematic account of the allegedly proper evidential beliefs (protocol statements) but because even if one could, it is not possible to reconstruct the rest of our knowledge on this basis as a deductive system. Evidence plays a role in determining belief, but so do other factors. With the re-adjustment of other beliefs, theism can be made compatible with whatever evidence is alleged to be against it. So, with suitable re-arrangement, the fictionalist could dismiss or explain away the evidence against theism and become an actual theist.

While there is something wrong with merely arranging one's beliefs so that the world comes out however one likes, the fictionalist thinks that believing really will make herself and/or the community better. It is rational for a person to believe in a worldview that she thinks is good for herself and her community. Why not really believe? If the fictionalist finds some psychological barrier within herself to believing, this barrier would equally undermine living as if reality were the way her religion said. If, on the other hand, one is psychologically open to the story of religion and views it as beneficial, one can take up religious practice in order to eventually attain real belief. Pascal's Wager is a serious objection to fictionalism.

Suppose that our fictionalist is an atheist because she thinks there is so much evil and suffering in the world that God would not make such a world as ours. There are many alleged ways of reconciling God with the presence of evil. Presumably, the fictionalist believes that these fail. But, the fictionalist wishes to live as if there were a God. However, in order to have an internally coherent fictionality, the fictionalist must adopt some theodicy. The reason for this is that the fictionalist takes up her real-world experiences within her fictionality – the world is her stage. Since the real world includes evil and suffering, there must be some way of integrating these within the larger fictionality. Ergo, a theodicy is needed as a plot device and it must be reasonable enough that one can live as if it were true. But, if it is this reasonable, why hold that attempts to reconcile God with the existence of evil fail?