Abstract
Pascal, with his famous wager, argued in favour of religious practice and faith by appeal to expected payoffs. Here I discuss an asymmetry in similar pragmatic arguments for decisions about God. I begin with the observation that apostates (those who abandon religious practice and/or belief) pay costs not shared by those who never adopt a religion in the first place. Noticing this asymmetry shows these arguments from a new perspective and may also contribute to an explanation of the endurance of religion.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
What I mean to say is that those writing on the benefits of religious belief and practice (Casey 2009), often discuss the benefits without taking the further step of drawing conclusions about what we should do.
One famous truth-independent argument for believing in God is James (1897). A more recent truth-independent argument for religious commitment is from Audi, Rationality and Religious Commitment, 2013: “What I want to stress is… that provided that a religious commitment, whether actual or prospective, is rationally permissible for one on non-pragmatic grounds… it may be rational to nurture or produce this religious commitment on the ground that it does or would greatly help one to do things which, for independent reasons, it is both rational and desirable to do. This practical consideration need not, then, be by itself an adequate reason for nurturing or producing theistic commitment; but it may be conditionally rational in the way just described.” [p. 126] Both James’ argument and Audi’s argument have more complicated forms than the simple truth-independent argument stated below.
See, for example, Casey, op.cit. Note that there are epistemic difficulties with distinguishing truth-independent benefits from truth-dependent benefits in studies such as this: it may be that some of the alleged truth-independent benefits such as the health benefits of prayer may actually be directly bestowed on the believer by God, rather than by the calming effect of the practice itself.
Please note that there can also be pragmatic arguments for religious practice (not involving any religious belief(/s)). Not all religious commitments involve a doxastic component.
I do not want it to appear that I am assuming that religious commitment always or only involves believing a proposition or propositions. As I said in the previous footnote, there are both doxastic and non-doxastic forms of religious commitment. I begin with simple examples only involving one religious belief. The form of argument I present later (on page 4) allows for non-doxastic commitment. (There I talk of the expected utility of religious belief/practice).
Ignoring putative truth-dependent costs and benefits (i.e., the costs and benefits bestowed by God) yields a simple truth-independent pragmatic argument for the belief in God. What would justify us in ignoring truth-dependent costs and benefits in a pragmatic argument for belief in God? One possible justification for this is the thought that Pascal’s wager, the famous truth-dependent argument for belief in God, is fatally flawed, due to the many gods objection or the mixed strategy objection. [For discussion of these objections, see Hayek (2003)]. But even if this is the case (a detailed examination of Pascal’s wager would be out of place here), it is not so straightforward to argue from there that one can simply leave putative truth-dependent costs and benefits out of account altogether.
Note that there are other justifications for ignoring the costs and benefits of belief in God that hold in special cases: in a case in which the truth-independent benefits already outweigh the costs where it is also true that any putative truth-dependent benefits would also outweigh the costs, leaving out those truth-dependent costs and benefits would not make any difference to the calculation. Similarly, in the case in which the truth-independent costs already outweigh the benefits where it is also true that any putative truth-dependent costs would also outweigh the benefits, neglecting the truth-dependent costs and benefits would also not make any difference to the calculation.
Note that for simplicity I am grouping together the costs/benefits of belief and the costs/benefits of religious practice. I am also ignoring cases where a theist chooses to convert to another religion rather than to become a non-theist.
Although the Quran does not explicitly state that apostates should be killed, it is explicitly stated in many Hadith (reports of sayings or actions of Muhammad that are considered by many Muslims to be essential supplements to the Quran), and the death penalty for apostasy was classically part of Shari’ah (Islamic law). For a discussion of the legal treatment of apostasy in Shari’ah, see, for example, Peters and Vries (1976–1977). Some other legal penalties for apostasy that were also part of Shari’ah relate to marriage and to inheritance. For example, the apostate cannot inherit, and Islamic marriages are considered ended by the act of apostasy.
Note that the penalties for apostasy are quite separate from actions (if any) that Islam prescribes towards those who have never been Muslim (for example for atheists who have never been Muslims or for Christians who have never been Muslims).
Note that if the benefits of believing only barely outweigh the costs of believing and the penalties for apostasy are not too large, then it is rational for a non-theist to estimate that when she revisits the calculation she will choose to become a non-theist again and it is rational to factor in a chance she will end up paying these costs. This then is another situation in which it would be rational for the non-theist to take the penalties for apostasy as a disincentive to adopting faith in the first place.
Contrast this with a decision structure we may call a Waver: suppose that Telecom rewards those who change their phone account to Telecom and Vodafone rewards those who change their phone account to Vodafone. In such a situation, it is rational to change company as often as possible, neglecting other possible costs such as time spent on paperwork.
I do not want to overemphasize this fact, as a full explanation the endurance of religion over world-history needs to look at details of each time-period, peoples, religions and details of actual causal mechanisms. I haven’t undertaken this here. Moreover, the claims I have made so far have involved normative decision theory: this aims to use calculations of expected utility to describe how decisions should be made. Real agents often do not live up to norms of rationality.
As I said above (footnote 2), once one includes truth-dependent costs, many people will think that one will run straight into famous problems with Pascal’s wager. Let us for the duration of this paper, suppose that there are no such things as infinite utilities, as these seem to reduce decision theory to absurdity. For example, see Hayek op. cit.).
I said that normative decision theory aims to use calculations of expected utility to judge how decisions should be made. The “should” can be interpreted in different ways here. It is sometimes interpreted in the narrow sense of practical rationality: the action one should perform in this sense is the action that is one’s own best self-interest. In this case, only expected self-interested costs and benefits to the agent are included in calculations of expected utility. Although this seems to be non-standard, we can also make sense of a wide normative decision theory which aims to use calculations of expected utility to judge how decisions should be made, in the widest possible sense of “should”. Wide normative decision theory will include moral costs and benefits in calculations of expected utility as well as self-interested costs and benefits, in order to generate arguments for what an agent ought to do in the widest sense. Supposing that shooting a bank employee in order to rob a bank is morally wrong, the moral cost of shooting a bank employee needs to be counted as a cost to the robber when evaluating the robber’s decisions, regardless of whether or not the robber has any preferences about morality. Difficult questions may arise in such a calculation, such as how to weigh moral costs against self-interested costs, but we can put such questions aside for now.
Note that expected utility is often judged by the satisfaction of the agent’s preferences, and both self and other-regarding preferences are counted as benefits. For example, when deciding whether or not to rob a bank, the fact the robber has a strong preference not to hurt others and the plan involves shooting a bank employee can be counted as a cost to the robber. I doubt that including both self-regarding and other-regarding preferences will correspond to the widest sense of “ought” described above. (Whether or not this is so depends on the truth of some controversial views: e.g. internalism about reasons, Humeanism about rationality and moral internalism). Hence, I prefer a formulation of Wide Normative Decision Theory where moral costs are explicitly included.
Even if the god(/s) of that religion really exist, it may still not be true, but let us grant this for the purposes of this discussion.
Matthew O’Leary has pointed out to me that non-theists may have related reasons for adopting religion. They may agree that only religion can ground morality and take this to be a reason for adopting religion. It seems to me that this would have to be an evidential reason, and would have to rely on a prior conviction that there is an objective morality. I should say that I personally disagree with both theists and non-theists who think that religion has moral authority, because of the Euthyphro Dilemma.
For example, Cardinal Newman writes that a catholic is not permitted seriously to question whether her beliefs are true:
The Church does not allow her children to entertain any doubt of her teachings; and that, first of all, simply for this reason, because they are Catholics only while they have faith, and faith is incompatible with doubt (Newman 1892).
Again, even if the god(/s) of that religion really exist, it may still not be true, but let us grant this for the purposes of this discussion.
Note that this is only true because we are supposing that the notion of infinite rewards or punishments is not coherent.
References
Casey, P. (2009). The psycho-social benefits of religious practice. Dublin: The Iona Institute.
Hayek, A. (2003). Waging war on Pascal’s Wager. Philosophical Review, 112(1), 27–56.
James, W. (1897). The will to believe. New York: Longmans Green & Co.
Johnson, D. D. P. (2009). Error of God: Error management theory, religion, and the evolution of cooperation. In S. A. Levin (Ed.), Games, groups, and the global good (pp. 169–180). Berlin: Springer.
Nola, R. (2013). Do naturalistic explanations of religious beliefs debunk religion? In McLaurin & Dawes (Eds.), A new science of religion (pp. 162–188). London: Routledge.
Peters, R., De Vries, G. J. J. (1976–1977). Apostasy in Islam. Die Welt des Islams, New Series, 17(1):1–25.
Newman, J. H. (1892). Discourse XI: Faith and doubt. In Discourses addressed to mixed congregations. London: Longmans, Green, and Co. (New Edition).
Acknowledgments
Thanks especially to Stuart Brock, Bradley Monton and Greg Dawes for helpful comments on drafts of this paper.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Maslen, C. Pragmatic decisions about god from different points of view: the costs of apostasy. Int J Philos Relig 80, 103–113 (2016). https://doi.org/10.1007/s11153-015-9555-9
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11153-015-9555-9