Faith is often defended not as blind belief, but as trust in God after belief that God exists has already been formed. On this view, trusting God is analogous to trusting a spouse, friend, doctor, or pilot. This paper challenges that analogy. I argue that even if God’s existence is granted, trust in God’s present guidance differs from ordinary interpersonal trust in three respects. First, interpersonal trust presupposes reasonable confidence in authorship. Before I decide whether to trust a message, I must first determine who sent it. Alleged divine guidance however, often arrives through thoughts, impressions, circumstances, and interpretations that are equally compatible with ordinary psychological explanations. Second, interpersonal trust is corrigible. Trust grows because experience can confirm or challenge our assessment of another person’s reliability. By contrast, trust in God is often insulated from disconfirmation by explanations such as “yes, no, or wait.” Third, trust in God is often rationalised retrospectively through selected memories and reconstructed narratives. Human relationships are also vulnerable to such bias, but they are constrained by observable behaviour, direct feedback, and third-party correction. Retrospective trust in God often lacks these constraints. Together, these asymmetries challenge the God/Spouse analogy.
One major approach to religious fictionalism is to highlight the moral benefits of religious practices, i.e., to argue that even non-doxastic acceptance of the content of religious claims may help practitioners with their moral growth, moral sensibility, moral motivation, and so on (cf. Eshleman 2005; Le Poidevin 2019, 2023; Jay 2014, Leng 2023). In this paper, I argue that this approach fails. Since our worldviews, value systems, and belief systems form an integrated whole in religious practice, it is difficult to isolate moral values (within which moral growth, moral sensibility, and moral motivation are all deeply involved) from this whole. Accordingly, moral benefits seem to be intimately connected with moral and religious belief and cannot be gained separately in religious practices. Instead, I suggest a ritualistic approach. I argue that religious rituals can be beneficial for attaining the meaningfulness of life. As Nozick (1981) argues, meaning consists in transcending limits: being part of God’s plan could be a way of pursuing the meaning of life. While the meaning of life may collapse with the rejection of religious belief, meaningfulness may nevertheless survive through non-doxastic acceptance alone. In this paper, I propose such a novel approach to religious fictionalism.
Religious fictionalism is roughly the view that our engagement with religious discourse, ritual and practise involves pretense. According to the fictionalist, religious talk does not involve assertion and religious thought does not involve belief. When we say things like ‘God is good’, we are merely expressing something like a make-belief that God is good (and perhaps inviting others to do the same). In this paper, I defend a version of hermeneutic religious fictionalism, suggesting that religious fictionalism is not just a practise we should adopt, it is a practise that many (if not most) religious practitioners currently adopt. If I am correct, religious practitioners are not in error; but many philosophers and athiests are.
Jason Thibodeau has recently developed a revised version of the Horrendous Deeds Objection against Modified Divine Command Theory (MDCT). On his formulation, if God has “moral grounding power”—the capacity for a being’s commands to constitute moral obligations—then any omnipotent being would possess the same power. This purportedly allows for a possible world in which a non‑omnibenevolent deity renders horrendous acts morally obligatory. I argue that this objection fails once the nature of moral grounding power is correctly understood. On standard versions of MDCT, moral obligation is identical to being commanded by God. When grounding is construed as identity rather than causal production, Thibodeau’s key premise collapses: identity is not transferable, and it is therefore logically impossible for the property of being morally required to be identical to the commands of any distinct agent, regardless of omnipotence. I further respond to two recent attempts to rehabilitate the objection, concerning alleged cases of type‑identical commands constituting the same normative phenomenon and the purported arbitrariness of restricting moral grounding power to God alone. I conclude that the revised Horrendous Deeds Objection does not undermine MDCT.
