Two challenges for ‘no-norms’ theism

Abstract

A number of theistic philosophers have recently denied that God is subject to moral and rational norms. At the same time, many theists employ epistemological and inductive arguments for the existence of God. I will argue that ‘no-norms’ theists cannot make use of such arguments: if God is not subject to norms – particularly rational norms – then we can say nothing substantive about what kind of worlds God would be likely to create, and as such, we cannot predict the likelihood of any particular evidence given theism. What is more, I argue that this lack of constraint on God's creative act raises a serious sceptical challenge for no-norms theism.

Introduction

The idea that God is exempt from moral and rational norms is not a new one in philosophy of religion; Marilyn McCord Adams, herself a proponent of this view, attributed it to ‘the great medieval philosophical theologians’ (McCord Adams (2013), 16), noting that in their minds, ‘God's aseity and infinity. . . keep God from having moral obligations to us’ (ibid.). More recently, a number of philosophers have embraced what I will call ‘no-norms theism’ (NNT) in order to avoid certain alleged paradoxes relating to the permissibility and rationality of God's creative act.¹

Elsewhere in the literature, one finds many examples of so-called ‘epistemological arguments’ for the existence of God, which typically work by presenting a sceptical challenge, and then offering theism as a potential solution. For example, Koons (2018) argues that the reliability of our rational facilities is to be expected given theism, but not to be expected given naturalism. As such, if we wish to suppose that our faculties are reliable (a necessity if we seek to avoid scepticism), then we ought to accept theism, or at the very least, reject naturalism. Similar arguments have been adduced by a number of theistic philosophers.²

Some of these epistemological arguments may be regarded as part of a larger category, namely inductive arguments, which present certain pieces of evidence as being more expected given theism than given naturalism. For example, the fine-tuning argument holds that the ability of our universe to support life is better predicted by theism than by its negation,³ while the argument from consciousness makes similar claims for the existence of its namesake phenomena.⁴ Some theistic philosophers have placed these inductive arguments at the centre of their philosophical project,⁵ while others have assigned them a less fundamental role; nevertheless, they are widely agreed to be an important part of the contemporary debate in philosophy of religion.

I will argue that one cannot simultaneously endorse no-norms theism and avail oneself of the above-discussed forms of theistic argumentation. What is more, I will argue that the lack of normative constraints upon God's creative act generates a substantial sceptical problem for NNT, based on a recent argument put forth by John Pittard (2021).

Divine norms and inductive arguments

Inductive arguments for the existence of God rely heavily upon the idea that we can say something about what kind of world a perfectly good creator would wish to make. For example, Richard Swinburne writes that ‘given some idea of moral goodness, we have some idea of the kinds of world that God, if there is a God, would be likely to bring about’ (Swinburne (2004), 114). This insight is crucial to his overall project; for instance, he uses it to establish his claim that ‘it is fairly probable that God will create humanly free agents. . . [and so] a beautiful physical universe; and not very improbable that he will create animals also’ (ibid., 123).

We can see that in order to establish his claim that God would be likely to create a beautiful, orderly universe containing human beings (as opposed to, say, a chaotic and ugly universe containing little besides dust and rock), Swinburne relies upon the idea that it is good for such things as beauty, order, and human life to exist. God, being subject to moral norms, ‘will do the best action or an action of the best kind’ (ibid.), and as such, we would expect such things as beauty, order, and human life to exist if there is a God. By contrast, such things are (Swinburne argues) relatively surprising on the assumption of atheism, in part because a naturalistic universe is one which came about via an unguided process (meaning that norms of goodness played no part in the coming-to-be and development of the physical universe).

Swinburne is far from the only theistic philosopher to rely upon considerations of the good when making inductive arguments for the existence of God; for example, Ben Page argues that the existence of consciousness is more likely given theism than it is given naturalism. Of the various reasons he gives in support of this view, nearly all of them involve some kind of normative claim; for example, he writes that ‘consciousness, and many of the things that come along with it, are very great goods. For instance, many moral goods appear to require conscious awareness, with these being goods of a different type, and seem more valuable than other types of goods applicable to non-conscious beings like trees’ (Page (2020), 342). In short, Page argues that God is likely to create consciousness because it is good that consciousness should exist, an argument that presupposes a concern for moral goodness – and, crucially, a rational pursuit of said concern – on the part of God. What is more, it is precisely the fact that a naturalistic universe would not be guided by norms of goodness that leads Page to conclude that consciousness provides evidence on behalf of theism.

With this in mind, it is easy to see how exempting God from moral and rational norms would radically undermine these sorts of inductive arguments for theism. If God is not bound by the norms of goodness, then what reason do we have for supposing that he would favour a world of beauty and order over a world of chaotic ugliness? Why should we expect the existence of conscious life, rather than the cold sterility of a dead cosmos? The answer cannot be that it is better for such things as beauty, order, and conscious life to exist; once God is exempt from the norms of goodness, words like ‘better’ and ‘worse’ lose all relevant meaning.

We can also see that such considerations would undermine the aforementioned epistemological arguments for the existence of God. For example, take Robert Koons's (2018) argument from intuition, which claims that the reliability of our cognitive faculties is to be expected given theism, but not to be expected given naturalism. The obvious question is: why should we expect the reliability of our faculties given the existence of God? If God is subject to moral norms, then the answer is simple: unreliable faculties would mean that most of what we consider our ‘knowledge’ is, in fact, not knowledge at all. But this would entail that we are radically deceived about the nature of our world, and ‘A good God would not subject us to such a radical deception’ (Swinburne (2004), 219).⁶ But of course, once God is exempted from moral norms, we lose any and all reason for supposing that he would ensure the reliability of our cognitive faculties. It may well be ‘better’ for our perceptions to reliably track the truth, but so what? Words like ‘better’ and ‘worse’ carry little weight with a being freed from all normative constraints.

The upshot, then, is that exempting God from moral and rational norms renders us almost completely unable to judge what kinds of worlds he would be likely to bring about. As such, many (if not all) inductive arguments for the existence of God will, given NNT, fall flat.

A sceptical challenge for NNT?

As if this were not enough, it seems that we can raise a sceptical challenge for the no-norms theist. John Pittard has recently argued that, in order to avoid radical scepticism, one must embrace some variety of axiarchism, defined as commitment to the following claim: ‘Because fundamental concrete reality is necessarily ordered toward goodness, the existence of a bad or not very good world is either metaphysically impossible or (antecedently) objectively improbable’ (Pittard (2021), 2). His argument for the anti-sceptical necessity of such a commitment goes as follows:

[Among] conceivable worlds that contain one's internal duplicate, ‘epistemically inhospitable’ worlds (i.e. worlds where all or most of one's internal duplicates are radically deceived) are predominant. This predominance of inhospitable worlds provides a prima facie reason for thinking that the actual world is probably inhospitable. To avoid skepticism, this prima facie support for inhospitableness must be countered by a good reason to think that the actual world is probably epistemically hospitable. (ibid., 1)

From this, it appears that we can derive a problem for the no-norms theist. Pittard argues that ‘if axiarchism is understood in a suitably general way, most theist philosophers can be understood as endorsing [it]’ (ibid.); however, if we reject the idea that God is subject to norms, then this seems to entail that ‘no world would be irrational or immoral for God to create. Even arbitrarily bad ones’ (Rubio (2018), 2988). But if this is the case, then surely God's existence does not render the existence of a subpar world metaphysically impossible, or even ‘(antecedently) objectively improbable’ (Pittard (2021), 2). As such, it is difficult to see how the no-norms theist can possibly rely on God's existence to ensure the epistemic hospitality of our world.

What is more, it seems that even if we assume that Pittard's argument is unsound, NNT may be sufficient to generate scepticism all on its own. The reason is simple: God surely has the power to create epistemically inhospitable worlds, and given NNT, there is no good reason for supposing that he would be unlikely to create them. What is more, if we assume that God's creative act involves him choosing from an infinite set of possible worlds, it follows that there is a potentially infinite number of epistemically inhospitable worlds. As such, even if the no-norms theist rejects Pittard's argument – perhaps because they think conceivability is not a legitimate guide to possibility – they nevertheless have a special reason for thinking that epistemically inhospitable worlds are metaphysically possible, and no corresponding reason to think that they would be improbable. To adopt no-norms theism is effectively to admit the existence of a slightly more benign version of Descartes's demon: a being with infinite power, and no obligation to refrain from global deceit.⁷

In some ways, the point here is similar to an objection raised by Ian Wilks against sceptical theism. Wilks argues once we begin to ‘think in consequentialist terms’ (Wilks (2013), 458), and accept the sceptical theist's claim that we lack ‘knowledge of what it would be evil for God to do or permit’ (ibid., 459), we become subject to radical scepticism, as God may have radically deceived us in the pursuit of some greater long-term good. It is only if there is a real obligation for God not to deceive us that we may be confident in our knowledge of the external world. It seems to me that NNT falls victim to a similar objection: once we concede that it is permissible for God to radically deceive us, we arguably lose all grounds for confidence in our everyday beliefs.⁸ What is more, we certainly lose the ability to invoke theism against Pittard's sceptical challenge.

Together, these considerations imply a serious sceptical challenge for no-norms theism. Once we grant the existence of an omnipotent creator, we gain a powerful reason to endorse the possibility of deceptive worlds.⁹ What is more, once we exempt this creator from moral and rational norms, we lose any grounds for thinking that such worlds are improbable. Therefore, even if Pittard is wrong in supposing that all non-axiarchists should have a prima facie disposition in favour of scepticism, it is certainly plausible to suppose that no-norms theists should.

Objections

These, then, are the two problems confronted by the proponents of NNT: the inability to make use of inductive arguments for the existence of God, and the potential scepticism that accompanies the denial of divine obligations. So, how might the NNT-advocate go about trying to respond? There are, it seems, a number of potential strategies.

Divine love

One initially appealing proposal is that while God may not be bound by normative goodness, he can nevertheless be said to love his creations. As such, we might suppose that God would prefer to create a world which allows for knowledge of, and communion with, himself. However, a brief investigation reveals that this cannot work.¹⁰ This is because there seem to be only two ways in which love might constrain God's creative act: either by providing reasons which would guide God's decisions, or by providing an entirely new form of normativity, distinct from moral and rational normativity. Unfortunately, both of these options are problematic. The first option requires that God be bound by the norms of rationality; after all, in order for us to use an agent's reasons as a guide to the likelihood of their performing a particular action, we must first assume that they are rational in their behaviour, and so will be influenced by their reasons in a manner befitting their rationality. As for the second option, while at least not immediately contradicted by NNT, it is extremely vague. There are currently no well-developed accounts of divine love which would enable it to serve as a distinct guiding norm – one which is sufficient to rule out epistemically inhospitable worlds, while also not falling prey to the general objections that NNT levels against divine morality and rationality.¹¹

Supererogation

Another option for the NNT is to appeal to supererogation.¹² Suppose I were to choose between donating $100 to charity, or donating $200. It seems clear that both of these actions are permissible for me; however, one is clearly better than the other. Hence, an agent can still be motivated by their goodness to choose one permissible action over another. As such, we might assume that the God of NNT, though all actions are permissible for him, would nevertheless choose to create an epistemically hospitable world, since this would be better than creating an inhospitable world. This reply might also allow the NNT to salvage various inductive arguments, since it would allow them to once again judge God's likelihood of performing actions based on how good those actions are.

Unfortunately, even if we grant the contentious assumption that supererogatory acts are possible,¹³ this reply would still (like that before it) require that God be bound by the norms of rationality. It is only if an agent is rational that we may use their reasons to evaluate their likelihood of performing a given action, and therefore, if God is (as NNT claims) free from rational norms, we cannot possibly hope to evaluate him on such grounds.

Divine holiness to the rescue?

Another potential escape route is via an account of divine holiness. For instance, Mark Murphy has recently put forward a view on which ‘the necessary motivational structure of God is better characterized in terms of God's holiness rather than morality or love’ (Murphy (2021), 137). This framework would allow one to deny that God is bound by moral norms, while still imposing some sort of constraints on the worlds that he is likely to create. Unfortunately, this response requires that God be bound by ‘perfect rationality’ (ibid., 125), which ‘entails that God is motivated by God's reasons in the way that those reasons entail that they should motivate God’ (ibid., 125). As such, the no-norms theist – who, as mentioned repeatedly above, denies that God is bound by the norms of rationality – cannot make use of such an account in replying to our sceptical challenge.

Accounts of rationality and reason-responsiveness

Perhaps the most promising escape route for the NNT is to adopt an account of reason-responsiveness which is incompatible with my earlier arguments. These arguments, they might say, depend upon the assumption that an agent is responsive to reasons iff they are rational, and that they are rational iff they are bound by rational norms. But this is not the only – or perhaps even the most popular – account of reason-responsiveness on offer: for instance, one might propose an account according to which to act rationally just is to act according to one's desires and beliefs.¹⁴ On this account, God might well be perfectly rational, without having to be subject to rational norms, so long as he acts in accordance with his own aims, properly informed by his own beliefs.

By my lights, there are at least two problems with this proposal: first, there are substantial objections to such a reductive account of rationality. For instance, John Broome argues that such an account, by defining rationality in terms of reasons, ‘in effect identifies rationality with normativity’ (Broome (2020), 293); however, this identity claim cannot be sustained, since ‘rationality supervenes on the mind whereas complying with normativity does not’ (ibid., 293). Hence, ‘it is false that rationality consists in responding correctly to reasons’ (ibid., 309). In order to uphold the clear intuitive link between rationality and normativity, without mistakenly identifying the two, Broome instead argues that ‘rationality may be a source of normativity’, which ‘requires you to respond to your normative beliefs in a particular way’ (ibid., 310).

A second (arguably more serious) problem is that even those who favour a reductive account of rationality have typically not managed to entirely free it from the spectre of normativity. For instance, Errol Lord's account holds that ‘[one is] irrational because [they] violate a requirement generated by possessed reasons’ (Lord (2017), 1123). Similarly, Benjamin Kiesewetter argues that epistemic reasons are normative in nature, since ‘just like practical normative reasons, epistemic reasons (i) provide partial justification for the responses they are reasons for; they (ii) constitute premises for good reasoning and (iii) are good bases for adopting these responses; relatedly, they (iv) can be the reasons for which agents give the response without thereby making any kind of mistake’ (Kiesewetter (2022), 687). Kiesewetter argues that each of these similarities contributes to a pair of serious problems for any anti-normative view of epistemic reasons; as he writes:

Conditions (i)–(iv) can be understood as instances of having property P. In each case . . . [anti-normativists] have to deny that being a reason that provides justification, constitutes a premise of good reasoning or good basis, or that can be acted upon without mistake is a sufficient condition for being a normative reason. But this raises two important problems. The first is that the relevant conditions are sufficient for normativity in the case of practical reasons. Anti-normativists thus owe us an explanation of why the same does not hold in the case of epistemic reasons. Secondly, the relevant property constitutes a significant analogy between (normative) practical reasons and epistemic reasons. This analogy is well-explained by normativism, which holds that both kinds of reasons are members of a common class, but it remains unexplained by anti-normativism, which denies this. (ibid.)

Hence, the notion of normativity is needed even within a reductive account, in order to draw a boundary between rational and irrational conduct. As such, it is difficult to see how such an account could save NNT, which denies that God is subject to any form of requirement or obligation. An account of rationality which is compatible with NNT would need to be one which makes no reference to any requirements or obligations of any kind. This raises the obvious question of how the dividing line between rationality and irrationality could possibly be drawn, if not via reference to some kind of requiring reasons.

Could one perhaps include normativity in an account of reason-responsiveness, without making reference to obligation? For instance, couldn't one say that an agent is irrational iff they do not act according to a rational norm, even if this does not constitute a failure to meet a requirement? After all, we say that a person who does not abide by Jewish dietary laws is non-kosher, but it does not follow that they have necessarily failed to meet an obligation, since they may not be obliged to follow such norms in the first place.¹⁵ Perhaps then the God of NNT can abide by rational norms even if he is not obliged to do so.

Unfortunately I think this proposal simply misses the point of our objection to NNT. The problem is not that NNT renders it impossible for God to act according to rational norms; rather, the issue is that we have no good reason to assume that he would. To run with the previous example, a non-Jewish person might conceivably decide to eat kosher, despite not being obliged to do so; however, it would be completely unjustified to assume that any given non-obliged person will voluntarily submit themselves to such a norm. Hence, even if such an account of rationality succeeds, the non-obligatory nature of its reasons leave us precisely where we began: with no way of guaranteeing that God, if such there be, will act according to rational norms.

Another possibility is simply to dispense with the concept of rationality altogether, and fall back on an account of reason-responsiveness which makes no reference to it.¹⁶ However, I find this strategy extremely unpromising. The most obvious problem is that there seems to be no available account of what this arational responsiveness would look like, making it rather difficult to judge its potential as a response to our arguments. In fact, it is not at all clear to me what ‘proper response to reasons’ could even mean when entirely divorced from rationality. Even if one rejects (as I do) the reductive account which identifies rationality with response to reasons, it is surely the case that there is some link between the two. Indeed, I find it difficult to make sense of either concept – rationality or reason-responsiveness – in the complete absence of the other.

In short, then, it seems that any account of reason-responsiveness which seeks to completely do away with the notion of normative obligation is bound to be beset by a number of serious issues. Certainly, the fact that NNT commits its adherents to such a view – combined with the fact that, given the success of our objection, a failure of their account will plausibly give way to global scepticism – is a significant result, one which ought to be a source of real concern among adherents of NNT.

Conclusion

We have seen from the above considerations that no-norms theists cannot appeal to epistemological and inductive arguments for the existence of God. Furthermore, they are forced to deal with a serious sceptical challenge, for which there is currently no sufficient solution. Any potential escape route would involve reimposing norms (particularly of rationality) onto God, and as such, would be unavailable to the adherent of NNT. The upshot, then, is that the theist who hopes to appeal to inductive arguments – not to mention avoid radical scepticism – should accept that God is, in fact, bound by norms, particularly rational norms.