The nature of the problem

Thinkers in the major monotheisms have conventionally taken the absolute perfection of God – representatively, God’s omniscience, omnipotence, and omnibenevolence – as the basis of God’s worship-worthiness. However, theists have gone further than to say that absolute perfection is a sufficient condition of worship-worthiness and have taken it as a necessary and sufficient condition. As Katherin Rogers puts it, ‘We must define God as best as possible, otherwise we are imagining a being woefully limited by our own imperfections and hence undeserving of our worship.’Footnote ³ Robert Adams characterizes this way of thinking about worship-worthiness (call it the maximal-excellence account) as the recognition ‘not just of God’s benefit to us, but of [God’s] supreme degree of intrinsic excellence’ (Adams Reference Adams1999, 14).

In Maximal God: A New Defense of Perfect Being Theism, Nagasawa challenges the notion that God’s worship-worthiness can adequately be grounded in terms of maximal excellence. Nagasawa’s maximal God approach leaves open the possibility that whatever the greatest metaphysically possible being is, it may not be what he terms ‘omni God’ – that being who possesses omnibenevolence, omniscience, and omnipotence. The core of the perfect being thesis, he contends, is that God is the greatest possible being – and while the greatest possible being may in fact be omni God, it could also be the case that it is metaphysically impossible for the omni attributes to be realized. If omni God is impossible, we ought to conclude then, not that God does not exist – or that there is no perfect being – but that the perfect being is something that falls short of omni God. The greatest possible being, on this account, is what Nagasawa terms ‘maximal God’. Maximal God is the maximally consistent set of great-making properties. If omni God is possible, then maximal God is omni God. If omni God is impossible, maximal God is something short of omni God.

Considering objections to his new defence of perfect being theism, Nagasawa notes the possibility that the maximal God approach might entail that God is not worthy of worship on the maximal-excellence account: ‘Few perfect being theists would be willing to accept the maximal God approach if it can entail that God is not worshipworthy’ (Nagasawa Reference Nagasawa2017, 104). Nagasawa offers the following response to the objection. Suppose that some X is omnibenevolent and omnipotent but only nearly omniscient. Suppose X knows all true propositions except one: p. However insignificant p is, says Nagasawa, X’s worship-worthiness becomes contingent on X knowing or not knowing p. X moves from not worship-worthy to worship-worthy as X comes to know p, and this is absurd. Accordingly, for Nagasawa, it would not necessarily follow that if God is less than omni-perfect, he is any less worthy of worship.

The problem, as Nagasawa defines it, appears to be of the standard soritical kind. We know that three, four, or five grains of sand do not amount to a heap. We know that at three, four, and five days old, a person is a child. That we cannot identify some nonarbitrary value n grains of sand that would turn a collection of grains of sand into a heap, or a value m days old that would turn a child into an adult, illustrates the problem Nagasawa is attempting to establish for the maximal-excellence account of worship-worthiness. Nagasawa’s one-trivial-proposition objection is meant to show that the maximal-excellence account of worship-worthiness would require that a given being go from not worship-worthy to worship-worthy with one uptick in value where the relative difference in value between these beings appears to be too inconsequential or arbitrary to justify the evaluative difference.

Ed Wierenga addresses Nagasawa’s objection head-on:

Wierenga’s reply may work as a narrow response to the specific ‘single proposition’ formulation of the objection. But there is a worry that replies like this amount to more of a fixation on the particular example rather on than the general point Nagasawa is making. Nagasawa’s objection raises an important question about the maximal-excellence account of worship-worthiness in particular, but it also functions as an important question for worship-worthiness generally. Is worship-worthiness all-or-nothing?

The heart of Nagasawa’s objection seems to be that we can conceive of a being so impressive as to land just one tier down on the scale of greatness from absolute perfection and that this being may even be God. Suppose absolute perfection entails the co-instantiation of omniscience, omnibenevolence, and omnipotence. Suppose also that absolute perfection turns out to be metaphysically impossible. If God is the greatest possible being and the greatest possible being is a being who falls just short of absolute perfection, Nagasawa wants to know what reason there is to believe that worship would be an inappropriate stance to take towards this being.

The intuitive appeal of the maximal-excellence account of worship-worthiness is often put this way:

Mark Murphy entertains the same thought: ‘It may well be that complete worship cannot be given to a being conceived as less than perfect–how can one worship God as fully as possible while thinking “but you could have been better” or “there could have been a being greater than you?”’ (Murphy Reference Murphy2017, 3). The problem with this, on Nagasawa’s view, is that the scenario Nagasawa envisions is one in which there could not have been a greater being. Nagasawa might say we have no more reason to believe that this being’s inability to instantiate an impossible set of properties is a disparagement of the value of this being than it would be a disparagement of the value of God’s power to say God cannot create logical impossibilities such as square triangles. In this scenario, God is still supremely valuable (in the sense that he actualizes the most possible value) even if God is not absolutely perfect. If absolute perfection is impossible, it would not be the case that there could have been a better and thus the ‘there could have been a better’ line of reasoning fails as a justification for the maximal-excellence account of worship-worthiness.

The sufficient value-inequality condition and the dual worship-worthiness threshold

Accordingly, it is worth exploring what the notion of worship-worthiness might entail apart from its association with absolute perfection and God’s worship-worthiness. In this section, I present an alternative version of the excellence account. In the following section, I argue this account avoids Nagasawa’s objection and preserves what is attractive about the intuition that God’s worship-worthiness is related to God’s absolute perfection.

To get our minds around worship-worthiness, it will be helpful to consider one of the concept’s most basic features: relationality. Worship is a phenomenon that fundamentally consists of two parties (Smart Reference Smart1972, 26–27; Rachels Reference Rachels1971, 331). There is a subject and an object: worshipper and worshipped. Worship-worthiness is not an intrinsic property of any one being. To talk about worship-worthiness, then, is to talk about a property that supervenes on the relation between two parties. To be worthy of worship is to be worthy of some agent’s worship. Accordingly, worship-worthiness would ostensibly include facts about the subject as well as the object of worship and the relation between them.

Further, to worship a being is to perform actions that express a range of beliefs and attitudes about that being. Following Murphy, I note that the content of such beliefs and attitudes ought to meet two conditions: ‘A being is worthy of worship by some person in this sense only when there is a massive inequality between that being and that person and the inequality is an inequality of a certain sort of value’ (Murphy Reference Murphy2017, 130). The value-gap relation must be of an ‘inescapable and overriding’ kind – one that bears practical priority over other kinds of value (instrumental, etc.). Similarly, the inequality condition is not one of mere degrees of value. Rather, the inequality is one in which the sorts of value are not meaningfully in competition with one another. Worship, then, is a response to a certain sort of value wherein the acts of worship amount to an elevation of the object and a minimization of the subject. It is a basic feature of worship-worthiness that the worship-worthiness of a being depends asymmetrically on the value of that being. This means that the value of a being is more explanatorily basic than whatever other features this being might have.Footnote ⁴

The maximal-excellence account posits a single threshold of greatness for worship-worthiness: absolute perfection. However, if the sufficient value-inequality condition is the key consideration for getting our heads around worship-worthiness, then the minimum amount of greatness requisite for appropriate worship may fall short of absolute perfection. It depends on just how great beings can be without being perfect beings.

Importantly, the sufficient value-inequality condition is not inconsistent with the maximal-excellence account of worship-worthiness. On the sufficient value-inequality condition, it may be that the level of greatness necessary for worship-worthiness just is absolute perfection. If this were the case and omni God is impossible, there simply would be no worship-worthy being. For any being who falls short of absolute perfection, one might withhold worship from this being, not because ‘there could have been a better’ but in virtue of its failing to have satisfied the sufficient value-inequality condition. Even so, it remains unanswered why absolute perfection is the requisite amount of greatness necessary for worship-worthiness. We would need independent reasons to think absolute perfection is a necessary condition and a solution to the vagueness problem presented by Nagasawa.

Let us say then that the minimum threshold requisite for worship-worthiness is met when the sufficient value-inequality condition is satisfied. Once the minimum threshold is met, the worship-worthiness of a being increases proportionately with the value of that being upwards to absolute perfection.Footnote ⁵

There are two things to highlight about the notion of absolute perfection functioning in the present account of worship-worthiness. First, as noted in the Introduction, it involves a relatively thin concept of absolute perfection. Some would say that other more controversial attributes, such as divine simplicity, are required for absolute perfection. With Nagasawa, the present account takes absolute perfection to refer to the omni God as a representative set of attributes. Second, absolute perfection in the standard sense entails an intrinsic maximum of value. To say there is an intrinsic maximum to the value a being might exhibit is simply to say it is not the case that the value of a being could increase indefinitely.Footnote ⁶ However, one construes traits such as omniscience, omnibenevolence, and omnipotence in their more technical definitions; the intrinsic maximum of each means that none could be realized in a more valuable way (Murphy Reference Murphy2017, 9–21).

This last feature of absolute perfection bears a relevant and parallel consequence for supreme worship-worthiness. If absolute perfection is the upper limit of value, and worship-worthiness is a property that increases proportionately with the value of a being to the intrinsic maximum of value, then worship-worthiness too has an intrinsic maximum. This is the second threshold. As we will see, the uniqueness of supreme worship-worthiness has several important consequences.

Vagueness and God’s worship-worthiness

An important aim of the last section was to distinguish the concept of worship-worthiness from the concept of God to clarify what the concept might mean apart from its association with the concept of God. This puts us in a position to see that the vagueness problem for worship-worthiness, as Nagasawa has laid it out, is only a problem for God’s worship-worthiness insofar as there is a singular minimally sufficient threshold of greatness in absolute perfection. In this section, I show that the dual-threshold account of worship-worthiness as I have presented it avoids this problem.

In positing a dual-threshold account of worship-worthiness where God’s worship-worthiness registers at the maximum level in supreme worship-worthiness, we create a distance between God’s worship-worthiness and the concept’s vagueness issues sufficient to dismiss the worries generated by Nagasawa’s objection. In the same way that we may not be able to know whether a man becomes bald at 400 rather than at 500 hairs, we may not be able to know which exact combination of great-making attributes is minimally sufficient for worship-worthiness. But the existence of borderline cases at the minimum level of worship-worthiness is no more a problem for maximal worship-worthiness than borderline cases of baldness are a problem for the legitimacy of predicating baldness of a man with zero hairs on his head.Footnote ⁷

But why think that any move from mere worship-worthiness to supreme worship-worthiness would be a large enough jump to avoid the threshold problems associated with the lower register of worship-worthiness? The answer lies in the special nature of supreme or maximal worship-worthiness. To see this, it will be helpful to distinguish between two ways of thinking about supreme worship-worthiness. Supreme worship-worthiness could be understood in a stronger normative sense or a weaker descriptive sense. Much like Nagasawa’s Maximal God, the weaker sense simply identifies maximal greatness as the upper limit of worship-worthiness, where maximal greatness is understood as the maximally consistent set of great-making properties. This leaves open the possibility that supreme worship-worthiness falls short of absolute perfection. It may be the case on the weaker notion that the supremely worship-worthy being would be worthy of the worship of any human being but perhaps not beings substantially greater than humans if the sufficient value-inequality condition did not obtain between them.

The normative sense is a much more expansive concept. For example, Murphy’s definition of supreme worship-worthiness highlights the expansive nature of the normative notion: ‘x has achieved this upper limit of worship-worthiness if necessarily, if y is an agent (or an agent capable of forming the beliefs, attitudes, etc. relevant to worship) and x ≠ y, then x is worthy of y’s worship’ (Murphy Reference Murphy2017, 130).

The expansive nature of this way of putting it can be illustrated by entertaining the following question: Could there exist a being so impressive as to fail to satisfy the value-inequality condition with respect to the supremely worship-worthy being? Or consider again Nagasawa’s objection to the maximal-excellence account. On the value-differential condition, could a being one downtick in value from absolute perfection appropriately worship the supremely worship-worthy being? On the stronger normative notion, it must be the case that any being who is not the supremely worship-worthy being, however impressive or absurdly valuable, could appropriately worship the supremely worship-worthy being and so satisfy the value-inequality condition. This sense of supreme worship-worthiness, then, presses outward on the boundaries of value. Analogously, there is rational pressure towards characterizing the value of the supremely worship-worthy being as pressing outward against the upper limits of value so that for any being who is not the supremely worship-worthy being capable of worshipping, to worship the supremely worship-worthy being would be a fitting thing to do.

The guiding thought for an account of supreme worship-worthiness is the notion that to be supremely worship-worthy means to be more worship-worthy than any other being. But it seems plain that to be worthy of any rational agent’s worship is to be more worship-worthy than to be worthy of the worship of only a limited range of worshippers. Here is one way to think about it. We can think of worship-worthiness as admitting varying degrees of modal stability. For any worship-worthy being just above the minimum threshold of worship-worthiness with respect to creatures such as humans (or whatever rational creature occupies the lowest status while still possessing the requisite capacities for worship), we can imagine a host of changes to the world in which they exist where they would lose their worship-worthy status with respect to some beings. This being would lack necessary worship-worthiness.

Say that a being just minimally qualifies as a worship-worthy being in a world where Superman will be born. Suppose also that, before Superman’s birth, the only other rational agents in this world are humans and this minimally worship-worthy being. On Superman’s birth, a being substantially more impressive than human beings has come to exist, and this minimally worship-worthy being loses its universal worship-worthy status. So, beings who exist in the lowest register of worship-worthiness have a maximally fragile worship-worthiness status insofar as these beings are the most susceptible to losing that status with respect to the largest range of kinds of beings.

To be supremely worship-worthy, on the other hand, is to have a maximally stable worship-worthiness status. There is no addition of value to any possible world that could result in the supremely worship-worthy being losing its worship-worthiness status with respect to any beings with the requisite capacities for worship.

I will say more shortly about why there is good reason to prefer the stronger to the weaker sense of supreme worship-worthiness. For now, though, it is sufficient to point out that the normative sense clearly secures the value gap requisite to distance supreme worship-worthiness from the vagueness problems in the lower registers of worship-worthiness.

Objections

Before moving on to consider the implications of what has been said so far for Nagasawa’s broader project, I will consider a few objections to the account of worship-worthiness I have been entertaining. First, if worship-worthiness has fundamentally to do with the sufficient value-differential condition, then it appears the minimum threshold requisite for worship-worthiness might be variable depending on the kinds of creatures that exist in the world. For example, if there existed beings sufficiently inferior to humans, it might be the case that these beings could appropriately worship humans as well as God, and this is a strange result.

However, any worship-worthiness relation necessarily presupposes a subject capable of forming the relevant beliefs, attitudes, and desires constitutive of worship. Any such subject must possess powers relevantly like those of the human intellect and will insofar as these powers are necessary for forming beliefs, attitudes, and desires like these. It is hard to envision, then, a being who meets these criteria and is sufficiently inferior in value to human beings so as to satisfy the value-differential condition. This rules out inanimate objects, plants, non-rational animals, and a whole host of other kinds of objects on the grounds that they are without intellect and will.

Importantly, the mere possession of intellectual and volitional powers is not sufficient to determine the lower bound. The central idea is that a being capable of the reflexive beliefs, practical commitments, and self-minimizing attitudes constitutive of worship must possess a substantial form of rational agency. On the account assumed here, agents possessing this form of rational agency remain within a range of value that is meaningfully competitive with that of ordinary human agents, even if their powers are considerably weaker. This is not to say that any two agents with intellect and will possess equal value. It is only to say that the differences among them are not of the inescapable and overriding kind required to make worship fitting.

This rules out rational creatures of lesser intellect and will than human beings as possible worshippers of human beings, since they would not stand in the requisite value-gap relation to us. Homo sapiens may be sufficiently impressive to members of Homo erectus, for example, to inspire worship-like attitudes in them. But the value gap between Homo sapiens and Homo erectus would no more provide grounds for fitting worship than would the value gap between human beings and angels. Human beings may fittingly adopt attitudes toward angels such as wonder, amazement, fear, or dread. But the value gap is not so great as to make fitting the attitudes constitutive of worship, since worship involves the recognition of kinds of value that, as Murphy puts it, are not meaningfully in competition with one another.Footnote ⁸

If this is correct, it provides evidence, in addition to that marshalled in the section ‘The sufficient value-inequality condition and the dual worship-worthiness threshold’, for the claim that worship-worthiness has an intrinsic maximum on value-gap-based accounts. In that section, I argued that there must be an upper bound to worship-worthiness because there is an upper bound to perfection. The divine perfections are such that they each have intrinsic maxima. Still, if it were possible for the value of subjects of worship to proceed downward indefinitely, this might provide an additional sense in which worship-worthiness would be without bound.

However, the considerations offered in reply to the objections that human beings could be worship-worthy on the present account establish that worship-worthiness is constituted both by an upper and a lower bound of value. Since there is both an intrinsic maximum to divine perfection and an intrinsic minimum to the value of the kinds of beings capable of worshipping, worship-worthiness has an intrinsic maximum.

What this means is that, whatever the range of possible minimum thresholds, each will fall within the same relatively narrow range of values insofar as, first, in all possible worlds, creatures similar to human beings in capacity hold the lowest position of value while still having the capacities requisite for worship and, second, worship-worthiness has an intrinsic maximum. So, it is not the case in my defence of the minimum threshold that there is a possible world where both human beings and God are worship-worthy.

Second, and relatedly, might the account on offer introduce a problem with the determinate content of the concept? If the supremely worship-worthy being is necessarily worship-worthy, then this would mean that being would be worthy of the worship even of other worship-worthy beings. But for the supremely worship-worthy being to be worthy of the worship of other worship-worthy beings, that being must satisfy the sufficient value-inequality condition with respect to these other worship-worthy beings. This would mean there is an absurd difference in value between any other worship-worthy being and the supremely worship-worthy being. The minimum threshold for worship-worthiness, then, would stretch the absurd range of value from minimally worship-worthy beings to the supremely worship-worthy being. So, while the account on offer does not allow for the possibility that creatures like humans and beings like God are worship-worthy, does it leave open a wide enough range of value so as to lose its determinate content?

There are two ways to think about the bearing on the range of value of existing beings in a world in which a supremely worship-worthy being exists. One would be that the existence of beings who qualify as worship-worthy but are not the supremely worship-worthy being presses ever outward on the value of the supremely worship-worthy being. The direction of conceptual influence, then, flows from non-supremely worship-worthy beings towards the supremely worship-worthy being such that for any worship-worthy being, the supremely worship-worthy being is always sufficiently more valuable so as to qualify as worthy even of their worship.

The problem with this characterization is that there is only so much space to move upwards in value. Worship-worthiness has an intrinsic maximum; thus, it cannot be the case that the threshold of value requisite for supreme worship-worthiness can be pressed ever outward without qualification.

The second way to think of this relation is that the existence of a supremely worship-worthy being places a conceptual restraint on the range of possible values for other rational creatures. The direction of conceptual influence then, flows from the supremely worship-worthy being downward such that the range of values of beings possessing the requisite capacities for worship is significantly lower than had a supremely worship-worthy being not existed. Putting things this way does not necessarily undermine the validity of the minimum threshold as a real concept; it only brings to light the interesting consequence that if it is possible for the supremely worship-worthy being to exist, it appears to follow that there would be no other worship-worthy beings.

This last point is related to a further consideration, which I cannot fully explore here but furnishes a reply to the objection that the dual-threshold account simply kicks the vagueness can down the road. Wouldn’t it still be the case, on the present account, that supreme worship-worthiness hangs on a difference in value that is too trivial to merit the evaluative difference?

I suspect that adequate handling of this concern would require us to move beyond the thin account of absolute perfection I have here been relying on. Do omniscience, omnibenevolence, and omnipotence secure necessary worship-worthiness? Some venerable traditions place these attributes downstream of more fundamental attributes like divine simplicity, necessary existence, or being the only being that possesses underived value.Footnote ⁹ On reflection, it may turn out to be the case that it is only attributes like these that can secure a maximally modally stable worship-worthiness status because they are the only kinds of traits that secure a value gap that is in principle not traversable. There is a qualitative difference between a being whose value is ‘through itself’, as Anselm says, and a being whose value is derived from that being whose value is underived. If this is right, then supreme worship-worthiness is not vague because it does not depend on attributes that are inherently gradable. But all I have attempted to show here is that on the thin metaphysical assumptions of Nagasawa’s Maximal God, there are resources to show that God’s worship-worthiness in general need not be vague on excellence-based accounts – even if supreme worship-worthiness is vague.

Supreme worship-worthiness and maximal God

Nagasawa presents the vagueness problem for God’s worship-worthiness as a response to the concern that theists would be unlikely to accept the maximal God approach to perfect being theism if it can entail that God is not worship-worthy. For Nagasawa, because it is absurd to believe that a being who is perfect in every respect, with the exception of not knowing a single proposition p, would go from not worship-worthy to worship-worthy as that being comes to know p, it does not follow that, if maximal God is not omni God, then God is not worship-worthy.

I have presented an account of worship-worthiness that dismisses the vagueness concerns for God’s worship-worthiness. Worship-worthiness may have borderline case issues at its lowest registers, but the theist is making no mistake in thinking that God is supremely worship-worthy in virtue of God’s absolute perfection. However, with respect to Nagasawa’s project, it is worth considering whether the reconfiguration of God’s worship-worthiness as I have presented it transposes the concern Nagasawa anticipates. Wouldn’t few people be willing to accept the maximal God approach if it might mean that God is not supremely worship-worthy?

If supreme worship-worthiness is a normative concept, to be supremely worship-worthy is to be (at least) omni God because supreme worship-worthiness corresponds to absolute perfection as the intrinsic maximum of value. If supreme worship-worthiness is a descriptive concept, on the other hand, the supremely worship-worthy being is whatever maximal God turns out to be simply in virtue of having the maximally consistent set of great-making properties – even if this set falls short of absolute perfection.

There are several reasons to prefer the normative construal to the descriptive construal in addition to those already mentioned. Most relevantly, as Jeff Speaks points out, whatever the best thing turns out to be – whatever the maximally consistent set of great-making properties is – depends on what the universe contains (Speaks Reference Speaks2017, 593–594). The greatness of the being with the maximally consistent set of great-making properties is contingent on the available modal space. Perhaps it is not possible, for example, for a being to be both omniscient and omnipotent.

If it turns out in the end that Michael Jordan is the being with the maximally consistent set of great-making properties, then on the descriptive construal of supreme worship-worthiness (which simply identifies the supremely worship-worthy being with the greatest possible being), Michael Jordan turns out to be the supremely worship-worthy being. It appears on this construal that God’s worship-worthiness is merely accidental. We would be compelled to conclude, says Speaks, that atheism is true rather than to conclude that Michael Jordan is God (Speaks Reference Speaks2018, 122–123).

What follows from all of this on the normative construal of supreme worship-worthiness is that if maximal God is not omni God, then a supremely worship-worthy being does not exist. On the descriptive construal, if maximal God is not omni God, we have the possible entailment that Michael Jordan is the supremely worship-worthy being. Now, if there could be beings sufficiently inferior in value to Michael Jordan to satisfy the value-gap criterion with respect to him, then perhaps there wouldn’t be anything especially counterintuitive about this result. But given what has already been said in the objections section concerning the range of the lowest possible value in creatures capable of worship, it seems to be the case that there simply won’t be beings sufficiently less valuable than Michael Jordan who also have the capacities that make worship possible. We have good reason, then, to prefer the normative construal of supreme worship-worthiness to the descriptive one.