Creator Theology and Sterba’s Argument from Evil

Abstract

In this paper, I reformulate Sterba’s argument from evil and consider the various ways theists might respond to it. There are two basic families of responses. On the one hand, theists can deny that God, as a perfect being, needs to act in accordance with Sterba’s moral evil prevention requirements (MEPRs). We can call these responses exceptionalist responses. On the other hand, the theist can deny that God’s acting in accordance with the MEPRs would imply an absence of significant and especially horrendous evil consequences of immoral actions in the world. We can call these responses compatibilist responses. I argue that the availability of both sorts of responses shows that Sterba’s argument should not be taken as a logical argument from evil. A good God is logically possible. However, this does not show that Sterba’s argument fails as an evidential argument from evil. In the second section, I argue that if we work within the framework of what Jonathan Kvanvig calls Creator Theology (CT), the force of Sterba’s argument as an evidential argument is greatly weakened.

In his 2019 book and again in his 2020 paper, both entitled Is a Good God Logically Possible?, James Sterba presents a new atheistic argument from evil (Sterba 2019, 2020). At the heart of Sterba’s argument against the existence of an all-good God is the idea that God would act in accordance with a set of moral principles called moral evil prevention requirements (MEPRs). Sterba claims that if God acted in accordance with these moral evil prevention requirements there would be no significant and horrendous evil consequences of immoral actions in the world. However, it is clear that there are significant and horrendous evil consequences of immoral actions. Therefore, Sterba claims, there is no all-good God.

In this paper, I call into question the cogency of Sterba’s argument.

The paper will proceed as follows. In the first section, I reformulate Sterba’s argument and consider the various ways theists might respond to it. There are two basic families of responses. On the one hand, theists can deny that God, as a perfect being, needs to act in accordance with Sterba’s MEPRs. We can call these responses exceptionalist responses. On the other hand, the theist can deny that God’s acting in accordance with the MEPRs would imply an absence of significant and especially horrendous evil consequences of the immoral actions in the world. We can call these responses compatibilist responses. I argue that the availability of both sorts of responses show that Sterba’s argument should not be taken as a logical argument from evil. A good God is logically possible. However, this does not show that Sterba’s argument fails as an evidential argument from evil. In the second section, I argue that if we work within the framework of what Jonathan Kvanvig (2021) calls Creator Theology (CT), the force of Sterba’s argument as an evidential argument is greatly weakened.

1. Sterba’s Atheistic Argument and Theistic Responses

Sterba presents a new argument from evil. Central to Sterba’s argument are two distinctions about goods. First, there are goods that we have a right to, and goods that we do not have a right to. Second, there are first-order goods, which are goods that do not logically presuppose the existence of some serious wrongdoing, and there are second-order goods, which do. Sterba gives the example of being free from a brutal assault as a first-order good that we have a right to; the good of coming to the aid of someone who has suffered a brutal assault is given as an example of a second-order good that we do not have the right to. With these distinctions in hand, Sterba begins to present a number of moral principles that determine when it is permissible to permit the significant and horrendous evil consequences of immoral actions to occur. First, Sterba notes that we are obligated to provide individuals with goods that they have a right to if we can easily do so without violating another individual’s rights. For instance, if we can easily provide the necessities of life to someone who lacks them without violating anyone else’s rights, we are obliged to do so. Furthermore, Sterba holds that not suffering the significant and especially horrendous evil consequences of immoral actions is a good that we have a right to. Thus, if we can easily make it such that someone does not suffer the significant and especially horrendous evil consequences of immoral actions without violating any else’s rights, we are obliged to do so. Sterba puts this as follows:

Moral Evil Prevention Requirement (MEPR) I: Prevent, rather than permit, significant and especially horrendous evil consequences of immoral actions without violating anyone’s rights (a good to which we have a right) when that can easily be done.

(Sterba 2019, p. 126)

Sterba notes that there are also second-order goods that we have a right to; we have, for example, a right to care if we have been physically assaulted. This particular good presupposes the serious wrongdoing of suffering a physical assault. However, Sterba takes it that no one would morally prefer to have these second-order goods as opposed to having the original wrongdoing never take place. As a result, it seems wrong to allow a physical assault to take place that one could have easily prevented without violating anyone’s rights, just to allow the victim of the assault to exercise their right to medical care after being assaulted. From this sort of case, Sterba generalizes to the following MEPR.

MEPR II: Do not permit, rather than prevent, significant and especially horrendous evil consequences of immoral actions simply to provide other rational beings with goods they would morally prefer not to have.

(Sterba 2019, p. 128)

Sterba then turns to goods to which we do not have a right. For both first- and second-order goods that we do not have a right to, Sterba takes the following MEPR to apply.

MEPR III: Do not permit, rather than prevent, significant and especially horrendous evil consequences of immoral actions (which would violate someone’s rights) in order to provide such goods when there are countless morally unobjectionable ways of providing those goods.

(Sterba 2019, p. 128)

With these MEPRs in hand, we can now state Sterba’s argument as follows:

P1. If God existed, then God would adhere to MEPRs I–III.

P2. If God adhered to MEPRs I-III, then significant and especially horrendous evil consequences of immoral actions would not obtain.

C1. Therefore, if God existed, then significant and especially horrendous evil consequences of immoral actions would not obtain (from P1 and P2).

P3. Significant and especially horrendous evil consequences of immoral actions obtain all around us.

C2. Therefore, God does not exist (from C1 and P3).

Sterba spends some time motivating premise P2 by having us reflect on God’s omnipotence. God could adhere to the Moral Evil Prevention Requirements I-III and permit significant and especially horrendous evil consequences of immoral actions only if he were constrained such that it was not easy for Him to prevent significant and especially horrendous evil consequences or if he were constrained in such a way that there were goods that one might morally prefer to have, but that God could not provide without permitting such evil consequences. Sterba argues that God is not constrained in this way. Oftentimes, we human beings find ourselves in a situation in which it is very difficult (if not impossible) for us to prevent some significant and especially horrendous evil consequences from obtaining. But this is because our causal powers are severely limited. For example, there are currently large amounts of significant and horrendous evil consequences in the world as a result of unjust wars. It would be extremely difficult, or rather impossible, for me to stop this. My causal powers are too limited. Thus, the fact that I “permit” the significant and especially horrendous evil consequences of unjust wars to take place is consistent with my scrupulously adhering to Moral Evil Prevention Requirements I–III. But whereas there are many things that stretch or go beyond my causal powers, it cannot be like this for God. God is omnipotent; therefore, God has the causal power to prevent all of the significant and horrendous evil consequences resulting from immoral actions. Furthermore, Sterba argues, it cannot be the case that God is somehow logically constrained from preventing these evil consequences; if this were the case, says Sterba, then God would be less powerful than we are. For while we are only causally impotent to prevent, for example, all the suffering caused by unjust wars, God would be logically impotent to do so. But it is absurd to think that omnipotent God is somehow less powerful than we are. Therefore, God cannot be causally or logically constrained in a way that makes it difficult or impossible for him to prevent significant and horrendous evil consequences. But given that God cannot be causally or logically constrained in such a way that makes it difficult or impossible for him to prevent significant and horrendous evil consequences, it follows that if God were acting in accordance with MEPRs I-III there would be no significant and especially horrendous consequences of immoral actions (See Sterba 2020, p. 205).

Sterba spends less time motivating premise P1; it seems however, that he takes P1 to fall out of God’s perfection. The idea here can be put into argumentative form. If God were maximally perfect, then God would adhere to the MEPRs; however, if God were to exist, God would be maximally perfect. Thus, if God were to exist, God would act in accordance with the MEPRs.

This line of reasoning seems to be in the background of Sterba’s engagement with Brian Davies. Davies denies that God is a moral agent (See Davies 2006, 2011). According to Davies’ conception of God, it may seem that the subjunctive conditional in P1 is false. If God were to exist, he may or may not act in accordance with MEPRs I–III since, not being a moral agent, God is not subject to the moral law. While Sterba looks to counter Davies’s arguments for an amoral God, at a certain point he is happy to let the point stand. What is at issue is not whether God is subject to moral requirements, but whether God is a perfect being. Says Sterba:

[T]he real problem with Davies account is not so much with his denial that God is subject to moral requirements. Rather, the real problem is that God, if he exists, and were not subject to such requirements, would still admittedly be permitting the horrendous evil consequences of all the immoral actions in the world when he could easily have prevented them without either permitting a greater evil or failing to secure a greater good, which is far more evil than that has been produced by all the great villains among us. That is the real problem.

(Sterba 2019, p. 117)

This is a problem, however, only if there is something in God’s nature that is incompatible with allowing such evil consequences. Presumably, what is incompatible with God’s permitting such evils is God’s perfection.

Sterba does not spend any time in defending P3. He does not need to since experience shows us that the significant and horrendous consequences of immoral actions regularly obtain.

This is Sterba’s atheistic argument. What are the possible responses theistic philosophers can give to it? Given the cogency of the inferences from P1 and P2 to C1 and from P3 to C2, theistic responses should focus on the premises of Sterba’s argument. This leaves theistic philosophers with the options of attacking P1 and P2 since experience confirms P3. We can call the responses that attack P1 exceptionalist responses. This is because, prima facie, the perfection of finite rational beings such as ourselves would require one to act in accordance with MEPRs I-III, but if P1 is false, divine perfection is an exception to this general rule. We can call responses that attack P2 compatibilist responses since the denial of P2 implies that God’s acting in accordance with MEPRs I-III is compatible with the significant and horrendous consequences of immoral actions regularly obtaining.

As an example of an exceptionalist response to Sterba’s argument consider the following.

Exceptionalism: While it is true that human beings should act in accordance with MEPRs I-III, this is because of a deeper moral principle, namely that we have duties of benevolence. We should act in such a way that promotes the good of others. God also has such duties. But we, unlike God, have severe limitations on our ability to promote the goods of others. If we allow others to suffer the significant and horrendous consequences of immoral actions when we could have easily prevented this, we cannot use this to bring about their greater good. Nor can we ever make it up to them so that they might rightly accept that our treatment of them was unsurpassably good (say by giving them an infinitely long life filled with infinite value). But God does not have these limitations, for God is all-knowing, all-wise, and all-powerful. Given God’s omniscience, omnisapience, and omnipotence, even if we have no idea how it is that God might use someone’s suffering for his or her own good, even if we have no idea how God could make it up to someone so that it could rightly be said that God’s treatment of them was unsurpassably good despite allowing this suffering to take place, this does not provide a reason for doubting that God can do it. Thus, while the perfection of men and God demands that they are both beneficent, only the perfection of limited creatures like us implies that one acts in accordance with MEPRs I–III. As a result, P1 of Sterba’s argument is false.

(Cf. Beaty 2021; other exceptionalist responses to Sterba’s argument include (Attfield 2020; Bishop 2021; Hasker 2021; Huffling 2021; Reichenbach 2021; Salamon 2021))

We can compare this exceptionalist response with the following compatibilist response to the argument.

Compatibilism:

Human beings are essentially radically interdependent. To be radically interdependent in the relevant sense is for one’s happiness and well-being to be dependent on the choices and actions of others, and for others’ happiness and well-being to be dependent on one’s choices and actions. Thus, my happiness and well-being is dependent minimally on others refraining from actions towards me that would harm my life, my health, or my psychological integrity. Also, in a more robust way, my well-being and happiness are dependent on certain individuals entering into and maintaining special relationships of love and trust with me. But it is not just that my happiness is dependent on others. My not being miserable is dependent on others too. If others harm my life, health, or psychological integrity, if those in special relationships betray my trust and reject me, it is not just that I will not flourish—I will be deeply unhappy. And what is true of me, is true of every other human being. Furthermore, this is not simply an accidental property of human beings so that in some possible world there are humans, just like you and me, who are not, and never have been, dependent on each other in this way. Such beings might be human-like, but they would be of a different kind. Given this, it is impossible, even for the all-powerful God, to create creatures such as us and to always or for the most part prevent the significant and horrendous consequences of immoral actions. For if God always, or for the most part, prevented the evil consequences of immoral actions from taking place, we would not be the radically interdependent creatures we essentially are. It is true that there would be no significant and horrendous consequences of immoral actions for us if God could easily prevent them all from happening; this is because God acts in accordance with MEPRs I-III. But even an omnipotent God cannot create essentially radically interdependent creatures and prevent all of the evil consequences their choices bring about for each other—in this case, the creatures would no longer be radically interdependent creatures! Because of this, P2 of Sterba’s argument is false. Furthermore, this does not have the absurd consequence that God is less powerful than us. The source of one’s inability to do something does not make one more or less powerful. It is logically impossible for me to become the first cherry tree to grow over 30 meters tall. It may merely be causally impossible for the cherry tree in my front yard to do so. It does not follow that my causal powers are more limited than the causal powers of the cherry tree.

I think that these responses go a far way towards showing that Sterba’s argument is not a successful logical argument from evil. It seems possible that there are beings whose perfection does not depend on their acting in accordance with MEPRs I–III and that the most perfect being is like this. Likewise, it is at least epistemically possible that human beings are essentially radically interdependent beings, and so, God could act in accordance with MEPRs I-III while permitting significant and horrendous evil consequences of immoral actions to take place. When we combine both of these exceptionalist and compatibilist responses to Sterba’s argument, we have good reason for thinking that the existence of significant and horrendous consequences of immoral actions seems to be consistent with the existence of God. This is not to say, however, that Sterba’s atheistic argument cannot be reframed as an evidential argument for atheism. According to evidential arguments from evil, God’s existence is not argued to be inconsistent with the existence of evil; rather, it is argued to be highly improbable. (For the loci classici of evidential arguments from evil, see Rowe 1979, 1991). On this reading of Sterba’s argument, the existence of significant and horrendous consequences of immoral actions, instead of being taken as strictly inconsistent with God’s existence, should be taken as providing us with evidence that God does not exist. How strong this evidence is a function of how likely P1 and P2 are. If we take P1 and P2 to both be likely, then the existence of the significant and horrendous consequences of immoral actions gives us strong reason for atheism.

2. Creator Theology and Sterba’s Atheistic Argument

In the last section, I argued that Sterba’s atheistic argument is not a successful logical argument. It seems possible that God, while being perfect, might not act in accordance with MEPRs I-III. And it is at least epistemically possible that God could act in accordance with MEPRs I-III, while allowing the significant and horrendous consequences of immoral actions to occur. Things may be possible, however, while being improbable. If it is extremely probable that if God were to exist that God would act in accordance with MEPRs I–III, and that if God were to act in accordance with MEPRs I–III there would be no significant and horrendous consequences, then the existence of such consequences would be strong evidence against the existence of God. In this case, while Sterba’s argument fails as a logical argument from evil, it works as an evidential argument.

In this section, I argue that if we work within the framework of what Jonathan Kvanvig has called Creator Theology (CT), as opposed to Perfect Being Theology (PBT), the strength of Sterba’s argument understood as an evidential argument is greatly weakened.

For Kvanvig, CT and PBT are approaches to theology distinguished by their assumptions about the fundamental nature of God. According to PBT, the fundamental nature of God is to be maximally perfect (Kvanvig 2021, p. 6). Traditionally, this conception of God involves identifying the set of perfections with intrinsic maxima and then attributing these to God. Thus, God is thought to be maximally knowing, maximally powerful, and maximally good (cf. Kvanvig 2021, p. 99). According to CT, on the other hand, the fundamental nature of God is to be the asymmetrical source of all that is (Kvanvig 2021, p. 8). Whereas everything that is not God is dependent on God, God is dependent on nothing else. It is important to note that the disagreement between PBT and CT has to do with the fundamental nature of God. Many philosophers who accept PBT will also accept that God is the asymmetrical source of all that is; many philosophers who accept CT will also accept that God is the maximally perfect being. The disagreement between these approaches to theology does not lie in the particular characteristics that they attribute to God, but rather in the characteristics of God that they take to be fundamental. The fundamental nature of God will be used to derive less fundamental aspects of the Deity. For example, we can read Saint Anselm, the prototypical perfect being theologian, as seeking to derive God’s asymmetrical sourcehood from God’s maximal perfection. Says Anselm,

WHAT are you, then, Lord God, than whom nothing greater can be conceived? But what are you, except that which, as the highest of all beings, alone exists through itself, and creates all other things from nothing? For, whatever is not this is less than a thing which can be conceived of.

(as quoted in Kvanvig 2021, p. 144)

A natural gloss on this passage is that Anselm is deriving God’s being the asymmetric source of all things from God’s maximal perfection. Anselm’s reasoning seems to run as follows: given that God is maximally perfect or, in Anselm’s words, that God is “whom nothing greater can be conceived” and, granted that anything lacking maximal independence and maximal creative power is less than maximally perfect, then God, as the maximally perfect being, will also be the asymmetric source of all else. According to this approach, God’s fundamental nature of being maximally perfect grounds God’s aseity and God’s absolute creative and sustaining power.

While PBT seeks to derive God’s being the asymmetric source of all things from God’s perfection, CT, in turn, proceeds from the opposite direction, seeking to derive God’s maximal perfection from God’s being the asymmetric source of all things. This derivation, unsurprisingly, is much more complicated, at least at first glance. Saint Thomas Aquinas, a paradigmatic practitioner of CT, for instance, takes the asymmetric source of everything else to be pure act. According to Aquinas, to be imperfect is to be potentially what one should be actually. Thus, the more actual one is, the more perfect one is. It follows that God, as pure act, is maximally perfect (See Aquinas 1888, Q4 A1). What is important here is not the details of Aquinas’s argument, nor its ultimate viability, but the way in which Aquinas embodies the project of CT. Starting from God’s fundamental nature as the asymmetric source of all things, Aquinas seeks to derive other divine attributes, such as absolute perfection. Thus, while Aquinas and Anselm both agree that God is maximally perfect and that God is the asymmetric source of all things, they embody two distinct approaches to theology.

This difference in starting points between CT and PBT has important consequences when assessing the force of Sterba’s argument as an evidential argument. Starting from PBT, the maximal perfection of the Deity is a given. But if PBT is to allow for any further theological reflection, one’s intuitions about what maximal perfection consists in, and what a maximally perfect being would do, need to be drawn upon. Otherwise, one would be incapable of deriving any other attributes of God from His perfection. Thus, for one working with the PBT framework, one’s conception of the Deity is built primarily upon the intuitions one has about maximal perfection and what this maximal perfection implies.1 It might seem highly probable to someone that a maximally perfect being would act in accordance with MEPRs I–III (even if it does not seem strictly necessary). If one then notices that in fact there is no all-powerful, all-knowing being who acts in accordance with MEPRs I–III, that instead we constantly see the significant and horrendous consequences of immoral actions taking place in the world, one may then take this as strong evidence that God does not exist.

Starting from CT, of course the maximal perfection of the Deity is not a given, but needs to be derived. What guides this derivation are intuitive connections between being the absolute source of all things and general ideas of perfection. We have already seen an example of this sort of reasoning in Aquinas. Aquinas connects being the asymmetrical source of all things with being pure act, and he then connects imperfection with potentiality. As there is no potentiality in God, as he is pure act, there is no imperfection in God either; he is pure act and so pure perfection. Another quite different example of this kind of reasoning can be found in Samuel Clarke (1767). We can paraphrase Clarke very loosely as follows. God’s being the asymmetrical source of all things implies that God is infinite, omnipotent, and omnipresent since nothing could constrain the asymmetrical source of all things, and nothing could exist apart from such a source. Furthermore, this asymmetrical source of all being must be intelligent given the ordered world that he has brought about. And given that this asymmetric source of all things is infinite, omnipresent, and intelligent, it is also all knowing since there is nothing that could limit its knowledge. Furthermore, given that all actions of intelligent beings are directed towards the good and fitting, except in cases of ignorance or weakness of the will, the asymmetrical source of all things always acts correctly since it is without ignorance, and being all-powerful, there is no way for it to exhibit weakness of will. But since no one but God is all-knowing and all-powerful, no one could exhibit as much moral perfection as God.

Suppose someone who followed this Clarkean line of thought also concluded that since God exhibits maximal moral perfection, it is extremely likely that God acts in accordance with MEPRs I-III. Is the fact that there are significant and horrendous consequences to immoral actions strong evidence for atheism? My claim is that it is much weaker for this individual as opposed to the individual working within PBT. This is because we can think of the evidential upshot of the existence of significant and horrendous consequences to immoral actions as being disjunctive. Something in the Clarkean chain of reasoning has gone wrong. But what exactly has gone wrong is not clear. Perhaps, we can move all the way from the fact that there are significant and horrendous consequences of immoral actions, all the way back to the denial that there is an asymmetrical source of all things. Such a line of reasoning seems precarious, however. Perhaps, one should instead take the existence of significant and horrendous consequences of immoral actions as evidence that God is not maximally morally perfect; or perhaps, one should take it as evidence that God’s being maximally perfect does not entail that God acts in accordance with MEPRs I–III or perhaps as evidence that God’s acting in accordance with MEPRs I–III is compatible with there being such consequences of immoral actions. The existence of such consequences causes a moment of cognitive dissonance, but the individual working within CT has a wide range of freedom in resolving this cognitive dissonance that does not necessarily involve rejecting the existence of an all-good God. An equal possibility is that she rejects her intuitions about what God’s moral perfection implies for His actions.

In the case I am considering, someone who adopts CT can find reason for being skeptical about her intuitions concerning what God’s moral perfection implies for His actions without adopting skeptical theism (or giving up on CT). (For a general overview of Skeptical Theism, see McBrayer 2010). Skeptical theism often motivates agnosticism with regard to what God would do if He exists by general appeals to our cognitive limitations. In the case I am envisaging, what motivates skepticism with regard to one’s intuitions concerning how a perfect God would act are other intuitions about the connection between God’s perfection, God’s being the asymmetric source of all being, and our evidence that God seemingly does not act in accordance with MEPRs I–III. Something has to give, and to the extent that the perfection of God is well founded on God’s being the asymmetric source of all being, it is right to give up our intuitions that a perfect God would act in accordance with MEPRs I–III or that God’s acting in accordance with the MEPRs is inconsistent with the world as we know it.

Things are different for one working in the framework of PBT. It is true that the epistemic upshot of the significant and horrendous consequences of immoral actions is also disjunctive for someone working in PBT. In responding to the fact that the horrendous and significant consequences of immoral actions exist, one might deny that God is maximally perfect, or one might reject one’s intuition that God’s perfection implies that God probably acts in accordance with MEPRs I-III. But to do either of these things would be to give up on PBT. Thus, for someone committed to PBT, if they find it intuitively probable that a perfect being would act in accordance with MERPs I–III, the existence of the significant and horrendous consequences of immoral actions provides them with strong evidence that God does not exist.

Of course, nothing I have said here shows that CT is a viable project. Whether it is depends upon the cogency of the arguments of philosophers such as Aquinas and Clarke working in the CT tradition. Even if Aquinas’s and Clarke’s arguments fail, however, this does not show that the project of CT is hopelessly flawed. If philosopher X fails in some project, this is weak evidence that the project cannot be successfully brought about. Furthermore, given the historical prominence of CT, it deserves careful consideration by contemporary philosophers of religion. Nor do these remarks show that philosophers working within a PBT framework will be unable to respond to Sterba’s argument construed as an evidential argument. Perhaps, further considerations about maximal perfection can dislodge the intuition that a maximally perfect being would most likely act in accordance with MEPRs I–III. Furthermore, it is always open to philosophers within the PBT framework to develop compatibilist responses to Sterba’s argument. I briefly sketched such a response above, but there are other possible compatibilist responses. For instance, one might question the coherence of holding that there can be immoral actions without the possibility of the significant and horrendous consequences to these actions. Nevertheless, any such responses provided by philosophers working in the PBT framework can be taken up by philosophers working in the CT framework who find the move from God’s perfection to God’s acting in accordance with MEPRs I–III attractive. As a result, to the extent that theorists working in CT can derive God’s perfection from his being the asymmetrical source of all things, CT provides more resources for dealing with Sterba’s argument from evil interpreted as an evidential argument.

3. Conclusions

In this paper, I have argued that Sterba’s argument from evil fails as a logical argument. Given the possibility of exceptionalist and compatibilist responses to the argument, it seems that a maximally good God is consistent with the existence of significant and horrendous consequences of immoral actions. Thus, we should interpret Sterba’s argument as an evidential argument. I have argued that for philosophers working in the CT tradition Sterba’s argument will not be as threatening as for those working within PBT. This provides further reason for exploring CT as an alternative to PBT.