Hostname: page-component-54dcc4c588-xh45t Total loading time: 0 Render date: 2025-09-22T11:41:12.658Z Has data issue: false hasContentIssue false

The opposition of omnibenevolence towards evil

Published online by Cambridge University Press:  18 September 2025

Yishai Cohen*
Affiliation:
Department of Philosophy, University of Southern Maine, Portland, ME, USA
Rights & Permissions [Opens in a new window]

Abstract

This article develops two logical arguments from evil that bypass Alvin Plantinga’s Free Will Defence through a critical examination of the relationship between freedom and value. The first argument assumes that morally innocent freedom is valuable, challenging the traditional emphasis on significant freedom. The second argument draws on an interpretation of J.L. Mackie’s underexplored ethical perspective, which highlights a form of evil that contrasts with the positive value of free will.

Information

Type
Original Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2025. Published by Cambridge University Press.

Introduction

In his landmark essay, ‘Evil and Omnipotence’, J.L. Mackie (Reference Mackie1955) echoes Epicurus, contending that the hypotheses that God exists and that evil exists are jointly inconsistent. Mackie maintains more specifically that this inconsistency is generated only through the addition of certain purported necessary truths, or, as he puts it, certain quasi-logical rules. Alvin Plantinga has been widely regarded as having refuted Mackie’s logical argument from evil by ingeniously showing that one of the argument’s quasi-logical rules is not a necessary truth. If Plantinga is correct, then Mackie did not demonstrate that God and evil cannot co-exist. Here is that rule, premise (7), within Plantinga’s (Reference Plantinga1974a, 21–22) rendition of Mackie’s Logical Argument from Evil in the form of a reductio ad absurdum:

The Logical Argument from Evil

  1. 1. God is omnipotent.

  2. 2. God is omnibenevolent.

  3. 3. God is omniscient.

  4. 4. Evil exists.

  5. 5. A being that is omnipotent, omniscient, and omnibenevolent eliminates every evil that it can properly eliminate.

  6. 6. There are no limits to what an omnipotent being can do.

  7. 7. If God is omniscient and omnipotent, then he can properly eliminate every evil state of affairs.

A being properly eliminates an evil state of affairs if it eliminates that evil without either eliminating an outweighing good or bringing about a greater evil (Plantinga Reference Plantinga1974a, 20).Footnote 1 In support of premise (5), Mackie (Reference Mackie1955, 201) says that ‘good is opposed to evil, in such a way that a good thing always eliminates evil as far as it can’. Premises (1)–(7) are supposed to be jointly inconsistent because (1)–(3) and (5)–(7) are supposed to jointly entail that there is no evil, contradicting (4). Assume for argument’s sake that premises (1)–(7) are jointly inconsistent. Presumably, the way in which Mackie wishes to resolve this contradiction, if it is a contradiction, is by denying that any of (1)–(3) are true because each of these premises presupposes that God exists. For the purposes of this discussion, let us provisionally accept this and direct our analysis to premise (7). Plantinga argues that premise (7) is not a necessary truth, and he thinks that he can show this by using several examples that make an assumption about the nature of value. Here is what Plantinga says about such examples:

Under what conditions would an omnipotent being be unable to eliminate a certain evil E without eliminating an outweighing good? Well, suppose that E is included in some good state of affairs that outweighs it. That is, suppose that there is some good state of affairs G so related to E that it is impossible that G obtain or be actual and E fail to obtain…Now suppose that some good state of affairs G includes an evil state of affairs E that it outweighs. Then not even an omnipotent being could eliminate E without eliminating G (Plantinga Reference Plantinga1974a, 22).

I wish to focus on Plantinga’s central example that presupposes the above description of the nature of value, the Free Will Defence. This defence says that, possibly, God is justified with respect to permitting moral evil, that is, an evil that results from free agency, because, although it is not up to God as to whether someone would freely commit evil if they were in certain circumstances, free will has a positive value, and the positive value of free will outweighs such moral evil.Footnote 2 Here is what Plantinga says:

A world containing creatures who are significantly free (and freely perform more good than evil actions) is more valuable, all else being equal, than a world containing no free creatures at all. Now God can create free creatures, but He can’t cause or determine them to do only what is right. For if He does so, then they aren’t significantly free after all; they do not do what is right freely. To create creatures capable of moral good, therefore, He must create creatures capable of moral evil; and He can’t give these creatures the freedom to perform evil and at the same time prevent them from doing so. As it turned out, sadly enough, some of the free creatures God created went wrong in the exercise of their freedom; this is the source of moral evil. The fact that free creatures go wrong, however, counts neither against God’s omnipotence nor against His goodness; for He could have forestalled the occurrence of moral evil only by removing the possibility of moral good (Plantinga Reference Plantinga1974a, 30).

Plantinga’s Free Will Defence presupposes incompatibilism about free will and determinism, a presupposition that Mackie rejects.Footnote 3 If Mackie were correct about the truth of compatibilism about free will and determinism, then Mackie’s Logical Argument from Evil would apparently not be undermined by Plantinga’s invocation of free will; for God would be able to prevent (or, to use the terminology of Mackie and Plantinga, ‘eliminate’) all moral evil simply by deterministically causing every obligatory free action.Footnote 4

I agree with Plantinga that, contra Mackie, incompatibilism is true, and that, as a result, Mackie’s support for the necessary truth of (7) based on compatibilism fails for the following reason. Given incompatibilism, Molinism, and the high value of free will, there may be no feasible world for God (a feasible world is, roughly, one that God can actualize on the condition that the contingently true counterfactuals of creaturely freedom – truths that are not up to God – are true in that world) in which all significantly free creatures never go wrong in the exercise of their freedom because every possible significantly free creature (or, more exactly, every creaturely essence) contingently suffers from transworld depravity, that is, every possible creature is contingently such that, if they had significant freedom, then they would go wrong with respect to their freedom at least once.Footnote 5 Consequently, every feasible world with significantly free creatures may contain moral evil. Nevertheless, given the high value of free will, God is, under the circumstances, justified in actualizing a world with moral evil, and thus premise (7) is not necessarily true (Plantinga Reference Plantinga1974a, 49–53, Reference Plantinga1974b, 184–189).

Despite Mackie’s failure to demonstrate the necessary truth of (7), I will investigate the relationship between freedom and value, and, through this investigation, present two logical arguments from evil that aim to evade the Free Will Defence. The first argument, the Judicious Argument from Evil, assumes that freedom with respect to what I shall call morally innocent actions is valuable. I also engage with the No Best World argument for atheism to consider and respond to an objection to this logical argument from evil. The second logical argument from evil, the Inexorable Argument from Evil, relies on my interpretation of Mackie’s underexplored ethical perspective, as I shall now explain. Mackie’s logical argument from evil relies not only on compatibilism but also on what I shall call Mackie’s value symmetry: for many higher-order goods invoked by the theist as a potential justification for God’s permission of evil, the atheist can invoke a contrasting higher-order evil that exists. Mackie’s value symmetry thus makes salient a potential evil that contrasts with the positive value of free will. Finally, there is some textual evidence suggesting that Mackie entertains, or at least has reason to entertain, a normative ethical principle that is similar in spirit to negative utilitarianism, which I shall call Mackie’s negative consequentialism. This second logical argument from evil that I will defend utilizes either Mackie’s negative consequentialism or a restricted version of Mackie’s negative consequentialism in conjunction with Mackie’s value symmetry.

Both logical arguments from evil have the advantage of employing premises that are consistent with incompatibilism, Molinism, and the high value that significant freedom must have to outweigh moral evil, which is why these arguments can circumvent the Free Will Defence. So, to emphasize, nothing argued for in this article relies either on a rejection of Plantinga’s doctrine of transworld depravity, or, more broadly, on a rejection of either Molinism or incompatibilism. Thus, my overall analysis of this debate is that a viable counter-response to Plantinga on behalf of Mackie involves scrutinizing in further detail the relationship between freedom and value rather than targeting the Molinist metaphysical framework underlying the Free Will Defence.

Before proceeding, I emphasize that Mackie (Reference Mackie1977, 15) is, of course, an error theorist in the realm of metaethics, according to which ‘There are no objective values’. Moreover, it is standard nowadays to understand the error theory as also saying, or implying, that no moral judgements are true (Shafer-Landau Reference Shafer-Landau2021, 333). In that case, such an error theorist would not affirm that evil exists. But it is dialectically permissible for such an error theorist to present a reductio for atheism with a premise that says that evil exists under the assumption that, in virtue of being a moral realist, the theist affirms, or should affirm, that evil exists. It is, in fact, not entirely clear whether Mackie himself thought that no moral judgments are true or whether he thought that, despite the non-existence of objective values, some moral judgments are true according to a revisionary interpretation. After all, in his Ethics: Inventing Right and Wrong, Mackie (Reference Mackie1977) engages extensively in normative ethics, which would suggest that he thinks that there is, at least in some important sense, a true normative ethical position (cf. Berker Reference Berker2019, 11–13). But regardless of how Mackie’s metaethics is ultimately interpreted, it is dialectically permissible for Mackie to present a reductio for atheism that presupposes that evil exists, given the theist’s commitment to moral realism. Moreover, when I present my interpretation of Mackie’s normative ethical perspectives, I will, for simplicity’s sake, skip over the issue of how to interpret Mackie’s normative ethical perspectives given his rejection of objective values.

The value of free will

Let’s begin by clarifying what we mean when we say ‘the value of free will’. The preliminary distinction we should make is between the mere possession of freedom and a free action. Given this distinction, we may say that either the possession of freedom has a positive value or a free action (or, more specifically, an obligatory free action) has a positive value, and each position makes a difference to what we might reasonably consider to be that which has the contrasting negative value, given Mackie’s value symmetry. For instance, if the mere possession of freedom has a positive value, then perhaps contingently lacking free will (lacking free will even though it is possible to have free will) has a negative value, or perhaps contingently lacking free will under certain circumstances, such as under manipulation, has a negative value. On the other hand, if a free action has a positive value, then perhaps an unfree action, or a contingently unfree action, has a negative value, or perhaps an action that is contingently unfree due to manipulation has a negative value. Or, if an obligatory free action has a positive value, then perhaps an impermissible free action has a negative value. I will primarily focus on the distinction between the positive value of a morally obligatory free act and the negative value of a morally impermissible free act, given that Plantinga’s remarks suggest that it is morally obligatory free acts that are valuable.

Two final points about value are in order. First, I will be using the notion of lexical priority in the following manner: if some good G is lexically prior to a different good G*, then no finite number of G*’s, and not even an infinity of G*’s, is such that the aggregative value of each of the values of G*’s outweighs the value of G. Whereas if G is not lexically prior to G*, then some finite number of G*’s, or an infinity of G*’s, is such that the aggregative value of each of the values of G*’s outweighs the value of G. Second, I assume for this discussion that the values of worlds are comparable (Chang Reference Chang, Hirose and Olson2015). I find such comparability plausible, but if the values of worlds are incomparable, then both the rationality of God’s decision to actualize a world and the accompanying explanatory power of theism seem to be jeopardized anyway (Kraay Reference Kraay2011). I now turn to defend my first logical argument from evil, an argument that relies on the value of morally innocent free actions.

Morally innocent free actions

A morally innocent free action is one such that the agent in question can do otherwise, but they cannot perform an alternative action that is either morally impermissible or morally obligatory, and so the agent is not significantly free in Plantinga’s sense. Accordingly, if, as Plantinga thinks, obligatory free actions are valuable, then I maintain that morally innocent free actions are also valuable. This is because the value of both kinds of actions is derived from their freedom and, more specifically, from an ability to do otherwise. For it is only by having an ability to do otherwise that we can be the ultimate source of our behaviour, possess authorship over our lives, and be a locus of creative activity (cf. Plantinga Reference Plantinga1974b, 171; Kane Reference Kane1996, 80; Swinburne Reference Swinburne1998, 84). God can thus create creatures who are capable of performing only morally innocent free actions, and so God can secure the value of free will without permitting any moral evil. In fact, in The Miracle of Theism, Mackie (Reference Mackie and Williams1982, 165) briefly entertains, without explicating further, the thought that God may create agents with what Kant called a holy will, that is, ‘agents with free will but only innocent inclinations’ (cf. Lewis Reference Lewis1993, 153). To be sure, such agents are not significantly free in Plantinga’s sense since they cannot go wrong in the exercise of their freedom, and perhaps significant freedom is more valuable than morally innocent freedom. However, unless the value of an obligatory free act is lexically prior to the value of a morally innocent free act, God could simply create additional beings who possess only morally innocent freedom, thereby offsetting the loss of value derived from significant freedom while also guaranteeing the absence of moral evil (and guaranteeing, more generally, the absence of all evil through God’s omnipotence). For instance, if a morally obligatory free act is twice as valuable as a similar morally innocent free act, then, instead of creating any number of agents with morally significant freedom, God could simply double the number of agents who possess morally innocent freedom, thereby counteracting any loss of value from morally significant freedom while still guaranteeing the absence of evil.

If the atheist can evade the Free Will Defence through the value of morally innocent free actions in the manner just described, then the technical definition of elimination that Plantinga employs in his construction of Mackie’s argument, proper elimination, prevents us from constructing a more charitable version of Mackie’s argument that adopts a different notion of elimination. Indeed, we should have already been suspicious of Plantinga’s employment of the notion of proper elimination due to a different example that Plantinga provides to articulate this notion:

Let’s suppose that E is Paul’s suffering from a minor abrasion and G is your being deliriously happy. The conjunctive state of affairs, G and E—the state of affairs that obtains if and only if both G and E obtain—is then a good state of affairs: it is better, all else being equal, that you be intensely happy and Paul suffer a mildly annoying abrasion than that this state of affairs not obtain (Plantinga Reference Plantinga1974a, 23).

Indeed, God cannot properly eliminate E because eliminating E also results in the elimination of the good conjunctive state of affairs, G and E. But that does not mean that we shouldn’t expect God to eliminate E since God can do so without eliminating the non-conjunctive good state of affairs, G, and the goodness of the conjunctive state of affairs, G and E, is presumably derived from the goodness of the non-conjunctive state of affairs G. Thus, ceteris paribus, God can eliminate E without decreasing the net value of the world, even though God cannot properly eliminate E. I thus suggest that the atheist adopt an alternative notion of elimination to construct a logical argument from evil: a being judiciously eliminates an evil state of affairs if (1) it eliminates that evil without bringing about either an equally weighty evil or a worse evil, and (2) it eliminates that evil without decreasing the net value of the world. God is thus able to judiciously eliminate E even though God cannot properly eliminate E; God can eliminate Paul’s suffering a minor abrasion without eliminating your being deliriously happy, and so God’s eliminating Paul’s suffering a minor abrasion does not decrease the world’s net value. Plantinga (Reference Plantinga1974a, 23) concedes that the above example may be perceived as ‘trivial, tricky, slippery, and irrelevant’ and then proceeds to provide a better example to illustrate the notion of proper elimination:

[T]here are people who display a sort of creative moral heroism in the face of suffering and adversity—a heroism that inspires others and creates a good situation out of a bad one. In a situation like this the evil, of course, remains evil; but the total state of affairs—someone’s bearing pain magnificently, for example—may be good. If it is, then the good present must outweigh the evil; otherwise the total situation would not be good. But, of course, it is not possible that such a good state of affairs obtain unless some evil also obtain. It is a necessary truth that if someone bears pain magnificently, then someone is in pain (Plantinga Reference Plantinga1974a, 23).

Let us suppose that Plantinga is correct that someone’s bearing pain magnificently is good and that this goodness outweighs the evil of such pain. God can presumably still actualize a world that has the same or more positive net value and that contains no evil by actualizing other goods instead that do not include any evils (cf. Schellenberg Reference Schellenberg2004). And if that is true, then even though God cannot properly eliminate this pain, God can still judiciously eliminate it by actualizing a world with a certain positive net value that does not contain this pain.Footnote 6

Mackie may thus contend that even if God cannot deterministically cause an obligatory free action given the truth of incompatibilism, God judiciously eliminates moral evil by, in conjunction with eliminating such evils, replacing the positive value of any number of obligatory free actions with the positive value of some number of morally innocent free actions to counterbalance any loss of value. This replacement does not result in a decrease in the world’s net value and thus seems unobjectionable and to be expected from an omnibenevolent being who is opposed to evil. So, in virtue of articulating what we should expect from an omnibenevolent being through the notion of judicious elimination, we can construct a stronger logical argument from evil.

The Judicious Argument from Evil

  1. 1. God is omnipotent.

  2. 2. God is omnibenevolent.

  3. 3. God is omniscient.

  4. 4. Evil exists.

  5. 5*. A being that is omnipotent, omniscient, and omnibenevolent eliminates every evil that it can judiciously eliminate.

  6. 6. There are no limits to what an omnipotent being can do.

  7. 7*. If God is omniscient and omnipotent, then he can judiciously eliminate every evil state of affairs.

Perhaps Plantinga is correct that there may be no feasible world for God in which significantly free creatures never go wrong in the exercise of their freedom because every possible significantly free creature (or, more exactly, every creaturely essence) may suffer from transworld depravity. But this does not mean that (7*) is possibly false since it is metaphysically necessary that, in virtue of God’s omnipotence, there are feasible worlds for God in which there is any number of creatures who possess morally innocent freedom. So, if the value of obligatory free actions is not lexically prior to the value of morally innocent free actions, then Plantinga’s Free Will Defence does not undermine the necessary truth of (7*), even if it undermines the necessary truth of (7). To be sure, the world might be radically different if it were to contain creatures with only morally innocent freedom and no significant freedom. Perhaps the world would be similar to the sort of afterlife conceived in the final chapter of Augustine’s (427/Reference Bettenson1972) City of God, in which agents cannot perform a morally impermissible action, and this in turn might require either different laws of nature or beings with different psychological architectures from our own.Footnote 7 But as long as this radically different world with morally innocent freedom does not result in a loss of some value that is lexically prior to the value contained in this world, it seems that we should expect God to actualize such a radically different world, given that it is, unlike the actual world, devoid of evil. As Mackie (Reference Mackie1962, 154) says in his extended discussion of his (Mackie Reference Mackie1955) argument: ‘If the theists tell us that God will eventually bring this utopia into being, the critics can hardly be blamed for wondering why he has gone such a long way round about it.’

Plantinga is not the sole proponent of the general contours of the Free Will Defence. After all, with some notable exceptions ranging at least from the classical Protestant tradition to the Ash’arite tradition, the absolution of God from moral evil through libertarianism can be found across all the Abrahamic traditions. So, the idea that significant freedom is immensely valuable should not be lightly dismissed. However, for the reasons previously explained, the Judicious Argument from Evil is not in tension with this idea. To undermine the necessary truth of (7*), a proponent of the Free Will Defence must not only affirm that an obligatory free action is immensely valuable, but rather must also affirm the much stronger claim that the value of an obligatory free action is lexically prior to the value of a morally innocent free action, in which case the best feasible world for God may contain moral evil, and thus (7*) is, in fact, possibly false. So, even if we were to concede for the sake of argument the position that obligatory free actions are both immensely valuable and more valuable than similar morally innocent free actions, this concession would not suffice to undermine the case for the necessary truth of (7*).

But I am not willing to concede this position anyway for a few reasons. First, several theists – including Augustine (427/Reference Bettenson1972, 1089), Anselm (1076/Reference Anselm and Williams1996, 103), and Leibniz (1710/Reference Leibniz and Austin Farrer1985, 386) – maintain that divine freedom, freedom without an ability to perform an impermissible act, is greater than significant freedom, the kind of freedom possessed by God’s creation. This makes sense given that God is that than which none greater can be conceived.Footnote 8 But this position is at least in tension with the idea that creaturely significant freedom is more valuable than creaturely morally innocent freedom. Second, if there is a difference in value with respect to these two kinds of freedom, then I suggest that it is, in fact, a morally innocent free action that is more valuable than a similar obligatory free action for the following reason. Consider the transformation of an unfree action into a free one by adding an alternative possibility.Footnote 9 If we add a sufficiently worse option than the actual action, then the actual action becomes an obligatory free action, and if we instead add an alternative possibility that is equally valuable to the value of the actual action, then the actual action becomes a morally innocent free action. Now, it strikes me that if there is an evaluative difference here, then adding an equally valuable option is better than adding a worse option. The supposition that adding a worse option is better than adding an equally valuable option seems to be emblematic of a reckless mindset that irrationally values taking unnecessary risks. This mindset also suggests, implausibly, that an agent’s significant freedom would be even more valuable if that agent possessed an even worse option than the worst option that they, in fact, have. Indeed, this implausibility seems to be a natural extension of one of Swinburne’s positions on free will:

[T]he good of freely forwarding the good is better if the agent has a free choice between a greater and a lesser good [as opposed to two equal goods]; better still if he has a free choice between the good and the bad, and even more so if the possibility of doing bad includes the possibility of doing wrong … not just between alternative equal goods (Swinburne Reference Swinburne1998, 84).

If it is even better for the possibility of doing bad to include the possibility of doing wrong, then, given that there are degrees of wrongness (and rightness), the value of free will presumably increases whenever there is an addition of an alternative possibility of committing a worse wrong than the worst wrong that one can, in fact, perform. And this is exactly what Swinburne (Swinburne Reference Swinburne1998, 85, 87) seems to think when says that an agent ‘is an ultimate source in an even fuller way if the choices open to him cover the whole moral range, from the very good to the very wrong’, and that ‘the more serious the free will and the stronger the contrary temptation, the better it is when the good action is done’. But it is doubtful that a virtuous individual has even a pro tanto reason to lament, whether for themselves or the world, the fact that they cannot perform far worse actions than the worst ones that they can, in fact, perform.

To illustrate these points with an example, consider the unfree action of deciding to give a friend an extra ticket to a rugby match (we may suppose, for example, that the action is determined). Let us first transform this unfree action into an obligatory free action (that is not determined) by supposing that exactly one other option is available to the agent, viz., deciding not to give anyone the extra ticket, and that acting on this option is morally impermissible. Next, let us instead transform this unfree action into a morally innocent free action by supposing that exactly one other option is available to the agent, viz., freely deciding to give the extra ticket to a different friend instead. This free action is morally innocent because, as a matter of stipulation, there are only two options available to the agent: deciding to give the extra ticket to the first friend or the second friend, and each of these two options is, by stipulation, neither morally obligatory nor morally impermissible. Since neither option is either impermissible or brings about an intrinsic evil, the agent in question is not capable of moral evil in these circumstances and thus is not significantly free in Plantinga’s sense. The friend who does not receive the ticket is indeed harmed, but we can suppose that, in this case, the harm merely consists in the absence of a benefit rather than bringing about an intrinsic evil for the friend, such as pain (cf. Bradley Reference Bradley2009, 47–48).

The obligatory free action of deciding to give the extra ticket thus does not appear to be better than, let alone lexically prior to, the morally innocent free action of deciding to give the extra ticket. For we started with the same unfree action and then rendered that action free by bestowing upon the agent either a worse option or an equally valuable option. And if we’re adding the same value of free will in each case in virtue of adding an alternative possibility, then, as I originally suggested, the value of the obligatory free action is, in fact, equal to, not better than, the value of the morally innocent free action. But if there is a difference in value here, it seems that adding an equally valuable alternative possibility is better than adding a worse alternative possibility. We might even maintain the stronger position that it is the value of a morally innocent free act that is lexically prior to the value of a morally significant free act if we agree with Susan Wolf’s (Reference Wolf1990, 79–80) position that there is no value in an ability to act badly or wrongly. If we wish to agree with Wolf while simultaneously preserving the idea that there is value in leeway freedom in virtue of being the ultimate source of one’s behavior, then we might say that this value is restricted to leeway freedom that involves options that are good and permissible, and that free actions are valuable insofar as they issue from such leeway freedom (Pereboom Reference Pereboom2021, 162, fn. 8). At any rate, even if we were to concede that adding a worse option renders the obligatory free action better than a morally innocent free action, given that both free actions have the same good outcome, viz. a friend receiving an extra ticket, it does not seem plausible that the obligatory free action’s value is so great as to be lexically prior to the value of a similar morally innocent free action.

Perhaps a more charitable interpretation of Plantinga’s position is that it is specifically the praiseworthiness of an action that is valuable, and praiseworthiness requires an ability to perform an alternative action that is morally impermissible. In other words, praiseworthiness requires morally significant freedom. In that case, the value of the praiseworthiness of an obligatory free act may be thought to be lexically prior to both the intrinsic and extrinsic value of a morally innocent free act, given that the extrinsic value of a morally innocent free act is not derived from the intrinsic value of praiseworthiness for that act. Call this the Praiseworthiness Objection to the necessary truth of (7*). This objection may be closed off to source incompatibilists who claim, typically based on Frankfurt cases, that moral responsibility does not require an ability to do otherwise, even though moral responsibility requires indeterminism. Moreover, not everyone agrees with the idea that praiseworthiness requires an ability to do otherwise, even if blameworthiness requires an ability to do otherwise (Nelkin Reference Nelkin2011; Wolf Reference Wolf1980). But let us admit, at least for the sake of argument, that praiseworthiness requires an ability to do otherwise. If we maintain that someone can also be praiseworthy for freely performing an action in virtue of that action being supererogatory, then praiseworthiness does not require significant freedom, even if it requires an ability to do otherwise. For instance, suppose that you can either benefit one person or benefit two people, and suppose that both options are morally permissible, but only the latter option is supererogatory. In that case, an agent may be praiseworthy for freely performing a supererogatory act, even though that agent could not have performed a morally impermissible act instead. So, if the value of an obligatory free action is derived from the agent’s praiseworthiness for performing that action, then God can secure this value by creating creatures who possess morally innocent freedom and who possess the ability to freely perform supererogatory actions. In virtue of creating creatures with morally innocent freedom and no creatures with morally significant freedom, God preserves the value of praiseworthiness while still eliminating all moral evil.

A proponent of the Free Will Defence might grant that praiseworthiness can stem from supererogation but insist that supererogation itself requires morally significant freedom. After all, supererogation may be colloquially articulated as an action that goes above or beyond the call of duty, which in turn implies that there is a duty, and according to standard deontic logic, if an agent can perform an obligatory action (a dutiful action), then that agent’s alternative non-supererogatory action is impermissible. My first response to the proponent of the Free Will Defence is that the nature of supererogation is controversial, given that neither it nor permissible suboptimality is representable in standard deontic logic (McNamara and De Putte Reference McNamara, De Putte, Zalta and Nodelman2022, 34–37). Moreover, McNamara (Reference McNamara and Timmons2010, 212) defines a supererogatory act as follows: ‘the agent did better than she would have done had she done the least she could have done’, according to which ‘could’ is to be understood in terms of moral permissibility. So, if McNamara is correct, then supererogation does not require morally significant freedom, in which case the Praiseworthiness Objection does not undermine the necessary truth of (7*).

Unfortunately for the proponent of the Judicious Argument from Evil, however, this supererogation response to the Praiseworthiness Objection fails for an entirely different reason. Even if Plantinga were to concede that one can be praiseworthy for performing a supererogatory act and that supererogation does not require morally significant freedom, Plantinga can still fall back on the Molinist modal landscape that motivated his doctrine of transworld depravity in the first place. Just as, given the nature of Molinism, it is possible that every possible (creaturely) person (or, more exactly, every creaturely essence) contingently suffers from transworld depravity, so too it is possible that every possible person contingently suffers from what I shall call transworld universal suboptimality: An essence E suffers from transworld universal suboptimality if and only if for every world W such that E entails the properties has an ability to perform a morally innocent free act that is supererogatory and performs at least one morally innocent free act that is supererogatory, there is a state of affairs T and no action A such that

  1. (i) T is the largest state of affairs God strongly actualizes in W,

  2. (ii) A is a morally innocent free action that is either permissibly suboptimal or supererogatory for E’s instantiation in W,

    and

  3. (iii) if God had strongly actualized T, E’s instantiation would have performed a morally innocent free action that is supererogatory by performing A.Footnote 10

Given the possibility that every possible person contingently suffers from both transworld depravity and transworld universal suboptimality, the best feasible world for God may contain moral evil because, possibly, the best feasible world contains praiseworthiness derived exclusively from obligatory free actions rather than from morally innocent free actions, despite this world also containing moral evil (that is brought about through morally significant freedom).Footnote 11

At this point, the defender of the Judicious Argument from Evil could resort to the position that only praiseworthiness does not require an ability to do otherwise. While I find this asymmetrical position on responsibility plausible, I prefer to respond to the Praiseworthiness Objection without relying on it. So, I respond to this objection in two other ways. First, the Praiseworthiness Objection at least suggests that, ceteris paribus, you are obligated to choose future situations in which you can perform the worst action among all your future possible actions to increase the degree to which you are praiseworthy for avoiding that worst future action; in other words, the worse that you can do, the greater your praiseworthiness for fulfilling your obligation instead. However, as discussed in my earlier criticism of Swinburne’s perspective on the value of free will, it is doubtful that a virtuous individual should lament, whether for themselves or the world, the fact that they cannot perform far worse actions than the worst ones that they can, in fact, perform.

My second response to the Praiseworthiness Objection employs Gary Watson’s (Reference Watson1996) distinction between two notions of moral responsibility: attributability and accountability. To be responsible for an action in the attributability sense is for that action to express one’s evaluative commitments or ‘what one stands for’ (Watson Reference Watson1996, 233). Responsibility in the accountability sense involves being imposed with certain demands and being liable to adverse sanctions, including being subject to reactive attitudes such as indignation and resentment (Watson Reference Watson1996, 235–238). While Watson focuses primarily on blameworthiness, the negative aspect of accountability, he does acknowledge a positive aspect of accountability, viz., praiseworthiness, in terms of a positive recognition of the agent as an appropriate response to the agent’s respect for moral ends. He is ambivalent about whether such praiseworthiness requires avoidability, given his sympathies with Wolf’s (Reference Wolf1990) asymmetry thesis. However, Watson is clear that blameworthiness in the accountability sense requires avoidability since it would be unfair to be subject to adverse sanctions for performing some action if that action is unavoidable (Watson Reference Watson1996, 239–242).

Given Watson’s distinction, the notion of praiseworthiness at issue in the Praiseworthiness Objection is arguably praiseworthiness in the accountability sense, given that this objection presupposes that praiseworthiness requires avoidability in virtue of presupposing that praiseworthiness requires obligatory free actions (or morally innocent free actions that are supererogatory). Alternatively, a similar notion of (positive) responsibility that may be thought to be at issue in the Praiseworthiness Objection is (positive) responsibility in the basic desert sense. To be responsible in the basic desert sense is to be worthy of praise or blame simply in virtue of wittingly acting rightly or wrongly, rather than in virtue of any consequentialist or contractualist considerations in favour of praise or blame (Pereboom Reference Pereboom2014, 2). Perhaps responsibility in the accountability sense is an extension of basic desert responsibility. At any rate, the important point for our purposes is this. Consider the possibility in which God is confronted with a modal disaster: every possible person contingently suffers from both transworld depravity and transworld universal suboptimality, and so every feasible world for God that contains praiseworthiness in the accountability sense also contains moral evil. Nevertheless, given Watson’s distinction, God can still actualize a world with any number of agents who are praiseworthy in the attributability sense for their morally innocent free actions, regardless of whether such actions are supererogatory or permissibly suboptimal. Therefore, even if we were to grant that the value of praiseworthiness for some action in the accountability sense is greater than the value of praiseworthiness for a similar action in the attributability sense, this will not suffice to undermine the necessary truth of (7*). Rather, the proponent of the Praiseworthiness Objection must assert the stronger claim that the value of praiseworthiness in the accountability sense is lexically prior to the value of praiseworthiness in the attributability sense.

To see why I am skeptical of this lexical priority claim, consider the following two scenarios. In the first scenario, your choice is between two morally innocent free acts, each of which is morally permissible: benefiting Alex or benefiting Bob. In the second scenario, your choice is between two morally significant free acts, benefiting Alex or harming Alex, such that the former is morally obligatory and the latter is morally impermissible. Now, suppose that in both scenarios, you freely benefit Alex. In that case, in the former scenario, you are praiseworthy only in the attributability sense, whereas, in the latter scenario, you are praiseworthy both in the attributability sense and in the accountability sense. Now, to repeat, perhaps the value of being praiseworthy in the accountability sense for benefiting Alex is greater than the value of being praiseworthy in the attributability sense for benefiting Alex. But given that each kind of praiseworthiness arises from similar actions (both of which are free) with the same good outcome, it is certainly not obvious that the value of the praiseworthiness in the accountability sense for benefiting Alex is lexically prior to the value of praiseworthiness in the attributability sense for benefiting Alex, and similar remarks apply to the relationship between praiseworthiness in the basic desert sense and praiseworthiness in the attributability sense. Hence, given the strength of the lexical priority claims in this section, the burden appears to be on the proponent of the Free Will Defence to argue for these claims. And unless this burden can be met, there appears to be good reason to affirm the necessary truth of (7*).

A ‘No Best World’ objection

If we suppose that there is a best possible world, then we may be inclined to think, perhaps alongside Voltaire, that such a world is devoid of evil. However, most philosophers seem to think, contrary to Leibniz (1710/Reference Leibniz and Austin Farrer1985), not only that the actual world is not the best one, but that there is, in fact, no best possible world. Instead, there is an infinite hierarchy of increasingly better worlds; for any world that we conceive, we can conceive of a similar but better world with additional intrinsic value. Therefore, since there is no best possible world, one might object to the necessary truth of (5*) on the grounds that God is not expected to actualize a world with no evil since for any world without evil that God can actualize, God can actualize a world with equal net value that nevertheless contains evil; all that matters is the net value of the world, whether or not that world contains evil. I shall argue that while this ‘No Best World’ Objection initially appears to undermine the necessary truth of (5*), a counter-response to this objection, in fact, strengthens the case for the necessary truth of (5*). To see this ‘No Best World’ Objection to the necessary truth of (5*) in further detail, we must take an extended detour through the No Best World Argument for atheism (Rowe Reference Rowe2004). Here is one formulation of the argument in the form of a reductio by Kraay (Reference Kraay, Draper, Talliaferro and Quinn2010, 483):

The No Best World Argument

  1. NBW For every world w that is within God’s power to actualize, there is a better world, x, that God has the power to actualize instead.

  2. P1 If it is possible for the product of a world-actualizing action performed by some being to have been better, then, ceteris paribus, it is possible for that being’s action to have been (morally or rationally) better.

  3. P2 If it is possible for the world-actualizing action performed by some being to have been (morally or rationally) better, then, ceteris paribus, it is possible for that being to have been better.

  4. G There possibly exists a being who is essentially unsurpassable in power, knowledge, goodness, and rationality.

These four propositions are jointly inconsistent, and so proponents of this argument conclude that G is the proposition that must go. This argument is importantly different from logical arguments from evil. It is a purely a priori argument from improvability for the necessary truth of atheism, whereas logical arguments from evil for atheism assert the a posteriori claim that evil exists. But the former argument is relevant to the Judicious Argument from Evil insofar as the theist’s rejection of either P1 or P2 may also constitute an objection to the necessary truth of (5*), as I shall now explain.

Consider an objection to P1 by Daniel and Francis Howard-Snyder (Howard-Snyder and Howard-Snyder Reference Howard-Snyder and Howard-Snyder1994, Reference Howard-Snyder and Howard-Snyder1996). They offer a two-step process for God’s actualizing a world. First, God must partition worlds between those that are acceptable and those that are unacceptable based on axiological considerations. Second, God must randomly pick one of the acceptable worlds to actualize. The Howard-Snyders claim that even though the world that God actualizes is surpassable given NBW, this two-step process that God employs is unsurpassable. So, God’s actualizing a surpassable world based on this two-step process does not imply that God’s act of actualizing that world is morally or rationally surpassable, in which case P1 is false.

If the Howard-Snyders are correct that there is an unsurpassable two-step process, then each step of that process must be unsurpassable. However, Kraay (Reference Kraay, Draper, Talliaferro and Quinn2010) is skeptical of either step being unsurpassable. He thinks that it is more plausible to suppose that there is an infinite hierarchy of increasingly better partition principles as well as of increasingly better forms of random selection. Regarding the former, Kraay (Reference Kraay, Draper, Talliaferro and Quinn2010, 486) says that ‘there are infinitely many points at which to partition the set of possible worlds’. Regarding the latter, Kraay (Reference Kraay, Draper, Talliaferro and Quinn2010, 487) invites us to imagine a randomizer A that God uses to randomly pick a world among the set of acceptable worlds. This randomizer is surpassable. Consider randomizer B that is just like A except that B adds one unit of value to a unit of value n that B initially randomly picks, and this addition of values represents the value of the world that B ultimately picks. Thus, randomizer B is more likely to pick a better world than randomizer A, and this, in turn, makes B a better randomizer than A. But, of course, randomizer B is surpassable by yet another randomizer C that adds two units of value to a unit of value n that C initially randomly picks, and this addition of values represents the value of the world that C ultimately picks, and so on for every greater unit of value that is added to a value n that is initially randomly picked.

It may appear that if a world-actualizer like God were in this predicament of increasingly better two-step processes, then this world-actualizer would become paralyzed, not in the sense that there would be no actual world, but rather that such a world-actualizer would not bring into existence any contingent beings, and so the actual world would be one devoid of any contingent beings. However, in response to Kraay, the theist might suggest that God’s two-step process is to be constrained by what I shall call energizing reasons for action. A reason is energizing if and only if it is not a paralyzing reason. A reason for action r is paralyzing if there is a similar reason for action r* that outweighs r, and acting on r* precludes acting on r, and there is yet another similar reason r** that outweighs r*, and acting on r** precludes acting on either r or r*, and this chain of increasingly stronger outweighing and contravening reasons is endless. Here is an example of a paralyzing reason under the assumption that there is no best possible world: world w* is better than world w. This reason to actualize w* is paralyzing because it is outweighed by a similar reason for action that precludes actualizing w* (and precludes actualizing w): world w** is better than world w*, and this chain of increasingly stronger outweighing and contravening reasons is endless.

Given that there is no best world, all reasons for actualizing a world that appeal merely to that world’s greater value in comparison to the value of some other world, without appealing to some axiologically relevant feature of either world, will count as a paralyzing reason for action. Thus, in response to Kraay, the Howard-Snyders might assert only the weaker claim that the two-step process for God’s actualizing a world must be unsurpassable with respect to its support from the strongest group of energizing reasons for action. Indeed, it seems that God has an energizing reason to actualize a world in which God creates contingent beings that have intrinsic value. That reason is that, ceteris paribus, a world in which God creates contingent beings that have intrinsic value is better than a world in which God does not create such beings. There is no relevantly similar outweighing and contravening reason for God to do anything other than actualize a world in which God creates contingent beings that have intrinsic value.

To see this response to P1 in more detail, let us begin by considering the second step of the two-step process, the step that involves randomly picking an acceptable world. It appears that God has no energizing reason to actualize one world rather than another among the set of acceptable worlds, and so God must randomly pick a world to actualize among the set of acceptable worlds. Therefore, the second step of randomly picking a world to actualize among the set of acceptable worlds is unsurpassable insofar as there is no energizing reason for a world-actualizer like God to do otherwise. Now consider the first step of the two-step process, the step that involves partitioning all possible worlds into the acceptable and unacceptable ones, respectively. Given Kraay’s plausible claim that there are infinitely many points at which God may partition the set of possible worlds, any reason to select one partition rather than another simply in virtue of the former being better than the latter appears to be paralyzing. But there may be energizing reasons to select certain partitions. For instance, consider the following examples of sorting principles that the Howard-Snyders (Howard-Snyder and Howard-Snyder Reference Howard-Snyder and Howard-Snyder1996, 424) offer:

  1. (i) No world in which beings live lives that are not worth living is acceptable.

  2. (ii) No world in which beings experience gratuitous suffering is acceptable.

  3. (iii) No world in which beings live lives which are not as happy and fulfilled as those lives could possibly be is acceptable.

  4. (iv) No world empty of sentient, rational beings is acceptable.

Does God have an energizing reason to adopt these sorting principles? Perhaps God’s reason for adopting sorting principle (i) is that, ceteris paribus, a world with lives that are worth living is better than a world in which it is not the case that there are lives that are worth living. God’s reason for adopting sorting principle (ii) may be that, ceteris paribus, a world with beings that do not experience gratuitous suffering is better than a world in which it is not the case that there are beings that do not experience gratuitous suffering. God’s reason for adopting sorting principle (iv) may be that, ceteris paribus, a world with sentient, rational beings is better than a world in which it is not the case that there are sentient, rational beings. All these reasons appear to be energizing. However, consider God’s purported reason for adopting sorting principle (iii): ceteris paribus, a world in which beings live lives which are as happy and fulfilled as those lives could possibly be is better than a world in which it is not the case that beings live lives which are as happy and fulfilled as those lives could possibly be. Perhaps this reason, if it exists, is also energizing. However, just as there is no upper limit to the value of worlds, there is plausibly no upper limit to the happiness and fulfillment of a life. If that is right, then the putative reason in favour of sorting principle (iii) simply does not exist.

We have seen that God may partition the acceptable and unacceptable worlds at least partly on the basis of sorting principles (i), (ii), and (iv) since the reasons for making a partition based on these sorting principles are energizing. If God can partition all worlds based on these sorting principles and any other relevant sorting principles that are also supported by energizing reasons, then we can say that God’s partition is unsurpassable insofar as it is supported by the strongest group of energizing reasons. If this is right, then P1 is false, and its apparent truth is due to a different true claim, viz. that if it is possible for the product of a world-actualizing action performed by some being to have been better by acting on some energizing reason that was not acted on, then, ceteris paribus, it is possible for that being’s action to have been (morally or rationally) better.

If the theist responds to the No Best World Argument in the manner that I’ve suggested, then at least some of the acceptable worlds may be thought to contain evil, in which case premise (5*) of the Judicious Argument from Evil is not necessarily true. After all, for any acceptable world without evil that God actualizes that has some net value, God might instead actualize an acceptable world with the same net value but that nevertheless contains evil. For instance, God might actualize an acceptable world with 100 hedons and no dolors, or God might actualize a similar acceptable world but with 200 hedons and 100 dolors. Either way, the net value of the world is the same, and so premise (5*) is not necessarily true given that God may be expected to actualize either world. This is how the No Best World Argument is relevant to the Judicious Argument from Evil; the theist’s response to the No Best World Argument constitutes a ‘No Best World’ Objection to the necessary truth of (5*). However, it is at this point that Mackie’s mantra that an omnibenevolent being is opposed to evil invites a counter-response. Just as God has energizing reasons to partition the acceptable and unacceptable worlds at least partly based on sorting principles (i), (ii), and (iv), God may also have an energizing reason to make a partition at least partly based on the following sorting principle:

  1. (v) No world that contains evil is acceptable.

The energizing reason for God to adopt sorting principle (v) is that, ceteris paribus, a world without evil is better than a world with evil. Marilyn McCord Adams (Reference Adams1999) might reject this energizing reason and replace (v) with a sorting principle that says instead that no world that contains undefeated evil is acceptable because, ceteris paribus, a world without undefeated evil is better than a world with undefeated evil. An evil is defeated when it is not just outweighed by some good, but rather the evil contributes to a good that outweighs that evil (Chisholm Reference Chisholm1968). Adams (Reference Adams1999, 167) proposes a theological form of defeat that involves the good of God’s empathetically identifying with the victim, such that the horrendous evil that the victim undergoes is defeated by this good for the victim: ‘by virtue of endowing horror with a good aspect, Divine identification makes the victim’s experience of horrors so meaningful that she would not retrospectively wish it away’. But this position suggests that a worse horror improves the value of divine identification, in which case not only would the victim not retrospectively wish the horror away, but rather the victim would retrospectively wish that their past horrors were worse than what they, in fact, were, which is surely doubtful if the victim is fully rational. Adams (Reference Adams1999, 167) might attempt to resist this suggestion in virtue of her claim that the incommensurate good of beatific intimacy with God does not require horrific evil. But then it becomes unclear as to why a victim would not retrospectively wish away a past horror if they could have attained beatific intimacy without this past horror (Pereboom Reference Pereboom2021, 65). I thus conclude that God’s energizing reasons for making a partition imply that every world that contains evil is unacceptable, and so premise (5*) of the Judicious Argument from Evil is, in fact, necessarily true. So much for P1.

Let’s now turn our attention to P2. Hasker’s (Reference Hasker2005) case against P2 relies on the following claim, which we’ll call ‘H’:

  1. H: If, necessarily, I fail to do better than I actually did, then failing to do better than I actually did is not a moral fault.

Hasker (Reference Hasker2005) argues for H by highlighting its similarity to another principle that is very plausible, which we’ll call ‘H*’.

  1. H*: If, necessarily, I fail to do the best I can, then failing to do the best I can is not a moral fault.

Given the plausibility of H* and its similarity to H, we should likewise accept H. However, Kraay (Reference Kraay, Draper, Talliaferro and Quinn2010, 488) rightly points out that, given that there is no best world, it is logically impossible for God to do his best. By contrast, it is logically possible for God to do better than he does. Therefore, in virtue of this difference between these principles, Hasker has not shown that the plausibility of H* transfers to the plausibility of H. But there is another principle in the moral responsibility, viz. Otsuka’s (Reference Otsuka1998, 688) Principle of Avoidable Blame, a variant of which may serve the purpose of Hasker’s H:

Principle of Avoidable Fault: An act of a given type is a moral fault for an agent only if they could instead have behaved in such a manner that that behaviour would not have been a moral fault.

Given that there is no best possible world, a world-actualizer like God cannot avoid actualizing a world such that there is a better world that that world-actualizer could have actualized instead. And since a world-actualizer like God cannot avoid performing an act of this type (even God’s refraining from creating any contingent beings still ‘results’ in some improvable world being actual), the Principle of Avoidable Fault implies that a world-actualizer’s act of this type is not a moral fault. And if that is the case, then we might think that God could not have been better, even though God’s world-actualizing action could have been better. So, given that the Principle of Avoidable Fault is more plausible than P2, it is P2 that must go.

A source incompatibilist such as Zagzebski (Reference Zagzebski2000) might be inclined to give up the Principle of Avoidable Fault rather than P2 due to Frankfurt cases (Frankfurt Reference Frankfurt1969). In that case, so much the worse for one way of resisting the No Best World argument for atheism. At any rate, while delving into Frankfurt cases goes beyond the scope of this article, we can at least agree that theists who are leeway incompatibilists such as Van Inwagen (Reference Van Inwagen1983, 161–182) and presumably Plantinga are likely to find the Principle of Avoidable Fault plausible. Moreover, granting this principle does not appear to stack the deck against theism within the context of this dialectic concerning the No Best World Argument for atheism. Rather, accepting this principle provides a way for the theist to resist this a priori argument for atheism.

However – and here’s the problem – there are surely some unacceptable worlds, to use the terminology of the Howard-Snyders, and that are such that a world-actualizer like God would be blameworthy, or morally or rationally criticizable, if that world-actualizer were to actualize one of them. Perhaps, following the previously mentioned energizing reasons that support sorting principles (i), (ii), and (iv), such a world-actualizer would be blameworthy if they were to actualize a world in which beings live lives that are not worth living, or a world in which beings experience gratuitous suffering, or a world that is devoid of sentient, rational beings. Such blameworthiness is compatible with the Principle of Avoidable Fault since this world-actualizer can avoid blameworthiness by actualizing a world that contains sentient, rational beings whose lives are worth living and who do not experience gratuitous suffering. This response to P2 might lead us to think that some acceptable worlds nevertheless contain evil, in which case premise (5*) of the Judicious Argument from Evil is possibly false. However, as previously discussed regarding P1, it also seems plausible that a world-actualizer like God would be blameworthy if they did not act in accordance with sorting principle (v) that is similarly supported by the aforementioned energizing reason. In other words, it also seems that such a world-actualizer would be blameworthy, or morally or rationally criticizable, if they were to actualize a world that contains evil. This verdict is consistent with the Principle of Avoidable Fault since this world-actualizer can avoid such blameworthiness by actualizing a world that contains no evil.

I conclude that my critique of the No Best World Argument on behalf of the Howard-Snyders and Hasker is, if successful, a Pyrrhic victory for the theist because the rejection of either P1 or P2 appears, in fact, to strengthen the case for the necessary truth of (5*) given that the theist’s most viable rebuttal to the No Best World argument employs the distinction between paralyzing and energizing reasons for action, and there is an energizing reason for God to actualize a world without evil. This concludes my defence of the Judicious Argument from Evil. I now turn my attention to developing the Inexorable Argument from Evil.

Mackie’s value symmetry

Recall that, according to Mackie’s value symmetry, for many higher-order goods invoked by the theist as a potential justification for God’s permission of evil, the atheist can invoke a contrasting higher-order evil that exists. To illustrate, consider Mackie’s response to theodicies that say that the first-order evils of pain and disease are required for certain second-order goods such as benevolence, sympathy, and heroism.

Our analysis shows clearly the possibility of the existence of a second order evil, an evil (2) contrasting with good (2) as evil (1) contrasts with good (1). This would include malevolence, cruelty, callousness, cowardice, and states in which good (1) is decreasing and evil (1) increasing … so evil (2) will, by analogy, be the important kind of evil, the kind which God, if he were wholly good and omnipotent would eliminate. And yet evil (2) plainly exists … (Mackie Reference Mackie1955, 207–208).

Mackie considers this response to the above theodicy, which is his third response, to be uniquely fatal. So, it is safe to say that Mackie’s value symmetry plays a pivotal role in his Logical Argument from Evil. Indeed, Mackie goes a step further by applying his value symmetry to a theodicy that invokes the third-order good of an increase in second-order goods (see also Mackie Reference Mackie1962, 154–155).

[W]e could hardly say that the really important good was a good (3), such as the increase of benevolence in proportion to cruelty, which logically required for its occurrence the occurrence of some second order evil. But even if evil (2) could be explained in this way, it is fairly clear that there would be third order evils contrasting with this third order good: and we should be well on the way to an infinite regress, where the solution of a problem of evil, stated in terms of evil (n), indicated in the existence of an evil (n + 1), and a further problem to be solved (Mackie Reference Mackie1955, 208).

Although Mackie does not elaborate on the notion of a contrasting value, we may propose that two values contrast with each other only if they share some important, similar characteristics. For example, pain and pleasure are sensations. Benevolence and malevolence are character traits. An increase in second-order goods and evils, respectively, is just that – an increase in value or disvalue over time. Shelly Kagan (Reference Kagan2014, 275) has independently explored a similar hypothesis, the symmetry thesis, according to which each element of an objective list theory of well-being has an opposite element within an objective list theory of ill-being. I share Kagan’s sentiment that the attraction of this symmetry thesis is due to the fact that pleasure and pain are opposite elements of well-being and ill-being, respectively, according to both hedonism and most objective list theories of well-being (and ill-being).

Now, we may wonder at this point in what way a contrasting evil is ‘a further problem to be solved’. After all, if the theist can highlight a possibility in which the best feasible world for God is one in which there are more instances of good of some order than some instances of evil of the same order, we might think that God would then be justified in actualizing a world with such evil, thus demonstrating that premise (7) is not a necessary truth. But at no point in this discussion does Mackie (Reference Mackie1955) seem to consider a quantitative difference between such goods and evils to bear upon that status of premise (7), although he does acknowledge this issue to some extent in Mackie (Reference Mackie and Williams1982, 154). In any case, Mackie seems to concede to the theist that the higher-order good of free will is, in fact, immune from Mackie’s symmetry strategy of invoking a contrasting higher-order evil that becomes ‘a further problem to be solved’ because at no point does Mackie consider a higher-order evil that contrasts with the higher-order good of free will. Here is what Mackie says.

[E]vil is not to be ascribed to God at all, but to the independent actions of human beings, supposed to have been endowed by God with freedom of the will. This solution may be combined with the preceding one: first order evil (e.g. pain) may be justified as a logically necessary component in second order good (e.g. sympathy) while second order evil (e.g. cruelty) is not justified but is so ascribed to human beings that God cannot be held responsible for it. This combination evades my third criticism of the preceding solution (Mackie Reference Mackie1955, 208).

When Mackie elaborates on this theodicy, he entertains the idea that freedom is a third-order good that renders second-order goods more valuable than they would be if they were deterministically produced. Strangely, however, Mackie still does not attempt to apply his symmetry strategy to a theodicy that invokes the higher-order value of free will. This omission seems to represent a concession to the theist that a free will theodicy – and, a fortiori, the Free Will Defence – is not susceptible to a corresponding third-order evil. Nor does Mackie explain in further detail why a free will theodicy is immune from his symmetry strategy. Instead, Mackie (Reference Mackie1955, 208–212; Reference Mackie1962; Reference Mackie and Williams1982, 166–168) defends compatibilism, partly by critiquing libertarianism, to sidestep this theodicy entirely as previously described; God guarantees the elimination of all moral evil by deterministically causing everyone’s obligatory free actions. But Mackie should have considered applying his symmetry strategy to the relationship between free will and value rather than resorting to compatibilism, as we will do now.

The evil of impermissible free actions

Plantinga’s emphasis on more morally good free actions than morally bad free actions in his Free Will Defence suggests that he thinks that obligatory free actions have positive intrinsic value. So, let us explore this option further from the perspective of Mackie’s value symmetry and thus assume that an impermissible free action is a higher-order evil that contrasts with the higher-order good of an obligatory free action. In that case, just as an obligatory free action is better than a similar action that is not free, so too an impermissible free action is worse than a similar action that is not free. One way to articulate this idea is that an action’s being free ‘amplifies’ the (positive or negative) value of that action. Laura Ekstrom (Reference Ekstrom2021, 51) shares this intuition concerning the more significant disvalue of an impermissible free action by considering a variety of such examples that conform to this intuition.

It is not entirely clear whether Swinburne agrees with this amplification approach to the (positive or negative) value of a free action. He does seem to think that an obligatory free action is better than a similar action that is not free (Swinburne Reference Swinburne1998, 84). Rather than amplifying the value of an action, however, Swinburne may think instead that an action’s being free is valuable, and that this value is added either to the positive value of a good action or to the negative value of a bad action which in turn depends on the goodness or badness of its effects (Swinburne Reference Swinburne1998, 93). On this additive approach, however, a question now arises as to whether it is possible for the net (intrinsic) value of an impermissible free action to be positive at least partly in virtue of the positive value of that action’s being free outweighing the negative value of the impermissibility or badness of that action. This possibility suggests that an increase in the number of certain impermissible free actions improves the world, ceteris paribus. But this does not seem particularly plausible (cf. Pereboom Reference Pereboom2021, 161–162); if free actions in and of themselves make any evaluative contribution to the world, then, ceteris paribus, impermissible free actions of any kind decrease the world’s net value. The amplification approach to the value of a free action thus seems to be more plausible than the additive approach.

Given the amplification approach, Mackie could employ his symmetry strategy to assert that the evil of an impermissible free action becomes ‘a further problem to be solved’ in the exact same way in which the aforementioned second-order evils become a further problem to be solved. However, even if an impermissible free action is an evil that contrasts with the goodness of an obligatory free action, I agree with Plantinga that God cannot deterministically cause free actions, in which case the best feasible world for God may contain more obligatory free actions than impermissible free actions. So, Mackie’s value symmetry alone does not suffice to undermine the Free Will Defence, and thus premise (7) has not been shown to be necessarily true. Nevertheless, as I aim to show in the next section, if we either accept a different principle that may be attributable to Mackie, viz. Mackie’s negative consequentialism, or conjoin Mackie’s value symmetry with a restricted version of Mackie’s negative consequentialism, then we can construct another logical argument from evil that circumvents the Free Will Defence and that does not rely on the assumptions employed in the Judicious Argument from Evil concerning the value of morally innocent free actions.

Mackie’s negative consequentialism

According to a normative ethical position termed but criticized by R.N. Smart (Reference Smart1958), viz. negative utilitarianism, an action is wrong just in case that action’s consequence contains more pain than the pain contained in the consequence of any other performable action.Footnote 12 I suggest that there is textual evidence that Mackie was inclined towards a view that is similar in spirit to negative utilitarianism. According to this view, which I shall call Mackie’s negative consequentialism, God, an omnibenevolent being who is opposed to evil, has most reason, all things considered, to eliminate every evil, even at the cost of eliminating any number of goods.

In his second critique of the aforementioned second-order goods theodicy, Mackie says that it follows from this theodicy that ‘God is not in our sense benevolent or sympathetic: he is not concerned to minimize evil (1), but only to promote good (2); and this might be a disturbing conclusion for some theists’ (Mackie Reference Mackie1955, 207). Mackie seems to think that some theists find this conclusion disturbing because the notion of benevolence that Mackie has in mind, a notion that emphasizes an opposition to evil, is one in which there is a kind of lexical priority (in the traditional Rawlsian sense) of the eradication of evil over the promotion of goodness. In support of this interpretation, consider Mackie’s (Reference Mackie1977) later work in which he criticizes normative ethical theories that cohere with the universalizability of moral judgments but that nevertheless allow for significant discrepancies in well-being.

Suppose that one of a pair of alternative sets of principles was likely to make most members of the society much happier than they would otherwise be, but at the price of making a few very unhappy, but not because either party deserved in any way this differential prosperity. As rational egoists, choosing from behind the veil of ignorance, we might well choose this set of principles…But this set of principles would not be one that we could endorse, even as an acceptable compromise…Indeed, we would not call them fair…Principles which it is egoistically rational to choose even from behind the equalizing veil of ignorance need not be intrinsically fair in their operation…a similar criticism can be made of any utilitarianism which takes only a quantitative account of interests or of happiness, and so allows the happiness of some to outweigh the undeserved misery of others (Mackie Reference Mackie1977, 95–96).

Mackie seems to think that a plausible normative ethical theory must be intrinsically fair and that no normative ethical theory that allows significant intrinsic evil to befall one person merely for someone else to enjoy an even more significant intrinsic good is intrinsically fair. Similarly, Mackie (Reference Mackie1977, 127, 145) critiques act utilitarianism on the grounds that it only takes into account an aggregative principle of well-being, but not a distributive principle of well-being, through the following question: ‘Is a state of affairs in which one person is supremely happy and nine are miserable better than one in which all ten are equally happy, provided only that the total balance of happiness is greater?’ (Mackie Reference Mackie1977, 127). This critical question is particularly forceful when considering an influential objection to utilitarianism that appeared in publication just a few years before Mackie’s (Reference Mackie1977) publication, viz. Nozick’s (Reference Nozick1974, 41) utility monster objection, which accuses utilitarianism of wrongly implying that a utility monster, someone who gets ‘enormously greater gains of utility from any sacrifice of others than these others lose’, is morally permitted to attain such gains even at the cost of others being harmed. Negative utilitarianism, by contrast, completely aligns with the commonsensical intuition that it is morally wrong for the utility monster to harm others merely for the sake of the utility monster’s benefit of acquiring additional intrinsic value.

Richard Yetter Chappell (Reference Chappell2021) attempts to undercut the force of Nozick’s utility monster objection by considering a variant of Nozick’s case in which the utility monster starts from a baseline status of enormous suffering. Under these circumstances, many of us probably have the intuition that it is, in fact, morally permissible for the utility monster to attain such gains at the cost of others who do not start from a baseline status of enormous suffering. Notice, however, that negative utilitarianism can accommodate our intuitions both in Nozick’s original case and in Chappell’s variant of the case because the gains of the utility monster in Chappell’s case are to be thought in terms of relief, that is, the removal of suffering, and that is precisely what should take precedence according to negative utilitarianism. This normative ethical theory should thus be taken seriously in virtue of its ability to accommodate our intuitions both in Nozick’s thought experiment and in Chappell’s variant of this thought experiment.

We should similarly take seriously the idea that, following Mackie’s negative consequentialism, God has most reason, all things considered, to eliminate all evil, despite the loss of any amount of goodness, and, a fortiori, despite the loss of goodness from obligatory free actions. This approach allows us to generate a logical argument from evil that is simpler than Plantinga’s construction of Mackie’s Logical Argument from Evil by replacing (5) and (7) with the following simpler propositions.

  1. 5’. A being that is omnipotent, omniscient, and omnibenevolent eliminates every evil that it can eliminate.

  2. 7’. If God is omniscient and omnipotent, then he can eliminate every evil state of affairs.

Call this the Simple Logical Argument from Evil. Alternatively, if we combine a restricted version of Mackie’s negative consequentialism with Mackie’s value symmetry, then we can adopt the more modest position that God is only primarily opposed to ‘the important kind of evil’, which is the evil contrasting with the higher-order good of an obligatory free action, viz. an impermissible free action. Given this normative ethical position, Mackie can construct a revised version of the logical argument from evil that is immune from Plantinga’s Free Will Defence even given the assumption that the value of obligatory free actions is lexically prior to the value of morally innocent free actions. This argument employs a different technical notion of elimination: a being inexorably eliminates an evil state of affairs if it eliminates that evil without bringing about either a worse evil or an equally weighty evil.

The Inexorable Argument from Evil

  1. 1. God is omnipotent.

  2. 2. God is omnibenevolent.

  3. 3. God is omniscient.

  4. 4*. Impermissible free actions exist.

  5. 5**. A being that is omnipotent, omniscient, and omnibenevolent eliminates every impermissible free action that it can inexorably eliminate.

  6. 6. There are no limits to what an omnipotent being can do.

  7. 7**. If God is omniscient and omnipotent, then he can inexorably eliminate every impermissible free action.

Plantinga’s Free Will Defence does not appear to undermine the necessary truth of either (5**) or (7**). Regarding the latter, an omnipotent creator clearly can actualize a world with a positive net value and no evil, regardless of whether every possible person contingently suffers from transworld depravity. For God can actualize a world with goods that do not give rise to evil, even if the (positive) value of free will is greater than these goods. A fortiori, God can actualize a world with a positive net value and no impermissible free actions, even if everyone suffers from transworld depravity. So, it is necessary that God can inexorably eliminate every impermissible free action.

Let’s now turn to premise (5**). According to my interpretation of Mackie’s normative ethical view, God – a being who is inexorably opposed to impermissible free actions – eliminates every such action even at the cost of eliminating any number of goods. So, even if the value of obligatory free actions is lexically prior to the value of morally innocent free actions, Mackie’s negative consequentialism, or a restricted version of Mackie’s negative consequentialism in conjunction with Mackie’s value symmetry, seems to preserve the necessary truth of (5**); God has most reason, all things considered, to actualize a world with a positive net-value but without significantly free creatures, at least if the creation of significantly free creatures would result in the existence of even one impermissible free action, an action towards which God is inexorably opposed.

The Inexorable Argument from Evil, if sound, does not show that God’s existence is incompatible with evil tout court, but its commitment to the existence of the evil of impermissible free acts does not seem much stronger than a commitment to the existence of evil tout court, particularly given the assumption that agents have morally significant freedom. Moreover, if Mackie’s negative consequentialism is true in its unrestricted form, then the Simple Logical Argument from Evil appears to be sound, and that argument does show that God’s existence is incompatible with evil tout court.

Conclusion

The arguments developed in this article effectively bypass Plantinga’s Free Will Defence in virtue of employing premises that are consistent with the doctrine of transworld depravity, Molinism, libertarianism, and the position that significant freedom is immensely valuable. An exhaustive defence of these arguments, which goes beyond the scope of this article, would require addressing the complete inventory of theodicies that attempt to provide plausible justifications for God’s permission of evil. Future research into Mackie’s ethical perspectives, and specifically Mackie’s value symmetry, may challenge some of these theodicies, particularly ones that appeal to the intrinsic or extrinsic value of free will, as I will now illustrate with two examples. According to Hick’s (Reference Hick1978, 276; Reference Hick and Davis2001) Irenaean soul-making theodicy, evil is required for the valuable process of moral and spiritual growth that, crucially, consists of morally significant free actions, and the value of that process outweighs the evils required for that process. Mackie may apply his symmetry strategy to Hick’s soul-making theodicy by highlighting the negative value of a process of freely performed soul-destroying actions, contrasting with the positive value of a soul-making process (cf. Phillips Reference Phillips2004, 60–67). Next, Rasmussen (Reference Rasmussen2013: 424–428) maintains that significant freedom (or, as he puts it, moral freedom) is necessary for the good of loving situations between God and creatures. Mackie can employ his symmetry strategy to suggest that the evil of a genuine adversarial situation with God is a contrasting evil that similarly requires significant freedom. Thus, in response to Hick, Rasmussen, and others, Mackie can invoke contrasting evils that become, as he would say, further problems to be solved. Whether these problems can be resolved remains a subject for future philosophical exploration.

Acknowledgements

For valuable feedback on this article, I am very grateful to Klaas Kraay, Ronald Lewis, Derk Pereboom, Holly Smith, Travis Timmerman, participants at the March 2025 Toronto Philosophy of Religion Working Group at Toronto Metropolitan University, participants at the SPAWN 2025 conference at Syracuse University, and the anonymous referees for this journal.

Footnotes

1. The very idea of elimination seems to suggest that that which is eliminated once existed. But this is not the sort of predicament that God is in with respect to evil. Rather, God must decide whether to allow an evil state of affairs to exist in the first place, in which case it may be better to use the concept of prevention or omission. Nevertheless, I will continue to use Mackie’s and Plantinga’s notion of elimination for simplicity’s sake. Also, since Mackie needs (5) to be necessarily true, then, in order to be maximally charitable to Mackie, Plantinga probably needs to revise his definition of proper elimination in the following manner: a being properly eliminates an evil state of affairs if it eliminates that evil (i) without eliminating either an outweighing good or an equally counterbalancing good, and (ii) without bringing about either a worse evil or an equally weighty evil (when Plantinga says ‘a greater evil’, I take him to mean a more significant evil, which I think should be described as a worse evil).

2. A defence in this context is an attempt to show that God and evil are compatible by constructing either a metaphysically possible story (Plantinga Reference Plantinga1974a, 28), or a story that is true, for all anyone knows (Van Inwagen Reference Van Inwagen2006, 66–71), or an attempt to show that it is likely that there are reasons, whatever they may be, that would justify God’s permission of evil if God were to exist (Tooley Reference Tooley and Zalta2015, 53). A theodicy, by contrast, attempts the more ambitious project of identifying the reasons that are likely to exist that would justify God. My primary concern is to engage with Plantinga’s minimal conception of a defence.

3. Plantinga’s Free Will Defence also assumes the truth of Molinism, which I shall understand as the view that there are true future-oriented counterfactuals of freedom concerning what someone would freely do in the future if they were in certain circumstances. For a presentation of a free will defence that does not presuppose Molinism, see Perszyk (Reference Perszyk1998). While I am inclined to reject Molinism on the grounds that there are no brute contingent truths, nothing argued for in this article presupposes that Molinism is false. For a discussion of the role that Molinism plays specifically in Plantinga’s Free Will Defence, as well as in some puzzles that arise from this defence, see e.g. Howard-Snyder and Hawthorne (Reference Howard-Snyder and Hawthorne1998), Sobel (Reference Sobel2004, 456–460), Otte (Reference Otte2009), Plantinga (Reference Plantinga2009), Almeida (Reference Almeida2012, 101–134), and Howard-Snyder (Reference Howard-Snyder, McBrayer and Howard-Snyder2014).

4. But see Turner (Reference Turner2013) who argues that even if compatibilism is true, God may not be able to create significantly free creatures in a deterministic universe who never go wrong with respect to the exercise of their freedom because freedom requires God’s not considering what such creatures will do. I would argue that the best explanation for why God’s foreknowledge (or timeless knowledge) is incompatible with freedom is that freedom requires an open future according to which there are no true future contingents, and such an open future precludes the possibility of foreknowledge (or timeless knowledge) of future free actions. Though, of course, this response is unavailable to the Molinist.

5. Here is Plantinga’s definition of an essence: ‘An essence simpliciter is a property P such that there is a world W in which there exists an object x that has P essentially and is such that in no world W* is there an object that has P and is distinct from x’ (Plantinga Reference Plantinga1974b, 187). A creaturely essence is an essence entailing is created by God (Plantinga Reference Plantinga1974b, 188, fn. 1).

6. Recall footnote 1 in which I said that although I am using Mackie’s and Plantinga’s preferred term, ‘elimination’, rather than either ‘prevention’ or ‘omission’, God’s eliminating some evil does not imply that that evil once existed.

7. But see Matheson’s (Reference Matheson2018) position according to which the theist should say that heavenly freedom preserves the ability to sin precisely due to the value of significant freedom according to the Free Will Defence.

8. Nagasawa’s (Reference Nagasawa2008) Maximal God thesis might be deployed to coherently maintain that the morally significant freedom of free creatures is indeed more valuable than God’s freedom – a freedom that is, in a sense, limited in virtue of God’s inability to perform morally impermissible actions – notwithstanding the fact that God is that than which none greater can be conceived.

9. Perhaps we also need to ‘add’ agent causation to render the action free. Nothing argued for here hinges on this issue.

10. Here is a rough account of the distinction between weak and strong actualization. God strongly actualizes only what God can cause to be actual, or only what God can deterministically cause to be actual. Since God cannot cause, or at least cannot deterministically cause, a free action, God cannot strongly actualize a state of affairs that contains a free action. By contrast, God can weakly actualize a state of affairs that contains a free action by creating the circumstances under which that free action may occur (cf. Plantinga Reference Plantinga1974b, 173).

11. A key structural difference between Plantinga’s (Reference Plantinga1974b, 186, 188) definitions of transworld depravity on the one hand and my definition of transworld universal suboptimality on the other hand is this. If an essence suffers from transworld depravity, then that essence’s instantiation would perform at least one morally impermissible free act if that essence’s instantiation were to possess morally significant freedom. By contrast, if an essence suffers from transworld universal suboptimality, then that essence’s instantiation would perform only morally innocent free actions that are permissibly suboptimal if, and whenever, it were to be the case that that essence’s instantiation possesses an ability to perform either a morally innocent free action that is permissibly suboptimal or a morally innocent free action that is supererogatory.

12. See Knutsson’s (Reference Knutsson2019) response to Smart’s world destruction argument against negative utilitarianism.

References

Adams, MM (1999) Horrendous Evils and the Goodness of God. Ithaca, NY: Cornell University Press.Google Scholar
Almeida, MJ (2012) Freedom, God, and Worlds. Oxford: Oxford University Press.10.1093/acprof:oso/9780199640027.001.0001CrossRefGoogle Scholar
Anselm, (1076/1996) Monologion and Proslogion. Williams, T (ed. and trans). Indianapolis: Hackett.Google Scholar
Augustine (427/1972) City of God. Bettenson, H (trans). London: Penguin Classics.Google Scholar
Berker, S (2019) Mackie was not an error theorist. Philosophical Perspectives 33, 525.10.1111/phpe.12123CrossRefGoogle Scholar
Bradley, B (2009) Well-being and Death. New York: Oxford University Press.10.1093/acprof:oso/9780199557967.001.1CrossRefGoogle Scholar
Chang, R (2015) Value incomparability and incommensurability. In Hirose, I, and Olson, J (eds), The Oxford Handbook of Value Theory. New York NY: Oxford University Press USA, 205224.Google Scholar
Chappell, RY (2021) Negative utility monsters. Utilitas 33, 417421.10.1017/S0953820821000169CrossRefGoogle Scholar
Chisholm, R (1968) The defeat of good and evil. Proceedings and Addresses of the American Philosophical Association 42, 2138.10.2307/3130021CrossRefGoogle Scholar
Ekstrom, LW (2021) God, Suffering, and the Value of Free Will. Oxford: Oxford University Press.10.1093/oso/9780197556412.001.0001CrossRefGoogle Scholar
Frankfurt, HG (1969) Alternate Possibilities and Moral Responsibility. Journal of Philosophy 66, 829839.10.2307/2023833CrossRefGoogle Scholar
Hasker, W (2005) Can God be free?: Rowe’s dilemma for theology. Religious Studies 41, 453462.10.1017/S0034412505007985CrossRefGoogle Scholar
Hick, J (1978) Evil and the God of Love, 2nd edn. San Francisco: Harper.Google Scholar
Hick, J (2001) An Irenaean theodicy. In Davis, ST (ed), Encountering Evil. 2nd edn: Louisville, KY: Westminster John Knox Press, 3952.Google Scholar
Howard-Snyder, D (2014) The logical problem of evil: Mackie and Plantinga. In McBrayer, JP, and Howard-Snyder, D (eds), The Blackwell Companion to the Problem of Evil. Malden, MA: Wiley, 1933.Google Scholar
Howard-Snyder, D and Hawthorne, J (1998) Transworld sanctity and Plantinga’s free will defense. International Journal for Philosophy of Religion 44, 121.10.1023/A:1003210017171CrossRefGoogle Scholar
Howard-Snyder, D and Howard-Snyder, F (1994) How an unsurpassable being can create a surpassable world. Faith and Philosophy 11, 260268.10.5840/faithphil199411222CrossRefGoogle Scholar
Howard-Snyder, D and Howard-Snyder, F (1996) The real problem of no best world. Faith and Philosophy 13, 422425.10.5840/faithphil199613333CrossRefGoogle Scholar
Kagan, S (2014) An introduction to ill-being. Oxford Studies in Normative Ethics 4, 261288.10.1093/acprof:oso/9780198722144.003.0012CrossRefGoogle Scholar
Kane, R (1996) The Significance of Free Will. New York, US: Oxford University Press USA.Google Scholar
Knutsson, S (2019) The world destruction argument. Inquiry 64, 10041023.10.1080/0020174X.2019.1658631CrossRefGoogle Scholar
Kraay, KJ (2010) The problem of no best world. In Draper, P, Talliaferro, C and Quinn, PL (eds), A Companion to Philosophy of Religion, 2nd Ed. Malden, MA: Wiley-Blackwell, 482491.10.1002/9781444320152.ch57CrossRefGoogle Scholar
Kraay, KJ (2011) Incommensurability, incomparability, and God’s choice of a world. International Journal for Philosophy of Religion 69, 91102.10.1007/s11153-010-9266-1CrossRefGoogle Scholar
Leibniz, G (1710/1985). In Austin Farrer, (ed.), Theodicy: Essays on the Goodness of God, the Freedom of Man, and the Origin of Evil. La Salle, IL: Open Court.Google Scholar
Lewis, DK (1993) Evil for freedom’s sake. Philosophical Papers 22, 149172.10.1080/05568649309506401CrossRefGoogle Scholar
Mackie, JL (1955) Evil and omnipotence. Mind 64, 200212.10.1093/mind/LXIV.254.200CrossRefGoogle Scholar
Mackie, JL (1962) Theism and utopia. Philosophy 37, 153158.10.1017/S0031819100036810CrossRefGoogle Scholar
Mackie, JL (1977) Ethics: Inventing Right and Wrong. New York: Penguin Books.Google Scholar
Mackie, JL (1982) The Miracle of Theism: Arguments for and against the Existence of God. Williams, B (ed), New York: Oxford University Press.Google Scholar
Matheson, B (2018) Tracing and heavenly freedom. International Journal for Philosophy of Religion 84, 5769.10.1007/s11153-017-9643-0CrossRefGoogle Scholar
McNamara, P (2010) Supererogation, inside and out: Toward an adequate scheme for common sense morality. In Timmons, M, Oxford Studies in Normative Ethics, Volume I. Oxford: Oxford University Press. 202235.Google Scholar
McNamara, P and De Putte, FV (2022) Deontic logic. In The Stanford Encyclopedia of Philosophy (Fall 2022 Edition). URL Zalta, EN, and Nodelman, U (eds 1118). <https://plato.stanford.edu/archives/fall2022/entries/logic-deontic/>.Google Scholar
Nagasawa, Y (2008) A new defence of Anselmian theism. Philosophical Quarterly 58, 577596.10.1111/j.1467-9213.2008.578.xCrossRefGoogle Scholar
Nelkin, DK (2011) Making Sense of Freedom and Responsibility. New York: Oxford University Press.10.1093/acprof:oso/9780199608560.001.0001CrossRefGoogle Scholar
Nozick, R (1974) Anarchy, State, and Utopia. New York: Basic Books.Google Scholar
Otsuka, M (1998) Incompatibilism and the avoidability of blame. Ethics 108, 685701.10.1086/233847CrossRefGoogle Scholar
Otte, R (2009) Transworld depravity and unobtainable worlds. Philosophy and Phenomenological Research 78, 165177.10.1111/j.1933-1592.2008.00236.xCrossRefGoogle Scholar
Pereboom, D (2014) Free Will, Agency, and Meaning in Life. Oxford: Oxford University Press.10.1093/acprof:oso/9780199685516.001.0001CrossRefGoogle Scholar
Pereboom, D (2021) Wrongdoing and the Moral Emotions. Oxford: Oxford University Press.CrossRefGoogle Scholar
Perszyk, KJ (1998) Free will defence with and without Molinism. International Journal for Philosophy of Religion 43, 2964.10.1023/A:1003040806970CrossRefGoogle Scholar
Phillips, DZ (2004) The Problem of Evil and the Problem of God. Minneapolis, MN: Fortress Press.Google Scholar
Plantinga, A (1974a) God, Freedom, and Evil. Grand Rapids: Eerdmans.Google Scholar
Plantinga, A (1974b) The Nature of Necessity. Oxford: Clarendon Press.Google Scholar
Plantinga, A (2009) Transworld depravity, transworld sanctity, and uncooperative essences. Philosophy and Phenomenological Research 78, 178191.10.1111/j.1933-1592.2008.00237.xCrossRefGoogle Scholar
Rasmussen, J (2013) On the value of freedom to do evil. Faith and Philosophy 30, 418428.10.5840/faithphil201330437CrossRefGoogle Scholar
Rowe, WL (2004) Can God Be Free? Oxford: Clarendon Press.10.1093/oso/9780198250456.001.0001CrossRefGoogle Scholar
Schellenberg, JL (2004) The atheist’s free will offence. International Journal for Philosophy of Religion 56, 115.10.1007/s11153-004-4626-3CrossRefGoogle Scholar
Shafer-Landau, R (2021) Fundamentals of Ethics, 5th edn. New York: Oxford University Press.Google Scholar
Smart, RN (1958) Negative utilitarianism. Mind 67, 542543.10.1093/mind/LXVII.268.542CrossRefGoogle Scholar
Sobel, JH (2004) Logic and Theism: Arguments for and against Beliefs in God. Cambridge: Cambridge University Press.Google Scholar
Swinburne, R (1998) Providence and the Problem of Evil. Oxford: Oxford University Press.10.1093/0198237987.001.0001CrossRefGoogle Scholar
Tooley, M (2015) The problem of evil. Stanford Encyclopedia of Philosophy (Winter 2021 Edition). Zalta, EN (ed), 193. https://plato.stanford.edu/archives/win2021/entries/evil/.Google Scholar
Turner, J (2013) Compatibilism and the free will defense. Faith and Philosophy 30, 125137.10.5840/faithphil201330210CrossRefGoogle Scholar
Van Inwagen, P (1983) An Essay on Free Will. New York: Oxford University Press.Google Scholar
Van Inwagen, P (2006) The Problem of Evil. Oxford: Oxford University Press.10.1093/acprof:oso/9780199245604.001.0001CrossRefGoogle Scholar
Watson, G (1996) Two faces of responsibility. Philosophical Topics 24, 227248.10.5840/philtopics199624222CrossRefGoogle Scholar
Wolf, S (1980) Asymmetrical freedom. Journal of Philosophy 77, 151166.10.2307/2025667CrossRefGoogle Scholar
Wolf, S (1990) Freedom Within Reason. Oxford: Oxford University Press.10.1093/oso/9780195056167.001.0001CrossRefGoogle Scholar
Zagzebski, L (2000) Does libertarian freedom require alternate possibilities? Philosophical Perspectives 14, 231248.Google Scholar