1. Oliveira’s symmetry argument and other similar arguments

In his article, Oliveira defines necessary evil as such: for all actual instances of evil e, the fact that e’s occurrence is not pointless gives God justifying reasons for allowing it if (1) e is necessary for bringing about some worthwhile greater good G or for preventing some unacceptable greater evil E, and (2) if e occurs, then either G in fact occurs or E in fact does not. He then formulates the ‘combination argument’, according to which one’s justified belief that God has sufficient reasons for allowing evil conjoined with a justified belief that I cannot mistakenly cause or allow a pointless evil yields a justification for moral inaction. One can justifiably believe they are permitted to stand by any and all evils they might otherwise permit precisely because they know that there is a good reason for them – for if they happen, they are indispensable to God’s plan.

Oliveira applies this line of argument to three proffered theodicies. First, some argue that some evil is necessary to some greater good, whether that greater good is freedom per PlantingaFootnote ² or an overarching organic whole in which an evil occurrence forms a unity that is on the whole good.Footnote ³ But if God only permits those evils from which he brings greater goods, then any evil that occurs in the actual world will be an evil from which God will bring about a greater good. Thus, I can be sure that if I myself allow some evil, if I believe that God has good reasons for permitting that evil to bring about a greater good, I can be assured that a greater good will be brought about. As such, I have a justifiable reason to permit a given evil to happen which I could have otherwise prevented. He raises similar arguments against ‘divine right’ formulations (e.g., God has the right to permit such evils and we do not) on the grounds that if some evil is necessary for some greater good or for the prevention of some greater evil, and A (an authority) has the right to allow it but not B, then B does not have a moral right to interfere with A’s allowing it. In other words, even if God has the right to permit evils whereas we do not, then I can always appeal to allowing God’s right to permit an evil for a greater good to stand in permitting any occurrence of evil. He then attacks open theism and divine duty theism (e.g. God simply has a sheer duty not to interfere with the evils he allows); as I think Oliveira’s arguments are successful against these options, I will not pursue his argumentation here.

Others have raised similar arguments.Footnote ⁴ The core idea of such arguments seems to be the same: if we know that any evil that happens will bring about a greater good, and if the occurrence of this greater good justifies God’s permission of this evil, then there’s no reason to think it wouldn’t justify our permission of said evil. Hence, the key premise to the argument is premise 2 above – the symmetry thesis. In developing my argument, then, I will aim to refute this symmetry thesis. But to do this, I will need to draw from an underlying view of providence that needs articulation.

2. A Reformed Thomist model of Providence (RTP)

It should be stated at the outset that, in articulating this vision of providence, I am not intending to reconcile these traditions so much as synthesize them. Hence, RTP will comprise the following claims:

1. (1) God decrees all that comes to pass, yet not in such a way that he authors or produces sin nor violates the freedom of his creatures (Westminster Confession of Faith 3.1).

2. (2) God executes his decree in time by actively moving his creatures to do whatever good they do while preserving their freedom, yet he ordains evil only by purposefully permitting them (e.g. permitting them in order to bring about higher-order goods).Footnote ⁵

3. (3) In accordance with (2), all the good of moral agents – whether moral or metaphysical – issues from God’s active causality, whereas all evil originates in the creature alone and is only permitted by God.

Each of these parts will benefit from explanation. (1) simply entails that God decrees, before the foundation of the world, what will and won’t happen in history. This definition is compatible with Thomism, Reformed, and even Molinist thought.Footnote ⁶ Indeed, it’s been argued elsewhere that the Westminster divines had something like libertarian freedom in mind when framing God’s relationship to evil.Footnote ⁷ This is why Jonathan Edwards model of the will received sharp criticism even from certain Calvinists;Footnote ⁸ Edwardsian determinism, in which the movement of the will is determined by factors external to the will, was not the intent of the framers of the Westminster Confession.Footnote ⁹

(2) states that God generates the good that any individual agent does, and yet only ever permits evil. Yet, with the Thomistic tradition, I argue that we should affirm that God produces good such that he moves creatures to freely perform the good whenever they do so. That means that there is a kind of co-operative movement of the will by which a creature joins with God’s operative grace so as to produce a given good. A more extensive treatment of how this might be the case can be found elsewhere.Footnote ¹⁰ For instance, God might give the grace of his influence in accordance with his knowledge that, if an agent is given x instance of grace, then the agent would freely do y in the circumstance of being given x. On the other hand, when God ordains that evil will occur, he only ever permits the evil.

(3), as a close corollary of (2), states that God is the source of all good in a creature’s will (ST. I.q105.a4). Yet he is never the cause or source of evil. Evil, rather, is what happens when a creature willfully distorts the order God has established in creation.Footnote ¹¹ Yet God permits such evil to bring forth higher order goods.Footnote ¹² Thus, RTP assumes that God decrees all that falls out in history, yet preserves creaturely free (even libertarian) will. He moves the creature to freely do all good that they do and only permits any evil he has decided from eternity to permit. Hence, creaturely good finds its source in God’s efficient causality (which does not override but works in and through creaturely willing), whereas evil occurs only when God permits the creature to will evil.

While a full description of the mechanics of divine and human interaction is outside the scope of this paper (and the sort envisioned here can be found elsewhere),Footnote ¹³ a few clarifications may help to elucidate RTP. First, when God generates good in his creatures, he does so such that the action is simultaneously God’s and the creature’s (as argued above). In envisioning this causal activity, we might describe it along the following lines. God is the radiant source of all good and beauty, such that all good and beauty in the world is in fact the emanation of God’s being on the human soul – like rays from the sun. Hence, whenever anyone feels stirred to good in recognition of the moral law, they are experiencing divine activity. That is, their conscience is a perception of the divine call and nature on the human heart in the process of moral decision making. God, as it were, intensifies the ‘radiance’ of his being in every perceived moral value, moving his creatures to a response.Footnote ¹⁴ For God to give some instance of grace, then, is just for God to give himself in some particular way to his creatures, where the self-giving of God is experienced as aesthetic delight or moral apprehension (or both). Where God permits evil, he simply chooses not to intensify his self-gift and lets the creature choose as she will. Thus, to say that God ordains all of history is just to say that God decides what will and won’t fall out in history and chooses to actively bring about the good by giving himself in particular ways (e.g. giving particular instances of grace) and permits specific evils in accordance with the purpose to bring about higher order goods.

3. The parity problem and the happiness consideration

Given the model of providence specified above, I argue that Oliveira’s argument fails owing to what I shall call the ‘parity problem’ and the ‘happiness consideration’ (PP and HC, respectively). Let us define A as a creaturely moral agent, and all evils – natural or moral – A is able to prevent as e_p: PP may be stated as such:

1. (1) If RTP is true, then A has no more providence-parity-considering reason to allow e_p as she does to prevent e_p.

2. (2) If a moral agent does not have any more providence-parity-considering reason to allow a preventable evil than to prevent it, then the consideration of this parity itself constitutes a reason to prevent the evil.

3. (3) Therefore (from 2 and 3), if RTP is true, then the consideration of this parity itself constitutes a reason to prevent the evil.

In other words, the consideration of reasons from divine providence seem to generate a parity between one’s reasons for allowing a preventable evil and their reasons for preventing it (the providence-parity considering reasons). This perceived parity then serves as a kind of parity-breaker by itself being a reason to prevent preventable evil.

In defense of premise 1, RTP stipulates that, for any good a creature enacts – including the prevention of evil – such good is sourced in God’s own movement and so expresses God’s will. Hence, if a creature prevents some evil, the creature does so through God’s grace. Thus, the prevention of e_p is enacted by God through the creature when the creature prevents e_p. If this is the case, then prior to e_p’s occurrence, the creature has no more reason from the consideration of providence itself to allow the evil’s occurrence than to prevent it. For if she prevents the evil, then she will discover in the act of prevention that God did not deem such an evil necessary to a higher-order good or, if it was, that such a higher-order good was not necessary to the perfection of the universe. Premise 1 obtains, then, because the creature is equally justified in holding two beliefs in tension from the consideration of providence alone:

Belief 1: God has sufficient reason to permit e_p, and so if I permit e_p, I will do so on appeal to God’s sufficient reason which guarantees some higher order good.

And

Belief 2: If I prevent e_p, then that will show there was no higher order good for which e_p’s occurrence was necessary or, if it was, then that such a higher order good is not necessary to the perfection of the universe.

Her consideration of providence lends equal support to either belief. Therefore, she has no more reason to think that she ought to permit e_p than prevent e_p. Hence, premise 1 obtains.

Premise 2 seems fairly intuitive. For it is another way of stating that, all things being equal, an agent should prevent evil rather than permit it. But beliefs 1 and 2 have epistemic parity from the consideration of providence alone and so are equal with respect to moral justification. For given RTP, in preventing e_p, A would in fact be the instrument by which God prevents e_p and demonstrates that such an evil was not necessary to the occurrence of a higher order good. If I have no more reason to think that permitting some evil will bring about a higher order good (or prevent a worse evil) than preventing it would prevent an evil unnecessary to the obtaining of that higher order good, then there could be no justifiable reason why I would opt to do the former rather than the latter.

This argument on its own, it seems to me, would be sufficient to show Oliveira’s argument to be amiss. For PP itself lends support to asymmetry, and so counters premise 2 in Oliveira’s argument. But there’s another reason why Oliveira’s argument does not succeed which, in fact, gives an agent good reason to do the good instead. I call this the Happiness Consideration (HC).

Many moral traditions consider ‘happiness’ as the last end of humanity. Pascal famously wrote that all people seek happiness without exception – even the person who hangs himself.Footnote ¹⁵ One finds a similar notion in Aristotle, with the acknowledgement that people mean different things by the word ‘happiness’.Footnote ¹⁶ Jonathan Edwards writes,

A desire for happiness – denoting something like general flourishing or fulfillment – seems to be inscribed into human nature. If there were someone who didn’t desire their general flourishing on any level, we’d likely think that something has gone seriously wrong with this person – that they would be afflicted with an illness of some sort. Happiness seems something like a general telos toward which human nature is oriented.

While one may not be able to derive an ought from an is (though this is debatable), one can certainly derive an ought from a goal or telos.Footnote ¹⁸ For instance, if I seek to make an omelet, then I ought to use eggs and I ought not use dirt. Oughts are indexed to goals. And as has been argued elsewhere, the goal of the moral life is happiness – e.g., flourishing, well-being, thriving, etcetera. Fitzpatrick argues that moral judgments are therefore judgments of those actions by which I uphold and promote my own human integrity.Footnote ¹⁹ Moral ‘oughts’ therefore exist in the domain of the pursuit of happiness. Now, one might worry this smacks of ethical egoism – as though all moral striving stems from self-interest.Footnote ²⁰ And it might in some sense. However, it should be importantly noted that there are ways to pursue one’s own joy that are, in fact, outward facing. John Piper in a defense of his philosophy, ‘Christian hedonism’, gives the following illustration. Suppose I buy flowers for my wife and book a spontaneous dinner date. When she asks me, ‘Honey, what’s the occasion? Why tonight?’, and I say, ‘Because it’s my duty’, I will have colossally failed as a husband. On the other hand, if I say to her, ‘Because nothing makes me happier than seeing you happy’, then my joy in her joy will honor her.Footnote ²¹ Now, I do not quote this example to advocate for Christian hedonism as such – I have reasons why I reject the label and the potential egoistic psychology of Christian hedonism. But it is nevertheless instructive, insofar as it shows that the pursuit of one’s well-being if such well-being and joy is outward facing needn’t be selfish.

What is the nature of moral oughts on a Christian vision of the world? It should be noted at the outset that Christianity is not committed to a strict eudaemonism that rules out ethical action for its own sake. Footnote ²² But it should also be noted, contra Mendham, that Christianity does not differ from eudaemonism in making something other than happiness the end of the moral life but by reconfiguring the very nature of happiness and ethical action. CS Lewis famously pointed this out in The Weight of Glory:

Fascinatingly, Christian ethics seems motivated largely by the promise of eternal happiness. This does not necessarily mean that the pursuit of happiness is the only motive in all Christian action but rather that it’s (at least) an essential component of all Christian moral action.

But what is the nature of happiness – of true flourishing – in a Christian vision of ethics? It seems to me, with the great tradition, that Christianity claims humans are fulfilled and only truly flourish in union with God.Footnote ²⁴ In other words, seeing God’s beauty and participating in that beauty – becoming truly one with God – is the chief end of human life and all creation. But the way in which a thing participates in God depends on the thing created, for each created thing’s perfection is its own unique assimilation to God’s divine excellence (ST. I.q44.a4). Just as dirt might cause a worm’s nature to flourish but not a human’s according to the different specifications of their nature, God has shaped creatures to reflect different aspects of his moral character. The Christian story, after all, teaches that angels are not subject to the same set of moral demands as humans (Heb. 2:16). Creatures, then, flourish when they reflect those aspects of God’s character which they were actually made to reflect. Humans must be attentive to the specifications of their nature in order to know how they ought to flourish.

But how do humans gain knowledge of their nature to know their moral obligations? God’s self-revelation in Jesus is, ultimately, the supreme norm by which humans learn who they are and what their nature is.Footnote ²⁵ This is why Calvin famously argues that the knowledge of ourselves is intimately tied to and clarified in the light of the knowledge of God.Footnote ²⁶ This is not, however, to cede to some divine command theory (DCT). According to its most sophisticated versions, DCT stipulates that wrongness or moral evil consists in disobedience to God’s commands, such that wrongness is defined by such disobedience. The reason, then, any individual action might be wrong is because it contravenes God’s commands – which commands perfectly express God’s nature.Footnote ²⁷ On the contrary, the account offered here stipulates that wrongness is constituted by distorting God’s self-communication in the structure of the world. Thus, an action is wrong just so far as it distorts the good – and therefore distorts God’s likeness in the nexus of creation.Footnote ²⁸ God’s commands, then, reveal the structure of human nature – and the good for humans qua humans – rather than define them as such.

Given this framework, certain refractions of God’s goodness might be appropriate for one creature that aren’t appropriate for another. This appropriateness can depend on either nature or role. For instance, it is appropriate for a judge to enact justice on a criminal but it is inappropriate for me to pursue vigilante justice owing to our difference in role rather than nature in the overall creational nexus. My nature is created and adapted to the overall nexus comprising created things and constituting creation; hence, I will flourish in accordance with God’s design for me to reflect his glory in particular ways rather than others. If true flourishing is constituted by participating in and reflecting God’s glory in my created nature, then my moral ‘oughts’ are constituted in relation to that particular telos.

Before specifying how this addresses Oliveira’s argument, one more qualification should be made. Some theorists resist teleological ethics insofar as it makes doing a good deed for its own sake either impossible or exceptionally difficult. Oughts are, it is claimed, means to an end.Footnote ²⁹ But this isn’t necessarily the case given every goal. For instance, if my wife and I aim to spend time with each other, then we ought to go on a date. But the date isn’t a means to an end external to the activity. The activity, rather, is an instantiation of the overall telos rather than an end achieved by some other means. It is in this sense that Christians ought to think of ‘moral oughts’ and their relation to union with God. Given RTP, any enacted good isn’t simply something we do to press into the life of God but is actually an assented-to union with God’s active presence, who is the origin of the good I do. Through a perception of the good, God shows an aspect of his life to my mind; through my enactment therein, I participate in God’s life according to the enactment of the perceived good. Every enacted good, then, is a way of pressing into the life of God. On the other hand, every act of evil is resistance to God’s activity in the world, and every natural evil is a disordering of God’s design of the natural world.Footnote ³⁰

This consideration gives us positive reason to opt to do the good we perceive, contra Oliveira’s argument. For through the parable of the Good Samaritan, Jesus’s instructions for his people – which instructions reveals God’s design for humanity at large – is for them to actively oppose any moral evil they see and seek to remedy the suffering of others to their ability. The role of humanity as ‘image bearers’ seems to describe human function in terms of reflecting God’s wise rule and order into the world – thereby remedying natural evil by restoring God’s order in creation.Footnote ³¹ Since moral ‘oughts’ are teleologically situated in the pursuit of union with God – which is the fulfillment of humanity’s nature – we ought to prevent any evil we see according to the way our natures are designed to reflect God’s own goodness. Now, one might object, ‘but if you’re seeking to prevent an evil that God is not, then you’re not reflecting God’. But this sort of objection forgets a key stipulation of RTP: any good I do and imbibe is a participation in God’s own goodness. Thus, my efforts to prevent any evil I see reflects a real aspect of God’s character that persists even if God chooses not to prevent an evil: namely, God’s grief over every instance of evil and his opposition to it. While God permits what he hates for morally sufficient reasons, he does not hate an evil any less for it. A disharmony is still a disharmony, even if it will eventually find resolution. Nevertheless, God reflects his true hatred of any and all evil through human beings who oppose any and all evils through participation in God’s goodness.

Let’s summarize the argument thus far, and then apply it to the various defenses addressed by Oliveira. Given RTP, I argue that, prior to the commencement of any preventable evil e_p some agent A is able to prevent, A is equally justified in holding two conflicting beliefs: ‘if I permit this evil, then this evil is necessary to some greater good or the prevention of some worse evil given God’s providence’ and ‘if I prevent this evil, then this evil is not necessary to some greater good or the prevention of some evil’. A, therefore, has no more reason to accept the first belief over the second, and consequently she has no more reason to act upon the first belief over the second. In other words, she has no more reason to permit the evil than to prevent it from consideration of God’s providence over evil. If A has no more reason to prevent an evil than to permit it – such that her reasons to permit an evil do not outweigh her reasons to prevent it – then she ought to prevent it. Second, A, as a human being, possesses the sort of nature designed to reflect God’s compassion and opposition to such evils. Her nature is fulfilled in participating in God, which happens by participating in his character. Each enactment of God’s commands is an enactment of God’s goodness, in which she participates in some limited way in God’s life. Since God’s self-revelation in Jesus tells her to ameliorate any evil within her power to do so, unless she has some outweighing reason not to, then she ought to ameliorate such evils – for moral oughts just are the enactment of the teleological aim of the moral life, which is flourishing, e.g. participation in God’s life. Premise 2 in Oliveira’s argument is therefore false.

4. Conclusion: False starts and answering potential objections

From the above, it should be clear that the answer given here rules out the free-will defense. Now, one might be able to take the basic insights here (e.g., the Parity Problem and the Happiness Consideration) and adapt them to the free will defense. For instance, it might be argued that God’s actualization of the best feasible world entails the Parity Problem, insofar as one might be in the same epistemic standpoint they’re in under RTP. Further still, HC is an entailment of a kind of natural-law Christian ethic at large plus the affirmation that God is the origin of all the good we do (no tradition need affirm that God is the sufficient cause of all good we do). The ‘Divine Right’ answer Swinburne gives might also be modified along similar lines. But as I reject the assumptions underlying either answer for reasons outside the scope of this paper, I will not pursue that line of thought here. It suffices to note that the free will defense is not open to me.

Thus, JL Mackie is right that nothing external to God (e.g., such as free will) prevented God from actualizing a world in which free agents only ever freely chose to do the good.Footnote ³² But his argument has a fatal flaw. Mackie argues that, if some greater good can come out of a horrible evil, then God would have no reason not to let the spiral of good and evil go on for eternity, in which ever increasing goods are brought forth from ever increasing evils. However, it might be the case that there is an ideal proportion of evil-defeating goods to such evils which provide a kind of limit. Thus, God may be after a particular kind of harmony as such, which harmony obtains given the proper proportion.

One might also argue that this response commits me to saying that God has created the best of all possible worlds. However, this is mistaken. While I do think that such an answer commits me to saying that God has created a maximally God-reflecting world – a maximally good world, the totality of world history considered – it is not clear to me that this world is the only possible world. However, it does seem to me that RTP commits me to arguing that any created world which God creates would contain evil which God would, ultimately, defeat in the best way, axiologically speaking. I am committed to such a position.Footnote ³³ However, if this argument is raised against my response to Oliveira’s argument, then it will not show that Oliveira’s argument itself is not defeated by the considerations sustained RTP; rather, it would be to raise a different argument. My underlying argument has been that PP and HC defeat Oliveira’s argument; RTP sustains PP and HC, but other models of providence might as well. However, I’ve opted to ground PP and HC in RTP given that I believe RTP is true and is the most theologically and philosophically defensible mode of providence. Yet other theists may well feel free to adapt PP and HC into another model of providence as they please in order to defeat the Symmetry Challenge.

Ultimately, from PP and HC, Oliveira’s argument cannot be sustained. We do not have morally sufficient reasons to permit those evils which God has morally sufficient reasons to, but instead must live in union with him, reflecting his nature in accordance with our purpose and order in the overall symphony of the cosmos.