Introduction

As is well-known, the Problem of evil asks why an omnipotent, omniscient, omnibenevolent God would allow the existence of evil and suffering in our world. Theodicy refers to those theories which seek to provide plausible and satisfactory answers to the Problem of evil, answers which leaves God free from blame. This is understood as “vindicating” or “justifying” God. As such, theodicy is often described with a reference to John Milton’s epic Paradise Lost, as the attempt to “justify the ways of God to men” (Milton [1667], 2005: Book I: 26). Anti-theodicy has been described as a protest against the construction of theodicies, and as such, is therefore quite distinct from attempts to debunk individual theodicies, since anti-theodicy rejects all theodical reflection. Nick Trakakis, for instance, describes the anti-theodical project as the attempt to show that theodicies are by their very nature fundamentally problematic.

An “antitheodicy” … aims to show that all theodicies are by nature defective in some important respects, and that these are not the kinds of deficiencies that could be remedied if only we knew more about, say, God’s intentions or the workings of the world. The very project of theodicy, on this view, is a nonstarter … In other words, the way in which the problem of evil is usually formulated rests on various presuppositions and principles, and once the dubious and questionable nature of these is made clear, the problem of evil dissolves back into the ether from which it arose (Trakakis, 2013: 363).

Some have found it useful to distinguish between two categories of anti-theodicy—moral and conceptual. As the title suggests, this paper focuses on moral anti-theodicy. In 2009, Robert Mark Simpson argued that moral anti-theodical arguments, which he also calls “global moral arguments against theodicy,” end up begging the question by discounting the possibility of there being a plausible explanation for why God permits evil, thus presupposing what they intend to establish. A decade later, in a debate with Toby Betenson, Lauri Snellman has likewise argued that moral anti-theodicy begs the question, by arguing that it presupposes that there are some evils that cannot be justified, thus ending up presupposing that theodicy is false. Others have also considered this critique, but in this paper, I have chosen to focus on Simpson and Snellman, since their papers are indirectly connected, as will be made clear below. I will proceed in chronological order, by first analysing Simpson’s, and then Snellman’s arguments. In this paper, the strength or success of moral anti-theodicy is not under consideration. Rather, I will be evaluating the arguments from two scholars, writing a decade apart, who conclude that moral anti-theodicy begs the question.

Simpson’s arguments

Simpson writes that “any global argument against theodicy must start from the position of agnosticism in relation to the tenability of theodical discourse. … What this means, then, is that a global moral argument against theodicy needs to grant (for the sake of argument, at least) the possibility that there is a theodicy whose claims about why God might permit the evils in our world … are actually plausible” (Simpson, 2009a: 160–161). He further states that if we assume, in advance, that there is some plausible explanation as to why God allows evil, then we grant theodicists too much, but if we rule out the possibility of there existing a plausible explanation, then we grant theodicists too little, and we end up begging the question against theodicists. From this demand, Simpson concludes that the anti-theodical argument from insensitivity, the argument from detachment, and the argument from harmful consequences all implicitly discount the possibility of there being a plausible explanation for why God permits evil. Furthermore, in relation to these arguments, Simpson does not feel the need to analyse any specific anti-theodical argument, writing that: “In my discussion I will not focus too heavily on the arguments of any one of these [anti-theodical] writers; instead, I will try to provide a general overview of the moral antitheodical position using examples from each of [the anti-theodicists’] work” (2009a: 155). As such, the anti-theodical arguments discussed by Simpson should be understood as types of arguments (i.e. arguments sharing similarities), rather than specific arguments raised by anti-theodicists.

Before considering the arguments and Simpson’s charge, it may be useful to consider Simpson’s reconstruction of moral anti-theodicy. The moral anti-theodical arguments Simpson evaluates (the argument from insensitivity, the argument from detachment, and the argument from harmful consequences) individually aim to support what Simpson calls the key premise of moral anti-theodicy.

  1. P1.

    There is an inherent moral impropriety in the discourse practice of theodicy (key premise of moral antitheodicy).

  2. P2.

    One ought not to engage in a discourse practice that constitutes or leads to some form of moral impropriety (putative ethical principle).

  3. C.

    One ought not to engage in the discourse practice of theodicy (ibid., 156).

As we can see, for Simpson, the key premise of moral anti-theodicy states that there is an inherent moral impropriety in the discourse practice of theodicy. From Simpson’s reconstruction, we can see that if this key premise was to be granted, the moral anti-theodicist could argue for the abandonment of the theodical project. Simpson’s reconstruction, especially what he names the “key premise,” will be relevant in my evaluation of Snellman’s argument, but for now we will focus on the moral anti-theodical arguments used to support this “key premise,” since Simpson believes these arguments beg the question. However, before we consider Simpson’s evaluation of the anti-theodical arguments, it will be necessary for us to account for the nature of anti-theodical critique.

Meta-theodicy and meta-critique

What is theodicy? What does theodicy aim to do? (How) Should theodicy be practiced? These are meta-theodical questions; that is, questions on the meta-philosophical level concerning the theodical enterprise in itself. The answer to the middle-question is often portrayed as the attempt to “justify the ways of God to men”—to vindicate God’s ultimate goodness or justice in relation to all the evil and suffering found in the world. This is done by providing plausible theories as to why God might be justified in allowing evil and suffering to exist. As such, the theodical discussion (or debate) concerns whether or not these theories are plausible or not. This is what is meant by ‘engaging in theodical discourse practice.’ Furthermore, discussions regarding the plausibility of theodical theories, along with the presuppositions these discussions and theories rely on, make up the theodical framework. As such, theodicists (like J. Hick and R. Swinburne) as well as critics of individual theodicies (like J. L. Mackie and A. Flew) are all engaging in theodical discourse practice, since in their discussion on the plausibility of theodicies, they are all relying on (and, at least, tacitly accepting) the presuppositions of the theodical framework. Put differently, anyone discussing the plausibility of a theodicy is relying on the presuppositions of the theodical framework, whether they are arguing for or against any given theodicy.

Now consider the three questions raised above again. Neither one of these questions provide a theory as to why God might allow evil, nor do they discuss the plausibility of any such theory. The three questions raised above are therefore questions of the theodical discussion or debate, not questions raised within the theodical debate. This is what makes them questions on the meta-philosophical level, or meta-theodical questions, and as meta-theodical questions they do not rely on, nor must they adhere to, the presuppositions or rules of the theodical framework. Rather, it is from questions like these that we can discover what the theodical framework presupposes.

Therefore, a good litmus test for whether a critique is a criticism of the theodical framework, or a criticism raised within the theodical framework, is whether or not the critique is discussing the plausibility of an explanation as to why God would allow the existence of evil. In other words, we need to ask: is the critique discussing the plausibility of a theodicy? If it does, then it is a criticism within the theodical debate, and is thus also relying on, or accepting, the presuppositions of that debate; it is operating within the theodical framework. If it does not, then it is a critique of the theodical debate, and as such is situated outside of the theodical framework. All of the criticisms Simpson considers in his paper—the argument from insensitivity, the argument from detachment, and the argument from harmful consequences—do not discuss the plausibility of a given theodical explanation of evil, and therefore Simpson has correctly identified them all as anti-theodical arguments; they are criticisms of theodicy, from outside the theodical framework.

Anti-theodicy is therefore a critique from outside of the theodical framework. Not only is this clear when one considers the nature of the arguments, but at times, anti-theodicists themselves have been quite explicit on this point. Consider notable anti-theodicist D. Z. Phillips, who writes: “I am not an antitheist, like J. L. Mackie, whose opposition is expressed within a theodicist framework. I want to probe the character of that framework” (Phillips, 2001: 146). Here, Phillips is explicit in that his critique is aimed at the theodical framework itself, as opposed to Mackie, whose critique of theodicy is expressed within that framework. Mackie’s critique of theodicy is therefore situated within the theodical framework as much as Hick’s or Swinburne’s construction and defence of theodicy, as I explained above.

We might also consider Toby Betenson, who explicitly and helpfully explains that “anti-theodicy might be better understood as a kind of meta-critique of the discussion concerning evil; it is a criticism of the discussion, rather than a criticism within the discussion” (Betenson, 2016: 57. Emphasis in original). And since anti-theodicy is a critique of the theodical discussion (or framework), it is and should be understood as a critique on the meta-philosophical level, and as such, it does not rely on, nor is it bound by, the presuppositions of the theodical framework.

So, considering the meta-theodical nature of anti-theodicy I have outlined here, I believe Simpson has not adequately considered what an argument on the meta-philosophical level entails. On my view, discussing or considering the plausibility of theodicies places a critique within the theodical framework, and so to insist that anti-theodical arguments “must allow for the possibility that there exists a plausible theodicy,” would be to place anti-theodical arguments within the theodical framework—i.e. within a framework where they, qua meta-philosophical arguments, do not belong.

The argument from insensitivity

Simpson first analyses the anti-theodical argument from insensitivity, which states that theodicists adopt an insensitive attitude toward the suffering of others. In evaluating this argument, Simpson presents us with two hypothetical examples—one which he believes is plausible and one which he believes is implausible. In the plausible example, a person is communicating a true statement to the family of someone who is suffering from a disease about the long-term benefits of the disease for the sufferer, and Simpson believes that the insensitivity of communicating such a true statement would depend on the situation; i.e., it is not necessarily insensitive. Next, in the implausible example, a person is communicating something which is false to the family of someone who is suffering from an illness. Here, Simpson believes that the person in this scenario is blameworthy regardless of the situation. Simpson writes that “[t]he problem with this response … is that it is false, and all the more insensitive for being false” (2009a: 162).

Thus, for Simpson, responses are (more) insensitive if they are false (like in the implausible example). It seems like Simpson believes that whether or not a theodicy is true or plausible will decide if a theodicy is sensitive or not. Simpson further asks us to imagine a theodicist who says that God is justified in allowing terminal illnesses because they improve the human gene pool, and states that communicating such a theodicy in the nicest way possible is still appalling, because it is obviously untrue (ibid.). He continues:

according to the argument from insensitivity, all theodicies … are in some way akin to the [implausible] example … they are insensitive no matter how and when they are espoused. … If the argument from insensitivity asserts that all theodicies are like the [implausible example], then it presupposes that the content of the claims in any given theodicy will be implausible, and … this assumption precludes the argument’s success (ibid., 162-3. Emphasis in original).

Firstly, in this argument, Simpson is assuming that a theodicy’s insensitivity is tied to its truth or plausibility. However, this is something the moral anti-theodicist would deny. Therefore, to assert that theodicies are insensitive is not to make any claims about their truth or plausibility, since even if a theodicy were true or plausible, the moral anti-theodicist would still find the attitude of theodicists insensitive. Secondly, Simpson has focused on the moral impropriety of communicating a theodicy. However, for moral anti-theodicists, the moral impropriety of theodicies still stands even if they are not directly communicated to those who are suffering. This is because moral anti-theodicists typically focus on theodicists’ attitude toward suffering, which remains the same even when theodicies are not communicated to those who are suffering. Consider Samuel Shearn’s response to discussions on communicating theodicies:

Theodicy does not stay in a university laboratory, sealed off from those who have suffered horrendous evils. Theodicy is written and spoken in the public arena, in the world of ideas, and to propagate a theodicy is to communicate it at once to potentially everyone, because the theodicy of academic journals finds its way to the pulpits and popular religious books, providing religious communities with theodical narratives for interpreting horrendous evils. Therefore … there is no strictly ‘theoretical context’ to which one can turn and safely do theodicy (Shearn, 2013: 444–445).

Thirdly, regarding the last sentence in the quote from Simpson, remember that Simpson has stated that if one rules out the possibility of there existing a plausible theodicy, then one begs the question against theodicy (Simpson, 2009a: 160). Simpson has then presented us with two examples, one which is true and not necessarily insensitive, and one which is false and insensitive. He further claims that the argument from insensitivity asserts that all theodicies are akin to the false and insensitive example. But Simpson’s examples are improper, because they highlight features of theodicy which are irrelevant for this anti-theodical argument: their truth or plausibility, and ways of expressing theodicies.

If we remember our discussion on meta-theodicy, then we see that Simpson is demanding too much from the moral anti-theodicist when he writes that any moral anti-theodicy must allow the possibility that there exists a plausible theodicy. This is because anti-theodicy is concerned with meta-theodical questions, e.g. whether or not we should be engaging in theodical speculation in the first place. The theodical framework is presupposing that the answers to this question is yes (or, at least, that there is nothing problematic about engaging in theodicy). I would agree that one answer, or clause, of the meta-theodical question “how should theodicy be done?” is the demand that one must allow for the possibility of there existing a plausible theodicy; I believe this demand is necessary for the existence of the theodical enterprise. But, meta-philosophically speaking, before we can answer the “how?”-question, we must answer the “should?”-question. That is, the question “should theodicy be done?” is meta-philosophically prior to the question “how should theodicy be done?” So, since the “should?”-question is meta-philosophically prior to the “how?”-question, and since anti-theodicy is concerned with the “should?”-question, the demand that anti-theodicy must allow for the possibility that there exists a plausible theodicy is misplaced, because this demand can only be applied to criticisms within the theodical framework, i.e., criticisms which have accepted the presuppositions of the theodical framework. For the anti-theodicist, these questions are irrelevant. To apply this demand to anti-theodicy would be to get ahead of oneself, to reverse the proper order in which questions are considered.

So, by insisting that an anti-theodicy must assume the possibility of there existing a plausible explanation as to why God allows evil, Simpson has placed anti-theodicy within the theodical framework. But since the question of the plausibility of any such explanation is irrelevant for anti-theodicy, Simpson is mistaken, as he has not considered the meta-philosophical nature of anti-theodicy. Because of this, Simpson has made the mistake of evaluating moral anti-theodicies by assuming the theodical framework, because it is an assumption of the theodical framework that it is possible that there exists a plausible theodicy; this is a rule which debaters must accept when arguing within this framework. But since anti-theodicy is situated outside of this framework, it is under no obligation to accept this rule. Importantly, this does not mean that anti-theodicy assumes its negation—that there is no possibility of there existing a plausible theodicy. Even though anti-theodicists may argue for it (or it being an implication of their arguments), they do not assume it. The same goes for the question of whether or not a given theodicy is true. It simply does not matter to the anti-theodicist. As Simpson himself writes,

Moral judgements about a discourse practice and its sensitivity towards the victims of suffering have no bearing on whether or not it is plausible to claim, say, that there will be a mass resurrection of human beings in the distant future. … since the argument from insensitivity can neither demonstrate [the] implausibility [of these kinds of claims], nor draw any other independent arguments which might demonstrate their implausibility, it cannot be used to establish the impropriety of theodical discourse in a global sense ( 2009a: 163).

While Simpson is right that anti-theodical arguments from insensitivity cannot say anything about the plausibility of theodical explanations of evil, this is not their function; it is not what they intend to do. By demanding that moral anti-theodicies allow for the possibility of there existing a plausible theodicy, and then rejecting them when he finds that they say nothing about this plausibility, Simpson has not only effectively rejected the entire anti-theodical enterprise, but he has also ensured that there can be no meta-critique of theodicy. That is, he has effectively shielded theodicy from any meta-philosophical criticism, because, according to Simpson, any critique of theodicy must allow for a presupposition of the theodical framework: that it is possible that a plausible theodicy exists. Again, I feel the need to stress the importance of the relation of anti-theodicy to Simpson’s demand. Just because moral anti-theodicy says nothing about this possibility, this does not mean that moral anti-theodicy rejects the possibility. These are not the only options available. By virtue of it being a meta-philosophical critique, the anti-theodical debate concerning if we should be engaging in theodicy in the first place, comes before any rules are established for engaging in theodical debate. As such, Simpson’s demand becomes irrelevant for the anti-theodicist. Because anti-theodicy is a critique on the meta-level, we must consider it in its proper, meta-philosophical context. We cannot demand that anti-theodicy adhere to theodical rules, which themselves are grounded in the assumption that the theodical enterprise is unproblematic, which is the very assumption being questioned by anti-theodicists! As such, theodical assumptions or presuppositions, which make up the theodical framework, already assume that anti-theodicy is false. As such, we cannot demand that anti-theodicy adhere to a framework which presupposes that anti-theodicy is false.

The argument from detachment

Further, in evaluating the argument from detachment, Simpson portrays this argument as a virtue-ethical objection, which argues that the discourse practice of theodicy “necessitates a vicious form of emotional detachment or hard-heartedness with regards to the suffering of others,” like “a military commander who calls death and suffering ‘collateral damage’” (Simpson, 2009a: 159).

Surprisingly, Simpson does not argue against the claim that theodicists utilize a viciously detached attitude when constructing theodicies—though he maintains that theodicists themselves will want to do so. Rather, he focuses on providing two ways through which a theodicist might respond to this anti-theodical argument, both of which would accept that detachment toward suffering is, generally speaking, a character vice. According to the first, if the facts about God and evil are as a theodicist believes them to be, “then the act of propounding a theodicy may be motivated solely by concern for the well-being of others. … The theodicist can simply argue that the requirements of benevolence trump the requirements of non-hard-heartedness” (ibid.: 164).

Firstly, as I see it, this hypothetical argument could be advanced by theodicists regardless what the facts are, since we are not in a position to know what the facts are, or even if a God exists. This means that a theodicist could appeal to benevolence, even if they are propounding a theodicy which is false. Secondly, if the motivation of concern for the well-being of others is dependent on the facts being as the theodicist believes them to be, as the argument suggests, then theodicists are not in a position to know whether or not they are, or could ever be, justified in espousing theodicies. Again, Simpson believes that whether a theodicy is true or not has any bearing on the anti-theodical argument. Not only are we not in a position to know the truth, but it is also irrelevant for the anti-theodicist, as is a theodicy’s (im)plausibility. We can also see, again, that Simpson believes the issue lies in communicating theodicies, which is not necessarily true; anti-theodicists are generally more concerned with the construction of theodicies.

The second way in which Simpson believes a theodicist might reply to the argument from detachment is by arguing “that detachment or hard-heartedness is a role virtue for philosophers or theologians seeking an adequately theoretical [theodicy]” (ibid.). Simpson sees this response illustrated in David O’Connor, who provides an analogy of a patient dying from an incurable illness.

To the victim, trapped in the sure prospect of death and for whom each day is a heroic struggle […] detached, disinterested research conducted wholly without reference to him in the particularity of his own anguish is beside the point […] But to conclude from this that pure medical research, with its inherent abstract conception of disease, immunity, life and death is irrelevant […] would be to misconclude (O’Connor, 1988: 65–66; Cited in Simpson, 2009a: 164. The emphasis is O’Connor’s).

By analogy, a theodicist’s hard-hearted attitude can be considered a virtue, like the medical researcher’s hard-hearted attitude toward the patient is a virtue for them, even though such attitude may be considered vicious outside of that context. This is because both the theodicist and the medical researcher have concerns for those who are suffering, which Simpson believes removes the viciousness from the detached attitude. However, there is a significant disanalogy between theodicists and medical researchers that we must consider. Firstly, an emotional detachment that can be perceived as cold-heartedness is not uncommon among medical professionals. For someone working with death and suffering every day, detachment is an effective coping mechanism to be able to deal with suffering more effectively and for longer periods of time. This is not the case for the theodicist. Secondly, in the analogy, the cold-heartedness of medical researchers is a virtue because they are trying to save lives. But the theodicist, is not trying to save people who are suffering, or people who will suffer in the future. Therefore, O’Connor and Simpson cannot conclude that detachment is virtuous in the case of the theodicist, since the goal of saving lives is precisely what does not make the detachment of medical researchers vicious. O’Connor can argue that theodicists make use of the same approach, but not that they have the same goal, which directly affects the viciousness or virtuousness of the detached attitude of medical researchers.

Simpson writes that both of his proposed responses to the argument from detachment

Proceed on the assumption that there could be a plausible explanation as to God’s permission of the evils in this world. If it is not the case that a plausible explanation exists, then it may be the case that any engagement in the discourse of theodicy constitutes a moral impropriety, due to the detached attitude it requires. Once again, however, to rule this possibility out would be to inappropriately presuppose the global failure of theodicy (Simpson, 2009a: 165).

As in the argument from insensitivity, Simpson believes that the argument from detachment also rules out the possibility that there exists a plausible theodicy, and that this would beg the question against theodicy. But it is unclear why Simpson believes this argument rules out the possibility that there exists a plausible theodicy, though he does provide a clue as to why in the antecedent of the conditional in the quote above. He writes that engagement in the discourse practice of theodicy could constitute a moral impropriety, if it is not the case that a plausible explanation exists. As in the argument from insensitivity, Simpson seems to believe that the existence of a plausible theodicy determines whether a discourse practice is vicious or not, and by simply arguing that a discourse practice constitutes a moral impropriety, one has effectively discounted the possibility of there existing a plausible theodicy. But as I explained when evaluating Simpson’s response to the argument from insensitivity, moral anti-theodicists would deny this. Further, the demand that anti-theodicy must allow for the possibility that there exists a plausible theodicy is misplaced, because this demand can only be applied to criticisms within the theodical framework. As stated above, since anti-theodicy is a meta-philosophical critique, we cannot demand that anti-theodicy adhere to the rules of a framework which presupposes that anti-theodicy is false. Thus, we see that Simpson is again demanding too much from the moral anti-theodicist when he writes that any moral anti-theodicy must allow the possibility that there exists a plausible theodicy.

The argument from harmful consequences

Lastly, Simpson describes the anti-theodical argument from harmful consequences in the following way.

According to moral anti-theodicists such as Terrence Tilley and Robert Mesle, theodicies tacitly sanction an acceptance of suffering and injustice, and thus they reinforce oppressive social structures. In other words, when theodicies posit evil as something that is ultimately for the good, they deter individuals and societies at large from working to overcome the sources of evil in our world. Theodicies tell us to tolerate suffering, these critics say, when we should oppose it. In responding to such claims, theodicists would generally agree that the passive acceptance of evil is a further evil in itself, and that the encouragement of this way of thinking constitutes a harmful consequence for a discourse practice (Simpson, 2009b: 341).

Simpson believes that this sort of anti-theodicy does not beg the question. However, he continues, this sort of anti-theodicy does beg the question in its presumption that harmful consequences are an inevitable result of theodicy.

Maybe theodicy does sometimes lead to the problems these critics identify, but this can be explained in any number of ways. Perhaps the theodicies that have been devised so far are wrong, but there is a right theodicy out there that is yet to be devised and disseminated. […] [I]t seems reasonable for the theodicist to suppose that if a plausible theodicy is in the offing, it would not only explain why there is evil in the world, but will also encourage a right moral orientation in our practical response to earthly evils. The consequentialist critique, however, rules this possibility out. As long as theodicies are espoused, the critique says, harmful consequences will follow (ibid.).

Firstly, Simpson does not say why “it seems reasonable to expect that a theodicy which ‘gets things right’ … will also support an upright moral orientation towards the evils that exist in our world” (Simpson, 2009a: 166). Theodicy says nothing about an upright moral attitude toward evil, since this is not its purpose, so it is unclear why it would be reasonable to expect this. Secondly, Simpson believes that even if we grant that theodicies may sometimes lead to harmful consequences, like those Tilley and Mesle identify, the question still remains whether any future theodicy will likewise have problematic consequences. He writes, “[p]erhaps if proponents of moral antitheodicy were to assume that any future theodicies will be flawed in their claim about God, evil, redemption and so on, they might be able to reasonably predict that the practical consequences of those future theodicies will be harmful. However, … such an assumption will be self-defeating” (ibid.).

Here, Simpson has essentially raised the problem of induction, and concludes that such reasoning is question-begging. The anti-theodicist has, thus far, observed and argued that theodicies have harmful consequences. From this observation, the anti-theodicist makes a general claim about theodicies, which Simpson believes goes beyond the observed. Since the observed theodicies have had harmful consequences, there is an expectation that the next theodicy will likewise have harmful consequences. But Simpson is objecting that the anti-theodicist cannot make inferences about theodicies that do not yet exist. In order to draw their conclusion, anti-theodicists must presuppose that future theodicies will have the same consequences as the previously observed. But if you presuppose that future theodicies will have harmful consequences as well, then you will have presupposed your conclusion, and you will have begged the question.

Would this mean that all inductive inferences, or general claims that go beyond the observed in trying to predict future observations, ultimately beg the question? All inductive inferences, as the problem of induction states, and begging-the-question are forms of circular reasoning. But in asserting that anti-theodicists beg the question by appealing to the problem of induction, Simpson has opened up a can of worms, and set a problematic standard for argumentation. Of course, one could try to meet the problem of induction head on, for example by arguing that inductive inferences from the observed to the unobserved still has pragmatic epistemic value, regardless of its circularity, or through the Popperian strategy, by arguing that the task of researchers is not to verify, but to see if theories live up to standards of falsification. Through the Popperian strategy, the anti-theodicist could assert that, since they have observed that past theodicies have had harmful consequences, they have developed the hypothesis that future theodicies will have harmful consequences as well. The task of the anti-theodicist would then be to test their hypothesis by seeing if it can be proved to be false by the next theodicy they encounter.

But there are other problems with Simpson’s argument. In the quote above, Simpson has extended the project of theodicy to include any potential future theodicies, and since anti-theodicists have not taken (indeed, could not take) future theodicies into account in their arguments and subsequent rejection of theodicy, Simpson accuses anti-theodicists of begging the question by presuming that future theodicies will fail as well. Simpson’s line of reasoning here endorses a standard for argumentation which I do not believe we should accept. On Simpson’s view, if you reject all of my theories, which favour my position, you can never conclude that my position is untenable, since you have not considered any future theory I might produce. For all we know, I might produce a successful theory in favour of my position tomorrow, and since this is always a possibility, you will not have considered all of my theories. Moreover, by including potential future theories into the conversation, you can never consider all of my theories. So, if you conclude that my position is untenable, you will have begged the question against my position by presuming that my future theories will fail as well (note that this line of reasoning can be applied to any debate). Indeed, this line of reasoning can be thrown back in the other direction: a defender of theodicy can never conclude that anti-theodicy is an untenable position, because, if they did, they would not have considered the tenability of any future anti-theodicy. Furthermore, simply by practicing or constructing theodicies, you will have begged the question against anti-theodicy, because you would be presuming that theodical practice is in itself unproblematic, and by extension, that all anti-theodicies—past, present and future—fail.

It seems to me that Simpson’s reasoning makes rejecting the theodical project an impossibility, since, according to his reasoning, even if all theodicies to date are wrong, the right one could be just around the corner. This reasoning could persist ad infinitum.

Snellman’s argument

In a 2019 debate with Toby Betenson, Lauri Snellman argues that moral anti-theodicies fail because they end up in question-begging moralism. The 2019 Snellman v. Betenson debate concerns anti-theodicies as described by Betenson in his 2016 paper, “Anti-Theodicy.” Below, Betenson provides a summary of Snellman’s argument:

[M]oral anti-theodicy stands accused of begging-the-question against theodicy by offering a moralistic judgement that is ‘detached from the system of reasons that exist in the world in question [i.e., the theistic world in which a theodicy is true]’. Simply put: if theism is true, and if a theodicy is correct, then the moral facts are not as the moral anti-theodicist believes them to be. There are, in fact, justificatory reasons for the permission of all evils. The moral anti-theodicist denies this. But if a theodicy is correct, the moral anti-theodicist is simply wrong. When they then denounce the morality of theodicy, they engage in a question-begging ‘moralising’ (Betenson, 2019: 214).

The core of the question-begging charge lies in the moral anti-theodicist’s denial that there are justificatory reasons for the permission of all evils, and when we read (Snellman’s, 2019) article, we find that he is relying on Betenson’s description of anti-theodicy when concluding that moral anti-theodicy begs the question. “[Betenson] admits that antitheodicy presupposes the moral claim that there are horrendous evils that cannot be justified” (Snellman, 2019: 215), he writes.

Snellman’s argument goes like this: in searching for God’s morally sufficient reasons for allowing evil and suffering to exist, the theodicist is presupposing the possibility that such morally sufficient reasons might exist. However, the anti-theodicist presupposes that there are some evils that cannot be justified; that, for some evils, morally sufficient reasons do not exist. This presupposition contradicts the theodicists’ presupposition. Therefore, the moral anti-theodicist is presupposing that the theodicist is wrong; that is, the moral anti-theodicist is presupposing that the theodicists cannot provide morally sufficient reasons for all evils.

In the following, I will first show why I do not agree with Betenson’s and Snellman’s understanding of moral anti-theodicy, i.e., that it presupposes that some evils cannot be justified. In viewing the context of how this understanding of moral anti-theodicy arose, I believe that this notion may have arisen as a misunderstanding. Secondly, I will argue that what Snellman refers to as a presupposition, is in fact the conclusion of anti-theodical arguments.

First, the claim that “there are evils that cannot be justified” is not a feature all moral anti-theodicies share, as seen above in the anti-theodical arguments Simpson has evaluated. Neither one of those moral anti-theodical arguments—the argument from insensitivity, the argument from detachment, and the argument from harmful consequences—rely on, or even discuss, whether or not all evils can be justified. Thus, the claim that moral anti-theodicy presupposes that there are evils that cannot be justified is not true. Furthermore, it is also noteworthy that Snellman does not provide any examples of actual moral anti-theodicies which do presuppose there are evils that cannot be justified. Rather, he is relying entirely on Betenson.

Betenson states that “the anti-theodical premise asserts that there are some evils that cannot be justified. As such, the key premise of anti-theodicy assumes that the conclusion of theodicy is false” (Betenson, 2016: 60). But it is unclear where he gets this notion of anti-theodicy. Before this quote, he discusses Kenneth Surin’s response to the notion “that we should remain open-minded about the possibility that horrendous evils … could be justified by appeal to greater goods” (ibid., 59). However, Surin is not stating that some evils cannot be justified. Rather, he is arguing that we should not engage in such speculation in the first place, because it is precisely such theoretical speculation that mediates a harmful praxis. Surin thus rejects all theoretical approaches to the problem of evil, and since entertaining the question of whether all evils can be justified or not would be to engage in such theoretical speculation, it is clear that Surin is not, in fact, entertaining such a question. Therefore, if I am right that Betenson get his notion of anti-theodicy’s key premise from Surin, then it seems that Betenson has misunderstood Surin’s objection.

But I do not, in fact, believe Betenson got his notion of moral anti-theodicy from Surin. Rather, I believe he got it from Simpson, since Betenson also briefly mentions Simpson’s argument that anti-theodicy begs the question. As shown above, Simpson also uses the phrase “key premise” of anti-theodicy, and since Betenson briefly refers to Simpson before himself using the same phrase, it is reasonable to conclude that Betenson got the phrase from Simpson. However, here is the problem: as shown above, for Simpson, the “key premise” of moral anti-theodicy is: there is an inherent moral impropriety in the discourse practice of theodicy. For Betenson, however, the “key premise” of moral anti-theodicy is: there are some evils that cannot be justified. So, even though both Simpson and Betenson are using the same phrase—the “key premise” of moral anti-theodicy—they are in fact referring to two different statements. So, from this discussion, I do not know where Betenson’s understanding of moral anti-theodicy comes from. What is clear, however, is that neither Betenson, nor Snellman, have provided any reasons for believing that their understanding of moral anti-theodicy is true.

Before continuing, I want to reiterate the importance of the distinction between the theodical debate and the anti-theodical debate. The former seeks to answer the Problem of evil and suffering. It constructs theodicies as theories or answers to the Problem, and critiques of these theories are also raised within this debate. Importantly, though, the critiques of theodicy raised within this debate try to debunk the explanations of evil that these theodicies provide, for example, by arguing that human free will is not sufficient to explain the incoherence between God and evil, since God could have created humans good to begin with.Footnote 1 As such, both theodicies and the criticisms proposed in this debate accept and rely on certain presuppositions—for example, (a) that it is unproblematic to attempt to find answers that vindicate God and justifies the existence of suffering; (b) that theoretical speculation on evil is harmless; (c) that we ought to strive for an objective perspective when theorizing about suffering—which make up the framework of the debate. Therefore, the theodical debate has, or is situated within, a certain framework, with certain rules. The latter debate, however—the anti-theodical debate—is not bound by the presuppositions (or rules) of the theodical debate, since it is situated outside of the theodical framework. The anti-theodical debate is concerned with the question of whether the theodical enterprise is problematic or not. It is therefore a meta-philosophical debate. As such, within the anti-theodical debate, the presuppositions of the theodical framework are still up for debate. The anti-theodical debate does not assume the negation of (a), or (b), or (c); if it did, then anti-theodicists would not feel the need to argue against them, just like theodicists do not feel the need to argue in favour of (a), or (b), or (c). Simpson, in his analysis of anti-theodical arguments, has argued in similar lines. He writes, “if we are willing to accept that all theodical explanations for evil are untenable, then there is just no need to bring further arguments against the discourse practice of theodicy” (Simpson, 2009a: 160).

Generally speaking, if you argue for something, then you do not presuppose it. So, since anti-theodicists actually do argue, for example, that some evils cannot be justified, that means that they do not presuppose it. Therefore, I believe that Betenson could be wrong when he states that “the anti-theodical premise asserts that there are some evils that cannot be justified. As such, the key premise of anti-theodicy assumes that the conclusion of theodicy is false” (Betenson, 2016: 60). This is not necessarily true. I would rather say that one anti-theodical conclusion states that there are some evils that cannot be justified, like the oft-quoted Ivan Karamazov, who argues that some crimes, like the imprisonment and unavenged torture of a small child, cannot be compensated for. Here, Ivan does indeed argue that some evils cannot be justified. But this is a conclusion, not a premise, as Betenson states. Rather, Ivan attacks a presupposition of theodicy, and by doing so, moves to a place where the presuppositions of theodicy are still up for debate – i.e., to a place outside of the theodical framework. Ivan’s argument thus illustrates Nick Trakakis’ understanding of anti-theodicy seen above, according to which anti-theodicies aim to show that theodicies are by their nature defective, and that their engagement with the Problem of evil rests on various problematic presuppositions.

Therefore, the claim that some evils cannot be justified is not a presupposition, as Snellman claims, but the conclusion of one argument, and to call it a premise, as Betenson does, might be misleading. If it is used as a premise in another argument, or a longer argument, which concludes that theodicy should be abandoned, then it is an intermediate conclusion of the whole argument. Therefore, moral anti-theodicy has not assumed that a presupposition of theodicy is false, but rather argued for it, and then using the conclusion of said argument to further argue that the claims of theodicy are false. What such an argument does presuppose, is that the issue of whether all evils can be justified or not is still up for debate.

Conclusion

In this paper, I have argued that Robert Simpson and Lauri Snellman have not managed to show that moral anti-theodicy begs the question. I have argued that Simpson’s charge that moral anti-theodicy begs the question stems from the mistaken assumption that both theodicy and anti-theodicy are situated in the same framework; that the theodical debate, and the anti-theodical debate, are one and the same. This mistake stems from not considering the meta-philosophical nature of anti-theodicy. When criticizing an anti-theodicy, we must recognize that we are engaged in the anti-theodical debate, which is situated outside of the theodical framework. This is because when discussing anti-theodicies or the presuppositions of theodical discourse practice, we are engaging in meta-theodical reflection, and thus not accepting the rules of the theodical framework. One cannot completely or adequately evaluate the theodical framework from within the theodical framework. This is because if we are situated within the theodical framework, then we are accepting that framework and its foundations, which are the very things being questioned by the anti-theodicist. Moreover, I have found that Simpson’s charge that anti-theodicy begs-the-question has the effect of shielding the theodical framework from criticism, essentially immunizing it from outside (meta)criticism, by demanding that any and all criticisms of theodicy must accept its rules and assumptions.

The problem I see with Snellman’s argument is his first premise—the claim that moral anti-theodicies presuppose that there are some evils that cannot be justified. However, he does not provide evidence of there existing such an anti-theodicy. To support his claim, he refers to Betenson, stating that Betenson admits that anti-theodicy presupposes the moral claim that there are some evils that cannot be justified. But as I argue, it is unclear where Betenson gets this notion, and Betenson does not provide evidence of there existing such an anti-theodicy either. I also, via my analysis of Simpson, provide examples of moral anti-theodicies which do not presuppose such a claim. Thus, Snellman’s and Betenson’s understanding of moral anti-theodicy relies on a claim that has not been substantiated or supported—a claim I have shown and argued to be erroneous.