A Wittgensteinian Antitheodicy

Abstract

Contrary to the majority of contemporary analytic philosophers of religion, James Sterba argues in his book Is a Good God Logically Possible? (2019) that Alvin Plantinga with his famous free will defense has not succeeded in solving the logical problem of evil. However, Sterba is not alone in disputing this generally accepted view in analytic philosophy of religion. D. Z. Phillips (1934–2006) has argued that the logical problem of evil has not been solved and he further holds that it has not even got off the ground. The aim of this article is to explore Phillips’ criticism of the free-will defense and mainstream theodicies. His critique is relevant for Sterba’s atheistic stance because Phillips’ arguments are partly applicable to forms of philosophical atheism that share the same assumptions with philosophical theism. In the first part of the article, I will briefly describe the starting points of the best-known solutions to the problem of evil in analytic philosophy of religion and refer to some aspects of Sterba’s arguments. After that I will explore Phillips’ ethical and conceptual criticism against frameworks used in the discussion of theodicy. Finally, I will pay attention to Phillips’ Wittgensteinian view of the task and the aim of philosophy in order to clarify some problematic aspects of his thought.

1. Introduction

In his book Is a Good God Logically Possible? James Sterba (2019) examines the problem of evil, which in the Judeo-Christian tradition is often understood as a question: Why does an almighty and benevolent God allow evil, suffering and injustice in the world? At different stages in the history of theology the problem has been addressed in various religious contexts. In the contemporary analytic philosophy of religion, which my paper deals with, the debate on the problem of evil reflects the questions of early modern and Enlightenment philosophy. The central question concerns the truth and rationality of theism in the light of the challenge of atheism and skepticism.

It is worth noting that the problem of evil does not necessarily have to be linked to the problems of Judaeo-Christian religion. Questions of evil and suffering also arise in other religious traditions—after all, teaching about suffering is at the heart of Buddhism, for example. On a more general level, the problem of evil as an existential problem affects people regardless of their religious beliefs. For example, since the Second World War, there has been much discussion about how extreme evil, such as the Holocaust, is possible (Bernstein 2002).

I will first briefly characterize different aspects of the problem of evil and proposed solutions to it in the contemporary analytic philosophy of religion. I draw special attention to the ways in which Sterba criticizes the apologetic arguments in this discussion. After that I explore D. Z. Phillips’ criticism of the mainstream approaches to the problem of evil. Phillips shares Sterba’s view that theodicies do not provide plausible moral grounds for an adequate solution to the problem of evil. However, Phillips’ criticism is more radical than Sterba’s, because according to him, ethical and religious perspectives related to theodicy are already so wrong in its starting points that they cannot be taken seriously. Phillips’ aim is not to defend atheism but to show that the debates regarding the problem of evil go astray and fail to have genuine religious and ethical meaning. In the last part of the paper, I will discuss Phillips’ Wittgensteinian “contemplative” conception of philosophy, according to which the task of philosophy is not to prescribe a certain form religious faith, but to do justice to the meaning of religious and ethical beliefs in human life.

2. Identification of the Problem of Evil

In analytic philosophy of religion the problem of evil is treated as an argument against the existence of God. The debate typically starts from certain shared assumptions adopted in both theistic and atheistic philosophy of religion. In his article “Evil and Omnipotence” J. L. Mackie (1955) formulated the problem of evil by referring to three propositions:

• God is omnipotent.
• God is wholly good.
• Evil exists.

Mackie argued that there seems to be a contradiction between these propositions; it is inconsistent to accept all three propositions at the same time. However, it is worth noting that he pointed out that:

the contradiction does not arise immediately; to show it we need some additional premises, or perhaps some quasi-logical rules connecting the terms ‘good’, ‘ evil’, and ‘omnipotent’. These additional principles are that good is opposed to evil, in such a way that a good thing always eliminates evil as far as it can, and that there are no limits to what an omnipotent thing can do. From these it follows that a good omnipotent thing eliminates evil completely, and then the propositions that a good omnipotent thing exists, and that evil exists, are incompatible.

(Mackie 1955, pp. 200–1)

This argument is called “the logical problem of evil”. It is often contrasted with “the evidential problem of evil”, according to which the existence of evil, e.g., the existence of actual human and animal suffering, counts against the truth of theism and makes it improbable, even if there is no obvious logical contradiction between propositions (1)–(3). There is also another important distinction that is worth keeping in mind. In analytic philosophy of religion, the problem of evil is understood as an intellectual—logical, metaphysical and epistemological—problem in contrast to the existential problem of evil. In the latter case, the problem of evil is seen as a personal, subjective and practical problem that has to do with how one can cope with evil and suffering in one’s life. When the problem is seen as an intellectual problem, it is treated from an abstract non-personal perspective; the intellectual problem concerns the validity and credibility of the argument from evil. The perspective of the debate is characterized by methodological ideals common in scientific inquiry, such as objectivity and the logicality of argumentation, coherence and the probability of claims that are used in developing philosophical arguments, etc. It is, indeed, striking that the debate on the matter has often been technically very sophisticated; philosophers have applied and developed various kinds of logical and analytical tools in defending their positions. The semi-scientific ideal linked with the use of technical logical and epistemological tools also characterizes works of philosophers such as Alvin Plantinga (2000) and Nicholas Wolterstorff (1983), who reject the idea of the Enlightenment scientific ideal of rationality and who think that Christian philosophers have their own methodological assumptions. Striving for objectivity in argumentation has to with the audience of debate. The audience of discussion involves philosophers whose worldviews differ. Although analytic philosophers may suspect that the arguments put forward in the debate succeed in convincing their opponents as to whether (or not) the existence of God is compatible with evil, this does not raise the question of the meaningfulness or ideological nature of the debate. Peter van Inwagen, who doubts that all philosophical arguments (in the light of the ideal demands he describes) are failures, however, points out that:

Now if it is indeed true that no philosophical argument for any substantive conclusion is successful in the sense that I have proposed, it immediately follows that the argument from evil is not a success in that sense—given, at any rate, two premises that I don’t think anyone would deny: that the argument from evil is a philosophical argument and that the non-existence of God is a substantive philosophical thesis.

(van Inwagen 2006, p. 53)

Thus, in this view, the existence and non-existence of God is a philosophical question that should be treated similar to other metaphysical questions in the field of philosophy. It is assumed that the idea of God—the idea of God’s existence and non-existence—is equally understood by the parties of the debate, and their task is to offer arguments which support the truth or rational acceptability of this “substantive philosophical thesis”. The existence of God is not a scientific question, but it is an intelligible metaphysical thesis that can in principle be solved by appealing to argumentative evidence.

3. Best Known Answers to the Argument from Evil and Sterba’s Response to Them

Plantinga (1974a, 1974b) has called into question the logical binding nature of Mackie’s argument by referring to the possibility that the theistic worldview includes a claim that does not create a logical contradiction between the allegations (1)–(3). According to his famous “free will defence”, it can be thought that a world with free beings is better than a world without free beings. In addition, a benevolent God does, of course, want to create that kind of world rather than a world of robots who always do good. Free will is of the utmost value, but it includes the opportunity to do both good and evil, and this provides a possible reason for God to permit evil in the world. The price is high, but God must pay it, because free choices are not possible without the possibility for evil acts. According to Plantinga’s libertarian view, free choices are not determined by previous events, circumstances and causal laws. It is person’s own will that determines his or her free choices. Plantinga does not seek to present the actual theodicy, that is, to provide a plausible or probable explanation why God allows evil, but he merely seeks to present a defense that shows that the logical problem of evil described above is not logically binding, for we can imagine reasons why benevolent God may permit evil. The purpose of this defense is to appeal to a possibility; Plantinga’s defense is speculation on logical possibilities. He only tries to show that the existence of evil is logically compatible with the existence of God.

Plantinga’s free will defense has its roots in Augustine’s theology. Another popular answer to the problem of evil is developed by John Hick (2001), who, in turn, uses Church Father Irenaeus’ thoughts in developing his theodicy. The starting point of Hick’s theodicy is the idea of the world as a “vale of soul-making”. The existence of evil things offers human beings a possibility to develop; the function of suffering is to refine people spiritually and morally. Richard Swinburne (1998), who, along with Plantinga and Hick, has been one of the most prominent figures in the debate, grounds his theodicy on the idea of free will and on the doctrine of the vale of soul-making. According to Swinburne, both moral and natural evil serve the overall purpose of God’s benevolence; they provide people with knowledge of good and evil, and provide an opportunity to learn how to help other people and learn to show compassion, etc. In Swinburne’s view, a benevolent God does not create a world without suffering. By allowing it, God gives both victims and their helpers the opportunity to grow morally. Tough conditions are required to develop our best features.

In addition to the free will defense and the vale of soul-making theodicy, one noteworthy apologetic solution in contemporary discussion is “sceptical theism”. Its basic idea can be explained by referring to an atheistic argument put forward by William Rowe (1979):

• If God existed, no intense unnecessary suffering would be experienced by humans and animals.
• Such suffering exists.
• Therefore, God does not exist.

Rowe does not claim that the argument is logically binding, but he says that premises (1) and (2) of the argument are likely to be true and rationally credible. However, skeptical theism denies premise (2). Although it seems that there is unnecessary intense suffering, we cannot conclude that there might not be reasons for allowing it. There is a great difference between the knowledge of man and that of an omniscient God. Stephen Wykstra (1984), a representative of skeptical theism, argues that an almighty and omniscient God, if he exists, knows more than a person whose cognitive abilities are limited and deficient. The epistemic situation between human beings and God is analogous to the situation between young children and parents. Parents know better than their children what is good for them. Although we, as cognitively limited human beings, do not know God’s reasons for allowing evil, he may have good reasons for it.

The above-mentioned topics have given birth to a wide-ranging discussion, and Sterba’s book is a contribution to it. According to Sterba, a major flaw in the current debate has been that many analytic philosophers of religion—both atheists and theists—believe that Plantinga’s defense of free will has succeeded in solving the problem of logical evil. Sterba does not, however, deny that Plantinga may be right in holding that it is logically possible that an all-good and all-powerful God could create a world with free creatures with some moral evil. In this respect, Plantinga’s defense may be relevant for Mackie’s atheological argument. However, Sterba thinks that free will defense is not plausible in the actual world or in possible worlds with significant horrendous evil:

My primary thesis here is simply that the freedom that exists, or has existed, in our world could not constitute a justification for the moral evil that exists, or has existed, in it. However, my secondary thesis is that Plantinga has not succeeded in showing that God is logically compatible even with some evil in the world, when that evil is taken to be, as it may well be, any of the significant and espe-cially the horrendous consequences of our immoral actions.

(Sterba 2019, pp. 11–12)

Sterba’s criticism is not only aimed at Plantinga’s views, for he also extensively criticizes other forms of contemporary theodicies. His central aim is to contribute to existing debates in this field by using and developing resourses from ethical theories. They provide tools that show the central weaknesses of theodicies; after all, the problem of evil is an ethical rather than a logical, metaphysical, or epistemological problem (Sterba 2019, p. 5). What is especially original in Sterba’s thought is the appeal to ideas of welfare liberalism in order to reveal the weakness of the free will defense (and weaknesses of theodicies). Sterba assumes that that there is a kind of analogy between the governance of society and the world governed by the ruler of the universe. Major problems with the free will defence emerge when we examine it and the actual world in the light of the ideals of a just society. Namely, the free will defense does not take into account the idea that the restriction of freedom has a central role in a society that works in the fairest way for everyone. According to welfare liberalism, a just society offers the right to significant freedom and well-being for everyone, e.g., for both rich and poor, and for this reason society restricts people’s freedom in various ways. In the light of this ideal, Sterba asks us to imagine “superheroes” who can influence society much more effectively than we can. By preventing various kinds of injustice, inequality, poverty, violence, and crimes, etc. Sterba believes that we expect these morally good fictional characters with their special abilities to make societies more just than our actual societies are. In this way they bring about a better distribution of significant freedom, for example, for those who are attacked by wrongdoers. This and much more is, of course, exactly what one would expect an all-good, all-powerful God to do. Thus, Sterba argues that:

the problem is not with God’s creating us and giving us free will. Rather, the real problem comes later in time when God fails to restrict the lesser freedoms of wrongdoers to secure the more significant freedoms of their victims. Hence, the world we live in cannot be justified by the distribution and amount of significant freedom that is in it.

(Sterba 2019, p. 29)

Sterba appeals to the idea that if there is a morally perfect and omnipotent God, then he would necessarily be expected to follow “Moral Evil Prevention Requirements”:

• Prevent rather than permit significant and especially horrendous evil consequences of immoral actions without violating anyone’s rights (a good to which we have a right) when that can easily be carried out.
• Do not permit rather than prevent significant and especially horrendous evil consequences of immoral actions simply to provide other rational beings with goods they would morally prefer not to have. [By these Sterba means goods such as receiving medical care after being brutally assaulted. No doubt, it would better to prevent the need for these kinds of goods rather than to have them.]
• Do not permit rather than prevent significant and especially horrendous evil consequences of immoral actions (which would violate someone’s rights) in order to provide such goods when there are countless morally unobjectionable ways of providing those goods. (Sterba 2019, pp. 126–28)

Sterba holds that an omnipotent God could easily prevent the evil consequences of immoral actions. Thus, if God were adhering to these requirements, then he has not permitted the existence of significant and especially horrendous evil consequences of immoral actions that obtain and have obtained in the actual world.

4. Antitheodicy

Sterba is not alone in disputing the generally accepted view in contemporary analytic philosophy of religion that Plantinga has succeeded in solving the logical problem of evil. D. Z. Phillips argues that neither Plantinga nor anyone else has succeeded in solving the problem as it was formulated by Mackie. Phillips’ approach, however, differs sharply from the way theistic analytic philosophers approach the problem of evil as he questions the basic assumptions of the debate:

if we stay within the terms of reference in which the logical problem of evil is usually discussed, we shall find that neither the proposition ‘God is omnipotent’, nor the proposition, ‘God is perfectly good’, can get off the ground—and that for logical reasons.

(Phillips 2004b, p. 5)

In what follows, I deal with Phillips’ critique of theistic analytic philosophy of religion. However, his critique is also applicable to philosophical forms of atheism that share the same assumptions. In this respect, his arguments are relevant also for Sterba’s approach, for it shares many philosophical and theological presuppositions that give a form for debates over the problem of evil.

In addition to Phillips, there is a minority of thinkers in Anglo-American philosophy of religion, such as Andrew Gleeson (2012), Terrence Tilley (1991), Kenneth Surin (1986) and John Roth (2001), who see serious problems in the way in which analytic philosophers usually deal with the problem of evil. The common feature of all these “antitheodicists” is that they hold that the problem with the contemporary analytic discussion of the problem of evil is not only that the particular solutions proposed to the problem of evil are unsatisfactory, but the whole project of trying to justify the existence of suffering and evil in the world is mistaken. An antitheodicist rejects the fundamental presuppositions and conceptual parameters that guide theodicies. In spite of the differences between their approaches, one could say that ethical criticism plays a key role for antitheodicists in rejecting mainstream theodicy. A good example of this is Kivistö and Pihlström’s (2016) extensive study, Kantian Antitheodicy: Philosophical and Literary Varieties. Their criticism against theodicy is based on Kantian “transcendental criticism” of the various forms of theodicy. The effort to justify the horrors of the world threatens the very possibility of a moral perspective. Kivistö and Pihlström use the term “theodicy” in a very broad sense to mean any kinds of views—religious as well secular—which try to offer justification and legitimation for horrors. Antitheodicy rejects all totalizing perspectives in which the suffering of individuals is instrumentally placed to serve some overall plan. Horrors have no sense and antitheodicy acknowledges the inconsolability of life and the futility of suffering.

Phillips is one of the best-known thinkers representing antitheodicism in contemporary Anglo-American philosophy of religion. His criticisms of theodicies is closely related to other aspects of his philosophical approach. His view of philosophy and the ethical and religious spirit of his thought—which is strongly influenced by Wittgenstein and Rush Rhees—differs sharply from contemporary analytic philosophy of religion. His most systematic presentation on the subject is The Problem of Evil and the Problem of God (Phillips 2004b). In the first part of the book, he deals critically with views of analytic philosophers of religion, such as Swinburne, Hick, Plantinga and Stephen T. Davis. The second part of the book deals with “a neglected” tradition in Christianity, which offers a different understanding of faith in God and its relation to human suffering. In describing this alternative approach, he often refers to Simone Weil’s writings. Phillips thinks that Weil’s struggle “to be nothing before God” expresses a deep understanding of what Christian faith can be without any expectations of compensation or consolations offered by theodicies. In this paper I will not deal with Phillips’ account of this neglected tradition and shall instead concentrate on criticisms of theodicies and defenses in contemporary analytic philosophy of religion.

5. Moral Issues

One of the main problems with theodicy, according to Phillips, has to do with the consequentialist and instrumentalist approach to moral issues. Analytic theists typically defend the view that the reason for God allowing human and animal suffering is the possibility of achieving a greater good. God cannot bring about this greater good unless he allows suffering. Allowing for evil is a necessary part of God’s overall plan, and this overall plan justifies the existence of evil in the created world. From this point of view, there is no unnecessary evil, because the possibility of moral and spiritual development requires that evil things happen in the world. Evil things that happen to others are potentially morally useful for us. Swinburne writes:

And when we have not ourselves had such experience we can freely choose to seek out those who have before coming to form a view about the moral principles involved. The suffering becomes the tool which we use for our growth of moral understanding, and so in yet another way the sufferer is of use to us in helping us so to grow.

(Swinburne 1998, p. 168)

Phillips finds this argument senseless. When sufferings of others are made instrumental to us, it can be said, for example, that God allows the man on the road to Jericho to be abused so that the Good Samaritan could act virtuously. The Good Samaritan could say “Thank you, God, for another possibility to be responsible” (Phillips 2004b, p. 59). It is difficult to see why one should take this kind of theodicy seriously. The effort to seek morally adequate arguments to show why God can allow (or could possibly allow) evil, and especially horrors, such as the Holocaust, is absurd. The problem is not in the details how this greater good can be conceived, for it is the whole moral framework of the debate that is distorted. Defending omni-God’s actions does not do justice to our customary moral perspectives.

Thus, both Sterba and Phillips share the view that theodicies do not provide plausible moral grounds for an adequate solution to the problem of evil. However, Phillips’ criticism is more radical than Sterba’s, because according to him, the consequentialist ethical perspective related to theodicy is already so wrong in its starting points that it cannot be taken seriously. The problem with the theodicist’s ethical system is not that it is prepared carelessly, but that its basic premises and ways of posing questions are totally incorrect. “The commitment to theological consequentialism, which asks us to be open-minded about the possibility of a balance for good being on the side of allowing the Holocaust, is itself a corruption of the notion of open-mindedness.” (Phillips 2004b, p. 120.).

One central problem with apologetic philosophers concerns the examples they have used in defending their theories. Although one can sometimes say that someone has learned morally important things through suffering, one cannot ignore the opposite cases. Phillips argues that the problem with the vale of soul-making theodicy is the use of one-sided examples. It is very problematic in general terms to talk about our own suffering and, in particular, about the suffering of others as an opportunity for ourselves to develop. Although in some cases this kind of talk may make sense, but it is obvious that suffering is not always linked to the realization of some greater good. It does not make much sense to say that statistically God’s programme for developing the ethical and spiritual life of the human race has worked well. Phillips invokes the counterexamples he finds abundantly in literature, (e.g., from Thomas Mann, W. Somerset Maugham, Primo Levi). One example comes from Maugham’s (1948) The Summing Up. Maugham recalls the times when he studied medicine:

At that time (a time to most people of sufficient ease, when peace seemed certain and prosperity secure) there was a school of writers who enlarged upon the moral value of suffering. They claimed that it was salutary. They claimed that it increased sympathy and enhanced the sensibilities. They claimed that it opened to the spirit new avenues of beauty and enabled it to get into touch with the mystical kingdom of God. They claimed that it strengthened the character, purified it from its human grossness and brought to him who did not avoid but sought it a more perfect happiness … I set down in my note-books, not once or twice, but in a dozen places, the facts that I had seen. I knew that suffering did not ennoble; it degraded. It made men selfish, mean, petty and suspicious. It absorbed them in small things. It did not make them more than men; it made them less than men; and I wrote ferociously that we learn resignation not by our own suffering, but by the suffering of others.

(Maugham 1948, p. 259; Phillips 2004b, p. 68; See also Rhees 1997, pp. 149–50)

In the light of Maugham’s observation, the greater good theodicy—the search for justified reasons for allowing suffering—would be ridiculous. Can it be imagined that Maugham would have changed his view after a careful study of current arguments in analytic philosophy? Phillips considers that you do not need philosophical arguments to show the absurdity of thought that horrors are justified and necessary, because of the great value of free will or because horrors offer us—but not necessarily to the victims—a possibility to become morally good and more morally developed persons.

There are, of course, many different ideas about the ennobling effect of suffering. An anonymous referee of this article made an interesting point about Nietzsche’s thought that suffering is necessary to human greatness. Nietzsche’s ethics has of course nothing to do with the justification of theism, but it brings out the complexity of the matter. However, Phillips does not to categorially deny that suffering can ennoble one’s character. For his purposes it is enough to point out that it is extremely difficult to find such a generalization credible, and it is especially incredible in the case of horrors.

One illuminating example of Phillips’ criticism concerns the use of the idea of free will in the theodicy debate. Phillips refers to a story told by Billie Holliday. It is about the fate of the wife of a famous jazz musician who was a drug addict. In a desperate situation, he needed drugs to perform a show and his wife was afraid that he would kill himself if he did not acquire the drug. She tried to protect her husband and went out to find the drugs for him and was arrested. She was not a user but in this difficult situation she took the drug in order to prove to the law that she was a user not a pusher. In this way she could protect her husband and also herself, because as a pusher, she would have received a longer jail sentence. Holliday ends the story, “And that’s the way she got hooked. She’s rotting in jail, right now. Yes siree bob, life is just a bowl of cherries” (Holliday and Dufty 1975, p. 93; Phillips 2004b, p. 73).

Billie Holliday sees the fate of the wife not as a story in which the wife used her freedom of will in a wrong way and but as a case where a terrible thing happened to her in trying to help her husband; Holliday says, the wife “was innocent and clean as the day she was born” (Holliday and Dufty 1975, p. 93). Richard Swinburne in commenting on this story said that the production of this example settles nothing, because we “don’t know some of the crucial factors involved”. For example, it is possible that the musician and his wife could have resisted (perhaps on some later occasions) the temptation to use the drug, and, if so, they share some responsibility for their troubles (Swinburne 1977, p. 129; Phillips 2004b, p. 73). Phillips says that his intention in the production of the example is not to settle anything, but simply to “put forward the possibility that things could be as Billie Holiday says they were” (Phillips 2004b, p. 74). The central weakness in Swinburne’s response has to do with the denial of things many of us have seen. In appealing to ignorance on the contingent facts or on ignorance of the philosophical problem concerning free will, a theodicist simply denies the possibility—which is obviously real for us in the context of ordinary life—that things sometimes go in a direction that is not under our voluntary control. Appealing to the ignorance of “the crucial factors” is a denial of these very common experiences. The abstract and general talk about certain moral principles that is meant to support theodicies are based on one-sided and problematic examples. This is also illustrated by the implausible parallels which theodicists have offered in order to defend their instrumentalist logic; for example, allowing a child to suffer during a visit to the dentist is equated with allowing horrors to happen (Phillips 2004b, p. 36; Swinburne 1998, p. 10).

Phillips points out that that the weakness of theodicists’ philosophical speculations is revealed in the fact that those who develop this line of thought do not speak of horrors in ordinary contexts in the same way as they speak about them when developing their theories. Think, for example, about the idea that the genocide of the Jews might be related to some greater good. Obviously, this would be morally offensive in any normal discussions; these kinds of theodicists’ speculations are out of the question in ordinary contexts. The contradiction between ordinary moral reactions and philosophical speculation shows that the language of theodicies loses its connection with ordinary moral realities. The central target of Phillips’ criticism is the confused role of scepticism associated with theodicists theories. The problem with these theories concerns the confused meaning of words not our ignorance of some hidden knowledge of God’s benevolent purposes for us. This confusion becomes apparent especially in the light of the darkest cases, such as the Holocaust, where the idea of evil as a “necessary means to a good end” makes no sense. We do know that the death camps were useless and unjustified. Questioning this, saying that we cannot be sure about that, is morally absurd. Thus, it can be said that Phillips relies on mundane moral realism and what we know about moral matters (Phillips 2004b, pp. 37–38). Speculation for reasons beyond comprehension that serve a good divine purpose leads to radical moral scepticism. This is an obvious problem with sceptical theism. Saying that God is justified in allowing horrors because it makes it possible for him to achieve the best possible result although we do not know what it is, signifies a shift beyond morality.

Phillips’ arguments and his ordinary moral realism is connected to his view of the relationship between metaphysical realism and radical scepticism. A radical global sceptic assumes that we can meaningfully question our practices from outside them; it makes sense to ask if a practice is true or false, rational or irrational independently of actual practices (Phillips 2000, Chap. 3; Ashdown 2002, Chap. 1). This is also metaphysical realism’s basic assumption; there is a truth as such independently of human practices. Most analytic philosophers of religion seem to assume metaphysical realism, and this assumption plays an important role in discussions concerning theodicies and theistic metaphysics in general.

A metaphysical externalist perspective on human practices is, in Phillips’ view, based on a confusion. It misleading to think that human practices—in the Wittgensteinian sense of the word, i.e., forms of life and language games that are part of them—are grounded in beliefs that give support to these practices from the external perspective. The central problem is the idea of objectivity that metaphysical philosophers try to seek in developing their theories. According to the externalist picture, the task of philosophy is to try to acquire a more objective understanding of some aspects of reality, e.g., morality, religion or science. This is accomplished by questioning what our established practices and ways of using language are. These practices belong to the Appearance which is contrasted to the Reality philosophy seeks to find. Wittgenstein rejected this externalist picture of the relation between ordinary life and philosophical theories and claims, “What we do is to bring words back from their metaphysical to their everyday use” (Philosophical Investigations, § 116). The crux of the criticism can be formulated as follows: the talk about God’s hidden, unknown moral reasons for allowing horrible things to happen is just idle talk that divorces moral concepts from the human life and practices in which these concepts have sense. It makes no sense to say that allowing the Holocaust is morally permitted in the light of moral standards that are wholly beyond our comprehension. The problem with theodicies is not the ignorance of God’s great plan, but meaningless talk about morality outside of actual moral practices and moral language.

6. The Concept of God

The notion of God is also at the heart of the difference between Phillips’ and mainstream analytic theists. According to Phillips, there are serious confusions in the way in which the concept of God is understood in contemporary philosophy of religion. The central problem lies in anthropomorphic ways of speaking about God as a moral person whose actions can be subject to our moral judgements. The anthropomorphic assumptions lead philosophers to ask questions that distort the meaning of religious concepts. It is a confusion to treat God as a being among other beings and as a moral agent among other moral agents. Therefore, it is not intelligible, for example, to speak about God’s covenant with his people as a contract in which God stands “in reciprocal relations of rights and obligations to the other parties to the contract”. It makes no sense to think of God as “part of a community of criticism and counter-criticism”, as theodicists do (Phillips 2004b, pp. 148–51).

The idea that God is identified with a morally perfect being plays a central role in apologetic theism. Phillips argues that the idea of a perfect being encounters severe difficulties when it is linked with an effort to show that this being has or might have morally acceptable reasons for his actions. Phillips illuminates these difficulties by referring to an example from William Styron’s novel Sophie’s Choice (Phillips 2004b, pp. 41–44; see also Roth 2001). In this fictional story, a Polish survivor of Auschwitz, Sophie Zawistowska, a mother of two children, Eva and Jan, is forced at the gates to Auschwitz by an SS official to choose life for one of her children and death in the gas chamber for the other. In this extremely dark situation, Sophie screams “I cannot choose”, but she lets Eva go. The story continues gloomily. Eva dies in a gas chamber and Sophie never learns what happened to Jan. Sophie cannot live with her decision and kills herself after the war. In considering the story, Phillips notes that we cannot, of course, condemn Sophie from the third person’s perspective. For outsiders, the proper attitudes for her are pity and compassion. However, things look different from Sophie’s own perspective, as her decision has stayed with her. “She is involved in a moral tragedy, where, whatever she did, would involve evil”. Phillips says that “Sophie never thinks of handing Eva over as an act to be excused in the light of the total situation” (Phillips 2004b, p. 22). Now, Phillips argues that under the assumptions of analytic theism, God as a morally perfect being is seen to be a member of a moral community, and in this sense is similar to Sophie. As a moral agent, God has allowed horrible things such as the Holocaust to happen, and according to theodicies and defences he has morally sufficient reasons for that. In the light of the parallel between Sophie’s case and God, the talk about God’s perfect goodness appears strange. Phillips asks:

Is God to be the object of pity? Is creation a moral tragedy in which God is necessarily involved in evil? And what of God’s view of what he has done? Does the Holocaust stay with him? Does he think that it can be excused in the light of the greater good that made it necessary, or does he recognize he has something to answer for? It will be obvious that within these moral parameters, there is no logical space for talk about God’s perfect goodness.

(Phillips 2004b, p. 43)

We are all familiar with situations in which there are only bad alternatives. However, the talk about a morally perfect person, who has allowed the Holocaust to exist, does not make sense.

Theodicists treat God as a person and as a moral agent among moral agents, and they believe that this person can be put on trial and that the task of a philosopher is to evaluate how well or how poorly God has coped with this test. This picture of God is based on conceptual confusions concerning the grammar of God in the Christian tradition. Phillips appeals to Rowan Williams, according to whom, in orthodox Christian thought, God is not understood as a being who acts in a punctiliar way or as a being who reacts to things. According to Williams, “God is not an item in any environment, and God’s action has been held, in orthodox Christian thought, to be identical with God’s being—that is, what God does is nothing other than God’s being actively real” (Williams 1996, p. 143; Phillips 2004b, pp. 150–51).

Catholic philosopher, Brian Davies, whose thinking is based on the philosophy of Thomas Aquinas, has criticized contemporary analytic discussion on the problem of evil partly on the same grounds as Phillips and Williams. Many contemporary analytic philosophers understand the notion of God in terms of “theistic personalism”, which is based on the idea that God is a person, and the concept of person, in turn, is explained by referring to characteristics of human persons. Davies argues that this anthropomorphic conception of God differs from classical theism, according to which God is not a person similar to us; the creator of the world is radically different from creatures. Therefore, Davies claims that it is a conceptual confusion to treat God as a moral agent and as a member of a moral community (Davies 2008). It is here worth noting that in many parts of his book Sterba assumes that theism includes the idea of God as a moral agent, and he relies on this idea in developing his criticism of theodicies. Sterba also devotes an entire chapter to a critique of Davies’ views (Sterba 2019, Chap. 6). However, Sterba’s criticism of Davies’ thought is not applicable to Phillips’ antitheodicy. Although there are interesting similarities between Phillips’ and Davies’ views, their philosophical traditions differ in many ways. Metaphysics plays a central role in Davies’ Thomistic thought, but Phillips, following Wittgenstein, has provided a radical criticism of metaphysics. Thus, Davies, for example, defends classical theism by appealing to a version of the cosmological argument. He argues for the truth of a metaphysical view associated with Christian theology. Phillips, in turn, does not speak about a belief in God in terms of a first cause, and he also holds that the task of philosophy is not to offer solutions to religious and theological questions. Instead, the task of philosophy is to offer an understanding of what kind of questions are being asked, and to expose various kinds of conceptual confusions concerning religious matters (Davies 2007; Koistinen 2017).

Although perhaps all analytic philosophers agree with Phillips and Williams that God is not similar to human beings or physical entities, the models of physical language nonetheless play significant roles in their discussions. Religious concepts are analysed by referring to the concepts we use when we speak about the actions of human beings, or religious concepts are analysed by referring to some abstract and general core meaning of these concepts. For example, Swinburne states that “An omnipotent being is one who can do anything logically possible” (Swinburne 1998, p. 3). Phillips argues that the abstract definition is inaccurate, for there are obviously plenty of things God cannot do. To use his (somewhat ridiculous) examples, God cannot ride a bicycle or eat ice-cream. God—or, perhaps one should add, the first person of the Trinity—who is spirit, cannot do these kinds of things (Phillips 2004b, p. 12). Phillips’ claims that it is a conceptual mistake to think of “God’s omnipotence” as if it had a meaning outside the context of its religious uses. It makes no sense to ascribe “all power” to God, because the term “power” means different things in different contexts. Therefore, it is misleading to start the discussion of what God’s omnipotence means by offering an abstract definition that is independent of religious contexts. This is, of course, based on Phillips’ Wittgensteinian view of meaning; words gain their meaning from the context in which they are used, and the aim of the philosopher is to remind us of the ordinary uses of these terms when we are philosophically confused about their meaning. Accordingly, Phillips points out that the meaning of God’s omnipotence is found in the religious life in which it is used. God’s power is a special kind of power. It is not the ability to do anything that someone x just happens to want to do, i.e., the abstract definition of power, such as x is omnipotent = x has an ability to bring about any logically possible state of affairs leads philosophers or religion astray. In a discussion concerning natural theology, Rush Rhees illuminates this problem as follows:

Suppose you had to explain to someone who had no idea at all of religion or of what a belief in God was. Could you do it in this way?—By proving to him that there must be a first cause—a Something—and that this Something is more powerful (whatever this means) than anything else: so that you would not have been conceived or born at all but for the operation of Something, and Something might wipe out the existence of everything at any time? Would this give him any sense of the wonder and glory of God? Would he not be justified if he answered, ’What a horrible idea! Like a Frankenstein without limits, so that you cannot escape it. The most ghastly nightmare!

(Rhees 1997, p. 36)

The power of God is not worldly power. It is not the power of a dictator or the devil, but God’s power is the power of love. As Phillips argues, “God does not have two separate attributes, power and love, but that the only power God has or is, is the power of love” (Phillips 2004b, p. 199). The way one uses the words “love” and “power” in religious language differs from the way they are used in mundane contexts. This is also related to Rhees’ and Phillips’ general idea of the special kind of logic of religious language. The logic of ordinary descriptive language which we use in speaking about physical objects and human beings is different from the logic of the spiritual language which we use in speaking about God. In describing a human person, we can make a distinction between a person and his contingent properties; a person may or may not be loving or red -haired, etc. However, the ordinary subject–predicate distinction does not apply in this way to God. God’s “goodness” and “love” do not work similar to descriptive predicates, but they are God’s “grammatical predicates”, which are not “related contingently to ‘God’ but are instances of what the word ‘God’ means” (Phillips 2004b, p. 188). As Rush Rhees points out, “Winston Churchill may be Prime Minister and a company director, but I might come to know him without knowing this. But I could not know God without knowing that he was the Creator and Father of all things” (Rhees 1997, p. 48). Similarly, one cannot know God without knowing that God is love. Phillips suggests that “the point could be put by saying that, in certain contexts, ‘creator’, ‘grace’ and ‘love’ are synonyms for ‘God’” (Phillips 2004b, p. 190).

Thus, in Phillips’ views, it is a conceptual confusion to start the discussion of the problem of evil in Christian theology by assuming the abstract idea of God as a limitlessly powerful Being who may cause or bring about any logically possible state of affairs. Instead, the key concept for an understanding of God’s power is love. Phillips uses the Christian story of the Suffering Servant and the Passion of the Christ as explanations for what this means:

The Crucified Christ is not resurrected with healed wounds. Those who taunt him on the Cross, urging to him to save himself and thereby prove that he is the Son of God, fail to understand the only omnipotence God has is the omnipotence of love. It is from such a love that the prayer for forgiveness for the oppressors comes from the Cross. ‘Forgive them, Father, for they know not what they do.’ But the price of such love is that it can be broken, sacrificed. And it is that sacrifice which is raised up, exalted for all, eternity.

(Phillips 2004b, p. 272)

7. Philosophy and Theology

Phillips’ approach to the problem of evil goes in a completely different direction than the mainstream of philosophy of religion. In his view, philosophical discussion of the problem of evil suffers from conceptual confusions and his criticism also includes ethical and religious aspects. His criticism of consequentialism in theodicies is infused with a strong moral tone. He argues that “friends of religions”, i.e., apologist philosophers, have caused great damage to religion (Phillips 2004a). His own personal spiritual and moral sensibilities are strongly present in these criticisms. Some have seen this as a problem with his “contemplative” conception of philosophy of religion (Phillips 1999, 2001). Namely, following Wittgenstein, Phillips considers that philosophy only describes the actual use of language and “leaves everything as it is” (Philosophical Investigations, § 124). Phillips contrasts his contemplative approach to philosophy with normative traditions in the philosophy of religion. The task of the Wittgensteinian contemplative philosopher is not to defend religious and moral views in the name of philosophy. In Phillips’ view, these kinds of efforts are grounded on a misleading conception of the task and limits of philosophical enquiry. Philosophy is neither a judge nor an arbiter of truth and rationality in religious and ethical matters. Thus, Phillips makes a sharp distinction between theology and philosophy. Theologians are “guardians of religious pictures”; their job is to take a stand on the truth of religious beliefs (Phillips 2004a, p. 87). A philosopher, in turn, is interested in the sense and intelligibility of religious language and beliefs. Therefore, as a philosopher he is only offering a clarification of religion and not prescribing a certain form of Christianity. The task of the philosophy is to do justice to the variety of religious beliefs in human life.

It has been argued, however, that Phillips is not practising what he preaches; according to several writers, Phillips’ criticism of theodicies and metaphysical theism and his own account of religious beliefs is an expression of one particular form of religious belief (Moore 2005; Swinburne 2001; Wolterstorff 2001). It has been argued that his account is not a description of the religious faith but a prescription of one form of faith, and analytic philosophers represent an alternative form of Christianity. There has also been discussion about how Phillips’ account of a neglected tradition in Christianity (an account strongly influenced by Simone Weil and Kierkegaard) is related to different doctrines and schools of thought in Christian theology (Thomas 2001; Kurtén 2007; Phillips 2007a, 2007b; Von der Ruhr 2007; Davies 2007, 2008; Schönbaumsfeld 2007; Koistinen 2017, 2018). I consider these questions relevant in evaluating Phillips’ views, but I cannot go deeper into such issues now. I will, however, refer to a few important points in order to clarify his position.

Here one must pay special attention to the idea of grammatical/conceptual investigation. According to the Wittgensteinian view, the ultimate appeal in conceptual inquiry is not philosophical theories but the ordinary or common uses of religious language. The task of philosophy is to look at how language is used in its ordinary contexts. This does not mean that a philosopher seeks to find the meaning of the religious beliefs and language by Gallup poll. In describing meaning one cannot proceed simply by asking religious people or theologians what they are saying. The conceptual investigation does not leave a confusion as it is. Thus, what believers (and philosophers who are believers) say when they try to give an account of their beliefs is not automatically warranted. Similarly, it would be absurd to give a philosophical account of ‘thinking’ simply by asking ordinary people what they mean by the term. In clarifying what religious beliefs amount to, reference is not made to the account believers would give if asked. Instead, reference is made to the role the words play in believers’ lives (Phillips 2004a, p. 7). In other words, there is a distinction between (a) simply describing what believers say when they try to explain the content of their faith, (b) describing “the logic of language” in the Wittgensteinian sense in order to look at the meaning of words and sentences from the contexts and practices in which the words and sentences play a role.

Philosophical debates differ from scientific debates, where interlocutors agree on the content of hypotheses or theories and disagree about their truth value according to the common measure of meaning. The starting point for philosophical discussion, in turn, is a situation where we are confused about the meaning of expressions. As Wittgenstein says, “Philosophy is a battle against the bewitchment of our intelligence by means of our language” (Philosophical Investigations, § 109). In battling against bewitchment, a contemplative philosopher of religion does not, as sociologists of religion may do, appeal to new scientific information or, similar to many philosophers and theologians have accomplished, appeal to ultimate metaphysical principles and categories. The criteria of meaningful and unmeaningful is found in the actual religious uses of language. “Yet”, as Phillips points out, “although it is our language that bewitches us, the remedy lies in reflection on that same language” (Phillips 2004a, p. 7). The clarificatory work is not carried out from “outside” religious language, for religious language games and forms of life are the given contexts in which the sense of religious beliefs is found. Philosophy offers no “deeper” explanatory theories that explain the meaning of religious beliefs or why people actually hold such beliefs.

Clarifying religious beliefs requires some understanding and sensitivity to religious forms of life. For his part, Phillips thinks that to understand religious faith is not to confess that faith and appropriate it personally. In this regard there seems to be a tension in Phillips’ thinking. This has to do with the possibility of transcending one’s own personal perspective when engaging in philosophical study. A philosopher is a human being whose own personal religious/non-religious perspectives and the particular cultural (religious) context in which he lives limit his understanding in many ways. Phillips was aware of this problem, but it is somewhat unclear whether Phillips had a satisfactory solution to the problem (Phillips 2001, pp. 318–19; Edelman 2009; Amesbury 2007; Koistinen 2012). Be this as it may, Phillips made it clear that he did not deny genuine differences between various forms of religious beliefs. Contemplative philosophers leave genuine religious differences and disagreement as they are; they do not try to solve these issues, but they do try to do them justice. Phillips, however, believes that contemporary debates on the problem of evil need other kinds of attention. He holds that contemporary philosophers of religion deal with the problem of evil in a way that distorts the meaning of moral and religious beliefs. In trying to show this, he appeals to examples intended to illuminate some actual features in religious and moral life that he assumes the interlocutor will recognize.