Comparing Three Twentieth-Century Philosophical Antitheodicies

Abstract

This paper compares three twentieth-century examples of antitheodicist thought in the philosophy of religion (and, more generally, ethics): William James’s pragmatism, D.Z. Phillips’s Wittgensteinianism, and Emmanuel Levinas’s post-Holocaust ethical reflection on suffering. It is argued that all three—despite their enormous differences, given that the three thinkers discussed come from distinct philosophical traditions—share the fundamental antitheodicist argument according to which theodicies seeking to justify God’s reasons for allowing the world to contain horrible evil and suffering amount to morally problematic, or even immoral, failures to acknowledge other human beings and their meaningless suffering. Furthermore, it is suggested that this antitheodicist line of thought shared by all three is based on a Kantian transcendental analysis of the necessary conditions for the possibility of occupying a moral perspective on the world.

1. Introduction

The tensions as well as interactions between different traditions or “schools” of twentieth-century philosophy have been as clearly present in a special field such as the philosophy of religion as they have been in the core areas of theoretical and practical philosophy. So-called analytic and Continental philosophers of religion have seldom engaged in fruitful communication with each other. Nor has, for example, pragmatism been taken very seriously among, say, mainstream analytic philosophers of religion over the past few decades, even though one of its major classics, William James, devoted a considerable chunk of his oeuvre to issues concerning religion.

This essay will offer a brief comparative examination of three important traditions—exemplified by three major individual philosophers—in relation to one of the classical issues in the philosophy of religion, viz., the problem of evil and suffering. I will try to show that all these three approaches to the problem of evil and suffering have defended what I call antitheodicism (to be explained in more detail in Section 2 below), in contrast to the theodicist orientation of most contemporary analytic philosophers of religion. I will also argue that all three philosophers have their Kantian roots, to the extent that there is a sense in which each of their antitheodicisms can, in a broad sense, be regarded as Kantian (cf. Kivistö and Pihlström 2016). All three can also be contrasted with the mainstream views they are criticizing, especially current Anglo-American analytic philosophy of religion focusing on arguments for and against metaphysical and evidential theism. Instead of evidentialist considerations, these three antitheodicists (and antitheodicisms) offer an essentially ethical argument focusing on the failures of theodicies to adequately acknowledge the meaninglessness of suffering.

The plan of the essay is, briefly, the following. Immediately after this introduction, Section 2 will explain what we should mean by theodicy, antitheodicy, theodicism, and antitheodicism. I will also briefly indicate how these notions are related to evidentialism and realism in the philosophy of religion. In addition, I will introduce the notions of recognition and acknowledgment in this context, because a recurring argument among the antitheodicists I will examine is that theodicies strikingly fail to recognize, or acknowledge, the reality of (meaningless) suffering. Section 3 is then devoted to an exposition of the three positions I am comparing (to be discussed in a chronological order): Jamesian pragmatist antitheodicism, with a focus on the pragmatist view of truth and truthfulness (Section 3.1), Wittgensteinian antitheodicism exemplified by D. Z. Phillips (Section 3.2), and Emmanuel Levinas’s Jewish “post-Holocaust” antitheodicism (Section 3.3). I will then explain why all these different approaches should be regarded as Kantian (in a broad sense) (Section 4), before drawing the argument to a close in the concluding Section 5.

2. Theodicy, Antitheodicy, Recognition, and Acknowledgment

We may characterize theodicy and theodicism with reference to the attempt to provide a justification for apparently purposeless (meaningless, absurd) suffering. Generally, theodicies seek a justification, legitimation, and/or excusing of an omnipotent, omniscient and absolutely benevolent God’s allowing the world to contain evil and for allowing humans and other sentient beings to suffer. Classical formulations range from Augustinian free will theodicy to Leibnizian “best possible world” theodicy and beyond. By “theodicism”, furthermore, I refer to all attempts to deal with the problem of evil and suffering that regard theodicy as necessary for theism, no matter whether they end up accepting theism or atheism. The theodicist can be an atheist, if s/he concludes that God does not exist (or probably does not exist) because the theodicist demand cannot be met.1 Moreover, theodicies can be secular, if God’s reasons and purposes are replaced by a metaphysical or other overall scheme explaining and legitimizing evil, such as (e.g., Hegelian) historical purposefulness or some other grand narrative. By “antitheodicism”, in contrast, I mean the rejection of not only all theodicies (religious as well as secular) but the very project of delivering a theodicy, including the attitude of encouraging or requiring philosophical engagement in such a project (no matter whether that attitude is accompanied by a belief in the ultimate success or failure of the theodicist project). The philosophers to be considered in the next section all deliver different antitheodicies—that is, arguments seeking to show that it is in some fundamental sense ethically and perhaps also theologically wrong or misguided to try to construct theodicies. While their specific antitheodicy arguments come from distinct philosophical traditions and employ very different philosophical vocabularies, they are all equally strongly committed to a general form of antitheodicism, understood as a refusal to play the theodicists’ argumentative game.

Theodicism and evidentialism are closely connected. As analytic philosophy of religion today is (albeit with significant exceptions) relatively strongly evidentialist (in a broad sense), it is no surprise that it is also often strongly theodicist when dealing with the problem of evil and suffering. That is, the reality of evil and suffering is usually seen as an empirical premise challenging theistic beliefs in an argumentative exchange seeking evidence for or against theism. This is so irrespective of whether the problem of evil is regarded as a logical or as an evidential problem (see, e.g., (Rowe 2001)); that distinction just basically concerns the strength of the evidence the atheist claims to possess, i.e., whether such evidence is claimed to be logically conclusive or inconclusive yet epistemically relevant. The theistic goal is, then, to respond to the atheists’ “argument from evil”—along with providing rationally acceptable evidential considerations in favor of theism more generally. Just as theodicism is a normative view according to which any rationally acceptable theism ought to formulate a theodicy (or at least take steps toward the direction of a theodicy by formulating a “defense”; cf., e.g., (Van Inwagen 2006)), evidentialism is a normative epistemological view according to which any rationally acceptable theism ought to be defended by means of evidence, or rational considerations more generally. Theodicism is, then, a specific dimension of evidentialism: it tries to tell us how we should discuss the problem of evil when evil is regarded as a piece of evidence against theism that the theist needs to deal with. A philosophical critique of theodicism is, therefore, a critique of the entire argument from evil and its use in both theistic and atheistic discourse.2

While the link between theodicism and evidentialism seems to be relatively clear, the relation between theodicism and realism, especially metaphysical realism, has not been studied in equally great detail. It could be argued not only that evidentialism is a central epistemological background assumption of theodicism but also that metaphysical realism remains a major presupposition of the entire theodicist discourse. By metaphysical realism I here simply mean the view that the world possesses its own fundamental ontological structure independently of our perspectives of conceptualization and inquiry and can, therefore, in principle be truly and completely described from an absolute standpoint (i.e., a “view from nowhere”, or a “God’s-Eye View”). We cannot discuss these general issues any further here, but when examining in some detail the three antitheodicist lines of thought to be compared in the next section, it will be useful to be aware of the evidentialist and realist presuppositions of theodicism. When rejecting theodicism, philosophers such as James, Phillips, and Levinas typically also reject both evidentialism and (metaphysical) realism (cf. Kivistö and Pihlström 2016, 2017).

Another concept to be employed (explicitly or implicitly) throughout my explorations of theodicies and antitheodicies below is recognition. As theorists of recognition (in German, Anerkennung) in the broadly Hegelian tradition famously developed by Axel Honneth and Charles Taylor (among others) have emphasized, a proper analysis of the concept of recognition, especially in its applications to interpersonal relations, crucially requires an “as” clause specifying the content of the relevant requests for recognition and the corresponding acts of granting recognition (cf. Honneth [1992] 2005; Koskinen 2017). In paradigmatic interpersonal (or inter-group) cases, A asks B to recognize it as something specific, say, as X, and then B either grants or refuses to grant such recognition of A as X. (By requesting such recognition, A also recognizes B as a potential recognizer of A as X.) Obviously, we are here dealing with a notion that has huge relevance to discussions concerning theology and the philosophy of religion (see especially (Saarinen 2016)). It is a continuous challenge to religious groups and orientations in a multicultural society to recognize each other—as religions, as sources of meaning and value for their members, and so on—but in our case the central question is whether theodicies (religious or secular) succeed in recognizing sufferers, the victims of evil, and/or their suffering.

The details of theories of recognition need not concern us here. What is important for the present inquiry is that there is, arguably, an interesting difference between the concepts of recognition and acknowledgment.3 This difference becomes relevant when we employ these concepts in our attempts to understand the ways in which we may and should respond to other human beings’ suffering. We may say that the concept of recognition inevitably invokes issues of truth, especially regarding the truth of the relevant “as” clause, that is, whether A is X, or is what s/he/it is recognized as. When we are interested in the recognition of another’s suffering, or of the victim of suffering and their unique experiences, we cannot avoid the question of what the sufferers or the victims, or their experiences, are to be recognized as. However, this already creates a (possibly also politically sensitive) field of power structures. The potential recognizer has, at least within certain limits (presumably constrained by the initial request for recognition), the power to determine as what the potential recognizee is to be recognized. Even the very act of recognizing someone, or some group, as a “victim” of some particular evil may be morally and politically problematic, if that person or group does not wish to be categorized as a victim, or if their wish to be so characterized plays a controversial political role. At least the potentially problematic aspects of “victimization” (either by oneself or by others) have to be taken into consideration whenever a question about recognizing someone or some group as a victim is raised.

However, no matter how we wish to fill in the relevant “as” clauses in our analyses of relations of recognition, we arguably also need a more purely philosophical (though not for that matter entirely non-empirical) notion of acknowledgment. It may be suggested that acknowledgment differs from recognition in not requiring, or possibly not even allowing, the same kind of “as” clause as recognition. Thus, the truth about the “as” of recognition, the truth about, say, A’s really being X in our schematic formulation, does not necessarily come to the picture, either. This does not mean that truth would be irrelevant to relations of acknowledgment, however. It might be suggested that we need a different notion of truth here, and this is where pragmatism—the first among the three antitheodicist ways of thinking to be considered in Section 3—will become relevant.

The observation that the notions of recognition and acknowledgment turn out to be useful in the comparisons between various different antitheodicy arguments of the twentieth century is itself a comparative gesture in the analysis of recent philosophers and philosophical traditions. Conceptual equipment drawn from a certain tradition (in this case, the neo-Hegelian tradition with a Frankfurt School critical theory twist, developed further in contemporary social ontology, among other contexts) may offer valuable ways of perceiving the similarities and differences between traditions that are distinct from the one within which that conceptual equipment has originally been developed. Engagement in such a comparative endeavor may also enhance our self-understanding as representatives of a certain philosophical tradition or framework.

3. Three Antitheodicies

3.1. Jamesian Pragmatist Antitheodicy: Widening the Concept of Truth

One may argue that an experientially deeper, more dynamic concept of truth than the one available in largely realistic, correspondence-theory-based analytic philosophy of religion—that is, a pragmatic concept of truth we may derive from William James’s pragmatism (see especially James 1907/1975 in (James 1975–1988), chp. 6)—plays a fundamental role in the structures of acknowledgment when it comes to developing an appropriate ethical attitude to the “truth” of others’ suffering. What we need is a broader, richer, and more clearly non-foundationalist—that is, preferably (but perhaps not necessarily) pragmatist—notion of truth in order to properly communicate experiences of suffering, and our acknowledgment of others’ suffering, in an ethically appropriate way across diverse religious, theological, and secular outlooks. This ethical challenge is ipso facto a challenge of developing the notion of truth itself in a pragmatist way, in a manner in which the acknowledgment of others’ experiences, such as experiences of suffering, particularly in their ethical dimensions—e.g., our future-oriented worry, care, or Sorge for the other—is an ineliminable part of the dynamics of truth in adequately responding to otherness. It is this essentially concerned or even “worried” future-directedness that the pragmatist conception of truth arguably takes into account better than any other, especially if we also read James’s writings on truth alongside his relational conception of identity based on his radical empiricism, as suggested in José Medina’s (Medina 2010) interpretation of his theory of truth. It is on these grounds that we may seek to develop a synthesis of pragmatism and recognition theory (or, rather, the theory of acknowledgment), focusing on the notion of truth.

It can be argued, furthermore, that James’s conception of truth already incorporates important aspects of the more explicitly ethical concept of truthfulness. Admittedly, the pragmatist theory of truth is far from uncontroversial, as anyone who ever read undergraduate textbooks on truth knows. We might here, however, approach it by referring to the distinction between truth and truthfulness, analyzed, e.g., in Bernard Williams’s (Williams 2002) important book with that title. These are clearly different notions. One may pursue truthfulness without thereby having true beliefs; one can be truthful also when one is mistaken, insofar as one sincerely seeks to believe truths and avoid believing falsehoods and also honestly seeks to tell the truth whenever possible (and whenever the truth to be told is relevant). Thus, clearly, whatever one’s theory of truth is, one should in some way distinguish between truth and truthfulness. On the other hand, certain theories of truth, such as the pragmatist theory, may be more promising in articulating the intimate relation between truth and truthfulness than some other theories—and this, we may suggest, is highly relevant to the theodicy vs. antitheodicy issue.

We might say that the distinction between truth and truthfulness is in a way “softened” in James’s pragmatist conception of truth, which rather explicitly turns truth into a value to be pursued in one’s (individual and social) life rather than simply a matter of propositional truth corresponding to facts that are mind- and discourse-independently “out there” in the world. Truth(fulness) in the Jamesian sense is, then, richer and broader than mere propositional truth. It is a normative property of our human thought and inquiries (in a wide, practice-embedded sense), not simply a semantic property of statements or beliefs. James’s pragmatic truth hence incorporates truthfulness, as truth belongs to the ethical field of inter-human relations of dependence and acknowledgment. It also incorporates an acknowledgment of the (at least potential) inner truth of others’ experiences, especially experiences of suffering.

In his discussion of James’s theory of truth, Medina (2010) also defends Jamesian pluralism in a politically relevant manner: in ethics and politics, we can never reach an “absolute” conception of what is universally best for human beings and societies, but different suggestions, opinions, experiential perspectives, and interests must have their say—or must be recognized or acknowledged (though this is not Medina’s terminology). James advocated not only pragmatic pluralism and individualism but also (on Medina’s reading) a relational conception of individual identities: nothing exists in a self-sustained manner but only as parts of networks of mutual interdependence. While James’s pluralism and relationalism are elements of a metaphysical view according to which “nothing can be understood in and by itself, but rather in relation to other things, in a network of relations”, they are irreducibly ethical and political, applying even to the reality of the self: “to have a sense of self is to have a sense of the dependences that compose one’s life” (ibid., p. 124). We are, Medina continues, “diverse and heterogeneous beings […] shaped and reshaped through diverse and heterogeneous networks of interpersonal relations”, and the Jamesian self is a bundle of such relations (ibid., p. 125; cf. also (Pihlström 2013)).

It is in this context that we should, according to Medina, understand and appreciate James’s theory of truth. True beliefs, as James himself says, are “good to live by”; when maintaining a belief, any belief, we are responsible for its consequences in our lives, and in those of others. The pragmatic “theory” of truth—which should not be called a “theory”, in order to avoid seeing it as a rival to, say, the “correspondence theory”—invokes not only, say, the agreeable consequences of true beliefs but also ethical ideas such as solidarity and justice. Therefore, we may say that truth (in the Jamesian pragmatic sense), truthfulness, and acknowledgment are conceptually tied to each other. One cannot really pursue truth in the Jamesian sense unless one also acknowledges, or at least truthfully seeks to acknowledge, others’ perspectives on reality—especially those defined by suffering.

If we take even a brief look at James’s various pronouncements of the significance of the problem of evil and suffering,4 we may note how thoroughly those discussions of the diversity of our human responses to such predicament are characterized by references to reality and truth. I suggest that these occurrences of such philosophical terms ought to be taken seriously as elements of James’s account of truth and reality—and thus as key elements of his pragmatism more generally. James argued, against “the airy and shallow optimism of current religious philosophy”, that what desperate and suffering human beings experience “is Reality”: “But while Professors Royce and Bradley and a whole host of guileless thoroughfed thinkers are unveiling Reality and the Absolute and explaining away evil and pain, this is the condition of the only beings known to us anywhere in the universe with a developed consciousness of what the universe is”. (James 1907/1975 in (James 1975–1988), pp. 20–21) Thus, idealist, optimistic (e.g., Hegelian) philosophers “are dealing in shades, while those who live and feel know truth”; a Leibnizian theodicy postulating a harmony of the universe is “a cold literary exercise, whose cheerful substance even hell-fire does not warm” (ibid., pp. 22, 20). What I have called theodicism is, for James, part of the “unreality in all rationalistic systems” of “religious” philosophy that remain “out of touch with concrete facts and joys and sorrows” (ibid., p. 17). James here even quotes at length from Leibniz’s Théodicée, concluding that “no realistic image of the experience of a damned soul had ever approached the portals of his mind” (ibid., pp. 19–20).5

The references to truth, facts, and reality in these formulations must be understood as elements of the ethically interwoven fabric of James’s pragmatism in general. James’s antitheodicist remarks can be seen as philosophically urging us to acknowledge the meaninglessness of suffering, i.e., the fact that in reality there is no harmonious world-system or a necessary divine reason for suffering. This is, in a sense, to recognize the sufferer as sufferer. As recognition requires the “as” clause—B recognizes A as something, as X—the one who employs the concept of recognition here must presuppose that there is something like the truth about the matter whether A is X, or can be construed as being X. This can, and in many obvious cases is, a truth created by the act of recognition itself; in our social life many important social facts, statuses, and institutions are created in this way. Thus, the fact (or the factual or propositional truth) that A is X, which enables B to recognize A as X, is either independent of B’s recognition act or constituted by it, by B’s “taking” A to be X. However, in both cases we are dealing with truths about the (social) world.

Now, it may be argued that the traditional correspondence theory of truth does not serve this situation very well. What we (arguably) need here is a more dynamic pragmatic notion of truth also covering cases where the truth of A’s being X is “made” by us. But even more importantly, we need a pragmatic notion of truth for those cases where there is no “as” dimension available at all—that is, for acknowledgment rather than recognition. This is another kind of ethical truth, truth conceived or reconceived as something like truthfulness. It is a conception of ethical truth compatible with, or perhaps even required by, James’s famous dictum in “The Moral Philosopher and the Moral Life” (1891), according to which there can be no “final truth” in ethics until the “last man” has had his experience and his say (James 1897/1979 in (James 1975–1988), p. 141). Thus, this is precisely the kind of ethical truth we need for a conception of ethics that admits, with James, that “there is no such thing as an ethical philosophy dogmatically made up in advance” (ibid.).

Acknowledging, with James, that there is no such final truth in ethics—or religion—is, we may argue, part of what we can, and should, mean by our being committed to ideals such as truthfulness and sincerity in our acknowledgment of others and their morally relevant perspectives on the world. There can, then, be something like the pursuit of truth, as well as something like truth itself, in the act of acknowledgment even if there is no narrow propositional sense of truth available regarding the “as” clause, as required in recognition. This pursuit of truth in a pragmatist sense is essentially the pursuit of a truthful attitude to the other person’s suffering (or one’s own suffering), which in the end will then lead into, or at least enable, a truthful attitude to experience in general—or the world in general.6 This yields an indirect (and reflexive) argument in favor of pragmatism: its conception of truth (and truthfulness) in the service of acknowledging diverse others and othernesses, and their various experiences of suffering and attempts to communicate those experiences, pragmatically works.

However, things are not as simple as that. While James’s pragmatism insightfully integrates truth and truthfulness, or incorporates elements of the notion of truthfulness into the pragmatically articulated notion of truth itself, and while (we may say) Jamesian pragmatism generally is framed by taking evil and suffering seriously in an antitheodicist manner (that is, without succumbing to the temptations of theodicy that would allegedly explain away the reality of meaningless suffering), it can be argued that the notion of objective truth itself is, in an irreducible way, required by the kind of antitheodicist attitude that James seems to be recommending as a sine qua non of a properly ethical acknowledgment of otherness.

Can truthfulness, in a sense fundamental to acknowledgment, be maintained even if mere propositional truth (fundamental to recognition, given the central role of the “as” clause) loses at least part of its significance in Jamesian pragmatism? Acknowledgment, rather than recognition, can be seen as the key to truthful explorations of suffering. This can also be regarded as a fundamentally Jamesian view—while employing terminology different from James’s own. In his ethical thought, James argued against our instinctive “blindness” and “deafness” in relation to other human beings and their individual ways of viewing the world. James may be read as presenting us with a profound, and endless, ethical challenge to acknowledge otherness, especially the other’s suffering—as the framing of the entire pragmatist project makes clear. It could even be argued that the way in which James’s pragmatic method is introduced (in James 1907/1975 in (James 1975–1988), chps. 2–3, in particular) as a method of “making our ideas clear” by (re-)entangling abstract metaphysical concepts with their conceivable ethical relevance (cf. Pihlström 2009) actually presupposes an antitheodicist understanding of the irreducible reality of suffering as something that needs to be taken fundamentally seriously in any ethical—or, therefore, metaphysical—world engagement.

We are here dealing with the need to acknowledge others’ suffering itself as well as others’ attempts to communicate their suffering, possibly across diverse religious and theological (but also non-religious) outlooks that are foreign to us. As James maintains in his 1898 lecture “Human Immortality”, encouraging us to acknowledge what he calls our “half-brutish prehistoric brothers”: “‘Tis you who are dead, stone-dead and blind and senseless, in your way of looking on. You open your eyes upon a scene of which you miss the whole significance. Each of these grotesque or even repulsive aliens is animated by an inner joy of living as hot or hotter than that which you feel beating in your private breast.” (James 1982 in (James 1975–1988), p. 99). The same line of thought is strongly present in James’s 1899 essay “On a Certain Blindness in Human Beings”. There James, observing the work of farmers and outdoor manual laborers, charges himself, and us, of being “as blind to the peculiar identity of their [those others’] conditions as they certainly would also have been to the ideality of mine, had they had a peep at my strange indoor academic ways of life” (James 1983 in (James 1975–1988), p. 134).

It is this kind of blindness that we have a moral duty to avoid, just as we are, according to Emmanuel Levinas (who, of course, comes from a quite different philosophical tradition), invited to appreciate the other’s face—as will be seen in a moment.7 It is, moreover, by self-critically countering such instinctive blindness that we may cherish the virtues of acknowledgment and truthfulness—and this, I am suggesting, is part of what it means to develop the notion of truth itself in a pragmatic manner. When developed in this Jamesian way, the notion of truth already has an ethics of truthfulness and acknowledgment built into it. This, I submit, is where its deepest antitheodicist significance lies.

3.2. Wittgensteinian Antitheodicy: D.Z. Phillips

Let us, before considering Levinas’s version of antitheodicy, turn from Jamesian pragmatism to a brief examination of the kind of antitheodicism that is central to the distinctive trend in the philosophy of religion inspired by Ludwig Wittgenstein. Theodicies have from a Wittgensteinian perspective been regarded as unethical violations of religious language use (even though Wittgenstein himself had little to say about this particular topic). Religiously speaking, theodicies could even be criticized as superstitious or blasphemous (cf. Phillips 1977, 2004).

Let us note, first, that several Wittgensteinian philosophers have argued that theodicies allegedly justifying “God’s ways to man”—that is, arguments seeking to make sense of apparently meaningless and absurd evil and suffering in God’s overall harmonious plan—amount not only to ethically insensitive use of language disregarding or misrecognizing others’ suffering in its utter pointlessness but also to conceptual confusion and pseudo-religious use of language. It is not only ethically wrong but in an important sense meaningless and conceptually confused, and thus beyond the meaning-constitutive grammar of religious language-games, to claim that others’ suffering has a metaphysical or theological meaning, function, or explanation. Such a “conceptual oddness” of theodicies has been noted not only by Phillips (see, e.g., (Phillips 1977, 2004)) but also by Ben Tilghman (Tilghman 1994, p. 192)—who uses this very phrase—and by Stephen Mulhall (Mulhall 1994, pp. 18–19), who even suggests that theodicies end up with blasphemy. In a similar vein, Andrew Gleeson (Gleeson 2012, especially chp. 1), also writing against theodicies in a broadly Wittgensteinian tradition, notes that theodicies should be criticized on both moral and conceptual grounds, while Mikel Burley (Burley 2012, §5) points out that the theodicist is “so confused as to be unaware of the degree of their own insensitivity” to pain and suffering—with both moral as well as logical and conceptual dimensions pertaining to this confusion.

Let me offer a few more specific observations on Phillips’s approach in particular. The Wittgensteinian method Phillips (along with many other Wittgensteinians) subscribes to carefully looks at the actual use of language in concrete human situations and practices, instead of any a priori rules or principles establishing linguistic meanings. Yet, he also emphasizes the general Wittgensteinian (and also, broadly speaking, pragmatist) ideas that “it is only in the context of [religious] language games that belief in God has any meaning” and “concepts have their life” “only in practice, in what we do” (Phillips 1993, pp. xi, xiii). In his criticism of theodicies, in particular, Phillips focuses on what goes wrong in the very form of the allegedly moral reasoning the theodicist engages in; he interestingly cites the Book of Job here: “Job cannot make sense of his afflictions in terms of the [theodicist] arguments of his would-be comforters” (ibid., p. 157). While those defending theodicies try to calculate what kinds of goods or benefits might outweigh or compensate for the evils and sufferings there are, the Phillipsian-cum-Wittgensteinian antitheodicist objects to “the concept of calculation in this context, because it excludes moral concepts” (ibid., p. 158; original emphases).

Phillips argues that the truly religious reaction to the contingencies and adversities of human life does not seek to “tidy up” messy human reality or to find explanations and understandings of suffering (see ibid., pp. 166–68). On the contrary, properly (genuinely, truly) religious uses of language, when addressing the problem of evil and suffering, recognize the limits of understanding and linguistic expression—not as contingent limitations that could in principle be transcended but just factually cannot be overcome by us, but as necessary limits defining the relevant language-game and therefore playing a (quasi-)transcendental role in constituting what is meaningful and possible for us (see ibid., p. 168), albeit in the end only contextually necessary limits that could in principle be redrawn as our lives change.

A transcendental critique of theodicies, when formulated from a Wittgensteinian perspective along Phillips’s lines, thus crucially focuses on the “grammar” constitutive of moral and religious language, that is, on the transcendentally meaning-structuring rules of the relevant language games—rules that might, however, themselves be historically transformed. If we take seriously the Wittgensteinian line of thought according to which there can be no meaning without practice-laden, habitual, world-engaging use of expressions within public human ways of acting, or language games, then we should also acknowledge the fact that the meanings of such expressions as “evil”, “suffering”, “God”, “meaning”, etc., are inextricably entangled with our using them in religious (and other) language-games and thus in our forms of life. If we do take this seriously, then it is conceptually, morally, and religiously misguided to seek to provide a theodicy—or to require one.

In many cases, for a genuinely religious person who sincerely attempts to speak about God in a religious way, belief in God’s reality is the necessary background of any potential theological or philosophical account of evil and suffering. Any possible argument, including the atheist’s argument challenging the theist to provide a theodicy by appealing to the problem of evil, will have to be evaluated against this background. The believer might point out, against both the atheist and the theist seeking to provide a theodicy, that it is strictly speaking nonsensical (i.e., beyond the meanings available in religious language-games) for human beings to try to evaluate God’s motives morally, or to seek to criticize or justify them. At the moment when the theodicist begins to engage in an argument, pro or contra, regarding the problem of evil conceived as an atheist challenge, the grammar of the religious language-game will already have been violated and the relevant expressions will no longer be used in a genuinely religious meaning. Therefore, the atheist argument starting from the problem of evil and suffering does not even get off the ground due to this confusion. But those theists who try to respond to such an argument by producing a theodicy are equally confused, or perhaps even more confused, because they do not perceive the atheists’ confusion any more than their own.

Theodicies should therefore be rejected as transcendentally confused misuses of the language of “God”, “evil”, and “suffering”. It can be suggested that one comes close to illegitimately transgressing the limits of meaningful discourse—the limits of religious language—simply in examining the problem of evil in terms of the alleged “argument from evil” and in attempting to respond to such an argument theodicistically. Both the theist and the atheist theodicists fail to follow the grammar of religious language religiously and thus breach the limits of language.

It is roughly in this way that I would like to suggest we can reinterpret Phillips’s and other Wittgensteinians’ perceptive remarks on theodicies being both morally and conceptually (or even logically) confused. The reason why these confusions are so deep is that they are transcendental in the sense of this word that remains available in the later Wittgenstein’s thought. This is the transcendentality of the constitutive (albeit historically transformable and reinterpretable) features of language-games and forms of life. It is only by violating the limits of language that the problem of evil and suffering construed as an atheist argument requiring a theodicy as a response can so much as be formulated. When we realize that such a (mis)formulation is based, precisely, on a violation of grammar, we realize that the entire business of theodicy is misconceived from the start. It is by means of a Wittgensteinian analysis of the limits of language that this point can be brought home.

However, as soon as we note all this we should re-emphasize that the grammar and meanings of our expressions may vary historically along with the changes and transformations taking place in our forms of life. The necessity of transcendental rules is, in a sense, based on the contingency of human life. This applies with full force to the problem of evil and suffering. Different reactions to this problem may become possible in different historical circumstances characterizing the forms of life through which the grammar of the relevant language-games is established. We may, for instance, find it necessary to examine the problem of evil and suffering after the Holocaust in a way essentially different from its pre-Holocaust articulations. In the contingent historical context in which the Holocaust actually took place and will therefore permanently, ineliminably, irredeemably, be part of our human history, it may seem that certain (new) limits of appropriate religious language have been established. It is no longer possible—morally or conceptually—to approach the problem of evil by providing a theodicy. We now—after the Holocaust—can see this as a striking, violent confusion. Moreover, we can now see—after the Holocaust—that it was never possible, even if that was not as clearly perceivable earlier. Historical contingencies may thus ground philosophical, ethical, and conceptual necessities—and this I take to be a fundamental Wittgensteinian message.

In the terms used in the previous Section 3.1, we might say that contextualizing philosophical inquiries into evil and suffering in this manner—that is, viewing historical events such as, paradigmatically, the Holocaust as crucial in shaping those inquiries or their very possibility today—also amounts to a truthful attitude to the inquiries at issue here, roughly in the sense in which James’s pragmatism was above claimed to incorporate an ethical notion of truthfulness in its account of truth. It is not only objective, propositional truth that we seek when seeking to understand evil and suffering in the context of the Holocaust (using that word metonymically to cover all instances of horrible, meaningless suffering). It is, more broadly, ethical truthfulness of proper acknowledgment that is at issue in our inquiry. It is right here that Jamesian pragmatism and Phillips’s Wittgensteinianism could plausibly join forces.

3.3. Jewish Post-Holocaust Antitheodicy: Emmanuel Levinas

Emmanuel Levinas, whom I am here discussing as a representative of the phenomenological tradition (distinct from both James’s pragmatism and Phillips’s Wittgensteinianism), compellingly argues, in “Useless Suffering”, one of his most important antitheodicist writings, that any theodicy is completely disproportional and out of balance in relation to the forms of suffering we know from the history of the twentieth century.8 Not unlike Phillips (1977), who reminds us that catastrophes may strike “without rhyme or reason”, Levinas emphasizes the “depth of meaninglessness” in suffering, the “for nothing” character of suffering, and the intimate relation between evil and suffering: “All evil relates back to suffering” (Levinas 2006, p. 79).

Some of the most central themes of Levinas’s philosophy as a whole are in fact strongly present in “Useless Suffering”. It is precisely the “attention to the suffering of the other” that (he argues) “can be affirmed as the very nexus of human subjectivity” and is the “supreme ethical principle” that is “impossible to question” (ibid., p. 81). Moreover, “my responsibility for the other, without concern for reciprocity”, the asymmetrical relation between the other and myself (ibid., p. 87), is a core idea of Levinas’s ethics as a whole, and it finds a particularly strong expression in his treatment of suffering and antitheodicy.

In the main work of his late thought, Otherwise than Being or beyond Essence, Levinas adds a footnote explaining further one of his best-known formulations of what it means to be an ethical subject, according to which “[t]o be oneself, the state of being a hostage, is always to have one degree of responsibility more, the responsibility for the responsibility of the other” (Levinas 1981, p. 117). The note explicitly links this with suffering:

The vortex—suffering of the other, my pity for his suffering, his pain over my pity, my pain over his pain, etc.—stops at me. The I is what involves one movement more in this iteration. My suffering is the cynosure of all the sufferings—and of all the faults, even of the fault of my persecutors, which amounts to suffering the ultimate persecution, suffering absolutely. This is not a purifying fire of suffering […]. This moment of the “for nothing” in suffering is the surplus of non-sense over sense by which the sense of suffering is possible.

(ibid., p. 196)

The (nonsensical) “sense” of suffering lies only in the ethical subject’s (my) absolute and infinite responsibility. While Levinas does not tell us expressis verbis what kind of impossibility we are dealing with in the impossibility of questioning our responsibility for the other, we may suppose that this impossibility of questioning the duty of attending to the other’s suffering is both ethical and metaphysical—in a sense in which the two are inseparably entangled (in a manner not very dissimilar from James’s pragmatism; cf. (Pihlström 2009, 2013)). The same inseparability of the metaphysical and the ethical can be seen in Levinas’s uncompromising rejection of any theodicy:

Perhaps the most revolutionary fact of our twentieth-century consciousness—but it is also an event in Sacred History—is that of the destruction of all balance between Western thought’s explicit and implicit theodicy and the forms that suffering and its evil are taking on in the very unfolding of this century. […] This is the century that is drawing to a close in the obsessive fear of the return of […] suffering and evil inflicted deliberately, but in a manner no reason set limits to, in the exasperation of a reason become political and detached from all ethics. […] [T]he Holocaust of the Jewish people under the reign of Hitler seems to me the paradigm of gratuitous human suffering, in which evil appears in its diabolical horror. […] The disproportion between suffering and every theodicy was shown at Auschwitz with a glaring, obvious clarity.

(Levinas 2006, pp. 83–84.)9

Levinas continues to note that the “suffering for nothing” of the Holocaust victims “renders impossible and odious every proposal and every thought that would explain it by the sins of those who have suffered or are dead” (ibid., p. 84). Again, it is the justification of the other’s (my neighbor’s) pain and suffering that is an “outrage” and “the source of all immorality” (ibid., p. 85). Such an attempt at justification can take very different forms—as recognized in Levinas’s reference to not only the explicit but also the “implicit” (e.g., presumably, secular) theodicies in Western thought.

We should note that Levinas constantly helps himself to modal formulations in his characterizations of the obscenity of theodicies. The “suffering in the other” is not only unjustified but “unjustifiable”, and the crematoria of the final solution render theodicy “impossible” (ibid.). It is in this modal sense, in particular, that we may view Levinas’s post-Holocaust antitheodicy as “Kantian” and transcendental—just as we may view Phillips’s Wittgensteinian antitheodicy as transcendental in its focus on the constitutive features of language games. Theodicies make our moral relations to other human beings (as well as our appropriately ethical discourse on such relations) impossible. This is a clear link between Phillips and Levinas: ultimately the issue concerning theodicy and antitheodicy pertains to the conditions of religious and ethical language and, therefore, to what needs to be presupposed in order for any humanly meaningful discourse in these areas to be possible for us.

In this sense, for Levinas as much as for James and Phillips, any theodicy is a form of ethical insincerity precisely because it fails to appropriately acknowledge the utter meaninglessness of suffering. This failure is manifested in theodicist language use violating the conditions constitutive of ethically appropriate (religious or secular) language games. There are, of course, major differences between the three philosophers I have cited as representatives of antitheodicism. Their philosophical methodologies are obviously very different. James’s pragmatic method (cf. James 1907/1975 in (James 1975–1988), chps. 2–3) is relatively far from (and indeed critical of) any Kantian transcendental methodology—even though it has been suggested in James scholarship that it is not implausible to read him as a Kantian thinker in a revised and reinterpreted sense (as proposed in (Pihlström 2013))—and while Phillips and Levinas are both concerned with the necessary conditions for the possibility of meaning and meaningfulness and have thus been engaged in something like transcendental investigation (albeit not in Kant’s original sense), their approaches remain quite distinct, as Phillips focuses on a Wittgensteinian “grammatical” examination of language games (religious ones included), while Levinas views meaning in relation to the ethical primacy of the other to and for whom we are infinitely responsible. Despite these obvious methodological differences, and the even clearer differences in philosophical style and content, the three thinkers considered share, however, a fundamentally antitheodicist attitude to what goes wrong in philosophical and theological attempts to justify others’ suffering.

4. Why Are These Antitheodicies “Kantian”?

Why exactly am I suggesting that the three different antitheodicies briefly explored in the previous section are “Kantian”? I have already in general terms referred to the “transcendental” character of this type of antitheodicist investigation. If one engages in Kantian antitheodicism, one asks how it is possible to so much as adopt a moral perspective on the world we live in with other human beings, and seeks the necessary preconditions for this possibility in the rejection of any theodicies, and of theodicist thinking generally. It seems to me that this is exactly what the three (above-discussed) very different philosophers are (were) doing when considering the problem of evil and suffering from their distinctive perspectives. The distinctive antitheodicies delivered by the three philosophers considered in Section 3 differ in many ways, including the degree to which they employ explicitly Kantian (transcendental) language in articulating the argumentation against theodicies and theodicism. For example, Phillips’s Wittgensteinianism is, arguably, rather clearly committed to the idea—redescribable in Kantian terms—that theodicies violate the necessary conditions for the possibility of using religious language (in a “genuinely” religious way). James’s pragmatist rejection of theodicies does not explicitly state anything like that, but it can be argued, as I briefly did above (though see Kivistö and Pihlström (2016, chp. 5), for a much more comprehensive discussion), that James is also concerned with the way in which we are able to adopt a truly moral perspective on other human beings and their distinctive perspectives on the world we share with them.

Kant’s relevance for an inquiry such as the present one is not restricted to his being arguably the first philosopher to explore evil from a resolutely human—non-speculative, non-rationalizing—point of view. This anthropocentric focus follows from his more general critical approach and set of problems—from the philosophical orientation that makes him the greatest of modern thinkers. For Kant, philosophical problems cannot be solved or even usefully discussed from an imagined God’s-Eye-View; they have to be reflected on from a limited, conditioned, and inevitably contextualized human point of view. That reflection unavoidably brings us to a situation full of tensions. A key Kantian issue is whether, and how, there can be any deeper meanings—ethical, aesthetic, teleological, religious—in a world which in a certain sense is, as an object of knowledge and science, devoid of such meanings.

For Kant, this issue can only be examined perspectivally, in terms of a plurality of critiques of reason. (Philosophers and philosophies such as pragmatism and Wittgenstein, as well as Levinasian phenomenology, are in my view basically variations on this theme and these tensions.) It is against this background of the lack of any super-human source of meanings, certainties, and legitimation that the problem of evil, and of (anti)theodicy, also receives its uniquely Kantian formulation—and is continued in, e.g., pragmatism, Wittgensteinianism, and Levinasian post-Holocaust antitheodicy. When seen in such a Kantian light, the problem of evil and suffering must be examined in a critical spirit abandoning any imagined “God’s-Eye-View” presupposed by the pseudo-consoling grand narratives of theodicism (typically assuming a kind of metaphysical realism equally firmly rejected by Jamesian pragmatists, Wittgensteinian philosophers such as Phillips, and Levinasian phenomenologists).

A further, perhaps more obvious Kantian feature of antitheodicism is the refusal to accept any theodicies that turn sufferers into mere means to some alleged overall good (as seen from the imagined “God’s-Eye-View”), failing to appreciate the “humanity formulation” of Kant’s categorical imperative, that is, the principle that we must always treat human beings, ourselves included, as ends in themselves, never as mere means. The antitheodicists examined are thus Kantian also in the sense that they, at least implicitly, argue for antitheodicism as a condition for the possibility of the moral perspective (or moral seriousness) itself. In a certain sense, then, we may see them as arguing that only an antitheodicist can so much as occupy a morally serious perspective on evil and suffering. This is a strong claim and cannot here be comprehensively defended or even properly articulated, but it must, in any event, be put forward as a transcendental thesis, not as a factual or empirical claim about people’s (theodicists’ or antitheodicists’) ability or inability to engage in moral deliberation.10

What the (quasi-)Kantian transcendental antitheodicists—such as, arguably, James, Phillips, or Levinas—try to argue is that antitheodicism is constitutive of a truly moral perspective on evil and suffering, and hence of morality itself, given that morality is largely (though not exclusively) a matter of responding to evil and suffering. Antitheodicism is not just one available approach among others but a standpoint needed for an ethical attitude to evil and suffering to be so much as possible. It transforms the way we view and engage with the world in general; as soon as we recognize the reality of meaningless suffering and seriously set aside the project of excusing it—or excusing the world, or God for allowing its existence—our entire perspective on reality, especially the reality of other human beings, dramatically changes.

Yet another Kantian (albeit not exclusively Kantian) dimension in arguments for antitheodicism is their frequent focus on the concepts of freedom and necessity. According to the theodicist logic, evil is in some sense necessary—if not metaphysically or theologically necessary (e.g., as an unavoidable element of the “best possible world” we live in according to Leibniz’s theodicy), then at least instrumentally necessary in order for some “greater good” to be available in the grand divine (or secular) scheme of world history. In contrast, antitheodicism in its different versions—perhaps most explicitly in James’s pragmatism urging us to do whatever we can to make the world better, given that neither cosmic salvation nor total destruction is guaranteed—refuses to accept such necessities and emphasizes, on the contrary, the radical contingency (non-necessity) of evil. The evil and suffering there are in our world are to a significant degree grounded in free human actions, as Kant himself maintained in his theory of radical evil. But this does not yield any “free will theodicy”, or justify evil and suffering simply with reference to the greater overall good of freedom. All claims about the necessity or unavoidability of evil, whether absolute or contextual, are either explicitly or implicitly theodicist, at least to some degree (assuming one can be a theodicist up to a degree), and it is this often hidden theodicism that ought to be repeatedly argued against—as Levinas perhaps most forcefully reminds us.

One more Kantian aspect of antitheodicist inquiry ought to be emphasized. When speaking of antitheodicism as a transcendental thesis, I am obviously comparing it with some of the key principles of Kantian critical philosophy that are also transcendental, i.e., not empirical or factual but providing the conditions for the possibility of anything’s being empirical or factual. However, it is with some caution that such comparisons should be made. We should not directly claim, for instance, that antitheodicism would have the same status as a transcendental principle as, say, the Kantian categories or the forms of pure intuition (space and time) that are, according to Kant’s First Critique, necessary conditions for the possibility of cognitive experience of objects and events (and thus also for the possibility of objecthood in general).

Possibly, a better analogy would be the somewhat weaker albeit distinctively transcendental status of what Kant (Kant [1788] 1983) in the Second Critique calls “postulates of practical reason”. They are, famously, the immortality of the soul, the freedom of the will, and the existence of God. According to Kant, these postulates cannot be transcendentally demonstrated in the same sense in which we can demonstrate, say, the universal applicability of the concept of causation (as a category) to all objects of possible experience. Yet, the postulates are transcendentally defended as necessary auxiliary presuppositions without which our practical commitment to morality (demanded by reason) and to what Kant takes to be its key principle, the categorical imperative, would not make sense. Similarly, antitheodicism can be defended transcendentally as a condition necessary for our being able to make sense of our commitment to the moral point of view (or what we may call its seriousness). It may not hold “fully objectively” as a condition for morality generally—for, e.g., imagined beings very different from humans—but it is a fundamental moral requirement comparable to a transcendental principle for beings like us, that is, the kind of rational yet finite creatures that Kant himself also addresses his critical philosophy to.

I have tried to show that as different as James, Phillips, and Levinas are as philosophers and moral thinkers, they all contributed to the theodicy vs. antitheodicy discussion along these fundamentally Kantian lines, though without explicitly invoking Kant in this context at all. Whether this makes them “Kantian” or “transcendental” thinkers in some robust sense is of marginal importance in comparison to the general conclusion that they all subscribe to a similar ethical campaign against the theodicist tendency to fail to acknowledge other human beings’ meaningless suffering. The occasionally very different ideas and arguments developed by these philosophers converge toward a shared conviction that it is ethically wrong as well as philosophically confused to view, in a theodicist manner, the absurd and pointless suffering some human beings have to go through as playing any instrumental role as an element of some imagined overall good. Antitheodicistically, all these philosophers maintain that it is a fundamental ethical and theological mistake to justify others’ suffering. Instead, they hold that such suffering, and its meaningless, must be fully acknowledged in order for us to be able to adopt a morally serious point of view on the world we live in. At the meta level, and often implicitly, this argument is transcendental in the sense that we may view all three philosophers as seeking to show that the theodicist mainstream of contemporary analytic philosophy of religion not only fails to adopt such a point of view but in a sense renders its adoption impossible.

5. Conclusions

Our consideration of three varieties of twentieth-century “Kantian” antitheodicies also highlights, as was already noted, a certain kind of contextuality of philosophical discussions of evil and suffering. The Holocaust was obviously a decisive turning point in philosophical and theological reflection on these matters. It is in this historical context that thinkers reflecting on evil and suffering started to turn to antitheodicy more explicitly than earlier. (This is also to emphasize the originality of James’s pre-Holocaust reflections on evil, which are as strongly antitheodicist as any later ones.) As philosophers deeply engaged with the Jewish context of their work—e.g., Bernstein (2002)—often emphasize, after the Holocaust it would simply be “obscene” to continue to develop theodicies.

Philosophy always develops in its contexts, whether as striking and out of the ordinary as the Holocaust was, or (usually) more ordinary ones. Acknowledging this contextuality may also be helpful when it comes to acknowledging the different yet converging ways in which thinkers representing quite distinct traditions but sharing the same historical reality may occasionally have come to emphasize similar philosophical ideas and arguments. It is something like this that I have tried to do in this brief comparative study of three twentieth-century antitheodicies.

Moreover, at the meta level we should also recognize the ways in which different assumptions and arguments concerning evil, suffering, theodicies, and related issues have shaped—or provided contexts for—the development of other philosophical ideas and arguments (for an account of the history of modern philosophy from the point of view of the problem of evil, see (Neiman [2002] 2004)). Thus, it is not only the problem of evil and suffering, or the argumentation pro and contra theodicies, that has developed in historical contexts; that problem and that argumentation have themselves provided contexts for the development of many other philosophical views and arguments. Indeed, as Neiman emphasizes, the very comprehensibility of the world is at issue when the problem of evil is seen as the core problem of the history of philosophy in the modern era.

I have not claimed that this would have been the core problem of philosophy for the individual thinkers I have discussed (except, perhaps, for Levinas, because his entire ethics of otherness certainly focuses on the problem of responding to others’ suffering), but certainly it has been a major concern for all the three approaches I have compared and has thus partly led to their developing into the kind of philosophical frameworks they have been, or into what they currently are. It is in this context that we should develop them further and encourage critical dialogues among their contemporary (and future) representatives.