SOPHIA (2008) 47:129–148 DOI 10.1007/s11841-008-0053-8
Why Theists Cannot Accept Skeptical Theism Mark Piper
Published online: 23 May 2008 # Springer Science + Business Media B.V. 2008
Abstract In recent years skeptical theism has gained currency amongst theists as a way to escape the problem of evil by invoking putatively reasonable skepticism concerning our ability to know that instances of apparently gratuitous evil are unredeemed by morally sufficient reasons known to God alone. After explicating skeptical theism through the work of Stephen Wykstra and William Alston, I present a cumulative-case argument designed to show that skeptical theism cannot be accepted by theists insofar as it crucially undermines epistemic license to the very theism it is invoked to defend. I also argue that attempts to defend a theism-friendly moderate version of skeptical theism either fail to halt the spread of damaging skepticism, or lack philosophical validity. Keywords Theism . Skeptical theism . Evidental problem of evil
Introduction The evidential problem of evil is a perennial problem for theism. Theism contains the claim that the world was created by an omnipotent, omniscient, and omnibenevolent God. If this were the case, however, it stands to reason that the M. Piper (*) Department of Philosophy, St. Louis University, 3800 Lindell Blvd., St. Louis, MO 63108, USA e-mail: [email protected]
world would contain no gratuitous evil; and yet a great amount of gratuitous evil appears to exist; hence God (so described1) must not. In the present essay I intend to revisit this time-worn dispute between theism and atheism. Specifically I will concentrate my analysis on one species of proposed solution to the evidential problem of evil (henceforth simply ‘problem of evil’): skeptical theism. Skeptical theism works by arguing that the putatively gratuitous evil we experience may in fact, for reasons beyond our ken, be a necessary correlate of the actualization of higher goods. The fact that we cannot discern these higher goods should not count against their possible existence and validity, however, for, given our limited cognitive capacities, we should not expect to be able to discern them. Because of this, the skeptical theist concludes, the evidence afforded by apparently gratuitous evil shouldn’t be taken to count against the perfect goodness of God, and consequently the problem of evil cannot get off the ground. In this essay I present a cumulative-case argument for the claim that theists cannot accept skeptical theism. The prevailing theme that unites my arguments can be expressed as follows: skeptical theism can be used to dissolve the problem of evil only at the cost of the maintenance of epistemic license to crucial aspects of the theism it is invoked to defend. I will proceed as follows. In the second section, The Case for Skeptical Theism, I explain skeptical theism (ST) in more detail through an elaboration of the arguments given in its support by Stephen Wykstra and William Alston. In the third section, The Case Against Skeptical Theism, I present my cumulative-case argument against ST as a viable theistic response to the problem of evil by focusing on three key difficulties for the maintenance of epistemic license to theism that consistent application of the principles of ST appears to entail. In the fourth section, Is a Theism-friendly Moderate ST Philosophically Defensible?, I outline three arguments that theists might employ in an attempt to limit the scope and force of the skeptical considerations advanced in the third section, and conclude that they all fail. The last section contains my final comments.
The Case for Skeptical Theism We can take William Rowe’s formulation of the problem of evil as our starting point. Let us call it Argument A.2 (1) There exist instances of intense suffering which an omnipotent, omniscient being could have prevented without thereby losing some greater good or permitting some evil equally bad or worse. 1
It should be noted that the problem of evil can be sidestepped (or at least seriously weakened) if one holds a different conception of God. If one holds, for example, that God is very good (but not perfectly good), and/or very knowledgeable (but not omniscient) and/or very powerful (but not omnipotent), then one can plausibly explain evil as part of God's plan or to some extent out of God's control. This strategy, however, has had limited appeal to theists who hold that God is essentially omnipotent, omniscient, and omnibenevolent. See Rowe, William, ‘The Problem of Evil and Some Varieties of Atheism,’ in D. Howard-Snyder (ed.), The Evidential Argument From Evil (Bloomington: Indiana University Press, 1996), pp. 1–11. First published in American Philosophical Quarterly 16 (1979), pp. 335–341.
Why Theists Cannot Accept Skeptical Theism
(2) An omnipotent, omniscient, wholly good being would prevent the occurrence of any intense suffering it could, unless it could not do so without thereby losing some greater good or permitting some evil equally bad or worse.3 (3) Therefore there does not exist an omnipotent, omniscient, wholly good being. Premise (1) of Argument A is supported by the following argument, which we will term Argument B: (i) There is an enormous variety and profusion of intense human and animal suffering in the world. (ii) Much of this suffering seems unrelated to any greater goods (or the absence of equal or greater evils) that might justify it. (iii) Such suffering as is related to greater goods (or the absence of equal or greater evils) does not, in many cases, seem so intimately related as to require its permission by an omnipotent being bent on securing those goods (the absence of those evils). (iv) Therefore, it is more reasonable to accept (1) than to withhold judgment on (1).4 The key to Argument B’s support of premise (1) is the principle of credulity, according to which it is reasonable to believe that what seems to us to be the case is in fact the case. The principle works in the present instance by allowing us to infer from (i), (ii) and (iii) – which hold that gratuitous evil seems to exist – to (1) – which holds that gratuitous evil does exist. And it is on the strength of the latter contention, combined with (2), that we can reasonably infer that God does not exist, according to Argument A. In response to Argument A, many philosophers5 have attacked (1) by taking issue with its supporting Argument B. These thinkers have suggested that human epistemic limitations are such that we aren’t in a position to infer from the fact that there don’t seem to be greater goods served in many instances of evil to the fact that there are no greater goods served in many instances of evil. For all we know, these thinkers contend, there exist greater goods that are beyond our ken which both
This principle, insofar as it involves weighing goods and evils, suggests the utilitarian calculus. It may be objected, however, that deontological or virtue-ethical conceptions of morality should be taken into consideration here (either exclusively, or in concert with consequentialist considerations). I do not believe that such modifications would affect my argument, but for the purposes of the present paper this possibility will not be explored.
Rowe, ‘The Problem of Evil and Some Varieties of Atheism,’ op. cit., fn 5, pp. 10–11.
See, for example, Wykstra, Stephen, ‘The Humean Obstacle to Evidential Arguments from Suffering: On Avoiding the Evils of “Appearance,”’ International Journal for Philosophy of Religion 16 (1984), pp.73– 93; Wykstra, Stephen, ‘Rowe’s Noseeum Arguments From Evil,’ in D. Howard-Snyder (ed.), The Evidential Argument From Evil (Bloomington: Indiana University Press, 1996), pp. 126–150; Alston, William, ‘Some (Temporarily) Final Thoughts on Evidential Arguments from Evil,’ in D. Howard-Snyder (ed.), The Evidential Argument From Evil (Bloomington: Indiana University Press, 1996), pp. 311–332, and Howard-Snyder, D., Bergmann, M., and Rowe, W., ‘An Exchange on the Problem of Evil,’ in W. Rowe (ed.), God and the Problem of Evil (Oxford: Blackwell Publishers Ltd., 2001), pp. 124–158.
entail the existence of the evil we experience and outweigh or trump that evil.6 And if this is the case, then God possesses a morally sufficient reason for allowing evil that is beyond our ken. Thus we cannot infer from the existence of apparently gratuitous evil to the non-existence of God. This response to the problem of evil has come to be called ‘skeptical theism’ because it defends theism against the problem of evil by invoking skepticism about our capacity to know the full range of goods, evils, and entailment relations between goods and evils. Variations on this theme have been written by Daniel Howard-Snyder, Michael Bergmann,7 Stephen Wykstra,8 and William Alston,9 among others. In the remainder of this section, I develop ST in more detail by elaborating the arguments given by Wykstra and Alston before proceeding, in in the third section, to tender a collection of criticisms of ST as a satisfactory response to the problem of evil. Stephen Wykstra’s Skeptical Theism: CORNEA and the Parent Analogy Wykstra’s skeptical theism involves a denial of Rowe’s premise (1) on the basis of an epistemic principle he terms CORNEA (condition of reasonable epistemic access). According to Wykstra, ‘CORNEA says that we can argue from “we see no
It could be claimed that ST need not be committed to the claim that evils might be necessary for the actualization of higher unknown goods. It could be held that God’s permission of those associated (but not necessarily associated) evils might be justified in the absence of this necessary connection. The key idea here is that it is God’s permission of the evil, rather than the evil itself, that is necessary for the actualization of a higher good. Furthermore, it could be held that God’s permission of an evil state of affairs could be justified even if the outweighing higher good is never actualized. This alternative understanding of ST can be given substance through construing the good in question as freely chosen goodness: God wills the actualization of freedom because it is necessary for the actualization of the great good of freely chosen goodness, and God permits the evils that accompany misuses of freedom because of the higher value of freely chosen goodness – a value which cannot exist without God’s permission of the evils accompanying the misuse of that freedom. The specific evils that (can) take place are not necessary for the actualization of freely chosen goodness, but the permission of their possibility is. Moreover, given the great value of freely chosen goodness, God is justified in permitting evils that attend the misuse of freedom, even in cases where freely chosen goodness is never actualized. Although this example is cast in terms of a good that is within our ken, it is possible to suppose that there are other such goods beyond our ken that may support a version of ST. Yet although this is a coherent alternative to the standard accounts of ST, I do not support it, primarily because it seems to run afoul of a necessary condition for a valid morally sufficient reason which would justify God in permitting suffering, viz., that the sufferer of the (possible) evil in question must be compensated for suffering that evil. In accepting this alternative account, one holds that God accepts the suffering of persons for the sake of a good which not only may not be actualized, but which, when not actualized, results in uncompensated suffering. That the suffering in question would be uncompensated is entailed by the claim that the suffering is justified by the possibility of a higher good – the compensating good – which may not be actualized. This possible result seems to seriously impugn God’s goodness. Much more could be said on this account, but for now I will lay the issue to one side. I am indebted to an anonymous referee from Sophia for bringing these considerations to my attention.
7 Howard-Snyder, D., Bergmann, M., and Rowe, W., ‘An Exchange on the Problem of Evil,’ in W. Rowe (ed.), God and the Problem of Evil (Oxford: Blackwell Publishers Ltd., 2001), pp. 124–158. 8
See footnote 5 for the Wykstra references.
Alston, William, ‘The Inductive Argument from Evil and the Human Cognitive Condition,’ in D. Howard-Snyder (ed.), The Evidential Argument From Evil (Bloomington: Indiana University Press, 1996), pp. 97–125.
Why Theists Cannot Accept Skeptical Theism
X” to “there is no X” only when X has “reasonable seeability” – that is, is the sort of thing which, if it exists, we can reasonably expect to see in the situation.’10 Thus, for example, it would be a violation of CORNEA for me to claim that there are no South Africans living in North Korea. I am not aware of any South Africans living in North Korea, but it would be going beyond my epistemic rights in this situation for me to positively claim that there are none there. I lack reasonable epistemic access to such data. The proper response to alleged facts which fail to satisfy CORNEA is neither affirmation nor denial, but agnosticism. Wykstra applies the CORNEA principle to Rowe’s premise (1). This premise, according to Wykstra, holds only if we have reasonable epistemic access to the complete or nearly complete range of possible goods that might justify God in allowing instances of intense suffering: ‘By CORNEA, Rowe’s noseeum argument works only if it is reasonable for Rowe to believe that if there were God-purposed goods served by all suffering, it is likely that we would have something different than his noseeum data.’11 Wykstra contends that it is highly doubtful that we should have epistemic access to such states of affairs: given the fact that God so greatly transcends us, we should not expect to know if there are any goods which could provide morally sufficient reason for God to allow the existence of horrendous evil. He illustrates our cognitive situation in relation to God with ‘the parent analogy’: [I]t might not be unlikely that we should discern some of [the goods which justify God in allowing certain evils]...But that we should discern most of them seems about as likely as that a one-month-old should discern most of his parents' purposes for those pains they allow him to suffer – which is to say, it is not likely at all.12 If this is correct, then there well might be morally sufficient reasons tied to the actualization of higher goods which are beyond our ken and which justify the existence of the evil we experience as gratuitous. Given that this is the case, we lack epistemic license to Rowe’s premise (1), and thus cannot use it in an argument from evil. William Alston’s Skeptical Theism: The Many Facets of our Cognitive Limitations Alston’s argument is essentially the same as Wykstra’s, but Alston’s nicely complements Wykstra’s by going into more detail concerning the ways in which we are epistemically challenged in relation to Rowe’s premise (1). According to Alston, the following six considerations suggest that our epistemic grounds for access to the full range of orders and modalities of goodness and evil are slight indeed: (i) Lack of data: We still know very little regarding such matters as ‘the secrets of the human heart, the detailed constitution and structure of the universe, and the remote past and future, including the afterlife if any...’ 10
Wykstra, ‘Rowe's Noseeum Arguments from Evil,’ op. cit., p. 126.
Ibid., p. 139.
Complexity greater than we can handle: ‘Most notably there is the difficulty of holding enormous complexes of fact...together in the mind sufficiently for comparative evaluation...’ (iii) Difficulty of determining what is metaphysically possible or necessary: ‘[I]t is notoriously difficult to find any sufficient basis for claims as to what is metaphysically possible, given the essential nature of things, the exact character of which is often obscure and virtually always controversial. This difficulty is many times multiplied when we are dealing with total possible worlds or total systems of natural order...’ (iv) Ignorance of the full range of possibilities: ‘If we don’t know whether or not there are possibilities beyond the ones we have thought of, we are in a very bad position to show that there can be no divine reasons for permitting evil...’ (v) Ignorance of the full range of values: ‘[W]e are...in a very poor position to answer the question [of whether there exist unknown goods that justify God in allowing apparently gratuitous evil] if we don’t know the extent to which there are modes of value beyond those of which we are aware. For in that case, so far as we know, [the evil] may be justified by virtue of its relation to one of those unknown goods...’ (vi) Limits to our capacity to make well-considered value judgments: we face tremendous ‘difficulty in making comparative evaluations of large complex wholes.’13 (ii)
Incorporating and expanding upon these factors, Alston diagnoses the atheist’s epistemic warrant for premise (1) as highly disputable on two fronts. Not only must the atheist be able to reasonably deny the possibility that there are goods beyond our ken which justify the allowance of evil, but he must also be able to demonstrate that theodicies based upon goods within our ken fail. The former seems patently impossible, given that such goods, if they exist, are beyond our ken; and the latter seems nearly impossible, given the wealth of theodicies that the atheist would have to undermine with his very limited epistemic powers. Alston concludes: Since it is in principle impossible for us to be justified in supposing that God does not have sufficient reasons for permitting [evil] that are unknown to us, and perhaps unknowable by us, no one can be justified in holding that God could have no reasons for permitting [any given case of apparently gratuitous suffering].14 Similar arguments have been given by different proponents of ST. The upshot of this line of reasoning is that we have no good reason to think that we’re epistemically situated to assert (1). And if this is not possible, the problem of evil is a non-starter.
See Alston, ‘The Inductive Argument from Evil and the Human Cognitive Condition,’ op. cit., p. 120.
Ibid., p. 119.
Why Theists Cannot Accept Skeptical Theism
The Case Against Skeptical Theism One of the dominant lines of criticism of ST – developed, for example, by Bruce Russell15 – has been a direct counterattack against ST consisting in attempts to strengthen the case for our epistemic right to (1). These attempts have in turn been met with bolstered ST arguments.16 However, the line of criticism that I will develop below is a bit different, and, I believe, more potent. For the purposes of my argument, I will grant that all arguments to the effect that we have epistemic license to (1) can be defeated. I will present ST in what its proponents would consider its strongest form. What I hope to show, however, is that the result is highly detrimental to theism. Ironically, it turns out that the extent to which ST is successful is directly proportional to the extent to which epistemic access to the theism it is invoked to defend is undermined. It should be noted that the cumulative-case argument that I provide is directed towards ST in conjunction with both expanded and restricted forms of theism.17 Neither ‘restricted’ nor ‘expanded’ theists can endorse ST without sacrificing epistemic license to a variety of core doctrinal commitments whose preservation is necessary for the maintenance of their belief-system. It is in this sense that ST, by opening a Pandora’s box of skepticism towards the epistemic rights to theism itself, is largely self-defeating. My cumulative-case argument consists of the following three points: (P1) ST has the effect of making confident knowledge (i.e., knowledge not vulnerable to reasonable skepticism) of the correctness of our own conception of goodness, and confident knowledge of God’s actual moral will, at the least highly questionable and at the most impossible; (P2) There is good reason to think that ST threatens a great deal of theological knowledge; (P3) The epistemic limitation considerations contained in ST can be consistently used in reverse: that is, they can be used to disclaim any ascription of certitude or likelihood that apparently good events are evidence of the work of a good God. A variant on point (P3) has been made before in the literature, but as it is especially germane to my argument, I believe it warrants revisiting.18 I will take each of these points in turn, but will give more attention to (P1) than to (P2) and (P3). 15 See Russell, Bruce, ‘The Persistent Problem of Evil,’ Faith and Philosophy 6:2 (1989), pp. 121–139; also Russell, Bruce, ‘Defenseless,’ in D. Howard-Snyder (ed.), The Evidential Argument from Evil (Bloomington: Indiana University Press, 1996), pp. 193–205.
See Wykstra, ‘Rowe’s Noseeum Arguments from Evil,’ op. cit., and Alston, ‘The Inductive Argument from Evil and the Human Cognitive Condition,’ op. cit., and ‘Some (Temporarily) Final Thoughts on Evidential Arguments from Evil,’ op. cit.
Here I am using a distinction first introduced by William Rowe. For his most recent use of this distinction, see Rowe, ‘Friendly Atheism, Skeptical Theism, and the Problem of Evil,’ International Journal for Philosophy of Religion 59:2 (2006), pp. 79–92, esp. pp. 83–4. I am indebted to two referees from Sophia for stressing the need for this clarification. See Wesley Morriston, ‘The Evidential Argument from Goodness,’ The Southern Journal of Philosophy Vol. XLII (2004), pp. 87–101.
(P1) ST works to the extent that it is reasonable to hold that our knowledge of the modalities and orders of goodness is limited. A proper discernment of this limitation, ST argues, entails suspension of judgment on the extent of possible modalities and orders of goodness and their relations to evil, and a correlative agnosticism regarding the first premise of Rowe’s argument from evil. The difficulty with this position, however, is that if we are to take it seriously as a defeater of the problem of evil, then we also have to take seriously two correlative possibilities: (A) Our cognitive limitations are such that we cannot have confident knowledge (i.e., knowledge invulnerable to reasonable skepticism) of the nature of goodness. (B) Our cognitive limitations are such that we do not know God’s relation to goodness. Certainly it seems that no theist can accept these claims in their strongest form and maintain theism at the same time. This is most clearly the case regarding expanded forms of theism. The Ten Commandments, for example, imply their rejection, as do the ethical claims in the Torah and the Koran. One cannot be Christian, Muslim, or Jewish without believing that one has access to God’s actual moral will (to a considerable extent) and to a correct understanding of goodness (at least most of it). Yet it seems that the very act of stressing our cognitive limitations in relation to the possible modalities and orders of goodness suggests A and B as live possibilities. To begin with (A), consider the following three accounts of the nature of our cognitive limitations in relation to the nature, modalities, and orders of goodness: (a) It is possible that our cognitive capacity is such that we cannot discern certain higher goods which somehow entail the evil we experience as gratuitous. On this account, the unknown higher goods do not conflict in any way with our understanding of the nature of goodness. None of them entail that what we consider good is in fact not good, or that what we consider evil is in fact not evil, or that the entailment relations we discern between known goods and evils are erroneous. Rather, what we consider good is in fact good, and what we consider evil is in fact evil, and the entailment relations between them that we discern are correct. The only difference is that there exists a higher but unknown order of goods whose (justified) actualization by God entails certain instances and sorts of evil that we, given our epistemic limitations, experience as gratuitous. This is the picture that the ST proponent paints. It is consistent with the epistemic maintenance of theism, and, supposing it to be true, it undermines Rowe’s premise (1), without which the argument from evil cannot get off the ground. (b) It is possible that our cognitive capacity is such that our perception of the correct hierarchical relations between goods is erroneous. On this account, what we consider good is in fact good, and what we consider evil is in fact evil; but our understanding of the correct hierarchical relations between goods – what trumps what – is, as entailed by an unknown higher good, faulty. For example, it might be the case that although we perceive free will to be a higher good than sexual pleasure, in fact some unknown higher good entails that the reverse is
Why Theists Cannot Accept Skeptical Theism
true. This certainly seems wrong to us: free will appears to be a higher good than sexual pleasure; but given the epistemic limitations introduced by ST, it seems we haven’t the warrant to judge the matter. It might be asked how an unknown higher good could have such an effect. In response it can only be said that a concrete example cannot be given - since giving such would involve specifying a good that is beyond our ken, which is patently impossible – but the possibility can be suggested through a description of a similar process at work in goods within our ken. Consider the following historical example. It used to be believed by many (and is indeed still believed by some) that forcible imposition of the benefits of advanced Western culture upon less technologically advanced peoples is a greater good than noninterference with such cultures. The subsequent perception and wide acceptance of the greater good of respect for persons, however, has suggested that this former understanding of the hierarchy of goodness is incorrect. It is suggested today that, contrary to what was previously believed, the higher value of respect for persons implies that it is in fact a greater good to leave such cultures (relatively) unmolested and to peacefully learn from them, and a lesser good to forcibly impose the benefits of technologically advanced Western culture upon them. If this suggestion isn’t palatable to some, then a few moments’ reflection should result in a suitable example. The replacement of individualistic Homeric virtues by later Greek virtues of active citizenship, or the replacement of Old Testament values with New Testament values might serve as suitable examples. In these cases and in countless others, our perception of normative orders has undergone modification from a state of prior error to a state of progress, and our ordering of the relative weight of certain goods has been recast or even reversed, as informed by a new appreciation of a higher good that implies such a reordering. Certainly not all historical modifications in the perception of the relations between values and normative orders have been progressive; but just as certainly some such modifications have been. The suggestion being made in connection with the current discussion is that the epistemic limitations suggested by ST open the possibility that there exist unknown higher goods which entail that our apprehension of the correct placement of the orders of goodness is substantially incorrect. The possibility of error mentioned above (in relation to our erroneous apprehension of the value of pleasure as against free will) is just one such case. (c) It is possible that our cognitive capacity is such that we perceive certain things to be goods which in fact are not goods in relation to – as entailed in some way by – certain unknown higher goods. In this scenario, what we take to be good only appears so to us; but the truth is rather that it either has no moral value, or it is in fact bad, or evil. For example, it might be the case that, for reasons beyond our ken, some unknown higher good or entailment relation between such a good and the values within our ken entails that concern for people beyond one’s immediate family has no moral value whatsoever. Or it might be the case that some unknown higher good entails that setting a day aside for prayer is positively wrong. Such notions certainly strike us as incorrect, but according to the epistemic perspective that informs ST we lack the warrant to conclude that they are in fact wrong. Again, concrete examples of this sort of
influence cannot be given, but examples from cases of such ‘good-trumping’ within our ken can serve to clarify what I am suggesting. Take, for example, the case of racism. For racists (past and present), it is considered a good thing to enforce inequality between certain races. This is based upon the conviction that the ‘lesser’ race has, in some way, inherently less value than the ‘greater’ race. Further ethical development, however, suggests that there is a higher good – again, we can use respect for persons – which shows racism to have not been a good at all, but just the opposite. In this case, a higher good entails that what was considered good is in fact evil. Further examples are not hard to find: cannibalism has been considered a good before the higher notion of the good of respect for human persons has recast the activity as evil. And so on. What has been suggested here is that, given the epistemic limitations in relation to the orders and modalities of goodness engendered by ST, it well might be the case that some of the things we strongly consider goods are, in a similar way, overturned by unknown higher goods and their entailment relations with putative goods (and evils) within our ken. If this is the case, our knowledge of the nature of goodness might be in large part erroneous. These three scenarios are all consistent with the possibilities for cognitive limitation contained in ST, and, given those limitations, we are not in a position to deny any of them. Now, it is clear that theists cannot accept many of the claims associated with (b) – and especially (c) – and retain epistemic license to core aspects of their very theism. To be a practicing theist of any stripe requires epistemic license to the claims that certain relations in the hierarchy of value are known and unassailable (e.g., that piety is a greater good than individual pleasure), and that our apprehension of goodness is, in certain respects, likewise known and unassailable (e.g., that it is good to help others). Yet to the extent that we are epistemically challenged in relation to the orders and modalities of goodness, it is possible that our cognitive situation is correctly described by (b) or (c). Which scenario correctly describes our cognitive situation? Although theists would of course choose (a), it would be more consistent with the epistemic limitation engendered by ST to say that we cannot say: this too must be grouped under judgments for which we lack epistemic license. Turning to (B), an even more problematic possibility suggests itself. If it is possible that the true orders and modalities of goodness are, given our epistemic limits, inaccessible, then, given that God takes these true orders and modalities into account, what confidence can we have that our conception of God’s relation to goodness is at all accurate? Consider these two possibilities: (α) God’s relation to goodness, for reasons beyond our ken, necessitates that God actively enjoys human suffering. (β) God’s relation to goodness, for reasons beyond our ken, necessitates that God cares not a whit for human affairs whatsoever. Neither of these alternatives are conclusively ruled out by the epistemic humility engendered by the considerations advanced in support of ST; and insofar as ST works to the extent that it relies upon the possibility that our conceptions of such things are limited (and hence possibly quite flawed), (α) and (β) – situations wholly
Why Theists Cannot Accept Skeptical Theism
anathema to any form of theism – well might be the truth. We simply do not know.19 These possibilities present themselves with all the greater force to the extent that the epistemic limitations invoked in ST are taken seriously. The harder a proponent of ST campaigns to show that we lack the epistemic access to make justified pronouncements about the nature, orders, and modalities of goodness, the more plausibility is given to the thesis that we well might not have the correct perspective on any of the matters described above. Note, however, that I am not arguing that it is in fact likely that we cannot know what goodness is or that we cannot have confident knowledge of God’s relation to goodness. What I am arguing is that if one accepts the epistemic constraints imposed by a strong rendering of ST, then one is not in a position to judge that the unwanted possibilities canvassed in this section are unlikely. (P2) The foregoing considerations suggest a more pervasive problem for the theist who wishes to make use of ST. My argument in (P1) suggests that ST precludes confidence in our cognitive capacity to grasp the true orders and modalities of goodness and the nature of God’s relation to goodness. Yet it seems that the epistemic situation might be even worse for the ST proponent, for it seems to undermine the possibility that we could have much confident knowledge of divine states of affairs whatsoever. As the proponent of ST stresses, God is the sort of being, regarding whom, in Alston’s words, ‘[we] have a very sketchy idea of what is in that territory, and...no sufficient basis for an estimate of how much of that territory falls outside [our] knowledge;’20 and therefore ‘It is surely the better part of wisdom to acknowledge that we are groping in the dark in assessing the extent to which we can survey the whole field.’21 Consequently, it is entirely possible that positive or even probable pronouncements concerning a great many of the claims about the nature of the divine are beyond our epistemic rights. Any attempt by a proponent of ST to make claim to knowledge of this sort can be met with the same sorts of doubtmakers that Wykstra and Alston employ in attempting to defuse the problem of evil. To echo Alston: Do we have access to all the relevant divine data? Doesn’t God possess more complexity than we can handle? Can we really determine here what is metaphysically possible or necessary? Do we have a confident grasp of the full range of possibilities regarding such an entity? And so on. What, then, can we say with epistemic confidence regarding any aspect of our knowledge of God? How can we consider ourselves sufficiently justified in most or even some of our claims concerning what God is, or what God is capable of, or what God has done, or what the exact nature of our relationship with God is, or what our relationship with God should be? Certainly no theist can maintain epistemic license to her theism while conceding that such broad swaths of subject-matter are beyond our epistemic 19
Richard Gale has written of a related practical problem for the theist which fits in well with my point that skeptical theism makes knowledge of God's moral will difficult. ‘The major problem faced by the moral-inscrutability-of-God version of defensive skepticism is that it seems to preclude our being able to enter into [loving] relationships with God, thereby undercutting the very purpose for which God created us according to theism, namely to enter into a communal relation of love with God.’ See Gale, Richard, ‘Some Difficulties in Theistic Treatments of Evil,’ in D. Howard-Snyder (ed.), The Evidential Argument From Evil (Bloomington: Indiana University Press, 1996), pp. 209–211. 20
Alston, ‘The Inductive Argument from Evil and the Human Cognitive Condition,’ op. cit., p. 120.
Alston, ‘Some (Temporarily) Final Thoughts on Evidential Arguments from Evil,’ op. cit., p. 318.
capacity to such an extent that reasoned or probable judgments about them are impossible. Such problems have bite in direct proportion to the extent that ST is taken seriously. Skeptical theism as a response to the problem of evil, like a virus, threatens to become skepticism about theism as a whole. Notice that I have not argued that the skepticism introduced by ST infects our access to knowledge across the board. Some thinkers22 have made this case, and I support Alston when he responds: [I have been arguing that] when [our normal epistemic standards] are applied to the kind of claim exemplified by Rowe’s (1), it turns out that this claim is not justified and that the prospects for any of us being justified in making it are poor at best. This is because of the specific character of that claim, its being a negative existential claim concerning a territory about the extent, contents, and parameters of which we know little.23 Although Alston only claims that skepticism is appropriate for negative existential claims concerning territories about which we know little, there seems no reason not to include skepticism for positive existential claims about such regions as well. And this response seems correct. There is an important difference in our epistemic access to certain facts about the world (e.g., that my car will not start) and our epistemic access to the transcendent. But I fail to see how Alston can suppose that the epistemic limitation introduced in relation to Rowe’s premise (1) can be restricted to Rowe’s premise (1). And in fact Alston gives no arguments by way of attempting to justify this restriction. He implies that the restriction is defensible – ‘my argument...is not based on, nor does it support, a general theological skepticism. It is compatible with our knowing quite a bit about the divine nature, activities, purposes, and relations with humanity.’24 – but he provides no developed defense. What he provides instead are the makings of a very sharp double-edged sword. (P3) The final argument combines elements from (P1) and (P2), and hence could have been mentioned above,25 but I have decided to present it separately because it appears to warrant special attention, given that it bears especially on the core theistic practice of assuming epistemic warrant in ascribing divine significance to historical events. If ST is taken seriously, what reason would a theist have for citing apparently good events and circumstances as evidence of the agency of a good God? Mightn’t it be the case that we lack the correct perspective to correctly judge these cases as well? This is especially problematic for those theists who accept the existence of an extremely powerful evil being such as Satan. Consider this scenario. We are presented with evidence of apparent goodness: say, what appears to be a miraculous
Richard Gale, for example, has made this case. See Gale, ‘Some Difficulties in Theistic Treatments of Evil,’ in D. Howard-Snyder (ed.), The Evidential Argument From Evil (Bloomington: Indiana University Press, 1996), pp. 209–211.
23 Alston, ‘The Inductive Argument from Evil and the Human Cognitive Condition,’ op. cit., p. 121. (Italics added.) 24
Alston, ‘Some (Temporarily) Final Thoughts on Evidential Arguments from Evil,’ op. cit., p. 321.
I am indebted on this point to an anonymous referee from Sophia.
Why Theists Cannot Accept Skeptical Theism
healing. This is adduced as evidence in favor of the existence and agency of a perfectly good God. For most theists, this move will be unproblematic, but not, it seems, for those who take ST (and Satan’s existence) seriously. For the latter, on pain of inconsistency, must admit that it well might be the case that what appears as an instance of great good is in fact an act of great evil orchestrated, in a way that we cannot understand, by Satan. The fact that this event appears good does not provide sufficient reason to believe that it is not actually a necessary correlate of evil, for precisely the same reasons cited in favor of the opposite thesis. Satan may have immorally sufficient reason(s) for bringing about what appears to be good, but which is a necessary correlate of an outweighing evil. We are precluded, by the very reasoning that supports the claim that we cannot cite apparent evil as evidence against the existence of a good God, from citing apparent good as evidence for the existence of a good God. As with (P1) and (P2), this conclusion appears unacceptable for theists, not only because it precludes confident ascription of God’s agency in their own lives, but more importantly because it precludes confident ascription of God’s agency in history. Christianity, Islam, and Judaism are in large part historical religions, and to be an adherent of one of them is, in part, to claim epistemic license to confident knowledge of certain pivotal historical events that substantively informed and shaped that religion, and without which one cannot consider oneself a true believer in that religion. Although differing in certain respects, all three of these arguments share a common (and oft-repeated) theme: skeptical theism can be used to dissolve the problem of evil only at the cost of the maintenance of epistemic license to crucial aspects of the theism it is invoked to defend. Unless a principled way can be found to keep the skepticism engendered by the ST position within theism-friendly borders, theists should abandon the ST strategy for defusing the problem of evil. Whether it is possible to defend such a restriction is the theme of the next section.
Is a Theism-friendly Moderate ST Philosophically Defensible? In response to an earlier draft of the present essay, one of the paper’s referees presented the following objection: [The problems that have been raised] could be defused by adopting a moderate form of skeptical theism. A theist who admits that he or she does not know all of God’s reasons for permitting evil...can still go on to claim that he or she knows something about God, such as that God became incarnate in Jesus Christ or that God answers our prayers. If the theist went on to cast doubt on our entire moral landscape, then that may have the kinds of consequences the author points out – but a skeptical theist need not go so far. Michael Bergmann and Michael Rea take a similar line in response to the critical arguments of Michael Almeida and Graham Oppy: [E]ven if [the tenets of ST imply] that we do not know much about the realm of value, they do not at all imply that we know nothing about that realm; and, in
particular, they do not imply that we lack knowledge of God’s commands as God’s commands.26 This form of response is very well taken. The primary objection that can be made to the arguments advanced above is that, contrary to the fears of a slippery slope of damaging skepticism, it is possible to defend a moderate or limited form of ST, one which supports the theism-friendly picture of cognitive limitation contained in (P1) (a) above, but which offers principled resistance against sliding into further forms of damaging skepticism. I will term such a position ‘moderate skeptical theism.’ This position might also be termed ‘just-so skeptical theism,’ in that it holds that the skepticism engendered by the theses constitutive of ST extends just to a particular realm of human knowledge (viz., human knowledge of possible God-justifying reasons to permit apparently gratuitous evil) but no farther; but as this formulation sounds uncharitable, I prefer the former term.27 In this section I canvass a number of ways in which moderate ST might be established, and argue that all of them are subject to serious difficulties. Specifically I will examine arguments to establish moderate ST that are based on the following three forms of justification: (i) the evidence afforded by faith-based pronouncements; (ii) the evidence afforded by independent philosophical argument for the existence of a God who would create a world in which skepticism would be restricted to that described in (P1) (a); and (iii) the evidence afforded by a proper understanding of the internal limits imposed by the theses that constitute the ST position. I contend that these responses either fail to extract ST from slipping into the damaging forms of skepticism canvassed above, or they fail to give us adequate philosophical reason to accept the limitation of skepticism within theism-friendly borders. I conclude that at the end of the day, the best recourse for a theist wishing to be a skeptical theist is to justify the restriction of her skepticism within theism-friendly borders through faith alone. True, to take this position is to give up the hope of providing a philosophical defense of ST as a response to the problem of evil, but given the difficulties that attend such an attempt, it well may be the least problematic option for theists who find the moderate ST position reasonable. (i) Grounds for moderate ST: Faith-based pronouncements Writing in response to criticism of ST tendered by Graham Oppy and Michael Almeida, Michael Bergmann and Michael Rea have drawn attention to the fact that when discussing ST one has to keep in mind that ‘Skeptical theists, after all, are theists. Thus, when [theists] consider the bearing of skeptical theism on their moral practice, they will inevitably and quite sensibly do so in a way that takes account of other things that they believe.’28 Although Bergmann and Rea speak specifically in response to putative ST-based difficulties for theism related to moral practice, a Bergmann, Michael and Rea, Michael, ‘In Defence of Sceptical Theism: A Reply to Almeida and Oppy,’ Australasian Journal of Philosophy 83:2 (June 2005), p. 244.
The Oxford Dictionary of Philosophy defines a ‘just so story’ as ‘An explanation of an admittedly speculative nature, tailored to give the results that need explaining, but currently lacking any independent rationale.’ Oxford Dictionary of Philosophy, ed. Simon Blackburn (Oxford: Oxford University Press, 1996), p. 204.
Bergman and Rea, ‘In Defence of Sceptical Theism: A Reply to Almeida and Oppy,’ op. cit., p. 244.
Why Theists Cannot Accept Skeptical Theism
widening of focus to encompass putative ST-based difficulties related to theism in general seems to be unproblematic and even warranted. This seems quite right. The reception and consideration of criticism of particular aspects of one’s belief-system does not take place in a vacuum, and conclusions suggested by isolated lines of argument may be trumped by the prevailing weight of competing argumentation based on other strands of one’s web of belief. In relation to the present topic, theists may well respond that isolated warnings about the danger of expanding skepticism engendered by acceptance of ST fail to consider competing aspects of theism which, when taken into account, serve to nullify such fears by providing grounds for the restriction of ST to moderate ST. One such response is based on faith – faith in the veracity and validity of religious doctrines that entail or imply that our cognitive situation is correctly described in (P1) (a) above, and that further skepticism of a theism-damaging sort is unwarranted. For greater specificity we can examine a faithbased defense of moderate ST that might be provided by a Christian theist. The key resource for such a defense is the (alleged) divine revelation contained within the Christian sacred scriptures. There are many passages within the Bible which could be used to support a moderate ST reading. The Book of Job recounts a conversation between God and Job, in which God berates Job for having the presumption to think that he can challenge God for seemingly abandoning him, or fathom the reasons behind his misfortunes. According to this revelation, it could be claimed, although we cannot fathom the reasons for our misfortunes, we can take heart there are indeed reasons.29 In the Letter of Paul to the Romans, Paul writes, ‘We know that in everything God works for good with those who love Him, who are called according to His purpose.’30 This revealed wisdom, it could be claimed, assures believers that although we don’t know God’s reasons for evil, everything – even apparently gratuitous evil – is ultimately for the good. Furthermore, in the First Letter of Paul to the Corinthians, Paul writes, ‘Has not God made foolish the wisdom of the world? ...in the wisdom of God, the world did not know God through wisdom...For Jews demand signs and Greeks seek wisdom, but we preach...Christ the power of God and the wisdom of God.’31 This revealed wisdom, it could be claimed, assures us that God has reasons beyond the reach of our human philosophy, reasons which satisfactorily explain all things, including why apparently gratuitous evil things must happen. Further passages could be cited as testimony that our knowledge of many of the most important hierarchies of goodness, of what we know to be good and evil, of God’s relation to goodness, of the core doctrines of Christianity, and of the attribution of divine agency to certain historical events, is all quite beyond question. All of this, it could be claimed, is known through faith – ‘the assurance of things hoped for, the conviction of things not seen’32 – and this faith-
This may not be the best story to use, insofar as the reasons behind God’s allowance of Job’s suffering seem to be given in the beginning of the Book of Job, and seem quite accessible to human understanding, but the general point is clear enough: humans lack the power to comprehend both God and God’s reasons for allowing misfortune to beset the faithful.
Letter of Paul to the Romans 8:28.
First Letter of Paul to the Corinthians 1:20–21.
Letter to the Hebrews 11:1.
based knowledge could be said to provide unshakeable assurance that moderate ST is the proper response to the problem of evil. As a first response to this argument, it can be seen that it appears to render the philosophical work contained in ST superfluous. The faith-based pronouncements contained in divine revelation seem to be more than capable of doing all of the work attempted by proponents of moderate ST: carving out a realm of human ignorance in relation to knowledge of God’s reasons for allowing apparently gratuitous suffering, and showing that skepticism about further key areas of knowledge is unwarranted. True, in taking this route, one would, as Almeida and Oppy have noted, ‘be giving up the skeptical theist ambition: it is no longer true that it is merely considerations about our cognitive limitations that yield the desired conclusion,’33 but as long as the desired conclusion is reached by valid means, that seems beside the point. But this leads to the second response to this form of argument: are purely faith-based means for establishing moderate ST valid? That depends on whether the claims contained in Christian revelation are in fact true. If they are true, and faith is the proper means to the appreciation of these truths, then it would seem that purely faith-based means are indeed valid. But are the claims true? Normally we as philosophers seek to determine truth on the basis of rational argument, but in this case rational argument appears to be inapplicable. The basis of this response to the problem of evil is precisely one of faith, and as a response of faith, rational argument or worldly wisdom seems to have no purchase; for ‘Has not God made foolish the wisdom of the world?’ At this point it appears that we as philosophers have no more to say. This faith-based defense of moderate ST is only valid if the religious revelations cited in its support are true; and yet the very nature of this defense precludes us from subjecting these putative truths to rational scrutiny. The effect of any purely faithbased defense of moderate ST, then, is to remove the discussion from the philosophical realm entirely, and to place it within the unassailable bounds of fideistic theology. For this reason it can be said that although this response could be valid, it cannot claim philosophical validity. For many theists this conclusion is quite in order; but for those who wish to provide a philosophical defense of moderate ST, this avenue is closed. (ii) Grounds for moderate ST: outweighing philosophical argument It is possible, however, to seek to give a defense of moderate ST that is grounded in philosophical argumentation. Importantly, none of the groundwork in this strategy is done by faith-based pronouncements. It begins with an accession to the skeptical theses that constitute the ST position and the correlative claim that, on the basis of that reasonable skepticism, the problem of evil is dissolved. The strategy then attempts to rebut the claim that ST has unwanted and damaging epistemic implications for theism by arguing that other philosophically justified aspects of theism are sufficient to show that the skepticism associated with moderate ST is correct, and that it extends no further. There are numerous ways that such a strategy could be developed. It could be argued, for example – and to continue speaking in terms of Christianity – that Almeida, Michael and Oppy, Graham, ‘Sceptical Theism and Evidential Arguments from Evil,’ Australasian Journal of Philosophy 81:4 (December 2003), p. 508.
Why Theists Cannot Accept Skeptical Theism
rational considerations support accepting Christianity on the basis of the historical evidence of the miracles that Jesus performed, and given this acceptance, it could be argued, we are rationally justified in taking the pronouncements in the Bible as true. This is not, it should be quickly noted, a reversion to the pure faith-based approach examined above; the key difference is that, on this approach, acceptance of the Bible’s authority is based, not in a pure act of faith, but in rational assent to the authority of Biblical pronouncements on the basis of the historical evidence for Jesus’ miracles. Alternatively, it could be argued that consideration of the world’s design supports the hypothesis that it was created by a God who would ensure that we do not suffer from skepticism in all of the potentially damaging forms mentioned above. Or it might be possible to fashion an argument from religious experience with essentially the same conclusion.34 Many other such arguments, or combinations of arguments, are possible – too many to be dealt with in detail within the bounds of the present essay. Yet it is possible, I believe, to cast reasonable doubt on the adequacy of all such attempts to attenuate the force or scope of the skepticism engendered by acceptance of the ST position without having to dive into extended consideration of the details surrounding these various arguments. The source of this doubt is, unsurprisingly, the skepticism generated by accession to the skeptical theses that constitute the ST position. The basic difficulty is this: given the initial accession to cognitive limitation, any attempt to employ philosophical considerations concerning that which we can know of the divine will suffer from the same difficulties that beset attempts to make positive pronouncements on the non-existence of higher goods beyond our ken that may justify God’s allowance of putatively gratuitous evils. Any such argument, at some point, will have to infer from considerations within our ken – be they the apparent veracity of historical accounts of miracles, the apparent order of the world, apparent experience of the divine, or any other such consideration – to a positive pronouncement about the nature and will of an entity that is said to transcend human cognition. But what assurance can we feel in the accuracy – or even probability – of such conclusions, once skepticism about human cognition regarding matters beyond our ken has been entrenched? The CORNEA principle is employed by Wykstra to argue that just because we don’t experience X doesn’t mean that X doesn’t exist; but there seems no principled reason why the CORNEA principle cannot and should not be used in support of the opposite sort of conclusion, namely: just because we think we experience X – directly or indirectly – doesn’t mean that X exists. There are plenty of quotidian examples of such CORNEA violations, including mirages, hallucinations, wishful thinking, and other forms of erroneous theory-laden perception. If God is transcendent, what epistemic right do we have to think that we have access to 34
It could be charged that, insofar as religious experience refers to (putative) divine revelation, an argument based on religious experience is essentially an argument based on faith. The difference, however, can be seen in this: whereas a pure faith-based defense of moderate ST is grounded in ‘the conviction of things not seen,’ a defense of moderate ST based on religious experience is grounded in a conviction formed on the basis of things that have been (putatively) directly experienced. The latter is, thus, at its core, a variant of an argument based on perception (similar to that which one might make to justify belief in the desk at which one works, or the trees outside one’s window), rather than an argument based on a pure leap of faith.
signs of God’s nature and will, or, even if we have such access, that we could interpret them correctly? An ant possesses a form of consciousness, but presumably it is not sophisticated enough to rightly grasp the import of what theists take to be evidence for the divine; and what reason do we have to believe that our situation is any different? True, our cognitive abilities are a quantum leap above any ant’s, but when the conclusions we seek to infer from phenomena within our limited experience and according to our limited cognitive abilities relate to a supposedly infinite entity, significant doubt about the validity of such inferences seems inescapable. For all we know, God exists, but possesses a nature entirely different from that which theists suppose it to be. This doubt is only intensified when we take into consideration the variety of reasons for epistemic humility tendered by William Alston. Theists will likely wish to respond that we have good reason for believing that God exists, loves us, and seeks to communicate with us; but on what can such claims be based? The philosophical form of approach, based wholly within the deliverances of human reason, seems manifestly incapable of justifying such claims with any degree of probability, given precisely the reasons for epistemic humility that form the backbone of the ST position. The faith-based form of approach could work – supposing, again, that certain revelations are true, and that faith is the proper avenue of appreciation for such truths – but, as we have seen, such an approach lacks any distinctively philosophical validity. What about a combination position between faith and reason? Perhaps the deliverances of reason can put us in a position to accept divine revelation without having to make a pure leap of faith – or perhaps an initial leap of faith can aid us in the appreciation of the deliverances of reason that support moderate ST? Yet this approach as well seems to fail as a philosophical defense of ST; for at the end of the day the key components in the defense of moderate ST will either be based on faith – which lacks philosophical validity – or reason – which, given admitted human cognitive limitations, lacks certifiable purchase. (iii) Grounds for moderate ST: internal limits of the ST position A final possibility for the defense of moderate ST has been provided by Bergmann and Rea. Writing in response to criticism of ST tendered by Almeida and Oppy, who argue that acceptance of ST leads to unacceptable skepticism about moral practice in general, Bergmann and Rea write, First, note that a sample of xs can be representative of all xs relative to one property but not to another...For [the theses that constitute the ST position], what we are interested in is whether our sample of possible goods, possible evils, and entailment relations between them...are representative of all possible goods, possible evils, and entailment relations there are relative to the property of figuring in a (potentially) God-justifying reason for permitting the evils we see around us. Although that property is not explicitly mentioned in [the theses that constitute the ST position], it is representativeness relative to that property that [the theses that constitute the ST position] are speaking of.35
Bergmann and Rea, ‘In Defence of Sceptical Theism: A Reply to Almeida and Oppy,’ op. cit., p. 242.
Why Theists Cannot Accept Skeptical Theism
The key to this response is that criticism of ST based on the possibility of extending skepticism into further realms of (possible) knowledge is entirely off base, insofar as ST is explicitly designed to apply to our knowledge of axiological states of affairs ‘relative to the property of figuring in a (potentially) God-justifying reason for permitting the evils we see around us,’ and to nothing else. On this approach, critics of ST can be likened, say, to those who, upon hearing someone make claims about the lack of aesthetic value of artwork X, complain that such pronouncements can’t be right insofar as they entail similar claims about the aesthetic value of other artworks Y and Z, which are (or should be) immune to such criticism. The proper response to such criticism, it may seem, is one of clarification: ‘I am not talking about Y and Z. Y and Z have to be judged on their own merits. I am talking about X, and my judgments concern X, and nothing else.’ Speaking in terms of ST, it could be said that attempts to apply the skeptical theses of the ST position beyond the indicated internal parameters in an effort to show that doing so has damaging consequences constitute an unjustified extension of their proper application, and thus have no force. There are two difficulties with this defense of moderate ST, however. The first difficulty is that, even supposing that we restrict attention to the application of ST in the precise manner specified above, the ST position is wrong, insofar as we might have good reason to think that skepticism in regard to knowledge of the indicated property is unwarranted. As others have written on this possibility, I will not treat of it again here. The second difficulty, however, is that it appears unjustifiably ad hoc. True, it is possible in some cases to criticize a position on grounds internal to that position alone, but when the criticism is based on principles which apply just as well to other similar positions, an extension of that criticism to those other positions seems more than warranted. It would be evidently inappropriate, for example, to reply to the criticism that applying the principle of non-contradiction to proof A leads to unwanted implications in relation to proofs B and C by saying that one is only dealing with the principle of non-contradiction in relation to A, and that consideration of the consequences of applying that principle to proofs B and C is beside the point. The principle of non-contradiction is a logical principle that applies across all possible proofs. In the present case, proponents of ST claim that we should be skeptical of attempts to positively pronounce on the parameters of certain types of knowledge, insofar that those types of knowledge well may be beyond our ken, given human epistemic limitations. To justifiably apply this principle to different types of knowledge only requires that those other types of knowledge, in a similar fashion, well may be beyond our ken, given human epistemic limitations. And it is on precisely this basis that critics of ST have extended the skepticism associated with ST, and have concluded that the consequences of doing so are detrimental to theism in various ways. Without further argument, at least, there seem to be no good reasons why a restriction of ST in the theism-friendly manner described above is anything more than an ad hoc and unjustified restriction of the application of the epistemic principles that constitute the ST position in an effort to save that position from unwanted consequences. Without such argument, the extension of ST principles to similar epistemic cases is quite warranted, and results in the detrimental consequences that have been tendered by critics of ST.
It seems clear that the best strategy in response to the sort of criticism of ST given in this paper is to seek to defend a moderate form of ST, according to which skepticism is appropriate only in relation to the realm of unknown higher goods that could justify God in permitting apparently gratuitous evils, but not in relation to other realms of knowledge (or at least not in relation to realms of knowledge where the consequences of not knowing could be detrimental for theism). In this section I have examined three possible arguments for establishing moderate ST, and have concluded that they either fail to provide sufficient reason to halt the spread of skepticism, or (possibly) succeed only at the cost of the hope for a philosophically valid defense of moderate ST. At this point, as mentioned above, a theist may well wish to abandon the hope for a philosophical defense of moderate ST, and rely instead on an entirely faith-based approach. Such a maneuver may have the (theistic) merit of placing the desired conclusions beyond the reach of philosophical doubt, but it is also an accession that theists cannot accept ST as philosophically valid.
Final Comments In this paper I have given a cumulative case argument in support of the contention that skeptical theism fails as a satisfactory theistic response to the problem of evil. If my arguments have been persuasive, then the need for theists to develop an alternate response to the problem of evil should be clear. The problem of evil requires a theistic response, but that response cannot, I have argued, be based upon ascriptions of the sorts of epistemic limitations that have populated the skeptical theist position without inviting the virus of expanding skepticism that threatens epistemic license to core elements of theism itself. It may be possible to bypass such fears by basing a moderate ST position on faith in allegedly divinely revealed truths about what we can know and what we cannot, but this response only works if those claims are true, and by its very nature such a response cannot be assessed for validity using philosophical methods. For those seeking a philosophical solution to the problem of evil, something else is required.36
I would like to thank Matthew Piper, Aaron Cobb, Timothy Pawl, and three anonymous referees from Sophia for helpful comments on earlier drafts of this essay.