International Journal for Philosophy of Religion 53: 1–23, 2003. © 2003 Kluwer Academic Publishers. Printed in the Netherlands.
1
Moral evil: The comparative response C. STEPHEN LAYMAN Philosophy Department, Seattle Pacific University, 3307 Third Avenue West, Seattle WA 98119, USA
Theists can respond to the problem of evil in four general ways. First, they can offer a theodicy, i.e., an attempt to provide God’s reasons (actual or possible) for permitting evil.1 Second, theists may seek to show that arguments purporting to demonstrate a problem of evil for theism are flawed, e.g., such arguments may involve a questionable inference from “As far as we know, there is no reason that would justify God in permitting horrors” to “There is no such reason.” This response to the problem of evil has sometimes been called “Skeptical Theism.”2 Third, theists may simply admit that the problem of evil is a special difficulty for theism (i.e., a greater difficulty for theism than for its metaphysical rivals – some of which may have no problem of evil); nevertheless, natural theology and/or religious experience provide adequate reasons or warrant for theism. We might call this response to the problem of evil, “The Overrider Response” as it involves the claim that some reason or warranting factor overrides the evidence evil provides against theism.3 Fourth, theists may argue that, although theism does not explain the presence of all evils well, it provides an explanation that is as good as (or better than) the explanation provided by some (or all) of theism’s metaphysical rivals. Let us call this fourth approach “The Comparative Response” since it involves comparing theistic explanations of evil with explanations provided by theism’s metaphysical rivals. The Comparative Response has received little attention in recent discussions of the problem of evil, and I propose to develop it in this paper. Specifically, I shall argue that although theism may not explain all evils well, on the whole theism explains the presence of moral evil as well as (or better than) naturalism does. (I suspect that theism can explain the totality of evil as well as, or better than, naturalism does, but a discussion of natural evil is beyond the scope of this paper.) It will prevent some misunderstandings if, at the outset, I make three observations about the nature and significance of the Comparative Response. First, what would be accomplished if one were to show that although theism does not explain all evils well, overall it explains evil as well as (or better than) its metaphysical rivals? Minimally, showing this would undermine
2
C. STEPHEN LAYMAN
the assumption that the problem of evil is a problem for theists only. For example, if naturalism explains evil no better than does theism, then if evil is a problem for theism, evil is a problem for naturalism too. Furthermore, if theism explains evil as well as naturalism does, then the phenomenon of evil does not provide a reason for accepting naturalism over theism. Finally, if theism explains evil better than naturalism does, then the phenomenon of evil might actually provide a reason to accept theism over naturalism. Second, the Comparative Response is, I take it, best pursued by comparing theism to one major rival at a time. At any rate, trying to compare more than two major metaphysical rivals at once, in regard to their explanations of evil, is apt to generate unmanageable complexity. Moreover, the Comparative Response may be successful in regard to some of theism’s rivals, but not in regard to others. For example, it might be that theism explains evil as well as naturalism does, but not as well as a “Finite Deity” hypothesis, i.e., the hypothesis of a Deity who lacks sufficient power to overcome evil.4 Third, in developing the Comparative Response, one unavoidably makes controversial assumptions about a series of issues, e.g., the nature of explanations, how best to formulate theism and its rivals, the nature of good and evil, and so on. That the assumptions are controversial is not, I take it, a good reason to reject the Comparative Response. As Marilyn Adams remarks, the “defense of any well-formulated philosophical position will eventually involve premises which are fundamentally controversial and so unable to command the assent of all reasonable persons.”5 This being so, both theism and its rivals will unavoidably make use of controversial premises. Therefore, in the absence of a case to the effect that theism employs more controversial premises than its rivals do, the observation that theism employs controversial assumptions is without force. This observation applies to the present discussion as follows: If the fundamental premises of theism as regards moral evil are no more controversial than those of naturalism, then the fact that theism’s assumptions regarding moral evil are controversial is not a serious objection to the Comparative Response. I shall now explore the Comparative Response, comparing the theistic explanation of moral evil to that of naturalism.
I. Explanations When is a hypothesis a better explanation of some phenomenon than a rival hypothesis? I take it we must ask two main questions of any hypothesis, H:6 1. Does H lead us to expect the phenomenon in question? 2. How probable is H independently of the phenomenon in question?
MORAL EVIL: THE COMPARATIVE RESPONSE
3
Some brief comments will help to clarify (1) and (2). When does a hypothesis, H1, lead us to expect the phenomenon in question better than another hypothesis, H2? Roughly speaking, when, on the assumption that H1 is true, the phenomenon is less surprising than on the assumption that H2 is true. The word “surprising” is here meant to indicate that the probabilities in question are epistemic (as opposed, e.g., to being frequency probabilities). This means we are comparing the likelihood of the phenomenon on multiple hypotheses, but not in a situation in which we can determine the result mathematically. (Think of jurors estimating how likely the various facts of the case are given the hypothesis that the defendant is the culprit.) The probability of a hypothesis independent of the phenomenon under consideration is its prior probability. The prior probability of a hypothesis is at least in part determined by its degree of fit with any relevant background knowledge. The background knowledge consists of any information (other than the information given in descriptions of the phenomenon) that may reasonably be taken for granted.7 Some hold that the prior probability of a hypothesis is determined in part by its degree of simplicity. Others regard simplicity as merely of aesthetic or instrumental value. I shall assume the Principle of Simplicity, by which I mean this: other things being equal, the simpler hypothesis is more probably true. And for present purposes, I think we can identify four main facets of simplicity:8 1. The first facet of simplicity is just a matter of the number of things postulated. To borrow an example of Swinburne’s, Leverrier postulated one planet to explain perturbations in the orbit of Uranus, not two planets, or three, and so on. 2. The second facet of simplicity is the number of kinds of things postulated. For example, a theory that postulates three kinds of subatomic particles is simpler than a theory that postulates, say, eight kinds. 3. The third facet of simplicity may be called simplicity of terms, and concerns whether a term used to state a hypothesis can be understood only by someone who understands some other term. To illustrate (again, borrowing from Swinburne), “All emeralds are green” is simpler than “All emeralds are grue,” because one must understand “green” in order to understand “grue” but not vice versa.9 This facet of simplicity will lead us to select terms that stand for objects or properties more readily accessible to our cognitive faculties. And thus it will lead us to avoid a more theoretical term (referring to underlying theoretical entities) unless we need it to obtain a hypothesis that yields the data. 4. The fourth facet of simplicity is the number of theses within a hypothesis that receive little or no probabilistic support from other theses belonging
4
C. STEPHEN LAYMAN
to that hypothesis. This facet is grounded in the rules of probability, in particular, the general conjunction rule: P(A & B) = P(A) x P(B given A).10 Roughly speaking, this rule ensures that the more one says, the more one is likely to say something false. For example, suppose we somehow know that the probability of A’s being true is 4/5, while the probability that (B is true on the assumption that A is true) is 3/5. Then the probability of the conjunction (A & B) is only 12/25, less than half. Because simplicity has multiple facets, there are many cases in which we cannot say which hypothesis is simpler all things considered. H1 may be simpler as regards facet one, while H2 is simpler as regards facet two, and we may see no way to determine whether H1 is simpler than H2 (or vice versa), all things considered. But I think the history of science reveals that, other things being equal, scientists accept the simpler theory. Of course, “other things being equal” is a crucial qualification. We reject overly simple hypotheses precisely because they do not lead us to expect the phenomenon or because they fit poorly with the background knowledge. But when two hypotheses lead us to expect the phenomenon equally well and fit equally well with the background knowledge, the simpler one is more likely true, according to the principle of simplicity.11 To determine which of two hypotheses is the better, then, we have to compare them on two points: (a) how well they lead us to expect the phenomenon in question and (b) their prior probabilities. The latter will be assessed by (i) fit with the background knowledge and (ii) the relative simplicity of the hypotheses. The better hypothesis, in any given case, better satisfies these criteria taken as a whole, and there can be trade offs, e.g., H1 may have a slightly higher prior probability than H2, but H2 may do a much better job of leading us to expect the phenomenon, so that we reasonably accept H2. For present purposes I shall assume that any given hypothesis, H1, provides as good or better an explanation of some phenomenon than a rival hypothesis, H2, if the following conditions are met: H1 leads us to expect the phenomenon in question as well as (or better than) H2; and the prior probability of H1 is equal to or greater than the prior probability of H2. (I am not suggesting that these are the only conditions in which one hypothesis is as good or better than another.) Let me close this section on explanation with two brief comments. First, following Swinburne, I hold that there is an important distinction between personal and inanimate explanations. Inanimate explanations are given in terms of initial conditions and natural laws. Personal explanations are given in terms of the beliefs, purposes, and powers of personal agents, including the power to make choices. For example, the stove is turned on because I wanted to make some coffee, I had the purpose of making coffee, I believed
MORAL EVIL: THE COMPARATIVE RESPONSE
5
that turning on the stove was a means to this end, I had the power to turn the stove on, and I chose to turn it on. I shall assume that both inanimate and personal explanations can be good ones. Second, no theory or worldview can explain everything. We may explain A in terms of B, B in terms of C, C in terms of D, and so on, but in the end we get back to an object or objects having such and such features, and for this we will have no explanation. In short we have to acknowledge something as ultimate in our attempt to explain as much as we can possibly explain.
II. Theism Discussions of natural theology can easily falter when insufficient care is given to stating rival hypotheses. So, I shall now offer explicit or “official” formulations of theism and naturalism. By theism I mean the view that (1) there exists exactly one nonphysical entity who is (2) perfectly morally good, (3) almighty, and (4) exists either eternally or necessarily. Also, (5) every ultimate, complete explanation is personal. Some clarifying comments are in order: Exactly one entity. Theism is simpler than polytheism in that theism postulates a single God rather than many gods. However, in postulating a nonphysical entity with moral qualities, theism postulates one entity more than naturalism does, and that entity belongs to a kind not postulated by naturalism. So, with regard to the first and second facets of simplicity, theism is less simple than naturalism. Perfectly morally good. If God is morally good, then God is a personal entity. A personal entity is one who has beliefs, purposes, and the power to make choices. And if God is perfectly morally good, then God has the moral virtues of love, wisdom, and justice. If God is wise, then God knows what is important and how to achieve it. If God is loving, then God cares about the long-term best interests of any persons he creates. If God is just, then God will act in accord with certain moral principles, e.g., he will not fail to keep any promises he makes. Many theists have held that God is essentially perfectly morally good, i.e., that the individual who is God is perfectly good in every possible world in which he exists. From this perspective, it is logically impossible for the person who is God to perform a morally wrong action. Typically, these theists also hold that essential moral goodness is the best kind of moral goodness.
6
C. STEPHEN LAYMAN
Call this combination of claims “the essential goodness thesis.” For purposes of this paper, I shall not regard this thesis as part of the theistic hypothesis. Although a discussion of the essential goodness thesis is outside the scope of this paper, I will here state very briefly why I do not build the thesis into my formulation of theism. First, many theists have suggested that without the freedom to choose between good and evil, humans would be like robots, having a “programmed” (i.e., unavoidable) goodness, which would be inferior to a goodness acquired (or maintained) via free choice among alternatives. I share this intuition. An agent who conforms to moral requirements freely (when not conforming is a genuine alternative) has a higher sort of goodness than an agent who cannot avoid conforming to moral requirements. And an agent who has essential moral goodness cannot (logically) avoid conforming to moral requirements. Of course, an essentially good Deity might choose between various supererogatory acts, but its avoidance of wrong or evil acts would be “programmed.” Second, if God is essentially morally good and almighty, then God is presumably able to create finite agents who are essentially morally good. But humans are not essentially morally good; in fact many humans are morally corrupt and some are moral monsters. I think it is indeed difficult to understand why God would create beings that have gone as far wrong (morally speaking) as humans have, if (a) God can create essentially perfectly morally good agents and (b) this type of goodness is better than goodness acquired (or maintained) through a free choice between good and evil. In short, the essential goodness thesis seems to exacerbate the problem of moral evil.12 Almighty. An almighty entity is one that has maximal power, but I assume that maximal power has some limits, e.g., it does not include the power to do the logically impossible. I take it that a maximally powerful being must be extremely knowledgeable, for knowledge is a form of power. So, an almighty being will have a tremendous amount of knowledge, at least enough to be in charge of the entire universe; but an almighty being cannot know what it is logically impossible to know. For example, it may be logically impossible to know infallibly what some agent will choose to do freely in the future. Furthermore, an almighty being will be non-physical, since an entity counts as physical only if it is governed by the laws of nature. And an almighty entity is not governed by the laws of nature. Eternal or necessary. God exists eternally or necessarily. God is eternal if God is either everlasting or timeless. God is everlasting if God existed at each past moment, exists now, and will exist at each future moment. God is timeless if God is outside of time, i.e., if there is no sequence, no “before
MORAL EVIL: THE COMPARATIVE RESPONSE
7
and after,” in the divine life. It is of course a matter of debate among theists whether God is everlasting or timeless, but for present purposes we need not be detained by this “family quarrel.” However, if God is neither everlasting nor timeless, then there must be situations (past, present, or future) not under God’s control. And then the theist would need additional, complicating metaphysical hypotheses to provide a complete explanation for the history of reality. It is also a matter of debate among theists whether God exists necessarily, in every possible world. I will avoid this debate simply by stipulating that I shall not assume that if God exists, then God exists of necessity. Every ultimate and complete explanation is personal. This is so, according to the theist, because the factors involved in inanimate explanation are due to God’s creative activity: God creates the physical universe and it is due to God that physical reality behaves in those patterned ways we refer to as natural laws. Thus, the factors involved in inanimate explanations are themselves susceptible to personal explanation, given theism.
III. Naturalism In the present philosophical climate, I take naturalism to be the primary metaphysical alternative to theism. By naturalism I mean the view that (1) there exists a physical reality, (2) it is self-organizing, i.e., it has an inherent structure as opposed to a structure imposed by a god or some other agent, principle, or force, (3) it exists either necessarily, eternally, or by chance, and (4) all property bearers are physical (leaving aside properties, sets, and numbers). Also, (5) every ultimate, complete explanation is inanimate. Again, some clarifying comments are in order: There exists a physical reality. What is meant by the word “physical”? It seems that the meaning of this word has gotten stretched over the history of science. As Nagel remarks, New properties are counted as physical if they are discovered by explanatory inference from those already in the class. This repeated process starts from a base of familiar, observable spatio-temporal phenomena and proceeds to take in mass, force, kinetic energy, charge, valence, gravitational and electromagnetic fields, quantum states, anti-particles, strangeness, charm, and whatever physics will bring us next.13
8
C. STEPHEN LAYMAN
I am not aware of any uncontroversial characterization of the physical. If we take the above quotation, not as a description of what has been counted as physical over the history of science but as a recommendation regarding what should be counted as physical, then it seems to have the counterintuitive implication that a successful argument to the best explanation in natural theology yields the further result that divinity is a physical property. For present purposes I shall employ the following, partial characterization of the physical: an entity is physical only if either (a) it is simple (non-composite) and has only mechanistic propensities or (b) it is composed entirely of entities with only mechanistic propensities. Mechanistic propensities may be either deterministic or probabilistic, but they are non-teleological in nature, i.e, they do not involve the power to act with a purpose or to make choices.14 Physical reality is self-organizing. This thesis is the heart of the naturalistic hypothesis. Physical things have propensities, i.e., they tend to act in certain ways. (For example, half of any given bit of Uranium-235 will decay into Lead-207 over a period of 704 million years.) And according to the naturalist, these propensities are not given them by a Creator or imposed from without; it is the nature of physical reality to be self-organizing. Physical reality exists either necessarily, eternally, or by chance. (a) Chance. The suggestion that physical reality in its entirety popped into existence uncaused is most implausible. And although some current cosmologists hypothesize that the big bang resulted from a random fluctuation in a preexisting super-space, such a preexisting super-space liable to fluctuations is plainly not nothing, and if it exists, it belongs to what I call “physical reality.” (b) Eternal. If physical reality is not eternal in some form (such as energy or a preexisting super-space) but had a radical beginning out of absolutely nothing, then its beginning to exist surely cries out for an explanation. This is the intuition behind certain forms of the cosmological argument and it places strong pressure on the naturalist to postulate that physical reality (in some form or through various phases) is eternal. (c) Necessary. If physical reality is necessary, it’s eternal, but if the naturalist claims that physical reality is necessary, he owes us an argument. Necessity is not an easy thing to establish outside of cases involving formal logic, math, or conceptual connections. So I shall assume that physical reality does not exist of necessity. All property bearers are physical (leaving aside properties, sets, and numbers). I take it that this exclusionary thesis – or something similar to it – is essential to naturalism. There are no nonphysical agents, e.g., souls or gods, according to the naturalist. If nonphysical entities such as sets or numbers are admitted into the naturalist’s ontology, it is with the understanding that
MORAL EVIL: THE COMPARATIVE RESPONSE
9
these entities cannot cause events.15 Also, since properties themselves can have properties (e.g., being reflexive, being mental), and since some naturalists hold that mental properties are not physical properties (though mental properties are properties of the brain), an exception in the case of properties is important to some naturalists. Every ultimate, complete explanation is inanimate. As I conceive naturalism, it regards personal explanations as non-ultimate. We explain that the water in the pot is boiling because Fred had the purpose of making coffee, believed boiling water was a means to this end, etc. But we can explain the very existence of Fred by way of evolutionary theory, and evolution is a process governed by natural laws. So, the presence of persons is to be explained by way of a more ultimate, inanimate explanation. Moreover, as I conceive naturalism, human behavior is ultimately to be explained in terms of natural laws and initial conditions. The various factors included in any given personal explanation are present and have whatever effects they have because of certain relevant natural laws (and initial conditions). We could in principle do without personal explanations for there is always a deeper, inanimate explanation of why people behave as they do. Now, it is possible for a naturalist to take the view that the factors involved in personal explanations cannot be explained in terms of inanimate explanations. We shall briefly consider this possibility in section VI, but for the time being I shall assume that naturalism includes thesis (5). Having spelled out the theistic and naturalistic hypotheses in some detail, it seems to me clear that naturalism is the simpler hypothesis. In particular, naturalism is simpler in terms of the first two facets of simplicity: fewer entities postulated and fewer kinds of entities postulated. Aside from these two facets, the two hypotheses seem to me roughly equal in simplicity, but because naturalism is clearly simpler in respect of these facets, it is the simpler hypothesis overall. Hence, as regards the attempt to explain evil, naturalism has a higher prior probability than theism unless theism fits the background information better than naturalism does. I shall let questions about background information arise at appropriate points in the discussion, but for the present, I shall leave them aside. And this means that, at the outset, I regard naturalism as having a distinct epistemological advantage over theism in regard to explaining the presence of evil.
IV. Evil For the sake of clarity, it will be helpful to have a brief description of the phenomenon of evil. By “evil” I mean badness in general, for example,
10
C. STEPHEN LAYMAN
moral wrongdoing, suffering, loss, insanity, and being under systematic illusion about important matters. For present purposes, it is useful to break the phenomenon of evil down into subcategories, as follows: The Phenomenon of Evil includes: (A) instances of human wrongdoing; (B) instances of suffering and loss that are due to human wrongdoing; (C) great wickedness and the total amount of suffering that is due to human wrongdoing; and (D) all instances of suffering and loss that are due to non-human causes. (A), (B), and (C) are categories of moral evil, i.e., wrongdoing for which humans are responsible and the suffering (and loss) that results from it. (D) encapsulates what is commonly called “natural evil.” For present purposes, we are discussing only moral evil, and leaving natural evil aside.
V. To what extent can theism explain moral evil? We should not be surprised to find moral evil in the world on the assumption that theism is true. Freely chosen love of God and neighbor is surely a much higher good than any automatic or robotic sort of love. Moreover, from the standpoint of theism, if the human responses to God and to God’s goodness are not free but “programmed,” not genuinely up to human agents, then those responses seem empty of meaning. Thus, a perfectly morally good Deity would have good reason to create beings possessing a capacity to choose between love and its alternatives, between good and evil. Moreover, there can be no significant choice between good and evil unless creatures have reasons and/or desires to do both good and evil. So, we should expect God to create agents having reasons and/or desires to do things that are evil – at least in the sense of desiring good things even when securing them will be at the expense of others. For example, it is good to eat for nourishment and enjoyment, but the desire to eat may still be present even when others need the food more and should have it. Or again, controlling others is good when it takes the form of appropriate leadership, but the desire to control can be present even when it does not serve the common good. Such motivational structures are not surprising if a loving God exists. For in a situation in which free creatures have a significant choice between good and evil, they must have reasons and/or desires to do wrong, and hence moral wrongdoing is not surprising. Thus, phenomenon A is not surprising, given theism. Now, logically speaking, we can have instances of wrongdoing without any resulting suffering. For example, if I selfishly hog the scarce food, God can create more food to prevent my selfishness from adversely affecting
MORAL EVIL: THE COMPARATIVE RESPONSE
11
others. However, there are problems with this type of inconsequential wrongdoing. The significance of a choice is surely linked to its consequences or expected consequences. If we can never benefit or harm others (or reasonably expect to do so), then we haven’t been given a significant degree of freedom or responsibility. Furthermore, if our actions never caused harm and suffering, we would surely fail to understand the seriousness of evil. Now, it might be replied that my choices will be significant provided I believe I can benefit or harm, even if the belief is false. And I will surely be apt to see the evil I do as egregious if I believe it causes others to suffer, even if it really doesn’t cause any suffering at all. However, if we believe we can benefit and harm others, when in fact we cannot, then we are systematically deceived about something extremely important and fundamental to our lives, and such massive deception would itself be an evil; hence it is plausible to suppose that a perfectly good God would not set up a world involving such deception. So, it seems to me that phenomenon B is not surprising given theism. How about phenomenon C – great wickedness and the total amount of suffering resulting from human wrongdoing? Once again, the significance of our choices is linked to the suffering and harm we can inflict. So, if a significant choice between good and evil is itself a good thing, as I take it to be, then I think some great wickedness (with correspondingly great suffering inflicted) is not surprising if there is a perfectly good God. However, the total amount of wickedness, together with the correlative suffering, requires special comment, especially because it seems plausible to suppose that, whatever the divine purposes are, they might surely have been achieved with less wrongdoing than is actually present. But this plausible supposition does not bear up well under scrutiny. First, I do not think it wise to assume that each wicked act is allowed as a means to some greater good. Instances of wickedness are better regarded as “collateral effects” or consequences of divine creative activity rather than necessary means to divine ends. That is, in actualizing a situation involving agents with significant free choice, God actualizes a situation in which wicked acts are an unsurprising result. There is no reason to suppose either that God wants any specific evil act to occur or that God plans for some greater good to arise from each such act. At any rate, I see no reason to make such assumptions. Wicked acts, taken collectively, are simply the unsurprising result of creating agents with significant free choice; and such acts, taken individually, need not be the necessary means to any greater good. (Taken collectively, the epistemological unsurprisingness of such acts is, I believe, a necessary condition of the presence of agents with significant free choice; and so of whatever purposes God has in mind for such free agents.)
12
C. STEPHEN LAYMAN
Second, the meaning or significance of human life is in large measure a function of the significance of the choices humans make. And if a loving God exists, he will certainly place us in a situation in which we can achieve lives that are rich in meaning and significance. For this reason, if there is a loving God, we ought to expect there to be many opportunities to make highly significant choices between good and evil. But the significance of the choices is in large measure a function of the benefit or harm the acts are apt to bring about (or prevent). Therefore, if a loving God exists, we ought to expect there to be plenty of opportunities to make choices between good acts that will provide great benefits (or prevent great harms), and evil acts that will inflict great harms (or obviate great benefits). It follows that if a loving God exists, we ought to expect there to be many opportunities to freely perform very wicked acts. Third, the suggestion that we could have as much meaning or significance in our lives as we actually do without certain types of wickedness, such as genocide, is problematic. If there were no genocide, no doubt some other example of wickedness would be under discussion, murder or rape, perhaps. And if murder and rape were beyond human capability, we would be discussing still other forms of wickedness. Some humans would probably be satisfied with a situation in which the consequences of our acts would be trivial; but such a world would pale in significance to the world we find ourselves in – a world full of drama, in which acts have enormous significance – significance that is partly a function of their consequences. A world with less significance has its attractions, because such a world contains fewer risks. But I believe it remains unsurprising that we find ourselves in a situation in which acts have enormous significance – assuming that God exists. Such a world has many risks, but it is also replete with meaning. What we do – and do not do – really counts for something. When we wish for this or that type of wicked act to be removed from the human repertoire, it seems to me that we are wishing for a world in which human action has less significance than it does in the world we have got.16 Now, at this point we have in effect added a conjunct to our official theistic hypothesis. Let us call it the free will thesis: T6. God has created human beings with significant freedom to choose between good and evil. In adding this thesis to theism, we complicate theism significantly, and so we undoubtedly lower the prior probability of theism somewhat, for reasons noted earlier. However, as we have just seen, the free will thesis itself is not surprising given the other claims involved in the theistic hypothesis. Therefore, I see no reason to suppose that adding T6 to theism will produce a dramatic drop in the prior probability of theism.
MORAL EVIL: THE COMPARATIVE RESPONSE
13
I do not think, however, that the theist can stop with T6. For while it is not surprising, given theism, that acts have great significance arising (in part) from their consequences (including great harm), we have to ask whether divine love can rest content with the destructive results of wickedness. An innocent child’s life is ended by violence or a victim of torture goes permanently insane – can divine love simply regard such wickedness as the price of significance – “end of story”? Surely not. An almighty and perfectly good being will not allow the effects of such wickedness to be the last word. Accordingly, the theist will add a seventh thesis, along these lines: T7. If any purposes that an almighty and perfectly good being has for its creatures have not been achieved prior to their death, then such purposes will be achieved – to the extent possible – after the death of the creatures. Let me comment briefly on T7. I assume that an almighty being is able to raise creatures from the dead; thus, life after death is possible, given theism. Moreover, life after death is probable, given theism, precisely because divine love will not give wickedness the last word. A loving God seeks the fulfillment of His creatures. The only thing that can prevent the fulfillment of an individual creature in the long run, I take it, is the individual’s free rejection of divine love. Therefore, T7 provides an assurance that those whose lives are wrecked by wickedness in this earthly life will have opportunities to reclaim the fulfillment denied them. Beyond this, T7 provides an assurance that all whose lives are marred by wickedness (i.e., everyone) will have an opportunity for complete fulfillment in a life to come, for surely perfect love could seek no less.17 Note that T7 is conditional in nature and as such it is very plausible independent of the claim that God exists: Assuming there is an almighty and perfectly good Being, it would certainly pursue its purposes to the limits of possibility, and a being with the power to create persons can surely secure their existence in a life after death. So, while T7 may complicate the theistic hypothesis slightly, its inherent plausibility allows it to be added to theism at little cost to theism’s prior probability. Terminological note. From here on, when I wish to speak of the original theistic hypothesis, with its five theses, I shall use the expression “theism (1–5).” I shall use the expression “theism (1–6)” to indicate a version of theism that includes theses 1 through 6, and “theism (1–7)” to indicate a version of theism that includes theses 1 through 7. Similarly, I shall have occasion to use expressions such as “naturalism (1–7)” to indicate alternative versions of naturalism.
14
C. STEPHEN LAYMAN
VI. To what extent can naturalism explain moral evil? A. Moral evil and life How surprising is moral evil given naturalism? First, naturalism explains no evil at all unless it explains the presence of life. How surprising is life given our naturalist hypothesis? We know that there can be no life unless the universe is “fine-tuned” to support it, i.e., had certain very fundamental features of the physical universe been only slightly different, there would be no life at all. For example, if the strong nuclear force had been stronger or weaker by as little as five percent, life would be impossible. And if the initial explosion of the big bang had differed in strength by as little as one part in 1060 , the universe would either have collapsed back on itself or expanded too rapidly for stars to form (and either way, life would be impossible). Also, if gravity were stronger or weaker by one part in 1040 , stars could not exist (hence, there would be no Sun). Over twenty such physical parameters must have values that fall within highly restricted ranges in order for life of any conceivable sort to be present.18 Does naturalism lead us to expect a fine-tuned universe? It seems to me that it does not. A self-organizing physical reality could take a multitude of forms, as far as we know. To illustrate, consider the force of gravity, which can be expressed as follows: F = Gm1 m2 divided by r2 . That is, the gravitational force F between two objects is equal to [(the gravitational constant G times the mass of the first object times the mass of the second object) divided by the distance between the two objects squared]. Now, to the best of our knowledge, the gravitational constant, G, might have been different from what it is. It might have been twice as strong as it is or only half as strong, for example. If we imagine ourselves able to set the value of G by a sort of dial having 1040 evenly spaced marks, then if the setting is off by even one mark, we get no stars, no Sun, and hence, no life.19 Similar observations could be made about the strong nuclear force, the electromagnetic force, the force of the initial big bang explosion, and so on. For each possible variation of these basic factors, there is an alternative possible physical universe, but the vast majority of the variations would not lead to a universe containing life. Accordingly, given merely the postulate of a self-organizing physical reality, we have no good reason to expect a fine-tuned universe. To explain the phenomenon of fine-tuning, naturalists will need to supplement their hypothesis. Often they adopt some form of the so-called “manyuniverses hypothesis” (MU, for short). According to MU, there are many distinct physical universes, millions of them, perhaps even infinitely many. (What makes them distinct universes is that the fundamental parameters of physics, such as the value of the gravitational constant G, vary significantly
MORAL EVIL: THE COMPARATIVE RESPONSE
15
from one to the next.) In one version of MU, the oscillating big-bang version, each universe begins with a big bang and collapses (sooner or later) back upon itself – a “big crunch.” There have already been infinitely many such bigbang/big-crunch cycles; therefore, our fine-tuned universe is not surprising after all. In fact, while most of the universes in the series have undoubtedly been without life, such a universe as ours was likely to turn up sooner or later. In another version of MU, the vacuum fluctuation version, the multitude of distinct physical universes need not arise in serial fashion: each universe is generated by a random fluctuation in a preexisting super-space. But again, the basic idea is that if there are millions or even infinitely many physical universes, we should expect one like ours to be among them. In adding MU to naturalism we greatly complicate it. For we are now saying not merely that physical reality is self-organized, but that this selforganization takes the form of an underlying mechanism that generates universes at random. Moreover, the number of universes generated at random is very large – millions, or perhaps even infinite. Thus, in adding MU to naturalism, we in effect add two crucial claims, resulting in the following modified formulation: Naturalism (second formulation): the view that (1) there exists a physical reality which is (2) self-organizing (i.e., it has an inherent structure as opposed to a structure imposed by a god or some other agent, principle, or force). (3) This self-organization includes a mechanism that generates universes at random and (4) the number of universes thus generated is very great – there are millions of them, perhaps even infinitely many. (5) Physical reality exists either necessarily, eternally, or by chance. (6) All property bearers are physical (leaving aside properties, sets, and numbers). (7) Every ultimate, complete explanation is inanimate. Hereafter I shall refer to this second formulation of naturalism as naturalism (1–7). In postulating millions (or even infinitely many) universes, the naturalist surely loses any advantage connected with the first facet of simplicity (i.e., number of objects postulated). In fact, as regards the first facet of simplicity, theism appears simpler than is naturalism (1–7). Moreover, the many universes postulated by the naturalist fall under an important new kind, namely, they are unobservable and not subject to empirical tests – for there is no way to observe these other universes or to run tests to verify their existence. This removes the naturalist’s original advantage in connection with the second facet of simplicity. So, by adding MU, the naturalist removes the grounds for claiming that naturalism is simpler than theism (1–5). Furthermore, in adopting MU, the naturalist adds two important theses to his overall theory – (3) and (4) above – thus complicating naturalism in regard to the
16
C. STEPHEN LAYMAN
fourth facet of simplicity. Taken together, (3) and (4) strike me as at least as substantial a claim as T6. If that’s right, then the prior probability of naturalism (1–7) is no higher than theism (1–6). But there’s another key factor to consider: Whereas we are led to expect T6 by the other theses composing the theistic hypothesis, we are not led to expect (3) and (4) by the other theses composing the naturalist hypothesis. Nor are (3) and (4) plausible in their own right. So, if we add (3) and (4) to the naturalistic hypothesis, I believe that its prior probability falls below that of theism (1–6). And since T7 is plausible in its own right, I see no good reason to assign theism (1–7) a lower prior probability than naturalism (1–7). Some naturalists may wish to avoid MU; they may instead wish to add the following supplements to naturalism (1–5), namely, that the self-organization of physical reality can take only one form and that form is precisely the form we find in our universe. Let us call these two theses (3*) and (4*) respectively, to distinguish them from (3) and (4) above. I do not see that adding (3*) and (4*) will be advantageous for the naturalist, however. First, if we add (3*) and (4*), we complicate naturalism with additional hypotheses that we are not led to expect given naturalism (1–5). Second, we in effect build the phenomenon of fine-tuning into naturalism, since in context the “one form” our universe can take has to involve the phenomenon of fine-tuning – otherwise adding (3*) and (4*) will not be useful. But then, adding (3*) and (4*) is less an explanation of fine-tuning than an incorporation of fine-tuning into the naturalistic hypothesis. And hence, if we add (3*) and (4*), we import all the complexity of the phenomena of fine-tuning into naturalism. (Recall that more than twenty basic physical parameters must have values that fall within highly restricted ranges in order for life to be present.) All in all, MU seems a more sensible supplement to me. But in any case, adding (3*) and (4*) isn’t better than adding MU. So, I shall assume that naturalism (1–7) provides as good an explanation of fine-tuning as any naturalism can offer. It is important to note that fine-tuning is not surprising given theism (1–5). In fact, many have supposed that a fine-tuning design argument provides significant probabilistic support for theism. But leaving arguments for theism aside, it is plausible to suppose that life (i.e., physically-based or biological life) is a good thing, and hence that God would have reason to create it. By a “good thing” I here mean one that merits responses such as wonder, admiration, and delight. It is even more plausible to suppose that intelligent life is a good thing, and extremely plausible to suppose that intelligent life capable of giving and receiving love is a good thing. So, if God exists, God has good reason to create life. But to the best of our knowledge, life requires a fine-tuned universe; hence, such a universe is not surprising, given theism.
MORAL EVIL: THE COMPARATIVE RESPONSE
17
(Incidentally, God may also have good reason to create unembodied persons, i.e., angels, but this in no way cancels the reasons for creating physical life.) To sum up, without MU, naturalism does not lead us to expect life. And if naturalism doesn’t lead us to expect life, then it leads us to expect no evil at all. But in adding MU, naturalism loses its claim to a higher prior probability than theism (1–7). B. The metaphysics of good and evil Even if naturalism leads us to expect life, it does not lead us to expect moral evil unless it provides a metaphysics of good and evil. How is it, on naturalistic principles, that certain acts are right or wrong, just or unjust? How is it that certain traits of character are virtuous and others vicious? If the naturalist responds by taking an anti-realist position, such as emotivism, most people, theist and non-theists alike, will not regard the explanation as adequate. For on such views, there are no moral facts, just human emotional responses, recommendations for action, and so on. And from the standpoint of moral realism, moral anti-realists do not explain any moral evil, they explain it away. There is, then, serious doubt whether naturalism explains any moral evil at all if naturalism is combined with moral anti-realism. Nevertheless, moral antirealism is common enough among naturalists, and indeed naturalism is often given as a reason for accepting moral anti-realism.20 But naturalists need not be moral anti-realists. To speak in general terms, they have at least two other options. They may either opt for a form of Platonism (i.e., certain moral truths obtain independently of the existence of physical reality) or opt for a form of moral supervenience (i.e., all moral truths supervene on sufficiently intelligent conscious beings, e.g., humans).21 I am not here concerned with the details of such views. What I do wish to emphasize is that a non-trivial thesis regarding the metaphysics of morality must be added to naturalism; otherwise naturalism will not lead us to expect the phenomenon of moral evil. Let us call this thesis the “Metaphysics of Morality” thesis, and we can state it as follows: N8. As regards the metaphysics of morality, some form of Platonism, supervenience, or anti-realism holds. As far as I can tell, neither Platonism nor supervenience receives any support from the other theses (1–7) of the naturalistic hypothesis. And it is a matter of controversy, even among naturalists, whether anti-realism receives support from the other theses of naturalism (1–7). Furthermore, N8 is by no means obvious. Hence, owing to the fourth facet of simplicity, to add N8 to naturalism is to complicate naturalism significantly. But I have already argued that naturalism (1–7) should not be assigned a higher prior probability than theism
18
C. STEPHEN LAYMAN
(1–7). So, if we add N8 to naturalism, then we ought to assign naturalism (1–8) a significantly lower prior probability than theism (1–7). Note that theism is in a radically different position than naturalism with respect to the metaphysics of morality. First, theism includes the thesis that God is perfectly morally good (hence, wise, loving, and just). Second, theism (1–7) includes T6, i.e., God has created human beings with significant freedom to choose between good and evil. Thus, theism (1–7) ensures the presence of whatever metaphysical factors are needed to make a choice between good and evil possible, while naturalism (1–7) does not ensure the presence of such factors. Again, the upshot is that if we add N8 to naturalism, we should assign naturalism (1–8) a significantly lower prior probability than theism (1–7). C. Free will and moral responsibility There is yet another barrier to the naturalist’s attempts to explain moral evil; for in order to explain the presence of moral evil, the naturalist must account for the presence of moral responsibility. But a plausible argument supports the conclusion that naturalism does not lead us to expect that humans are morally responsible for their actions. All ultimate explanations are inanimate ones, according to the naturalist. Thus, the ultimate explanation for the existence and behavior of beings having moral responsibility must be given solely in terms of initial conditions and natural laws. For the naturalist, such laws are, in the final analysis, grounded in the mechanistic propensities of the fundamental physical entities. Now, natural laws hold no matter what any human being does. For example, one can jump out of a window, but that won’t alter the law of gravity, only illustrate it. I emphasize that natural laws are beyond human control whether they are deterministic or statistical. Deterministic laws are of the form, “Given p, q” where p is a proposition summarizing the initial conditions and q describes the result. Given the initial conditions, only one result can occur (in a closed system), if a deterministic law applies. For present purposes, we can regard statistical laws as having this form: “If p, then q has probability n and r has probability 1 – n.” To illustrate, suppose there is a statistical law that connects brain states B1, B2, and B3 as follows: “If B1, then the probability of B2 is 0.75 and the probability of B3 is 0.25.” No matter what I do, this law holds. No matter what I do, the probabilities remain as indicated. And suppose that, given B2, I’ll tell a lie on a certain occasion but given B3, I won’t tell the lie. There are two possible futures for me, but do I have any control over which of them becomes actual? It seems to me that the answer is no, given naturalism, since on this view, human behavior is fully explicable in terms of natural laws. We may speak of people forming intentions and acting accordingly,
MORAL EVIL: THE COMPARATIVE RESPONSE
19
but this is at best short hand for the underlying (and correct) natural-law explanation. But if the human lack of control over natural laws is a problem for naturalism, isn’t it a problem for any science-friendly metaphysical view? I think not. In general, a view that allows for basic causal powers of a non-mechanistic sort, such as agent causation, has more room to maneuver. By agent causation I mean the view that persons can bring about certain outcomes simply by choosing to do so. (Presumably, one will have a reason for choosing, but there may be equally good reasons for choosing an alternative.) And the power of choice is here regarded as an inherent power of particular persons (rather as mechanistic propensities may be regarded as inherent powers of particular physical particles). With such an additional form of causation in the picture, a person’s act can be up to her (or under her control) in a significant sense, and this need imply no violation of natural laws, under two understandings. First, if the natural laws are statistical, outcomes can occasionally be influenced without changing the relevant probabilities, and hence without creating exceptions to the laws. The laws are not violated as long as the frequency of outcomes conforms to the relevant probabilities over the long haul. Second, natural laws may simply be regarded as having an implicit rider along these lines: if there is no causal input from a non-mechanistic source, given p, q. From this perspective natural laws tell us how things will go when only mechanistic factors operate. But if agent causation will enable the naturalist to avoid the problems I’ve been pressing regarding moral responsibility, why can’t the naturalist simply incorporate agent causation into her view? First of all, I do not think that most naturalists want to adopt agent causation. As I see it, the preference for explanations in terms of natural laws runs very deep in most naturalists. So, I believe that the version of naturalism described in this paper, which does not leave room for agent causation, is a standard version. Second, adding agent causation would lower the prior probability of naturalism considerably, for two reasons. (1) If the naturalist adopts agent causation, then the naturalist postulates a kind of explanatory dualism. Ultimate explanations will then be of two radically different kinds, neither explicable in terms of the other. To add this explanatory dualism to naturalism is to build into it a very strong element of mystery. (2) Do the other theses of naturalism lead us to expect agent causation? I see no way to argue for a “yes” answer to this question. The answer is either inscrutable or negative; and either way, adding agent causation to naturalism lowers its prior probability. But by my lights, naturalism (1–8) already has a lower prior probability than theism (1–7). So, complicating naturalism by adding an agent-causation thesis will place naturalism’s prior probability quite significantly below that of theism.
20
C. STEPHEN LAYMAN
By contrast, given theism, all ultimate explanations are personal. And talk about ultimate personal explanations is simply another way of talking about agent causation. If God exists, God is an agent cause. And it is hardly surprising that God should create other agent causes if that’s what it takes to create beings that are morally responsible.22 To sum up, in order to explain the phenomenon of moral evil, both theists and naturalists are forced to complicate their hypotheses. The elaborated hypotheses are as follows: Theism: (1) there exists exactly one nonphysical entity who is (2) perfectly morally good, (3) almighty, and (4) exists either eternally or necessarily. (5) Every ultimate, complete explanation is personal. (6) God has created human beings with significant freedom to choose between good and evil. (7) If any purposes that an almighty and perfectly good being has for its creatures have not been achieved prior to their death, then such purposes will be achieved – to the extent possible – after the death of the creatures. Naturalism: (1) there exists a physical reality which is (2) selforganizing (i.e., it has an inherent structure as opposed to a structure imposed by a god or some other agent, principle, or force). (3) This self-organization includes a mechanism that generates universes at random and (4) the number of universes thus generated is very great – there are millions of them, perhaps even infinitely many. (5) Physical reality exists either necessarily, eternally, or by chance. (6) All property bearers are physical (leaving aside properties, sets, and numbers). (7) Every ultimate, complete explanation is inanimate. (8) As regards the metaphysics of morality, some form of Platonism, supervenience, or anti-realism holds. I have argued that theism (1–7) has a higher prior probability than naturalism (1–8). And because naturalism falters in explaining moral responsibility (and hence in explaining moral evil), I conclude that theism (1–7) explains moral evil as well as (or better than) naturalism (1–8). Therefore, if my arguments to this point are on the right track, the presence of moral evil in the world does not provide a good reason to accept naturalism over theism. Indeed, it may be that moral evil presents a greater problem to naturalism than it does to theism.23,24
MORAL EVIL: THE COMPARATIVE RESPONSE
21
Notes 1. For example, Richard Swinburne, Providence and the Problem of Evil (Oxford: The Clarendon Press, 1998). 2. For a series of fine essays for and against skeptical theism, see Daniel Howard-Snyder, ed., The Evidential Argument from Evil (Bloomington and Indianapolis: Indiana University Press, 1996). 3. See Alvin Plantinga, Warranted Christian Belief (New York: Oxford University Press, 2000), pp. 473–481. 4. Of course, even if a Finite Deity hypothesis explains evil better than theism, “Finite Deism” is apt to run into special difficulties of its own, e.g., since a Finite Deity presumably cannot account for the existence of the physical universe, one must add complicating supplements to Finite Deism to account for the presence of the physical universe. And such complicating supplements are apt to decrease the prior probability of Finite Deism. 5. Marilyn M. Adams, Horrendous Evils and the Goodness of God (Ithaca: Cornell University Press, 1999), p. 180. 6. My remarks on explanation are heavily indebted to Richard Swinburne, Simplicity As Evidence of Truth (Marquette University Press, 1997) and Is There a God? (Oxford University Press, 1996), p. 26. However, I have modified this material to suit my own purposes, so I must take responsibility for any errors in the presentation. 7. The relationship between a hypothesis H, some specific evidence for the hypothesis E, and the background knowledge K, can be expressed formally using Bayes’ theorem, which is derived from the elementary rules of probability: P[H/(E & K)] =
8.
9.
10.
11.
P(H/K) × P[E/(H & K)] {P(H/K) × P[E/(H & K)]} + {P(∼H/K) × P[E/(∼H & K)]}
P(H/K) = the probability of the hypothesis given the background knowledge = the prior probability of H (i.e., the probability of H independent of the evidence E). Here I am again borrowing heavily from Swinburne, Simplicity As Evidence of Truth, op. cit., pp. 23–27. But I have altered this material for my own purposes (especially as regards the fourth facet of simplicity), and I have also omitted (as irrelevant for present purposes) some of Swinburne’s facets of simplicity, e.g., those having to do with the simplicity of any mathematics involved in a hypothesis and with the simplicity of scientific laws. The term “grue” is borrowed from Nelson Goodman, who defined “grue” as “examined before future time t and found to be green or not so examined and blue.” All emeralds so far observed have been both green and grue. Nelson Goodman, Fact, Fiction, and Forecast (Cambridge: Harvard University Press, 1955; reprint, Indianapolis: Bobbs-Merrill, 1973), p. 74 in the reprint. That is, the probability of a conjunction (A & B) equals the probability that conjunct A is true times the probability that (conjunct B is true on the assumption that A is true). For present purposes, I regard entailment as the highest degree of probabilistic support possible, so that if A entails B, the probability of (B on the assumption that A is true) = 1. The intuitive plausibility of the Principle of Simplicity may be brought out by the following sort of case – again borrowing from Swinburne, Simplicity, op. cit., p. 45. Suppose we are plotting the course of an asteroid. And let’s suppose (however unrealistically) that the data falls into the following pattern, with x being the location of the asteroid at one time and y being the asteroid’s location at a later time. x = 0 y = 0
1 2
2 4
3 6
4 8
22
C. STEPHEN LAYMAN
Both of the following hypotheses lead us to expect our data: Hypothesis 1: y = 2x Hypothesis 2: y = 2x + 3x(x – 1) (x – 2) (x – 3) (x – 4)
12.
13. 14.
15.
16.
H1 and H2 each yield the observations made so far, and H1 is clearly simpler than H2. Let us assume H1 and H2 fit equally well with the background information. Note that for x = 5, H1 and H2 diverge in their predictions. For x = 5, H1 tells us y = 10 while H2 tells us y = 370. Let us suppose that this means H1 predicts the asteroid will strike Cleveland, Ohio, while H2 predicts the asteroid will strike Columbus, Ohio. (Assume the asteroid will inflict great damage on a city, but not on a wider area.) As time is very limited and emergency medical supplies are limited, we must focus our logistical efforts on either one of the cities, but not both. Isn’t it very plausible to suppose that we should focus our efforts on Cleveland, because that’s where the asteroid will probably strike? The reasoning here is unavoidably oversimplified in the interests of brevity, and for more rigorous argumentation I must refer my readers to the following authors, to whom I am heavily indebted: Theodore Guleserian, “Divine Freedom and the Problem of Evil” Faith and Philosophy 17(3) (July, 2000), pp. 348–366; Wesley Morriston, “Is God ‘Significantly Free’?” Faith and Philosophy 2(3) (July 1985), pp. 257–264; and Robert F. Brown, “God’s Ability to Will Moral Evil”, Faith and Philosophy 8(1) (January 1991), pp. 3– 20. For a defense of the essential goodness thesis, see Thomas V. Morris, Our Idea of God (Downers Grove, Ill: InterVarsity Press, 1991), pp. 47–64. Defenders of the essential goodness thesis may argue that even if God could create essentially good creatures, it is morally permissible for God to create beings possessing an inferior type of goodness, namely, contingent goodness. (Compare: Even if chimps are more wonderful animals than frogs, it surely wouldn’t be wrong for God to create frogs but not chimps.) I have serious doubts about this line of reasoning, but even if it works, more fundamental issues remain: Is a logically necessitated goodness really superior to the type of goodness that involves a free choice between good and evil? Is an agent who cannot choose to do wrong properly regarded as having moral goodness at all? I favor a view which might naturally be called Theistic Platonism, involving these elements: (a) No entity counts as divine (i.e., worthy of worship) unless it is perfectly morally good; (b) there are some necessary truths about right and wrong, good and evil (e.g., that love is a virtue); (c) the individual who is God is contingently perfectly morally good; and (d) were the individual who is God to do wrong, that individual would no longer be divine. For a defense of the thesis that God is contingently morally good, see Timothy P. Jackson, “Is God Just?” Faith and Philosophy 12(3) (July, 1995), pp. 393–408. Thomas Nagel, Mortal Questions (London: Cambridge University Press, 1979), p. 183. This characterization of the physical is inspired by a discussion of physical causation in William Hasker, The Emergent Self (Ithaca: Cornell University Press, 1999), pp. 62–63. However, I have modified this material for my own purposes, so I must take responsibility for any deficiencies in it. I am aware that Peter van Inwagen, Material Beings (Ithaca: Cornell University Press, 1990) has argued that the only composite material objects are living organisms. So far as I can see, nothing in my argument turns on a further specification of the physical than the one I’ve provided. There are various reasons a naturalist might wish to admit abstract entities into his ontology. For example, the statement, “Five is an odd number” seems to involve a reference to the number 5 that cannot be eliminated. And it is implausible to suppose that the number five is a physical object. So, will heaven be a relatively meaningless mode of existence because no opportunities for wickedness are present? I think not, for two reasons. First, heaven should be viewed as the
MORAL EVIL: THE COMPARATIVE RESPONSE
17.
18.
19. 20.
21.
22.
23.
24.
23
denouement of a larger story. So, its significance is, in part, a function of the significance that is grounded in the wide range of challenges and dangers of this earthly life. Second, I am not sure it is quite right to say that there will be no opportunities for wickedness in heaven. Those in heaven may look back on their earthly lives rather in the way that most of us can look back on some earlier stage of our earthly life, and see that we were foolish or engaged in a pattern of wrong actions. We could in principle return to that foolishness or wrong behavior, but we’ve “learned our lesson.” We could turn back, but we won’t. Does T7 imply that the wicked will be punished in the afterlife? Only if such punishment is demanded by perfect goodness. The doctrine of hell is outside the scope of this paper and it is highly contested territory even among Christian theists. For a range of positions, see Kallistos Ware, “Dare We Hope for the Salvation of All?” The Inner Kingdom (New York: St. Vladimir’s Seminary Press, 2000), pp. 193–215; Clark Pinnock and Robert Brow, “Hell: Rejecting Love”, Unbounded Love (InterVarsity Press, 1994); Michael J. Murray, “Heaven and Hell” in Michael J. Murray, ed., Reason for the Hope Within (Eerdmans, 1999), pp. 287–317. The specific examples of fine-tuning listed here are gleaned from Robin Collins, “A Scientific Argument for the Existence of God” in Michael J. Murray, ed., Reasons for the Hope Within (Grand Rapids: Eerdmans, 1999), p. 49. A more complete list is provided in Hugh Ross, The Creator and the Cosmos, third expanded edition (Colorado Springs: NavPress, 1993), chapter 14. For a more rigorous and detailed discussion of these issues, see John Leslie, Universes (New York: Routledge, 1989), esp. chapters 2 and 3. The illustration is borrowed from Collins, op. cit., p. 67. For works that appeal to naturalism to support moral anti-realism, see: J. L. Mackie, Ethics: Inventing Right and Wrong (New York: Penguin Books, 1977); Bernard Williams, Ethics and the Limits of Philosophy (Cambridge: Harvard University Press, 1985); Allan Gibbard, Wise Choices, Apt Feelings: A Theory of Normative Judgment (Cambridge: Harvard, 1990); Simon Blackburn, Essays in Quasi-Realism (New York: Oxford University Press, 1993). For an interesting exploration of issues related to naturalism and morality, see George Mavrodes, “Religion and the Queerness of Morality” in Rationality, Religious Belief and Moral Commitment: Essays in the Philosophy of Religion, eds. Robert Audi and William Wainwright (Cornell University Press, 1986); anthologized in Louis P. Pojman, ed., Philosophy of Religion (Wadsworth, 1991), pp. 573–581. As noted previously, theists disagree about the scope of God’s freedom. Many have held that God is essentially perfectly morally good, and hence unable to do wrong, but even these theists are apt to regard God as having the power to choose between various supererogatory actions. See, e.g., Thomas V. Morris, Our Idea of God (Downer’s Grove: InterVarsity Press, 1991), pp. 56–64. If the arguments of this paper show that theism explains moral evil better than naturalism, then they amount to a kind of moral argument for theism. The argument would be an “argument to the best explanation” in favor of theism over naturalism, with moral evil as the phenomenon to be explained. I wish to thank the following for helpful comments on earlier drafts of this paper: Terence Cuneo, Phillip Goggans, Kenneth Einar Himma, and Patrick McDonald.
Address for correspondence: C. Stephen Layman, Philosophy Department, Seattle Pacific University, 3307 Third Avenue West, Seattle WA 98119, USA Phone: 206-281-2036; Fax: 206-281-2335; E-mail:
[email protected]