Strategically placed at the core of the problem of evil—of how suffering, dysfunction, the unnecessary triviality of unrewarding experiences, and discord can exist in the presence of a God who has perfect power, knowledge, and goodness—lies an understanding of what God is able to do and in fact does. Because God has the power both to effect and to affect events in the world, both directly and indirectly, one must wonder why there is evil at all, or at least so much evil.
Attempts have been made by process philosophers to reconceive the traditional understanding of divine power and thereby avoid, if not resolve, the incompatibility or tension that exists between God’s existence, God’s possession of perfection properties, and the presence of evil in the world. As David Griffin puts it, “A central, and perhaps the central, notion in the traditional idea of God in Western thought has been the notion that God controls, or at least could control, every detail of the events in the world. This has been taken as belonging to the defining essence of ‘God’…. God is ‘omnipotent’ in the sense that nothing happens which God does not either cause or at least permit. ‘Permission’ is used here in the sense of allowing something to happen when one has the power to prevent it” (GPE, 16-17). This is the classic or traditional idea of omnipotence. Griffin holds that adopting this traditional view problematically precludes any successful resolution of the problem of evil, for it leads to the inference that God unilaterally could remove all evil if God desired to do so, and, presumably, as perfectly good, God would want to remove all evil.
To resolve the problem of evil, process philosophers use the general strategy of clarifying the type of power possessed by God and other actual entities, which in process thought are delineated by having their own unity of experience. “The crucial issue is taken to be the nature of power. Specifically, the question is: if an actual world necessarily has power, is it possible for God to have the type of power in relation to the world which was ascribed to God by the traditional doctrine of omnipotence?” (GPE, 12-13). Process philosophers roundly reject the idea of omnipotence classically understood as God being able to unilaterally cause or permit every event. Rather, since God is an actual entity like other world constituents, God is part of the nexus of the world and cannot override the freedom of self-determining entities. Self-determination is an intrinsic feature of the actual world and not a gift derived from a divine creative act; an actual occasion or entity that loses its self-determination is no longer an actual occasion or entity. God’s power is maximized in terms of God’s ability to causally affect other actual entities by presenting ideal possibilities (final causation) and by persuasion (efficient causation) while accommodating their noneliminable freedom. The problem of evil can only be addressed by recognizing God’s “limits”: in terms of efficient causation, God can only persuade; he cannot unilaterally resolve issues by coercive power. This “cannot” may look like a limit, but in actuality it is not, for God still can do what is possible for God to do. And if God necessarily cannot unilaterally resolve situations involving evil, God cannot be held morally accountable for this impossibility. With the removal of belief in God’s unilateral coercive power, held by both traditional theodicists and their critics, the problem of evil disappears.
This may look like the classical theist’s free will theodicy, where God is responsible for, but not morally responsible for, evil because he grants humans free will. But this apparent similarity is a grave misconception, for the free will theodicy assumes that freedom is largely a human feature God granted in creation that can in principle be overridden or removed by its donor. For process thinkers, God did not give individuals or actual entities freedom, and God cannot override their intrinsic freedom. Self-determining power is an intrinsic property of actual entities. Hence, process thinkers cannot (and do not want to) use the classical version of the free will theodicy.
Process thinkers want their readers to see that the traditional solutions to the problem of evil have not worked and will not work because they cannot work. Traditional theodicies cannot provide an adequate solution because advocates of these solutions have seriously misunderstood God’s power and the nature of the world. It is this formidable challenge to the classical solutions to the problem of evil and their critiques that we will undertake to carefully consider.
THE PARADOX OF OMNIPOTENCE
Some philosophers have sought to establish the unintelligibility of the concept of divine omnipotence by appealing to the paradox of omnipotence. Suppose that we define “omnipotence” as
(D) A being x is omnipotent if and only if (1) it is capable of bringing about any contingent state of affairs (a) whose description does not contain or entail a contradiction, and (b) whose description does not exclude or entail the exclusion of x or any omnipotent agent from among those that may have brought about that state of affairs, and (2) no being y greater in power than x can be conceived.
Does (D) fall prey to the paradox of omnipotence?
We might formulate the paradox as follows, where x is any being:
(3) Either x can bring about the existence of an object that x cannot lift, or x cannot bring about the existence of an object that x cannot lift.
(4) If x can bring about the existence of an object that x cannot lift, then, necessarily, there is at least one state of affairs that x cannot actualize (namely, the lifting of the object in question).
(5) If x cannot bring about the existence of an object that x cannot lift, then, necessarily, there is at least one state of affairs that x cannot actualize (namely, the bringing about of the existence of an object that x cannot lift).
(6) Hence, there is at least one state of affairs that x cannot bring about.
(7) If x is an omnipotent being, then x can bring about any state of affairs that meets conditions (1) and (2) in (D).
(8) Therefore, x is not an omnipotent being.
In the argument above, (3) is true by virtue of the principle of excluded middle, while (6) follows from (4) and (5). (7) follows from our definition of omnipotence, while (8) follows validly from (6) and (7). Since for the paradox to succeed it must be shown that the states of affairs described in (4-6) meet conditions (1) and (2) of definition (D), let us focus our attention on premises (4) and (5). Since (5) parallels (4), what we have to say about (4) likewise will apply to (5). To see whether (4) meets condition (1) of (D), let me first suggest three possible descriptions of the state of affairs indicated in the antecedent of (4) insofar as they relate to (6).
(9) An omnipotent being creates an object its maker cannot lift.
(10) God creates an object that God cannot lift.
(11) An object is created that its maker cannot lift.
Considering (9) first, that (9) fails to meet condition (DI) can be seen as follows. The maker of the object referred to in the predicate is the omnipotent being mentioned in the subject of the sentence; further, as omnipotent, this being can lift anything. Hence we can reformulate (9) as
(9*) An omnipotent being creates an object that a being that can lift anything cannot lift.
But (9*), containing a contradiction, describes a logically impossible state of affairs and hence, in failing to meet (1a) in (D), lacks any relevance to the question whether a being is omnipotent.
The same analysis results from the consideration of (10), provided God is considered necessarily omnipotent, for we can substitute for (10)
(10*) A necessarily omnipotent God creates an object that God, who can lift anything, cannot lift.
As with (9*), (10*) likewise contains a contradiction and hence fails to meet condition (1a) of (D). Thus (10*) is irrelevant to the issue of omnipotence.
Interpretation (11) depends on the referent of “its maker.” If “its maker” refers to an omnipotent being, the above problem resurfaces. It speaks of an object that a being that can lift anything cannot lift. If “its maker” refers to a finite being, then an omnipotent being can indeed bring about a state of affairs that would make (11) true, namely, bring about an x that cannot bring about a particular state of affairs. For example, an omnipotent being might create a beaver and cause it to construct a twenty-five-foot dam of sticks and mud across a stream, a dam that, once constructed, the beaver as the maker would be unable to lift. But the maker (the beaver making the dam) is finite, and hence so understood (11) is irrelevant to omnipotent beings. We can conclude that in invoking such an antecedent, (4) is irrelevant to the omnipotence of any being.
A similar case can be made for (5). If interpreted along the lines of a negation of (9) or of (10), (5) fails to meet condition (1a) of (D) and hence is irrelevant to the issue of omnipotence. If interpreted like the negation of (11) along the lines of an infinite referent, a contradiction results. If interpreted along the lines of a finite referent for x, (5) is plainly false: an object can be created that its (finite) maker cannot lift, as shown above. Thus, again, the antecedent is irrelevant to the question of omnipotence. Hence, proposition (6) does not follow from (4) and (5), and the paradox of omnipotence fails.
The above discussion suggests that there are certain limits to God’s power, namely, the limits of the rational. Although from the perspective of some thinkers, even the imposition of this condition restricts God’s freedom, a correct counterargument is that to do the logically impossible is to do nothing. Consequently, this stipulation of the limit of the rational does not really impose a limit on God’s power. But are there other limits to God’s power? One such candidate has been proposed by process thought, namely, that limits are imposed by the world itself as composed of necessarily self-determining actual entities.
THE OMNIPOTENCE FALLACY
The process philosopher David Griffin has serious objections to (D) and thereby (7). For him, the advocacy of this notion of omnipotence leads to the unresolvability of the problem of evil, for if God can unilaterally bring about any state of affairs, God can unilaterally remove genuine evil, something that God has not done. Griffin terms what lies behind (7) the “omnipotence fallacy” (GPE, 263):
P: An omnipotent being can unilaterally bring about any state of affairs that it is logically possible for a being unilaterally to bring about.
R: An actual world (i.e., one with a multiplicity of actual beings) devoid of genuine evil is a logically possible state of affairs.
S: Therefore, an omnipotent being could unilaterally bring about an actual world devoid of genuine evil.
Unfortunately, he continues, the argument itself is not formally valid, since it contains an “oscillation between a logically possible action (P, S) and a logically possible state of affairs (R)” (GPE, 264), therefore committing the fallacy of four-terms. Thus, Griffin notes, to make the argument valid, another premise—affirming that all states of affairs that are logically possible are states of affairs that it is logically possible for one being unilaterally to bring about—is necessary. This premise Griffin labels
Q: It is logically possible for one being unilaterally to bring about a state of affairs among a multiplicity of actual beings.
Suppose we grant that PQRS is formally valid. Is there any reason why the traditional theist who believes that God is omnipotent must accept the argument? Griffin claims that all traditional theists are committed to Q, although in a parenthesis he qualifies this statement when he admits that some theists limit the scope of Q. But if Q’s scope is limited, then theists are not committed to this argument. One can see that theists are not committed to the argument by considering P. P simply is false; there are logically possible states of affairs that an omnipotent being cannot bring about, because for an omnipotent being to bring them about involves a contradiction. Consider, for example, the proposition
(12) George freely performs the action of yelling at his dog.
George’s freely yelling at his dog is a logically possible state of affairs that a being unilaterally could bring about; George himself can actualize it. At the same time, it is a state of affairs that an omnipotent being (given that George is not omnipotent) cannot unilaterally bring about, for to do so would be unilaterally to bring about the state of affairs of George’s being caused by another being freely to yell at his dog. Since freely bringing about a state of affairs entails that one was not caused by another to bring it about, (12) entails
(13) No other being causes George to freely perform the action of yelling at his dog.
Thus, it is the case that the state of affairs described in (12) fails to meet condition (1b) of our definition (D) of omnipotence, for (13) explicitly excludes any being other than non-omnipotent George, including an omnipotent being, from bringing about the state of affairs in question. Consequently, though (12) describes a logically possible state of affairs, it is not the case that an omnipotent being would be able or expected to bring about that state of affairs. Accordingly, proposition P is false, and no libertarian theist can be expected to assent to the argument PQRS.
Griffin, however, rejects the argument PQRS for a different (though related) reason. It is instructive to see the argument he presents, for it will provide us with insight into how process philosophy looks at the critical question of the nature of God’s power vis-à-vis the doctrine of omnipotence. Griffin rejects argument PQRS—and along with it traditional theism—because in invoking Q both the argument and traditional theism (presumably) presuppose the key premise
X: It is possible for one actual being’s condition to be completely determined by a being or beings other than itself (GPE, 264).
And X, he holds, is meaningless because determinism by any being is incompatible with the self-determination of actual beings or entities. However, traditional Christians who hold a libertarian view on human freedom likewise reject X in the case of actions of free beings or the condition of being free, for this would remove the freedom that characterizes humans.
Where traditional Christians and process thinkers differ concerns the scope of X: is “actual being” a technical term or a general term referring to anything that exists? Answering this question will help us to understand the process view of divine power.
LIMITATION BY NECESSARILY SELF-DETERMINING BEINGS
Process theology holds that the world is composed of actual entities or actual occasions, each of which is essentially and necessarily self-determining. God does not and cannot control actual entities, not because of any act of divine self-limitation, but because of metaphysical necessity—because of what it is to be an actuality. “It is necessarily the case that God cannot completely control the creatures…. To be an actuality is to exercise creativity” (GPE, 276), to have power, to be a self-determining being. Thus any world that God could have created (or, better, brought out of chaos, since there have always been finite actual occasions) would necessarily contain actual entities or occasions that were self-determining (possessing power). These self-determining actual entities include the entire continuum from inanimate electrons, atoms, molecules, and animate cells to animal and human psyches (minds, souls) (ER, 216). In effect, the beings referred to in Q and X are these self-determining actual entities or occasions.
Griffin’s argument fundamentally proceeds as follows.
(14) Necessarily, no actual entity is devoid of power.
(15) Necessarily, if an actual entity can be completely determined by another, it must be devoid of power (GPE, 266).
(16) Therefore, necessarily no actual entity can be completely determined by another.
That is, X is false.
The critical premise in Griffin’s argument is (14). In its defense, Griffin argues as follows:
(7) The (only) actual entity I directly experience is my mind or soul.
(18) My mind or soul has some power to determine itself.
(19) All actual entities are analogously the same.
(20) Therefore, what I mean by actual entity is a being with some power to determine itself.
(21) Therefore, all actual entities necessarily have some power to determine themselves.
He contends that this is not an inductive argument, but an argument relying on an “empirical criterion of meaning” according to which “the meaningful use of terms requires an experiential grounding for those terms” (ER, 141). The term here is “actual entity,” and my experience of myself as having power provides the experiential grounding. We have “an experiential basis for speaking meaningfully of an actual entity which I can then transfer by analogy to other things” (GPE, 267).
Several criticisms of this argument might be raised. First, regarding (19), why must all actual entities be analogous? Griffin rejects any inductive defense of the claim; instead it seems to be a reasonable a priori presumption of our language usage. Second, supposing that all actual entities are analogous, why should one think that they are analogous in one respect or property possession rather than another? Griffin holds that they are analogous with respect to self-determination; but why not also think that actual entities, as individuals, are analogous with respect to being animate or rational (which properties we experience of our minds but would not apply to electrons)? That is, why universalize one property rather than another? Third, even if the property is possessed universally, why should one think that this property is part of the meaning of the term? Griffin does not think that animation or sentience is part of the meaning of “individuality” (ER, 217). Finally, the very experiential criterion of meaning is problematic. We might not experience properties found in the meaning of a term; yet we can understand and use a term meaningfully. For example, it is part of the meaning of “electron” that it is a negatively charged particle; and though being a particle with a negative charge is not something I can experience, I can know what an electron is and use the term meaningfully.
In short, even if (19) is true, (20) is suspect in that we have moved from an empirical feature of my singular experience in (17), through the a priori universality of (19), to the necessity, embodied in the meaning of the term “actual entity,” found in (20). But universality by itself cannot establish the necessity of a property unless the a priori metaphysical is introduced. And this runs the risk, not of providing support for (14), but of begging the question.
What does the falsity of X tell us about God’s power? Griffin argues for what he calls C omnipotence, that is, “coherent, creationistic omnipotence.” Power is a relational concept, such that any “delimitation of perfect power requires a discussion of the nature of ‘world.’… Hence, before drawing implications as to what a being with perfect power could do, the nature of the things upon which power is to be exerted must be considered” (GPE, 265). Since the actual world—indeed, any actual world—is necessarily composed of self-determining actual entities or occasions that possess and actualize power outside the control of other beings, a being that is perfect in power cannot possess “a monopolistic concentration of power.” Actual entities have “both the power of self-determination and the power to exert causal influence upon others” (ER, 57), but not the power to determine other actual entities unilaterally. Thus, God does not have all the power there is, but God does possess all the power a being can conceivably possess consistent with there being other actualities necessarily having power. Consequently, this power does not allow God to deterministically and unilaterally bring about any state of affairs in the world, including the absence of evil.
GOD’S PERSUASIVE POWER
What kind of power does God possess? God’s power is not the power of coercion, “measured by the incapacity to resist on the part of that on which it is wielded.” However, God uses his power in a persuasive manner. Whereas coercive power or compulsion is employed upon the less powerful or powerless, persuasion is employed upon those with sufficient power to resist. We have seen that all actual occasions are essentially powerful. For God to coerce actual occasions to perform in a certain manner would be to remove them from being actual occasions, which is metaphysically impossible. Therefore, God’s power—and corresponding action—cannot be understood in terms of compulsive, coercive, deterministic efficient causation. This means that to speak about God unilaterally bringing about states of affairs, as does (D) above, is fundamentally mistaken. Rather, God has the power of influence or persuasion: God “exercises the optimum persuasive power in relation to whatever is.”
What, precisely, is persuasive power? Some process thinkers have held that God’s power must be thought of in terms of final causation, wherein God introduces ideal initial aims and richer possibilities of order and complexity for actual occasions to realize. The process God calls forth novelty and, through conceptual innovation, free thought and free imagination; and ultimately he provides for the freedom of moral responsibility. God’s power is seen in God’s calling actual entities to the full realization of their potentialities. “By the way God constitutes himself he calls us to be what we can be and are not. He constitutes himself so as to provide each occasion with an ideal for its self-actualization, and it is in relation to that ideal that each human energy event forms itself.” Thus, God lures each and every actual entity to self-realization according to its nature.
Griffin demurs from this identification of God’s persuasive power with final causation. He distinguishes between final causation, which is the causation of self-determination exerted intrinsically within a single actual entity, and efficient causation, which is the causal influence that one actual entity has (extrinsically) on another. The final cause for actual entities, on the one hand, is the initial aim supplied by God as an ideal possibility; on the other hand, it is the internal, subjective aim of the actual entity.” God elicits novelty, increasing the enjoyment experienced by actual entities in realizing their possibilities, but whether the actual occasion adopts the initial aim is a matter of self-determination.
It is the efficient cause, the causal influence exerted between actual entities, that concerns the doctrine of omnipotence. Griffin distinguishes two kinds of efficient causation: persuasive and coercive. These, he notes, can be understood both metaphysically and psychologically. In a metaphysical understanding, the actual entity or effect involved in persuasive efficient causation is not completely determined by the efficient causation brought to bear upon it; it partly determines itself in various degrees, which depend, for example, on whether it is inanimate or animate. In coercive efficient causation, the cause completely determines the effect. These two kinds of efficient causation are mutually exclusive: the effect is either completely determined (coercive causation) or not completely determined (persuasion). Understood psychologically, coercive and persuasive causation are not mutually exclusive but exist on a continuum; they seem synonymous, respectively, with extrinsic (coercively deterministic) and intrinsic (self-determinative) influences.
Applying these distinctions to the power God can exercise, Griffin holds that God cannot exert psychologically-understood coercive efficient causation since, as externally determinate, this depends upon metaphysical coercion, of which God is also incapable. God also would be incapable of psychologically-understood persuasive efficient causation, since that is intrinsic and not extrinsic to the actual entity. When it comes to a metaphysical understanding of efficient causation, neither God nor anything else can conduct coercive efficient causation on actual entities, since they are essentially self-determinative. Hence, the only power of efficient causation that God has is persuasive (nondeterministic) efficient causation.
The language here is confusing, not least because in ordinary language, persuasion is typically psychological (ER, 102), while coercion differs from persuasion not in degree but in kind. Perhaps we would be better served and the discussion significantly simplified to talk about whether the cause is sufficient to produce an effect in another. In fact, Hartshorne suggests this approach. This fits well with Griffin’s talk of God “using persuasion” (GPE, 276). While a coercive efficient cause would be sufficient to cause the actual occasion, a persuasive efficient cause cannot be sufficient to cause the actual occasion. But Griffin probably would not deem this substitution satisfactory, for persuasion requires that the actual entity (the effect) determine itself at least in part, whereas language about sufficient conditions does not require this; what prevents something from being a sufficient cause might be the necessity of other external causal conditions. Griffin also might reject this approach for another reason. If the “in part” is very small (Hartshorne “approximately” in referring to “conditions sufficient to determine approximately what happens”), and if natural laws (even if only limiters and not determiners) are operative, talking about sufficient conditions comes close to allowing God as a sufficient condition to approximately or mostly affect nature as God desires, even if only in a limiting way; this might allow the reintroduction of the problem of natural evil that process thought attempted to avoid. (I shall return to this point below.)
One might think that in departing from definition (D) of omnipotence, Griffin advocates a conception of a finite God whose power is limited by the power possessed necessarily and essentially by other actual occasions. Griffin rejects this contention on the grounds that God has all the power that God conceivably could have; no other being could possess more power.
THE CENTRAL PROCESS ARGUMENT
Griffin’s avoidance of the view that God’s power is finite (finitism) depends upon Griffin’s contention that the following propositions are necessarily true (in virtue of their being metaphysical principles applicable to all possible worlds).
(22) There is an actual world (GPE, 279).
(23) To be a world is to contain actual entities or actual occasions (GPE, 269, 279).
(24) All actual members of a world either are or are composed of actual or entities or actual occasions (GPE, 249, 277).
(25) All actual entities or occasions are essentially self-determining.
(26) The only power self-determining actualities can have consistent with there being other essentially self-determining actualities is persuasive power (GPE, 280-81).
(27) God is an actual entity (GPE, 281).
(28) God possesses the maximal possible persuasive power (GPE, 273).
But are these propositions not only true but necessarily true? (22) obviously is true. Griffin thinks it is necessarily true because God is essentially creative and hence cannot exist without a world with which to engage creatively. Since God necessarily exists, actual entities necessarily exist and hence always have existed. God does not create ex nihilo; rather, God’s creating activity brings order out of chaos, harmony out of discord (GPE, 286). Classical theists agree that (22) is true, but not necessarily true; creation as bringing about the universe ex nihilo is a free act of God. But since the process view of how God creates is a consequence of this set of necessarily true propositions, we shall begin elsewhere in the set. Griffin holds (23) and (25) to be true by the definitions of “world” and “actual occasion.” Whether or not (24) is true depends upon how one interprets “actual entity,” and this is clarified in (25). So understood, (24) asserts that all constituents of any world either are or are composed of experiencing, power-exercising, self-determining, creative actualities.
Here, one might suggest, lies part of the core of the dispute between tradition and process theism. On the one hand, traditional theists restrict self-determination to conscious beings, and some even more narrowly to moral agents. For traditional theists, if (25) is true, (24) is false. On the other hand, process thought contends that the universe is totally composed of self-determining actual entities, both inanimate and animate. Both (24) and (25) are true.
The other crucial difference between process thought and traditional theism lies in (26). Traditional theists allow coercion of self-determining beings when their power of self-determination is limited, obstructed, or temporarily removed. Of course, while coercively limited or obstructed, the being is not self-determining, and if self-determination is completely removed over a significant period of time, the being is a robot or is nonfunctioning (e.g., dead). Perhaps coercion can be more clearly shown in the claim by traditional theists that God allows or permits beings to exercise their power.
Process thought also allows coercion of entities, but this is possible only because process thought makes a further distinction between two types of entities: individuals and aggregates. Individuals are either actual occasions (or actual entities) or serially ordered societies of such (ER, 216). Since individuals have a unity of experience, they are self-determining and hence cannot be coerced. Neither can they coerce; only by persuasion can they be efficient causes of other individuals.
Aggregates (roads, balls, rocks, buildings, corpses) are enduring things, the objects of the world, with “no self or unity of experience and therefore no power of self-determination” (ER, 102). “The only experiences in the aggregate are the experiences of the actual occasions constituting the various enduring individuals.” They do have a coherence because they are composed of societies of self-determining individuals. But even so, they are not even in part self-determining, for the whole need not have the same properties as its parts. Consequently, they are only coercible; “aggregates as such can be moved only by another aggregate” (ER, 217). Individuals cannot coerce either aggregates or other individuals; they can only persuade. Neither can aggregates coerce individuals. And since God is an individual, not an aggregate, God cannot causally coerce or determine anything.
What is interesting is what happens to X—”It is possible for one actual being’s condition to be completely determined by a being or beings other than itself”—when these categories are developed. X is meaningless for Griffin only if “actual being” refers to actual occasions or entities. But if “actual being” is expanded to include aggregates, which wait “upon some external force for movement” (ER, 217) and can be coerced (ER, 103), then Griffin must think, like the traditional theist, that X is true. So the movement of the discussion from actual entities to individuals and aggregates is not inconsequential to the process argument.
APPLICATION TO REAL LIFE
How does this metaphysic work out in real life? How do these abstract categories explain how fathers move their children from room to room, how people move their bodies, how billiard players control cue sticks to move balls, or how one person affects another person (to use Griffin’s examples)? Griffin gives the example of a father who picks up his young daughter to carry her against her will to another room. Her father can coerce her body when he picks her up, but he cannot coerce his daughter to want to move; only she can want to move. We might say that the father can coerce her body but can only persuade her will. One would think that the father, as a person choosing to carry his recalcitrant daughter, would be a prime example of an individual capable of self-determination. But since bodily coercion is involved, and coercion occurs only between aggregates, the father must be an aggregate. How can this be understood?
Griffin holds that human beings are psychophysical entities composed of billions of actual entities. Yet they are not mere aggregates, for they have a series of dominant occasions of experience, called a soul or mind. Because a human being has a soul, it “is an example of that special type of structured society called a monarchical society” (ER, 111); “the psyche (as a serially ordered society) is the monarch, insofar as it synthesizes the various experiences of the cells into a unity of experience” (ER, 218). Having a soul, humans (and animals with central nervous systems) are considered as self-determining and cannot be coerced. At the same time, we also can consider humans only in terms of their bodies. Qua bodies, humans are aggregates and hence can be coerced. So the father as deciding to carry the girl is viewed as an individual; and the father reasoning with his daughter and the daughter agreeing to move are regarded as individuals. But the father carrying the girl and the carried girl are viewed as aggregates, for the girl’s body is coerced to go to another room, and only aggregates can coerce aggregates.
What can be said about the billiard player? The 8-ball is moved by the cue ball, obeying the laws of motion, and the cue ball is moved by the cue stick, which has been moved by the player’s hand. All of these are aggregates and hence can be determinately coerced (ER, 103). But how does the soul (the self-determining aspect) relate to the hand (or, more generally, to the body that is an aggregate)? Since the soul is an individual and individuals cannot move or coerce aggregates, it would seem that the soul cannot move the hand.
Not so, says Griffin; the relation between soul (mind, psyche) and body is persuasive. The mind has persuasive power over the body because the body can make a self-determining response. But, one might reply, the individual qua body was treated as an aggregate, and since aggregates cannot apprehend, they cannot be persuaded. Why then should one think that the body or part of the body that one wants to move persuasively and not coercively is an individual that can make a self-determining response? Griffin appeals to the fact that the body may not respond to our wishes to move it when it is “impaired by injury, drugs, alcohol, or weariness” (ER, 103), such that it is in part self-determining.
But though the body or body part may not respond as we wish or at all, the fact that the body or a body part does not move when we want it to or in the way we desire fails to establish that the body has self-determining power; it only shows that the power of the mind and will can be contravened by other causal conditions, such as drugs or alcohol, which themselves originate externally to the body. Moreover, if the mind cannot coerce the body because the body can make a self-determining response, the body is no longer causal story is no longer consistent; problems arise at the point of the causal relation of mind to body, a difficulty not unknown to philosophers.
Griffin also holds that “the body can never, in the strict sense, coerce the psyche” (ER, 219). This is because the body as an aggregate cannot coerce an individual. But there exist many cases, especially in cases of physiologically based mental diseases, where the psyche is so coerced. Experiences of schizophrenic and autistic persons and of persons under hallucinogenic drugs are prime examples. Such persons have experiences which they neither might want nor can control, and these experiences determine features of their personality, behavior, and character. Or again, neurologists can stimulate the brain in particular places and ways to create particular experiences, which refutes Griffin’s position.
To address these problems, Griffin’s discussion of the “organizational duality” between individuals and aggregates is couched, not in terms of what is the case, but of what is regarded as being the case. Here he introduces again the concept of “the monarchical society, which combines features of both aggregates and genuine individuals. Whether or not a monarchical society can be coerced depends on how it is being regarded” (ER, 218). The father is viewed as an aggregate when he carries his daughter; he is regarded as an individual when he decides to carry her. The billiard player’s hand is viewed an aggregate when it is moving the cue stick; it (or the central nervous system that moves it) is considered an individual when it is moved by the soul or mind. But the question is not “how we treat them”; the question is what these really are. If fathers and billiard players are aggregates, they can move aggregates and manipulate cue sticks; they are coercible and not self-determining. They are more like cues, balls, and hands than individuals. If they are individuals, they are self-determining and not coercible, but are incapable of acting in the real world composed of aggregates, which are coercible but not persuadable. The field of action for the psyche must be regarded as persuadable and hence self-determining. There are inanimate aggregates that can coerce and be coerced, and there are individuals who can persuade and be persuaded, but where individuals come into causal relation with aggregates and vice versa, Griffin’s explanation seems to depend on how aggregates are viewed in any given instance, not in terms of what they are.
If the father viewed as an individual can move his daughter viewed as an aggregate coercively, why cannot God who is an individual be viewed as moving aggregates coercively? Griffin’s response is that we view the father moving the daughter because the father has a body, but we cannot view God as moving aggregates because God has no body: “Absolute coercion occurs when body acts on body, aggregate on aggregate” (ER, 112). God can only exercise persuasive causation on individuals or psyches. But the psyche moves the body not by means of the body, but directly. And if, as we argued above, the mind can coerce the body (as is often the case, fortunately, in our daily motor existence), there is no difficulty in seeing God as moving nonsentient entities by coercion. God can be a sufficient condition for effects in aggregates. Griffin’s view of God is narrowed by his view that one individual can coercively affect or effect another only by means of a body. That is why he has a problem with creation ex nihilo. But if we do not impose on God the necessity of using materialist sufficient conditions in order to act, the way is open for God to be active in other ways, to be a sufficient condition for bringing about events. As Genesis states, “God said and it was so.” Genesis does not tell us how God acts, but then again, Griffin cannot tell us how the psyche as an individual moves the central nervous system. Acknowledgement of the fact does not necessitate understanding the how.
Griffin wants us to believe that the difference between classical theism and process theism is metaphysical (GPE, 265-67). He is correct: the dispute ultimately is not about paradoxes or fallacies but about the way one deems the world to be, about competing worldviews. It is about propositions (22-28) and about how classical and process theists view existents in general. And it is about how they view human beings—their soul and their body and their respective powers. Contrary to Griffin, the evidence is clear that the mind can (and often does) coerce the body, and the body and external factors can coerce the mind or soul. And if this is so, there is nothing metaphysical to prevent God from causally determining the world or exercising his permissive power to allow self-determining moral agents to act.
In summary, one’s initial commitments usually determine the outcome of the discussion. The process philosopher, in affirming propositions (22-28), sees the world totally filled with self-determining actual occasions that either are or compose all that is. To account for what surely seems like coercion, these actual entities are conjoined into aggregates that lack self-determining power. But in moving from actual beings to aggregates, as we noted above, the causal relations of coercion and persuasion become problematic. X, as applied to aggregates, is no longer false, and the possibility of other causal relationships, viewed in terms of sufficient conditions, emerges. In short, the scope of X matters. The traditional theist sees the world as containing not only self-determining beings but also things or events whose natures, states, and existence can be and are determined through external and internal causation. One cannot read scripture without seeing God as personally working in the world in such a fashion. For process thought, God is not a person but an actual entity that can influence persuasively but not determinately produce effects by efficient causation. For classical theism, God is a person who can bring about his purposes, though often God chooses to do so through persons who exercise morally significant freedom. Though process thought provides significant insights that can and have been incorporated into traditional theism, the traditional theist finds little compelling reason to adopt a process metaphysic of self-determining actual entities or occasions. Given the problems we delineated above, the classical theist will strongly dispute Griffin’s bold contention that “the metaphysical position [of process thought] provides a more consistent, adequate, and illuminating account of experience and reality in general than any metaphysical position starting from the contrary hypothesis” (ER, 118).
THE NECESSITY OF THE POSSIBILITY OF EVIL
Griffin contends that a middle ground lies between the claim that genuine or gratuitous evil is necessary (which traditional theists and process philosophers would deny) and the claim that genuine or gratuitous evil is impossible (which atheists and many traditional theists affirm should follow from the Anselmian concept of a perfect God). The middle ground is that the possibility of genuine or gratuitous evil is necessary (GPE, 269). It is this that provides a resolution to the problem of evil. God must allow (for the traditional theist) or cannot prevent (for the process thinker) the possibility of evil in order to guarantee human freedom or genuine self-determination. The possibility of evil is necessary.
Griffin’s distinction between the necessity of evil and the necessity of the possibility of evil reveals a point generally overlooked by theists and atheists alike. Not all evils must be justified, in the sense of being shown to be preconditions for or means to a good, in order for theists to make their case that God is not morally responsible for evil. The presence of gratuitous or genuine evil—evil that of itself does not lead to or produce any greater good—does not jeopardize claims that God exists or has certain perfection properties, so long as the possibility of the existence of evil is justified either as logically or metaphysically necessary (process theology) or as being necessary for a greater good (free will theodicy).
This insight suggests that, in some respects, the difference between Griffin and certain traditional theists is not all that large. The difference between them lies not in the denial of the necessity of the possibility of evil but in the basis of the necessity, or, as Griffin puts it, in the metaphysical status of the necessity. For Griffin, the fundamental components of the world are self-creative. As Whitehead states, “All actual entities share with God this character of self-causation. For this reason every actual entity also shares with God the characteristics of transcending all other actual entities, including God.” The possibility of evil is necessary because actual entities have intrinsic powers that they can freely exercise. “One being cannot guarantee that all the other beings will avoid all genuine evil…. This position follows from the meaning of ‘world’ as containing self-determining beings, since it is not logically possible for one being completely to determine the activity of another entity that by definition has activity that is underived from any other being” (GPE, 269). From this Griffin concludes that the existence of genuine evil does not disprove the existence of an all-good and omnipotent (as he understands “omnipotent”) being. Though a world without evil is logically possible, a noncoercively omnipotent God cannot guarantee it or bring it about. God can persuade but not compel to that end because of the intrinsic nature of the world’s components as self-determining.
For the traditional theist, likewise, a world without evil is logically possible. However, the possibility of evil is necessary because God desires rational creatures to attain moral good, which requires that they have morally significant freedom. And the possession of morally significant freedom makes possible both moral and natural gratuitous evil. That is, God’s voluntary granting of morally significant freedom and the requisite conditions for it (e.g., natural laws) makes it possible that humans perform evil acts or that suffering results from our physical involvement with physical nature, an involvement requisite to carrying out moral intentions. Gratuitous evil does not directly lead to, and hence cannot be justified by, any good. That is why this evil is gratuitous or, in Griffin’s language, genuine. The presence of gratuitous or genuine evil does not entail anything about the existence of God or God’s lack of certain properties, so long as the possibility of there being evil is justified as being necessary for a greater good. The possibility of gratuitous evil results from conditions necessary for the existence and functioning of moral agents and the good they realize, and can realize, only as free agents. Hence, since the greater good realized by free agents requires morally significant freedom, which in turn makes possible moral and natural evil, and since an individual cannot be held morally accountable or blameworthy for what it is impossible to have done otherwise, it follows that God cannot be held morally accountable or blameworthy for gratuitous or genuine evils. By invoking a God who voluntarily limits his power to attain a greater good, we have a morally sufficient reason for allowing genuine or gratuitous evils without introducing a world totally composed of essentially self-determining actual entities or actual occasions. The difference, then, is that whereas the process theist sees God’s power as necessarily limited by the necessary self-agency of actual entities that compose the world, the classical theist sees God as voluntarily limiting his power in order to grant certain beings the morally significant freedom requisite for attaining a greater good. For the former, the scenario is a matter of metaphysical necessity; for the latter, it involves a voluntary act of self-limitation to bring about the conditions for realizing a greater good. But in both cases the explanation serves the same function: providing the possibility and space for self-determining entities.
There is little practical difference, I suggest, with regard to resolving the problem of moral evil: evil for which self-determining human agents are responsible. Both the process theist and the classical theist contend that in order to have self-determining individuals or beings capable of achieving moral good, these beings must be free to a significant extent. The difference really affects natural evil, evils that, on the classical theist’s view, are not the moral responsibility of self-determining human agents. Process theists contend that the same answer suffices for both kinds of evil: since everything is either an actual entity or composed of such, and since all actual entities are essentially self-determining, God is not morally accountable for evil—moral, natural, or aesthetic. But as we have seen, with the introduction of aggregates and the complex way individuals interact with them, we need a more nuanced discussion of natural evil. The classical theist can appeal to the self-determination of moral agents as a reason for God allowing natural evil, but not in the way that the process theologian does. For the classical theist, in order for there to be the possibility of morally significant choice and action, the world must have a general order, but this order makes natural evil possible.
Griffin raises several objections to this traditional theistic free will theodicy: here I will consider only the first. He argues that, on this view, God could be more proactive in preventing the worst evils. The required violations of human freedom would be minimal and an insignificant price to pay for this protection. Traditional theists can make two replies. First, the same objection can apply mutatis mutandi to Griffin. On Griffin’s view, God is persuasively causative in the world, just not coercively causative. But then one might wonder whether God could be more persuasive, taking causally efficient action that is short of being sufficient or coercive to remove evil but that is approximately effective. For example, following Hartshorne’s suggestion regarding natural laws, could not God be more limiting in their effects on nature in a given circumstance? After all, actual entities have varying degrees of self-determination. Griffin’s response sounds familiar to the theist: God exercises his persuasive power maximally, but God’s maximal exercise cannot remove all evil, for evil arises from the self-determination of all actual entities. It might not appear that God is maximally persuasive, but that does not mean that God is not so.
The traditional theist’s second response follows on the first: it is perfectly possible that God is already involved in removing the worst evils from the world. We do not know what a world might be like without a providential God, but very likely it would be worse than this world. But, Griffin replies, there are a significant number of horrendous evils that God seemingly could prevent with his coercive power, if God had such power as the classical theist suggests. The principle behind Griffin’s reply is that a coercively omnipotent God could and should always intervene to remove the worst evils, even at the price of removing some freedom. But what the worst evils are is a relative notion, relative to all the other evils. Hence, a consistent application of this principle would necessarily lead to a world without evil, without freedom, and without morally significant beings. This does not mean that we know where the line of maximal divine intervention lies. Indeed, petitionary prayer makes sense for us because we do not know this line. But our ignorance of this line of divine intervention does not mean that such a line, beyond which significant human freedom is in jeopardy, does not exist. The same issue arises for Griffin as well, for one might suggest that God is not persuasively enough involved in the world, especially with regard to horrendous natural evils or discord. Granted, God cannot coercively eliminate the evils, but it seems that God could exercise his persuasive causation to a greater extent to limit horrendous evils. After all, God could have tried to persuade others to assassinate Hitler; or he could have been more persuasive in encouraging the Corps of Engineers to build higher dikes in New Orleans. Griffin responds that God not only persuades people, but God is using God’s persuasive power already to prevent or minimize horrendous evils (ER, 115). Like the traditional theist, Griffin does not know where the line of maximal persuasive involvement lies, except to say that from the amount of evil present in the world we cannot calculate that line of maximal persuasion (ER, 115-16), which is the same reply the libertarian theodicist gives to the initial criticism raised by Griffin.
CONCLUSION
From the foregoing discussion it is clear that libertarian traditional theists are able to resolve the paradox of omnipotence and do not commit the fallacy of omnipotence. It might be argued that process theodicy advances the discussion, in that it makes it logically and metaphysically impossible for God to unilaterally eliminate the evil in the world, rather than invoking a voluntary divine act of self-limitation. But as we have seen, this advantage quickly disappears when process thought allows God to be a persuasive efficient cause. A process theodicy not only faces similar objections to those raised against a libertarian theodicy but also commits us to a metaphysic of actual occasions that is implausible when conceived of universally and is fraught with problems when implemented practically on the scale of individuals and aggregates. But on one major issue Griffin is correct: the debate between a libertarian and a process theodicy is at heart a metaphysical debate regarding which worldview one believes it is most rational to adopt. And if one argues, as Griffin does, that the decision between worldviews should be based on which worldview is “more coherent, adequate, and illuminating” (ER, 119), traditional theists are confident in the logical coherence and rational adequacy of their view and ready to engage in the ongoing debate.
Bruce R. Reichenbach, “Evil, Omnipotence, and Process Thought.”