The challenge

In advancing his defence, Plantinga aims to show that the following two propositions are consistent:

Plantinga's strategy is to find a third possibly true proposition that is consistent with (G) and, together with (G), entails (E). He suggests that the following proposition does the trick:

Notice that (TWD) specifies that God actualizes a world containing moral goodness, but it does not specify why God actualizes such a world. Without some indication of what might justify a wholly good God in actualizing a world containing moral goodness, there is little reason to accept Plantinga's claim that (G) and (TWD) are compatible. To see this, suppose that Plantinga had proposed the following as his third proposition: God actualizes a world containing some evil. That proposition clearly entails (E), but there is little reason to accept that it is compatible with (G). As it stands, (TWD) has much the same weakness. Fortunately, there is a straightforward way of expanding (TWD) to solve the difficulty. Plantinga's discussion of the free-will defence suggests that Plantinga sees moral goodness as an intrinsic good. In offering what he describes as a ‘preliminary statement’ of the free-will defence, Plantinga says that: ‘[a] world containing free creatures who are sometimes significantly free (and freely perform more good than evil actions) is more valuable, all else being equal, than a world containing no free creatures at all’ (Reference Plantinga1974, 166). That passage, together with the fact that Plantinga never speaks of some other good to which moral goodness is a means, suggests that Plantinga's view is that the presence of moral goodness in a world increases the value of that world because moral goodness itself is intrinsically valuable. Furthermore, Plantinga must hold that any intrinsic evils entailed by the existence of moral goodness are outweighed or defeated (in the sense explained in Chisholm (Reference Chisholm1968) ) by the intrinsic value of moral goodness. For if that were not the case, then a world containing free creatures who are sometimes significantly free and freely perform more good than evil acts might fail to be more valuable than a world containing no free creatures in virtue of the existence of intrinsic evils entailed by moral goodness. Accordingly, it is reasonable to expand (TWD) this way:

In the rest of this section I make the case that given various claims that are plausible and/or to which Plantinga appears to be committed, (G) and (TWD-E) are incompatible. First, I make the case that (TWD-E) together with various Plantingian/plausible claims entails that possibly God is capable of engaging in moral goodness (as defined below). Then I make the case that (G) together with various Plantingian/plausible claims entails that necessarily God is not capable of engaging in moral goodness. If (G), (TWD-E), and the Plantingian/plausible claims to which I appeal generate a contradiction, then many theists who would endorse the free-will defence as a solution to the logical problem of evil have some thinking to do. Specifically, if they wish to endorse theism and Plantinga's free-will defence, then they must either reject at least one of the relevant Plantingian/plausible claims or identify a weakness in my argument.

In (TWD-E), ‘moral goodness’ indicates the performance of a morally right action in a situation in which one possesses morally significant freedom.Footnote ⁶ To have morally significant freedom is to be in a situation in which it is true both that (a) you are free to perform a morally right action and (b) you are free to perform a morally wrong action (Plantinga (Reference Plantinga1977), 30). But what, by Plantinga's lights, is required for an agent to be free to perform a given action? Plantinga identifies two factors that he claims would prevent an agent from being free to perform a given action: (i) it being the case that God determines or causes the agent not to perform the action (ibid.) and (ii) it being the case that antecedent conditions and/or causal laws determine that one does not perform the action (ibid., 29). But why, according to Plantinga, are those factors a threat to free action? Plantinga doesn't explicitly say, but a remark he makes in his brief critique of compatibilism suggests an answer. In his evaluation of the suggestion that a given action could be determined and yet free, Plantinga points out that the fact that it is causally determined that a certain person performs a particular action implies that there is ‘some event beyond his control’ that causally necessitates that the person perform the action in question (ibid., 31, emphasis added). And, indeed, one thing that is common to cases in which one's action is determined by God and cases in which one's action is determined by antecedent conditions and/or causal laws is that in both cases one's action is causally determined by conditions or events that are entirely beyond one's control. Accordingly, it is plausible to suppose that Plantinga holds something like the following:

While condition (ii) of (P) focuses on causal determination, an even stronger relation is entailment. If one accepts that causal determination by an event outside of one's control is a threat to free action, then it is hard to see how one could reasonably deny that entailment by a truth outside of one's control is similarly a threat to free action. Thus, Plantinga's rejection of compatibilism points us in the direction of the thought that free actions are ones over which the agent has at least partial control and that the complete absence of control is transmitted by or closed under entailment. The following principle captures that idea:

Combining (FAC) and Plantinga's understanding of morally significant freedom yields the following principle:

Thus, (TWD-E) entails that God actualizes a world in which there is at least one creature that satisfies the right-hand side of (MSF) and performs a morally right act.

Outside of the context of the free-will defence, Plantinga argues that the claim that God exists entails that God is maximally excellent in every possible world (Plantinga (Reference Plantinga1977), 107–109). Plantinga further claims that in order to be maximally excellent in a given world, God must possess in that world the maximum degree of each great-making property – properties like wisdom, knowledge, and power (ibid., 91). In his brief discussion of moral perfection (a property that a maximally excellent being would possess), Plantinga suggests that a morally perfect being is one such that ‘it would not be possible for it to be exceeded’, morally speaking (ibid.). (TWD-E) entails that there is a possible world in which moral goodness is intrinsically valuable and some free creatures engage in it. Furthermore, Plantinga holds that there are possible worlds in which God creates beings with morally significant freedom who engage in moral goodness and never act wrongly (ibid., 44). Consider one such being; call it ‘Shmod’. Suppose that God and Shmod both exist, neither Shmod nor God performs any immoral actions, and Shmod engages in moral goodness, an intrinsically good sort of action, whereas God is never capable of engaging in that sort of intrinsically good action. Suppose also that Shmod is such that there is no possible situation in which he would act immorally (though there are many possible situations in which he is free to act immorally). Finally, suppose that, as (TWD-E) implies, the intrinsic goodness of Shmod's acts of moral goodness outweighs or defeats any intrinsic evils entailed by Shmod's acts of moral goodness. Given all of this, it is prima facie plausible that, all things considered, Shmod surpasses God, morally speaking. For while neither Shmod nor God does or would act immorally, Shmod does something God never does: he freely acts rightly when he was free to act wrongly, thereby making the world overall better than it would have been otherwise. At the very least, anyone who would maintain that God is not morally surpassed by Shmod has some explaining to do. It does not seem plausible to suggest, for example, that Shmod does not morally surpass God because the fact that Shmod is free to act immorally and God is not entails that Shmod's character has some sort of defect that is absent from God's character. The problem with that suggestion is that (TWD-E) entails that the intrinsic value of Shmod's acts of moral goodness outweighs or defeats the evil of any defects in Shmod's character entailed by his ability to act immorally. (TWD-E) implies that any such defects are for the greater good; how, then, could it also be true that the presence of such defects prevents Shmod from morally surpassing God? It appears that if God is to avoid being exceeded morally and qualify as maximally excellent in worlds containing creatures with morally significant freedom who engage in moral goodness and never act wrongly, then He, too, must engage in moral goodness in such worlds. Therefore:

That completes the first half of my argument; if I am right, then (TWD-E) together with various plausible/Plantingian claims entails (GG). I turn now to the second half of my argument, making the case that (G) together with various plausible/Plantingian claims entails the denial of (GG).

As I noted above, Plantinga holds that the proposition that God exists entails that He is a maximally excellent being in every possible world, and that entails that He is morally perfect in every possible world. Plantinga also holds that God's moral perfection entails that He never performs any morally wrong actions (Plantinga (Reference Plantinga1977), 91).Footnote ⁸ Thus:

1. (1) That God exists entails that He never performs any morally wrong actions.

As the preceding discussion indicates, Plantinga holds that if God exists, then He exists in every possible world – and so, naturally, Plantinga holds that God exists in every possible world (ibid., 112; Plantinga (Reference Plantinga1980), 140). Furthermore, that God exists is not something that He has even partial control over. He is, we might say, simply stuck with the fact that He exists – and necessarily so. His own existence is one of the necessary constraints within which He must work.Footnote ⁹ So, assuming the truth of (G), we have:

1. (2) Necessarily, that God exists is never even partially under God's control.

From (G), (1), and (2) we can derive:

1. (3) Necessarily: there is some true proposition, that God exists, such that (i) that God exists is never even partially under God's control and (ii) that God exists entails that God never performs a morally wrong act.

From (3) and (MSF) it follows that:

1. (4) Necessarily, God lacks morally significant freedom,

from which (together with Plantinga's definition of ‘moral goodness’) it follows that

1. (5) Necessarily, God is incapable of moral goodness.

And (5) is incompatible with (GG). Thus, from (G), (TWD-E), and various claims that are plausible and/or to which Plantinga appears to be committed we can derive a contradiction. This result presents at least a strong prima facie case that, given Plantinga's various commitments, it is impossible that (G) and (TWD-E) are both true. But Plantinga's free-will defence requires that the conjunction of (G) and (TWD-E) is possibly true. Strictly speaking, this argument does not establish that Plantinga's free-will defence on its own is internally inconsistent. But if the argument succeeds, it does show that the free-will defence is incompatible with a certain collection of claims that many theists – including Plantinga – find attractive.

In the next section I consider one interesting strategy for resisting my argument.

The ultimate source strategy

According to a recent proposal by Kevin Timpe, the essence of God's freedom is that He is the ultimate source of all of His own actions in that none of His actions are causally determined by factors over which He has no control (Timpe (Reference Timpe2012), 97 and Idem (2014), 104; see also Lehe (Reference Lehe1986), 322). God can freely do the right thing because that He always does the right thing is not causally determined by factors outside of His control and is the result of His own essential nature (Timpe (Reference Timpe2012), 98 and Idem (2014), 109). In fact, according to Timpe, the capacity for immoral action is (or at least indicates the presence of) a defect in one's character and hence maximal excellence is incompatible with the capacity for immoral action (Lehe (Reference Lehe1986), 321–322; Pawl & Timpe (Reference Pawl and Timpe2009), 415; Timpe (Reference Timpe2014), 113). On this view, God's morally right actions flow from His essential perfect character and are not causally determined by anything external to His ‘volitional structure’ (Timpe (Reference Timpe2014), 103), and hence those actions are free and God is maximally excellent in every possible world. Indeed, ‘God's freedom is the most perfect freedom there can be’ (ibid., 113).Footnote ¹⁰

Timpe's proposal suggests a strategy for answering my argument in the previous section. I should emphasize that I do not claim that Timpe would endorse this strategy; I claim only that it is inspired by some aspects of Timpe's work and is sufficiently interesting to merit consideration here. Recall that, according to Plantinga, to have morally significant freedom is to be in a situation in which it is true both that (a) you are free to perform a morally right action and (b) you are free to perform a morally wrong action (Plantinga (Reference Plantinga1977), 30, emphasis added). According to the Timpe-inspired ultimate source strategy, although there is no possible world in which God acts immorally He is nevertheless free to act immorally because His refraining from acting immorally is not causally determined by factors over which He has no control. God thus possesses morally significant freedom and is capable of engaging in moral goodness (again, despite the fact that there is no possible world in which He acts immorally). According to this strategy, (MSF) is false and hence the move from (3) to (4) in the preceding section fails.Footnote ¹¹

In the previous section, I identified a proposition that (i) is true if God exists, (ii) is entirely outside of God's control if God exists, and (iii) entails that God never acts immorally: that God exists. Thus, even if it is not the case that there are factors outside of God's control that causally determine that He always does the right thing, if God exists then there is a true proposition outside of God's control that entails that He always does the right thing. Consequently, the ultimate source response to my argument is plausible only if the following claim is plausible:

But it is hard to see any reason for thinking that (C) is plausible and easy to see a reason for thinking that it is not. One of the philosophers to whom Timpe looks for inspiration in developing the ultimate source proposal, Robert Kane, writes that in free actions ‘the ultimate control of our actions lie in us and not outside us in factors beyond our control’ (Kane (Reference Kane2005), 6, emphasis added). The italicized phrase is the crucial one here. It seems that the fundamental intuition that underlies the claim that causal determination by factors outside an agent's control eliminates freedom is something like the following: if, given some factor over which you never have any control, it's impossible that you do not do A, then you don't do A freely; whether the factor is inside or outside the agent is entirely irrelevant – what matters is whether the agent ever has control over the factor in question. And that intuition equally supports the claim that entailment by factors outside an agent's control eliminates freedom.Footnote ¹²

Thus, the defender of the ultimate source strategy faces the challenge of motivating an account of freedom according to which (i) God couldn't create beings that (at some time) possess morally significant freedom and also guarantee that those creatures never act wrongly and (ii) God is essentially morally perfect and yet freely acts morally.Footnote ¹³ If one rejects the claim that entailment by factors beyond the agent's control prevents an action from being free, then it is hard to see how (i) can be plausibly defended. On the other hand, if one accepts the claim that entailment by factors beyond the agent's control prevents an action from being free, then it is hard to see how one can consistently endorse (ii).