The Free Will Defense Refuted and God’s Existence Disproved


Raymond D. Bradley

1. The Down Under Logical Disproof of the Theist’s God

     1.1 Plantinga’s Attempted Refutation of the Logical Disproof

     1.2 Plantinga Refuted and God Disproved: A Preview

2. Plantinga’s Formal Presentation of his Free Will Defense

3. First Formal Flaw: A Non Sequitur Regarding the Consistency of (3) with (1)

4. Further Flaws Regarding the Joint Conditions of Consistency and Entailment

     4.1 A Non Sequitur Regarding the Entailment Condition

     4.2 Telling the Full Story in Order to Satisfy the Entailment Condition

     4.3 Two Non Sequiturs Involving the Consistency Condition

     4.4 The Russell Scenario: A Refutation by Logical Analogy

5. Plantinga’s Leibnizian Definition of Omnipotence

     5.1 Logical Consequences of the Definition

     5.2 Why Leibniz Didn’t Lapse

     5.3 A Fatal Equivocation Over the Scope of Transworld Depravity

     5.4 It Was All “Up to God”

     5.5 The Distinction Between Absolute and Consequential Modalities

     5.6 Plantinga’s Ploy

     5.7 Plantinga’s Modal Muddle

     5.8 Interlude on the Prevalence of the Modal Muddle

     5.9 Back to Plantinga’s Modal Muddle

     5.10 Interim Verdict

6. Judgment Day: God Almighty and All-Knowing Not Morally Perfect

     6.1 The Trial of Dog Almighty and All-Knowing

     6.2 The Trial of God Almighty and All-Knowing

7. Final Verdict: The Triumph of Down-Under Logic and Morals

     7.1 The Failure, and Irrelevance, of the Free Will Defense

     7.2 Evil and God’s Command Responsibility for its Occurrence

     7.3 A Formal Proof that if Evil Exists then the Theists’ God Does Not

     7.4 Conclusion

1. The Down Under Logical Disproof of the Theist’s God

More than half a century ago, John L. Mackie laid the charge.[1] Forcefully restating a problem that has troubled theists at least since The Book of Job, he argued in 1955 that the proposition

(1) God is omniscient, omnipotent, and wholly good.

cannot logically be reconciled with the proposition

(2) There is evil.

and that, since (2) is undeniably true, it follows that (1) is false. Two other Down Under (Australian or New Zealand) philosophers followed. H. J. McCloskey published another strong case in 1960.[2] Then, in 1967, I published “A Proof of Atheism” using an argument that sidelined the idea of free will.[3]

1.1 Plantinga’s Attempted Refutation of the Logical Disproof

Enter Alvin Plantinga, apologist in chief for the theist’s God. First in God and Other Minds (1967), then in The Nature of Necessity (1974), Plantinga spelled out his so-called Free Will Defense. He claimed it be a type of consistency proof that would prevail for all time over any such further attempts to disprove God’s existence. Not only, in his view, is it “extremely difficult” for anyone produce a proposition (or set of propositions) whose conjunction with (1) and (2) produces a formal contradiction. He thinks he has shown it to be impossible.

The myth of his rebuttal’s impregnable status has been growing for well over thirty years, even among many philosophers of an atheistic bent.

I propose to put an end to that.

1.2 Plantinga Refuted and God Disproved: A Preview

In sections 2 through 5, I’ll show that Plantinga’s Free Will Defense contains, as Hume would say, “nothing but sophistry and illusion.”

On the face of it, it is God’s perfect goodness rather than his omniscience or omnipotence that is compromised by the presence of evil in the universe he purportedly created. Thus it is natural for even the philosophically untutored to reason thus

(a) that being omniscient, God would have known down to the very last detail all the evils (natural as well as moral) that would bedevil the world he planned to create, including all the evils his creatures would bring about;

(b) that being omnipotent, God need not have created that world but could have chosen to create one containing no evil whatever; and

(c) that by virtue of his failure to exercise that option, God should be held responsible for every evil that exists in the world he did create.


Since evil exists, an omniscient, omnipotent God who is also perfectly good, does not.

Plantinga, however, ignores clauses (a) and (c), and targets only clause (b), that involving God’s omnipotence. He sketches a scenario according to which God did his best to create a world without evil but had his plans thwarted by the freedom-abusing creatures he had created. “Given these conditions,” he argues, God could not have created a world free of evil. This “despite” his omnipotence. True, moral and natural evil exists. But that’s up to us, and Satan, respectively. It isn’t “up to God.” So Plantinga claims.

But Plantinga’s Defense, I show, is a failure. Not only is his argument formally flawed. It is made plausible only by virtue of his well-concealed failure to abide by his own definition of omnipotence. His scenario shows that God “couldn’t” have created a world without evil only in the sense that his inability to do so is a consequence of his own advance plans for its creation. It doesn’t show that he couldn’t have created such a world had he availed himself of the full resources of his omnipotence. Plantinga’s reasoning to the contrary involves a logical muddle about the modal notions of possibility and impossibility.

Having effected a major breach in the logic of Plantinga’s Free Will Defense, I then–in sections 6 and 7–present a new version of the Down Under Disproof by focusing once more on clauses (a) and (c): those to do with God’s omniscience and moral goodness, respectively.

Thus I produce a proposition of the kind that Plantinga thinks it “extremely difficult to find”: a proposition, that is, whose conjunction with (1) and (2) produces an explicit contradiction. It is a proposition so fundamental to both morals and international law that allowing any exception would absolve many of the greatest criminals of human history from responsibility for evils committed under their command. God can no more be excused than can they. It is a proposition from which, in conjunction with the uncontested fact of the existence of evil, we can deduce that–irrespective of whether or not we have free will–any being who is both omnipotent and omniscient should be held responsible for all the evil that exists. Hence from the existence of evil, together with God’s supposed omnipotence and omniscience, we can deduce that the theist’s God is not the morally perfect being theists define him as being. Ipso facto, their God does not exist.

2. Plantinga’s Formal Presentation of his Free Will Defense

Plantinga presents his Defense most persuasively in his Chapter IX of The Nature of Necessity.[4] Announcing it as “The Free Will Defence Triumphant,” Plantinga writes:

Put formally, you remember, the Free Will Defender’s project was to show that

(1) God is omniscient, omnipotent, and wholly good.

is consistent with

(2) There is evil.

by employing the truth that a pair of propositions p and q are jointly consistent if there is a proposition r whose conjunction with p is consistent and entails q. What we have just seen is that

[(3)][5] Every essence suffers from transworld depravity.

is consistent with God’s omnipotence. But then it is clearly consistent with (1). So we can use it to show that (1) is consistent with (2). For consider the conjunction of (1), (3), and

[(4)] God actualizes a world containing moral good.

This conjunction is evidently consistent. But it entails

(2) There is evil.

Accordingly (1) is consistent with (2); the Free Will Defence is successful.

Come again? For what purports to be a rigorous formal consistency proof, this has got to be a joke. Even a first year logic student would be warranted in describing it as “bluster and bombast” to use Plantinga’s unwarranted description of Richard Dawkins’ The God Delusion.[6]

So what’s wrong with it? There’s nothing wrong with the overall strategy of trying to find a proposition (or set of propositions) r whose conjunction with a proposition p is consistent and entails q. That is indeed a way of proving that p is consistent with q. The problems lie, not in his strategy, but in its execution.

The demands placed on any argument that purports to be sound, especially one that is supposed to constitute a formal proof, are quite exacting. First, it must be formally valid: the conclusion has to follow logically from the premises as stated without depending on premises that have been concealed, taken for granted, simply because they were thought intuitively obvious.[7] Second, its premises must be true. And third, it must not contain any informal fallacies of equivocation in which one and the same term is used with different meanings.

Plantinga’s purported consistency proof fails on each of these three counts.

3. First Formal Flaw: A Non Sequitur Regarding the Consistency of (3) with (1)

First, Plantinga claims that proposition (3) (that every essence suffers from transworld depravity) is consistent with God’s omnipotence and then goes on to claim “But then it is clearly consistent with (1).”

Yet that does not follow. For consider what (1) states. It states that God is not only omnipotent but also omniscient and wholly good. Now it is true that the simple conjunction of (3) with the claim that God is omnipotent does not generate a formal contradiction. But that doesn’t mean that (3) is also consistent with God’s omniscience and his perfect goodness as well.[8] This is sloppy, patently invalid, reasoning.

4. Further Flaws Regarding the Joint Conditions of Consistency and Entailment

Second, he claims that the conjunction of (1), (3) and (4) satisfies two conditions: it both entails (2) and is “evidently consistent.” That is why, in his view, the conjunction of (3) and (4) plays the role of r in the proof schema he is following.

4.1 A Non Sequitur Regarding the Entailment Condition

Consider the first of these two conditions: that of entailing the existence of evil. Formally, the conjunction of (1), (3) and (4) falls far short of entailing (2). After all, it is a logical rule of thumb that in a formal proof you can’t get more in your conclusion than you’ve explicitly put into your premises.[9] One can’t, therefore, formally derive the existence of evil from premises that don’t even mention evil. We need an explicit account of how evil gets into the act. Only when the missing premises are brought into the open can we determine whether, when added as further conjuncts to the conjunction of (1), (3) and (4), we have a proposition or set of propositions that plays the role of r by satisfying both the entailment condition and the consistency condition.

Remember that Plantinga is acting as defense counsel for God. It is easy for him to tell a story on God’s behalf that will seem consistent if he confines himself to the consistency of a restricted subset of the propositions involved in the story. But how about the logical situation if he were forced to tell the story in full; if he were, so to speak, to tell the truth, the whole truth, and nothing but the truth? We need to listen to the whole story, especially that part which is supposed to satisfy the entailment condition, in order to judge whether the case for the defense hangs together in the way Plantinga says it does.

4.2 Telling the Full Story in Order to Satisfy the Entailment Condition

Let’s start, provisionally, with the proposition that Plantinga describes as being “the heart of the Free Will Defence,” namely,

(5) God could not have created a universe containing moral good (or as much good as this one contains) without creating one containing moral evil. (p. 167)

This helps to fill one of the gaps in his argument since, if taken together with

(4) God actualizes a world containing moral good.

it does in fact imply

(2) There is evil.

So now we know how evil got into the act. It got there because God created a world containing it.

But how does Plantinga get to (5)? His appeal to

(3) Every essence suffers from transworld depravity.

tells part of the story. Yet only part. And the part it plays is not the one Plantinga assigns it in his formal presentation. Plantinga claims that from (1), (3) and (4) taken together one can formally derive (2). But clearly that’s wrong since, once again, there’s no explicit mention of evil in any of them. So, by the logical rule of thumb I invoked earlier, this inference is unwarranted. It is premises (5) and (4) that entail (2). Premise (3) does indeed play a role in Plantinga’s argument. But that comes earlier, as it were, as one of the premises he needs in order to establish (5). So we need to back up still further to see how he gets to (5).

The needed premises lie scattered throughout his informal presentation. I’ll state them in his words and number them (6) through (10).

Let’s start with the story of God’s rationale for creating the kind of world in which we find ourselves.

(6) “God thought it good to create free persons” (p. 170) [because] “[a] world containing creatures who are sometimes significantly free (and [who] freely perform more good actions than evil ones) is more valuable, all else being equal, than a world containing no free creatures at all.” (p. 166)

Plantinga gives no argument for (6). He seems to think that it states an obvious truth about which possible worlds would be most valuable for both God and us. Yet in God’s scheme of things, the existence of significantly free creatures who can abuse their free will ranks highest in his order of priorities; higher than their health and happiness, for instance. Plantinga even goes so far as to state that “the price for creating a world in which such persons produce moral good is creating one in which they also produce moral evil.” (p. 186f). No mention here of the fact that it is we, not God, who pay the price. No moral qualms on that score. An unfazed anticipation of Madeleine Albright’s chilling remark, “We think the price is worth it,” when confronted with the fact that the sanctions imposed on Iraq during the 90s had caused the death of over half a million children.[10]

Plantinga provides his Defense with a generality it would otherwise lack when he adds that the free creatures God had in mind weren’t limited to humans. As he puts it, “possibly natural evil is due to the free activity of a set of non-human persons.” (p. 192) This move enables Plantinga to treat both moral and natural evil as “special cases of what we might call broadly moral evil–evil resulting from the free actions of personal beings, whether human or not.” (p. 193) The sort of nonhuman personal beings he has in mind are those to whom St. Augustine attributed natural evil, viz., “Satan and his cohorts.” (p. 192) The Satanic hordes feature importantly in that branch of orthodox Christian theology known as demonology. But presumably Plantinga would not exclude the angelic choir from his list of significantly free persons.

All God’s creature, according to Plantinga, are significantly free, where “significant freedom” is to be construed in contracausal terms. On his account, free will is incompatible with any kind of causality, natural or supernatural:

If a person S is significantly free with respect to a given action, then he is free to perform that action and free to refrain; no causal laws and antecedent conditions determine either that he will perform the action, or that he will not. (p. 165f)

And, singling out God in particular, he writes

if I am free with respect to an action A, then God does not bring it about or cause it to be the case either that I take or that I refrain from this action; he neither causes this to be so through the laws he establishes, nor by direct intervention, nor in any other way. (p. 171)

I am free with respect to action A if I bring it about that A occurs and no one and nothing brings it about that I bring it about that A occurs. This sort of free will is commonly called “libertarian free will.” It is supposed to be “buck-stopping” freedom: the kind of freedom that, since it has no causal antecedents, allows attributions of responsibility for every evil to be leveled against us–not God–for whatever we freely bring about. Or so Plantinga argues.

Plantinga’s next premise is

(7) God did in fact create significantly free creatures. (p. 167)

This, too, he takes to be a simple matter of fact.

Then comes

(8) Since every person is the instantiation of an essence, he can create significantly free persons only by instantiating some creaturely essences. (p. 188)

Given Plantinga’s metaphysical view, according to which a potential person (that person’s “essence”) is an abstract entity, he can plausibly argue that (8) follows from (7).

It is at this point that (3) assumes its role in Plantinga’s overall proof. Referring to the essences mentioned in (8), he tells us

(3) Every essence suffers from transworld depravity. (p. 189)

However, there’s a crucial ambiguity in (3). Does he intend us to understand it as true in all possible worlds, i.e., as applying to all possible creaturely essences? Or does he intend us to understand it as true only in the actual world, the world that God chose to create? On either interpretation, as we’ll see, premise (3) generates huge problems for Plantinga’s Defense. It makes his argument unsound.

But on with the show. All Plantinga needs now is one more premise, namely,

(9) If every such essence suffers from transworld depravity, then no matter which essences God instantiated, the resulting persons [whether human or nonhuman], if free with respect to morally significant actions, would always perform at least some morally wrong actions. (p. 189)

from which he obtains its generalization

(10) If every creaturely essence suffers from transworld depravity, then it was beyond the power of God himself to create a world containing moral good but no moral evil. (p. 189)

Given Plantinga’s definition of “transworld depravity,” both (9) and (10) are supposed to follow from the conjunction of (7), (8), and (3).

None of the propositions (6) through (10) features in Plantinga’s presentation of his formal proof. But now we have them in hand it is clear how we are supposed to fill the gap leading to the previously cited

(5) God could not have created a universe containing moral good (or as much good as this one contains) without creating moral evil. (p. 167)

For (5) follows transparently from (3) and (10). And, as previously noted, from (5) together with

(4) God actualizes a world containing moral good.

we can indeed formally derive

(2) There is evil.

Plantinga had claimed that we could deduce (2) from (3) and (4) alone. He was wrong. We can deduce (2) from the conjunction of (3) and (4) of his premises only if we expand this conjunction to include propositions (5) through (10) as well.

Recall, however, that there is a second condition that must also be satisfied if this expanded conjunction is to play the desired role of r in Plantinga’s consistency proof. Not only must it entail (2); it must be consistent with (1).

4.3 Two Non Sequiturs Involving the Consistency Condition

The problem is that, once we’ve told the full story so as to satisfy the entailment condition, it is far from clear that the expanded conjunction of (1) and (3) through (10) is consistent.

First, it isn’t evident that all the propositions in the conjunction of (3) through (10) are consistent even with God’s omnipotence. The problem isn’t hard to see. Plantinga himself defines God’s omnipotence as the absence of any nonlogical limits to God’s power, meaning that God’s power to act or refrain from acting is constrained only by the laws of logic. So consider premise

(5) God could not have created a universe containing moral good (or as much good as this one contains) without creating one containing moral evil. (p. 167)

This, remember, is one of the conjuncts in the expanded conjunction, the conjunction that is supposed not only to entail (2) but also to be consistent with God’s omnipotence. Yet (5) asserts that God is unable to do something logically possible. And on the face of it, this is not consistent with his omnipotence. For what are the laws of logic that prevent God from creating a morally perfect world?

Second, even if we turn a blind eye and suppose, for the sake of argument, that (5) is consistent with God’s omnipotence, Plantinga’s argument would still fall far short of demonstrating that the whole expanded conjunction is consistent also with God’s omniscience, and total goodness. Once more his Defense, as presented, involves a non sequitur. He had previously claimed that because (3) was consistent with God’s omnipotence, it was also consistent with all the attributes listed in (1). In order for the consistency condition to be satisfied, he’d have to say the same about (5).

4.4 The Russell Scenario: A Refutation by Logical Analogy

Consider what I call “the Russell scenario.” In his essay “A Free Man’s Worship,” Bertrand Russell imagines Mephistopheles telling the story of creation thus:

The endless praises of the choirs of angels had begun to grow wearisome; for, after all, did he not deserve their praises? Had he not given them endless joy? Would it not be much more amusing to obtain undeserved praise, to be worshipped by beings whom he tortured? He smiled inwardly, and resolved that the great drama should be performed.[11]

Simply add the premise that gratuitous torture is evil and the Russell scenario all by itself entitles us formally to derive the truth of (2), the existence of evil, and thereby satisfy the entailment condition. What is more, the Russell scenario is also consistent with God’s omnipotence. Recall, then, that Plantinga had said that (3)–the proposition that every essence suffers from transworld depravity–is “consistent with God’s omnipotence,” and had gone on in the very next sentence to claim, “But then it is clearly consistent with (1),” viz., with the proposition that God is also omniscient and wholly good. If the consistency of some proposition p with God’s omnipotence sufficed to establish its consistency with God’s omniscience and total goodness as well, then the consistency of Russell’s scenario with God’s omnipotence, taken together with the fact that it entails evil, would also suffice to establish its consistency with his omniscience and total goodness. Yet the God of Russell’s scenario is clearly evil, not good.

What has gone wrong here? The answer, as I said before and as Russell’s scenario so vividly demonstrates, is that from the consistency of some proposition or set of propositions p with God’s omnipotence, it does not follow that p is consistent with the full constellation of God’s attributes as stated in (1).

5. Plantinga’s Leibnizian Definition of Omnipotence

To understand exactly what is going let’s consider the concept of omnipotence more carefully. Plantinga defines omnipotence thus:

the theist says that God is omnipotent–which means, roughly, that there are no non-logical limits to his power. (p. 167).

The implications of Plantinga’s definition should be obvious. It allows that there are indeed certain logical limits to God’s power. It is logically impossible for him to make a necessarily true proposition false, and logically impossible for him to make a necessarily false proposition true. But that’s the full extent of the logical limits to God power. Since contingent propositions are neither necessarily true nor necessarily false, i.e., are both possibly false and possibly true, there are no logical limits to God’s capacity to make any contingent proposition true or to make it false by bringing about the state of affairs that would make it true or make it false, as the case may be.

And since (by definition) there are no nonlogical limits to his power either, it follows that within the realm of the contingent, God’s power to choose what states of affairs to actualize or refrain from actualizing is absolute. If God chooses to place certain nonlogical limits on his omnipotence–by choosing to create only a certain kind of world or by delegating some of his powers to other beings, for instance–that’s entirely up to him. And if he subsequently can’t do certain things as a consequence of those choices, that too is up to him. But he cannot then claim, nor can Plantinga justly claim on his behalf, that he really couldn’t have done otherwise in the absolute sense attributed to him by the definition of omnipotence.

5.1 Logical Consequences of the Definition

There should be no dispute about it. From the definition of omnipotence it follows that in an absolute sense of “can,” God can actualize any contingent state of affairs that he chooses. Plantinga’s Down Under critics have only to add that since there is no contradiction involved in the conception of a world in which it is contingently the case that there are significantly free creatures who always do what is right, it follows that, in the omnipotence-relevant sense of “could,” God could have created just such a world instead of the evil-containing world in which we find ourselves. This is why Mackie, for instance, asserts

(11) God could have actualized worlds in which significantly free persons always do what is right.

and deduces, contrary to Plantinga’s (5),

(12) God could have actualized a possible world in which evil does not exist.

If Mackie is right about (11) and (12), Plantinga is wrong about (5), the “heart” of his Free Will Defense.

Both Leibniz and Mackie held that there are many possible worlds in which evil does not exist. As Plantinga himself observes, Leibniz therefore reasoned that since God is wholly good and would have created the best world he could have created, the actual world must be the best of all possible worlds despite the evil it contains, while Mackie reasoned that since the actual world is clearly not as good as one in which there is no evil whatever, God cannot be as good as his supporters make him out to be.

5.2 Why Leibniz Didn’t Lapse

Plantinga, however, claims that both Leibniz and Mackie made a mistake. Mackie, he allows, is “right in holding that there are many possible worlds containing moral good but no moral evil.” (p. 184) But then, he says, it does not “follow” that God could have actualized any of those worlds. To suppose otherwise, he claims, is to commit what he calls “Leibniz’s Lapse.” (p. 184)

As he sees it both Leibniz and Mackie have overlooked the fact that under certain conditions it was impossible for God to create a world in which there is good but no evil. After specifying those conditions, he asserts

Under these conditions God could have created a world containing no moral evil only by creating one without significantly free persons. (p. 189)

Well, what are these conditions, the ones that are supposed to make (5) true? The ones that do the work in his argument turn out to be those spelled out in premises (6), (7), (3) and (4). Yet all these are nonlogical conditions, conditions that are imposed on God not by logic but by himself.[12]

Consider them in turn.

Premise (6) tells us that God thought that a world containing significantly free creatures would be better “all else being equal” than a world lacking free creatures. But no logical considerations compelled him to think that way. Logically speaking he could have come to a different conclusion, especially in light of the fact that, by virtue of his omniscience, he knew that the ceteris paribus clause would not be satisfied. There is no moral equivalency between, on the one hand, a world in which we suffer from every evil known to mankind, and, on the other hand, a “heavenly” world that is free from all evils. Most of the victims of his choice would willingly sacrifice the supposed benefits of libertarian freedom in order to avoid the hellish conditions that prevail in much of the world. Although there are possible worlds in which (6) true, there are also ones in which it is false. God was under no logical constraints to make it true. Premise (6) merely tells about certain nonlogical conditions that God imposed on himself.

Likewise with premise (7), the statement that God in fact created significantly free creatures. Clearly (7) could have been false. Otherwise God wouldn’t have had the choice that (6) envisages him as having. So, if God did what (7) says he did, it can only be because he freely chose to make it true despite knowing the consequences of his choice.

5.3 A Fatal Equivocation Over the Scope of Transworld Depravity

How about premise (3), the claim that all God’s creatures suffer from transworld depravity? Couldn’t God have actualized some other creatures? Plantinga’s response is curious: “Perhaps. But then again, perhaps not.” (p. 187)

But what does Plantinga mean when he says “perhaps”? Is he saying that it is logically possible that God could have actualized creatures whose essences aren’t depraved? It would seem so. For later, when he discusses his favorite candidate for someone suffering transworld depravity, Curley Smith, the hypothetical mayor of Boston, he writes

Note that transworld depravity is an accidental property of those essences and persons it afflicts. For suppose Curley suffers from transworld depravity; then so does his essence. There is a world, however, in which Curley is significantly free but always does what is right. If that world had been actual, then of course neither Curley nor his essence would have suffered from transworld depravity. (p. 188)

Here Plantinga is implicitly acknowledging that his own premise (3), the thesis of transworld depravity, is a contingent proposition, both possibly true and possibly false. That is to say, he is acknowledging that there are indeed some persons like Curley who, in at least one nonactual world, always do what is right because they possess saintly essences. As he says, “If that world had been actual, then of course neither Curley nor his essence would have suffered from transworld depravity.” But this means that (3) doesn’t impose a logical constraint on God’s omnipotence after all. So we’re back to square one as it were: Why couldn’t God have actualized some such world? Why does Plantinga, say that “perhaps” God couldn’t have done so?

Plantinga’s discussion of transworld depravity is stunning in its audacity and fallaciousness. At one point he says, without fear of contradiction from his critics,

It is possible that every creaturely essence suffers from transworld depravity. (p. 188)

There’s no problem with this since here he is claiming only that it is there is at least one possible world in which (3) is true.[13]

But then, without notice, he switches to the supposition that the possible world he is talking about is the actual world, and writes

If this possibility were actual, then God could not have created any of the possible worlds that include the existence and freedom of just the persons who do in fact exist, and also contain moral good but no moral evil. (p. 186, my emphases)[14]

Here he invites us to share his supposition that the possible world in which (3) is true is the actual world, the possible world that contains actual persons. However, if the scope of “every” in (3) is thus taken to be only the essences that are instantiated in the actual world, then (3) is true only because God chose to make it so, not because he was logically obliged to make it so.[15]

If, on the other hand, the scope of “every” is taken to be essences in all possible worlds, then (3) is neither true nor even possibly true. It isn’t true since there are possible worlds in which creaturely essences don’t suffer from depravity. Worse still, (3) isn’t even possibly true since it excludes the logical possibility of saintly and angelic essences. Mackie had pointed out in (12) that it is logically possible that there be worlds in which significantly free person always do what is right. And Plantinga had concurred. So if (3) were taken to deny such a logical possibility, (3) would be necessarily false.

Plantinga’s equivocation on the scope of “every” in (3) gives the game away. For the question was why God would choose to actualize a world containing depraved creatures when he didn’t have to. It won’t do to answer that such world is the one that he did in fact actualize. For the actual world is actual only because God chose it from the set of all possible worlds as the one he would actualize. The dispute between Mackie and Plantinga is over what worlds God could have created, not what world he did create. The universality of transworld depravity in the actual world doesn’t show that God found it logically impossible to actualize a possible world free from such depravity. Given his omnipotence, God could have created a world containing moral good but no moral evil. So Plantinga’s premises (6) through (8) could well be true without it being true that evil exists. It is his premise (3) that creates the problem. A problem for Plantinga, that is, not for Leibniz or Mackie.

My point can be presented in the form of a dilemma. If the scope of the universal quantifier “every,” as it occurs in (3), were taken to be creaturely essences in every possible world, then (3) would deny the logical possibility of worlds in which humans like Curley lack transworld depravity. Yet Plantinga himself allows that such worlds are indeed logically possible. So, since whatever is logically possible is necessarily possible, to deny that such a world is possible would be to assert something necessarily false. And if premise (3) were false (let alone necessarily false) then Plantinga’s argument would be unsound. But if, on the other hand, the scope of “every” is taken to be the creaturely essences that exist in the actual world, then (3) leaves open the possibility of God actualizing a world containing humans like a virtuous Curley. But in that case Plantinga’s contrary claim, in (5), that it is “possible” that God could not have created a world containing moral good but no moral evil, is false and his argument shown once again to be unsound. On either interpretation of (3), Plantinga’s argument is unsound.

5.4 It Was All “Up to God”

In any case, God was not under any logical compulsion to instantiate any essences at all as (4) presupposes he did. By definition, his omnipotence is curtailed only by logic, not by his own choices as to which kind of world to create or whether to create any world at all. Plantinga effectively concedes the point when he tells us

Of course it is up to God whether to create free creatures at all; but if he aims to produce moral good, then he must create significantly free creatures upon whose co-operation he must depend. (p. 190, my emphases)

And he continues with the explicit acknowledgement that in so doing God is placing himself under limitations that logic itself does not dictate:

Thus is the power of an omnipotent God limited by the freedom he confers upon his creatures. (p. 190)

Confers without our consent or approval, of course.

5.5 The Distinction Between Absolute and Consequential Modalities

By now it should be clear that the expression “could not,” as it occurs in (5), conveys nothing more than that (5) is a logical consequence of the self-imposed, and therefore nonlogical, conditions specified in (3), (4), (6) and (7) in particular. We need to distinguish, therefore, between two very different senses of modal expressions like “could” and “could not,” “possible” and “impossible,” and their cognates. In the absolute sense of the word, God’s omnipotence consists in there being no nonlogical limits to his powers. In the absolute sense of the words “could not,” as they occur in (5), (5) is not just false but necessarily false since there are no nonlogical limits on God’s ability to create a world containing moral good but no moral evil. The only sense in which (5) is true is that in which the expression “could not” is used in what I shall call the consequential sense, that in which his inability to create a world without evil is a consequence of his curtailing his absolute powers by placing himself under certain nonlogical constraints.

But there’s a problem here. On the consequential interpretation, Plantinga’s argument does nothing whatever to address the Down Under objection that if God had availed himself of the absolute powers that, by definition, his omnipotence confers on him, he “could” have created a world containing moral good but no moral evil. Being consequentially impossible is compatible with being absolutely possible. Hence, Plantinga’s argument for the former does nothing to refute Mackie’s argument for the latter.

5.6 Plantinga’s Ploy

In order to sustain his charge that Leibniz had lapsed, Plantinga has to resort to what I call Plantinga’s Ploy: that of equivocating on the notions of possibility and impossibility.[16] He wants us to read “could not” as it occurs in (5) in its consequential sense so as to make (5) consistent with God’s omnipotence–as it has to be in order for his proof strategy to be successful. At the same time, he wants us to read this expression as it occurs in (5) in its absolute sense so as to make (5) contrary to one of the implications of God’s omnipotence–as it has to be if it he is to refute the Down Under Disproof. But he can’t have it both ways. No proposition can be both consistent with, and a contrary of, another. His Free Will Defense is a logical fraud perpetrated on the unsuspecting by playing fast and loose with words. It wasn’t Leibniz whose logic lapsed. It was Plantinga. He didn’t stick with the definition of omnipotence to which he explicitly subscribed.

5.7 Plantinga’s Modal Muddle

Is Plantinga’s Ploy a case of deliberate disingenuousness–the sort of thing one expects from an overzealous defense counsel? Mimicking Plantinga himself, one might say, “Perhaps. But then again, perhaps not.” Perhaps his usual astuteness had abandoned him. Perhaps Plantinga simply failed to perceive the distinction between consequential and absolute uses of such modal terms as “possible,” “impossible” and their cognates: expressions such as “can,” “could,” “cannot,” and “could not.” Perhaps he should be seen as joining ranks with countless other metaphysicians who have committed a fallacy so endemic that it deserves to be called ‘the’ Modal Muddle.

5.8 Interlude on the Prevalence of the Modal Muddle

A case in point is all those who suppose that the necessary truth of the statement “The future will be what it will be” commits us to believing that the future must be what it is going to be and it is impossible for us to divert the future from its predetermined course. They suppose that logic itself commits us to fatalism.

On analysis, their reasoning goes like this. Consider the proposition

(13) If P then P.

where P is a contingent proposition such as Aristotle’s “A sea battle will occur in the Bay of Salamis.” Since (13) is a truth of logic, and hence necessarily true, it is also true that

(14) It is necessary that if P then P.

In (14) the modal property of being necessarily true is attributed to (13), and the expression “necessary” is being used in the absolute sense to mean that there are no logically possible conditions under which (13) is false. Now (14) lends itself to being expressed by sentences such as

(15) “If P then it is necessary that P.”

and its syntactic equivalent

(16) “If P then it is impossible that not-P.”

But in (15) and (16) we have a potential source of logical confusion. On the one hand, we can think of each as merely expressing (14) in other words. And in that case nothing remotely fatalistic even seems to follow from the necessary truth with which we started. But on the other hand, we can erroneously think of (15) and (16) as attributing absolute necessity or impossibility to the consequent clause or its denial, respectively.

That’s the fallacy committed by many metaphysicians when discussing Aristotle’s problem of future contingents. Aristotle had posed the question whether, if it is true that a sea battle is going to occur in the Bay of Salamis, it follows that such a sea battle must occur, and cannot but occur. To answer “Yes” would seem to commit one to saying that the logical truth of (13), as stated in (14), entails that the future is fated and that there is nothing one can do about it. It is to suppose, as I once put it, that logical determinism–the logical truth of (13)–entails logical fatalism.[17] But, of course, logic itself does not dictate that the proposition P, as it occurs in the consequent clause of (15) and (16) is itself “necessarily true” or that its denial, not-P, is “not possibly true” or “impossible.” These modal expressions, as they occur in the consequents of (15) and (16), should not be understood in an absolute sense, but in a consequential sense. For the proposition P, remember, is a contingent proposition and hence not necessarily true and not such that its denial is impossible. That is to say, because P–by hypothesis–is contingent, it could be false (where “could” is to be understood in the absolute sense). To suppose that P can’t be false on the basis of the infelicitously expressed sentences (15) and (16) is to confuse the consequential uses of these modal expressions with their absolute uses. It is to be guilty of The Modal Muddle. All that follows from, is entailed by, the truth of the proposition that a sea battle will occur is that it will occur, not that it “must” occur or that its nonoccurrence is “impossible.”

A parallel modal fallacy infects the thinking of those who suppose that the doctrine of theological fatalism or predestination follows from the doctrine of God’s omniscience. The doctrine of omniscience says that for any true proposition P God knows “from all eternity” that P. Now consider any proposition of the form

(17) God knows that P.

where P is a (supposedly) true contingent proposition such as “Judas will betray Jesus.” Then since God’s knowing P to be true entails the truth of P, both the following propositions will be true:

(18) If God knows that P then P.


(19) Necessarily if God knows that P, then P.

Now consider the sentences

(20) “If God knows that P then necessarily P.”


(21) “If God knows that P then it is impossible that not-P.”

We can construe the sentences (20) and (21) innocuously as expressing nothing more than (19). And when we so construe them, they carry with them no hint of the idea that God somehow predestined Judas’s betrayal. Yet we can construe these same sentences as lending support to the doctrine of predestination.

Thus some theologians have thought that God’s foreknowledge of future events, such as the supposed betrayal of Jesus by Judas, necessitates the occurrence of those events and makes their nonoccurrence impossible. They have thought that God’s foreknowledge of what Judas would do made Judas a pawn in God’s plan for salvation. More generally, they conclude, God’s foreknowledge of the future makes the future inevitable; his foreknowledge determines not just Judas’s future but also predestines the futures of all of us. Yet it is easy to see that this account of the relationship between foreknowledge and predestination gets the logic of the situation wrong.[18] Like the argument for logical fatalism this argument for foreordination falls foul of The Modal Muddle: that of supposing that the consequent clause all by itself is necessary (as in (20)) or its denial impossible (as in (21)) when, as a matter of logic, each is contingent.

5.9 Back to Plantinga’s Modal Muddle

In general terms, what I am calling The Modal Muddle is that of attributing to the consequent Q of a necessarily true proposition having the form “If P (the antecedent) then Q (the consequent)” a modal property that it does not have. In logical contexts, it arises when a sentence of the form “If P then necessarily Q” is taken to attribute necessity to the consequent when the consequent is not necessary but contingent and hence possibly false; or again when a sentence of the form “If P then it is impossible that not-Q” is taken to attribute impossibility to the denial of the consequent when the denial of the consequent is not impossible but contingent and hence possibly true.[19] The consequent will, of course, be consequentially necessary in the sense of being a logically necessary consequence of the antecedent premises from which it is derived; and the denial of that consequent will, correspondingly, be consequentially impossible. But it does not follow that the consequent is absolutely necessary or its denial absolutely impossible. To suppose that it does follow is to be guilty of The Modal Muddle and to violate what has aptly been called “The Principle of the Fixity of Modal Status”: in particular, the metalogical principle that no contingent proposition ever “becomes” necessary or impossible just by virtue of appearing in the conclusion of a valid argument.[20]

By now it should be clear that Plantinga’s Free Will Defense is bedeviled by The Modal Muddle. It has the general form of an argument from

(22) Necessarily, if God actualizes a world in which propositions (3) through (10) are true, then evil exists.


(23) If God actualizes a world in which propositions (3) through (10) are true, then it is impossible that evil not exist.

For the purposes of his Defense, Plantinga wants us to conclude that the impossibility of evil’s nonoccurrence, as asserted in the consequent of (23), is absolute. But all he has shown is that the occurrence of evil is a logical consequence of God’s actions.

5.10 Interim Verdict

Summing up so far: Leibniz didn’t lapse; neither did Mackie. Either Plantinga indulged in deliberate Ploy or he fell foul of the Modal Muddle.

Even when charitably reconstructed, the most that Plantinga’s argument shows is that God couldn’t have done otherwise in the sense that he is unable to evade the logical consequences of the nonlogical conditions he imposed on himself by exercising his own free choices.

But that doesn’t absolve God of responsibility for the consequences of those choices.

With the collapse of Plantinga’s “he couldn’t have done otherwise” defense, the issue of God’s culpability comes sharply into focus. What are we to say of a hypothetical God who wantonly inflicts every evil known to humanity on his victims when he could have chosen otherwise? What are we to say of a God who fits the Russell scenario to a tee?

6. Judgment Day: God Almighty and All-Knowing Not Morally Perfect

In order to guide our intuitions, I will tell a parable about another fictional character: God’s earthly analogue, Dog.

6.1 The Trial of Dog Almighty and All-Knowing

“Beware of the Dog,” said his neighbors. They were referring to Dog Almighty and All-Knowing.

Dog, you see, was a dog breeder who reigned supreme on an island of his own making. Despite his name, he wasn’t really almighty. Not in the sense in which God was almighty, anyway. His powers were limited not only by the laws of logic, but also by those of nature. Within those bounds, however, he had the ability to do anything physically possible. And he wasn’t all-knowing, either. But he did know enough of the laws of nature and their application to be able to predict, without fail, what the future would be.

One day Dog had an idea. He’d create a breed of dogs in his own image and give them the precious gift of freedom. For, as his defense counsel said, when Dog went on trial for gross cruelty to animals, Dog had judged it better, “all else being equal,” to have his island inhabited by free dogs rather than by no free dogs at all.

Now Dog had a choice. The only way he could create his chosen breed of dogs was to animate some of the frozen embryos (the earthly correlate of Plantinga’s abstract essences) that were stored in his cryonic freezer. Some of these, he knew, were embryos of a gentle breed that would, if animated, never behave badly. On the other hand, some were of a breed of savage dogs whose actualization, he knew in advance, would lead to the most horrific outcomes he could envisage. Super pit bulls, they were. And among them a giant of a super pit bull, Satan, a dog that would kill and maim many lesser dogs and infect them with his own evil disposition.

Which breed would he choose? For some unexplained reason, Dog chose to animate the savage breed and leave them significantly free, unleashed, unmuzzled, and undisciplined. The outcome was predictable. Indeed he knew what would happen. Satan and his cohorts made the life of less powerful dogs, and other creatures on the island, a living hell. And in these conditions, these weaker dogs did unto others as others had done unto them. Starvation and suffering, murder and mayhem, prevailed throughout Dog’s world. Yet Dog Almighty did nothing to intervene. Like the God of Genesis chapter 1:13, Dog Almighty “saw everything he had made, and behold it was very good.” Like the God of Russell’s scenario, he enjoyed having his dogs worship him despite their suffering. Like Madeleine Albright, he thought the price was worth it.

Mind you, Dog’s counsel–Plantinga’s junior partner as it happens–did put up a defense. Reading from the brief Plantinga had already prepared (his “preliminary statement of the Free Will Defence”), and substituting only two words in Plantinga’s text (“Dog” for “God,” and “pit bulls” for “creatures”), he argued thus in court:

A world containing pit bulls that are significantly free is more valuable, all else being equal, than a world containing no free pit bulls at all. Now Dog could create free pit bulls, but he could not cause or determine them to do only what is right. For if he did so, then they would not be significantly free after all; they would not do what is right freely. To create pit bulls capable of good, he had to create ones capable of evil; and he couldn’t leave these pit bulls free to perform evil and at the same time prevent them from doing so. Dog did in fact create significantly free pit bulls; but some of them went wrong in the exercise of their freedom; this is the source of all the evil. The fact that these free pit bulls sometimes went wrong, however, counts neither against Dog’s omnipotence nor against his goodness; for he could have forestalled the occurrence of these evils only by excising the possibility of good. (p. 167 with specified substitutions)

Counsel rested the case for the defense by saying that “under these conditions,” and “despite” his near-omnipotence, Dog Almighty “couldn’t” have actualized an island in which no pit bulls ever misbehaved.[21] Snickers were heard in Court.

The flaws in Dog’s defense were evident. The prosecutor argued that Dog, knowing what he did from the start, was guilty of psychopathic criminal intent when he chose to animate the pit bull embryos; guilty of recklessness when he set the pit bulls free; guilty of negligence when he failed to intervene; and guilty of moral turpitude when he failed to attend immediately to those animals who had suffered as a consequence. Dog was free to do otherwise at any point. True, Dog didn’t do any of the biting, mauling, or killing himself. He wasn’t the direct cause of these evils. But he was responsible for them nonetheless.

Judge and jury agreed. Dog was found guilty, sentenced, and put away.

Now all the while, another case was proceeding, with another accused on trial.

6.2 The Trial of God Almighty and All-Knowing

“Beware of the God,” his accusers had long been saying. They were referring to God Almighty and All-knowing, the God of theistic faith.

In addition to the charges of which his human counterpart had been found guilty, he faced the graver charges of committing crimes against humanity, war crimes, and using the powers of nature–disease and disasters–as his weapons of mass destruction. Indeed, every calamity and crime, every sin and misdemeanor, had been laid at his feet as Creator and Supreme Commander of the universe: the primary planner and deliberate initiator of all that’s wrong with it.

Mark Twain had made the accusation back in 1909. Referring to “Disaster, Disease, and the rest” as “the Creator’s Grand Army” and to God himself as “Commander in Chief,” he’d argued

The Christian begins with this straight proposition, this definite, this inflexible and uncompromising proposition: God is all-knowing and all-powerful.

This being the case, nothing can happen without his knowing beforehand that it is going to happen; nothing happens without his permission; nothing can happen that he chooses to prevent. That is definite enough, isn’t it? It makes the Creator responsible for everything that happens, doesn’t it?[22]

A specially convened tribunal of the International Criminal Court was hearing God’s case, and Plantinga was Chief Counsel for the defense. Plantinga advanced his carefully prepared Free Will Defense–a covert form of the classic Defense of Necessity or “duress of circumstances.” But the judges noted that the only conditions limiting God’s actions were self-imposed, circumstances entailed by his own plans for creation. Adopting a verbal distinction that Plantinga’s acolytes later used when commenting on the case, they noted that the unfeasibility of God’s doing otherwise did not entail the logical impossibility of his doing otherwise.

They wondered at the moral myopia of someone who would advance a defense which, if accepted, would absolve Hitler of all his crimes on the grounds that he himself never fired a shot in anger or pulled the levers that released gas into the chambers of Auschwitz. The free will of Hitler’s subordinates, they noted, wasn’t “buck-stopping”; it did nothing to absolve Hitler as the person who planned and initiated all operations with full knowledge of their outcome. Would defense counsel, they asked, mount the same argument on behalf of other moral monsters of human history, men who have brought about suffering on an incalculable scale: men like Genghis Khan, Stalin, and Pol Pot, or others like George W. Bush and the mythical Moses, both of whom claimed God’s own sanction for their crimes? Was he suggesting they, the judges, adopt double standards, holding God Almighty and All-Knowing to lower standards than his weaker, benighted subordinates?

They reminded defense counsel that the accused was being judged according to the moral principles concerning Command Responsibility as recognized by Ping Fa around 500 B.C.E., principles that were eventually enshrined in the Hague Conventions of 1907, the Geneva Conventions of 1948, and the Nuremberg Charter of 1950 (Principles III and VI of which explicitly assign responsibility to Heads of State who have “planned” and “initiated” crimes against humanity). And, quoting from Article 7 (3) of the International Criminal Tribunal for the former Yugoslavia, they pointed out that the fact that a subordinate committed crimes

does not relieve his superior of criminal responsibility if he knew or had reason to know that the subordinate was about to commit such acts or had done so and the superior failed to take the necessary and reasonable measures to prevent such acts.[23]

This principle, they observed, is a particular instance of the more general moral truth:

If a person knows that a morally reprehensible act or state of affairs will occur or continue to occur unless they prevent it from occurring, and has the ability to prevent it from occurring, then that person is morally culpable for the occurrence of that event or state of affairs.

By virtue of his omnipotence and omniscience, God was found to be an accessory, before, during, and after, the fact of all evils.

The judges found God guilty on all counts: criminal intent, recklessness, negligence, and moral turpitude as well as the major charge of crimes against humanity.

Before passing sentence, they asked the accused, God Almighty and All-knowing–previously also known as the “Good” Lord–whether he had anything to say. But, like his son before Pilate, he was virtually mute.

Sentence was pronounced and carried out. He was transported to the graveyard of the gods, thence to fade through the mists of forgetfulness into the final oblivion that all gods deserve.

And all people wise and good said “Amen.”

7. Final Verdict: The Triumph of Down-Under Logic and Morals

“God is dead,” announced Nietzsche, indulging the fiction that God was once alive. I have entertained that same fiction; but only in order to demonstrate that, given the theist’s concept of God and the undeniable existence of evil, he never was.

7.1 The Failure, and Irrelevance, of the Free Will Defense

Plantinga’s defense of the theist’s God is a logical and moral fraud. The logical fraud has to do with what God “could” have done when conceiving his plans for the world. The moral fraud has to do with God’s misuse of his free will when executing those plans. The fact, were it a fact, that God’s subordinates also misused their free will–no matter how “free will” is construed–is nothing but a side show. Even if the causal links between God’s creative acts and ours were broken in the kind of way Plantinga’s contracausal account of free will envisages, God’s ultimate responsibility for evil would not be expunged.

7.2 Evil and God’s Command Responsibility for its Occurrence

Plantinga asked for, but did not want, a proof that God does not exist if evil does. He challenged someone to produce a proposition the conjunction of which with the indisputably true proposition

(2) Evil occurs.

would logically demonstrate the falsity of the proposition

(1) If God exists, then he is omniscient, omnipotent, and wholly good.

Ask and ye shall receive.

The Principle of Command Responsibility, as recognized in international law, deals expressly with the kind of scenario that Plantinga has developed in God’s defense, one in which he argues that sole responsibility for both moral and natural evils should be laid at the feet of God’s subordinate creatures (human and Satanic) and none at the feet of the Commander in Chief who planned and brought about the situation in which they find themselves.

By way of contrast, the Generalized Principle of Command Responsibility spells out sufficient conditions (as expressed in (3) below) for holding any person morally culpable, and criminally responsible, for the occurrence of evils, whether those evils have been brought about by that person himself or by that person’s subordinates or by any other agency, and whether or not any of that person’s subordinates brought about those evils of their own free will. It not only allows for the Plantinga scenario in which natural evils such as disease and disaster are attributable to the agency of nonhuman subordinates such as Satan, but also allows for the biblical scenario in which natural evils are attributable to God himself, i.e., are, as we so aptly describe them, “acts of God.”

The Generalized Principle of Command Responsibility, namely (3), I would argue, is an unassailable moral truth any exception to which, either in this world or in any other possible world, would be morally unconscionable and would undermine morality itself. Hence, if any moral truths are candidates for the status of necessary truth, this is one of them.

7.3 A Formal Proof that if Evil Exists then the Theists’ God Does Not

Here, then, is a the logical disproof of (1) given the truth of proposition (2). The Generalized Principle of Command Responsibility features as proposition (3).

(1) If God exists he is omniscient, omnipotent, and wholly good.

[Hypothesis that the theists’ God exists]

(2) Evil occurs.

[Statement of the undisputed fact of evil]

(3) If someone did not prevent the occurrence of evil despite having full knowledge in advance that it would occur if he were not to prevent it and despite also having unlimited power to prevent it, then that person is morally culpable for its occurrence.

[Generalized principle of command responsibility]

(4) By virtue of his omniscience, God knew in advance that evil would occur unless he was to prevent it.

[From 1 by definition of omniscience]

(5) By virtue of his omnipotence, God had the ability to prevent the occurrence of evil.

[From 1 by definition of omnipotence in terms of absence of nonlogical limits to God’s ability]

(6) God did not prevent the occurrence of evil.

[From 2 by double negation]

(7) God had the ability to prevent evil from occurring and knew it would occur if he did not prevent it.

[From 4 and 5 by conjunction]

(8) God is morally culpable for the occurrence of evil.

[From the conjunction of 3, 6, and 7 by modus ponens]

(9) God is not wholly good.

[From 8 by definition of “wholly good”]

(10) God does not exist.

[From 1 and 9 by modus tollens]

7.4 Conclusion

The theist’s God was supposed to be morally perfect as well as omnipotent and omniscient. But from the undisputed fact that evil exists in the world whose existence he supposedly brought about, it follows–by the unassailable moral truth expressed in the Generalized Principle of Command Responsibility–that he can’t have all three properties at once. Ipso facto, such a God does not now, and never did, exist. It is the logic of the new Down-Under Disproof, not of Plantinga’s Free Will Defense, that triumphs.


[1] John L. Mackie, “Evil and Omnipotence,” Mind, 64 (1955). His central thesis–that God could have created persons who always freely choose the good–is elaborated at length in Chapter 9 of his The Miracle of Theism, Oxford at the Clarendon Press, 1982.

[2] H. J. McCloskey, “God and Evil,” Philosophical Quarterly, 10 (1960).

[3] R. D. Bradley, “A Proof of Atheism,” Sophia, Vol. VI, No. 1, April 1967, pp. 39-45. My own argument for logical inconsistency was closer to Hume’s than to Mackie’s. I simply took it for granted that the issue had little to do with our exercise of our own free will (however that be construed) and much to do with God’s unconstrained exercise of his own–much to do, that is, with what I now refer to as his Command Responsibility.

[4] Alvin Plantinga, The Nature of Necessity, Oxford at the Clarendon Press, 1974, p. 167.

[5] For the sake of simplicity, I have changed the numerical designations of most of the propositions that appear in Planting’s presentation. Thus his original “(31)” becomes “(3)”; his “(32)” becomes “(4)”; and so on.

[6] Alvin Plantinga, “The Dawkins Confusion,” Books and Culture, March/April 2007.

[7] Its validity should be evident to someone who doesn’t even understand what the premises mean.

[8] Note further that even if (3) were consistent with God’s having each of these other properties severally, it still would not follow that (3) is consistent with God’s having all three properties jointly.

[9] It is, of course, a logical fact that from any conclusion C one can, by virtue of the logical rule known as “Addition,” validly deduce C or X, where the expanded conclusion can contain any concept one cares to put into it. But deducing the disjunction C or X is not at all the same as deducing C alone. It is the latter sort of deduction to which my rule of thumb applies.

[10] May 10, 1996, on 60 Minutes.

[11] Bertrand Russell, “A Free Man’s Worship,” Mysticism and Logic, Doubleday Anchor Books, 1957, p. 44.

[12] The other premises, (8), (9) and (10), are either ones that Plantinga regards as necessarily true (as in the case of (8)) or are statements of the inferential links between the other premises.

[13] I am presuming here that when Plantinga says that (3) is possibly true, he is not using the term “possibly” in its purely epistemic sense, i.e., that in which it means something like “for all I know to the contrary.” Otherwise, he would be guilty of still another kind of equivocation on the meaning of “possibly.”

[14] See also “For suppose all the people who exist in alpha [his name for the actual world] suffer from transworld depravity.” (p. 187)

[15] It is worth noting that on page 188 Plantinga writes, “it is possible that every creaturely essence suffers from transworld depravity” and adds, as a footnote, the qualification “i.e., every creaturely essence entailing is created by God.” Clearly creaturely essences having the property of being created by God are those persons only who exist in the actual world.

[16] For a careful discussion of these and other problematic uses of modal expressions see Bradley and Swartz, Possible Worlds: an Introduction to Logic and its Philosophy, Indiana, Hackett Publishing Company, and Oxford, Basil Blackwell Publisher, 1979, pp. 328-332. There we distinguished between absolute and relative uses of modal expressions. Here, for reasons of clarity, I have chosen to draw the distinction between absolute and consequential uses.

[17] “Must the Future be What it is Going to Be?,” Mind 68; reprinted in The Philosophy of Time, ed. Richard M. Gale, 1967; New York, Anchor Books.

[18] There is a better warrant for saying that God’s determining what the future will hold for all of us by having our destinies “written in the book of life” from “the foundation of the world” is what gives him foreknowledge of our futures.

[19] It is worth noting that an analogous modal fallacy can occur in reasoning involving nonlogical uses of modal terms, e.g., their uses to express causal modalities (modal concepts involving not logical but causal laws). Example: the invalid inference from “As a matter of (causal) necessity, if a rocket does not reach escape velocity it will fall back to earth” to “If a rocket does not reach escape velocity then it is (causally) impossible for it not to fall back to earth.”

[20] I am indebted to my friend and philosophical collaborator (as coauthor of Possible Worlds) for providing this memorable formulation of the metalogical principle violation of which generates what we earlier had christened “The Modal Fallacy.”

[21] The fact that Plantinga’s Defense lends itself so readily to parody tells one something about just how shallow it really is.

[22] Mark Twain, Letters from the Earth, Fawcett Crest Book (7th printing New York, 1968), p. 33. Twain continues scathingly: Then, having thus made the creator responsible for all those pains and diseases and miseries above enumerated, and which he could have prevented, the gifted Christian blandly calls him Our Father … What do you think of the human mind? I mean, in case you think there is a human mind.

[23] Additional Protocol 1 of 1977 to the Geneva Conventions of 1949, Article 86 (2).

Copyright ©2007 Raymond D. Bradley. The electronic version is copyright ©2007 by Internet Infidels, Inc. with the written permission of Raymond D. Bradley. All rights reserved.