+
The argument just defines God into existence using fancy modal logic. I can define anything into existence this way.
1. The argument's real move is a question about coherence, not a definitional trick.
The ontological argument does not say "by definition God exists" and stop there. It asks whether the existence of a certain kind of being, a maximally great necessarily existent being, is even coherent as a concept. If the concept is coherent, it is possible. And if it is possible, the logic of P3 takes over: a being with necessary existence cannot be possible without being actual. The critical step is P1, and P1 is a claim about conceptual coherence, not a definition smuggling in a conclusion.
2. Parody attempts fail because their concepts lack genuine intrinsic maxima.
The parody objection was first made by Gaunilo of Marmoutier, an 11th-century monk, who responded to Anselm's original argument with a "most perfect island" example ("On Behalf of the Fool," c. 1078). The same structure has since been applied to maximally great pizzas, unicorns, and similar objects. These parodies all fail for the same reason.
Consider a maximally great island. What would that even mean? You could always add one more palm tree, one more waterfall, one more beach. Island greatness is not the kind of property that has an intrinsic ceiling. The same goes for pizzas, which can always have more toppings, and unicorns, which can always be imagined slightly faster or more elegant. These are all contingent, finite objects whose defining properties admit of open-ended improvement. There is no natural stopping point.
The properties of a maximally great being are entirely different. Omniscience means knowing all truths: you cannot know "one more truth" than that. Omnipotence means having all logically possible power: you cannot have "a little more power" than every possible power. Moral perfection means being perfectly good: there is no increment beyond perfection. These represent genuine absolute maxima that cannot be raised even in principle.
3. Necessary existence is what makes the parodies collapse entirely.
Even granting a parody of "maximally great island," the concept of such an island does not include necessary existence as one of its defining properties. Islands, however great, are the kind of things that depend on geological conditions, sea levels, and contingent physical facts. They exist in some possible worlds and not others. The modal argument applies only to a being for whom necessary existence is built into the very concept of maximal greatness. No island, pizza, or unicorn, even in their most inflated parody forms, is the kind of thing that could exist necessarily.
4. Kant's version of this objection targets Anselm's argument more than Plantinga's.
The most philosophically developed form of the "defining into existence" objection is Immanuel Kant's argument that existence is not a predicate. Kant's objection has genuine force against Anselm's original formulation. Plantinga's modal version is specifically designed to address it, because it does not argue that existence adds greatness to a concept. It argues that necessary existence is a genuine modal property that is part of what maximal greatness means. See Defeater 5 for a full treatment of Kant's challenge and why Plantinga's version responds to it effectively.
+
If you can argue for a maximally great good being, you could run the same argument for a maximally great evil being. Why should we prefer one over the other?
1. Maximal greatness includes moral perfection as a constitutive property.
The argument is specifically about a maximally great being. Moral perfection, meaning maximal goodness, is part of what maximal greatness means. A "maximally evil" being would, by definition, lack the moral perfection that constitutes one of the absolute maxima of greatness. It could not therefore be maximally great. The two concepts are not interchangeable. The argument cannot be run for an evil being without first gutting the concept of what "maximal greatness" means.
2. Evil is a privation, not a positive excellence: the classical tradition of privatio boni.
Augustine of Hippo argued in his Confessions and City of God that evil is not a positive property or substance in its own right. It is a privation, a lack or corruption of good. Aquinas developed this further in the Summa Theologica (I.48.1): evil has no independent existence; it is always a deficiency in something that should have goodness.
On this account, properties like justice, wisdom, love, and holiness are genuine perfections: they add something positive to a being. Cruelty, malice, and injustice are defects: they represent the absence or corruption of something that ought to be present. You cannot construct a "maximally evil" being by analogy with a maximally great being, because evil is not the kind of thing that has a positive maximum. There is no "absolute ceiling of evil" in the way there is an absolute ceiling of omniscience, because evil is defined by what is missing, not by what is present.
3. The concept of a necessarily and maximally evil omnipotent being is internally incoherent.
Suppose, granting the parody for the sake of argument, that a "maximally evil" being were omniscient and omnipotent. Such a being would know all truths, including every truth about goodness, beauty, and moral perfection. It would have all power, including the power to recognize and pursue the good. A will that is necessarily and maximally opposed to all good, while also being all-knowing and all-powerful, is not a coherent maximization of anything. It describes a being permanently in conflict with its own full knowledge and power. This is not a mirror image of maximal goodness. It is an incoherent hybrid.
+
The reverse modal ontological argument works just as well: if it is possible that a maximally great being does not exist, then since such a being would exist necessarily if it existed at all, it follows that it is impossible for a maximally great being to exist, and therefore God does not exist. There is no reason to prefer the positive possibility over the negative one.
1. The reverse argument is logically valid but its premise is not equivalent to the original.
The reverse argument is a real logical challenge and Plantinga himself acknowledges it directly. It is valid within S5 modal logic: if it is metaphysically possible that no maximally great being exists, then no maximally great being exists. The two starting premises, "possibly a MGB exists" and "possibly no MGB exists," cannot both be true at the same time. If a MGB exists in any possible world, it exists in all of them; and if it fails to exist in any possible world, it fails to exist in all of them.
The question is therefore not which argument form is valid, but which starting premise has better support.
2. "Possibly no MGB exists" requires demonstrating impossibility, not merely expressing uncertainty.
The symmetry between the two arguments is only apparent. The two starting premises carry very different evidential burdens.
To say "it is metaphysically possible that a MGB exists" requires only that the concept contains no internal contradiction. Positive conceivability without demonstrated incoherence is the standard evidence for metaphysical possibility.
To say "it is metaphysically possible that no MGB exists," however, means that there is a genuine possible world in which a MGB is absent. But a MGB, if coherent, exists necessarily in every possible world. So the only way a possible world could exist without a MGB is if the concept of a MGB is genuinely incoherent, meaning necessarily impossible. The reverse argument's starting premise therefore silently assumes that the concept of maximal greatness is incoherent.
This is the asymmetry: the positive premise requires only coherence, while the negative premise requires demonstrated incoherence. Demonstrated incoherence is a much heavier burden, and no philosopher has successfully discharged it.
3. Epistemic uncertainty is not the same as metaphysical possibility.
An atheist might say: "I don't know whether a MGB exists. For all I know, it might not. So it is possible it does not exist." But this confuses epistemic possibility (I don't know) with metaphysical possibility (there is a genuine possible world in which it does not exist).
Consider pi again. Before a proof of its irrationality existed, a mathematician might have said "for all I know, pi might be rational." That would be epistemic possibility. But there is no possible world in which pi is rational, because its irrationality follows necessarily from its mathematical nature. The mathematician's uncertainty did not create a genuine metaphysical possibility.
Similarly, an atheist's uncertainty about God does not create the metaphysical possibility of God's non-existence. If the concept of maximal greatness is coherent, then in every possible world a MGB either necessarily exists or necessarily does not. Uncertainty about which of these is true is an epistemic condition in the mind of the doubter, not a fact about the modal structure of reality.
4. The theist has positive, converging reasons to accept the possibility premise.
The positive premise, that a MGB is possible, is supported by several independent considerations. The concept has survived centuries of philosophical scrutiny without a demonstrated incoherence. Other theistic arguments (cosmological, moral, fine-tuning) independently suggest that a being with the relevant properties exists, which is strong evidence that the concept is genuinely possible. And the intuitive coherence of the attributes, omniscience, omnipotence, and moral perfection, gives direct support to P1.
The negative premise, that possibly no MGB exists, has no comparable support. It relies solely on the atheist's uncertainty, which, as shown above, is not evidence of metaphysical possibility. Whoever has better reasons for their possibility claim has the stronger argument. The theist's reasons are substantive; the atheist's are merely the absence of conviction.
5. Plantinga's own honest acknowledgment of the argument's limits.
Plantinga himself does not claim this argument can compel rational assent from everyone. A person who begins with genuine doubt about whether the concept of maximal greatness is coherent, and who therefore withholds the possibility premise, is not making a logical error in the same sense that misapplying a syllogism is an error. What the argument does establish is this: the two options, God necessarily exists or God is necessarily impossible, are the only alternatives available once the modal structure is accepted. The argument forces the atheist to choose between a demanding position (maximal greatness is incoherent) and theism. Most people, on reflection, find the first option harder to maintain than the second.
+
P1 assumes that it is possible a maximally great being exists. But maybe God is metaphysically impossible due to hidden contradictions in the concept. Granting P1 already begs the question.
1. The debate becomes a question about whether the concept of God is coherent.
The argument forces the discussion onto precise philosophical ground: is the concept of a being that is omnipotent, omniscient, morally perfect, and necessarily existent internally consistent? If critics claim it is impossible, they must identify an actual contradiction, not merely express suspicion. This is a much more demanding intellectual task than simply saying "I don't believe in God."
2. Specific alleged contradictions have received careful philosophical responses.
Philosophers have proposed several specific challenges to the coherence of maximal greatness. The paradox of the stone asks: can God make a stone so heavy he cannot lift it? If he can, there is something he cannot do (lift it); if he cannot, there is something he cannot do (make it). This appears to show omnipotence contradicts itself.
The standard resolution is that omnipotence means having all power over logically possible states of affairs. Creating a stone so heavy that an omnipotent being cannot lift it is a logical contradiction, like asking for a square circle. The limit is not in God's power; it is in the concept being described. Omnipotence does not include the power to bring about logical contradictions, because there is no genuine "power" that corresponds to an impossibility. The paradox disappears once omnipotence is defined with this qualification.
For challenges involving the compatibility of God's attributes with each other, such as omniscience and human freedom, see Defeater 7, which addresses these in full.
3. The default rational stance favors possibility in the absence of demonstrated incoherence.
In philosophy, the absence of a demonstrated contradiction in a concept is the standard prima facie evidence that the concept is coherent and therefore possible. We do not withhold the judgment "this concept is coherent" until a proof of coherence is provided. We withhold it only when a contradiction has actually been identified.
Think of the concept of a black hole. Before Einstein's general relativity, the idea of an object so massive that light could not escape it might have struck many people as strange or counterintuitive. But strangeness is not contradiction. Once physicists showed no internal contradiction in the concept, the question became an empirical one: do they exist? Similarly, once no internal contradiction in maximal greatness is demonstrated, the question becomes: is the concept actual? And the modal argument shows that, given the special nature of necessary existence, actual follows from possible.
4. The argument shifts the philosophical burden decisively.
The atheist cannot simply sit back and deny that God exists. To resist this argument's conclusion, they must actively defend the claim that God's existence is metaphysically impossible by identifying a genuine contradiction in the concept of maximal greatness. Simply saying "I find it unlikely" or "I don't see the evidence" is insufficient. This is a substantially heavier burden, and the argument forces it into the open. Until that burden is discharged, it remains rational to accept P1 and, by the logic of the argument, C1.
+
At best this argument shows some "maximally great being" exists, not the specific God of Christian revelation. The gap between a philosophical abstraction and the God of the Bible is enormous.
1. This is correct, and it is not what the argument is designed to do.
The modal ontological argument is designed to establish one specific conclusion: that a necessarily existent, personal being with maximal greatness, meaning omnipotence, omniscience, and moral perfection, exists in every possible world including the actual one. This is what it establishes, and it establishes this rigorously.
2. What this conclusion already rules out.
The conclusion is not theologically thin. It immediately eliminates several significant positions: atheism and robust naturalism, the view that only physical reality exists; finite godism, the view that a god exists but is limited in power, knowledge, or goodness; and polytheism in its classical forms, since multiple beings cannot each independently possess maximal greatness (two omnipotent beings each claiming all power is incoherent). The argument clears a great deal of philosophical ground.
3. Further arguments build on this foundation to identify the God of Scripture.
The modal ontological argument is one piece of a larger cumulative case. Other arguments contribute additional attributes. The cosmological argument establishes that the necessary being is the cause of all contingent existence. The moral argument establishes that this being is the objective ground of moral value. The fine-tuning argument establishes that this being is a rational, intelligent designer of extraordinary precision. The argument from consciousness establishes that this being is personal and the source of mind.
Together these converging arguments paint an increasingly specific portrait that aligns with the God of classical theism. Evidence from Scripture and the historical resurrection then addresses the question of which religious tradition accurately describes this God's self-revelation. The ontological argument provides the philosophical foundation; the other arguments and the historical evidence fill in the portrait.
+
Immanuel Kant argued in the Critique of Pure Reason (1781) that existence is not a real predicate. When we say "God exists," we are not adding a property to the concept of God. We are merely saying the concept has an instance in the world. You cannot make a concept "greater" by adding existence to it, because existence is not a property that belongs to concepts. This destroys the ontological argument at its foundation.
1. Kant's objection, stated precisely.
In the Critique of Pure Reason (A598/B626), Kant argues that "Being is evidently not a real predicate; that is, it is not a concept of something which could be added to the concept of a thing." When I think of 100 thalers in my mind and then say they exist in my pocket, the 100 thalers in my pocket have exactly the same properties as the 100 thalers I imagined. Existence adds nothing to the concept; it merely says the concept is instantiated. Applied to the ontological argument: adding "exists in reality" to the concept of God does not make God greater in any genuine sense, it only says the concept has an instance. The argument's move from concept to existence is therefore illegitimate.
This objection carries genuine force against Anselm's original formulation in the Proslogion, where the key move is that a God existing in reality is greater than a God existing only as an idea. If existence is not a predicate, the comparison does not make sense.
2. Plantinga's modal version does not argue for bare existence as a great-making property.
The argument presented here is Plantinga's reconstruction, not Anselm's original. Plantinga's version does not claim that bare existence, the mere fact of being instantiated, is a great-making property. Kant is entirely correct that bare existence is not a property you can add to a concept to make it greater.
What Plantinga argues is that necessary existence is a genuine modal property, and that it is a great-making property. Necessary existence is not the same as bare existence. It is the property of existing in every possible world, of being such that one's non-existence is a logical impossibility. This is a genuine and substantial property, one that can distinguish beings from each other in terms of their modal status.
3. The difference between bare existence and necessary existence is real and significant.
Consider two mathematical objects: the number 7 considered as something that might or might not exist depending on circumstances, versus the number 7 considered as a necessarily existing abstract object whose non-existence is logically impossible. Even Kant would grant that these are different modal descriptions. The second adds genuine information about the object's modal profile.
Think of a promise. "I will help you" is a contingent commitment that can fail for many reasons. "I necessarily will help you under any circumstances whatsoever" is a qualitatively stronger commitment. Necessary existence works analogously. It is a richer and stronger modal property than bare existence, not merely existence with extra emphasis.
4. Kant's objection therefore does not defeat Plantinga's version of the argument.
Kant showed that bare existence cannot be used to compare concepts in terms of greatness, and he was right. Plantinga's argument does not use bare existence in that way. It uses necessary existence, which is a genuine modal property that Kant's critique does not reach. The two arguments are targeting different properties. Kant's refutation is powerful but its scope is limited to formulations, like Anselm's original, that rely on comparing a concept-only God with a reality-including God. Plantinga's version operates on different grounds and survives Kant's critique.
+
The argument moves from "we cannot find a contradiction in the concept of maximal greatness" to "a maximally great being is genuinely metaphysically possible." But conceivability is a poor guide to metaphysical possibility. Before non-Euclidean geometry, Euclidean space seemed necessarily true. Before quantum mechanics, classical determinism seemed obviously possible. Our failure to find a contradiction might mean only that our conceptual powers are limited, not that the concept is genuinely coherent.
1. The objection is genuine and Plantinga acknowledges its force.
Peter van Inwagen ("Ontological Arguments," Noûs, 1977) and others have pressed this challenge. It is correct that conceivability is not an infallible guide to metaphysical possibility. History offers examples of ideas that seemed coherent but turned out to contain hidden inconsistencies. This means P1 cannot be established with absolute certainty by mere failure to find a contradiction.
2. The default rational stance favors possibility absent demonstrated incoherence.
However, the objection sets an unreasonably high bar. We routinely and rationally accept that concepts are possible when we cannot identify a contradiction in them. The concept of a black hole, before it was observed, contained no demonstrated contradiction, and physicists were rational to treat it as possible and investigate it. The concept of a living organism on another planet contains no demonstrated contradiction, and scientists are rational to treat it as possible and search for it.
The same standard applies to maximal greatness. We cannot be certain the concept is coherent, but the rational default, in the absence of a demonstrated contradiction, is to treat it as possible rather than impossible. The skeptic cannot point to an identified contradiction; they can only gesture at the theoretical risk that one might be lurking. That is insufficient grounds for denying the possibility premise.
3. Conceivability is a genuine, defeasible guide to possibility.
The examples of Euclidean geometry and classical determinism illustrate that conceivability can mislead. But these cases involved concepts that turned out to conflict with deep mathematical or physical facts. The skeptic owes an account of what deep fact makes maximal greatness impossible. Without a specific candidate for the hidden contradiction, the risk that one exists is speculative rather than evidenced. Speculative risk is not sufficient to defeat a reasonable possibility claim.
4. The concept of maximal greatness has independent support, not just absence of refutation.
The case for P1 does not rest solely on the failure to find a contradiction. Other theistic arguments provide independent convergent evidence that a being with the relevant attributes exists. The cosmological argument suggests a necessarily existing first cause. The moral argument suggests a necessary ground of objective moral value. The fine-tuning argument suggests a rational designer of extraordinary intelligence. If these arguments have force, they provide positive evidence that a being with maximal greatness exists, which is strong evidence that the concept is metaphysically possible. The modal ontological argument does not stand alone; it is supported by a wider body of independent evidence that converges on the same conclusion. See Defeater 3 for the related discussion of how the burden of proof falls on those who assert impossibility.
+
Even if each attribute of maximal greatness is individually coherent, the combination might be incoherent. Philosopher Wes Morriston and others argue: Can omnipotence and moral perfection coexist? If God can do anything, He can do evil things, but a morally perfect God cannot do evil. Can omniscience and genuine human free will coexist? If God knows infallibly what I will freely choose, my choice seems determined in advance. If the attributes jointly incoherent, P1 is false.
1. The challenge requires showing a genuine contradiction, not a surface tension.
Wes Morriston ("Omnipotence and Necessary Moral Perfection: Are They Compatible?" Religious Studies, 2001) presses this challenge carefully. It deserves a careful response. The key issue is whether the apparent tensions between attributes rise to the level of logical contradiction or whether they dissolve once the attributes are defined more precisely.
2. Omnipotence and moral perfection: the apparent conflict resolved.
The apparent conflict runs: an omnipotent being can do anything; a morally perfect being cannot do evil; therefore these attributes cannot coexist. But this argument trades on an imprecise definition of omnipotence.
Omnipotence means having all power over logically possible states of affairs. Bringing about a logical impossibility, such as a married bachelor or a round square, is not a genuine power, because there is no coherent state of affairs to bring about. Similarly, a morally perfect being sinning is not a coherent state of affairs; it is not a limitation on power but a description of something that has no coherent content. A perfectly good will acting against perfect goodness is a contradiction, like a perfectly round object having corners.
The standard illustration is the paradox of the stone: can God make a stone so heavy He cannot lift it? The answer is that "a stone too heavy for an omnipotent being to lift" describes a logical impossibility. There is no such thing. God's inability to create this non-thing is not a limitation on His power; it is a recognition that logical impossibilities have no content for power to act on. Far from undermining omnipotence, this clarification shows that omnipotence and moral perfection are compatible once both are defined with precision.
3. Omniscience and human freedom: the main proposals for resolution.
The tension between God's foreknowledge and genuine human free will is perhaps the most debated issue in philosophy of religion. Several serious proposals exist.
Molinism, developed by the 16th-century Jesuit Luis de Molina, holds that God has "middle knowledge": knowledge of what every possible free creature would freely choose in every possible situation. God can have complete foreknowledge while creating a world in which creatures genuinely choose freely, because His knowledge tracks what they would freely do rather than determining it.
Open theism holds that God voluntarily limits His foreknowledge of future free choices in order to preserve genuine creaturely freedom. On this view, omniscience is defined as knowing all truths that can be known, and future free choices are not yet determined truths.
Simple foreknowledge models hold that God's timeless knowledge of the future is compatible with free will because God's eternal perspective does not impose causal determination on temporal events, any more than watching a recording of a past event retroactively determines it.
These proposals are contested among philosophers and theologians. But their existence demonstrates that the tension between omniscience and freedom has not produced a decisive proof of incoherence. The challenge remains an open philosophical debate, not a closed refutation.
4. The burden of demonstrating incoherence remains unmet.
Each specific challenge to the joint coherence of the divine attributes has generated extensive philosophical literature with serious responses. None of these challenges has achieved the status of a consensus demonstration that maximal greatness is logically impossible. The concept continues to be defensible by serious philosophers, which means P1 retains its standing as a rational premise.
+
Jordan Howard Sobel argues in Logic and Theism (Cambridge University Press, 2004) that the modal machinery of the ontological argument risks proving too much. If we accept that the apparent coherence of a concept justifies inferring its possibility, and if possible necessary existence entails actual existence, then we might be able to use similar reasoning to derive the existence of multiple or conflicting necessary beings, or to generate other problematic conclusions the argument cannot distinguish from theism.
1. Sobel's objection, stated carefully.
In Logic and Theism (Cambridge University Press, 2004), Jordan Howard Sobel raises the concern that the modal argument's machinery, if valid, is not specific enough to single out the God of classical theism. If "apparently coherent concept of a necessarily existing being" plus "S5 modal logic" yields existence, the question becomes whether the machinery can be applied to other apparently coherent concepts of necessary beings in ways that generate conflicts or absurdities. If it can, the argument proves too much and its conclusions are suspect.
2. Maximal greatness is constrained by uniqueness in a way arbitrary concepts are not.
The key restraint on the modal machinery is the requirement that the concept in question be internally coherent and represent a genuine absolute maximum. Maximal greatness is constrained by this requirement in a way that prevents escalation.
A maximally great being, by definition, has all logically possible power, all knowledge of truths, and perfect goodness. Two beings each of whom independently has all logically possible power would conflict: if each has all power, neither can limit the other, yet omnipotence would seem to imply the ability to limit any other being. This means the concept of multiple maximally great beings is not coherent in the way the concept of a single maximally great being is. The modal argument, applied with the coherence requirement intact, generates at most one maximally great being, not a set of competing ones.
3. The coherence requirement filters out arbitrary necessary being concepts.
Not every description of a "necessarily existing being" passes the coherence filter required for P1. A "necessarily existing perfectly evil being" is incoherent for the reasons given in Defeater 1. A "necessarily existing being that is maximally powerful but ignorant" is incoherent because maximal power would seem to include the power to know anything knowable. The machinery only runs for concepts that pass genuine internal coherence scrutiny, and maximal greatness as defined by omnipotence, omniscience, moral perfection, and necessary existence is the concept that classical theism proposes and defends.
4. The argument's power lies in the uniqueness of the concept it employs.
Sobel's challenge is most powerful if we could easily generate multiple coherent necessary being concepts and use the same modal machinery to derive their existence in conflict with each other. But the burden lies on the challenger to produce such a concept that is genuinely coherent under scrutiny and genuinely conflicts with maximal greatness. The centuries of philosophical debate surrounding the concept of God have not produced a successful competing concept that passes these tests. The escalation concern remains a theoretical risk rather than a demonstrated problem.
