Arguments from Reason

Anselm's argument appears inside a prayer. His project is often summarised as fides quaerens intellectum, faith seeking understanding: he begins as a believer asking to understand what is already confessed. This context does not make the reasoning invalid, but it changes what success might mean. The argument may reveal implications of a theistic concept to someone who possesses it rather than compel every hearer to adopt that concept. The “fool” of Psalm 14 denies God while nevertheless understanding the words “a being than which nothing greater can be conceived”.

The argument is a priori because its essential premises concern concepts and comparative greatness rather than observations about the world. It is also deductive: if the definition and greatness premises are true, denial is meant to generate contradiction. That distinguishes it from cosmological and design arguments, whose premises are empirical and whose conclusions compete as explanations. Anselm is not asking which hypothesis best explains an observed effect; he is asking whether the greatest conceivable being can coherently be conceived as non-existent.

The definition requires care. “That than which nothing greater can be conceived” does not mean the greatest thing anyone happens to imagine, nor an indefinitely improved superhero. It is a limiting concept: no possible being could exceed it in greatness. If “greatness” is left as a collection of preferences, the argument stalls. Anselm's later chapters associate unsurpassable greatness with wisdom, goodness, necessity and aseity. Critics can therefore challenge whether these perfections are coherent together or whether a maximum exists for every great-making property.

The Chapter II reasoning can be reconstructed without the misleading slogan “existing things are always better”. First, even the fool understands the description, so the being exists in intellectu, in the understanding. Secondly, suppose it exists only there. Thirdly, a being otherwise alike that exists in re, in reality, would be greater. The supposed greatest conceivable being could therefore be exceeded. Since that contradicts its definition, it cannot exist only in understanding.

The painter analogy distinguishes possessing a concept from assenting to its instantiation. Before painting, the picture exists as an idea; after painting, the painter both has the idea and knows that an object answers to it. Gaunilo's objection is that understanding words does not guarantee possession of a coherent object. People understand reports, rumours and fictional descriptions without a corresponding reality. A round square fails to identify a possible object however fluently the phrase is repeated. The defender must show that Anselm's concept is coherent and genuinely understood before the reductio begins.

The comparative premise also needs refinement. A real disease is not better than an imaginary cure. Anselm means greatness appropriate to an unsurpassable being, not value in every context. Yet treating existence as a perfection seems to move from a description to instantiation. The argument's defender can say that the contradiction is not “God lacks one ordinary property”; it is that unsurpassable reality cannot be merely conceptual. The critic replies that a conditional is all that follows: if such a being exists, it does not exist merely in thought. The disputed step is whether the concept itself guarantees an instance.

Chapter III is not simply a repetition. Anselm distinguishes a being which can be conceived not to exist from one which cannot. Contingent existence could have failed to obtain; necessary existence cannot fail in any possible circumstance. A necessarily existing unsurpassable being would be greater than an otherwise equivalent contingent being. Therefore the greatest conceivable being cannot be conceived not to exist.

This modal version helps answer parody. Islands, however excellent, are contingent objects with no intrinsic maximum: another beach or tree can always be added, and there is no contradiction in there being no island. A maximally great God is alleged to be different because necessary existence belongs to the relevant kind of greatness. Anselm's reply to Gaunilo effectively invites him to apply the reasoning to anything which genuinely cannot be conceived not to exist; the challenge is finding a second such candidate.

Norman Malcolm later argues that ordinary existence is not a perfection, but necessary existence or absolute independence is. If God is possible, God cannot come into or pass out of existence, since that would make God dependent. The resulting modal structure is powerful: possibly necessary existence entails necessary existence in standard modal system S5. But the weight moves to the possibility premise. A critic may say that a necessarily non-existent greatest being is equally formulable, or that logical possibility cannot be read directly from a positive description. Modal restatement clarifies the argument; it does not provide a cost-free proof of possibility.

Anselm addresses the "fool" who says in the heart that there is no God. The fool can still understand the phrase "that than which nothing greater can be conceived" when it is heard, so the description exists in the understanding even if its object is denied. This opening move concerns conceptual competence, not yet belief. A person can understand a description of a golden mountain without accepting that any such mountain exists.

The definition is also a rule of maximality. It does not ask a thinker to picture an infinitely large object, because pictorial imagination always supplies finite features. Instead, it directs thought towards unsurpassable greatness: any candidate than which a genuinely greater being is conceivable fails to be God. The relevant greatness includes power, knowledge, goodness and independence rather than spatial size. This makes the concept evaluative and modal from the outset.

A first difficulty is whether all perfections are commensurable and jointly coherent. Greater power might conflict with the freedom of creatures; perfect justice and mercy can appear to recommend different outcomes; timelessness may sit uneasily with responsive personal action. Unless the set of great-making properties can be exemplified together, the description may fail to determine a possible object. Understanding the words separately is not yet grasping one coherent essence.

A second difficulty concerns reference. A definite description can be meaningful even when nothing satisfies it. Anselm's later reasoning must therefore do more than repeat the definition with emphasis. His argument is strongest when read as a reductio: suppose the unsurpassable being exists only as an understood possibility, then derive a conflict with unsurpassability. Its vulnerability lies in the principle used to generate that conflict—whether existence in reality or necessary existence makes a being greater in the relevant way.

The Chapter II argument can be displayed without devotional language. First, God is understood as that than which no greater can be conceived. Secondly, this being exists at least in understanding. Thirdly, existing in reality as well as in understanding is greater than existing only in understanding. If God existed only in understanding, a greater being could be conceived—one otherwise alike but real. That contradicts the definition, so God exists in reality.

The logical form is valid if its premises and comparison are granted. Most criticism therefore targets the third premise or the assumption that the description successfully picks out a possible maximum. "Existing in the understanding" is ambiguous. The thought exists as a mental event, but the described being does not thereby possess a diminished mental mode of existence. Comparing a real God with a merely conceived God may compare one possible object with no object, not two beings with different degrees of greatness.

Existence also differs from an ordinary perfection. A compassionate person and a more compassionate person can both be objects, allowing a comparison. A hundred real coins and a hundred imagined coins may have the same specified concept; reality affects whether the concept is instantiated, not which properties the coins would have if instantiated. If so, adding "and exists" does not make a concept internally greater. It asserts that something answers to it.

A defender can reply that Anselm is not attaching existence to an arbitrary description. Absolute greatness includes independence from thinkers: a reality-dependent-on-conception would be inferior to an otherwise unsurpassable reality. This shifts the issue towards necessary existence and Chapter III. It may save the intuition that God cannot be a fiction while conceding that ordinary existence is not one quality among others. The success of Chapter II then depends on a specifically divine mode of existence rather than a general rule that reality makes every object greater.

Chapter III says that the unsurpassable being cannot be conceived not to exist. A being whose non-existence is impossible is greater than one which happens to exist but could fail to exist. If God could be conceived as contingent, a necessarily existing being would be conceivable as greater, contradicting maximality. This is not identical to Chapter II. The comparison concerns modes across possible circumstances, not a transition from a mental object to an external one.

Necessary existence avoids some of Gaunilo's parody. Islands, however excellent, are contingent complexes. Their features admit indefinite improvement and their non-existence is coherent. Anselm can therefore say that the reasoning applies only to a candidate whose nature is unsurpassable and whose mode of being is included in that unsurpassability. The reply narrows the argument rather than licensing definition of every desirable object into existence.

Yet a modal gap remains. From the claim that a maximally great being would exist necessarily if possible, it does not follow without a possibility premise that such a being is possible. Necessary beings differ from contingent beings because their existence does not vary across possible worlds; an impossible being also fails to vary, since it exists in none. The decisive work is establishing that the divine attributes form a coherent possible essence.

This explains why modal ontological arguments often look powerful to someone who already accepts divine possibility and unpersuasive to someone who does not. The inference from possible necessary existence to necessary existence can be valid in a strong modal logic, while the premise remains religiously substantial. A high-level judgement should not call the argument a verbal trick if its modal logic is sound, but neither should it treat "possibly" as a harmless concession. In this context possibility contains most of the metaphysical dispute.

Gaunilo writes on behalf of Anselm's fool. His objections are often reduced to the lost island, but he first questions whether hearing a description places its object in the understanding in the required way. A person can repeat unfamiliar or mistaken words without possessing an adequate concept, and a reported object can be entertained without being represented as real. If God's essence exceeds finite comprehension, Anselm cannot assume the clear intellectual possession needed by the reductio.

The island is a parody of form. Imagine an island more excellent than every other island. If existing in reality is greater than existing only in thought, the greatest conceivable island must exist. Because this conclusion is absurd, the general move from conceived excellence to instantiation is invalid. The objection tests whether Anselm has supplied a rule special to God or a pattern that generates non-existent perfect objects.

Anselm replies that islands do not possess a maximum. More fertile land, another beautiful bay or safer climate can always be added, and the qualities lack one intrinsic scale. More importantly, an island is contingent by kind. Only that which cannot be conceived not to exist is at issue in the Proslogion's strongest reasoning. This answer defeats a simple one-to-one parallel, but it risks circularity if necessary existence is introduced merely because it protects God from parody.

Gaunilo's most durable challenge is therefore not that every argument can be copied with an island. It is that conceptual analysis needs a justified bridge to reality. If the special divine concept is independently coherent and necessary existence is genuinely a perfection, the parody fails. If those claims are secured only by defining the target as unparodiable, the argument assumes what it seeks to establish. An essay should assess the reply at that level rather than awarding the debate to whichever example seems more memorable.

Aquinas agrees that the proposition "God exists" is self-evident in itself if God's essence is existence. In a self-evident proposition the predicate belongs to the subject's essence. Human knowers, however, do not comprehend the divine essence directly. The proposition is therefore not self-evident to us, and its denial can be entertained without formal contradiction. We must reason from effects better known to us towards their cause.

This is an epistemic objection rather than a denial that God is metaphysically necessary. It blocks Anselm's assumption that finite thinkers possess a concept rich enough to derive instantiation. The word "God" may signify an ultimate reality while our understanding remains analogical and incomplete. A definition tells us how believers use the term; it does not guarantee insight into the essence which would make existence inseparable.

Anselm could answer that complete comprehension is unnecessary. A thinker need understand only the formal description of unsurpassable greatness and the relation between contingent and necessary being. We reason validly with concepts whose objects we do not exhaust. Aquinas' response is that the description does not establish that the predicates converge on one essence or that what it signifies is possible. Effects provide the required existential contact.

The contrast reflects wider methods. Anselm seeks what reason can unfold from faith's object; Aquinas controls theological predication through creaturely effects. An ontological argument promises certainty independent of changing observation, but risks moving within concepts detached from reality. An a posteriori route makes contact with existence, but yields a conclusion no stronger than its causal premises. A strong assessment can accept divine necessity within Aquinas' system while rejecting its availability as the starting premise of a proof.

In the Fifth Meditation, Descartes compares God's existence with geometrical properties. Just as understanding a triangle involves its three angles, clear and distinct understanding of a supremely perfect being involves every perfection. Necessary existence is a perfection, so it cannot be separated from God's essence. Descartes does not claim that thinking causes God to exist; rather, he claims that the immutable nature encountered in the idea constrains correct judgement.

The analogy distinguishes invention from discovery. A thinker cannot decide that a triangle has four angles, and Descartes says the divine idea similarly presents necessary relations not subject to will. Yet geometry normally establishes conditional truths: if a triangle exists, it has the relevant properties. It does not establish that a triangle is instantiated outside thought. The critic argues that God's concept likewise yields at most: if a supremely perfect being exists, it exists necessarily.

Descartes replies that necessary existence differs because it is itself among the perfections. The response makes the predicate issue decisive. If existence is not a property contributing to a thing's qualitative content, it cannot be included alongside knowledge and goodness. If necessary existence instead means independence and impossibility of non-being, it is a mode of exemplification which may distinguish a divine candidate from contingent objects without behaving like an ordinary quality.

There is also a circularity concern in the wider Meditations. Clear and distinct perceptions are secured by a non-deceiving God, while a clear and distinct idea helps establish God. Descartes distinguishes the immediate certainty of a present intuition from remembered reasoning, but the dependence remains debated. The ontological argument is most effective inside his rationalist system, where essences can be intellectually perceived. It is less compelling to an empiricist who treats definitions as human classifications lacking existential import.

Kant argues that being is not a real predicate: it does not add a determination to a concept. One hundred real thalers do not contain one more conceptual coin-property than one hundred possible thalers. The difference is that the concept is instantiated in reality. To assert existence is to posit an object corresponding to the concept, not to enlarge the object's list of qualities.

In modern logical form, "dragons exist" does not assign the property existence to a dragon already identified. It says that at least one object satisfies the dragon-description. If nothing does, there is no object waiting to possess all properties except existence. This diagnoses the Chapter II comparison: a merely conceived God is not a lesser being, but a concept with no instance. No increase in descriptive greatness can quantify an instance into being.

Necessary existence requires a subtler response. It may not be a first-order quality, but a proposition's modal status can still differ: some truths hold in every possible world and others in only some. Modal defenders therefore concede much of Kant's ordinary-existence objection and argue that maximal greatness includes necessary exemplification. Kant can reply that necessity belongs to judgements, not things considered outside conditions of possible experience, and that no existential proposition is analytic.

The objection decisively blocks a crude "existence is better" argument. Whether it defeats every ontological argument depends on whether modal versions repeat that error or shift to possibility and necessity. If the latter, the battle moves to the coherence and possibility of maximal greatness. Kant remains relevant because one cannot move from a complete qualitative description to instantiation without an existential premise. Necessary-existence language must not conceal that premise inside a definition.

Norman Malcolm distinguishes two ontological arguments in the Proslogion. He accepts that Chapter II is vulnerable if it treats existence as a perfection, but reads Chapter III as concerning necessary existence. God's existence cannot be contingent: an unsurpassable being would not depend on causes or circumstances which might have failed. Therefore God's existence is either impossible or necessary. If the concept is not self-contradictory, necessary existence follows.

This reconstruction answers the claim that God could happen to come into or go out of existence. A being produced by another or vulnerable to non-being would not be the independent object of worship. Necessary existence is not one more admirable feature like courage; it expresses the absence of dependence. Malcolm therefore avoids directly ranking a real object above an imaginary counterpart.

The weakness is the transition from "not impossible" to "possible" and then to actuality. Showing that no contradiction has been found is not always showing positive metaphysical possibility. Hidden tensions among omnipotence, necessary goodness, freedom, knowledge of contingents and timelessness may remain. Furthermore, a critic can formulate symmetry: if a necessary God possibly exists, God exists in every possible world; if it is possible that no necessary God exists, there is no such God in any. Modal logic makes the possibility judgement more, not less, important.

Malcolm's version is therefore valuable as clarification. The classical God is not proposed as an unusually durable contingent entity. The argument forces critics either to challenge the coherence of that concept or deny that conceivable coherence establishes possibility. It does not compel a neutral thinker unless the possibility premise has support independent of the desired conclusion. Its achievement is to locate the real disagreement in modal metaphysics rather than in a fanciful comparison with islands.

Gaunilo, a fellow believer, speaks “on behalf of the fool”. His lost-island parody is not merely “imagine an island and it appears”. It tests the inference pattern: if conceiving the greatest conceivable instance and treating reality as greater were sufficient, parallel reasoning would establish an island whose excellence guarantees its existence. Since no definition manufactures an island, something has gone wrong. The parody pressures Anselm to show why God is unique rather than exempting the conclusion by stipulation.

Several further objections follow. Words can be held in the mind without a clear object; a report or piece of gossip may be entertained and even believed without being true. Different people may also lack a shared, adequate conception of God. The painter knows what will be painted because the image is derived from a possible artefact and a plan to cause it; the fool has no comparable route from concept to divine reality. Finally, Gaunilo says God's non-existence appears as conceivable as one's own, challenging Chapter III at its modal starting point.

Anselm's central response is that the island is the wrong kind of thing. Its excellence is indefinite and contingent, whereas God is unsurpassable and necessary. That response is effective if necessary existence is independently established as part of coherent maximal greatness. Otherwise it risks circularity: the parody fails because God is necessary, while God's necessity is what the argument was supposed to prove. A balanced verdict distinguishes Chapter II, against which the parody has intuitive force, from Chapter III, which narrows the relevant comparison but inherits a demanding modal premise.

Aquinas does not deny that God's essence, if known adequately, would include existence. He denies that finite human knowers comprehend that essence directly. “God exists” may be self-evident in itself but not self-evident to us. We therefore reason from effects in the world, as in the Ways, rather than begin with a definition available to every mind. This is an epistemic objection: Anselm assumes knowledge of what God is which Aquinas thinks unaided human reason lacks.

Kant offers a logical objection: existence is not a real predicate which adds a feature to a concept. One hundred real thalers do not contain more conceptual content than one hundred possible thalers; the difference is that the concept is instantiated. Denying a triangle together with its angles creates no contradiction because rejecting the subject removes its predicates. Likewise, “God does not exist” need not attribute and withhold existence from one object; it can deny that anything answers to the concept.

A modal defender can reply that necessary existence is not an ordinary first-order predicate like colour. It concerns the way a being exists across possible worlds. Kant's point still applies at the level of instantiation: analysing a concept tells us what anything satisfying it would be like, but why think the concept is possibly instantiated? The argument's deepest value may therefore be conditional. It shows that classical God, if possible and genuinely maximally great, is not a contingent optional inhabitant of reality. Whether such a being is possible remains a substantive metaphysical question.

Alvin Plantinga defines maximal excellence in a possible world as omnipotence, omniscience and moral perfection there. Maximal greatness is maximal excellence in every possible world. The argument then proposes that it is possible that a maximally great being exists. If such a being exists in any possible world, it exists necessarily and therefore in every possible world, including the actual world. In modal system S5, possible necessity entails necessity.

The inference is formally powerful. It does not compare mental and real objects, and it does not use ordinary existence as a property. It also shows why the conclusion cannot be "God probably exists". Given the stipulated kind of being and S5 accessibility, either the possibility premise is true and necessary existence follows, or the premise fails. The argument is a test of rational acceptability of divine possibility.

Parodies must match the modal structure. A maximally excellent island is not necessarily instantiated because islands are essentially contingent and lack intrinsic maxima. More serious reverse arguments propose the possible non-existence of a necessary perfect being. They reveal that competing modal intuitions cannot both be used without support. One side must establish which world-description is genuinely possible rather than merely grammatically expressible.

Plantinga himself presents the reasoning more modestly than a coercive proof: it can show that belief in God is rational if the possibility premise is rational. Evidence may come from coherence, religious experience or a broader cumulative case, but then the argument is not independent of all experience in practical persuasion. The best judgement is that validity is not the problem. Soundness turns on a premise which a theist has reason to accept and a critic has reason to withhold, so the proof clarifies commitments more readily than it converts opponents.

Ontological arguments require more than a dictionary definition. Maximal greatness must describe a metaphysically possible nature. Omnipotence is normally refined as power to do all that is logically possible, because making a square circle is not a task awaiting more strength. Necessary moral perfection then raises whether God is praiseworthy or free if wrongdoing is impossible. Timeless knowledge raises how an immutable knower knows changing tensed facts. Each refinement aims to preserve a coherent maximum.

Some apparent conflicts result from imprecise concepts. Foreknowledge need not cause a free action; inability to do the logically contradictory is not weakness; justice and mercy can apply under different descriptions. Other tensions are substantive. If God necessarily chooses the best possible world, divine freedom appears constrained; if several equal worlds are available, a sufficient reason for this choice is unclear; if no best world exists because improvement is always possible, unsurpassable creative action becomes difficult to state.

A defender need not solve every theological puzzle before regarding the concept as possible. Ordinary modal reasoning often treats absence of a demonstrated contradiction as provisional support. The critic replies that extraordinary necessary-existence conclusions require more than provisional conceivability, especially because human concepts may hide inconsistency. Positive models of how attributes relate strengthen the premise more than bare assertion that God is "by definition" coherent.

This makes the topic connect with attributes of God rather than ending at formal logic. If a coherent maximally great being is genuinely possible, modal reasoning yields a dramatic conclusion. If the concept is impossible, no definition can create an instance. If its status is epistemically undecided, the argument remains conditional. A mature essay can therefore conclude that ontological reasoning identifies the exact hinge—divine possibility—without pretending that formal validity settles that hinge.

Possible-world language does not propose distant planets. A possible world is a complete way reality might have been. A proposition is necessary if true in every possible world, possible if true in at least one, and contingent if true in some but false in others. The S5 system treats accessibility among worlds symmetrically: any genuinely possible world is relevant from any other. Under S5, if a proposition is possibly necessary, it is necessary. If there is a possible world in which a maximally great being exists necessarily, that being exists in all worlds.

The controversial step can be shown with a contrast. It is possible that a contingent inventor exists, but this does not place the inventor in the actual world because existence in the selected world carries no claim about other worlds. A maximally great being is defined to possess necessary existence, so its presence in one world reaches every world. The dramatic result comes from combining the possibility premise with the modal status of the candidate, not from moving an object physically between worlds.

Critics challenge both S5 and the premise. Weaker modal systems do not automatically validate possible necessity as necessity. Even within S5, a reverse claim can be framed: if it is possible that no maximally great being exists, then a necessarily existing maximally great being is impossible. The two possibility assertions cannot both describe genuine worlds. Intuition alone supplies no neutral rule for choosing between them.

This symmetry is intellectually useful. It shows that the ontological debate is not resolved by placing "possibly" before a theological sentence. Positive reasons are needed to regard maximal greatness as coherent and possible, while positive incompatibility arguments can support impossibility. Modal notation sharpens where disagreement lies; it does not remove the need for metaphysical judgement. Students can therefore use S5 to explain validity without presenting a contested axiom system and premise as universally compulsory.

Anselm's original title and prayer place the reasoning within "faith seeking understanding". Beginning from worship, he asks what must be true of the one addressed as unsurpassable. This can refine theology even if a non-believer is not compelled. A deity who depends on a greater cause, accidentally acquires goodness or can simply cease to exist would not be the ultimate object Anselm intends. Necessary existence expresses the radical non-dependence implicit in that practice.

As internal reasoning, the argument tests consistency between beliefs. Someone who affirms an unsurpassable God while describing God as contingent holds unstable commitments. The Proslogion can expose and repair that instability without deriving a being from an empty concept. This role resembles conceptual analysis in ethics: defining justice does not create just institutions, but it can reveal contradictions in a political theory.

The limitation is apologetic. Internal clarification cannot be advertised as a premise-free demonstration to anyone whatsoever. If the maximality framework is warranted by revelation or religious experience, existential support has entered from outside pure reason. That does not make the reasoning worthless; it changes its audience and conclusion. The proof may show what classical theism entails more securely than it shows that classical theism is true.

A balanced judgement can therefore distinguish three achievements: formal validity, metaphysical soundness and theological illumination. A version may be valid yet depend on a disputed possibility premise; it may fail to persuade a neutral critic yet illuminate why believers call God necessary; or it may reveal that a proposed divine attribute set is incoherent. Assessing these separately is more precise than declaring that the argument either "proves God" or has no value.