Omnipotence

The Research Problem and Its Philosophical Stakes

Omnipotence enters philosophical inquiry wherever an attribution of ultimate power is offered as an explanation of reality or as an essential feature of divinity. Traditional Western theism normally treats omnipotence as one of the attributes of a maximally perfect God. Yet an attribution of maximal power is intelligible only when the nature of power, the objects of exercise, and the standards of comparison have been fixed. An ordinary claim that someone can perform every task is plainly false of every finite person; a philosophical claim that an agent possesses every power can fail to express even a consistent candidate property unless the range of relevant tasks is specified.

The question has three logically distinct forms. The semantic question asks what 'omnipotent' means. The metaphysical question asks whether anything could have the property so defined. The theological question asks whether the being described by a particular religion actually possesses it and whether that attribution is compatible with further doctrines. A definition does not establish metaphysical possibility, and a proof of possibility would not establish actual existence. Conflating these levels is responsible for many superficial victories in debates over divine power. This study keeps them separate while investigating the relations that motivate each inference.

The classic challenges have a remarkable internal diversity. The ability to make a square circle asks whether contradictions are feasible. The power to change the past asks whether the object of the proposed operation is metaphysically available once an event has occurred. The stone too heavy for its maker to lift tests the interaction of an agent's power with a self-referential description of the agent. The power to sin tests a relation between ability and moral defect. Two beings attempting opposed outcomes test the compatibility of simultaneous supreme agencies. The problem of evil adds moral expectations about how the allegedly supreme agency would exercise power. The issues are connected without being interchangeable.

The proper research object is therefore omnipotence as a structured philosophical predicate, rather than a slogan of unrestricted ability. This means examining the analytical roles of power, action, opportunity, agency, modality, history, and excellence. It also means taking both the strongest classical defenses and the strongest contemporary objections seriously. The central question is: under what noncircular conditions could a power count as maximal, and which conditions make that property fit or fail to fit a larger religious conception?

Primary evidence considered here includes accessible English translations of Anselm's Proslogion VII, Maimonides' Guide for the Perplexed III.15, Thomas Aquinas's Summa Theologiae I.25, Augustine's City of God V.10 and Leibniz's Theodicy; Descartes' 1630 correspondence is approached through cited historical scholarship. Contemporary arguments are read directly where the author's paper or open scholarly reference is available (particularly Flint and Freddoso, Hoffman and Rosenkrantz, and Wreen 2022). Publisher abstracts are used at abstract scope for Lembke 2012, Wreen 2018, Morriston and selected other paywalled discussions. The work marks this evidentiary distinction rather than attributing unavailable details to a source.

Omnipotence as a Concept of Power

An ability is normally an agent-relative capacity to perform an action under appropriate circumstances. A power is often understood more broadly as a capacity together with the effective opportunity to exercise it; a person might retain the skill to swim while imprisoned where there is no water, but that competence is not an opportunity to swim. The Stanford Encyclopedia of Philosophy distinguishes power from bare ability in precisely this connection (Hoffman and Rosenkrantz, 'Omnipotence'). A theory that speaks of an agent's total capacities without specifying its circumstances may evaluate latent abilities; a theory of effective sovereignty must consider whether anything can obstruct their exercise.

The adjective 'maximal' admits more than one interpretation. It may mean that every individual action-type is within the agent's repertoire; that the agent can bring about every possible outcome; that no other agent can exceed the agent's total power; or that the agent's power is essential and unlimited throughout its existence. These are different properties. A being able to produce countless microscopic variations might have infinitely many possible effects while remaining unable to move a large object. Numerical infinity of outcomes does not settle comparative superiority. Likewise, being stronger than every actual rival leaves open whether a stronger possible rival could exist.

The domain of power also matters. Physical force is measured by parameters such as work or energy only within a physical model. The concept of divine creation, if coherent, concerns bringing about the existence of a world, an action whose supposed object is not simply a physical body already available for acceleration. Moral permission is another dimension: an agent's power to act cannot be inferred from the moral permissibility of the action, and an agent's refusal to act is not immediately evidence of incapacity. The philosophical investigation must give each relation its own criterion.

A state of affairs is an obtaining or nonobtaining condition, often represented by a proposition. On a result-centered analysis, to have power with respect to a state of affairs is to be able to bring it about, prevent it, or maintain it. Action-centered analyses instead ask whether a specified action is executable by the agent. These frameworks overlap, but they diverge when a task description includes the history, identity, weakness, or essential nature of its performer. 'A human being learns that they are fallible' may describe a feasible action, while 'an essentially infallible agent learns that it has always been fallible' incorporates conditions incompatible with the agent's characterization.

An adequate analysis must avoid vacuity. Defining an omnipotent being as one that can do everything it can do provides no test of maximality. Defining omnipotence as the power to do absolutely anything, including contradictions, produces a striking formulation whose logical content needs independent defense. Defining it merely as being unsurpassed invites problems of incomparability: one agent may control every physical event while another alone can determine whether a distinct class of facts obtains. If neither capacity includes the other, a unique maximum need not follow from a partial ordering of power profiles.

There is a further temporal question. Does omnipotence belong to an agent at a moment, during an interval, throughout its life, or essentially in every possible world in which it exists? A transiently unlimited power is distinct from a necessarily everlasting supreme power. An objection against necessary omnitemporal omnipotence may have little bearing on momentary omnipotence, though religious doctrines often require the stronger claim. The Internet Encyclopedia of Philosophy rightly emphasizes this logical difference in assessing the stone paradox. A general account of power must say which temporal predicate it analyzes.

Method and Evidentiary Architecture

The monograph uses historical reconstruction, close examination of accessible primary passages, conceptual analysis, informal modal reasoning, and comparison of explicitly stated necessary and sufficient conditions. Its object is a philosophical concept and associated arguments; no experimental study is being reported. Historical claims are tied to identifiable texts and passages. Modern arguments are assessed by their logical form, the presuppositions they make, and how their results vary under competing accounts of possibility and action.

Three distinctions govern source assessment. First, direct public access to a historical English translation is different from collation of a critical original-language edition. An English passage can support a careful reconstruction of its displayed argument while leaving textual variants and difficult translation choices for later review. Second, a publisher's abstract establishes the author's advertised claim and bibliographic identity, rather than independently validating every step of the full argument. Third, a scholarly encyclopedia's reconstruction supplies valuable context, but its interpretation must be distinguished from the primary writer's own wording. These levels are reflected in the editorial evidence accompanying this dossier.

The study also deliberately restricts its cross-cultural scope. 'Omnipotence' in its modern analytic sense is principally a predicate in arguments about the Western personal God, and the primary corpus here follows that trajectory. Jewish and Islamic philosophical resources are directly represented by Maimonides and the larger medieval discussion. Traditions in which ultimate reality is conceived impersonally, or where cosmic agency is distributed, demand different criteria of comparison; they should not be reclassified as failed versions of Western perfect-being theism. A genuinely comprehensive comparative theology would require additional primary-language work beyond this monograph's evaluated evidence.

Finally, the Artificial Era is a horizon of inquiry into forms of agency, not a warrant to transfer theological omnipotence to new technical systems. Aisentica's distinction between Homo and Artificial can sharpen the methodological requirement to specify the bearer and domain of an attribution of power. It provides no reason to call technological systems omnipotent. This research keeps the philosophical analysis of divinity central, and uses the wider issue of agency only when a general conceptual result genuinely follows.

The Ancient and Patristic Background

Ancient philosophy supplied notions of potentiality, actuality, necessity, causation, and divine perfection without producing one uniform formal doctrine of omnipotence. The Aristotelian distinction between dynamis and energeia explains how an agent may possess a capacity without continually exercising it. It also warns against identifying every conceivable outcome with an actual causal power. From this standpoint a power is directed toward an intelligible object of operation, and the object's conditions matter to the identity of the power. This conceptual inheritance remains evident when medieval theologians ask whether contradictory descriptions correspond to possible effects at all.

The biblical and patristic language of the almighty adds a different dimension: universal dominion and providential freedom. A ruler of all that exists is not automatically identical to a performer of every grammatically describable action. P. T. Geach's twentieth-century differentiation between 'almighty' and 'omnipotent' underscores the historical significance of this contrast. Claims about divine lordship, causality, and worship may require sovereign efficacy without resolving an artificially sharpened modern demand to execute every task whatsoever.

Augustine provides a foundational treatment in City of God V.10. Defending the compatibility of providential foreknowledge and freedom, he argues that divine inability to die or err does not diminish divine power. He describes divine omnipotence in terms of doing what God wills, while the capacity to suffer unwillingly would indicate subjection rather than supreme agency. This is an argument about the relation of power to deficiency: certain human abilities arise from vulnerability and are inappropriate measures of divine greatness. The claim is stronger than the observation that God happens to avoid error; the inability to err is presented as integral to the perfection attributed to God.

That proposal also raises a problem that subsequent philosophy cannot evade. If the power to act wrongly is excluded from the list because the ideally powerful agent is essentially good, the criterion of what counts as power now depends on a theory of perfection. A critic can coherently inquire whether this makes the account circular. One reply is that an ability to deteriorate from an ideal condition is not a pure increase in an agent's effective control. Another is that the set of options an agent can realize still becomes narrower even if the narrowing has a morally desirable explanation. Both considerations have substance; identifying one as a conceptual truth requires a further argument.

Augustine's discussion anticipates a general methodological distinction: an inability may concern a failure of resources, or it may concern the incompatibility between an act description and the form of agency under discussion. The former is a direct deficit of power; the latter can be an incoherent comparison if the excluded performance requires losing the agent's defining power. Yet this distinction must be applied carefully. A rule that dismisses every counterexample as contrary to divine essence would make omnipotence untestable. The subsequent tradition is best read as a continuing attempt to determine principled boundaries.

Anselm and the Difference Between Power and Defect

In Proslogion VII, Anselm asks how God can be omnipotent while unable to be corrupted, lie, or reverse a completed event. His answer reinterprets the ability to undergo corruption as a kind of weakness. Certain capacities express an agent's exposure to misfortune or moral failure rather than superiority. The paradox therefore results in part from an equivocation in 'can': being liable to an adverse event is verbally an ability, but its possession may reduce rather than increase power (Anselm, Proslogion VII, English version in the Christian Classics Ethereal Library).

Anselm's move introduces a normative comparison into a modal problem. In ordinary human contexts the power to lie may indeed provide an instrumental option even though lying is wrong. A person physically capable of lying can choose not to do so. For an essentially truthful being, however, the corresponding 'ability' may not be an executable option compatible with that being's essence. Anselm's proposal therefore needs a defensible account of how power is individuated and ranked, not merely the linguistic substitution of 'weakness' for every unwanted case.

His answer also connects to the broader Anselmian understanding of divine greatness. If divinity is understood through unsurpassable greatness, then maximal power is one member of a network including perfect knowledge, goodness, and immutable excellence. Questions about omnipotence can no longer be decided by counting options in isolation. An agent with more sheer options may be inferior under a total assessment of greatness. The argument will later reappear as an explicit dispute about whether maximal power and the greatness of a maximally perfect being must be the same property.

These early sources supply two durable insights. Power attribution must distinguish active capacity from susceptibility, and it must identify whether the test measures sheer possibility of acts or a normatively specified perfection. The crucial open question is how to formulate those distinctions in a way that allows substantive, potentially adverse testing. The historical tradition does not eliminate that burden; it gives the philosophical debate its most precise starting point.

Maimonides and the Status of the Impossible

Moses Maimonides offers a direct boundary claim in Guide for the Perplexed III.15: the impossible as such is not a thing that an agent could produce. This statement rests on the view that a contradiction fails to describe a coherent object of power. The accessible Friedländer translation identifies examples involving a single substratum's possession of opposed properties in the same respect. Maimonides also notes that philosophers may disagree about which particular states are impossible. That qualification is essential. Agreement that impossibilities cannot be achieved does not settle the controversial metaphysics determining what is possible.

There are at least two defensible interpretations of this argument. On the first, logical impossibility is conceptually prior to any exercise of causal power. An expression such as a square circle fails to specify an achievable effect under the ordinary meanings of its constituents. On the second, the nature of reality imposes the relevant structure of possibility, and our logical descriptions are evidence about that structure but may not exhaust it. The first yields a semantic diagnosis; the second makes a metaphysical claim. Maimonides' own text invokes what is impossible as a constant condition, while acknowledging dispute over its extension. It therefore supplies an especially useful warning against pretending that a list of examples proves a complete modal theory.

A further issue concerns the direction of explanation. If an impossibility is not a feasible effect, saying that God cannot produce it adds nothing to an inventory of limitations. If the impossibility is instead a substantive metaphysical boundary independent of God, one may ask whether the theology has admitted a reality prior to or above divine will. That question motivates voluntarist positions. But the claim that logical principles are not independent objects superior to God remains distinct from the claim that contradictions can literally obtain. The dispute is partly about explanatory priority, not merely a test of technical performance.

Maimonides' contribution is historically significant beyond the single paradox. The argument belongs within a systematic Jewish philosophical project shaped by Aristotelian thought and theological commitments, and its careful formulation of unachievable objects became part of a wider medieval field of modal analysis. A description of medieval philosophy as a straightforward rejection of unlimited power would distort that tradition. The central work is to specify what the word 'all' legitimately quantifies over.

Aquinas and Absolute Possibility

Thomas Aquinas's Summa Theologiae I, question 25, is one of the clearest primary sources for the classical account. Article 3 explicitly asks whether God is omnipotent and rejects the definition that God can do whatever lies within God's power because the resulting circularity explains nothing. Aquinas distinguishes what is possible relative to a particular creature's power from what is possible absolutely. A person may lack the strength to move an object even though moving it is absolutely possible. The scope of divine power is then connected to the absolutely possible, characterized through the compatibility of a proposed subject and predicate.

The contradiction example is decisive. An effect that would require a thing both to be and not to be in the same respect does not fall within the scope of any power, since it lacks the character of a possible effect. Aquinas therefore does not construe the law of noncontradiction as a stronger agent resisting God. He construes the impossible description as failing to designate a producible reality. This is a positive theory of action's objects: every active power has a correlative effect-type, and the range of absolute divine power is determined by the entire range of effects that could have being.

The argument has substantive strengths. It avoids defining omnipotence by appeal to itself; it distinguishes natural incapacity from general possibility; and it supplies a reason why contradictions are excluded from the domain. It also encounters difficulties. Predicative noncontradiction is at best an initial test of possibility. Some coherent descriptions may conceal metaphysical impossibilities involving identity, essence, or incompatible conditions. Conversely, in unusual logics some descriptions with classically contradictory forms may be treated differently. The philosophical adequacy of Aquinas's account depends on the relation between intelligibility, being, and logical consequence.

Article 4 addresses the past. Aquinas denies that God can make what has been done never to have been done. This is not an account of the weakness of God against time as an external stronger force. It is connected to the conditions under which a proposition about an already actual event is assessed. To bring about that a past event did not occur while continuing to affirm that it did occur would demand incompatible historical states. Yet modern debates about backward causation and alternative histories distinguish several different tasks hidden in ordinary talk of 'changing the past.' Aquinas's conclusion should be located in the medieval conceptual framework that motivates it rather than treated as an experimental finding about physical time travel.

Article 5 draws a further distinction between absolute and ordained divine power. Considered absolutely, divine power ranges over all effects that could have being; considered in relation to a wise will and established order, it is described according to what God actually wills and executes. This distinction allows theologians to explain why a state of affairs that does not happen may nevertheless have been within divine power. Actual history is not automatically coextensive with divine capability. The distinction later acquired multiple technical forms across scholastic traditions and cannot be equated with the same doctrine in every author.

Article 6 asks whether God could make things better than they are. The question illustrates the importance of specifying comparison classes. An agent's power to produce another arrangement does not by itself imply that every possible object can be improved while preserving its identity and nature. Better worlds, better individual things, and better realizations of a chosen plan require different evaluative metrics. The philosophical achievement of question 25 is not a final elimination of paradoxes; it is a disciplined analysis of the relation among potency, possibility, will, actuality, and excellence.

The Medieval Difference Between Absolute and Ordained Power

The distinction commonly labeled potentia absoluta and potentia ordinata should not be treated as an authorization for two different gods or two competing sources of power. It concerns ways of describing the one divine power relative to what could be done and what has been willed within an ordered economy. Its historical articulations vary. Some medieval theorists emphasize alternative possibilities under divine freedom; others work through dependence, causal order, or the limits imposed by logical and metaphysical identity.

This variation matters to contemporary discussions of miracles and laws. A claim that God can act otherwise than under the ordinary causal order is not the same as a claim that contradictory propositions can both be true. Natural laws may be contingent within a theological theory even if logic and identity retain their modal status. An account of absolute power must therefore identify which background conditions it suspends and which it preserves. Without such specificity the phrase becomes a rhetorical substitute for a theory of possibility.

The medieval tradition also complicates claims about divine freedom. If a perfectly rational God necessarily chooses the best, is there genuine alternative power? A distinction is required between being able to realize an alternative and being able, consistently with every essential attribute, to will its realization. Some accounts locate freedom in the absence of external compulsion even where a perfectly rational decision is necessary. Others require alternate possibilities. This problem anticipates the contemporary compatibility dispute between unlimited power and necessary moral perfection.

Historical attribution demands attention. One should not ascribe a complete modern possible-world semantics to Aquinas or Maimonides merely because they discuss contingency. Contemporary reconstructions can clarify what their arguments would imply under specified modal assumptions, but those reconstructions are subsequent philosophical developments. In the history of omnipotence, the operative problem changes with the underlying theory of modalities.

Descartes and the Creation of Eternal Truths

René Descartes gives the classical analysis its most difficult rival. In his 15 April 1630 letter to Marin Mersenne, he affirms that mathematical truths commonly called eternal depend on God; the 27 May letter reiterates the dependence in terms of divine efficient causation. The historically established thesis concerns God's creation or establishment of eternal truths, including mathematical truths. It follows neither automatically nor trivially that contradictions can be true within the logic of Descartes' own philosophical system. Scholars continue to debate the interpretation of his theological voluntarism, the role of human cognitive limitations, and the status of counterpossible descriptions (Descartes' correspondence, AT I:145, 151-152; SEP, 'Descartes' Modal Metaphysics').

An intuitively powerful objection to Aquinas asks whether treating necessary truths as fixed independent constraints makes them a superior framework within which God must operate. Descartes' reply refuses to picture logic as a separately existing authority to which a creator submits. The conceptual move is radical: the deepest necessities known to us depend on the ultimate creative source. Yet the difference between explanatory dependence and possible alteration is decisive. Even if a necessary proposition depends on God for its truth, it does not follow without argument that there exists a coherent possible world where that proposition is false.

The popular characterization that Descartes simply lets God make a square circle is therefore too crude unless accompanied by a careful account of what 'can,' 'could have,' and 'truth' mean in the cited correspondence. Our human inability to conceive of a different mathematics might be epistemic; alternatively, the supposed counterpossibility might collapse into an incoherent description. Descartes' position tests a boundary between epistemic humility about the scope of reason and a substantive metaphysical claim of divine freedom over necessity.

This dispute matters for the foundations of reasoning itself. If a premise such as the law of noncontradiction may fail in circumstances realized by omnipotence, an argument that omnipotence is coherent must explain which consequence relation it employs. A proof that appeals to standard modal logic while granting unlimited power to overturn that logic risks depending on principles its own thesis destabilizes. Voluntarism can remain a serious metaphysical option, but it carries a burden of supplying a stable inferential metalanguage in which claims about it can be evaluated.

Leibniz and the Rational Ordering of Possible Worlds

Gottfried Wilhelm Leibniz's Theodicy (1710) places divine power within a rational structure of possibles, divine wisdom, and the selection of a world. The availability of other possible worlds establishes a distinction between the capacity to realize an alternative and the judgment that makes one alternative preferable. Divine power is accompanied by intelligence about possibilities and goodness in selecting among them. The resulting account has immense consequences for the problem of evil and for the freedom of divine creation.

A classical Leibnizian claim is that God chooses the best possible world. One must not infer that omnipotence alone produces a best world. The inference requires a ranking of worlds, the existence of an optimum under that ranking, and an account of divine rational choice. If the admissible worlds are not well ordered by value or no maximum exists, the demand to create 'the best' may fail to identify a possible performance. For this reason the problem of no best world is distinct from the problem of creating any possible world.

Leibniz also confronts an important form of modal consistency. Not every collection of separately possible things is jointly compossible. Two complete histories may differ in incompatible propositions. The power to actualize each history under distinct alternatives is not the power to actualize both at once in a single consistent history. This helps explain why the total range of possible outcomes should be modeled as a structured domain of alternatives rather than a bag of independently realizable isolated propositions.

A further issue concerns choice and necessity. A being that always chooses the best can be understood as necessarily wise without being externally constrained; critics object that if the choice is fully determined by the uniquely best alternative, the open possibility of doing otherwise disappears. The controversy parallels Anselm's account of perfect agency, but now operates through a theory of world values. A theory of omnipotence that simply says God can actualize all possible worlds must confront the possibility that not all worlds can be actualized given the same divine nature and intentions.

The conceptual achievement of early modern philosophy is thus a transformation of the original question. Aquinas asks what may have being as the object of action. Descartes asks whether necessary truths themselves depend on divine will. Leibniz asks how worlds of possibility are ordered for a rational chooser. Each inquiry reframes the quantifier 'all.' An agent may possess all power over a domain defined by a modal structure, while that structure itself remains controversial.

From Power Over Things to Power Over States of Affairs

Twentieth-century analytical philosophy increasingly replaces concrete task descriptions with propositional states of affairs. The shift is useful because it provides a general language for effects, necessities, contingent histories, and conditions of agency. An agent's power can be assessed with respect to a state that obtains, fails to obtain, or depends on choices. But this abstract language has its own hazards: the proposition expressing that an agent was never omnipotent may be possible for some agent, yet impossible for a necessarily omnipotent one to make true of itself.

The theory of states of affairs therefore requires indexation. Is the state possible simpliciter, possible relative to a particular history, possible for an agent of a given nature, or possible for any agent whatsoever? A definition of omnipotence based on 'all logically possible states of affairs' is underspecified if the relevant predicates refer to the performing agent and the time of action. Introducing states of affairs improves precision only when these hidden parameters are made explicit.

The contrast between 'can bring about' and 'can coexist with' is also essential. An agent might coexist with a mathematical necessity without bringing it about. An agent may bring about a contingent condition by creating a relevant causal antecedent without controlling every independent aspect of the resulting world. A proposition describing a contingent history may be true in a possible world that is inaccessible to the agent from the actual past. These distinctions prepare the analysis of self-referential paradoxes and modern shared-history accounts.

The Stone Paradox and Reflexive Tasks

The familiar formulation asks whether an omnipotent agent can create a stone so heavy that the agent cannot lift it. A negative answer seems to identify an undone task: creating that stone. An affirmative answer seems to identify a resulting limit: lifting it. The argument works rhetorically because it creates an apparently exhaustive dilemma. Philosophically, its power depends on the precise scope of the task description and on whether omnipotence is required to persist after any conceivable act.

Let S be a being and R a stone. The task may be expressed as: S creates R and, as a consequence of R's properties, S cannot lift R. Under a description in which S remains omnipotent throughout, the completed task appears to entail both continued maximal lifting power and a limit to that power. The description itself may thus be inconsistent relative to the intended agent profile. This is the point associated with George I. Mavrodes' 'Some Puzzles Concerning Omnipotence' (1963): an apparent performance test can specify a self-contradictory pseudo-task. The existence of a grammatical description does not by itself establish a genuine object of power.

A complication emerges as soon as we distinguish necessary from temporary omnipotence. Imagine an agent omnipotent at time t1 which creates at t2 a state that removes or weakens its own power. The agent's ability to do this at t1 may be consistent with a definition of temporally local omnipotence, while undermining the doctrine that the agent is essentially omnipotent at every time. The task therefore tests different attributes depending on the temporal index. A conclusion valid against necessary omnitemporal omnipotence should not be silently generalized to a merely momentary version. The IEP discussion makes this distinction explicit.

Harry Frankfurt's 'The Logic of Omnipotence' (1964) responds to Mavrodes by pressing a voluntarist alternative. If an omnipotent agent can perform what ordinary logic classifies as contradictory, it might both make the apparently unliftable stone and lift it. This reply reveals that the argument against omnipotence cannot simply assume without discussion that standard consistency conditions govern divine action. However, the voluntarist reply then assumes a special explanatory burden: what counts as an entailment, an action, or the same stone in a framework where logical contradiction can be actualized?

There is also a danger of switching between a sentence expressing an individual task and a sentence expressing a conjunction of tasks. 'Create a stone too heavy for its creator to lift' is possible for an ordinary finite maker. 'An essentially omnipotent creator creates a stone it cannot lift while remaining essentially omnipotent' is a different proposition. The second may be impossible without implying any lack in the creator, but one needs a principled account of why essential omnipotence belongs in the task's background. Reclassifying everything a candidate cannot do as conceptually incompatible with its essence would beg the question.

The stone puzzle accordingly teaches four lessons. A test's hidden agent-reference must be made explicit; the relevant time must be specified; joint performability cannot be inferred from the separate possibility of each component action; and the standard of logical possibility must be identified. The argument remains instructive even for theories that ultimately dissolve its contradiction. It exposes the formal architecture that any adequate analysis of maximal power must possess.

Contradictions, Logical Consequence, and Metaphysical Possibility

A square circle is a standard example of a contradictory description, understood with the ordinary geometrical meanings of the words. On a classical analysis a thing cannot be both square and circular in the same plane, respect, and sense. The impossibility differs from an extremely difficult task: immense but finite force does not help, because the obstacle is not a shortage of physical energy. The question is whether the description identifies any realizable condition.

Logical possibility and metaphysical possibility should nevertheless be distinguished. A proposition can be consistent with a restricted formal system yet be metaphysically impossible due to essential identity conditions. Conversely, ordinary language may seem contradictory because it compresses distinct times, respects, or senses into one phrase. 'This is both hot and cold' can describe a composite object with different temperatures in different regions. Before invoking a logical limit of power, the analyst must disambiguate the alleged object.

Formal logic makes the issue transparent. In standard modal reasoning, if p is impossible, then not-possibly-p holds under the given interpretation. To require an agent to realize p would require possible success at a condition excluded from every admissible alternative. This creates a contradiction inside the explanatory framework, independently of the agent's physical magnitude. Whether the framework itself is the correct metaphysics is a further philosophical question. It is a category mistake to infer a discovery about divine existence from a chosen semantic convention alone.

The historical voluntarist objection raises the meta-level. If divine agency creates every necessary truth, why should logic govern that very agency? One reply denies that a necessity is an independent authority or object and claims that logic articulates the conditions of any coherent proposition about action. Another rejects the independence of logical constraints and grants divine sourcehood over necessity. Both can be stated, but the second must specify what inferential rules remain available when debating its own coherence. Without that account, claims of a contradiction's possible truth are at risk of becoming immune to evaluation rather than intelligibly powerful.

Nonclassical logics complicate the rhetoric further. Paraconsistent consequence relations can prevent an inconsistency from entailing every sentence. That mathematical fact alone does not show that an actual square circle is metaphysically possible or that divine power realizes contradictions. A choice of consequence relation is a formal resource; a metaphysics of impossible objects requires additional commitments. An informed analysis identifies the logic in use and what the theory purports to say about the world.

Necessary Truths and the Problem of Bringing About

Suppose that an agent is omnipotent and that the arithmetic proposition 2 + 2 = 4 is necessary. Can the agent bring about that proposition? It is tempting to answer yes because the proposition holds everywhere, but that response confuses an agent's presence alongside a truth with its causal production. Normally to bring about s is to make a difference to whether s obtains or occurs, relative to some relevant alternative. Necessary truths do not admit alternatives under the necessity relation that makes them necessary.

Aquinas avoids one version of the problem by requiring possible effects with the nature of being. Contemporary theories of states of affairs often exclude necessary truths from the class of things an agent must be able to actualize. Yet the exact exclusion can be contested if necessity itself is understood as dependent on divine agency, as in some interpretations of Descartes. One cannot simply impose the counterfactual model of causation on all theological views without examining the intended relation of sourcehood.

The word 'make' also marks distinct relations. A mathematician makes a theorem known; a legislator makes a rule operative within a legal institution; a carpenter makes a table; a deity in classical creation theology makes dependent things exist. The proposition that a theorem is necessary is different from the event in which someone proves or recognizes it. An omnipotence test must say which making is at issue. Ambiguity here gives rise to apparent impossibilities that reflect relations among concepts rather than limits of an agent.

A stronger consequence follows. Omnipotence as maximal causal efficacy does not automatically imply metaphysical authorship of truths. Conversely, metaphysical dependence of truths upon an ultimate source does not entail that the source has every empirically describable capacity. This distinction prevents the inflation of power into explanatory priority and vice versa. It also bears on later comparisons between divine power and mere technological achievement: unprecedented agency in one domain does not by itself determine the modal status of all other domains.

Action Theories, Result Theories, and Individuation

Action-centered accounts begin with tasks: an omnipotent being can perform every action of a relevant type. Their difficulty is the ability to act under reflexive descriptions, such as learning one has never been omnipotent, committing a mistake, or becoming subject to a force stronger than oneself. Some actions are possible for ordinary agents precisely because of properties an essentially omnipotent agent lacks. To make every executable description obligatory risks attributing defects as positive powers.

Result-centered accounts assess possible states of affairs. Their advantage is abstraction from the agent's bodily repertoire: changing an outcome can be an object of power even if the mode of doing so is unlike human action. But these accounts face causal and modal exclusions. A condition may be logically possible in a universe whose past differs from ours while becoming inaccessible from the actual history. A state may also be contingently possible yet depend upon another agent's libertarian choice, for which unilateral production is in dispute.

The measure of comparative power depends on the individuation of results. If an agent can make a light flash red, it can bring about infinitely many descriptions of that result: that the light flashes, that a visual event occurs, that a red event occurs, and so forth. Counting propositions as separately acquired powers risks inflation by linguistic redescription. The metaphysical unit of comparison should be an appropriately individuated achievement or power-family, not every sentence true as a consequence of the achievement.

In their influential analyses, Hoffman and Rosenkrantz examine repeatable states of affairs to avoid some of these problems, while Flint and Freddoso and Wierenga use relations between worlds that share histories or relevant background conditions. The technical details are distinctive and not interchangeable. What unifies the approaches is recognition that a flat quantification over anything conceivable fails to capture the modal and agential structure of omnipotence.

Shared Histories and the Fixity of the Past

The shared-histories approach asks what an agent could bring about in a world with the same history up to a specified time. An alternate future may be accessible under such an account even if an alternative past is excluded by the stipulated historical index. This makes a formal distinction between a being's prospective causal reach and a logically describable world differing in its earlier events.

An illustrative case is a person who did not go to a city yesterday. A world where the person went may be metaphysically possible. But an agent acting today may not have the power to bring about that the person went yesterday in the actual past. If past-directed causation is allowed by one's metaphysics, more sophisticated analysis is needed: causing an event in the past and changing a completed past are different possibilities. The classical denial of changing the past should be compared with, rather than presumed to decide, modern accounts of temporal causation.

The formulations associated with Flint and Freddoso (1983) and Wierenga (1989) also address the behavior of free creatures and facts counterfactually dependent on their actions. Their shared-history formulations help exclude some accidentally impossible outcomes from omnipotence tests. Yet the choice of which facts are held fixed is controversial. If the framework freezes too many facts, maximal power may become too easy to satisfy; if it freezes too few, it may require control over independent choices or settled events. A defensible model needs an independent account of the admissible background.

Even under fixed-history semantics, a problem remains about divine actions that establish the history itself. If the deity is described as the source of time and the world, treating the actual past as a fixed constraint can seem question-begging. The theorist may respond by distinguishing creative power over entire histories from power at a time within an actual history. The same agent may have different modal relations to world-creation and to later interventions, without losing an essential capacity. An adequate model must articulate both levels rather than force one temporal predicate to carry all metaphysical work.

Two Omnipotent Agents and Conflicting Endeavors

Imagine two agents A and B, both omnipotent, attempting contradictory results at the same time. A tries to move a feather; B tries to keep it still in the same respect and interval. If both are genuinely able to realize their goals despite the other's opposition, the combined description appears inconsistent. This is the multiple-omnipotence problem. It is a more general version of the self-limitation puzzle because it tests the relation between incomparable supreme agencies.

One response defines omnipotence as unique and thereby excludes two omnipotent agents. This may fit a monotheistic theology but does not independently demonstrate that maximal power implies uniqueness. Another response claims that two omnipotent wills would necessarily be harmonized, so the hostile scenario is impossible. This shifts the burden to the metaphysics of willing and perfection. A third treats power as the independent capacity to effect a result, even where its manifestation is prevented by another free power, but then it risks collapsing omnipotence into merely maximal ability without opportunity.

The example demonstrates why comparison must include interference and opportunity. Power over all others cannot be understood only as a list of things an agent could do in an empty universe. If the domain includes other agents, the theory must specify how their causal contributions interact. A statement that an omnipotent being is unsurpassable in overall power does not automatically entail a single universal controller unless a total order of powers and the relevant conflicts have been established.

Self-Limitation, Abdication, and the Continuity of Power

A familiar theological proposal holds that an omnipotent creator voluntarily limits its own exercise of power for the sake of creaturely independence. If self-limitation means refraining from intervention despite retaining capacity, there is no immediate contradiction. If it means irrevocably giving up a capacity while necessarily remaining omnipotent, the doctrine is much more difficult. It combines an essentially stable maximal attribute with an irreversible loss of a constitutive part of that attribute.

Two descriptions must therefore be tested independently: a being that has unlimited power at one time and surrenders it later, and a being that necessarily possesses unlimited power at every time yet surrenders it. The first is conceptually less demanding; the second requires a coherent sense of self-surrender that preserves what is allegedly surrendered. The stone paradox makes this issue vivid, but the general topic is a theory of agency over its own future capacities.

The philosophical point is not to deny the possibility of religious accounts of divine restraint. Such accounts can distinguish willingness from ability and voluntary commitment from external compulsion. The analytical task is to determine whether what is called a restraint constitutes a limit in causal possibility, a norm that guides choice, a self-imposed promise, or a metaphysical change in being. Those differences must remain visible whenever 'omnipotence' is defended through self-limitation.

Omnipotence and Necessary Moral Perfection

The compatibility of omnipotence with moral perfection presents a challenge distinct from the stone paradox. The question is whether the most powerful possible agent must be able to commit a morally wrong action, and whether a being that is necessarily good could have that capacity. Augustine and Anselm favor a conception on which inability to err can signify excellence; contemporary philosophers ask whether the elimination of options nevertheless diminishes maximal power.

The issue becomes precise when perfection is indexed to possible worlds. A necessarily morally perfect being acts rightly in every world in which it exists. A being capable of wrongdoing has at least some admissible circumstance in which it could do otherwise. If the former cannot choose a wrong act while the latter can, then the latter appears to possess an additional ability. It does not immediately follow that it possesses greater overall power. The comparison presupposes a rule for ranking a set containing a morally defective option against a set containing other capacities and excellences.

Wes Morriston (2001) argues explicitly that necessary moral perfection and absolute maximality of power may diverge. His published abstract states that a necessarily good deity might fall short of the maximum conceivable amount of power and that a being with somewhat less than maximal power might nevertheless be the greatest possible being. This is a powerful challenge to a common shortcut in perfect-being theology: greatness and power cannot be treated as analytically identical merely because each is associated with a supreme being. The argument requires a substantive theory of what powers are, what makes one greater than another, and how total greatness is evaluated.

Richard Swinburne's approach, represented by his Coherence of Theism, analyzes omnipotence with restrictions involving the logical structure of actions, time, and freedom from non-rational influence. The 2016 second-edition chapter abstract gives a time-indexed formulation. An omnipotent being must possess the relevant knowledge and rational independence as well as the ability to bring about appropriate logically contingent events. This richer conception places the property within a theory of agency instead of defining it as pure physical force. The 1993 first-edition abstract additionally records a restriction concerning overriding reasons not to perform certain actions. These are different editorially identifiable formulations; they should not be conflated into a single timeless quotation from Swinburne.

A central analytical choice is whether the inability to do evil is a failure of capacity, a matter of motivational impossibility, or a consequence of rational self-governance. Someone might be physically capable of causing harm while finding it morally unthinkable to do so. Another might have a nature that makes deliberate wrongdoing metaphysically impossible. The first has a capacity combined with a stable disposition; the second has a restricted possibility profile. The distinction is significant because a claim of essential goodness is stronger than a claim of exceptionally reliable good action.

The synthesis must preserve both sides of the issue. Moral excellence is relevant to the ranking of persons and divine candidates, while maximal power remains an independently assessable attribute. An analysis that deliberately defines everything a perfect agent cannot do as no power at all might secure consistency by fiat. An analysis that counts any additional sinful option as an unqualified increase in power might ignore the relevance of autonomy, rational control, and vulnerability. A sound theory makes its value assumptions public and retains the disagreement as a substantive dispute about the nature of perfection.

Omnipotence, Freedom, and Counterfactuals of Creaturely Choice

The relation between divine power and creaturely freedom poses a further problem. Can an all-powerful creator bring about every possible world, including a world in which created agents always freely choose the good? The word 'freely' matters. Under libertarian interpretations, some choices are not determined by prior causal conditions; producing a particular free decision by unilateral determination may eliminate the very freedom that the desired result describes.

Alvin Plantinga's God, Freedom, and Evil (1974) distinguishes the logical problem of evil from the project of explaining the actual reasons for permitting evil. In his free-will defense, the possibility of significant creaturely freedom is used to contest an alleged contradiction between a good omnipotent God and the presence of evil. The chapter structure includes a direct examination of whether God could create any possible world and introduces 'transworld depravity.' The substantive point is that a broadly possible world need not be actualizable by an omnipotent creator in circumstances where particular free decisions have counterfactual profiles independent of unilateral divine choice.

This argument must be assessed with care. A defense aiming to establish consistency needs only to show a possible conjunction of claims that blocks the proposed logical contradiction. A theodicy attempts to identify actual justifying reasons and is much more ambitious. Plantinga's defense does not prove that free will accounts for every actual evil. Nor does its structure automatically account for suffering from natural processes, predation, or events unrelated to human choice. The treatment of 'transworld depravity' is part of a formal defense, and a historical claim about its success should be separated from the philosophical merit one assigns it.

The counterfactuals of freedom used by some accounts of divine providence are contested. A subjunctive such as 'If this agent were placed in circumstances C, the agent would freely choose A' may be taken as true independently of the creator's determination. The issue is whether such conditionals have determinate truth values and what grounds them. Flint and Freddoso's account uses counterfactual and world-history machinery; critics question whether the needed conditionals are coherent or sufficiently independent. The SEP entry on omnipotence presents this as an important debate rather than a settled technical solution.

There is no single way to combine omnipotence with freedom. Compatibilists may describe a free action as one arising from the agent's motives under appropriate conditions, even where those motives are causally determined. Libertarians normally attach stronger significance to open alternatives or ultimate sourcehood. Under either approach, one must distinguish the power to create a free agent, the power to set circumstances, the power to influence a decision, and the power to guarantee a particular decision while preserving the relevant freedom. Each is a different claim about causal scope.

The Problem of Evil and the Exercise of Power

The logical problem of evil takes omnipotence, omniscience, and perfect goodness as premises and argues that their conjunction conflicts with the existence of evil. J. L. Mackie's 'Evil and Omnipotence' (1955) is a seminal analytic statement of the issue. Its significance for this research is that an apparently unlimited capacity to prevent suffering becomes part of an argument about what a perfectly good agent would do. The inference from power to action requires moral and explanatory premises; omnipotence alone is insufficient to establish it.

The argument can be reconstructed conditionally. If a being knows about evil, has power to prevent it, and has decisive reason to prevent it, then its nonprevention calls for explanation. To turn that observation into a contradiction, one must establish that no morally sufficient reason for permission is compatible with the remaining attributes. A free-will defense, soul-making account, or greater-good proposal may challenge that additional premise. Their adequacy remains debatable, particularly concerning horrendous suffering and the distribution of harms.

An evidential argument is different. It can contend that actual patterns of suffering reduce the probability of traditional theism even if no strict contradiction follows. Showing that divine omnipotence is formally coherent would not thereby defeat an evidential argument based on the apparent gratuitousness of specific evils. Conversely, the existence of a severe problem of evil does not by itself demonstrate that the very predicate 'omnipotent' is logically incoherent. Philosophy of religion benefits when conceptual compatibility, explanatory adequacy, and comparative probability are assessed separately.

The question also tests accounts of divine power that emphasize permission and cooperation. If an all-powerful agent voluntarily refrains from intervention, the existence of evil may still pose a moral challenge to that choice. If the agent lacks coercive power by metaphysical necessity, the challenge changes: a traditional claim of universal control is abandoned or revised, but a framework with constrained power may interpret suffering differently. It is therefore essential to identify the conception of divinity involved before claiming that a particular answer resolves 'the problem of evil' as a whole.

Process Theism and Persuasive Power

Process theism, associated with Alfred North Whitehead and Charles Hartshorne, presents a major alternative to classical omnipotence. In its characteristic metaphysical picture, actuality consists of events or occasions with their own forms of activity. Divine influence is often construed through persuasion, invitation, and responsive relation rather than unilateral determination of every concrete decision. The SEP entry 'Process Theism' explains the specific metaphysical claim that no actual entity can simply take over every decision of another actual entity.

This conception changes both theology and the philosophy of power. If another creature's self-determination is built into the metaphysical structure of existence, the absence of unilateral divine control is not a voluntary temporary restraint on an otherwise exhaustive coercive power. It expresses a different kind of ultimate relation. Some process theologians argue that persuasive influence is a superior form of power to domination. A traditional critic can respond that the ability to persuade plus the ability to intervene coercively is still an expanded power set, and that a wholly persuasive God may lack capacities expected of the classical creator.

The disagreement should not be hidden by a shared use of words such as 'supreme,' 'universal,' or 'maximal.' Classical and process theists may rank modes of power differently and adopt different metaphysical constraints on what a single agent can bring about. The debate is partly about whether an ultimate being should be modeled through exhaustive unilateral control at all. That is a substantive dispute about the structure of relational reality.

Process theism also poses an ethical question for traditional accounts. If perfect benevolence could operate through influence without guaranteeing outcomes, what would justify assigning an even greater coercive capacity to the deity? Conversely, if an agent could prevent horrendous suffering without erasing meaningful agency, why would restraint count as perfect benevolence? A philosophical comparison must examine both the metaphysics of creaturely self-determination and the moral responsibilities that follow from each theory's claimed powers.

Omnipotence and the Existence of Ultimate Reality

In perfect-being theology, omnipotence is rarely an isolated predicate. It is attached to a proposed ultimate source that is also necessary, good, intelligent, and the creator or ground of contingent reality. The interaction of attributes generates a further family of problems. If divine knowledge includes a complete future, is there room for genuine alternative divine choice? If divine simplicity denies internal composition, what is the ontological status of distinct divine powers? If God is timeless, how should the 'at time t' clauses of contemporary omnipotence definitions be interpreted?

The fact that an attribute can be formulated without direct contradiction does not prove that all divine attributes can be instantiated jointly. A proof of joint possibility must address cross-attribute constraints. Nor is it enough to say that each perfection has its own domain. An account of perfect knowledge that includes facts about future free choices can alter the possible extent of freedom; an account of perfect goodness that excludes certain intentions can alter the accessible acts; and an account of necessary existence may render self-annihilation incoherent. These dependencies are among the central reasons omnipotence is a difficult philosophical attribute.

The theory of maximal greatness offers one response: the supreme being possesses those great-making properties that are jointly compossible, considered through an appropriate account of excellence. But compossibility and ranking require arguments. A conception that automatically labels any inconvenient incompatible property a non-perfection may become analytically insulated from criticism. An adequate perfect-being methodology needs an independent criterion for greatness and a demonstrated relation between that criterion and the powers claimed.

A complementary question concerns the alleged creation of physical laws, time, and modal structure. Creating a physical world may entail a power over natural law not available to any creature within that world. It does not follow that the same power extends to the creation of every logical truth, nor that a world can be both totally determined and libertarianly undetermined in the same respects. The metaphysical architecture must identify each level of dependency and preserve the inferential distinctions among them.

Comparison Beyond Classical Theism

A research dossier on omnipotence should recognize that 'God' is not a culturally uniform concept. Jewish, Christian, and Islamic thinkers have developed distinct interpretations of omnipotence, necessity, revelation, divine freedom, and causality within diverse schools. Even within traditions associated with the classical Abrahamic God, there is no single agreed relation between divine will, possibility, and creaturely independence. The medieval evidence considered here is only a defined part of that intellectual history.

Religions with multiple divine figures, dualistic cosmic principles, or an impersonal ultimate may not attribute a singular agent's power over every state of affairs. Applying the same property without considering local ontological and religious meanings would impose an external vocabulary. A careful comparative project would need original-language primary corpora and precise descriptions of the traditions studied. The present monograph confines detailed argument reconstruction to the sources actually inspected and marks wider comparison as a distinct future research task.

This restraint also applies beyond religion. Political claims of absolute sovereignty and technological rhetoric about unlimited control can borrow the language of omnipotence metaphorically. Such uses can be examined as analogies about the distribution and limits of agency. They do not establish that a state, person, or artificial system possesses the metaphysical attribute defined in traditional perfect-being theology. The philosophical framework developed here is useful for identifying when comparisons change the subject.

Contemporary Directions and Unresolved Debates

Several contemporary directions remain live. One concerns formal power-orderings and whether there can be an upper bound on the powers possible for an agent. Another concerns the ontology of states of affairs and their individuation: if outcomes are defined at different levels of description, comparative power may vary. A third concerns the role of impossible-world frameworks in representing counterpossible reasoning without granting that impossible states are metaphysically producible.

A fourth direction connects omnipotence to theories of agent-causal control. Control may be exercised by determining outcomes, choosing among possibilities, sustaining background conditions, or empowering independent actors. A fifth concerns the structure of time and history: models of backward causation may alter how a past-directed task is classified. A sixth concerns perfection: a moral option can be treated as an added possible action, a susceptibility to defect, or a failure of rational self-command. Different conclusions follow from these theories.

The central methodological danger remains easy to state. Every proposed solution can be made trivial by changing the definition after confronting an objection. A more rigorous practice records the intended scope, identifies which tests the account can actually pass, and acknowledges which cases it excludes on independent grounds. The standard is neither an impossible demand for unanimity nor the rhetorical appearance of a solved paradox. It is a transparent relation between definition, modal theory, and consequences.

No philosophical result surveyed here establishes the actual existence of an omnipotent agent. Conversely, no single isolated paradox settles every coherent analysis of maximal power. Contemporary debate has progressed by replacing the uninformative phrase 'can do anything' with increasingly precise questions about what is possible for which agent, in which circumstances, with what history, and under which constraints. That progression provides the basis for the independent synthesis developed next.

Recent Definitions and the Scope of the Literature

The comparison of contemporary positions must begin with the strongest documented predecessors rather than treating a list of analytical dimensions as an unprecedented discovery. Flint and Freddoso's “Maximal Power” (1983) is particularly important. Their author-hosted complete transcription starts by separating religious almightiness, construed relative to a divine nature, from maximal power as a philosophical property that ought to be analyzable without building a particular theology into its definition. They articulate five adequacy conditions and work with states of affairs, time, shared histories, and strong and weak actualization. An analysis that merely lists domain, agency, temporal context, causation, and evaluation would therefore recapitulate much of their groundwork. Their historical priority and detailed technical work must be acknowledged at the point of comparison, rather than buried in a general bibliography.

The first challenge is the description of power's object. A state-of-affairs analysis asks whether an agent can make a proposition true in the appropriate sense; an action analysis asks whether an agent can perform a certain act. These objects cannot be casually interchanged. A state that obtains without an agent's contribution is not thereby an effect of the agent. Conversely, different action descriptions may designate one occurrence while having different conceptual implications. In their open 2026 encyclopedia revision, Hoffman and Rosenkrantz explicitly treat the difficulties with unrestricted repeatability, history, and divine moral perfection. Their article illustrates both the power and the limits of state-of-affairs reconstruction: even a clear state description needs a principled account of possible actualization by a particular agent.

Flint and Freddoso's distinction between strong and weak actualization matters whenever independently free agents enter the argument. Strong actualization involves causally determining an outcome. Weak actualization, on the relevant account, permits an agent to bring about circumstances within which another freely brings about a result. The difference is more than terminology. Under a libertarian premise, a creator's establishing opportunities for another's free choice can be compatible with that choice's independence, whereas the creator's unilaterally determining that very free act would withdraw the independence stipulated in the object. Whether any particular free choice would occur under selected circumstances depends on further counterfactual assumptions. The conceptual separation neither secures a favorable counterfactual nor excuses an agent from explaining a prevented or permitted harm.

Martin Lembke's “Omnipotence and Other Possibilities” (2012) explicitly proposes, according to its Cambridge publisher abstract, a distinctive metaphysical account after examining the difficulty of reconciling omnipotence with divine impeccability. This establishes a significant alternative to accounts that concentrate on enumerating executable actions. The abstract alone cannot warrant a reconstruction of every clause of Lembke's proposed definition or prove that it succeeds. Its warranted contribution here is bibliographic and methodological: a metaphysical analysis was already offered in the contemporary literature, so the present work cannot represent the turn to metaphysical conditions as an original founding move. The additional issue of how an omnipotent agent may bear on other agents' possibilities deserves study in his complete article rather than a speculative paraphrase.

Michael Wreen's “Omnipotence” appeared online on October 24, 2017 and in volume 84 of the International Journal for Philosophy of Religion in 2018, pages 205–221. The Springer publisher abstract describes seven unsatisfactory candidate definitions before a proposed adequate definition. These are claims about Wreen's stated research program, not a claim that the present researcher independently verified his seven definitions. The distinction matters because a bibliography that cites only the published volume year may conceal the work's earlier electronic publication date, and an analysis that announces a simple definitive solution would need to confront the sophisticated alternatives already on record.

Wreen's subsequent open article, “The Contradiction Approach to Solving Problems about Omnipotence” (2022), is available as a complete publisher-hosted PDF and gives a sharper test. He begins with the apparent elegance of excluding contradictory tasks and examines an action involving a stone too heavy for “the last personal being Wilfrid thought about” to lift. Under the described circumstances, that person is God, but the action sentence's indirect identifying phrase does not itself exhibit the contradiction that a simple exclusionary rule would require. A restriction on actions therefore depends partly on how actions and descriptions are individuated. Merely deleting syntactically inconsistent requests from a list of tasks does not guarantee a satisfactory theory. Wreen presses the issue through extensional and intensional descriptions and concludes that a metaphysical account remains necessary. This is a source-specific reconstruction of his 2022 argument, not an attribution of the same example to his inaccessible 2018 full article.

One useful cross-reading is that the intellectual distance between medieval and analytic treatments is smaller than their vocabularies suggest. Aquinas already separates absolute possibility from possibility relative to a finite agent's power. Maimonides already insists that contradictions concern the object described rather than a deficiency in a divine cause. The contemporary question is how far that insight can carry an analysis once action identity, possible histories, independently free choices, and necessary divine attributes are specified in finer detail. The historical authors did not possess a present-day possible-worlds semantics; the correspondence concerns a shared explanatory problem and does not transfer modern notation backward.

A separate source-critical issue concerns claims that mathematics or logic themselves could be created or modified. Descartes' letters of 1630 are genuinely important to the discussion, but different scholarly interpretations of the doctrine of eternal truths should not be collapsed into a single unqualified thesis. Treating the position as a challenge to the independence of necessary truths is warranted by the consulted historical scholarship. Treating it as a proved demonstration that contradictions are actualizable is not. Leibniz's public English Theodicy translation, in turn, grounds the contrast between mere imaginable variants and complete, mutually connected possible worlds. The ethical and providential argument cannot be detached from the compossibility of the entire world being evaluated.

These comparisons limit the present monograph's claim of originality. The original contribution consists in independently evaluating which consequences follow when widely documented distinctions are applied together in specific hard cases. Its comparisons are answerable to Flint–Freddoso, Hoffman–Rosenkrantz, Lembke, and Wreen. It supplies neither an invented new technical definition nor proof of research priority over those predecessors. Its quality rests on whether it identifies a real logical dependency, keeps evidence at its demonstrated level, and improves the reader's ability to recognize when two purportedly opposing answers address different propositions.

Comparative Analysis of Omnipotence Claims

An attribution of unlimited power has to specify what its grammatical object is, who exercises it, at what time, by which causal relation, and under which interpretation of maximality. These are standard distinctions in the literature, especially the state-of-affairs, history-relative, and actualization discussions of Flint and Freddoso and of Hoffman and Rosenkrantz. The point of combining them here is diagnostic: several apparent disputes exchange one background assumption for another mid-argument. Reconstructing those transitions makes the inferential commitments inspectable.

The domain question comes first. The set of logically describable states can differ from the set of metaphysically possible states, from possibilities physically available in a particular world, and from outcomes feasible given a fixed past. One cannot say that an agent is omnipotent because it can achieve all outcomes possible for it without offering a separate criterion for which outcomes count as possible. The formula otherwise assigns maximality even to a creature whose nature allows only one trivial operation. An independent domain criterion is necessary for the attribution to discriminate meaningful degrees of power.

The bearer question concerns the attributes that enter a proposed identity. To ask whether an essentially omnipotent being can make itself permanently impotent is to ask about the joint possibility of two descriptions, not solely about strength. An agent that possesses power only at an initial time may relinquish it later without contradicting that temporally limited predicate. An agent that is stipulated necessarily and eternally omnipotent cannot both retain that defining attribute and permanently abolish it. The latter exclusion does not yet demonstrate that the necessarily omnipotent being is possible; it shows the contradiction within a more elaborate action description.

The historical question prevents a simple transfer of power between complete possible worlds and an agent acting from a given present. A complete history could have included an event that did not occur. If the past is fixed, an agent now lacks an available action that both preserves the actual history and makes that same history different. A backward-causation theory may introduce other causal relations, but it must specify whether an earlier event is changed, explained, or merely reached from a later time. Flint and Freddoso's shared-histories analysis confronts exactly this level of qualification. No proof of a universal temporal prohibition follows from identifying a history-dependent conflict.

The efficacy question distinguishes producing a result from creating an opportunity for someone else to produce it. Suppose a world includes a free decision by another agent. Setting the circumstances in which that decision occurs differs from determining the free decision's content. If the relevant freedom is libertarian, the latter description may be incoherent; if it is compatibilist, it may be coherent. In either case, a philosophical theory must determine which power relation is under discussion before treating an instance of free action as a counterexample to supreme power.

The evaluative question concerns the relation between a repertoire of actions and perfect agency. If a being's essential impeccability rules out morally wicked actions, a simple count of executable actions may rank a corruptible agent above it in a narrow ability measure. Yet a theory of maximal perfection may treat the possibility of moral failure as a defect, as Anselm's argument expressly does. It remains necessary to explain why the moral restriction does not make the degree of power arbitrary. The compatibility of maximal power and perfect goodness must be argued as a substantive metaphysical proposition rather than introduced by definition.

The distinction between a unique greatest power and an unsurpassed power is also decisive. In a partial ordering, two agents may possess non-nested capacities: each has an ability the other lacks. Neither is exceeded by the other, but this does not establish that a single greatest bearer exists. The inference from “no actual agent is more powerful” to “this agent is the greatest possible being” therefore needs a completeness or comparability premise. Adding such a premise is permissible in a defended theory; smuggling it into the word “maximal” is not an argument. This is an elementary ordering observation, not a new formal theorem.

A robust assessment holds these choices stable while testing examples. When a stone paradox is posed, its agent and time assumptions must be fixed before an answer is judged. When freedom is discussed, the concept of freedom must remain constant between premises and conclusion. When the problem of evil is invoked, one must distinguish the factual capacity to prevent a particular suffering, the knowledge of it, and the reasons for permitting it. A framework that silently redefines possibility whenever a difficult task arises has immunized its conclusion against assessment. The appropriate scholarly response is to make the changed premise explicit.

An example involving two agents clarifies the significance of causal opportunity. Imagine agents A and B, both claimed to be able to ensure opposite results at the same moment, and assume that each has unconditional power to guarantee its own choice independently of any rival. The conjunction of opposed successful outcomes is inconsistent. Accordingly, the joint attribution of unconditional simultaneous outcome-guaranteeing power cannot hold in that setup. It does not follow that no two beings can have every other form of maximal power; a qualified conception might restrict conflicting exercises or define their shared agency differently. The verdict depends on the unconditional guarantee in the stipulated case.

A case involving mercy illustrates the value of separating dimensions. A perfectly good agent may refrain from a permissible intervention because it judges another value to be greater. Its abstention does not show physical inability, but the bare assertion of an undisclosed superior reason cannot establish moral adequacy. Evidence of preventable and apparently pointless suffering can remain a serious objection. The power analysis supplies a predicate about capacity and causal efficacy; moral evaluation adds a predicate about reasons. Both must survive scrutiny when the argument concerns a perfectly good omnipotent creator.

These comparisons are intentionally neither a sufficient definition of omnipotence nor a new named doctrine. They synthesize documented lines of analysis, expose hidden transitions between modal, temporal, causal and moral premises, and establish which further work is required to defend a particular conception. As a research finding, their contribution lies in transparent comparative reconstruction rather than a claim to an unprecedented taxonomy.

Case-Based Consistency Examination

The following informal cases serve as an internal consistency check on the claims made in this monograph. Their purpose is to identify invalid implications. They do not provide an axiomatized modal logic, a completeness proof, or a novel definition of maximal power. Each conclusion is conditional on its stated premises.

Case one concerns a contradiction. Suppose an action requires a single object to instantiate a property and its negation in the same respect and at the same time. Within classical logic, the described result is inconsistent; the inability to produce it is therefore no evidence of inferior causal force under an account that independently limits power to compossible results. The restriction becomes circular if a theorist instead defines “possible” as simply “whatever the favored agent happens to do.” The independent logical criterion is doing the necessary work. A defender of the creation of eternal truths must explain what logic governs this purported creation and its description; changing the modal premise changes the question.

Case two concerns the stone. Let A be stipulated omnipotent at time t, while leaving open whether its omnipotence persists. The task “at t, create a stone that A at a later time t2 cannot lift” is not automatically equivalent to the task “at t, create a stone that A cannot lift at t while retaining unlimited lifting power.” An answer about permanent omnipotence must therefore specify why a later loss of power is excluded. If it is essential by hypothesis, the self-limitation description conflicts with the essential property. If only present omnipotence is assumed, that specific contradiction no longer follows. Neither branch settles the independent possibility of momentary omnipotence.

Case three concerns an oblique identity description. Let the “last being considered by Wilfrid” designate A. An action ascribed to A may be described using that indirect phrase and may appear free of overt contradiction, though the designation picks out the supposedly omnipotent bearer. Wreen's 2022 example shows why an argument cannot depend only on the surface wording of the task. Identity and intensionality matter: the apparent executable task can change its philosophical interpretation when co-designating expressions replace one another. A serious account must justify its treatment of action individuation, not rely on linguistic censorship.

Case four concerns historical possibility. Let a past event e have occurred. The proposition “there exists a complete possible history in which e never occurred” can be consistent with “from the currently fixed history, no agent can now bring about e's nonoccurrence while preserving that history.” The first concerns alternative whole histories; the second concerns present intervention relative to a given past. Conflating them introduces a scope error. A backward-causation hypothesis needs a distinct causal analysis and cannot be disposed of merely by repeating the word “past.”

Case five concerns free choice. Suppose C chooses freely in libertarian circumstances. A creator may strongly actualize the availability of those circumstances, but whether that suffices for C's actual free choice depends on counterfactual assumptions that are not settled by the power predicate. If a critic demands the creator strongly determine the free choice as free in the stipulated libertarian sense, the proposed outcome contains incompatible causal conditions. If instead the challenge concerns why the creator did not prevent known suffering through some separate permissible intervention, the contradiction response alone is inadequate. The two challenges must be addressed independently.

Case six concerns goodness. Let an agent have the causal power to prevent harm H and also possess perfect goodness. There is no valid deduction from the power proposition alone to the conclusion that the agent ought to prevent H in every circumstance; an argument about moral reasons is needed. Equally, there is no valid deduction from the existence of logically conceivable justifying reasons to the conclusion that any such reason actually obtains. Therefore, a power analysis can clarify the logical structure of the problem of evil while leaving the evidential and axiological question open.

Case seven concerns comparative maximality. Let the powers of A and B be represented by non-nested sets of successful effects. From the fact that A and B are incomparable, it does not follow that either possesses every possible power. A maximal element within an ordering is not automatically a greatest element. Applying this elementary distinction to agent capabilities helps prevent unwarranted claims about an unsurpassable bearer. Whether a universal greatest power is metaphysically possible requires assumptions about the space of compossible powers and their bearer.

Taken together, the tests show a defensible but limited conclusion. Several standard paradoxes reveal a hidden contradiction, a shift in time, or an unexamined theory of agency. Other objections—especially incompatibility with moral perfection, feasibility of worlds with creaturely freedom, and whether a greatest possible power profile exists—remain substantive even after those ambiguities have been removed. The research thus establishes a reliable separation of explanatory burdens without pretending to settle every metaphysical dispute.

Criteria for a Successful Theory of Omnipotence

An adequate account should meet six conditions. It should provide an explicit, noncircular definition of the relevant ability or efficacy relation. It should give an independently motivated account of the modal domain. It should state whether omnipotence is momentary, persistent, essential, or necessary. It should explain how tasks are individuated and how agency enters their descriptions. It should identify the relation between power and further divine attributes, especially goodness and freedom. Finally, it should survive standard adversarial cases without modifying its central definitions merely to evade the cases.

These are evaluative requirements, not a proof that only one formal definition is acceptable. In some philosophical projects a state-of-affairs analysis will be appropriate, while an action-capacity analysis is better suited to others. A historical reconstruction of almightiness in a religious tradition may be guided by sovereignty and providence rather than the possible-world scope of late analytical philosophy. The requirement is that a model be clear about its object of analysis and not present a conditional result as an unconditional answer.

The criteria also identify frequent failures. A theory defining omnipotence as 'being able to do everything possible for an omnipotent agent' risks pure circularity unless the reference to omnipotence is eliminated by independent modal conditions. A theory requiring actualization of every logically consistent sentence risks including states with impossible identities or incompatible histories. A theory counting every linguistically distinct consequence as an independent power risks artificially inflating the number of powers. A theory excluding all self-referential tests simply because they threaten its conclusion removes informative challenges from the domain. These failures follow from independently assessable defects of analysis, not from hostility to any religious position.

At the same time, there is a legitimate role for conceptual exclusions. A square circle, under its ordinary definitions, is not an extraordinarily challenging physical construction; it is a contradictory task description. Making the past different from how it actually was may likewise be contradictory under a fixed-history interpretation while remaining distinct from causing an event at an earlier time under a more complex temporal model. A careful theory can exclude pseudo-tasks and retain demanding genuine tasks; the epistemic burden is to demonstrate which is which.

The comparison of religious positions also requires a common measurement of the attribute being compared. Classical theism may view sovereignty over created history as a perfection. Process theism may describe the capacity to elicit creative self-determination as the relevant ideal. Neither framework can be judged solely by importing the other's evaluative standard E without an additional argument for the standard itself. Genuine comparison examines the consequences of each model, its explanatory work, and its consistency with the tradition's other commitments.

Analytical Synthesis: Maximal Power, Modal Scope, and Effective Agency

The central result is a defended methodological judgment: an adequate attribution of maximal power must distinguish the possibility of an outcome, the ability of a given agent to bring it about, the opportunities available at a time, and the standard by which one power profile is compared with another. This follows from the failures of vacuous agent-relative definitions, from history-sensitive cases, and from the difference between causal production and merely permitting a state to obtain. The constituent distinctions have substantial established predecessors, particularly Flint and Freddoso and Hoffman and Rosenkrantz. Their comparative application in this monograph is an original interpretive analysis, with the source lineage explicit.

The distinction between the limits of description and the limits of causal agency is decisive. A contradictory task has no coherent success condition under classical logic; a physically demanding but logically consistent task can have a clear success condition that exceeds an actual agent's power. A task requiring an essentially impeccable agent to be corruptible has an identity-relative difficulty. A task requiring an independently free actor to choose under total external determination challenges a theory of freedom. One word, “impossible,” conceals these different explanatory grounds. Restating the grounds exposes the legitimate inferential possibilities and prevents a theorist from passing from one meaning to another.

A further distinction concerns maximality. The range of a being's capacities can be ordered by inclusion, effectiveness, or some conception of greatness, and these comparisons need not coincide. If two agents' ranges are incomparable, maximal elements may exist without one universal greatest element. A perfect-being argument must therefore justify why its power standard has the kind of upper bound it requires. The existence of that upper bound, its compatibility with other perfections, and its actual instantiation are three independent propositions. A coherent description alone can establish none of the latter two.

The historical lesson is one of continuity and sharpening. Augustine and Anselm connect power to freedom from defect. Maimonides and Aquinas explain the exclusion of contradictory effects in terms of the nature of the possible. Descartes challenges assumptions about divine dependence on eternal truths. Leibniz analyzes world-creation under the interconnectedness of possibilities and divine wisdom. Contemporary discussions specify state descriptions, times, shared histories, and the kinds of causal relation involved. The resulting vocabulary is more fine-grained, but the central philosophical aspiration persists: to make supreme power an intelligible, nonvacuous perfection.

The analysis also illuminates a limit on theological explanation. A claim that omnipotence extends to every genuine possibility cannot by itself decide what the world's genuine possibilities are. A claim that a perfectly good agent has sufficient reason to permit suffering cannot by itself identify that reason. A claim that supreme power allows free creation cannot by itself decide the metaphysics of libertarian action. Where further premises are required, a philosophical monograph serves its subject by identifying them precisely. That clarification is its original contribution; it is an analytical conclusion rather than a new theory, named model, or proof of novelty.

Research Limits and Evidentiary Confidence

The historical judgments in this dossier draw on directly displayed English translations of Augustine, Anselm, Maimonides and Aquinas, a public English Theodicy, and scholarly historical reconstructions for Descartes and disputed medieval contexts. The study does not present itself as a collation of every critical Latin, Judeo-Arabic, or French edition. Its claims about specific historical positions are confined to the accessible passages identified in the reference list. This boundary is editorially material to precision but does not prevent responsible conclusions about what those selected passages actually argue.

Flint and Freddoso's complete author-hosted article supplies direct evidence for the technical distinctions attributed to them. The Stanford Encyclopedia of Philosophy's 2026 Omnipotence entry supplies a current full scholarly synthesis. Wreen's 2022 open journal article permits passage-level discussion of the indirect-description and contradiction challenge. The 2012 Lembke and 2018 Wreen articles are incorporated at the level of publisher abstracts and bibliographic metadata; no detailed unverified definition from either is reconstructed. Other books and paywalled articles are treated at the source-access level documented in the accompanying evidence file.

The modal analyses are deliberately conditional and informal. Each case states assumptions and tests a particular inference, rather than claiming to decide the consistency of all modal logics, all conceptions of free agency, or every theory of divine perfection. The new editorial pass checks the cases for a consistent scope of quantification and for the separation of necessary, contingent, historical, and causal claims. It establishes a manuscript-level internal logic review appropriate to these carefully restricted arguments; it is not an external specialist review or a machine-checked formal derivation.

The scope leaves open substantial philosophical disagreements, including whether any coherent greatest power profile can be instantiated and how divine goodness interacts with preventable suffering. These are conclusions about the state of the debate, not unresolved factual or citation defects in this edition. The research package records actual primary, full-text scholarly, abstract-level, and secondary evidence separately, preserving transparency for later rereading.

Conclusion

Omnipotence is a rigorous philosophical problem concerning the relation of power to possibility, agency, time, and perfection. The classical tradition establishes the main terms: Augustine and Anselm distinguish power from defect, Maimonides and Aquinas examine the possible object of divine action, Descartes challenges the independence of eternal truths, and Leibniz situates creation within an order of compossible worlds and rational selection. Contemporary analytical philosophy sharpens the discussion by disambiguating task descriptions, worlds, histories, freedom, and the scope of divine attributes.

The apparent contradictions surrounding omnipotence have different sources. A self-referential stone puzzle can combine incompatible descriptions of continuing omnipotence and self-limitation. A contradictory figure may fail to describe any realizable state. An alternative past may be metaphysically possible in a broad sense yet inaccessible from an already fixed history. An outcome described as freely chosen may cease to have its intended property if produced by unilateral determination. In each case the evaluation depends on an explicitly identified domain and power relation.

The relationship between omnipotence and necessary goodness remains a substantive philosophical dispute. A being with additional immoral options may have more possibilities in a narrow sense without being greater overall. Conversely, eliminating those options by an evaluative stipulation cannot by itself prove that maximal power and moral perfection are jointly instantiated. The problem of evil likewise asks both what an omnipotent agent could prevent and what a perfectly good agent would have decisive reason to prevent; the propositions must be examined together without confusing their roles.

The comparative argument developed here requires each disputed claim to specify its modal domain, the agent's identity, the temporal and historical context, the mode of efficacy, and the criterion of greatness. These methodological requirements have recognized antecedents in Flint and Freddoso and in Hoffman and Rosenkrantz. This dossier's original analytical synthesis applies them across historical interpretations and contemporary cases to identify what further premises are necessary to advance from conceptual coherence to modal possibility and from possibility to an existence claim.

The resulting conclusion is a qualified but determinate one: omnipotence can be formulated as a coherent candidate perfection under independently defended modal and agential conditions; no unrestricted formula such as 'the power to do absolutely everything' resolves those conditions on its own. A complete theology must justify the chosen analysis, show its compossibility with the other attributes of the divine, and address its implications for freedom, history, and suffering. Philosophical progress consists in making those commitments explicit enough to test.