The Coherence of Theism

William Lane Craig considers one of the central concerns in the Philosophy of Religion - the coherence of theism. The first of these two highly philosophical articles considered the theistic attributes of necessity, aseity, incorporeality, omnipresence and eternity. This second article considers God's omniscience, simplicity, immutability, omnipotence and goodness. The talk covers only the attribute of omniscience, but can be accessed from Part 1 or Part 2 of the articles.

"If the object of worship in the Judeo-Christian tradition is indeed intended to be God--the ultimate reality responsible for the existence and activity of all else--and if the Anselmian conception is coherent, then it can be quite reasonable to hold that the God of Anselm is one and the same as the God of Abraham, Isaac, and Jacob, the God and Father of Jesus the Christ."
--Thomas V. Morris, "The God of Abraham, Isaac, and Anselm," Anselmian Explorations.


On the standard account of omniscience, for any person S, S is omniscient if and only if S knows every true proposition and believes no false proposition.  The standard account entails that if there are true propositions expressed by future-tense sentences, then God, since He is omniscient, must know those propositions.  If God exists in time, then He has literal foreknowledge of the events described by such propositions.  So if it is true that "Jones will mow his lawn next Saturday," then God, being omniscient, must know and have always known the proposition expressed by this sentence.  But this raises two difficult questions:  (1) If God has always believed this proposition and God cannot be mistaken, then is not Jones fated to mow his lawn on Saturday?  (2) If Jones's action is truly free, then how can God foreknow it?

The first question raises the issue of fatalism, the view that everything that happens happens of necessity.  Ancient Greek thought was infected with fatalism, and the Church Fathers felt obliged stoutly to resist it.  Greek fatalism was purely logical:  if it is true that some event will happen, then it will necessarily happen.  For the Church Fathers fatalism took on a theological coloring:  if God foreknows that some event will happen, then it will necessarily happen.  Almost every major Christian philosophical theologian after Origen had something to say about this question, the vast majority defending freedom and contingency, but some like Martin Luther and Jonathan Edwards, who denied libertarian freedom, endorsing it.

Aristotle had sought to avoid fatalism by denying the validity of the Principle of Bivalence for future contingent propositions; that is to say, he held that propositions about future contingents are neither true nor false.  Such a position would be compatible with divine omniscience, since no truths remain unknown by God; but such a solution was not open to the Church Fathers in light of the biblical doctrine that God has foreknowledge (Greek:  prognosis [Acts 2.23; I Pet. 1.1-2, 19-20]) and the many biblical examples of detailed prophecies of future events (e.g., Mk 14.18, 30).  Some contemporary philosophers, notably the Polish logician Lukasiewicz, have followed Aristotle's lead, but few have found this course attractive in view of the logical dislocations and implausibilities attending this position.[1]

 Theists who deny God's knowledge of future contingents have therefore felt obliged to re-define omniscience in such a way that God's ignorance of true future contingent propositions does not count against His being omniscient.  For example, it is typically proposed that S is omniscient if and only if S knows only and all true propositions which are such that it is logically possible for them to be known.  But it is not clear what more beyond truth is required for a proposition to be logically possible to know, in which case the revision is pointless.  Revisionists will say that true future contingent propositions are logically impossible to know, for if one knows them, then they are not contingently true.  But the revisionist's reasoning is fallacious.  For any future contingent proposition p, even if one grants that

    1.  Not-possibly (God knows p, and p is contingently true)


    2. p is contingently true,

it does not logically follow that

    3. Not-possibly (God knows p). 

It only follows that God does not in fact know p.  But it is still possible for Him to know p.  Thus, even on the revisionist definition of omniscience, God must know future contingent propositions, since it is logically possible for Him to know them.

But if God knows future contingent propositions, does this not imply fatalism?  The question here is why we should think that (1) is true.  The basic form of the fatalistic argument on behalf of (1) is as follows:

  4.  Necessarily (If God knows p, then p).

  5.  God knows p.


  6.  Necessarily (p).

Since p is necessarily true, it does not describe a contingent event.  In virtue of God's foreknowledge everything is fated to occur.

The problem with the argument is that it is just logically fallacious.  What is validly implied by (4) and (5) is not (6) but

  6'.  p.

It is correct that in a valid, deductive argument the conclusion follows necessarily from the premises.  That is to say, it is impossible for the premises to be true and the conclusion to be false.  But the conclusion itself need not be necessary.  The fatalist illicitly transfers the necessity of the inference to the conclusion itself.  What necessarily follows from (4) and (5) is just the contingent proposition (6').  But the fatalist confusedly thinks that the conclusion is itself necessarily true and so winds up with (6).  In so doing, he simply commits a common logical fallacy.

Undoubtedly a major source of the fatalist's confusion is his conflating certainty with necessity. One frequently finds in the writings of contemporary theological fatalists statements which slide from affirming that something is certainly true to affirming that it is necessarily true.  This is a mistake.  Certainty is a property of persons and has nothing to do with truth, as is evident from the fact that we can be absolutely certain about something and yet turn out to be wrong.  By contrast, necessity is a property of propositions, indicating that a proposition cannot possibly have a different truth value.  We can be wholly uncertain about propositions which are, unbeknownst to us, necessarily true (imagine some complex mathematical equation or theorem).  Thus, when we say that some proposition is "certainly true," this is but a manner of speaking indicating that we are certain that the proposition is true.  People are certain; propositions are necessary.

By confusing certainty and necessity, the fatalist makes his logically fallacious argument deceptively appealing.  For it is correct that from premises (4) and (5) we can be absolutely certain that the events described by p will happen.  But it is muddle-headed to think that because they will certainly happen they will necessarily happen.  We can be certain, given God's foreknowledge, that the events foreknown will not fail to happen, even though it is entirely possible that they fail to happen. They could fail to occur, but God knows that they will not.  Therefore, we can be sure that they will happen—and happen contingently.

Contemporary theological fatalists recognize the fallaciousness of the above form of the argument and therefore try to remedy the defect by making premise (5) also necessarily true:

  4.  Necessarily (If God knows p, then p).

  5'. Necessarily (God knows p). 


  6.  Necessarily (p).

Thus formulated, the argument is no longer logically fallacious, and so the question becomes whether (5') is true.  Now at face value, (5') is obviously false.  Christian theology has always maintained that God's creation of the world is a free act, that God could have created a different world, where p is false, or even no world at all.  To say that necessarily God knows p implies that this is the only world God could have created and thus denies divine freedom.

But theological fatalists have a different sort of necessity in mind when they say that necessarily God knows p.  What they are talking about is temporal necessity, or the necessity of the past.  Often this is expressed by saying that the past is unpreventable or unchangeable.  If some event is in the past, then it is now too late to do anything to affect it.  It is in that sense necessary.  Since God's foreknowledge of future events is now part of the past, it is now fixed and unalterable.  Therefore, it is said, (5') is true.

Unfortunately, theological fatalists have never provided an adequate account of this peculiar modality.  What is temporal necessity anyway, and why think that God's past beliefs are now temporally necessary?  We have yet to encounter an explanation of temporal necessity, according to which God's past beliefs are temporally necessary, which does not reduce to either the unalterability or the causal closedness of the past. 

But interpreting the necessity of the past as its unalterability (or unchangability or unpreventability) is clearly inadequate, since the future, by definition, is just as unalterable as the past.  By definition the future is what will occur, and the past is what has occurred.  To change the future would be to bring it about that an event which will occur will not occur, which is self-contradictory.  It is purely a matter of definition that the past and future cannot be changed, and no fatalistic conclusion follows from this truth.  For we need not be able to change the future in order freely to determine the future.  If our actions are freely performed, then it lies within our power to determine causally what the course of future events will be, even if we do not have the power to change the future. 

The fatalist will insist that the past is necessary in the sense that we do not have a similar ability to determine causally the past.  The orthodox theologian may happily concede the point:  backward causation is impossible.  But the causal closedness of the past does not imply fatalism.  For freedom to refrain from doing as God knows one will do does not involve backward causation.  The orthodox theologian may grant that there is nothing I can now do to cause or bring about the past.  In particular, I cannot cause God to have had in the past a certain belief about my future actions.  Nevertheless,  as the medieval theologian William Ockham saw, it may well lie within my power to freely perform some action A, and if A were to occur, then the past would have been different than it in fact is.  Suppose, for example, that God has always believed that on August 23, 2010 Jones would mow his lawn.  Let us suppose that up until the time arrives Jones has the ability to mow or not mow his lawn.  If Jones were to decide not to mow his lawn, then God would have always held a different belief than the one He in fact holds.  For if Jones were to decide not to mow his lawn, then different future contingent propositions would have been true, and God, being omniscient, would have known them.  Thus, He would have had different foreknowledge than that which He in fact has.  Neither the relation between Jones's action and a corresponding future contingent proposition about it nor the relation between a true future contingent proposition and God's believing it is a causal relation.  Thus, the causal closedness of the past is irrelevant.  If temporal necessity is merely the causal closedness of the past, then it is insufficient to support fatalism.

No fatalist has successfully explicated a conception of temporal necessity which does not amount to either the unalterability or the causal closedness of the past.  Typically, fatalists just assert some sort of "Fixed Past Principle" to the effect that it is not within my power to act in such a way that, if I were to do so, the past would have been different--which begs the question.  On analyses of temporal necessity which are not reducible to either the unalterability or the causal closedness of the past, God's past beliefs always turn out not to be temporally necessary.[2]   Thus, (5') is false, and the argument for theological fatalism is unsound. 

If divine foreknowledge and future contingency are compatible, the question remains as to how could God know future contingent propositions.  Process theologians typically deny divine foreknowledge because,  given the contingency of the future, it is impossible for anyone, even God, to have knowledge about what will happen.  In assessing this objection, it will be helpful to distinguish two models of divine cognition:  the perceptualist model and the conceptualist model.  The perceptualist model construes divine knowledge on the analogy of sense perception.  God looks and sees what is there.  Such a model is implicitly assumed when people speak of God's "foreseeing" the future or having "foresight" of future events.  Given a dynamic theory of time, the perceptualist model of divine cognition does run into real problems when it comes to God's knowledge of the future, for, since future events do not exist, there is nothing there to perceive.

By contrast on a conceptualist model of divine knowledge, God does not acquire His knowledge of the world by anything like perception.  His knowledge of the future is not based on His "looking" ahead and "seeing" what lies in the future (a terribly anthropomorphic notion in any case).  Rather God's knowledge is self-contained; it is more like a mind's knowledge of innate ideas.  As an omniscient being, God has essentially the property of knowing all true propositions; there are true propositions about future contingents; ergo, God knows all true propositions concerning future contingents. So long as we are not seduced into thinking of divine foreknowledge on the model of perception, it is no longer evident why knowledge of future events should be impossible.  Unless the detractor of divine foreknowledge can show some incoherence in the notion of innate knowledge, his objection cannot even get off the ground.

We can push the conceptualist analysis a step further.  According to the Counter-Reformation theologian Luis Molina, logically prior to the divine decree to create a world, God possesses not only knowledge of everything that could happen (natural knowledge) but also everything that would happen in any appropriately specified set of circumstances (middle knowledge).  God's natural knowledge is His knowledge of all necessary truths.  By means of it God knows what is the full range of possible worlds.  He knows, for example,  that in some possible world Peter freely denies Christ three times and that in another world Peter freely affirms Christ under identical circumstances, for both are possible.  God's middle knowledge is His knowledge of all contingently true counterfactual propositions, including propositions about creaturely free actions.  A counterfactual proposition is a conditional proposition in the subjunctive mood.  For example, logically prior to His creative decree, God knew that if Peter were in circumstances C, he would freely deny Christ three times.  These counterfactuals serve to delimit the range of possible worlds to worlds which are feasible for God to actualize.  For example, there is a possible world in which Peter freely affirms Christ in precisely the same circumstances in which he in fact denied him; but given the counterfactual truth that if Peter were in precisely those circumstances he would freely deny Christ, then the possible world in which Peter freely affirms Christ in those circumstances is not feasible for God.  God could make Peter affirm Christ in those circumstances, but then his confession would not be free.  By means of His middle knowledge, God knows what is the proper subset of possible worlds which are feasible for Him,  given the counterfactuals that are true.  God then decrees to create certain free creatures in certain circumstances and, thus, on the basis of his middle knowledge and His knowledge of His own decree, God has foreknowledge of everything that will happen (free knowledge).  In that way, He knows, simply on the basis of His own internal states and without any need of any sort of perception of the external world, that Peter will freely deny Christ three times.

Thus on the Molinist scheme, we have the following logical order:

  Moment 1: . . . O  O  O  O  O  O . . .

  Natural Knowledge:  God knows the range of possible worlds

  Moment 2: . . . O  O  O . . . 

  Middle Knowledge:  God knows the range of feasible worlds

 Divine Creative Decree

  Moment 3: O

  Free Knowledge:  God knows the actual world

Of course, basing divine foreknowledge in divine middle knowledge raises inevitably the question of the basis of God's middle knowledge.  The Molinist can respond either that God knows the individual essence of every possible creature so well that He knows just what that creature would do under any set of circumstances God might place him in, or that God, being omniscient, simply discerns all the truths there are and, prior to the divine decree, there are not only necessary truths, but counterfactual truths and therefore God possesses not only natural knowledge, but middle knowledge as well.  Thus, a conceptualist model along the lines of middle knowledge furnishes a perspicuous basis for God's knowledge of future contingents.

Does God, then, have middle knowledge?  Consider the following argument:

  1. If there are true counterfactuals about creaturely free choices, then God knows these truths.
  2. There are true counterfactuals about creaturely free choices.
  3. If God knows true counterfactuals about creaturely free choices, God knows them either logically prior to the divine creative decree or only logically posterior to the divine creative decree.
  4. Counterfactuals about creaturely free choices cannot be known only logically posterior to the divine creative decree.
  5. Therefore, God knows true counterfactuals about creaturely free choices. (MP,1, 2)
  6. Therefore, God knows true counterfactuals about creaturely free choices either logically prior to the divine creative decree or only logically posterior to the divine creative decree. (MP,3, 5)
  7. Therefore, God knows true counterfactuals about creaturely free choices logically prior to the divine creative decree. (DS,4, 6)

-- which is the essence of the doctrine of divine middle knowledge.

The truth of (1) is required by the standard definition of omniscience.  As for (2), a little reflection reveals how pervasive and indispensable such counterfactual truths are to rational conduct and planning.  We sometimes base our very lives upon them.  Moreover, Scripture itself gives examples of such true counterfactuals (I Cor. 2.8). 

The most common objection urged against (2) is the so-called "grounding objection." Detractors of middle knowledge typically claim that if such counterfactuals have any truth value, they are uniformly false, since there is no ground of their truth.  Grounding objectors have never clearly articulated or defended the theory of truth which the objection tacitly presupposes.  It appears to assume some version of what is called "Truthmaker Theory," according to which true propositions are made to be true by certain entities in the world.  Truthmaker Theory is a controversial position, however, and even its proponents typically reject "truthmaker maximalism," the doctrine that all types of true propositions have truthmakers.  No grounding objector has yet to answer Plantinga's retort:  "It seems to me much clearer that some counterfactuals of freedom are at least possibly true than that the truth of propositions must, in general, be grounded in this way."[3] Moreover, acceptable truth-makers for counterfactuals of creaturely freedom are available.  Alfred Freddoso suggests, for example, that counterfactuals of creaturely freedom are grounded by the fact that a relevant indicative proposition would have grounds of its truth.  Thus, the truthmaker of the counterfactual If Peter were in C, he would deny Christ three times is the fact or state of affairs that the proposition Peter denies Christ three times would have a truthmaker under the relevant condition.

Premise (3) of the argument states logically exhaustive alternatives for an omniscient deity and so must be true.  Finally, (4) must be true because if counterfactuals of creaturely freedom were known only posterior to the divine decree, then it is God who determined what every creature would do in every circumstance.  Augustinian-Calvinist thinkers bear witness to the truth of this premise in their affirmation of compatibilist theories of creaturely freedom.  They thereby testify that God's all-determining decree precludes libertarian freedom, which is the sort of freedom with which we are here concerned.  Thus, if God knows counterfactual truths about us only posterior to His decree, then there really are no counterfactuals about creaturely free choices.  If there are such counterfactuals, they must be true logically prior to the divine decree.

Given the truth of the premisees, the conclusion follows that prior to His creative decree God knows all true counterfactuals of creaturely freedom, which is to say that He has middle knowledge.  If this conclusion is correct, then we have a theological tool of remarkable fecundity when we come to deal with other questions, such as the nature of divine providence.


Divine simplicity is a doctrine inspired by the Neo-Platonic vision of the ultimate metaphysical reality as the absolute One.  It holds that God, as the metaphysical ultimate, is an undifferentiated unity, that there is no complexity in His nature or being.  As such, this is a radical doctrine which enjoys no biblical support and even is at odds with the biblical conception of God in various ways.  According to the doctrine of divine simplicity God has no distinct attributes, He stands in no real relations, His essence is not distinct from His existence, He just is the pure act of being subsisting.  All such distinctions exist only in our minds, since we can form no conception of the absolutely simple divine being. While we can say what God is not like, we cannot say what He is like, except in an analogical sense--which must in the end fail, since there is no univocal element in the predicates we assign to God--, leaving us in a state of genuine agnosticism about the nature of God.  Indeed, on this view God really has no nature; He is simply the inconceivable act of being.

Why should we adopt so extraordinary a doctrine?  It will be recalled that Thomas Aquinas argued for the existence of such a simple being by means of his cosmological argument.  But Thomas’s famous argument from contingent beings (beings whose essence is distinct from their existence),  forces us at most to postulate the existence of a being whose essence is such that it exists necessarily, a metaphysically necessary being.  It need not commit us to divine simplicity.

The doctrine is open, moreover, to powerful objections.  For example, to say that God does not have distinct properties seems patently false:  omnipotence is not the same property as goodness, for a being may have one and not the other.  It might be said that God's  omnipotence and goodness differ in our conception only, as manifestations of a single divine property, just as, say, “the morning star” and “the evening star” have different senses but both refer to the same reality, namely, Venus.  But this response is inadequate.  For being the morning star and being the evening star are distinct properties both possessed by Venus; the same entity has these two distinct properties.  In the same way being omnipotent and being good are not different senses for the same property (as are, say, being even and being divisible by two) but are clearly distinct properties.  In God's case the same entity possesses both of these distinct properties. 

Moreover, if God is identical with His essence, then God cannot know or do anything different than what He knows and does.  He can have no contingent knowledge or action, for everything about Him is essential to Him.  But in that case all modal distinctions collapse and everything becomes necessary.  Since "God knows that p" is logically equivalent to "p is true," the necessity of the former entails the necessity of the latter.  Thus, divine simplicity leads to an extreme fatalism, according to which everything that happens does so, not with temporal necessity, but with logical necessity.

It might be said that Thomas could escape this unwelcome conclusion by his doctrine that God stands in no real relations to creatures.  As a simple being, God transcends all the Aristotelian distinctions among substance and accidents, and since relations are one type of accident, God has no relational properties and stands in no real relations to things outside Himself.  Things stand in real relations to God, but the situation is not symmetrical:  God's relations to creatures are just in our minds, not in reality.  Thus God is perfectly similar in all logically possible worlds which we can imagine, but in some worlds either different creatures stand in relation to God or no creatures at all exist and are related to God.  Thus the same simple cognitive state counts as knowledge of one conjunction of propositions in one world and another conjunction of propositions in a another world.  Similarly, the same act of power (which just is the divine being) has in one world effects really related to it in the form of creatures and in another world no such effects.

But Thomas's doctrine only serves to make divine simplicity more incredible.  For it is incomprehensible how the same cognitive state can be knowledge that "I exist alone" in one world and that "I have created myriads of creatures" in another. Moreover, what God knows is still different, even if God's cognitive state is the same; and since God is His knowledge, contingency is introduced into God.  It is equally unintelligible why a universe of creatures should exist in some worlds and not others if God's act of power is the same across worlds (for more on this see chapter 28 on the doctrine of creation)..  The reason cannot be found in God, since He is absolutely the same.  Neither can the reason be found in creatures themselves, for the reason must be explanatorily prior to creatures.  Thus, to contend that God stands in no real relations to things is to make the existence or non-existence of creatures in various possible worlds independent of God and utterly mysterious. 

Finally, to say that God’s essence just is His existence seems wholly obscure, since then there is in God’s case no entity that exists; there is just the existing itself without any subject.  Things exist; but it is unintelligible to say that exists just exists. 

In short, we have no good reason to adopt and many reasons to reject a full-blown doctrine of divine simplicity.  Still, that does not mean that the doctrine is wholly without merit.  For example, it is surely correct to emphasize that God is not composed of mind and body; rather He is pure mind.  As such, God is remarkably simple in that such an immaterial substance is not composed of pieces, or separable parts in the way that a material object is.  Moreover, some thinkers such as William Alston, while rejecting complete simplicity, have advocated that God's knowledge be construed as simple.  On Alston's view God has a simple intuition of all of reality, which we human cognizers represent to ourselves propositionally.  Such a view is in line with Aquinas's adaptation of the Augustinian notion of the Divine Ideas.  In order to preserve divine aseity in the face of Platonism, Augustine located the Platonic forms in God's mind as the Divine Ideas.  Aquinas went further by contending that God does not, strictly speaking, have a plurality of Divine Ideas but rather an undifferentiated knowledge of truth.  We finite knowers break up God's undivided intuition into separate ideas.  Similarly, Alston maintains that God's knowledge is strictly non-propositional, though we represent it to ourselves as knowledge of distinct propositions.  Thus, we say, for example, that God knows that Mars has two moons , and He does indeed, know that, but the representation of His knowing this proposition is a merely human way of stating what God knows in a non-propositional manner.  Such a conception of divine knowledge has the advantage that it enables us to embrace conceptualism without committing us to an actual infinite of divine cognitions or Divine Ideas.

A modified doctrine of divine simplicity might be useful in other ways.  For example, we might think of God's act of creation or of conservation of the world, not as a multiplicity of individual acts of creating this or that thing, but as a single act comprising all that exists outside Himself.  This is an area which deserves further exploration by Christian thinkers.


For Aristotle God was the Unmoved Mover, the unchanging source of all change.  God's immutability is also attested in Scripture (Mal. 3.6; Jas. 1.17).  But the biblical authors did not have in mind the radical changelessness contemplated by Aristotle nor the immutability required by the doctrines of essential divine timelessness or simplicity.  They were speaking primarily of God's unchanging character and fidelity.  But if God is essentially timeless and simple, He must be utterly incapable of change.  We can distinguish usefully between two types of change:  intrinsic change and extrinsic change.  An intrinsic change is a non-relational change, involving only the subject.  For example, an apple changes from green to red.  An extrinsic change is a relational change, involving something else in relation to which the subject changes.  For example, Jones becomes shorter than his son, not by undergoing an intrinsic change in his height, but by being related to his son as his son undergoes intrinsic change in his height.  Jones changes extrinsically from being taller than his son to being shorter than him because his son is growing.  Divine timelessness or divine simplicity would require that God undergo neither intrinsic nor extrinsic change.  For in either case a relation of before and after would be generated by such changes which would serve to locate God temporally with respect to those changes.  Thus, God would have to be incapable of even the slightest alteration.

We have seen reason, however, to reject God's essential timelessness and simplicity.  Is there some other reason for thinking God to be immutable?  Sometimes divine immutability is said to be a necessary correlate of divine perfection.  For if God were to change, it could only be for the worse.  Being essentially perfect, God must therefore be changeless.  But this reasoning seems clearly incorrect.  A perfect being need not change “vertically,” so to speak, on the scale of perfection and, hence, for the worse but could change “horizontally,” remaining equally perfect in both states.  For example, for God to change from knowing “It is now t1” to knowing “It is now t2” is not a change for the worse in God; on the contrary, it is a sign of His perfection that He always knows what time it is. 

We have argued that in virtue of His real, causal relation to the temporal world, God must minimally undergo extrinsic change and therefore be temporal--at least since the moment of creation.  Moreover, God's knowledge of tensed facts, implied by His omniscience, requires that since the moment of creation He undergoes intrinsic change as well, since He knows what is now happening in the universe.  Thus, God is not immutable in a strong sense.

Rejection of radical immutability leaves it open for us to affirm nonetheless that God is immutable in the biblical sense of being constant and unchangeable in His character.  Moreover, He is immutable in His existence (necessity, aseity, eternity) and His being omnipresent, omniscient, and omnipotent.  These essential attributes are enough to safeguard God's perfection without having Him frozen into immobility. 

Finally, if we do adopt our proposed construal of divine eternity as timelessness without creation and temporality with creation, then we can affirm God's changelessness, if not His immutability, in the timeless state of existing alone without the world.  It bespeaks God's freedom and condescension that He should quit such a state for the sake of the creation and eventual salvation of temporal creatures like ourselves.  God could have remained changeless had He wished to; the fact that He did not is testimony to both His love and freedom.


Although one of the biblical names of God is El-Shaddai (God Almighty), the concept of omnipotence has remained poorly understood due to its recalcitrance to analysis.  Few thinkers, aside from Descartes, have been willing to affirm that the doctrine means that God can do just anything--for example, make a square triangle.  Such a view has been construed as affirming universal possibilism, the doctrine that there are no necessary truths.  For on this view an omnipotent deity could have brought it about that even logical contradictions be true and tautologies be false, as inconceivable as this may seem to us.  But such a doctrine seems incoherent.  For is the proposition There are no necessary truths itself necessarily true or not?  If so, then the position is self-refuting.  If not, then that proposition is possibly false, that is to say, God could have brought it about that there are necessary truths.  Using possible worlds semantics, we may say that there is, therefore, a possible world in which God brings it about that there are propositions which are true in every possible world.  But if there are such propositions, then there is no world in which it is the case that there are no propositions true in every possible world, that is, it is not possible that there are no necessary truths, which contradicts universal possibilism.

Moreover, Descartes's position is incredible.  It asks us to believe, for example, that God could have brought it about that He created all of us without His existing, that is to say, there is a possible world in which both God does not exist and He created all of us.  This is simply nonsense.

One must therefore delineate more carefully what is understood as omnipotence. Unfortunately this has not been easy.  Philosophical  discussion appeared to be stalled until the remarkable piece "Maximal Power" by Thomas Flint and Alfred Freddoso appeared in 1983.[4]   Their analysis is closely connected with the Ockhamist solution to theological fatalism and Molina's doctrine of middle knowledge.  A key insight into the concept of omnipotence is that it should be defined in terms of the ability to actualize certain states of affairs, rather than in terms of raw power.  Thus, omnipotence should not be understood as power which is unlimited in its quantity or variety.  Such an understanding leads immediately to the hoary problem of whether God can make a stone too heavy for Him to lift.  If God possesses the power to do the one, then He lacks the power to do the other.  Such conundrums arise, not from some lack of power on God's part, but from a faulty concept of omnipotence.  If we understand omnipotence in terms of ability to actualize states of affairs, then it is no attenuation of God's omnipotence that He cannot make a stone too heavy for Him to lift, for, given that God is essentially omnipotent, "a stone too heavy for God to lift" describes as logically impossible a state of affairs as does "a square triangle" and thus describes nothing at all.

Shall we say, then, that an agent S is omnipotent if and only if S can actualize any state of affairs which is broadly logically possible?  No, for certain states of affairs may be logically possible but due to the passage of time may no longer be possible to actualize.  For example, "The Cubs' winning the 1968 World Series" describes a logically possible state of affairs; but it is no longer possible to actualize this state of affairs, at least directly.  To actualize directly such a state of affairs now would be tantamount to changing the past or to backward causation, both of which are broadly logically impossible.  Therefore, the inability of S to actualize directly such a state of affairs at this time should not count against S's omnipotence.  But could S perhaps actualize such a state of affairs indirectly?  Suppose there were some action A which S could take, and if S were to take A, then God would have foreknown this and brought about a Cubs victory in 1968.  Then S would now seem to have the power to actualize the state of affairs "The Cubs' winning the 1968 World Series" indirectly.  Whether S in fact has such a power depends on whether certain "backtracking" counterfactuals, that is, counterfactuals which involve the past's being different than it was, happen to be true.  Let us call past states of affairs which are indirectly actualizable by someone the "soft" past and states which are not indirectly actualizable by someone the "hard" past.  Accordingly, we must amend our analysis of omnipotence in such a way that S's inability to actualize states of affairs inconsistent with the hard past does not count against S's omnipotence.

Shall we say, then, that an agent S is omnipotent at a time t if and only if S can at t actualize any state of affairs which is broadly logically possible for someone sharing the same hard past with S to actualize at t ?  It seems not.  For counterfactuals about creaturely free actions raise a further problem.  The counterfactual If Jones were in C, he would freely write to his wife is within Jones's control, that is to say, how Jones freely decides to behave in C entails whether that counterfactual is true or not.  By contrast that counterfactual does not lie within Smith's control:  if that counterfactual is true, then it is logically impossible for Smith to do anything that would entail its falsity.  He could prevent Jones from being in C, but he can do nothing which would entail the falsity of the counterfactual.  Thus, one has control over counterfactuals about one's own free decisions but not over counterfactuals about the free decisions of others.  That implies that an adequate definition of omnipotence cannot require S to be able to actualize states of affairs described by counterfactuals about the free decisions of other agents.  If the above counterfactual is true, S cannot be expected to actualize a world in which Jones is in C and freely refrains from writing to his wife, for this is to demand the logically impossible of S.  Our provisional definition would require S to have the ability to actualize all the same states of affairs as Jones if they share the same hard past at t, which is to ask the logically impossible of S.  We must therefore amend our analysis of omnipotence so as not to require such an ability in S.

Shall we say, then, that S is omnipotent at a time t if and only if S can at t actualize any state of affairs which is broadly logically possible for S  to actualize, given the same hard past at t and the same true counterfactuals about creaturely free acts of others?  This seems almost right.  But it is open to the complaint that if S is essentially incapable of any particular action, no matter how trivial, than S's inability to perform that action does not count against his omnipotence.  For in that case it is logically impossible for S to perform the act in question, since he is essentially incapable of it.  Therefore we need to broaden the definition so as to require S to perform any action which any agent in his situation could perform.  The trick is that we must not thereby require S to actualize states of affairs described by counterfactuals about the free acts of others.

The following analysis would seem to do the trick: S is omnipotent at a time t if and only if S can at t actualize any state of affairs which is not described by counterfactuals about the free acts of others and which is broadly logically possible for someone to actualize, given the same hard past at t and the same true counterfactuals about creaturely free acts of others.  Such an analysis successfully sets the parameters of God's omnipotence without imposing any non-logical limit on His power.


Believers in the monotheistic tradition have always held that God is perfectly good, and Christians have thought of God as the fount of all varieties of goodness, whether moral, metaphysical, aesthetic, and so forth.  Here our interest is in God's moral goodness.  Some versions of the axiological argument for God's existence imply that goodness is somehow rooted in God Himself.  But ever since Plato the claim that moral values and duties are founded in God has been criticized as problematic.  In a famous dilemma in his dialogue Euthyphro, Plato asks, in effect, whether something is good because it is approved by God or whether something is approved by God because it is good.  Either horn of this dilemma has been said to lead to untenable consequences.  If we say that some action is good or right for the mere reason that God wills it, then morality is fundamentally arbitrary.  God could have willed that cruelty be good and love evil, and we should have been obliged to hate others and seek to do them harm.  Not only is this unconscionable, but it appears to make the claim that God is good vacuous.  To claim that God is good seems to mean no more than that God does whatever He wants!  On the other hand, if God wills that we perform some action because it is the right or a good thing to do, then moral values are not based in God after all but exist independently of Him.  Such an alternative is taken to be incompatible with classical theism because it compromises the sovereignty and aseity of God.  God is Himself duty-bound to obey certain moral principles not of His own creation, but, as it were, imposed on Him.  Evidence that moral value is independent of God is sometimes said to be found in the fact that we can apprehend moral values and duties quite independently of belief in God.

In sorting through the tangle of issues raised by this objection, we shall find it helpful to distinguish clearly various areas of Moral Theory:[5]

                                         Moral Theory

             Normative Ethics                             Meta-Ethics

Particular   General        Moral         Definition       Moral          Moral
 Claims      Systems    Linguistics   of Morality   Ontology   Epistemology

The claim that moral values and duties are rooted in God is a Meta-Ethical claim about Moral Ontology, not about Moral Linguistics or Epistemology.  It is fundamentally a claim about the metaphysical status of moral properties, not a claim about the meaning of moral sentences or about the justification or knowledge of moral principles.

These distinctions serve to sweep away in a single stroke all those objections to theistic Meta-Ethics based on linguistic or epistemological considerations.  For example, the theist should not be understood to be offering a definition of "good" or "right" in theistic terms (e.g., "willed by God").  The theist does offer a Definition of Morality in the sense that moral values and duties are to be explicated in terms of God's nature and will rather than of self-interest, social contract, common happiness, or what have you.  But his aim is not to analyze the meaning of moral terms; rather he aims to provide an ontological grounding for objective moral values and duties.  Therefore objections to the effect that we can understand the meaning of statements like "Torture of political prisoners is wrong" without reference to God are quite beside the point.  Similarly, when it is said that the statement "God is good" becomes trivial on theistic Meta-Ethics, this objection misconstrues the theistic position as a meaning-claim.  A statement like "God is good" may be taken as a synthetic statement expressing a proposition which is metaphysically necessary both in the sense that the proposition is true in all possible worlds and in the sense that goodness is an essential property of God.  Or again, the theist will agree quite readily (and, if he is a proponent of the axiological argument for God's existence, even insist) that we do not need to know or even believe that God exists in order to discern objective moral values or to recognize our moral duties.

Although theistic Meta-Ethics assumes a rich variety of forms, there has been in recent years a resurgence of interest in Divine Command Morality, which understands our moral duties as our obligations to God in light of His moral commands, for example, "You shall love the Lord your God with all your heart," "You shall love your neighbor as yourself," and so on.  Our moral duties are constituted by the commands of a just and loving God.  For any action A and moral agent S, we can explicate the notions of moral requirement, permission, and forbiddenness of A for S as follows:

  A is required of S if and only if a just and loving God commands S to do A.

  A is permitted for S if and only if a just and loving God does not command S not to do A.

  A is forbidden to S if and only if a just and loving God commands S not to do A.

Since our moral duties are grounded in the divine commands, they are not independent of God nor, plausibly, is God bound by moral duties, since He does not issue commands to Himself.

If God does not fulfil moral duties, then what content can be given to the claim that He is good?  Here Kant's distinction between following a rule and acting in accordance with a rule has proved helpful.  God may act naturally in ways which for us would be rule-following and so constitutive of goodness in the sense of fulfilling our moral duties, so that God can be said similarly to be good in an analogical way.  This fact also supplies the key to the arbitrariness objection.  For our duties are determined by the commands, not merely of a supreme potentate, but of a just and loving God.  God is essentially compassionate, fair, kind, impartial, and so forth, and His commandments are reflections of His own character.  Thus, they are not arbitrary, and we need not trouble ourselves about counterfactuals with impossible antecedents like "If God were to command child abuse ...." God may be said to be good in the sense that He possesses all these moral virtues--and He does so essentially and to the maximal degree!  Thus, God's axiological perfection should not be understood in terms of duty-fulfillment, but in terms of virtue.  This conception helps us to understand the sense in which God is to be praised:  not in the sense of commendation for fully executing His duties or even for His acts of supererogation, but rather in the sense of adoration for His axiological perfection.

Nonetheless, the fact that God is not duty-bound should alert us to the fact that He may well have prerogatives (for example, taking human life at His discretion) which are forbidden to us.  Taking the life of an innocent person is something we have no right to do; but God is not similarly restricted.  God's having no duties also helps to explain how God can command a person to perform an action which would be sinful were the person to undertake such an action on his own initiative, but which is his moral duty in virtue of God's command.  The most celebrated example is Abraham's sacrificing his son Isaac, an act which would have been murder in the absence of any command of God but which became Abraham's moral duty in light of the divine command given him.  This is not to say that God can bring it about that murder be good, but rather that He can command an act which would have been murder had it been undertaken in the absence of a divine command.  This also not to say that God could have brought it about that it be a general moral duty for people to kill one another.  The case of Abraham and Isaac is the exception that proves the rule.  Issuing a general command that we should seek one another's harm would be contrary to God's loving nature, but in the extraordinary case of Abraham and Isaac, it was not unloving of God to so try Abraham's devotion, and God had good reasons for testing him so severely.

The question might be pressed as to why God's nature should be taken to be definitive of goodness.  But unless we are nihilists, we have to recognize some ultimate standard of value, and God seems to be the least arbitrary stopping point.  Moreover, God's nature is singularly appropriate to serve as such a standard.  For by definition, God is a being worthy of worship.  And only a being which is the locus and source of all value is worthy of worship.

Go to Part 1


analogical sense
commands of a just and loving God
conceptualist model
counterfactual propositions
Definition of Morality
Divine Command Morality
extrinsic change
free knowledge
grounding objection
individual essence
intrinsic change
middle knowledge
moral goodness
Moral Linguistics
Moral Ontology
moral requirement
moral virtues
natural knowledge
perceptualist model
Principle of Bivalence
states of affairs
temporal necessity
truthmaker maximalism
Truthmaker Theory
universal possibilism


[1] For discussion, see William Lane Craig, Divine Foreknowledge and Human Freedom: The Coherence of Theism: Omniscience, Brill's Studies in Intellectual History 19 (Leiden:  E.J. Brill, 1991), ch.4.

[2] See, for example, Alfred J. Freddoso, "Accidental Necessity and Logical Determinism," Journal of Philosophy 80 (1983): 257-78.  The implications of this, though startling, are not unique to divine foreknowledge but also follow from retro-causation, time travel, precognition, and the Special Theory of Relativity.

[3] Alvin Plantinga, "Reply to Robert Adams," in Alvin Plantinga, ed. James E. Tomberlin and Peter Van Inwagen, Profiles 5 (Dordrecht:  D. Reidel, 1985), p.378.

[4] Thomas P. Flint and Alfred J. Freddoso, “Maximal Power,” in The Existence and Nature of God (Notre Dame:  University of Notre Dame Press, 1983), pp.81-113.

[5] See Walter Sinnott-Armstrong, "Moral Skepticism and Justification," in Moral Knowledge?, ed. Walter Sinnott-Armstrong and Mark Timmons (New York:  Oxford University Press, 1996), pp.4-5.