The purpose of this part of the course is fourfold. First, we continue to investigate the concept of God by looking carefully at the central divine attributes of providence, perfect goodness, omniscience, and omnipotence. More specifically, we will be concerned with some of the philosophical difficulties that arise when we try to give adequate philosophical accounts of these attributes. (Question: Why undertake such an investigation? Answer: Same reasons St. Thomas gave in SCG I, chapter 2.) Second, again building on what went before, we see more specifically how the twin dangers of anthropomorphism and obscurantism can enter into an examination of the divine nature and attributes. Third, we see several cases in which even the strict bounds of orthodoxy leave open important questions about how to understand doctrines and leave open interesting and exciting philosophical possibilities. Fourth, we see some of the most recent philosophical work on the divine nature; this is an ongoing project which, for various reasons, has migrated from theologians to philosophers in recent years.

Readings: St. Thomas Aquinas, Summa Contra Gentiles III, chap. 64 Thomas P. Flint, "Two Accounts of Providence"

A. The Elements of Providence

1. SCG III, 64:

    a. [2]: God is the one who orders all things to their end, in much the same way that the general orders the parts of the army together to accomplish His end. Since it is God who knows, loves and possesses [or better: is] goodness, it is God who governs all things.

    b. [4]: All created things have being, power and action only insofar as God gives them being and power, conserves them and their power in being, and acts with and through them when they bring about their characteristic effects. So everything that occurs in the universe is or involves an action of God's (even if this action results from only a permission instead of an intention on God's part). From this it follows that God, who acts by understanding and will, controls the universe as one whose effect the universe is.

    c. [9]: God's end in creating the universe is to establish a universe whose order reflects the divine goodness. So it is such an order that God primarily intends (though it is not known to us why God should choose this particular order rather than others that would also have reflected His goodness in an appropriate way). But to impose order on things is just to govern them. Therefore, God governs all things.

    d. [12]: Here point out that God's main concern is the order of the whole, so that certain things might fail to achieve what is their end by nature, even though this failure is itself incorporated into the goodness of the whole.

2. Providence and the other attributes:

Divine providence thus involves God's knowledge, since God by His intellect formulates a plan for the created universe in every detail; God's power, since God by His will exercises His power in order to create, conserve and act with created things in bringing about their effects; and God's goodness, since God's formulated and executed plan is good both in its end (the order of the whole) and in its means to that end.

(Notice, by the way, that God's foreknowledge is a result of His providential governance of the world and not the result of His observing the world, including its future, as cosmic bystander, as it were. This is one reason why the threat that this foreknowlege poses to human freedom is much more serious than that posed by the mere fact that future contingent propositions are true.)

3. Six theses about providence that are common to Molinists and Thomists:

From this we may conclude that the following thesis is part of the orthodox conception of divine providence:

Thesis One: Everything that occurs in the created world is either specifically intended by God or specifically permitted by God.

    Say something about the consequences of denying this, e.g., (i) that God might be continually and everlastingly frustrated by the free refusal of his creatures to cooperate with Him, and (ii) that prophecy about the future could not be certain ahead of time if it involved free human decisions and actions (such as Peter's denial and other prophecies concerning Christ), and (iii) the fascination with risk-taking is highly anthropomorphic, and (iv) that the reasons usually cited by such a view (such as a concern to preserve human freedom) beg the question against both Molinism and Thomism.

    Remembering the via remotionis and being mindful of the danger of anthropomorphism we can tentatively imagine a picture that goes along with the doctrine of providence. The picture is basically that of God formulating one plan from among many available to Him and then executing that plan. (Orthodoxy dictates that God could have chosen not to created anything at all.) This picture essentially involves God's doing this freely, and so we rule out the view that given God's nature, there is just one plan that God could have carried out. Let us call this view necessitarianism, according to which there is just one possible world, viz., the actual world. So this gives us:

Thesis Two: Necessitarianism is false. That is, there are many (indeed, uncountably many) possible worlds that God could have willed to actualize, including that world in which God never creates any creatures at all.

    Likewise, there is another extreme to be avoided; this would be the claim that there are absolutely no antecedent constraints on God's choice of a plan and of a world to actualize. According to this view, there are no necessary truths which would have remained true no matter what God had decided to do and no necessary falsehoods which would have remained false no matter what God had decided to do. On this view, it is up to God to decide whether, say, two plus two equals four, or whether it is possible for something to both exist and not exist at the same time, or whether He Himself is perfectly good, or perfectly knowledgeable, etc. Let us call this view pure possibilism.

Thesis Three: Pure Possibilism is false. That is, there are true propositions whose truth or falsity is not within God's control and which would have been true no matter what God had decided to do.

    Now from these two theses a certain picture emerges. God acts by understanding and will and makes a choice, but the choice is among various alternatives that God chooses from among but that God does not freely constitute as alternatives. Rather, the alternatives are set independently of God's free act of will. So, for instance, it is not up to God that every alternative available to Him is one in which two plus two equals four. Likewise, it is not up to God that the alternatives open to Him are just these and no more.

    So here is the picture. Before God formulates or chooses a plan He is faced with a set of alternative choices, each resulting in a particular world. It is not up to God to decide which worlds are alternatives; rather God decides which of the alternatives will be actualized.

Thesis Four: Before formulating or choosing a comprehensive world-plan, God is in a Creation Situation, where a creation situation is a set of truths which God knows prior to making any decision and which defines the alternative world-plans that God is able to choose and actualize. (It is not up to God to determine the elements of the creation situation He finds Himself in.) Let us say that the worlds that embody the world-plans defined by a creation situation (CS) are feasible with respect to (CS).

Thesis Five:If God finds Himself in a given creation situation (CS)1, then He is able to actualize a given possible world w only if w is feasible with respect to (CS)1.

Since I have been referring to possible worlds, I should introduce and define some technical notions:

    A possible world is a complete way things might have been, including a whole history of events from the beginning to the end. The actual world is one of the possible worlds; all the other possible worlds are merely possible.

    A proposition p is (metaphysically) necessary just in case p is true in every possible world.

    A proposition p is (metaphysically) possible just in case p is true in some possible world.

    A proposition p is (metaphysically) contingent just in case p is true in some possible worlds and false in some possible worlds, i.e., just in case p is possible but not necessary.

    A proposition p is (metaphysically) impossible just in case p is true is not possible.

Thesis Six: Any possible creation situation includes every necessary proposition.

B. Providence and Freedom

So the picture is that "before" the free act of the divine will God has knowledge of a sort that makes clear to Him His options in creating. This sort of knowledge includes all those propositions whose truth or falsity is not under His control. Let us call this knowledge prevolitional. Then God chooses one world-plan and wills to execute it. Given His prevolitional knowledge and His knowledge of what He Himself has willed, He has postvolitional knowledge of what will in fact happen. This postvolitional knowledge is called God's free knowledge.

However, this is at least an apparent problem with this picture. From what we have said so far, it seems that God's prevolitional knowledge includes only those propositions which are metaphysically necessary. Now the problem is that this seems insufficient to guarantee comprehensive divine free knowledge of the future and hence to guarantee that Thesis One above is satisfied. Why is this? The problem at least appears to stem from the claim that agents act freely only if they could have done otherwise. Consider the following:

God is contemplating creating Adam and leaving Him free with respect to the action of eating the forbidden fruit in a set of circumstances C. Now God's creation situation, call it CS25, includes all the necessary truths. But the only necessary truths that are relevant here are the following:

    (1) It is possible that if Adam is placed in C, then he will freely obey God's commands.

    (2) It is possible that if Adam is placed in C, then he will freely disobey God's commands.

So suppose that (1) and (2) are all that God has to go on by way of relevant prevolitional knowledge, and suppose that God decides to place Adam in circumstances C. Does He know what Adam will do on that basis? It seems not. Let us say that worlds in which Adam is placed in C and freely eats the fruit, thus disobeying God, are D-worlds, and let us say that worlds in which Adam is placed in C and freely refrains from eating the fruit, thus obeying God, are O-worlds. Then it seems that God cannot know, just on the basis of (1) and (2) and His own decision to place Adam in C, cooperate with his action, etc., whether the resulting world will be an D-world or an O-world. But if God cannot know this, then He cannot be perfectly provident. For suppose that Adam will disobey God by eating the forbidden fruit; then, it seems, God did not know for sure what was going to happen, and so He could not have beforehand permitted Adam's disobedience as part of His eternal plan.

Obviously, prior to deciding whether or not to place Adam in C, God must also know what Adam would freely do if placed in C. That is, in addition to knowing (1) and (2), He must also know which of the following is true:

    (3) If Adam were placed in C, he would freely obey God's commands.

    (4) If Adam were placed in C, he would freely disobey God's commands.

Propositions such as (3) and (4) are called counterfactuals of freedom (I prefer to call them conditional future contingents). Both Molinists and Thomists believe that God knows propositions such as (3) and (4) for every possible free agent and every possible free action. They disagree, however, as to whether or not it is up to God whether such propositions are true. Let us now examine Molinism and Thomism.

C. Molinism

Molinism holds the following:

    (i) the truth or falsity of counterfactuals of freedom is not up to God. This is because if Adam is free, then it is up to Adam alone whether he freely eats the forbidden fruit.

    (ii) Of course, it is true that he could not so much exist or have power or freely exercise this power if God did not act in various ways to supply the circumstances for Adam's action. But the action is Adam's, and so whatever God does must, if the action is truly free, be compatible both with Adam's freely eating and with his freely refraining from eating.

    (iii) By the same token, however, propositions like (3) and (4) are contingent. Suppose that (3) is in fact true and (4) is in fact false; all that follows is that Adam would freely eat the fruit. But it is easy enough to imagine a creation situation in which (4) would be true and (3) false instead. What's more, if (4) is true, then it is not within God's power to actualize a world in which Adam is in C and freely obeys God. So it follows from Molinism that not every possible world is a feasible world.

So according to the Molinist God's prevolitional knowledge includes both God's knowledge of necessary truths (this is called God's natural knowledge) and God's knowledge of the true counterfactuals of freedom (this is called God's middle knowledge, because it comes between God's natural knowledge and God's free knowledge). So we have:

Prevolitional Knowledge
(not under God's control)


Free Act of Will
(under God's control)


Free Knowledge

Natural Knowledge
(neceessary truths)

Middle Knowledge (counterfactuals of freedom)

Knowledge of
own causal contribution

Knowledge of
absolute future contingents

So according to Molinists, then, God can know the future free actions of those creatures of His who are endowed with freedom only if He has from eternity as part of His prevolitional knowledge a comprehensive knowledge of all counterfactuals of freedom.

D. Thomism

The Thomist disagree with this picture for several reasons. The two most important are these:

(i) First, whereas necessary truths do not require an explanation for why they are true, this is not so with contingent truths. Every contingent truth needs some sort of explanation of why it is true rather than not true, and this explanation can only be that someone or something made them true. But according to Molinism counterfactuals of freedom are contingent propositions that were true from eternity even though no one made them true from eternity (God did not make them true because, according to Molinists, this is incompatible with human freedom; and the relevant agents obviously did not make them true from eternity, since they did not exist).

(ii) Second, according to Molinism we--and not God--are responsible for the good that we do (contrary to what St. Paul says). For what God does is wholly compatible with both our doing good and our doing evil. So if we do good instead of evil, it is not God's doing, but wholly ours.

Our alternative picture is this:

Prevolitional Knowledge
(not under God's control)


Free Act of Will
(under God's control)


Free Knowledge

Natural Knowledge (neceessary truths)

Knowledge of
counterfactuals of freedom

Knowledge of own causal contribution

Knowledge of
absolute future contingents

So on this alternative picture, the truth of propositions like (3) and (4) is up to God, since it is up to God whether or not He would give free agents the sort of cooperation (or: concurrence) necessary in order for them to perform given actions freely. That is, God decides how He would cooperate with Adam if Adam were placed in C, and, according to the Thomists, this is sufficient to determine what Adam would freely do in C. So we might think of God's first decision as making a complete complement of conditional future contingents free by deciding to give or withhold efficacious concurrence in every possible decision situation for every possible secondary cause, including free beings. So unlike on the Molinist picture, it is not the case that God's total causal contribution is compatible both with Adam's eating the forbidden fruit and with his not eating the fruit.

E. The Metaphysics of Freedom and Goodness

1. Freedom

How is one to choose between Molinism and Thomism. It seems that there are a number of considerations favoring each of the two sides. Part of the dispute has to do with conflicting philosophical analyses of freedom. Flint contrasts Libertarian with Compatibilist accounts of freedom. A libertarian insists that the following is true:

    (C) Necessarily, for any human agent S, action A, and time t, if S performs A freely at t, then the history of the world prior to t, the laws of nature, and the actions of any other agent (including God) prior to and at t are jointly compatible with S's refraining from performing A freely.

However, many would insist that (C) is too strong, that an action can be free even if it is determined by causes outside the agent, as long as the agent wants to perform the action or as long as the agent wants to perform the action and the agent's want has been caused "in the appropriate way," where certain kinds of deviant psychological, biological, sociological causes are ruled out.

Now Thomists do not endorse (C) as it stands but would endorse it if it were altered just slightly to read

    (C*) Necessarily, for any human agent S, action A, and time t, if S performs A freely at t, then the history of the world prior to t, the laws of nature, and the actions of any other created agent prior to and at t are jointly compatible with S's refraining from performing A freely.

So the main question separating Thomists from Molinists on this front is one about the relationship of divine causation to free human action. Interestingly, here is one point where worries about anthropomorphism and obscurantism come to the fore. The Thomists think that Molinists are anthropomorphists by virtue of the fact that they treat God's causal influence on the human will as of a piece with the influence of created agents; but Molinists think of the Thomists as obscurantists by virtue of the fact that they can give no really clear account of just how God's influence differs from that of created agents.

2. Goodness

The Thomists have a theory according to which all good effects in the world depend upon God's intrinsically efficacious concurrence; created causes would themselves be the source of goodness otherwise. For they would be the reason why intrinsically neutral concurrence should lead to good acts in some cases and evil acts in other cases.

The Molinists, on the other hand, allow that God might influence the will by antecedent internal and external assistance that falls short of determining the will to the good or being incompatible with its going wrong.

There are other issues as well, e.g., the vexed question of the truth-values of subjunctive conditionals, which lurk in the background. It is important to see how complicated and interesting the issues are, and also to see how important disputes can take place within the bounds of strict orthodoxy.