The Argument from Nonbelief : A Rejoinder (2003)
Abstract: In his book, Nonbelief and Evil, Theodore Drange presents an argument against the truth of the existence of God. In this argument, which he calls the Argument from Nonbelief, he claims that the fact that all people do not believe the gospel message before they die provides grounds for denying that the Christian God exists. I will attempt to show that there are good reasons to deny that this inference goes through. I will argue that given the nature of free persons, it is not within the set of logically possible states of affairs that God is able to actualize. Further, I believe Drange has an inadequate understanding of religious belief that should be rejected and replaced with a more robust formulation. My treatment will be a rejoinder only–I will not attempt to argue positively for the truth of the existence of the God of evangelicalism.
In Nonbelief and Evil, Theodore Drange sets out the following argument that he calls The Argument from Nonbelief (ANB):
The gospel message = the following three propositions:
(a) There exists a being who rules the entire universe
(b) That being has a son
(c) The ruler of the universe sent his son to be the savior of humanity
Situation S = the situation (or situation-type) of all, or almost all, humans since the time of Jesus of Nazareth coming to believe all the propositions of the gospel message by the time of their physical death.
(A) if the God of evangelical Christianity were to exist, then he would possess all of the following four properties (among others):
- being able to bring about situation S, all things considered
- wanting to bring about situation S, i.e., having it among his desires;
- not wanting anything else that necessarily conflicts with his desire to bring about situation S, as strongly as he wants to bring about situation S;
- being rational (which implies always acting in accord with his highest purposes).
(B) If there were to exist a being that has all four properties listed above, then situation S would have to obtain.
(C) But situation S does not obtain. It is not the case that all or almost all, humans since the time of Jesus of Nazareth have come to believe all the propositions of the gospel message by the time of their physical death.
(D) Therefore [from (B) & (C)], there does not exist a being who has all four properties listed in premise (A).
(E) Hence [from (A) & (D)], the God of evangelical Christianity does not exist.
The argument is constructed as a deductive argument concluding that the God of evangelical Christianity does not exist. I will attempt to demonstrate that there are good reasons to believe that the argument does not go through. I believe Drange has an inadequate understanding of religious belief and makes some fatal errors in his understanding of the nature of God at least insofar as he is attempting to reflect the Evangelical Christian views about God.
Drange’s argument is constructed in such a way that given (A), God actualizes situation S by logical necessity. That is, Drange includes in his conception of God the property such that God must actualize S if he is to exist. I will attempt to argue that (A) is faulty in two respects. First, (A4) assumes an inadequate view of God’s freedom, namely that God is determined in some way by his belief-desire states. I will show that this view is wholly inadequate and misguided. Second, I will argue that (A1) is not a property that God possesses. In short, I will argue that though God may be desires that S obtain, he is unable to bring S about as S is written. Specifically, I will try to show that “actualizing situation S” is not among the set of logically possible acts within God’s ken. Since (A) is faulty, ANB does not go through as Professor Drange lays it out. First I will attempt to surface what Drange is claiming in (A) then I will examine the premise in detail.
The Nature of God
Before I begin my analysis, I want to offer a reformulation of ANB that I believe will help surface important aspects of ANB and make my critique of the argument more efficient. Drange’s formulation seems to me to be cumbersome and bit confusing. For example, premise (A) seeks to establish some fundamental properties that must God possess but it does so in reference to S. A clearer way to construct the argument would be to assert properties God must possess without reference to S then argue that those properties are necessary for the actualization of S.
Premise (A1) simply asserts that God has a given ability to do something, namely bring about S. Premise (A2) and (A3) attempt to establish that God has what Drange calls a “nonoverridden” desire to bring about S. There is some ambiguity about whether this means that the object of God’s desire is the bringing about of S or S itself. Is God’s desire for some action on His part or that some state of affairs (S) obtain regardless of how it’s brought about? Drange argues that the latter entails the former and so there is no distinction between the two. That is, given that God has “bringing about S” as His greatest desire, “then there is little difference between God wanting situation [S], and God wanting to bring about situation [S] himself.” If I understand Drange’s claim here, he appears to be saying that given that God has the complete ability to act on his greatest desire, anything He desires in a nonoverridden kind of way entails a desire to bring that desire about. The entailment comes by way of (A2) and (A3). Add to this premise (A4) and God not only desires to bring about S but also necessarily acts to bring about S. In a word, (A) states that God is a being that always acts on his highest desire (in this case, S) for which he has complete ability to bring about.
If this is a correct way of understanding (A), ANB could be constructed more clearly by restating (A) in terms of unqualified properties God must possess. Subsequent premises then could make claims about S in light of those properties. Given S, ANB could be constructed as a reductio ad absurdum argument as follows (ANB-R):
(A′) The God of Evangelical Christianity exists (assumption)
(B′) The God of Evangelical Christianity is a being such that for any given desire He possesses, necessarily God acts on that desire iff (1) it is predominant and (2) he has the ability to bring it about
(C′) God has the ability to bring about situation S
(D′) God predominantly desires to bring about situation S
(E′) Necessarily S must be actualized (entailment from B′- D′)
(F) Situation S is not actualized (fact)
(G) The God of Evangelical Christianity does not exist (not-A′) by reductio
I believe this formulation accurately captures ANB but lays out the premises more clearly. First, it does not explicate certain properties God possesses in terms of situation S. This articulation separates the properties of God from an explicit instance in which those properties might be exemplified. Second, formulating the argument in this way surfaces the modality of the argument in a clearer way. For example, the modality of (E′) is clearly that of necessity. Though I think Drange’s original argument demands such a modality, his original formulation conceals that fact. Therefore, in addressing ANB I will work off of this formulation since if I were to critique the original ANB I would restate each premise individually in these terms anyway.
On the face of it, the premises of ANB seem to be ones with which Evangelicals can agree and Drange spends many pages in his book arguing that, in fact, Evangelicals can and should agree with them. Yet most Evangelicals would want to patently deny the conclusion. So where have things gone wrong? There must be a problem with one of the premises despite their prima facie strength and their ultima facie support that Drange culls together. But which premise (or premises) is faulty? I will argue that the core problem with Drange’s argument lies in premise (C′), namely that it is false.
Does God possess compatibalist freedom?
First, a point of contention over the modality the argument possesses given the way Drange lays it out. Under ANB-R Drange’s view of God’s freedom immediately surfaces. Premise (B′) claims that God necessarily acts to bring about those states of affairs for which his desires are strongest and for which he has the ability to bring about. According to Drange, if God has a predominant desire d and he has the ability to bring about a state of affairs a that would fulfill d, necessarily God brings about a. This, according to Drange, is the definition of rationality. Accordingly, God’s action is the necessary result of His strongest desire along with the ability to act to bring the object of that desire about. Drange claims that for God to do otherwise would make him irrational.
On the most generous reading, Drange is committed to a compatibalist view of freedom with respect to the will of God. A compatibalist view of God’s freedom seems to entail that God’s will is determined in some sense by his nature and desires. An argument for a compatibalist view of God’s freedom might look like this (call it GC):
P1 God has a predominant desire for x
P2 God has the ability to bring about x
C1 Therefore, God brings about x
Given the entailment, it is not possible for God not to bring about x given P1 and P2. What are we to think of this view of God’s freedom? ANB certainly seems to require it but is it the proper way to view God’s ability to act? Many Evangelicals hold that man is determined at some level but few (perhaps none) hold that God is determined in this way. Few classical renditions of Christianity would endorse the compatibalist notion of God’s freedom that Drange is promoting. According to ANB, God appears to be an automaton who must act according to his strongest desires. However, according to classical theism, God is libertarianly free with respect to his will. That means that God is not caused to act by anything–not even his own beliefs and desires. If God is determined in the way Drange seems to imply that he is, we need an argument to demonstrate this. God is the archetype of libertarian freedom if anyone is!
Still, it does seem right to hold that even if God is libertarianly free with respect to his desires, Drange’s rationality constraint still holds. It seems right to say that God will want freely to actualize his greatest desires (assuming that all of his desires are maximally good). In other words, God can be libertarianly free yet will always freely will to actualize his highest desires. At the end of the day, the result is the same whether God has compatibalist or libertarian freedom: God actualizes his greatest desires over which he possesses the power to actualize. So why address the issue?
The relevance of this qualification primarily has to do with the modality of the argument. If God possesses compatibalist freedom with regard to actualizing his greatest desires then God’s desires function as efficient causes of His actions over which God ultimately has no control. This is especially true if God’s desires are a direct result of His nature. This view doesn’t allow God to willfully constrain his actions in light of His greatest desires. It seems that it’s not even possible for God to exercise willful control over his actions given certain desires and power that he possesses.
If, however, God possesses libertarian freedom where reasons function as final causes rather than sufficient causes, God acts not because his desires cause him to act but because he wants to accomplish some end. As such, God exercises his will according to his desires not because of them. Of course it is true that God always performs perfectly good acts and so his desires too must also be perfectly good. The broader point however is that because God’s desires function as final causes and not sufficient causes, God is not necessitated by his desires and thus might be motivated to exercise his will for other reasons.
More importantly for the present purpose, if God possesses true libertarian freedom the modality of ANB cannot be that of necessity. That is, though it is highly probable that God will act on His greatest desire, the fact that he possesses true active power over his will, it is still possible in the broadly logical sense that God will not act on his greatest desire and so he is not necessitated to act on his greatest desire. There may be other reasons, epistemic or moral, why God refrains from acting. I think this satisfies Drange’s rationality constraint without requiring the modally necessary form of the argument.
If one rejects the deterministic understanding of God’s will, the modality of (B) changes from necessity to possibility. The modality of the argument then changes as well and the conclusion does not necessarily follow. The best Drange can claim at this point is that possibly God is a being of whom compatibalist freedom is true and such a claim would need strong arguments and a thorough evaluation of the Christian scriptures to support it. However the modality of the argument hinges on (B′) so if (B′) is only possible, so is (G). Thus the best Drange can conclude then, is that possibly the God of Evangelical Christianity doesn’t exist.
It is at this point that the question of the intended modality of the argument itself becomes important. If the argument is not meant to prove the nonexistence of God (as Drange seems to indicate in places ), perhaps Drange would be happy to restate the argument in terms of probability instead of necessity. However, he most likely wants to argue that the conclusion should not be understood to be merely possible but the stronger “probably true” where the probability quotient is quite high that the conclusion is true. This can be done, he might argue, by claiming that although (B′) isn’t true of God, (B′) could be restated in terms of probability that would make it true. For example, Drange could argue that though God doesn’t necessarily act according to his strongest desire and for which his ability is efficacious, he probably does. He might claim that certainly it is reasonable to believe that God as a perfect being has pure desires and so most likely acts on those desires since there would be no moral reason he shouldn’t (especially if a desire is predominant for God and he has the complete ability to bring it about). He might offer (B″) in place of (B′).
(B″) The God of Evangelical Christianity is a being such that for any given desire, God probably acts on that desire if (1) it is predominant and (2) he has the ability to bring it about.
He then would have to modify (E′) to assert
(E″) Probably, S is actualized (from B″- D′)
and conclude with (G′)
(G′) The God of Evangelical Christianity probably does not exist (probably not-A′)
I think this is a reasonable way to state the argument and employs a notion of God’s freedom that more Evangelical theists could accept. Of course the conclusion is not nearly as strong as the original ANB but given what Drange has to work with, the revised conclusion is the best he is able to do. If one were to grant Drange (B″), the onus of the argument would then rest on premises (C′) and (D′). What are we to make of these premises? Are they claims consistent with Evangelical Christianity?
ANB: A Second Look
Belief and the Will
In examining ANB-R, I will start with comments on premise (D′) and examine (C′) in light of those comments. Premise (D′) states that God predominantly desires to bring about S. What are we to make of this premise? First it will be important to drill down on exactly what God is said to want to actualize. I believe there is a serious problem with Drange’s conception of S (at least as far as evangelicalism is concerned) primarily because he equivocates on the scope of the meaning of belief. Drange is quite clear in places that he intends the term “belief” to refer only to a propositional attitude towards the gospel and not to a stronger notion of commitment to the truths (or the person) propositions are about (I will adopt Audi’s taxonomy here and call the first type of belief propositional belief and the second type attitudinal belief).
He makes a lot of hay in chapter 5 arguing that God’s actualizing S would in no way interfere with human free will because belief formation is not voluntaristic. Drange argues that God clearly could actualize S without needing to do anything regarding subsequent action on or commitment to those beliefs with respect to the believer. In actualizing S, God is concerned only with propositional type belief not attitudinal type. But certainly no evangelical would agree that this constitutes God’s highest desire. Evangelicalism emphasizes God’s desire that people accept the gospel and commit to it. Certainly the desire to actualize a world where all persons merely believe the gospel does not constitute the God’s highest desire needed to support (D′).
It seems to me that Drange equivocates on the meaning of belief and the source of the equivocation (or perhaps it’s simply lack of clarity) centers around the relation of belief formation to acts of will. In Appendix C, Drange discusses the notion of belief and the will. Though he is unclear about this, Drange appears to hold that some form of direct voluntarism is possible in certain cases but that the case of God actualizing S would not be an interference of an individual’s free will. He appears to hold this mainly because he does not think that belief formation generally comes voluntaristically. At the very least, Drange appears to be an indirect doxastic voluntarist at most he is a strict determinist. God, then, could cause belief in a person entirely retaining the freedom of that person. I agree with Drange that direct doxastic voluntarism probably is not true but I hold that indirect voluntarism is probably the way many of our beliefs are formed.
I think it would be uncontroversial to say that beliefs are mental states that are formed in the mind of an individual by way of some phenomenological event (call it e). By “phenomenological,” I mean to include not only sensory perceptions like sight and sound but internal perceptions such as reflections on ones own mental states, memory and the like. I want to add that though beliefs contain a propositional component, they cannot be fully represented linguistically by sentences. This is because beliefs are inherently phenomenological. They cannot fully be captured in a sentence that can represent more than one proposition. The propositional aspect of a belief is not identical with any other proposition but an unqualified single sentence can serve as a linguistic representative of many propositions. So beliefs, on my view, are inherently phenomenological mental events or propositional attitude if you will and cannot be fully represented linguistically.
There seems to be another element crucial to belief: acceptance. It is not enough to say that one believes something just in case one has a mental state of a certain type. Not only must one have the mental state formed in them, but they must also accept that the mental state is a true representation of the world. For example, I can have the mental state that Santa Claus lives at the North Pole but not believe it because I don’t accept that “fact” as true about the real world. Drange has a mental state the content of which is, in some sense, the Christian God but he doesn’t believe that this mental state is of anything in the world. Given the above, I offer the following definition of belief: S believes P just in case S has a mental state that is formed by way of some perceptual event, that mental state is propositional but cannot fully be represented by a sentence, and S accepts her mental state as true or as the way the world actually is. (Another way to put the latter part of this definition is to say that S has a disposition to act on her mental states. )
Of course, one might qualify this by noting that the event of belief formation may best be captured by the phrase “one finds oneself believing.” This includes not only the formation of the mental state but the acceptance of that mental state is true of the world. For example, as I reflect upon my mental states right now, I find myself believing that there is a computer monitor in front of me and a purple-covered book next to it. This mental state was formed in me (I did nothing actively to form that mental state) but it also seems that the acceptance that the mental state is true of the world seems to be outside of my ability to reject or deny. This is what involuntarism means–I involuntarily accept the mental state as true about the world. In fact, the truthfulness of the mental state seems to be a property of that state and can’t be separated or “assigned” to it. In this sense then, acceptance would not be the same as a choice.
Still, not all beliefs are formed in this direct way. Some beliefs are formed inductively or deductively as a result of an examination of evidence. The belief is still formed involuntarily where the mental (or propositional) content of the belief and its acceptance are formed in me apart from an act of my will. But the act of belief formation is the result of me voluntarily subjecting myself to evidence that causes the belief to be formed in me. On this notion, evidence plays a crucial role in whether or not the belief is accepted as true.
If I look out my window, I don’t form any direct belief about the specific age of the earth (other than that it is more than a few minutes old). I don’t form the direct belief that the earth is x number of years old. However, I can choose to do research that will subject me to evidence that may cause a belief to be formed in me regarding the age of the earth. This scenario captures what is typically meant by indirect volunteerism.
There is one more aspect of belief that I need to mention. Though belief may be formed (either directly or through an inference) and accepted involuntarily, there is a voluntary aspect to one’s “doxastic life” that cannot be ignored. This aspect can affect belief formation, the acceptance aspect of belief, or both. Regarding the first of these, there is a real sense in which most of us choose to accept certain sorts of evidence over other sorts or to act according to one set of truth claims as opposed to another. Most people contain within their noetic structure a set of false beliefs. Some of them may be trivial like believing that one is better looking than one actually is or more important like belief or disbelief in God. False beliefs are not formed directly or based on evidence at least not in the strict sense since, ostensibly, these beliefs do not satisfy the truth condition. So how are they formed? One answer is that they are formed by some act of the will on the part of the believer.
Drange might say for example that religious believers choose to exercise faith in something that doesn’t exist. They aren’t compelled to belief by way of evidence or a direct experience of God (since God doesn’t exist). Rather, they exercise their will and cause themselves to form belief in God. This may be called the “psychologization of belief” where the person forms belief in God as an heuristic device rather than naturally as a response to the real world.
Faith as a type of belief
However even if a given belief is formed in the right way, the dispositional component of that belief varies in strength for each person and if the belief is to have any substantive impact on the person’s life, it seems that the person must choose to act on the belief or commit oneself to the truth that grounds the belief. In other words, it is not sufficient simply to accept that some proposition is true about the world. This is particularly true in certain beliefs involving relationships with others and, more the present point, what I will call mere belief is not sufficient for religious belief. By religious belief, I’m referring specifically to the kind of belief Drange is talking about in S: belief required for salvation in the Evangelical sense. A key notion in evangelical soteriology is the notion of what Jonathan Edwards calls “the affections.” This type of belief involves a change of the heart and is displayed in action.
Consider a person (call her Addie), who is preparing to go to college. Addie has developed the belief over the years that her parents love her. Additionally, she has come to have the belief that because of that love, her parents have set aside a large amount of money that will allow her to attend college without any financial need. Let’s stipulate that the gift cannot be applied to Addie’s college expenses without Addie’s approval. Even though Addie is aware of this gift, she has chosen to reject her parent’s love and all the benefits that love provides and ends up going into huge debt and having to work part time while in college. Both choices end up affecting her grades in a negative way. Ultimately, she ends up doing poorly in many of her classes and is unable to graduate. It should be fairly clear to see, in this case, that Addie has the relevant beliefs that would enable her to have a successful college career: her parents, out of their love, have provided for her financially. However, Addie has chosen not to commit to that love and act on the belief she possesses. In essence, the belief does her absolutely no good. For her parents to make the gift efficacious, they would not only have to give her the gift, but ensure Addie would use it by modifying her will in such a way that she entered into a relationship with her family such that the gift would be used.
For salvific purposes, God wants the will to be turned such that the believer not merely believes the truths of the gospel but willingly submits to them in an act of love. So if God merely actualized S, it would be wholly ineffective for salvific ends and thus would be pointless. Surely, belief that the propositions of the gospel message are true is necessary for salvation, but few if any evangelicals would agree that it is sufficient. Thus if S is to have any relevance at all, it has to include the idea of what the Bible calls faith and not merely belief (as the book of James says, even the demons believe that God exists but certainly that believe doesn’t do anything for them with regards to salvation). So the key feature in actualizing S for person P would have to involve changing the will of P and not simply altering her doxastic structure.
Does ANB require mere belief or faith?
Drange appears to agree. Though his treatment of the Free-will Defense in chapter 5 seems to ignore this component, he clearly sees the need for it in chapter 2 in his comments on ANB. In support of his premise (A2) and (A3) he offers 7 arguments to show that God predominantly desires to actualize S. In his conception of belief here, he clearly brings in the notion of commitment, trust, and love. All these concepts strongly involve the will. I will briefly survey his arguments in order to surface this idea.
In argument (1) he cites 1 John 3:23 as evidence that God predominantly desires to actualize S. This text is one of the clearest in the New Testament about salvation and so clearly involves commitment not merely the will. In argument (2) he cites Biblical passages that command that people love God maximally. This too clearly involves a commitment to God not merely a belief in him. Arguments (3) and (4) refer to missionary activity on the part of Christ and the apostles and to the great commission. The mission of Christ and his followers was to “bring salvation to all of the world” and not merely true belief. If Drange wants to cite missionary activity as evidence of God’s desire to actualize S, the belief in S has to refer to a commitment to the gospel and not merely belief in it. Argument (5) references 1 Timothy 2:4 which states that God “wants all men to be saved and to come to a knowledge of the truth.” Again the clear reference here is that God’s desire (predominant or not) is for people’s salvation not merely that they possess certain noetic content.
In the 6th argument, Drange clearly states that God’s desire in (A2) is both for people’s salvation and for their acceptance of Christ as savior. He writes, “One of [the premises of (A2)] is the claim that, according to the Bible, God wants all humans to be saved…. But in order for people to be saved they must believe in God’s son, which is usually taken to include accepting him as the savior of the world. Hence, God must want people to believe in his son in that sense (accepting him as savior of the world), which entails believing all the propositions of the gospel message. It follows that (A2) must be true.” God’s maximal desire is for people’s salvation not merely that they belief the gospel with no commitment. Argument (7) adds to this by claiming that because God wants people to love him, he must also want them to believe the gospel. Love involves a commitment and self-sacrifice. Certainly this is stronger than mere belief.
Drange’s purpose appears to be to argue against God’s existence based on the fact that not all people are Evangelical Christians by the time they die. Surely Drange would have to agree that being an Evangelical Christian goes far beyond merely believing the propositions of the Evangelical gospel and involves a commitment to them (to the God to whom those propositions refer). For if the propositions of the Evangelical gospel had no implication for this or otherworldly commitment (they are just propositions that are added or not added to one’s noetic structure) then believing them or not believing them is inconsequential. It would be similar to the doxastic posture most take towards belief in the existence of UFOs: belief or disbelief really is not relevant for this world or the next. For Drange’s argument to qualify as a form of the argument from evil, he has to understand belief in the gospel in a stronger sense than merely value-neutral content of one’s noetic structure.
Though his purposes are different than mine, Robert McKim in his excellent Religious Ambiguity and Religious Diversity states a similar point. “For us to have a personal relationship with God, we would need to be aware in a constant and forceful way of God’s relating personally to us; we would need to have a constant and palpable sense of God’s presence. We would need to be intimately aware God in a permanent and ongoing way. This would be very invasive. And the more invasive it would be, the easier it is to make a case against its being desirable that we should currently be in such a state.” Where I disagree with McKim is in that it is not the having of a personal relationship with God (and the having of the states McKim’s describes as important towards accomplishing that end) that is invasive. It is only invasive if God forces us to have such a relationship. As many theists have argued, being free to “choose God” is a virtue.
Still, McKim’s point is well taken: if God forced a relationship onto humans, that relationship would not be a matter of choice for us. So, my argument goes, in order for us (as the type of creatures we are) to have a truly free relationship with God, our belief, our faith, in Him must be libertarianly free at some point in the process of belief formation.
So if we are to take Drange’s notion of (religious) belief as presented in chapter 5 as propositional type belief, then clearly (D′) is false. For no evangelical would grant that such a notion of belief is God’s predominant desire (or even minimal desire). Further, Drange provides no argument for accepting (D′)-cum-propositional-type belief since his arguments for (D′) in chapter 2 strongly involve attitudinal-type belief. Which type of belief is Drange talking about in situation S? I’m inclined to think Drange would rather support (D′) rather than his argument against the Free-will defense. For all intents and purposes then, the notion of belief presented in chapter 5 (and appendix C) does not address the notion of belief needed to support ANB. For the purposes of my treatment then, I will understand Drange’s notion of “belief” in situation S to essentially involve an act of willful commitment to the propositions (attitudinal-type belief) as described in his defense in chapter 2 and not as merely a disposition to act as if they were true. Given this, is God able to actualize S?
Persons, Properties, and Plantinga
I believe a plausible argument can be made that God is constrained by the nature of free beings to create those beings with a certain noetic structure such that imposing unilateral beliefs of type attitudinal on that structure would not be possible. An argument of this sort would consist in trying to show that God could not create a human creature that was not libertarianly free with respect to attitudinal-type beliefs. He could create a nonfree creature but it would not be human. The argument would consist in showing that a necessary condition for being human is that the creature has the imago dei, the image of God as a part of its nature. An essentially necessary property of the imago dei is libertarian freedom. Such an argument, if successful, would preserve God’s omnipotence (thus satisfying Drange’s premise (A1)) by removing the actualization of situation S from the set of logically possible acts God could perform.
Perhaps the most famous and successful rendition of this type of argument is Alvin Plantinga’s version. Plantinga argues for libertarian freedom for the purpose of demonstrating that God could not actualize a possible world where a person is truly free yet does not do at least one wrong act. His argument is subtle and important. However I want to surface what I believe to be the main feature of the argument (a feature many seem to miss). The argument is not that God simply prefers, due to some ungrounded value judgment he places on freedom, to actualize a world which contains free creatures that may do wrong. Rather, it is that God is unable to actualize such a world. Plantinga explicates this in terms of creaturely essence and possible world semantics.
Plantinga considers a person, Socrates, who possesses properties that are not essential to him, that is, properties that are not necessarily true of him in every possible world. Among these would be his height (he could have been taller or shorter than he is), his weight (he could be heavier or lighter than he is) and his eye color (he could have had a different eye color than he actually has). However there are some properties Socrates possesses that he must possess in every possible world in order to remain Socrates. Many properties are “trivially essential” like being self-identical and being unmarried if a bachelor. Other essential properties would include Socrates’ humanness. For instance if in world W1 Socrates was made of ice, such an entity would no longer be essentially the same creature as the one in W that is human. Even those who deny the existence of natures or are not realists regarding properties would agree that there is a point where a human person, upon the removal of enough properties (regardless of what they turn out to be), ceases to be a human person.
Freedom and the imago dei
Evangelicals believe that human persons are created by God and that part of what makes them human is that they contain aspects (properties) that are also true of the divine essence known as the imago dei (the image of God). If the imago dei is removed, the being would cease to be human. If a necessary aspect of the imago dei is libertarian freedom and a necessary condition for creating a human person is that he or she is created with the imago dei, then it should be clear that God could not have created a human, nonfree person. God may be able to create other kinds of persons that are not free (some argue angels are just such persons). But it would not be possible for God to create nonfree, human persons. An argument for such a conclusion might look like this:
(H) God has created human persons
(I) Necessarily human persons possess certain properties of the divine nature (they are created in the divine image)
(J) It is not possible for God to create human persons that do not possess properties they possess necessarily.
(K) Therefore, it is not possible for God to create a human person that does possess features of the divine nature. (from (H) and (I))
This argument (call it HF) as it stands just argues that humans possess certain essential properties that are also true of the divine nature. Certainly this argument falls well within orthodox Christian doctrine and should be fairly uncontroversial. However as it stands, it is insufficient as an argument whose conclusion is not-S. Other premises need to be added to include libertarian freedom as an essential part of the imago dei possesses by humans.
(L) Necessarily human persons possess libertarian freedom as a result of being created in the divine image
(M) Therefore, it is not possible for God to create a human person that does not possess libertarian freedom (from (K) and (L))
(N) For God to be able to actualize situation S, he must be able to create human persons that do not possess libertarian freedom
(O) Therefore it is not possible for God to actualize S (from (K) and (N))
This argument is a possible way God could be omnipotent in the way Drange (and orthodox Christianity) requires and still not be able to actualize S. It is important to understand the ontology of freedom leveraged in HF as an essential property of human creatures. HF is not simply saying that God must create creatures free and then, after the moment of creation, that freedom need not be an essential part of the ontology of the person. As such, HF describes an essential aspect of what it means to be human and that includes being libertarianly free with respect to one’s choices. However there’s another worry related to the scope of the freedom one has over one’s will.
Does one have to be actually free with respect to one’s will or only possibly free? If freedom is a capacity of some sort, couldn’t it function as a second-order capacity rather than a first-order capacity? If so, one could possess libertarian freedom but not actually have any control over one’s will. If we claim that one has to have a first order capacity over one’s will in order to be considered libertarianly free, a host of further questions arise. For example, over how much of one’s will does one have to have a first-order capacity for freedom? Certainly everyone lacks control over some area of their lives. How does one go about defining how much freedom one can lose and still remain “a free creature” and thus a human creature. Another problem is what to do with infants or the mentally ill or those with severe brain damage who lack or lose this first-order capacity?
If arguments like this succeed, then actualizing S is not logically possible in the same way that it is not logically possible that God create square circles. For God to actualize S he would have to create a human person that does not possess the properties necessarily true of human persons. Thus, so long as omnipotence is defined broadly as the ability to do whatever is logically possible, God remains omnipotent with regards to actualizing S since S is outside of the logical possibility of what God (or anyone) could actualize.
It is important to reiterate here that (K) is plausible based upon the definitions of belief and faith I gave above. My claim is that the type of belief God would need to give to persons in order to actualize S would have to be of the attitudinal-type not of the propositional-type and in so doing, God would have to essentially modify the way humans come to form beliefs. Namely, he would have to create persons as nonfree creatures.
On Missing the Point
I anticipate at this point that Drange might respond by saying, “All this talk about God creating free creatures is fine but that’s not the point. My argument isn’t that actualizing S is a function of God’s ability to create free creatures but of causing the relevant beliefs in those creatures regardless of whether they’re free or not.” The objection then is that even if humans have been created with libertarian freedom, could not God cause the relevant beliefs to be formed in them for salvific purposes and still retain their status as “free” in the broad sense? Surely this is broadly logically impossible.
I want to say three things in response to this question. First, it seems fairly uncontroversial to say that God could violate human free will under various circumstances where the person remains a free creature in the broad sense but is not free for certain choices. Setting aside the theological question, most humans do in fact lack control to varying degrees over certain moral and physical areas of their lives. We might say a person that is a compulsive liar or a compulsive thief lacks control over their will in the area of truth-telling and property rights. In a sense, they are determined by the compulsion to lie or steal and have abdicated their ability to control their tongue or their hands. It seems possible that God similarly could direct the will of a libertarianly-free human person.
Yet because they lack control in this one area, does that mean they are no longer free? Certainly not. Being free is a function of a property or capacity a human being possesses not a function of the number of free choices one makes. In the current case then, it certainly seems broadly logically possible to hold that God could cause a person to believe in Christ while maintaining that person’s status as a free individual. In such a case, it would seem possible for God both to cause certain beliefs to be formed in an individual and avoid the objection I raised above.
Given that then, though it seems prima facie possible for God to actualize this state of affairs for some, it probably would only be possible for a very few. It doesn’t seem possible that God could do this for many humans and especially not for every human (which is required by the actualization of S) and maintain humanity’s status as creatures that possess libertarian freedom essentially. The reason for this is that the type of belief that is needed for the full acceptance of the gospel message (what I’m calling attitudinal-type belief) is not an isolated mental disposition but a whole host of dispositions that affect one’s entire noetic structure. It wouldn’t be enough to cause a person (call him Bob) to have the given belief and relational attitude without building in the beliefs and dispositions needed to maintain that those things in a coherent worldview. Consider the following scenario:
Suppose that God wanted to actualize attitudinal-type belief in Bob and preserve Bob’s status as a truly free person. We might say that at bare minimum God would have to cause Bob to come to have the following beliefs and dispositions (among others):
- Bob believes that God exists
- Bob believes that the Bible is true
- Bob believes that the Biblical account of Christ’s work in the world and the nature of who he is are true
- Bob loves God and loves Christ
- Bob loves other Christians
- Bob is an active part of the Christian church
So in this scenario, God would maintain Bob’s freedom (and ostensibly the freedom of everyone else in Bob’s life) in every other part of his life but cause him to have the beliefs and dispositions in the list above. But is this possible? It doesn’t seem so. For God to cause Bob to have the types of dispositions and beliefs noted in this list has substantial implications for all of Bob’s character and so would have to impact the way Bob views the world at a very fundamental level. For Bob to love Christ according to standard Christian doctrine, Bob would have to admit that he is not sovereign over his life, that he should serve the creator of the universe and give his life to that service, that God has asked certain things of His creatures and that God wants his creatures to live a certain way, that persons have meaning in the broader context of their being created by God and so his relationships with others would become more meaningful, and so on.
For these things to be true (and for Bob to remain rational where he has a coherent worldview), Bob has to have another set of dispositions, beliefs, desires, and attitudes that contribute to making these coherent and sustainable. Robert McKim describes this well when he writes, “If we need to have certain attitudes or a certain character in order to recognize certain facts that are easily missed, or relevant beliefs are in part a product of something about us. In particular, in the case of theism, this would make belief in God, and failure to believe in God, an expression of our judgment and character (etc.). This would be consistent with, and may illuminate, the claim sometimes made by theists, that faith is a response of the whole person; and the same would hold for not having faith.” If this is true then it would seem that God would not have to cause only the beliefs and dispositions noted above but a host of others. A fortiori He’d essentially have to affect a large part of Bob’s character that normally would be under Bob’s direct or indirect responsibility to change. It would seem hard to continue to view Bob as a free creature.
So if God were to cause the relevant beliefs and dispositions without directly changing them, perhaps God could cause them to change indirectly. But here the problem grows exponentially. In order for God to cause belief in this way, it seems he would have to violate the freedom of numerous other people to indirectly cause a given person to come to believe in the necessary propositions. For example, it is commonly understood that our dispositions and attitudes relevant to belief formation may be heavily influenced by others. Experts commonly cite parental influence as a major factor in influencing the direction of our character. If God were to indirectly cause Bob to have the relevant beliefs and dispositions indirectly he might do so through Bob’s parents. But how would this work? Wouldn’t God have to directly control the wills of Bob’s parents such that they would have to have a complete set of beliefs dispositions and attitudes much like God would have had to cause in Bob so that they would properly influence Bob in the right way? And wouldn’t their parents have to be so caused? There seems no end to it.
McKim makes this point in an even stronger way. He states that it seems likely that minimal theism (what I’m calling bare theism) isn’t even possible. He gives some compelling reasons to think that anyone who might possess bare theism (which is not under one’s direct control) would be strongly inclined towards full theism (over which one at least ought to exercise control). McKim concludes, “Minimal theism, in short, exists only in theory, and is to all intents and purposes, out of the question; a belief would not count as belief that God exists unless it involved full belief and trust.”
It seems that in order to preserve our status as free creatures it would not be possible to actualize S. My argument then is that God is constrained by his nature both to create us humans as free and in order to maintain that freedom is unable to actualize S. ANB fails because C′ is false.
Two More Possible Responses
ANB and “bare” theism
Suppose at this point Drange admits that under the notion of belief offered above, there is no logical inconsistency with understanding how an omnipotent God would be unable to instantiate S. But also suppose that he denies that bare theism is impossible. He might want to reject the notion of faith as described above and go for the lot less demanding notion of bare theism. If Drange wanted to preserve the idea of faith he might want to argue that though mere belief isn’t sufficient for S to obtain, it is at least necessary. As such there still is a formulation of ANB that skirts the free will objection above.
To argue in this way Drange would have to present a new version of ANB that cashes situation S in terms of a bare-bones theism instead of in terms of Christianity specifically. In order to do this, he would he would have to sufficiently modify S so that it is not the truths of the Evangelical gospel that God needs to cause people to believe but simply his existence. S would then read:
Situation S′ – the situation (or situation-type) of all, or almost all, humans since the time of Jesus of Nazareth coming to believe the proposition that God exists.
Of course, the notion of belief here does not mean that humans actually submit to God or are willing to do his will. Certainly the notion of belief here could include a whole range of actions towards God including hating God, rejecting God, and hating and rejecting anything to do with God. The substantial aspect of S′ is that humans believe God exists in the sense of having a mental state plus acceptance in the sense described above.
In fact, Drange addresses an argument of this type in chapter 14 of his book. His modified ANB applied to what he calls theism in general is constructed in the same way as the original ANB except situation S is modified in a way similar to my situation S′. Drange’s version, however, describes God as “a being who rules the entire universe.” For Drange, this conception of God rules out pantheistic or nonpersonal religions. Essentially Drange’s argument in chapter 14 consists in claiming that is that it is fairly obvious that much of the world’s population doesn’t belief in the relevant sort of God. Drange then spends most of the ink in chapter 14 applying ANB to that claim. In the end, ANB for theism in general turns out to function in much the same way as the original ANB for evangelical Christianity: the fact that many people don’t possess belief in God provides good grounds to deny his existence. What are we to make of this argument? Is ANB stronger when applied to theism in general?
I tend towards a different response to this argument than the one I gave against the original ANB. Whereas against the original ANB I argued that situation S isn’t actualized (agreeing with Drange) but showing why we shouldn’t expect it to be (arguing against premise (A2)), against the modified ANB I’m more inclined to argue that situation S′ is actualized. My inclination is to argue that God has given all people some sort of knowledge or, at the very least, a capacity to know him. Defending this claim is beyond the boundaries of this paper and will have to be left for another time. However, I can make a few brief comments.
First, there is a whole Christian tradition that has argued for the claim that all men possess knowledge, or a capacity, or a latent desire to know God. As mentioned above, many philosophers, primarily in the Reformed tradition, have argued and do argue that all people do have an idea of God–that, in fact, S′ does obtain. Drange seems to have totally ignored this tradition. In fact, Drange dialogs very little with philosophers that have written in this area. Instead he chose to interact with a couple of brief statements by Daniel Howard-Snyder and Robert McKim. The statements he examines are not even related directly to this issue.
Second, in disagreeing with Drange on the whether or not S′ has been actualized, the argument becomes an evidential one not a logical one. That is, the argument stands or falls on the strength of the evidence that accrues for or against it. The philosophers alluded to above argue that most people do have some sort of belief in God and provide philosophical and evidential reasons for believing this. Some people cultivate this belief, others suppress it but, according to these philosophers, belief in God does exist. There does seem to be empirical data to support this as well.
Finally, even if one rejects the claim that everyone possesses some sort of knowledge of God, ANB as applied to theism in general has another difficulty. One might be able to produce a fairly substantial argument that having merely propositional knowledge of God could end up producing the opposite goal God has for each person and so God does not give person’s mere belief in him for that reason. God’s desire is to enter into a mutual loving relationship with each person. If all individuals had merely propositional belief without the opportunity to accept or reject the relationship, that belief may become cold, repressed, and comfortable. Put another way, the belief would end up being as familiar as the belief, say, that one can walk–a rather complex task physiologically speaking and, at one time in our development, a difficult one–and the belief would lose its existential importance. As Screwtape wrote to Wormwood, “Indeed the safest road to hell is the gradual one—the gentle one, soft underfoot, without sudden turnings, without milestones, without signposts.” When belief in God becomes a familiar part of one’s intellectual landscape, one is much less likely to confront that belief for salvific purposes. It seems fairly clear that modifying S to include only belief in God doesn’t help Drange’s case at all. The existential problem Even if one comes to agree with the logical conclusions I’ve drawn in this paper, there still may be another way to reply to my rejoinder by way of surfacing what a colleague called “the existential tension” latent in ANB. It might go something like this. Given what I’ve argued above, one might conclude that at the very least every person has some awareness of God though that awareness may be hidden, suppressed, ignored or denied in some sense. Still, everyone has this awareness yet God is not able to cause any person to believe in Him if that person is to remain a human person created in the image of God. My interlocutor may want to response by saying that the next logical step in this argument is to talk about culpability. If the burden for nonbelief doesn’t lay with God, then it must lay with each person who does not believe (or it might lay secondarily with a community of people, whether that community is as large as a nation or as small as a family, who foster nonbelief).
“So,” the atheist might continue, “if all this is true, then it would follow that I have this suppressed belief in God and that I am somehow culpable or responsible for my disbelief. Further, if I am to come to know the truth, there is something I must do to change or cause a change in my noetic structure and affective will so that I come to love God.” At this point it would be easy to see how the atheist would simply deny that she in fact does have this belief (“I’ve checked thoroughly,” she might proclaim, “and it’s simply not there”) and that her nonbelief has nothing whatever to do with her will. Rather, she might argue that she is “compelled by the lack of evidence” or “so thoroughly convinced of the incoherency of the concept of God” that she can’t help but be an atheist.
Of course the point of this exercise is to attempt to force the theist’s hand and try to get him to concede by way of a sort of reductio ad absurdum: it would be ridiculous to suppose that all atheists have some internal knowledge of God that they’re suppressing and that if they would just discover and submit to that God, they would be Christian theists. An existential argument like this deserves an existential reply. Actually there are two ways to respond to this. First, if there actually is a God and he has given all persons a sense of himself, then it is likely that the scenario the atheist paints above is true. In other words, for the above scenario to be as ridiculous as the atheist wants to make it sound, it would have to be assumed that indeed God does not exist and such a claim clearly would be question-begging. To avoid the petitio, the atheist would have to allow that the scenario is at least possible and she might be suppressing or rejecting or somehow failing to recognize something that is present within her. Consider the following analogy.
Suppose a large body of scientists held that the human persons are hylomorphic beings made up of a body and nonmaterial substance or soul. The soul, they argue, contains all the mental activity of the person. They claim that such a truth is plainly obvious to anyone who considers the matter because, for example, by simple reflection one plainly can see that the objects of their mental life do not appear to have the same properties of the material world. Among this body of scientists is a small subset that believes differently. They hold that human persons are identical to the physical components that make them up and that there is no further nonphysical entity that need be included in human psychology to account for all mental phenomena one might encounter. The hylomorphicytes claim that for the ahylomorphicytes to sustain their belief, they have to deny what is plainly evident: that mental phenomenon does not have physical properties. The ahylomorphicytes respond by saying that the hylomorphicytes are attempting to claim that every ahylomorphicyte is somehow intellectually dishonest (by denying what is plainly evident) and that to attempt to sustain such a claim is foolish and ridiculous.
The claim of the ahylomorphicytes would be valid only if humans indeed did not possess a soul. However, if humans did possess a soul then the hylomorphicyte story may be correct and the response of the ahylormorphicytes should be expected. Why? Because if human beings do possess a soul, then there is a fact to the matter and it just is the case that the ahylomorphicyte fails to recognize that fact. Further, one couldn’t rule out the possibility that the ahylomophicyte could somehow be culpable for her disbelief simply because that claim is “uncomfortable.” If the ahylomorphicyte is somehow responsible, it is possible, perhaps probable, that she fails to recognize that fact.
The point of the analogy is that disbelief in what believers say is apparent may imply certain things about the disbeliever and seems to put an epistemic burden on the nonbeliever, those facts themselves should not constitute grounds for denying the believer’s claim. The theist may claim that God has made himself known to all persons by way of some internal capacity. The atheist may deny that such a capacity exists and more importantly that he doesn’t possess the awareness that capacity is supposed to produce. Still, there is a fact to the matter. Either God exists and has given each person a glimpse of Himself (however small or vague) or he hasn’t. The fact alone that a relatively small minority of people throughout history have failed to recognize God’s presence does not count against the truth that it is there (if it is a truth). The atheist cannot point to her lack of awareness as an argument against the theist’s claim. For her lack of awareness may be precisely what is at issue.
The second response I would make is this. Empirical evidence of the kind demanded by those who prize it seems neither to favor atheism nor theism. At best, the evidence for both sides stands in rough epistemic parity. My own take is that an atheist that claims to be an atheist solely on evidential grounds probably is being dishonest. But I would say the same for the theist as well. Though empirical and philosophical evidence plays an important role in belief (or disbelief) formation, most people are not in an epistemic position to clearly adjudicate between the evidence for and against. Other things are going on in belief formation (at least in the area of theology) and some of those things may involve culpability on the part of the inquirer.
I have argued in this paper that ANB suffers from a number of maladies. First, the modality of ANB is too strong. At best ANB would be a probabilistic argument and not a deductive proof against God’s existence. Second, I tried to show that a key premise of ANB–the one that argues that God has the ability to actualize S–is problematic and should not be accepted. If one rejects this premise, ANB does not go through. Finally, I claimed that even if S was softened such that the object of belief was God’s existence rather than the Evangelical gospel, ANB fares no better. Since there seems suffers from critical problems and there doesn’t seem to be any good way to bolster the argument, I conclude that ANB should be abandoned as an argument against the existence of God. 
 Drange does qualify this. He states that the argument is “an evidential” argument and it should only be taken as probabilistic. On page 71 he states, “As with AE, when all the support for all the premises is included, ANB needs to be classified as an inductive argument. It is an evidential argument, not intended to be conclusive, since the support for its premises is of an inductive or evidential sort. However, ANB itself, disregarding the support for its premises, is a deductive argument, and in my opinion, sound.” I’m not quite sure what to make of this. He seems to intend this in his treatment in Chapter 15. In this chapter, he seems to be interested in the strength of the evidence that supports the premises not in the structure of the argument itself (such that the latter is deductive but the relative evidential strength for each premise be less than 1 (or 100 in his convention). Regardless, he presents it as a deductive proof (avoiding the language of probability) and that’s how I’ll treat it.
 I understand a nonoverridden desire as a desire that has causal efficacy with regards to God’s action here in reference to his actions towards humanity. In reviewing an earlier draft of this paper, Professor Drange made the following comment about my understanding of the strength of God’s desire in this context. “Although ANB does claim that God desires situation S, it does NOT imply that that desire is particularly strong or that situation S is high on God’s list of priorities. All that ANB says is that God desires situation S and does not have some other desire which conflicts with and overrides that one. That could be the case with God even if situation S is relatively low on his list of priorities.” So it seems that Drange wants to say that the strength of God’s desire for S is not one of degree but of priority. But how does this work? How does one understand “not very strong” here? Not very strong relative to what? It seems like it would have to be relative to other desires. But then we face a contradiction: God’s desire isn’t very strong relative to the other desires he has but he also has no other desire that is stronger than his desire to actualize S. This would seem to make his desire to actualize S his strongest desire.
Drange himself seems to say that actualizing S is God’s strongest desire. He says in his review that God’s desire for S is not particularly strong however premise (A3) says very clearly that God’s desire for S is the strongest desire God has for humanity. It reads, “not wanting anything else that necessarily conflicts with his desire to bring about situation S, as strongly as he wants to bring about situation S” (emphasis mine).
Further, premise (A4) states that rationality means that God always acts according to his highest purposes. What does “highest” mean here? If God’s highest purposes means a desire that is “relatively low on his list of priorities” then Drange seems to be using “highest” in somewhat of a nonconventional way.
I think the best way to understand what Drange is after is to contextualize the desire. In this scenario, we’d have to contextualize God’s “greatest desire” to his greatest desire for humanity. That is, there is no greater desire that God has for humanity than to actualize S. This could be true but still not be God’s maximal desire vis-à-vis every desire in his noetic structure. We could say for example that the desire to exemplify His own goodness is his maximal desire while maximizing goods for humanity is relatively weak vis-à-vis that desire.
If that’s correct, then nonoverridden should not be understood as the greatest desire God possesses with respect to his entire noetic structure (what Daniel Howard-Snyder calls an “all-things-considered desire”) but his greatest desire with respect to human beliefs. I prefer to use the term “predominant” desire to capture this idea. Though it can’t be fully developed here, I’d offer the following as an explanation of predominant in this context.
(M) A property p of some member m is predominant in class c just in case there is no other member within c that exemplifies p to a greater degree than m.
So if the class is redness (R), and the member is a certain shade of red (call it R(1) and the property is hue h, R(1) has h predominantly in R just in case R(1) exemplifies h to a greater degree than R(x) where x is any other member of R.
As is, however, (M) will not capture what Drange seems to be after. Drange is clear that God does not need to desire S greater than he desires anything else such that God’s desire for S is the greatest desire in his overall noetic structure. However, it seems fairly clear that S has to be God’s greatest desire with respect to some object, namely humanity. As such, (M) would have to be modified in the following way.
(M1) A property p of some member m is predominant in class c with respect to object o just in case there is no other member within c that exemplifies p with respect to o to a greater degree than m.
So, for example, if the class is light, L, and the member is a certain light source (say a light bulb), L(1), and the property is brightness b, and the object is a given room, o, L(1) has b predominantly in L with respect to o just in case L(1) exemplifies b in o to a greater degree than L(x) in o where x is any other light source in o.
In this example, L(1) is predominant with regards to all lights but just to the lights within a certain spatial context namely a given room. I think the same can be said for God’s desire for S. In that case the class under consideration would be desire, the class member would be “desire for S”, the property would be strength, and the object would be humanity. If God’s desire for S is not predominant in the way just described, that desire would not be God’s “highest purpose” and (B) would not follow from (A). Drange wants to claim that (B) follows from (A) because (A4) claims that God always acts according to his highest purpose and God’s desire for S qualifies as his highest purpose. God’s desire for S has to be predominant (in the way just described) with regards to humanity otherwise his desire for S would not be his highest purpose and (B) would not follow.
Given the above then, I will go with the term predominant to refer to God’s desire for S with respect to humanity.
 Nonbelief and Evil, 58
 It is important here to understand ability as hypothetical not categorical. That is I don’t think Drange could claim that, given certain belief-desire-ability states that God possesses, God has the ability to act and to refrain from acting. Drange’s notion of God’s ability only includes the idea that God is not prevented from acting and so ability only means the ability to act.
 The following is a good, formal definition of this type of freedom: Person S is free with respect to an action A only if (1) it is within S’s power to perform A and it is within S’s power to refrain from performing A, and (2) no collection of necessary truths and causal laws–causal laws outside S’s control–together with antecedent conditions outside S’s control entails that S performs A. (Plantinga, Alvin. “On Ockham’s Way Out.” In God, Foreknowledge, and Freedom, John Martin Fischer, 178. Stanford, CA: Stanford University Press.) It is plausible to believe that such a definition of freedom is true of God for any logically possible acts he might perform.
 Certainly there are theological issues that surround the idea that God possesses libertarian freedom. Drange seems to incorporate the compatibalist (or stronger, determinist) view of God’s freedom as an essential component in his argument. My point is that such a view is not necessitated by an evangelical view of God and that understanding God as having libertarian freedom is not only possible but preferred despite whatever theological issues surface on such a view.
 This point was made to me by Steven Duncan.
 William Lane Craig suggests in his “Divine Timelessness and Personhood” that God’s choice to create the world can be described solely in terms of an exercise of will without having to include notions of desire. Craig says, “existing changelessly alone sans creation, God may will and intend to refrain from creating a universe. God’s willing to refrain from creation should not be confused with the mere absence of the intention to create. A stone is characterized by the absence of any will to create, but cannot be said, as God, to will to refrain from creating. In a world in which God freely refrains from creation, that abstinence is the result of a real act of the will, choosing between two available alternatives.” Though Craig doesn’t speak to the issue of desire specifically, it seems consistent with what he says here (and with classical theism) to take it that God did not create the world because he was necessitated to do so by some need or want on His part. His goodness was wholly satisfied just as much before creation as it was after. So why did God create? A plausible answer is that He created the material world ad hoc: by way of his own fiat, he created the world as an end in and of itself.
 As I noted in footnote 1, Drange is at best unclear about what he intends the logical force of the argument to be.
 In dealing with Hick’s “epistemic distance” argument, Drange references arguments made by Schellenberg and Keller that claim that God could cause beliefs without subsequent commitment to those beliefs. For example, “nontheists could become theists without thereby strongly committing themselves to God or obeying his laws. They could become ‘nominal Christians,’ who believe the gospel message as a kind of background assumption, but who do not have a religious attitude and do not practice the religion very faithfully.” (126). He goes on to strengthen this idea and claims that actualizing S doesn’t need to include the idea of salvation or commitment at all. “Although [people] would be caused to believe the gospel message, they would not thereby be caused to “accept Christ” and become religious Christians, so that would still be a matter of their free will and would still be up to them. And that sort of free choice . . . is really all that God is concerned about with regard to free will, according to evangelical Christianity.” (127, emphasis mine).
 It should be noted that the Bible does teach that all persons do have some knowledge of God but suppress and distort it (see Romans 1:18-23) and that is why they are not given more information about him (if one rejects a little knowledge of God, why would he give more?). Though this knowledge is not the Gospel as outlined in S, it certainly would satisfy the idea that God has given all men enough knowledge to make them culpable. So a convincing case could be made that God has brought about S at least to a minimal degree. More will be said about this below.
 The idea that beliefs can be formed directly by acts of the will. See Nonbelief and Evil, p. 331 ff.
 Ibid., 338-340
 On page 347 he writes, “I lean toward the view that no one is ever culpable for any action whatever. There would be different ways to support such a claim, but one way would be through a restricted form of determinism, with which I tend to agree.”
 For example, if at a given time t it is raining in Seattle Washington and Cedarville Ohio, a Washingtonian can utter the phrase, “It is raining.” That sentence is a linguistic “instantiation” of the proposition that represents (or is identical to depending on your view of propositions) the state of affairs that it is raining in Seattle and Cedarville. Though the propositions that stand behind that sentence are different (since they relate to two different states of affairs) the sentence does not distinguish between them. This ambiguity, in my opinion, is a key driver of Gettier-type counterexamples.
 See Keith Lehrer. Theory of Knowledge. Boulder, CO: Westview Press, 1990., pp. 7-10
 The stronger form of this type of belief is generally understood as propositional knowledge or knowledge that something is true. The belief we’re talking about here is propositional in this sense.
 I think this is along the lines of what Plantinga addresses in chapter 5 of his Warranted Christian Belief although his treatment is much broader than what I’m suggesting here. Certainly, there can be all sorts of dysfunction, none having to do with acts of the will, that may cause one to belief certain things. Some people in mental institutions suffer from such dysfunctions. As Plantinga points out, people like Freud held that belief in God is a coping device and in some sense a function of an immature mind. It would be hard to cash that in volunteeristically. Still, I do think it is fairly plain that many of us do will to believe various things and this fact can have a significant impact on our noetic structure.
 Jonathan Edwards, Religious Affections (Carlisle, PA: Banner of Truth, 1986). See Alvin Plantinga’s treatment of this in his Warranted Christian Belief, chapter 9.
 Drange may complain at this point and say the analogy is no good: for in it Addie has knowledge of the gift and of her parents love. The atheist (and ANB) wants to claim that people do not have knowledge of God or evidence of his existence. For Addie to act on the gift, she has to first know about it. This is all true enough (though I’ll argue later that it may not be the case that all men are ignorant of God). However the point of the analogy is to demonstrate the insufficiency of mere belief vis-à-vis a belief that involves a subsequent commitment. For all intents and purposes, having the former is functionally equivalent to having no belief at all since both do not result in the desired end: salvation. Belief in God has to be accompanied by an act of the will to commit to the object of that belief. This, I propose, is what God is unable to do.
 I hold that this is a type of knowledge by acquaintance. However I also hold that true religious belief involves a propositional knowledge-cum-knowledge by acquaintance. Both are necessary for salvation in the Evangelical understanding of that concept.
 For an interesting proposal on how nonbelief might universally be culpable and unreasonable, see Douglas V. Henry, “Does Reasonable Nonbelief Exist?,” Faith and Philosophy 18, no 1 (January 2001): 75-92.
 See Nonbelief and Evil, 63-69 for these arguments.
 Nonbelief and Evil, 66 (emphasis mine)
 In the introduction, Drange states that at root ANB is an argument against God’s existence based on the fact that there are so many nontheists in the world. However ANB as stated clearly argues using different premises. ANB clearly states that God does not exist because many people do not believe in the Evangelical gospel by the time they die. This is a more specified argument and demands a different sort of critique.
 Robert McKim, Religious Ambiguity and Religious Diversity (Oxford: Oxford University Press, 2001), 102. McKim makes an excellent distinction between what he calls “full theism” and “minimal theism.” The former “takes seriously the belief that God exists and focuses on and pays attention to the nature of God. This belief is central to, rather than peripheral to, his system of beliefs.” (58) It is this type of belief that I believe is necessary to actualizing S.
 See Jonathan Kvanvig’s insightful article, “Theism, Reliabalism, and the Cognitive Ideal” in Michael D. Beaty, Christian Theism and the Problems of Philosophy (Notre Dame: University of Notre Dame Press, 1990), 71-91.
 see his God, Freedom and Evil (Grand Rapids: Eerdmans Publishing Company, 1974) and Nature of Necessity (Oxford: Oxford University Press, 1982) chapter IX
 Plantinga, God, Freedom, and Evil, 49-53
 For example, we might say that infants possess few, if any, first-order capacities with respect to the freedom of their will. However, most people would (pace Peter Singer and the like) would not want to say infants aren’t human.
 McKim, Religious Ambiguity and Religious Diversity, 67,68
 Ibid., 61
 Of course this does not rule out the possibility that God greatly influences our belief formation and character nor the possibility that God directly changes our beliefs and character under certain circumstances. Neither does it rule out the idea of efficacious grace (a doctrine held by many in the Christian tradition). All this rules out is the possibility that God can actualize S.
 Philosophers arguing along these lines include William Alston, Nicholas Wolterstorrf, and Alvin Plantinaga. Plantinga makes his strongest case yet for this position in his latest opus Warranted Christian Belief. In this book, he builds upon the strong epistemological system he developed in Warrant and Proper Function and applies that system to Christianity. Of course this line of argument is not new. It was adopted in one form or another by the greatest and influential thinkers of history including Anselm, Augustine, Calvin, Luther, and Thomas Aquinas. See Paul Helm. Faith and Understanding. Grand Rapids: Eerdmans, 1997 and Dewey J. Hoitenga, Jr. Faith and Reason from Plato to Plantinga. New York: State University of New York Press, 1991.
 On page 139 Drange notes that according to the 1996 World Almanac, two-thirds of the earth’s population is “non-Christian.” However if we analyze the demographics in terms of theism, the numbers change significantly. In 1996 the US Department of Commerce, Bureau of the Census lists the world population in 1996 at 5,772,000,000. The 1996 Encyclopedia Britannica Yearbook lists atheism at 219,925,000 adherents or 4% of the world’s population. Certainly I’m not attempting to claim that the rest of the world population is theistic in terms friendly to the Judeo-Christian concept of God. But I don’t think S’ demands such a conception. Drange’s figures are misleading in that they don’t allow for a knowledge of a higher power or “transcendent” being (or beings). This is the sense of theism that I believe is relevant in this case.
 I want to thank Paul Herrick for invaluable feedback on an earlier draft of this paper especially for help with stylistic improvements and Steven Duncan for his penetrating analysis and feedback. I also want to thank David Sanders for hours of helpful discussion on matters of religion and philosophy and against whom I was able to bounce many of the ideas presented in this paper. Two anonymous reviews of this paper saved me from embarrassing mistakes and pointed out weaknesses in the paper that I was able to address. I wish to thank the two individuals who wrote those reviews as well. Finally, I want to thank Jeffrey Lowder who was the impetus for this response and who handled getting peer reviews and feedback from Professor Drange.