Is divine foreknowledge compatible with free will?

The objective of this essay is questing accounts in order to object theological fatalism and evaluate the potential solutions for the problem which theological fatalism brought to free will. Theological fatalism is the argument saying that God’s foreknowledge[ In this essay, God’s foreknowledge is interchangeable with divine foreknowledge and omniscience.] of human’s behavior leads to a result that human lack of free will and leads to determined outcomes of every event in the past and future. (Zagzebski, 2017). If this doctrine is correct, then there is God who cannot be wrong about his belief about the future, and hence, this being’s foreknowledge determines the entire history of human and human action is not free. Since compatibilism believes that free will is compatible with divine foreknowledge, theological fatalism is the denial of compatibilism. Firstly, I am going to introduce the theological fatalism, and then consider evaluate how two accounts argue against the theological fatalism. Later, I will evaluate the pros and cons of Frankfurtian solution and Aristotelian solution respectively, then compare their plausibility. Finally, I will conclude with the best possible solution: Aristotelian accounts.

Introduction of theological fatalism

Firstly, theological fatalism offers an account to explain the incompatibility between omniscience and will of freedom. Tapp (2011, p.66) pointed out that generally, theological fatalism argues that we are not capable of changing what God believes to be true, nor can we alter that God is infallible. The features of God’s past beliefs of future events and the infallibility brings about the unchangeable future. If we cannot alter the truth value of God’s past belief, then we have no control over the future.

This is a basic deduction based on some of the principles to yield the argument for theological fatalism:

Let T be the statement that you will answer the phone at 9 am tomorrow.

“(1) Yesterday God infallibly believed T. [Supposition of infallible foreknowledge]”

“(2) If E occurred in the past, it is now-necessary that E occurred then. [Principle of the Necessity of the Past]”

“(3) It is now-necessary that yesterday God believed T. [from 1, 2]”

“(4) Necessarily, if yesterday God believed T, then T. [Definition of “infallibility”]”

“(5) If p is now-necessary, and necessarily (p → q), then q is now-necessary. [Transfer of Necessity Principle]”

“(6) So it is now-necessary that T. [from 3,4,5]”

“(7) If it is now-necessary that T, then you cannot do otherwise than answer the telephone tomorrow at 9 am. [Definition of “necessary”]”

“(8) Therefore, you cannot do otherwise than answer the telephone tomorrow at 9 am. [from 6, 7]”

“(9) If you cannot do otherwise when you do an act, you do not act freely. [Principle of Alternate Possibilities]”

“(10) Therefore, when you answer the telephone tomorrow at 9 am, you will not do it freely. [from 8, 9]” (Zagzebski, 2017).

Compatibilists’ response is against the argument of theological fatalism and they offer some solutions below.

Frankfurtian solution

introduction of PAP

Firstly, before I say what is wrong with the process to yield the argument of theological fatalism, I want to introduce the principle of alternate possibility (PAP). The content in 9th premise of the basic argument above was supported by PAP which “states that a person is morally responsible for what he has done only if we could have done otherwise.” (Frankfurt, 1969, p,829). Avila-Peregrina (2011) asserted that the will of freedom is equivalent to bearing moral responsibility, i.e., we can say that free will requires moral responsibility, which means that you are free only if you are morally responsible for what you do. Then, PAP implies that free will requires the ability to do otherwise. If we put Avila-Peregrina’s assertion and PAP together, we can have that PAP proves the validity of premise 9 because what PAP further implies is what premise 9 states: “if you cannot do otherwise when you do an act, you do not act freely.”

PAP proves that 9th premise is true for now, but there is something we can do to argue against the theological fatalism. If we can find the flaw of PAP, then we can successfully argue against the argument of theological fatalism, or if we can prove that there is some circumstance which we can be free even if we do not have choice in front of us, then we can say the determinism is compatible with free will.

Frankfurt’s objection to PAP with hypothesis

It might seem PAP is plausible at first glance, however, Frankfurt pointed out that there is falsity in PAP, because Frankfurt argued that there may be a circumstance that someone bears moral responsibility for what he did even though it is not within his power to do otherwise and because there is a possibility that sufficient conditions are met for him to do a certain act and hence there is no way for him to do otherwise, but that he is not forced to produce his action. (Frankfurt, 1969, p.829-830). If the falsity of PAP is proved, then theological fatalism would turn out to be false and it would make compatibility between divine foreknowledge and free will plausible.

Harry Frankfurt ran a thought experiment to prove the falsity of PAP:

Assume that there is someone named Black, he wants Jones to act in a certain way. Black had a remarkable plan to be ready to get it done without showing any evidence of his existence. So, until Jones is about to decide what to do, he does nothing. He is superb in knowing when to start the plan. He will manipulate Jones to decide to do what black desires and actually do it instead of what Jones wants to do once Black notices that Jones is going to decide to do something else instead of what Black wants. Black’s plan works every time. He has never been noticed before and will not be noticed by Jones in the following future. (Frankfurt, 1969, p.835).

Let us suppose that Black decides to leave Jones to act on his own, because there could be a case that Jones decides to perform the certain behavior which is what Black wants. In this situation, Jones for sure ought to hold this moral responsibility for what he does because he acts on his own and no one impels him to do. It would be implausible to lessen the blame for his act, just because of the fact that he does not have alternate choices in front of him. With or without the fact, Jones would have acted the same based on his inclination or under Black’s manipulation. (Frankfurt, 1969, p.835).

The above hypothesis offers environment which is ample for Jones to do the act being considered. The certain action he tends to perform does not depend on himself because Jones must end up performing a certain action Black desires and Black is not going to reveal himself. The only thing which is up to him is which one he chooses among the choices between doing the act willingly or as an outcome because of Black. The choice is made by what action he chooses to perform. However, whatever he chooses eventually, he must act in the same way. He gets no choice other than performs what Black desires. In spite of the presence of Black lurking at the background, if Jones acts willingly, moral responsibility for the act will not be erased since Black did nothing. (Frankfurt, 1969, p.836).

Nevertheless, some might think that the efficacy of the case is ruined by its dependence on this schemer (Frankfurt, 1969, p.863). So, perhaps you can replace Black’s control by natural forces, or imagine that a man decides to remain in a room without noticing the door is sealed, instead of deciding to go out (Pike, 1965, p.33), then in these cases, we can easily tell that he is morally responsible because he acts on his own. For these situations, according to the assumption, Black has nothing to do with what he did. Jones would be led to do the exact same thing. (Frankfurt, 1969, p.837). Hence, Frankfurt concluded that there are some circumstances which someone bears moral responsibility for what he did even though he has no ability to do otherwise (Frankfurt, 1969, p.829-830). It seems that Frankfurt suggests that moral responsibility should be considered as a consequence of what actually builds up the performance, but not on Jones’ chance to do otherwise. Hence, there is some circumstance which someone does not have second choice but is still morally responsible for his action. (Naylor, 1984, p.250). Frankfurt proved that not being able to do otherwise can neither entail the absence of moral responsibility nor then entail the absence of free will.

Therefore, Frankfurt gave a solution for the problem of PAP by replacing PAP with the new principle: “a person is not morally responsible for what he has done if he did it only because he could not have done otherwise” (Frankfurt, 1969, p,829), instead of keeping PAP, “a person is morally responsible for what he has done only if we could have done otherwise.” (Frankfurt, 1969, p,829). PAP implies that the necessary condition for holding moral responsibility is having alternate possibilities, which also means that the ability to do otherwise alone can be the necessary and sufficient condition to produce the result that he for sure bears moral responsibility. And hence, he cannot have moral responsibility if he could not have done otherwise. The new Frankfurtian solution implies that bearing moral responsibility does not entail that the possibility to do otherwise. Since moral responsibility requires will of freedom, what the account shows, then, is that the ability to do otherwise are not always relevant to free will. (Zagzebski, 2017).

Therefore, Frankfurtian solution proves the falsity of PAP in the above sense, hence the content in 9th premise in the argument of theological fatalism is no more valid. Thus, the solution is against the theological fatalism. PAP shows the incompatibility between the determinism and moral responsibility and since theological fatalism is a denial of compatibilism, now we got the Frankfurt’s argument to prove the falsity of PAP and to account the compatibility between the divine foreknowledge and the free will. It also allows the compatibility between the determinism and the free will.

Naylor’s objection to Frankfurtian solution

However, Frankfurtian solution to PAP does not seem convincing enough. Naylor’s argument against Frankfurt’s objection to PAP may be sound. As we have seen that, in Frankfurt's counterexample, we need to consider the case whether he has the ability to do A willingly or not. In another sense, Naylor argues that we can also consider that either (1) that “it was within Jones's power not to do A”, or (2) that “it was not within Jones's power not to be compelled to do A”. Clearly, we are sure that (1) must not happen in the case because Jones always act A in Frankfurt's counterexample. Then, the outcome must meet the situation (2) in the thought experiment of Black, i.e. Jones has no ability not to be led to do A. Yet, Naylor contends that for Jones to hold moral responsibility, it requires that he is not led to do A. Thus, Jones is not morally responsible for doing A, because the least necessary condition that Jones has not been led to do A regardless of him doing A willingly or not has not met in Frankfurt’s thought experiment. The Frankfurt's counterexample to PAP shows that “it is not the case in which someone is morally responsible for performing an act when he could not have done otherwise.” Therefore, Naylor argued that Frankfurt did not prove the falsity of PAP. (Naylor, 1984, p.256-257).

The conclusion which Frankfurt's case brings and proves according to Naylor, is that we bear moral responsibility for doing A on our own, whereas we are not morally responsible for being compelled to do A. Someone holds moral responsibility for doing A because he has the power to do otherwise on his own; but bears no moral responsibility for doing A because he has no ability not to do A. Furthermore, Naylor suggests that from the moral prospect, Jones should be blamed or praised if he did something willingly or by choice. If this is the situation, then we can say Jones is morally accountable for what he did because he had what he wanted even though he is not morally responsible because there is no alternate possibility for him. (Naylor, 1984, p.256-257).

conclusion of Frankfurtian solution

Thus, it turns out that the thought experiment of Black’s intervention is not effective enough. Naylor concludes that Frankfurt's objection to PAP does not show that people bear moral responsibility for doing A when there is no alternate possibility. (Naylor, 1984, p.256-257) because Frankfurt considers moral responsibility as a consequence of what made up by the performance, but not on Jones’ chance to do otherwise. The conclusion is not equivalent to PAP stating that “a person is morally responsible for what he has done only if we could have done otherwise.” (Frankfurt, 1969, p,829). In addition, Naylor suggests that if somebody desires to do A, then he is morally accountable for what he did on his own even if he is not morally responsible for what he eventually did when he cannot do otherwise. The objection to the solution seems plausible enough to suggest that the Frankfurtian solution is not effective enough to prove the falsity of PAP and cannot support the compatibilism between determinism and free will.

Although Naylor offers a plausible account, it does not mean it is persuasive enough to make me totally support it. Many others would like to object the account Naylor offers by saying, for example, that the desire to act could be determined as well, even there is a second order desires, their inclination to act in a certain way can still be determined. I believe other solution can help us make a breakthrough to attack the theological fatalism.

Aristotelian solution

Another way to object the argument of theological fatalism is by accepting Aristotelian solution by claiming that the statement about contingent future has no truth value because no one can know the truth value of contingent future until it happens and becomes a fact. If it is true, then statement in premise 4 will be false. The statement says that “necessarily, if yesterday God believed T, then T” which is a future event that has not happened yet.

Since Aristotle is in no doubt that it is not the case that everything that happens, happens necessarily. He accepts that “What is, necessarily is, when it is; and what is not, necessarily is not, when it is not.” Then, he added that “not everything that is, necessarily is; and not everything that is not, necessarily is not.” (Rice, 2018). Aristotelian solution is about to disagree the idea that the certainty of occurrences of events is made to be sure even when those events associate with the things that do not happen necessarily. (Rice, 2018).

Purtill’s view supports Aristotelian solution

To support Aristotelian solution, first, I need to introduce a view Purtill offers and he calls it the commonsensical view, although it is not common sense to me, it is still plausible. The view states that some propositions of future always are true or false, whereas truth values of others about the future remains unknown and they are only probabilities which have a range of zero to one before they happen. “This probability varies from day to day, indeed very likely from instant to instant”, the truth values of the statements about future will not reveal themselves until the moment the events occur. (Purtill, 1988, p.187). Purtill asserts that any version of bible and Christian tradition about omniscience is compatible with the view which future cannot completely be known by God because some statements of future events are neither true nor false. (Purtill, 1988, 189). So, he seems confident that this view is the plausible one and I think the view is compatible with Aristotelian account: some statements about future are not either true or false.

The view argued that God tends to come up with probabilities of any occurrences which have no truth value instead of truth value of the statement about future. If the view is true, it seems like Purtill uses it to account that God can foresee the probability of occurrence of events and always only believe what is true, instead of those without truth value for now but will acquire one in the future. It infers that human’s free will and action can determine truth value of some future events when God can only have probability of occurrence of events. It implies that free will is compatible with divine foreknowledge and is compatible with determinism. Aristotelian solution claims that some propositions do not have truth value yet will get one in the near future. The view is also coherent with Aristotelian solution.

Thus, according to Purtill’s view, I think it is appropriate that the omniscience can be defined as the ability of knowing every fact that happened and will happen necessarily and the probability of occurrences of every event that is neither true nor false. According to the definition of infallibility introduced before, God’s belief is infallible, so there is no way for God to believe in something false. (Zagzebski, 2017). It implies that God does not believe in the future which is false or is neither true nor false, which seems that God’s infallibility of divine foreknowledge is compatible with the view.

To better understand how the view is compatible with Aristotelian solution and divine foreknowledge. Let us assume a statement, “Geraldine Ferraro will be the president of the United States in 1990.” be F1. Now by the end of 1990 this statement, F1 must has its truth value if the view is right. (Purtill, 1988, p.185). For example, God might make some infallible statements by the end of 1st June, 1890, “on 1st June, 1890, Geraldine Ferraro is likely to be the President of the United States in 1990.” We can see that this view support what Aristotle believes – “not everything that is, necessarily is”, it is not always the case that everything happens necessarily. The above entails that those events whose truth value is neither true or false will obtain one in the near future (Zagzebski, 2017).

What’s more, this solution allows defenders to come up with accounts to maintain the compatibility between divine foreknowledge and free will and is also coherent with God’s infallibility in knowledge. There is the example offered by Purtill for accounting the Aristotelian solution. Take the example of Ms. Ferraro, the probability of F1 is neither zero nor one until Ms. Ferraro succeed in the campaign according to the view. At this point, Purtill suggests an idea that the chance for F1 to be true will rise if Ms. Ferraro and George Bush come up with something after an argument and will decline if she profits from a commercial. (Purtill, 1988, p.187). This is compatible with God’s being infallible in everything he does believe because God does not believe those false and those without truth value. It is also compatible with God’s omniscience if omniscience is the ability including foreseeing those events that will happen necessarily and knowing the probability of occurrence of some future events.

However, some might say that the modified definition of omniscience sets the limit for the range of God’s knowledge, so not only it challenges how we see God traditionally, but also the view shows its drawback of laws of logic (Zagzebski, 2017).

Flaw of Aristotelian solution about laws of logic

Traditionally, Divine omniscience used to be considered as a property of knowing everything that includes every detailed statement about the future which God believes to be true and everything is present to God. Yet, Aristotelian solution shows that statements about contingent future can neither be true nor false because what happened might not have happened necessarily according to what Aristotle believes. So, the Aristotelian solution seems to reject the law of bivalence: “The law of bivalence: every proposition is either true or false.” (Rice, 2018).

The Aristotelian solution makes some sense to me, because there is a counterexample to law of logic, when we consider a proposition “Christmas Day is the best festival”, it cannot either be true or false, it is just an expression of opinion. It reminds us that there are only limited circumstances to which propositional logic can be only applied. (Purtill, 1988, p.189).

Flaw of Aristotelian solution about limitation of divine foreknowledge

Aristotelian solution might seem to restrict the range of God’s foreknowledge. Conventional Christianity tells us that God is the omniscient being, i.e., “he knows every true statement,” (Purtill, 1988, p.189), till the solution claims that some events of contingent future are unknown to the omniscient being. So, it seems to run counter to the traditional picture of God. According to the solution, God would not have believed that she will be president in 1990 if God knew it will not happen, because God is infallible about the true statement and God does not believe any such false statement and other statement without truth value. Hence, the defender of the solution admits that God does not know about some contingent futures such that God cannot predict the exact result that whether Ms. Ferraro will be president or not in 1990 until it happens. It implies that Aristotelian solution seems to restrict the range of God’s divine foreknowledge. God does not always know the truth of every statement, but that does not mean God’s omniscience is restricted. God is capable of knowing the truth value of all true statements, the falsity of all false statements and of calculating the chance of occurrences of all events which are lack of truth value. Thus, He knows precisely how many chances that Ms. Ferraro will be president in 1990. We can consider that God cannot know that she will be or that she won’t be because there is no such a fact to know about, and thus it is plausible to consider that it is not a limitation of God’s omniscience not to know it. (Purtill, 1988, p.189). Purtill’s view argued that God tends to come up with probabilities of any occurrences which have no truth value. If it is true, then the omniscience will be defined as the ability which includes knowing about every determined fact that will happen necessarily and knowing the probability of occurrence of some future events.

conclusion for Aristotelian solution

Therefore, in my point of view, Aristotelian solution is persuasive when it points out the flaws of the argument of theological fatalism and shows the compatibility between free will and divine foreknowledge.

Conclusion

In conclusion, theological fatalism is attacked by two solutions: Frankfurtian solution and Aristotelian solution. Above, I have evaluated the Frankfurtian solution and Aristotelian solution and compare their plausibility. Firstly, Frankfurtian solution offers the objection against PAP. It seems not convincing enough that we are morally responsible even when we are left with no choice, because Naylor suggests that the thought experiment of Black is not effective enough to object the PAP, the external force gave us no choice but leading the agent to the same action and same outcome. Since the necessary condition for someone to hold moral responsibility is for someone to be within his power to resist to act what the external force desires. Eventually, Frankfurtian solution did not solve the problem that is left to compatibilism between divine foreknowledge and free will. On the other hand, Purtill’s view seems to support Aristotelian solution and helps define divine foreknowledge as the ability which includes knowing about every determined fact that will happen necessarily and knowing the probability of occurrence of some future events which may happen or may not happen unnecessarily. That leaves room for human to act freely so as to influence the probability of occurrences of any events that are lack of truth value and to choose to actualize those events or not in the contingent future. Each choice humans make can affect the probability. For God’s omniscience not knowing about those undetermined contingent future, it is because there is no such fact for God to know about future event until it happens. God’s foreknowledge might seem to be limited, but I think it is still plausible and persuasive as long as God is still the greatest possible being. Therefore, by comparing their plausibility, it is reasonable to think that the Aristotelian solution is more convincing than Frankfurtian solution so far.

==========================================================================================================================

Reference:

Avila-Peregrina, Karla, ‘Is Free Will an Illusion?’ (Lake Forest College, 2011)

Frankfurt, Harry G, ‘Alternate Possibilities and Moral Responsibility’, The Journal of Philosophy, 66.23 (1969), 829–39 <http://www.jstor.org/stable/2023833>.

Naylor, Margery Bedford, ‘Frankfurt on the Principle of Alternate Possibilities’, Philosophical Studies: An International Journal for Philosophy in the Analytic Tradition, 46.2 (1984), 249–58 <http://www.jstor.org/stable/4319705>

Pike, Nelson, ‘Divine Omniscience and Voluntary Action’, The Philosophical Review, 74.1 (1965), 27–46 <http://www.jstor.org/stable/2183529/>.

Purtill, Richard L, ‘Fatalism and the Omnitemporality of Truth’, Faith and Philosophy, 5.2 (1988), 185–92.

Rice, Hugh, "Fatalism", The Stanford Encyclopedia of Philosophy (Winter 2018 Edition), Edward N. Zalta (ed.), URL = <https://plato.stanford.edu/archives/win2018/entries/fatalism/>.

Tapp, Christian, and Edmund Runggaldier, God, Eternity, and Time (Ashgate Publishing, Ltd., 2011) URL = <https://books.google.co.uk/books?hl=en&lr=&id=gK2RR6TEMPwC&oi=fnd&pg=PR7&dq=God,+Eternity,+and+Time&ots=CTAS7sNY18&sig=8hky6h7hwuMV5fIFVAJK3EroVbM&redir_esc=y#v=onepage&q=eternity%20and%20fatalism&f=false>

Zagzebski, Linda, "Foreknowledge and Free Will", The Stanford Encyclopedia of Philosophy (Summer 2017 Edition), Edward N. Zalta (ed.), URL = <https://plato.stanford.edu/archives/sum2017/entries/free-will-foreknowledge/>.