Part 1: The Form of the Arguments
Problem of Evil
Since we are looking at the logical Problem, we will be looking at a logical argument. Many people state the Problem in an over simplified manner, such as; “If God exists, then there is no evil,” “there is evil,” therefore God does not exist. The problem with this overly simplified presentation is that, though the conclusion logically follows from the premises, the first premise is not self-evidently true. We must ask; does God’s existence really mean that evil cannot exist? There are unstated premises that must be elucidated before this argument can be effective. In academic and apologetics literature, it becomes clear that the perceived contradiction results from God’s character according to the Bible. Flushed out, the argument looks like this;
(1) If God was all-powerful, He could stop evil.
(2) If God was perfectly good, He would want to stop evil.
Conclusion: If an all-powerful and all-good God existed, then evil would not exist. (From 1 & 2)
(3) Evil exists.
Conclusion: An all-powerful and all-good God does not exist (From 1 + 2 & 3)
In the history of apologetics, theologians have unsuccessfully tried to deny Premise 1, 2, and 3. To deny that God is all-powerful or could stop evil is to eviscerate the biblical picture of God, similarly with denying the second premise. To reject the third premise, that evil exists, is the route that some cults and early church thinkers have taken. To do this is to make a sham of our lives, for we frequently experience evil; to reject this premise is contrary to all human experience and the testimony of Scripture, so this answer must also be rejected.
Proponents of the various theodicies and the Free Will defence seek to uphold the character of God and the existence of God by pointing out that the atheologian, with this argument, is assuming that God has no justified reason for allowing evil to exist while remaining perfectly good and all-powerful. They suggest that this is an unwarranted assumption, if it can be shown that God has a good reason (theodicy) or even that there could be a logically coherent reason (defence), then it can be shown that the existence of God and of evil is not a logical contradiction. The FWD suggests that we are justified in believing that there is a reason why God allowed evil, though He is perfectly good and all-powerful.
The Free Will Defence
The origins of the FWD are found in the early writings of Augustine. Dialoguing with Evodius, Augustine suggests that there is a sufficient reason for evil in the world in that we need free will to live aright and a consequence of having a free will is the possibility of sinning, this consequence is justified by the obvious good that results from God’s justice displayed through punishment and rewarding. Augustine latter adopted a different position on free will in his writings against Pelagius and this early theodicy did not gain traction as the answer to the Problem of Evil.
Following Augustine’s example, philosopher Alvin Plantiga introduced the FWD in the 20th century; it is his argument that has received much attention in apologetics over the last 50 years and that has been lauded as the answer to the logical Problem of Evil. With Augustine, Plantiga’s argument stands or falls on 1 key presupposition. It assumes that the nature of freedom is what has been called Incompatibilism. Incompatibilism says that having freedom is to have the ability to choose freely between choices (choice A or not-A) without any sufficient condition present to guarantee that one makes a specific choice. This view of free will means that to be held responsible for an action, an agent needs to possess free will in the sense of Incompatibilism; if someone’s action has a sufficient condition then he cannot be held responsible. This writer is convinced that this presupposition is necessary to every presentation of the FWD. Admittedly; I have not examined every presentation of the FWD since Plantiga first introduced it, but to attempt to make this argument without the Incompatibilist presupposition is to eviscerate its apologetic power. The FWD is the name given to the specific formulations of a defence that follow or have been molded by Plantinga’s argument. Key to this defence is the idea that meaningful creaturely freedom is a greater good than a world without this freedom, and that it is plausible that this freedom introduces the possibility of free choices for evil. As we examine the argument below, this will become clear. The minute compatibilism is conceded, these aspects of the defence fall apart; compatibilism furnishes God with no limits on the possibilities He could create, other than those that are necessitated by His nature. Just as God will one day create a new heavens and new earth free from sin, within a compatibilist model He could have created His initial creation this way.  Someone may formulate an alternate apologetic defence that involves free will, but without these presuppositions, there is no justification for calling it “the Free Will Defence.”
What the FWD attempts to do is show that God—though omnipotent and omniscient—could not have created a world without evil that contained as much good as this one, which contains evil. When reading this argument it must be kept in mind that it does not have to present the actual reason why God allowed evil, all it has to do to succeed is to give a logically plausible scenario in which God allows evil and maintains all His attributes.
The argument supposes that “A world containing creatures who are significantly free (and freely perform more good than evil actions) is more valuable, all else being equal, than a world containing no free creatures at all.” A creature is significantly free when he is free in a circumstance, according to the incompatibilist presupposition, to perform a morally significant action. An action is morally significant when to do it is objectively wrong and to refrain is objectively right, or vice versa.
The FWD argues that it is possible that God was unable to create a world where significantly free creatures would not choose to perform evil acts. If a world of significantly free creatures is a greater good, all else being equal, than a world without significantly free creatures, then God would be justified in creating this world with significantly free creatures and evil introduced by their free choices. The FWD would then look like this:
(1) All else being equal, a world containing significantly free beings is a greater good than a world without significantly free beings.
(2) God, even though He is omniscient and omnipotent, is unable to create a world containing significantly free creatures in which they do not choose to do evil.
(3) Therefore, God is justified in creating a world that contains both evil and significantly free creatures.
If (2) can be shown to be plausible, then the FWD shows that God’s character and the existence of evil are not logically contradictory. Is it true that, assuming incompatibilism, an omnipotent God could not have created world where free creatures always choose what is right?
Those who would disagree with this premise argue that an omnipotent God could create any world that He pleases. This is where FWD proponents suggest that atheologians go wrong. It is not a logical contradiction for a world to exist where free creatures only do what is right, but is it possible that God could not have created this world? Using the doctrine of middle knowledge and possible worlds, it is argued that it is possible that the only possible world that God could create is one in which free creature will inevitably fall. For the sake of brevity, we will not outline Plantinga’s whole argument, but just enough to show that—according to the presuppositions of the FWD—it is possible that God could not have created a world where free creatures do not fall.
Let us suppose that if God knows the truth of proposition (4): in circumstance C, person P will freely choose to plant a flower in a flowerbed. Because (4) is true, then in every possible world where circumstance C obtains, person P will freely choose to plant the flower. Because (4) is true, it is impossible for God to create a world in which person P does not plant the flower. Either (4) or the contrary proposition (5) has to be true; if one is true, then all worlds where the opposite action by person P obtained would be impossible for God to create. If it is possible that God is unable to actualize worlds that contain significantly free creatures, could not the logically possible world where every free creature does only what is right be one of them?
We have seen that—supposing (4) to be true—it is impossible for a possible world to exist where person P freely chooses not to plant the flower in circumstance C. It would then seem possible for there to be no possible worlds where a significantly free person does not perform and evil act. Plantiga describes the condition of the unfortunate individual who does an evil act in every possible world as suffering from “transworld depravity.” If it is possible that one person suffers from transworld depravity, then it’s also possible for everybody to suffer from transworlddepravity. If it is possible that even one person suffers from transworlddepravity, let alone everybody, then it is plausible for (2) to be true. If (2) is even plausibly true, then the FWD succeeds, for the character of God is shown to be compatible with the existence of evil.
The FWD has led many philosophers to accept the Christian faith, but many Christians have suspected that a few of the arguments presuppositions are its downfall. The 2 biggest questions, which are interrelated, are; “is Incompatibilism the only possibility for free and responsible action?”, and “does God possess middle knowledge?” It is to these questions that we must now turn our attention. We will first examine the objections to Incompatibilism from a Christian worldview, then we will turn to those from the realm of philosophy, and lastly we will examine the possibility of middle knowledge.
John M. Frame, Apologetics to the Glory of God: An Introduction (Phillipsburg, N.J: P&R Publishing, 1994), 150; Hick, “Evil,The Problem of,” 136. J.L. Mackie, in his powerful presentation of the argument, included God’s omniscience with premises 1 & 2. Howard-Snyder, “God, Evil, and Suffering,” 83.
 A sufficient condition, contrasted with a necessary condition, is such that if the condition is fulfilled, y will obtain; y being any effect resulting from the fulfillment of the sufficient condition x. An example would be; if I hold a lighter to a match (the sufficient condition=x), it will ignite (the result=y). A necessary condition is such that its presence is necessary for the consequent, but its presence does not guarantee that the consequent will obtain. For example; a necessary condition for me to start and drive my car is that there is gas in the tank, but having gas in the tank does not automatically result in my car starting and driving. Having gas is a necessary, but not sufficient, condition.
James Porter Moreland and William Lane Craig, Philosophical Foundations for a Christian Worldview (Downers Grove, Ill.: InterVarsity Press, 2003), 270; John S. Feinberg, “God Ordains All Things,” in Predestination & Free Will: Four Views of Divine Sovereignty & Human Freedom, ed. David Basinger and Randall Basinger (Downers Grove, Ill: InterVarsity Press, 1986), 20–21.
Norman L Geisler, Systematic Theology : In One Volume (Minneapolis, Minn.: Bethany House, 2011), 518; Howard-Snyder, “God, Evil, and Suffering,” 88; Michael J. Murray, “Heaven and Hell,” in Reason for the Hope Within, ed. Michael J. Murray (Grand Rapids, Mich: W.B. Eerdmans, 1999), 295; Craig, On Guard, 156; Alvin Plantinga, “The Free Will Defense,” in Christian Apologetics: An Anthology of Primary Sources, ed. Khaldoun A. Sweis and Chad V. Meister (Grand Rapids, MI: Zondervan, 2012), 422. It would seem that Ronald Nash—and maybe Scott Oliphant (though after presenting Plantiga’s argument with compatibilist clarifications, he seems to move towards an agnostic defence such as Frame’s)—attempts to use this defence while adhering to a Compatibilist understanding of free will, but this nullifies the core of Plantiga’s defence, as should become obvious throughout this paper. Nash, Faith & Reason, 185–194; Oliphint, Covenantal Apologetics, 161–192.
 Carson agrees that this understanding of human responsibility leaves this defence without any ground. D.A. Carson, Divine Sovereignty and Human Responsibility (Eugene, Oregon: Wipf and Stock Publishers, 2002), 208–209.
 Along with the incompatibilist presupposition of this argument, middle knowledge is thought to be necessary for this argument to work. Some may suggest that God does not need middle knowledge to know what his significantly free creatures will do; if this is the case, then the above answer given by FWD proponents can, with some modification, also work for these defenders. The possibility of middle knowledge or God’s knowledge of significantly free creatures’ choices will be discussed below. For an explanation of the doctrine of possible worlds see Plantinga, “Christian Apologetics,” 426.