LECTURE #23: The Free Will Defense

I. The Tender-Hearted Response

The tender-hearted responses depend on challenging premises 4 and 5 of the standard argument, namely,

4. A world with the greatest possible surplus of good over evil would be a world devoid of evil.
5. An omnipotent God could actualize any world. By rejecting one or both of these premises, the tender-hearted theist is, in effect, arguing for some relevant limitation on what an omnipotent God could do. If we reject premise 4, we are suggesting that the range of possible worlds is narrower than we might have thought. Since whatever God can bring about must ipso facto be possible, any limitation on the range of possibility is also a limitation on the range of God's omnipotence. If we reject premise 5, we are very obviously placing limits on God's omnipotence, limits that are additional to the limits of the possible.

II. Four Accounts of Free Will

A. The compatibilist/soft determinist view. Free-choice is an extrinsic type, one that involves how an action is caused/determined.

  1. A soft-determinist account of freedom: Agent x's action A in circumstances C is free iff x was caused to do A (determined with probability 1 to do) in circumstances C by a process of type F.
  2. What is involved in being a "process of type F"? We have little idea. All we have are particular cases of F and of non-F (such as duress, schizophrenia. drug-induced mania, etc.). Compare: pre-scientific understanding of what "water" is.
  3. A "counterfactual of freedom": (1) In circumstances C, if Charles were offered a bowl of chili and if he were to respond freely, he would accept it.
According to the compatibilist/soft-determinist, (1) is either necessarily true or necessarily false. Hence, God cannot affect its truth-value.

B. The objective-probability indeterminist view. Like the soft determinist, the OP indeterminist believes that what makes something a free choice is how the action was caused. Unlike the soft determinist, the OP indeterminist thinks that free choice and predetermination (determination with objective probability of 1) are incompatible: there must be a range of possible choices, each having a non-zero probability of being caused.

  1. An OP indeterminst account of freedom: Agent x's action A in circumstances C is free iff x was indeterministically caused to do A in circumstances C (x's act of A-ing was the result of an objective probability in C of strictly less than 1) by a process of type I.
  2. Consider again the counterfactual of freedom (1): (1) had a certain objective probability p (p < 1), given the nature of process I and of the circumstances C. That fact that (1) has this objective probability is itself a necessary truth, even though (1) is only contingently true or contingently false.
  3. Two possibilities:
  4. God did not bring about the actual truth-value of (1): He simply left it up to chance -- with probabilities p and (1-p) -- whether (1) would be true or false.
  5. God did bring about the actual truth-value of (1), but He did so in a way that somehow preserved the objective probability p that (1) should turn out true.
C. The Molinist view. The Molinist adopts a contra-causal or agent-causal theory of freedom: what one freely does is not in any sense caused (either deterministically or indeterministically) by one's prior state plus one's circumstances. However, God knows with certainty the actual, contingent truth-value of all counterfactuals of freedom, including (1). God has, however, no control whatsoever over the truth-value of these counterfactuals.

D. The Anti-Molinist libertarian. Exactly like the Molinist, except the anti-Molinist libertarian holds that counterfactuals like (1) have no truth value (or, are always false) -- so God cannot know them (no Middle Knowledge). God has no control overy how free creatures use their freedom and cannot even anticipate how they would use that freedom under various scenarios.

III. Four Corresponding Versions of the Free Will Defense

Crucial questions: could God have actualized a world containing freedom but no moral evil? Could He have actualized a world with as much freedom as the actual one but with fewer or lesser moral evils?

Could it be that every possible free creature suffers from trans-world depravity: the condition of being a creature that would, in any possible world, perform at least one evil action?

A. From the soft-determinist perspective. It is quite possible that every creature would, if free, perform at least one evil action. It might be that the nature of processes of type F is such that no creature could have some of its actions caused by a process of type F without at least one of those actions being morally evil. We don't know enough about type F to rule out this possibility. So, contrary to what Kelly Clark says, Plantinga's free will defense does not depend on a contra-causal theory of freedom.

B. From the OP-indeterminist perspective. Here it seems clear that no creature suffers from trans-world depravity. However, it might be that every possible creature suffers from a trans-world propensity to depravity: being a creature such that the objective probability that it would, in any possible world, perform at least one evil action is very high.

This would mean that God could not actualize a world containing free creatures without running a high risk of moral evil. There is nothing God could do to lower arbitrarily this risk factor, since if He did so, He would thereby eliminate the freedom of the creatures involved. Hence, the moral evil occurring in the actual world is to be expected, and we cannot hold God responsible for it.

C. From the Molinist perspective. Again, as Plantinga and Clark argues, it is possible that every creature suffers in fact from trans-world depravity, so, although there are possible worlds where creatures are free but commit no moral evil, these were not feasible worlds for God to actualize, since the truth-values of the relevant counterfactuals of freedom were not under His control.

D. From the anti-Molinist libertarian perspective. In this case, the very question of trans-world depravity makes no sense. However, when God actualized a world-segment containing free creatures, He lost some of His control over the flow of events. Hence, He cannot be held responsible for the moral evils resulting.

IV. How is this Relevant to the Argument from Evil?

A. The soft-determinist has reason to reject premise 4: that the best possible world contains no evil. Any possible world containing no evil also contains no free actions. The value of existence of freedom outweighs and absorbs the lesser negative value of moral evil

B. The other three positions have reason to reject premise 5: that God could (with certainty) actualize any possible world.

  1. According to the Molinist, there are possible worlds God knows He cannot actualize, because He knows that the free creatures involved would not cooperate appropriately.
  2. According to the anti-Molinist libertarian, God could have tried but failed to actualize a better world, God's failure again being due to the uncooperativeness of creatures.
  3. According to the OP-indeterminist, there are two possibiliites. God might have tried and failed to actualize a better world, the failure being due to the uncooperativeness of chance (i.e., to bad fortune). Alternatively, it may be that God actualized exactly the world He intended, but He was unable to actualize a better one, since this would have required His raising the objective probability of morally good actions, and this He could not do without destroying the significant freedom of His creatures.

V. The Explanation of Natural Evil

A. Plantinga suggests that all natural evil might be the result of morally evil actions by superhuman creatures (Satan, fallen angels).

B. A response: this hypothesis seems extremely unlikely, so we can still draw the conclusion that God's existence is unlikely, given the existence of evil

Plantinga has two replies:

  1. What makes the hypothesis unlikely? The mere existence of natural causes of natural evils does not exclude the possibility that the natural evils also have an ultimate, supernatural cause.
  2. Plantinga asserts that, since belief in God is properly basic, it can have a probability of 1 for the rational believer. Something that has a probability of 1 cannot be made unlikely by additional evidence, so the problem of evil has no affect on the probability of God's existence.
C. Another problem: Plantinga's hypothesis does not explain why God made us vulnerable in this way to the free choices of superhuman creatures.

Similarly, Plantinga's free will defense doesn't explain why God made us as vulnerable as we are to the actions of other humans.

Some explanation in terms of the overriding and aborbing value of this vulnerability would seem to be necessary.