I am journalling through Gordon H. Clark's "God and Evil: The Problem Solved", as carefully and thoughtfully as I can.
The current section is called "Omniscience". In it Clark attempts to show that foreknowledge is not compatible with human freedom. By "foreknowledge" Clark means "omniscience", or more specifically "prescience", and not the traditional understanding of 'relational' foreknowledge said to be meant by Paul's use of the Greek "prognosko". From Clark's prose I can pick out four distinct arguments.
The first argument is written ambiguously. Either Clark is arguing that foreknowledge requires or implies causation of future events, or that it merely precludes the freedom of future events. At first I thought he intended the latter, but just after it he says "Suppose it be granted, just for the moment, that divine foreknowledge... does not cause the foreknown event". This implies that his argument was in fact asserting that foreknowledge causes future events. Additionally, he then goes on to argue that foreknowledge precludes freedom, even if it isn't causal. For these reasons I consider his first argument to be that foreknowledge is causal.
The argument itself goes something like the following. If God knows a future action, then the agent cannot chose otherwise than they will. This means that they have no free will. If free will doesn't exist, then God is the one causally determining everything.
The second argument is that even if foreknowledge isn't itself causal, it is nevertheless incompatible with free will. Clark says that if man has free will, then things can be different. If things can be different, then God cannot know how things will be for sure.
The third argument is that if God is not the one causing every event, then He cannot know what events will take place, and therefore cannot be omniscient. To be fair I cannot tell if this is meant to be a distinct argument from the second one.
The fourth argument is that if God is not the one causing every event, then His knowledge is dependent.
Even though I am anticipating the part where Clark gets to positive arguments for deterministic Calvinism in general, scriptural exegesis in support of Calvinism in particular, and the explanation of how Calvinism solves POE (though it seems to exacerbate it), I was nevertheless excited to see that he was getting into some metaphysics. The excitement quickly faded as I realized that he was only rehearsing old arguments without interacting with their classical rebuttals. While he seems well-versed in the history of Calvinist theology, he seems ignorant of the history of the free will discourse in the philosophical literature.
The flaw in the first argument is the confusion over what the determiner of a future free event is. The determiner of a future free event is the agent who will perform it. The event is thus foreknown because it is determined by the free agent, and not the other way around as the first argument goes. Just as knowledge of the past is not causal and knowledge of current events is not causal, foreknowledge is not causal. It is difficult to understand at first when it comes to foreknowledge because of the time disjunct, but it is nevertheless a mistake to think that any kind of knowledge can be causal.
The flaw in the second and third arguments is the conflation of two senses by which an event "can be different". In one sense, a free event can be different precisely because it is free. Or perhaps more accurately, the ability for an event to be more than one way is what makes it "free". Let's call this principle of alternate possibilities the "freedom sense of PAP".
But in another sense, a free event cannot be different than it is. This can be most easily understood by thinking about free events in the past. A free event that occurred in the past cannot now be different than it occurred. But this does not imply that it was not at the time performed freely, with the real possibility of alternatives. Let's call this principle of alternate possibilities the "matter of fact PAP".
Free events in the future are similar to free events in the past in their inability to be different. But this is merely because they lack "mater of fact PAP". That is, free events will unfold one way as a matter of fact and not another. Put another way, future contingents have a truth value. This is what makes foreknowledge possible: the fact that the future will be a certain way. But free events of all kinds also share their ability to be different. That is, all free events have the "freedom sense of PAP". That's what makes freedom possible.
Conflating the "freedom sense of PAP" with the "matter of fact PAP" is what leads to Clark's confusion. He mistakenly thinks that because future contingents have a truth value, they cannot be freely caused, and the inverse, that if events can be freely caused, then they cannot have a truth value in advance. My good friend Derek von Barandy believes arguments that presuppose foreknowledge alone precludes free will commit "Sleigh's Fallacy", and I am inclined to agree [Update 01/23/10: Derek articulates something similar here, which is a helpful read. Thanks, Derek!]. As I understand it, Sleigh's Fallacy goes something like the following.
1. P is true.
2. Therefore, P is true by necessity.
What should be inferred instead is:
1. P is true.
3. Therefore, P is necessarily true.
The difference lies in the sense of necessity. Given the fact that P is already true, it is now necessarily the case that P is true. But this doesn't mean that P is a necessary truth. True propositions don't bootstrap themselves into necessity.
Asserting that foreknowledge precludes free will makes the same mistake as inferring (2) above from (1), but adds tense.
To avoid this error, we need to focus on what the truth-maker of a given proposition is. The truth-maker of the truth of P is the state of affairs P describes. The truth of P is logically dependent on the truth of P. Similarly, the truth-maker of a future free event is the free agent who will perform it, and then as a logical result, God will foreknow the future free event.
Because God is God, He is able to have knowledge that logically proceeds future events, even though such knowledge chronologically precedes them. One way of making sense out of this is by maintaining that God is outside of time.
Boethius explains all of this in Book V of "The Consolation of Philosophy", especially from section 145 on.
Providentially, William Lane Craig's most recent Q&A addresses this very issue. It is the current question on the main Q&A page, but will eventually be archived as question 124. His explanation of the problem and its resolution are a little bit more philosophically technical than mine, but probably more clear in the long run. He does have 20+ years of PhD+ research under his belt after-all. (Here's his page on divine omniscience again, for reference.)
As for the fourth argument, I wholly concede its conclusion: that God has a type of knowledge which is dependent. This is classically called "scientia media", or "middle knowledge". This is a reference to the three categories of knowledge that God is traditionally regarded as having. The first category of knowledge concerns necessary truths. That is, propositions that cannot fail to be true. These would be things like "2+2=4", and "God exists". The middle category of knowledge concerns contingent truths and counterfactuals that God does not cause to be true. God's middle knowledge is logically dependent, but nevertheless complete and eternal. The last category of divine knowledge is God's "free knowledge", which includes those things that are true because God causes them to be true. (Here's the relevant Wikipedia entry, and here are articles on middle knowledge in the Stanford Encyclopedia of Philosophy, if you're interested.)
It may yet be the case that God deterministically causes everything and that Free Will doesn't exist, but Clark's arguments in this section do not constitute good reasons to think so.
The current section is called "Omniscience". In it Clark attempts to show that foreknowledge is not compatible with human freedom. By "foreknowledge" Clark means "omniscience", or more specifically "prescience", and not the traditional understanding of 'relational' foreknowledge said to be meant by Paul's use of the Greek "prognosko". From Clark's prose I can pick out four distinct arguments.
The first argument is written ambiguously. Either Clark is arguing that foreknowledge requires or implies causation of future events, or that it merely precludes the freedom of future events. At first I thought he intended the latter, but just after it he says "Suppose it be granted, just for the moment, that divine foreknowledge... does not cause the foreknown event". This implies that his argument was in fact asserting that foreknowledge causes future events. Additionally, he then goes on to argue that foreknowledge precludes freedom, even if it isn't causal. For these reasons I consider his first argument to be that foreknowledge is causal.
The argument itself goes something like the following. If God knows a future action, then the agent cannot chose otherwise than they will. This means that they have no free will. If free will doesn't exist, then God is the one causally determining everything.
The second argument is that even if foreknowledge isn't itself causal, it is nevertheless incompatible with free will. Clark says that if man has free will, then things can be different. If things can be different, then God cannot know how things will be for sure.
The third argument is that if God is not the one causing every event, then He cannot know what events will take place, and therefore cannot be omniscient. To be fair I cannot tell if this is meant to be a distinct argument from the second one.
The fourth argument is that if God is not the one causing every event, then His knowledge is dependent.
Even though I am anticipating the part where Clark gets to positive arguments for deterministic Calvinism in general, scriptural exegesis in support of Calvinism in particular, and the explanation of how Calvinism solves POE (though it seems to exacerbate it), I was nevertheless excited to see that he was getting into some metaphysics. The excitement quickly faded as I realized that he was only rehearsing old arguments without interacting with their classical rebuttals. While he seems well-versed in the history of Calvinist theology, he seems ignorant of the history of the free will discourse in the philosophical literature.
The flaw in the first argument is the confusion over what the determiner of a future free event is. The determiner of a future free event is the agent who will perform it. The event is thus foreknown because it is determined by the free agent, and not the other way around as the first argument goes. Just as knowledge of the past is not causal and knowledge of current events is not causal, foreknowledge is not causal. It is difficult to understand at first when it comes to foreknowledge because of the time disjunct, but it is nevertheless a mistake to think that any kind of knowledge can be causal.
The flaw in the second and third arguments is the conflation of two senses by which an event "can be different". In one sense, a free event can be different precisely because it is free. Or perhaps more accurately, the ability for an event to be more than one way is what makes it "free". Let's call this principle of alternate possibilities the "freedom sense of PAP".
But in another sense, a free event cannot be different than it is. This can be most easily understood by thinking about free events in the past. A free event that occurred in the past cannot now be different than it occurred. But this does not imply that it was not at the time performed freely, with the real possibility of alternatives. Let's call this principle of alternate possibilities the "matter of fact PAP".
Free events in the future are similar to free events in the past in their inability to be different. But this is merely because they lack "mater of fact PAP". That is, free events will unfold one way as a matter of fact and not another. Put another way, future contingents have a truth value. This is what makes foreknowledge possible: the fact that the future will be a certain way. But free events of all kinds also share their ability to be different. That is, all free events have the "freedom sense of PAP". That's what makes freedom possible.
Conflating the "freedom sense of PAP" with the "matter of fact PAP" is what leads to Clark's confusion. He mistakenly thinks that because future contingents have a truth value, they cannot be freely caused, and the inverse, that if events can be freely caused, then they cannot have a truth value in advance. My good friend Derek von Barandy believes arguments that presuppose foreknowledge alone precludes free will commit "Sleigh's Fallacy", and I am inclined to agree [Update 01/23/10: Derek articulates something similar here, which is a helpful read. Thanks, Derek!]. As I understand it, Sleigh's Fallacy goes something like the following.
1. P is true.
2. Therefore, P is true by necessity.
What should be inferred instead is:
1. P is true.
3. Therefore, P is necessarily true.
The difference lies in the sense of necessity. Given the fact that P is already true, it is now necessarily the case that P is true. But this doesn't mean that P is a necessary truth. True propositions don't bootstrap themselves into necessity.
Asserting that foreknowledge precludes free will makes the same mistake as inferring (2) above from (1), but adds tense.
To avoid this error, we need to focus on what the truth-maker of a given proposition is. The truth-maker of the truth of P is the state of affairs P describes. The truth of P is logically dependent on the truth of P. Similarly, the truth-maker of a future free event is the free agent who will perform it, and then as a logical result, God will foreknow the future free event.
Because God is God, He is able to have knowledge that logically proceeds future events, even though such knowledge chronologically precedes them. One way of making sense out of this is by maintaining that God is outside of time.
Boethius explains all of this in Book V of "The Consolation of Philosophy", especially from section 145 on.
Providentially, William Lane Craig's most recent Q&A addresses this very issue. It is the current question on the main Q&A page, but will eventually be archived as question 124. His explanation of the problem and its resolution are a little bit more philosophically technical than mine, but probably more clear in the long run. He does have 20+ years of PhD+ research under his belt after-all. (Here's his page on divine omniscience again, for reference.)
As for the fourth argument, I wholly concede its conclusion: that God has a type of knowledge which is dependent. This is classically called "scientia media", or "middle knowledge". This is a reference to the three categories of knowledge that God is traditionally regarded as having. The first category of knowledge concerns necessary truths. That is, propositions that cannot fail to be true. These would be things like "2+2=4", and "God exists". The middle category of knowledge concerns contingent truths and counterfactuals that God does not cause to be true. God's middle knowledge is logically dependent, but nevertheless complete and eternal. The last category of divine knowledge is God's "free knowledge", which includes those things that are true because God causes them to be true. (Here's the relevant Wikipedia entry, and here are articles on middle knowledge in the Stanford Encyclopedia of Philosophy, if you're interested.)
It may yet be the case that God deterministically causes everything and that Free Will doesn't exist, but Clark's arguments in this section do not constitute good reasons to think so.
I'm just getting around to reading all of your posts on this book... I sensed the 'wink' as you typed "providentially" in the third-to-last paragraph above. But then, I don't know what you really mean by that anymore ;-)
ReplyDeleteNevertheless, I'm enjoying reading through these and all the ensuing comments.