It is no secret to most that I hold strongly to the Reformed doctrines of grace. But it is equally no secret that I have deep respect for the godly character and scholarship of many of the Arminian persuasion that believe differently than I. The issues that unite us a greater and more substantial than those that divide us. In other words, the Calvinism/Arminianism divide is over non-essential issues in my opinion. What I am saying is that this article is in no way meant for to put an essential line of demarcation concerning the issues of Calvinism and Arminianism. However, just because something is not essential does not mean it is not important. Therefore, I continue to write on these about such.
Yesterday, I wrote that I believe that the doctrine of Prevenient grace is the Achilles heel of Arminianism, Catholicism, and Eastern Orthodoxy (although, less so with Eastern Orthodoxy and Catholicism since they don’t have such a strong stance on depravity). Prevenient grace literally means “grace that comes before.” Prevenient grace is the Arminian counter to the Calvinistic doctrine of Irresistible grace.
It is important to note at the outset that both Calvinists and Arminians believe that people are born sinful. To make this a little more clear, both sides agree that all people are born with an inherent disposition of antagonism toward God. Both Calvinists and Arminians reject what is know as Pelagianism. Pelagius, a fifth-century British monk, taught that people are born neutral, neither good nor bad. Pelagius believed that people sin as a result of example, not nature. Augustine, the primary opponent of Pelagius, responded by teaching that people are not born neutral, but with a corrupted nature. People sin because it is in their nature to sin; they are predisposed to sin. Both Calvinists and Arminians agree with Augustine believing the Scriptures to teach that people are born with a totally (radically) corrupt spiritual nature, making their disposition toward God perpetually antagonistic. Therefore, according to both sides, people are absolutely helpless without God’s gracious, undeserved intervention. This is an important mischaracterization of Arminian theology that adherents to my position often fail to realize. Arminians believe in the doctrine of total depravity just as strongly as Calvinists. In contrast, Eastern Orthodox and Roman Catholics both hold out some sense of natural ability. Therefore, they don’t believe that the will is as depraved as traditional Protestants.
This adherence to total depravity makes the Arminian doctrine of Prevenient grace necessary. A former Wesleyan theology professor of mine who believed in Prevenient grace once called it the “ingenious doctrine.” Why? Because according to Arminians it allows them to hold to the biblical position of total depravity, yet also allow true free will. You see, according to Calvinists such as myself, if people are in such desperate condition, being inclined toward enmity with God from birth, and unable to change their condition on their own (as a leopard cannot change its spots – Jer. 13:23), having no “free will” to choose against this depraved nature, then the only way to answer the question, How is anyone saved? is to answer that the will of God saves them. In other words, if our will could not change our disposition, then God must have changed our will. Up to this point, both Calvinists and Arminians agree. But the Calvinist will say that God’s intervention is radical. In our depraved state, God comes into our lives and opens our eyes to His beauty. This intervention happens by means of saving or “irresistible” grace. In our helpless and antagonistic position, while shaking our fists at God, God sovereignly and autonomously regenerates us. Once regenerated, we trust and love the Lord because our nature has been transformed by Him. Therefore, God is the only one to credit for our salvation seeing as how we did not play any part in its genesis (this is sometimes referred to as monergism). But, according to Calvinists, God does not give this gift of saving grace to all people, only the elect. Otherwise, all would be saved.
How do Arminians deal with our depraved condition? Well, they reject the Calvinistic doctrine of “irresistible” grace believing that it does violence to the necessary freedom that must exist for God to have a true loving relationship with man. But something, nevertheless, must make belief possible. In comes Prevenient grace. This is an enabling grace that comes to the aid of all people so that their disposition can be made capable of receiving the Gospel. It does not save them as the Calvinist doctrine of irresistible grace, but it makes the savable. In essence, Prevenient grace restores people to a state of ability. As Adam before the fall was not predisposed toward a willful rejection of God, being able to make a true free will decision, so people, once affected by Prevenient grace are brought dispositionally to Garden of Eden type conditions. God’s grace comes to the aid of all fallen sinners restoring freedom of the will. Now, it is up to the individual to make an unbiased untainted choice for or against God. Voila! With the doctrine of Prevenient grace, total depravity and true freedom can be harmonized. Ingenious, right?
I agree with Calvinist commentator and theologian Tom Schreiner that “Prevenient grace is attractive because it solves so many problems [for the Arminian] . . .” but I also believe that it creates more problems than it solves. I am going to briefly list the two major problems that I see with the doctrine of Prevenient grace, but I, as always, want to remind you that there are many great men in the history of the church and in contemporary Evangelical philosophy and theology that do not see things the way I do. I encourage you to seek out their position from them in addition to reading my analysis.
1. Lack of Scripture: The biggest issue that Calvinists have traditionally had with the doctrine of Prevenient grace is its lack of biblical support. Tom Schreiner’s quote above is incomplete; it concludes with this, “. . . but it should be rejected because it cannot be exegetically vindicated.” While Prevenient grace may solve problems and allow Arminians to hold to a biblical understanding of depravity, the biblical support for the doctrine is very difficult to find. Most Arminians would agree that direct and explicit support from Scripture is not there, but they would say that the concept is necessitated from other explicit teachings. Most importantly, God commands and desires that all people are to repent of their sin (Acts 17:30, 2 Pet. 3:9, et al) and holds them responsible if they do not. This assumes that “all people” have this ability, otherwise God’s desire is hopeless and His command is useless. While there may be some mystery in the fact that God desires the salvation of all and commands all to repent, this does not necessitate nor justify, in my opinion, the insertion of a fairy complected and even more mysterious doctrine of Prevenient grace. In other words, it could be conceded that God commands all people to repent because sin is at issue. People have violated God’s law. This necessitates God to act as God in accordance with His righteous character and reveal the violation of sin, even to those who have no ability to change on their own. In this case, God’s command is true and genuine. Even if no one were to respond, their sin is made manifest and God’s righteousness is exposed through God’s command. It can also be conceded that God does truly desire the repentance of all people, even if people do not have the ability to repent. God’s desire in this case is mysteriously not going to be an active agent in bringing about the salvation of some. Why? I don’t know. But my ignorance in this matter does not justify the implication of Prevenient grace. God can passively desire things that He does not actively will to come about.
2. It does not really solve any problems: Lets assume that we could overcome the difficulties of the lack of Scriptural support of Prevenient grace. Let’s say that I give the Arminians the benefit of the doubt and say that it is possible to interpret the biblical data in such a way that all people receive an enablement that neutralizes their antagonistic disposition toward God. God then would come to each person sometime in their lives and graciously restore their will to the point that they don’t have any predisposed inclination toward rejection or acceptance of the Gospel. What would this look like?
First, this “balancing the scales” of the will makes any choice, good or bad, for God or against, impossible. Why? Because each person would be suspended in a state of perpetual indecisiveness. They would have no reason for choosing A rather than B. Even Arminian theologian Roger Olson admitted to this in a recent post: “One thing I wrestle with about Arminianism is the mystery of free will. I don’t know how it works. There does seem to be an element of uncaused effect in it” (source). If there is no reason to choose one over the other, then all choices, if they were made, would be completely arbitrary (“uncaused effect”).
You see, we make choices according to who we are. If “free will” of the Arminian variety is going to be responsible for making the choice, and this will is neutralized by Prevenient grace, then there is nothing compelling you (character, upbringing, disposition, the Holy Spirit, genetics, etc.) to make any decision whatsoever. Who you are, the primary factor behind every choice, is taken away. There is no “you” to make the choice. It is arbitrary. It does not solve the “loving relationship”problem to say that God is pleased to have a relationship based upon the arbitrary decisions of people. Therefore, in order to hold to the doctrine of Prevenient grace, one is left with either perpetual indecisiveness or an arbitrary choice. Neither of which solves any problems.
Not only this, but lets do the math. Prevenient grace neutralizes the will, making the will completely unbiased toward good or evil. Therefore, this restored “free will” has a fifty-fifty shot of making the right choice. Right? This must be. The scales are completely balanced once God’s Prevenient grace has come upon a person. What would you expect to see if this were the case? Well, I can flip a coin and pretty much expect that the coin would land on heads just as many times as tails. The same should be the case with salvation. You should expect that just as many people to trust the Lord as those that don’t. But just a cursory look through Scripture tells us that this is not the case. For the most part the number of unbelievers has been dramatically higher than that of believers. Take the time of the flood for instance. How is it that out of millions of people (probably much more), there was only one who was found to be righteous? That would be like me flipping a coin a million (or more) times and it landing on tails 999,999 times and only landing on heads once. Impossible. Christ even explicitly said that there will be and always have been many more people who don’t believe than those that do (Matt. 7:14). How can this be if Prevenient grace created a situation of equal opportunity for all people? It can’t.
Now I don’t want to be accused of building a straw man here so I will attempt to represent how Arminians would respond to this. They would say that the contributing factors that influence people’s freedom are those in the outside world. As the snake came from the outside and influenced Adam’s otherwise neutral will, so also outside influences such as culture and family influence people’s will. Therefore, in the time of Noah, the reason why there was only one righteous person on the earth is because the culture had become so corrupt that God could not be found. This is why God destroyed everyone with the flood. This makes some sense, but in reality it simply re-introduces the same problem that Arminians are desperately attempting to avoid – divine unconditional election.
Let me explain. If outside influences play such a large role in influencing Prevenient-grace-restored-people in their choice for or against God, doesn’t that make God the determining factor in whether they are saved or not? If you had a choice, knowing that outside influences were going to play such a big role in the decisions you make, would you want to be born to a family of believers who teach and live the Gospel in a culture of believers that do the same, or would you rather be placed in a committed Muslim home in a Muslim country where the Gospel is unable to give a testimony of God? In other words, would you rather be placed in a Garden with the snake or without the snake? Of course you would say you want to be placed in the environment where the outside influences for belief in God would be most prominently exemplified. Why? Because you have a better chance. Maybe the odds are not perfect, but they would still be much better. Let’s face it, if you were in the preflood world at the time of Noah, as nice a person as you are today, I seriously doubt that you would have followed Noah rather than the rest of the world.
The problem is that you do not decide where you live or when you will be born. You do not determine your outside influences, God does.
Acts 17:26 26 And He made from one man every nation of mankind to live on all the face of the earth, having determined their appointed times and the boundaries of their habitation.
This passage tells us that God determines the outside influences that are the ultimate influence, the determining factor, in our choice. God chose where you would be born, when, and to what family you would belong. Therefore, God’s sovereign unconditional choice is still the ultimate and determining cause in our salvation. This is the very problem that Arminians seek to avoid with the doctrine of Prevenient grace.
If Arminians were to respond by saying that God gives more grace to those in the most depraved conditions, this would not explain why it is that people in cultures and families that are godly have a higher percentage of believers. We are back to flipping the coin. It does not work either way.
In conclusion, I don’t believe that there is a reason for to entertain the doctrine of Prevenient grace outside of a presupposed view of what some believe must be in order for the truth to be palatable. More importantly, since it really does not solve any problems, it is, in my opinion, superfluous and confusing. Even if it may seem more palatable to say that all people have equal opportunity to accept the Gospel, the palatability of a doctrine does not determine its veracity. This is why I reject the doctrine of Prevenient Grace.
Whether you agree with me or not, I hope that I have been able to give you an appreciation of why Calvinists such as myself have issues with the libertarian freedom inducted by Prevenient grace.
360 replies to "Why I Reject the Arminian Doctrine of Prevenient Grace"
[continued from above]
You may say, “No, it’s not possible to know a CCF without determinism!” but it’s easy to give a counterexample. Remember, I know that if a defenseless person right next to me were to Spill a glass of water on Mt. Rushmore, I would not Murder that person just for doing that, i.e. S[]->~M. Because of my free will, I know I could do so if it were to happen, i.e. S[]->{}M; just as I know that when the clock strikes Ten p.m. tomorrow, I could Kill myself but I know I will not do so (i.e. T->~K and T->{}K). While S[]->~M and S{}->M are inconsistent, S[]->~M and S[]->{}M do not appear to be for the same sort of reason that T->~K and T->{}K are not. The idea that p []-> q entails p []-> ~{}~q is just as modally fallacious as p -> q entailing p -> ~{}~q (as if by arguing “…if it were true then it couldn’t possibly be false…”). Knowing the consequent of a would-counterfactual does not entail that the consequent could not have been otherwise if the antecedent were actualized.
So here we have a counterexample to the claim that “(S&C)[]->A entails (S&C)[]->~{}~A.” Your argument appears to make a fallacy of modal logic.
We’re using the word “know”, which has some difficult philosophical ambiguities; I wonder if we’re running into one. Clearly you’re accepting a 0.0…01% probability of error as acceptable; but a probability of error in a binary choice doesn’t mean that your opinion of the future has a slight error that might be within tolerance. Rather it means that something ENTIRELY DIFFERENT than you believed might happen, and whether or not that happens is completely out of your control. You really shouldn’t call that “knowledge”; it’s actually just belief.
So no, you are not justified in “accepting the true belief” [sic] that B will be the consequence of A. You are free to believe that the risk of ~B is low, but you’re not free to believe that it’s zero.
Even if the part you quoted were irrelevant (arguable!), the part you didn’t quote directly contradicted what you just said. Let me say the same thing it another way, though. What you’re doing by timetravel is actualizing for yourself the person and the circumstances you want data about, and allowing the person to make their own choice. This is not the same thing as being able to predict the outcome without the actual person being available, which is what middle knowledge does.
-Wm
I never used “could not” (I think that’s what ~{} means, right?). On the contrary, I find it a baffling irrelevancy, because it’s throwing in completely unnecessary modalities into an argument that’s already complicated enough.
OTOH, I do assume that (S&C)[]->A implies that ~((S&C) {}-> ~A) — in English, I intended to say that asserting a CCF implies the denial that S “might do otherwise” in those circumstances. This is not me bringing in an extra modality; it’s simply a law of modal logic that asserting a “would” produces a denial of a “might not”.
-Wm
It does, by 2.
This is true, but it is not analogous. As I keep saying: If you accept that God’s knowledge of the future is unmediated/immediate, then it follows from #2 that God does not “predict” future human choices, but rather simply experiences them. He knows about them because they’re real. Molinism accepts that this is God’s “free knowledge”. But this is not available prior to the creative decree.
#2 is fairly general — I intended it to fit a number of fields, including the mathematics of “deterministic systems” and the study of “deterministic algorithms” (Wikipedia has articles on both, but I dunno if they’ll be helpful). So a “process input” is “whatever the process asks for as part of how you start it”. In this case, we’re talking about the “S” and “C” of the CCFs: if God sets a C, and plonks down a S into it, the outcome will always happen, and will always be A.
I’m not sure what you mean here. Note, though, that “inputs” is being used for two totally different purposes, so that might be a problem. “Process inputs” means “inputs to a process”, while “God’s doxastic inputs” means “what God knows”.
In this case, the “process” is an unknown thing called “S’s free choice in C” whose outcome is described by a CCF. The process inputs are S and C; vary those and God gets different outcomes, hold…
(Interrupted!) … The process inputs are S and C; vary those and God gets different outcomes, hold them constant and God gets the same outcome.
Unfortunately, nobody is willing to define what “free will” is, so I’m forced to simply observe how people describe it. Watching the Molinists, it’s very clear that for them, free will is a deterministic process — a thing whose outcome can in principle be predicted from a knowledge of the initial conditions.
-Wm
I’m not sure whether it’s possible to know a CCF. I do know that whatever a CCF is, if it’s true at all it describes a deterministic process.
Same as before: you know those things not as CCFs, but rather as facts about your personal character. You know that you “wouldn’t do that” because you know that you’re not the kind of person to do that. Your knowledge expressed as a would-statement is grounded in a real human being.
That’s no part of my argument.
That is not my claim.
-Wm
I’m not sure that I’ve mentioned before… I’m making a distinction between “prediction” and “knowing” that I think is important. My definition of prediction is given in point #3 of my argument above. By this definition, according to Molinism, God doesn’t predict the future. We might say that God designed the future, so His knowledge flows from His “decree” (a word that’s normally used by Calvinists, but which is appropriate here); or, if we accept that the future is real (theory-B, IIRC) we would say that God simply knows the future.
This is just another way to argue that a man cannot have faith unless he is first regenerated. It is a big smoke screen to keep men from the truth. However…
When we understand what Biblical regeneration is, then this topic becomes irrelevant. Biblical regeneration occurs when God sends Jesus Christ into the heart of a man (Eph. 3:17) to give him the gift of eternal life (1John 5:11, 12). This is the act of God that saves a man—we are saved by His life (Rom. 5:10), saved by regeneration (Titus 3:5), made alive with Jesus Christ by grace we have been saved (Eph. 2:5). No man has life if he doesn’t have the Lord (1John 5:12). And if a man does not have Jesus Christ, he is not one of His (Rom. 8:9). Before a man is given new life in regeneration, he is made to die in conjunction with Jesus Christ via baptism (Rom. 6). A man is crucified with Christ before Christ lives in him. The new life he is given in regeneration is eternal life when Jesus Christ comes to live in his heart, and when he is made alive with Jesus Christ he is saved.
When did men begin to be baptized by Jesus Christ? When did men begin to be indwelt by Jesus Christ? Isn’t it interesting that Peter tells us that men were born again through the resurrection of Jesus Christ from the dead (1Peter 1:3). Could men be born again through the resurrection of Jesus Christ before Jesus Christ rose from the dead?
cont…
Romans Chapter 5 gives us some very important details. Note the end of the chapter—“grace reigns through righteousness into eternal life” (Rom. 5:21). Note eternal life, remember that is the indwelling of God’s Son. Romans 5 tells us about many things that happen in a man before Jesus comes to dwell in him. Everyman is born a sinner (Rom. 5:8, 19), ungodly (5:6), under judgment, condemned to die (Rom. 5:16, 18). A man is reconciled to God through the death of His Son (Rom. 5:10) via the baptism into Christ’s body, into Christ’s death. He goes from many transgressions to justification (5:16). He was justified in His blood (Rom. 5:9) when the blood of Jesus Christ cleansed him from all unrighteousness (1John 1:7). Justified by faith he has peace with God (Rom. 5:1). He went from being an enemy to being reconciled to God. A man is first cleansed and at peace with God before Jesus comes to live inside of him. Therefore he was introduced by faith into the grace (favor) in which he stands (Rom. 5:2). He receives a gift of righteousness (5:17), an acquittal (5:18), made righteous (5:19) –grace reigns through righteousness into eternal life. Grace came by Jesus Christ (John 1:14). Jesus Christ is the gift of grace. Romans 5:15 literally says, “much more, the grace of God and the gift in grace, the one man Jesus Christ into the many abounds”.
In Reformed Doctrine regeneration is the first grace (irresistible) and without it no man can have faith. In Reformed Doctrine you must have regeneration before faith. If faith comes before regeneration then there is no Reformed Doctrine of predestination. In the Bible however, men had faith long before men began to be regenerated. After God began to regenerate men, faith still preceded regeneration (faith precedes Christ dwelling in hearts, and faith precedes the eternal life and salvation given in regeneration). Faith introduces us to grace (Rom. 5:2), not grace introducing us to faith.
I’m not sure why the [sic] is there, but here you seem to be suggesting that knowledge requires certainty of the belief being true. But is that really the case? I do not know for certain that the Holocaust occurred; it is not an analytic truth like “all bachelors are unmarried.” Nonetheless, that the Holocaust happened is very, very likely. Should we say we don’t know it happened simply because of the 0.0…1% that it didn’t?
Which (again) is irrelevant to my point; I wasn’t talking about middle knowledge here. My point: what I described is at least a logically possible way for one to know a future libertarian choice via doxastic inputs despite the choice being indeterministic.
Well, this is what you seemed to be arguing. You say that Molinism specifies that God uses information about a pair of inputs (person S, circumstance C) to get information about the output (action A) and seem to say that God knowing (S&C)[]->A somehow makes the process deterministic, which would imply that (S&C)[]->A entails (S&C)[]->~{}~A, whereas such entailment is fallacious.
Unfortunately, in your argument you didn’t explain which lines followed from which, and this contributed to (1) your argument appearing invalid; (2) your line of reasoning seeming a little hazy.
8 doesn’t follow from 7 and 2 either. To illustrate:
(2) Determinism means that predictions can be made about a process using only information about its inputs.
(7) It’s possible in principle to construct a model to predict human choice under Molinism [exactly as though human choice were a deterministic process].
(8) Therefore, human choice under Molinism is a deterministic process and therefore not compatible with LFW (supposedly follows form 2 and 7).
Unfortunately you still haven’t explained what inputs line (2) refers to. If it’s simply the initial conditions of the system, then I’ll grant for sake of argument that line (2) is true (assuming predictions are “that which is known with certainty”). But even the combination of (2) and (7) doesn’t yield (8). One reason is that God’s doxastic inputs aren’t limited to just initial conditions. Another is a clear counterexample of knowing a CCF where one’s choice isn’t deterministic.
It has similarities to a deterministic process, but it isn’t quite the same thing, since the outcome isn’t a necessary result of previous conditions. It’s easy to construct a counterexample where a CCF is known (i.e. it’s true and one is properly justified in believing it to be true) and yet the choice is not deterministic.
Are the two inconsistent? Can’t knowledge of the CCF arise from knowledge of personal character? More to the point, doesn’t my counterexample give an instance where a CCF is true and one is properly justified in believing it to be true? Isn’t it the case that “If a defenseless person were to spill a glass of water on Mount Rushmore, I would not murder that person just because of that” is true and that I am properly justified in believing it to be true?
No. To know something, that thing must actually be true — it must correspond to reality. (The rest of the definition of “know” gets very complicated; I typically use knowledge == “justified true belief”, but it’s complicated.) With God it’s simple; He knows all knowable things. But something that’s only probable is not knowable.
No. It’s not irrelevant. Please read what I wrote. You’re gathering data FROM THE ACTUAL PERSON. Not from information about the person. I don’t know why you started using the phrase “doxastic inputs” (are you confused by my use of the term “process inputs”, which is a completely different concept?), but the vital concept is that anything that can be completely modeled using only information is a deterministic process (I’m being terse and possibly incomplete — see my argument).
Furthermore, even if I completely ignore that you’re getting your data by “cheating” and peeking, you’re still making the completely unwarranted assumption that the person will make the same choice the same way “next time” history rolls around — so your assumption that you’re gathering useful data is unsupported.
-Wm
The similarities are proven, and are actually part of both mathematical and computer science definitions of “deterministic process”. The differences are entirely hypothetical, based on unsubstantiated claims. And worse yet, what you forgot to mention is that the “proper justification” for believing the thing is your direct experience with your own character and choices. Prior to creation nobody has direct experience with any humans.
I said: “Same as before: you know those things not as CCFs, but rather as facts about your personal character.”
If mere data (about the character or about anything) is sufficient to determine ALL CHOICES, then choices are deterministic processes. If it’s only enough for some choices, Open Theism is right, and Molinism is wrong.
More…
-Wm
Well, let’s say that if we ever hold a RTM meet’n’greet at Mt. Rushmore I’ll be sure to stay away from glasses of water. 😉
But honestly, this is harder to answer than I’ve been indicating. It’s not because it’s obviously right; it’s more because it’s so obviously a misfit, not at all similar to the cases under discussion. In addition to the serious philosophical problems I mentioned (why it’s not parallel to the case under discussion), there are obvious reasons why this example *appears* to be so likely true.
1. The conditions (C) are (deliberately) immensely narrow (location, stimulus, and even your internal mental state).
2. The outcome (~A) is a negative rather than a positive, meaning that it’s infinitely broad (“I might do anything, aside from murdering that person”).
You’ve chosen an example where I, a person who knows close to nothing about you, feels willing to agree that you won’t do this (under that specific mental state). Doesn’t that trouble you? Don’t you expect a tiny bit more from Molinism than an example so utterly trivial that it takes NO data whatsoever to solve?
At the same time, since I know nothing about you, I have to admit that I might be wrong. We know that humans DO sometimes crack for non-reasons, and the odds of even this one being false is nonzero — so I can’t simply say that I know it; I can say that I’d trust it without giving it a second thought, but I also trust my own driving, so now you know what my trust is worth.
-Wm
Agreed.
Disagree. As I’ve said, it’s not “God knowing” that makes the process deterministic. It’s the “fact” (under Molinism) that God can model you as a process with predictable outputs that proves that you are deterministic (under Molinism).
Of course, if I used this logic I would be making a fallacy. But I didn’t draw the conclusion that we “could not do otherwise because God knows that we would do this.” You’re assuming that the only way to prove that something is deterministic is to show that its outcome is necessary, and you already know that Molinism doesn’t have a detailed enough model to make any claims about necessity. Unfortunately for Molinism, mathematics has made advances since Molina’s day; the rules of philosophy still apply, but additional and more specific rules can apply where there’s special cases.
And Molinism provides exactly the special case we need, by claiming that ALL human action can be reduced to a lookup table.
-Wm
WM,
“And Molinism provides exactly the special case we need, by claiming that ALL human action can be reduced to a lookup table”
Said it before will say it again. No Molinist of which I am aware claims what you assert they claim. Now it may be that Molinism logically necessitates this, but you haven’t proven that. Every Molinist of which I am aware would say that God knows what person X will do in situation Y in Universe Z, but at the same time maintain that the choice in question was still free and indeterministic. No offense, but I have yet to see you make a convincing argument proving that this is logically incoherent.
Michael, I admit that the sentence you quoted is perhaps a bit of an exaggeration. The lookup table needed would presumably be infinitely large, making it NOT comparable to a normal data structure.
Nonetheless, I have posted an argument, and I’ll next expand and annotate it (Wade was right to complain that it had no attempted annotation). I don’t know why you’re worrying about persuasiveness; it would seem that validity is more of a pressing concern :-). I’d like to draw your attention, however, to the enormously complex claim Molinists make without any argument or evidence whatsoever. The challenge of the Open Theists (that would-counterfactuals destroy actual LFW) should have been taken seriously rather than being brushed off by Dr Craig because they’re unpopular (?).
Really, look! http://www.reasonablefaith.org/site/News2?id=6803&page=NewsArticle
His first objection is that Boyd hasn’t captured many followers. His second point is that “if mights exist, why shouldn’t woulds” — but by saying that he completely ignores Boyd’s _reasons_ for proposing this. The question isn’t whether counterfactuals “exist”, but rather whether they can be known to be _true_ along with LFW. And “might” can obviously be known in the presence of the ability to do otherwise, while “would” is questionable. Apparently Craig is content with the mere logical possibility of universes in which one does otherwise; Craig refers to that as “the ability to do otherwise”, even though this “ability” need not attach to the actual person making the choice or be available in the actual universe.
I simply don’t see how Craig’s argument leaves him with any objection to Calvinism, which after all does the exact same thing — it’s logically possible that we could have done otherwise than follow God’s predestinate plan in this universe (because it’s logically and metaphysically possible that God could have willed otherwise).
-Wm
I’ve rephrased some of my points to allow me to more clearly cite sources (and to use terms you’ve used). The intent is the same, but if the meaning has changed I’ll stand by this one in place of the previous one.
1. LFW means determinism is not true of at least some choices made by humans. (axiom: given)
2. A deterministic system is a system for which a model could exist. (axiom: definition of “deterministic system”, Wikipedia)
3. A model could exist if, given information about the initial conditions for an event, the outcome can in principle be perfectly predicted even in the absence of the system. (axiom: definition, Wikipedia “deterministic system”, and the link to “model” if you want details)
4. Molinism specifies that God uses information about the complete initial conditions for a “free choice” (person S, circumstance C) to perfectly predict a single outcome (action A). (definition: “molinism”, Craig)
4a. God’s prediction occurs prior to His creation. (definition: “molinism”)
4b. Therefore, God’s prediction requires only information, not any existing free creatures. (from 4a)
5. Molinism then claims God chains the results of these together to perfectly model all “feasible worlds” (Craig’s term), chooses the most satisfactory one for His purposes, then creates the first link in the timeline, and then observes everything turn out exactly as specified, proving the perfection of the predictions. (definition: “molinism”, Craig)
—
6a. Therefore, it’s possible in principle that a model would perfectly predict each human choice under Molinism. (from 3, 4, 4b)
6b. Therefore, the same model would perfectly predict every human choice that occurs in the actual world. (from 5, 6a)
7. Therefore, every human choice under Molinism is a deterministic system. (from 2, 5, and 6)
8. Therefore, Molinism is not compatible with LFW. (from 1 and 7)
I apologize for the sublettering scheme, but I felt it was right to break down the compound…
I said: “You’ve chosen an example where I, a person who knows close to nothing about you, feels willing to agree that you won’t do this (under that specific mental state).”
I was thinking about this some more. Not only are the preconditions narrow and the outcome broad, but the implicit assumptions also constrain the probability of the statement being true. When you specify “on Mt. Rushmore” as the location, this appears to be an irrelevant addition; it seems intuitively likely that the probability of the statement is only slightly reduced by this addition. If I then think about the statement WITHOUT that additional specificity, it becomes clear that you’ve probably already been in those circumstances; and the fact that you’re going to church with Michael means you’re probably not posting from criminal psych confinement, as one would expect for a person who would murder only because someone accidentally spills a glass of water in any location.
So I adopt this as my prior probability. so, given the fact that you in fact do not murder when people spill water in any of the places where you’ve been in the past, what is the probability that you would murder on Mt. Rushmore, specifically?
The answer is: so low that I can safely accept the risk (especially when I’m not at Mt. Rushmore with you!).
But all of this is inherent in the structure of the question, not in my own knowledge. I know the upper bound on this probability not because I have a special type of knowledge, but rather because the question itself contains information about its probability. The answer would be exactly the same for any equally precise definition of “S” and “C”.
The CCFs of Middle Knowledge are not this type of question; not only do they contain precise information about the initial state, they also contain precise and unique information about the outcome.
-Wm
(Huh, my argument post got truncated — by my software I was well under 2000 characters. What I said was:)
I apologize for the sublettering scheme, but I felt it was right to break down the compound statements.
-Wm
WM,
1. When I talk about pursuasivness I meant I found your arguments against the logical incoherency of Molinism thus far unpursuasive and in particular your arguments for Molinism logically entailing things Molinists don’t claim.
2. On your latest formulation I don’t see how the simple fact that something can be modeled implies determinism. After all can model Quantum Fluctuations even though they are ultimately indeterminate. Furthermore a Molinist would probably dispute the term “models” as it applies to God. A Molinist would say that God simply knows because he possesses Middle Knowledge in the same way that God knows because he possesses foreknowledge. Molinists do not appear to to posit that God knows X through some intermediate modeling process. He simply knows because by the definition of omniscience he knows all truths.
3. On logical argumentation – if you are going to prove something to be logically incoherent you need to rephrase your argument to remove any equivocation. As long as any statement in your proof is eqivocal it remains possible for something to be logically coherent. I found you’re arguments using “could”, “possible”, etc. very confusing given the context.
I’ve made it very clear that my argument is intended to prove (entail) something that many Molinists (Craig in particular) not only do not claim, but actively deny.
I think that you MEANT to say that my arguments REQUIRE things that Molinists do not claim. This is correct; Molinists do not take positions on the definition of a deterministic system, both because that definition is a settled matter of fact and thus not arguable; and also because they assume that Molinism doesn’t deal with deterministic systems. My intent is to question that assumption based on Molinist’s own models.
Perhaps I should have said “deterministic model”; but my definition as it stands is unambiguous, since it requires a single outcome for a completely specified initial state.
I don’t purport to provide a mechanism for God, nor did I say that God used a mathematical model. I simply stated that the Molinist’s definitions imply the possibility, in principle, for a model to exist.
That’s a great suggestion, and I welcome any pointing out of possible ambiguities and/or equivocation. I’ll try to figure out what you’re referring to; could you attempt to be as specific as possible?
-Wm
Here are the phrases where I use “could” or “possible”.
2. A deterministic system is a system for which a model could exist. (axiom: definition of “deterministic system”, Wikipedia)
3. A model could exist if, given information about the initial conditions for an event, the outcome can in principle be perfectly predicted even in the absence of the system. (axiom: definition, Wikipedia “deterministic system”, and the link to “model” if you want details)
These are not equivocation; the second precisely defines the first.
6a. Therefore, it’s possible in principle that a model would perfectly predict each human choice under Molinism. (from 3, 4, 4b)
The only possible dual meaning this could have actually would weaken my argument, not make it apparently stronger. But I don’t want you to be able to do that, so I’ll rephrase it:
6a*. Therefore, a model could exist for human choice under Molinism. (from 3, 4, 4b)
I’ve rephrased it to precisely match the definition in 3, to which it directly refers.
As a side comment, I don’t mean for “existence” to imply created or uncreated status. I’m happy to leave that unspecified, and simply note that most mathematicians think mathematics exists uncreated, just as Molinists think the CCFs exist uncreated.
-Wm
That’s irrelevant too. Remember, I’m making a rather modest point here: what I described is at least a logically possible way for one to know a future libertarian choice via doxastic inputs despite the choice being indeterministic. The fact that I gathered data from the actual person etc. doesn’t make this point any less true. But since this is almost to the point of a digression by now, I may stop speaking about this further.
Even if true, it’s irrelevant to the point I’m making. Here, my point is modest: there is a real example a CCF (“If a defenseless person were to spill a glass of water on Mount Rushmore, I would not murder that person just because of that”) that is true even though my choice is indeterministic. In any case, you didn’t quite answer my question. Isn’t it the case that “If a defenseless person were to spill a glass of water on Mount Rushmore, I would not murder that person just because of that” is true and that I am properly justified in believing it to be true?
I wouldn’t say it’s unwarranted; time travel entails the B-theory of time; the choice in effect has already been “made” and so ex hypothesi there is no “next time” history rolls around. That said, you could argue that observing it somehow changes the time-stream, but we can add a law of physics in our logically possible scenario such that this sort of thing doesn’t happen.
I agree, but it’s possible for a belief to be both true and probably true. Probability is, after all, a measure of likelihood of the proposition being true; it is not a degree of the proposition’s truth. Hence a statement like, “But something that’s only probable is not knowable” doesn’t appear to make much sense without the “knowledge requires certainty” criterion. Suppose a proposition is (1) true; (2) one believes it is true; (3) one’s justification for accepting the proposition is that (from her evidential base) the probability of it being true is 99.99999%. Isn’t she properly justified in believing the true proposition? If so, isn’t it true that she knows it to be true?
Which sounds very much like saying that (S&C)[]->A entails (S&C)[]~{}~A, and that of course is fallacious. True, under Molinism God knows (S&C)[]->A and in this sense can be said to model a process with predictable outputs: God knows if he put someone S in conditions C, action A would result; this follows necessarily from (S&C)[]->A. But this does not at all entail that (S&C)[]->~{}~A is true. If you do not mean “God can model you as a process with predictable outputs” in the sense that I described, what exactly do you mean?
1 and 2 risk equivocating definitions of “determinism.” Determinism with respect to metaphysics means that effects are necessary results from antecedent causes. What do you mean by “model” in line 2? You didn’t specify, but if it is “given information about the initial conditions for an event, the outcome can in principle be perfectly predicted even in the absence of the system” (you said this entailed a model in 3, but you didn’t say whether this was a necessary condition) then I disagree that God can model human behavior, nor does Molinism entail such a thing can be done. So this premise:
appears to be false. Certainly if I were a molinist I would never agree to this, and you gave no citation for Craig thinking this. True, under molinism God knows (S&C)[]->A, but that doesn’t imply that God’s knowledge of the initial conditions is sufficient for Him to know this. To use an analogy, God knows whether I will drink a root beer tomorrow, but God’s knowledge of the initial conditions is insufficient for him to know what I will freely choose tomorrow; God’s doxastic inputs extend beyond mere initial conditions.
WM,
Couple things and then I’ll sit back and eat popcorn again while I watch you and Wade debate.
“Perhaps I should have said “deterministic model”; but my definition as it stands is unambiguous, since it requires a single outcome for a completely specified initial state.”
1. Can a system which is ultimately indeterminate in the final analysis behave and be modeled in a determinate manner?? I think quantum mechanics undermines your point.
2. Furthermore, I think your point would apply to the general foreknowledge of God in addition to Middle Knowledge which gets us back to the modal fallacy. Think about it – even traditional Arminians hold that God knows the single outcome that will occur given a specific person in a specific circumstance. Yet as we have established to equate this knowledge to determinism is logically fallacious.
You know, I did completely forget to work the results of that discussion into my argument. I’ll do that, and post a new version. Of course, you’ll recall that although the two definitions are not in general equivalent, the “necessary” definition is the more general one, and the “capable of being deterministically modeled” is a more specific one. So I’m proving a stronger result, not a weaker one. (But I still need to make that explicit in the proof.)
I didn’t give a real definition, I admit. I did mention that Wikipedia’s article on “Deterministic system” gives a link to “Model” (the article’s name is “Mathematical model”). But I don’t need a necessary condition for being a model; all I need is a sufficient condition.
(Michael also pointed out that I’m talking about “deterministic models”, rather than models in general. Oops, good catch.)
Molinism does in fact entail that it can be done; and in fact God must effectively model human choice a large number of times prior to creation, in order to view the final outcomes of all feasible worlds (surely a large number!) and select the one with the best outcome.
-Wm
Molinism is quite explicit that it’s not sufficient; God also uses middle-knowledge (and Molinism doesn’t specify where that comes from). But I’m puzzled now as to why you think this affects my argument. Why does that make #4 false? Are you thinking that a model can never use anything more than the raw information contained in the initial conditions given to it? All real models have to be constructed, and that construction alone adds information that isn’t contained in the inputs.
You really like the words “doxastic inputs”, don’t you. 🙂
-Wm
Hours of entertainment — and such a low cost! 🙂
As I mentioned (and am putting into my rework), I should have specified “deterministic model”, not just any model; I know of several nondeterministic models, I’ve worked on several different military ones (although, amusingly, all of my programming was done on one of the only deterministic ones).
Nope, not at all. Point #3: if the system is present, the existence of a model won’t prove anything, since the system may provide corrective feedback to the model.
Agreed. But it’s not a fallacy because it’s impossible that the thing be deterministic; rather, it’s a fallacy because there are other possible ways the things could be associated to produce that valid logical relationship. It’s like the rules of statistics: never confuse correlation with causation. But, at the same time, it’s still possible to prove causation — you just can’t do it with statistics.
-Wm
I would like to say this is all very interesting….but I’m having a really hard time following most of this post….you guys appear to know what your talking about…..but I really don’t know.
Pass me the popcorn and I’m going to try another channel. It’s like in the 60’s when Robin says in reply to one of Batman’s philosphical pronouncements which is way over his head……Holy Cow Batman!!!
A quick comment… I’m probably not going to have time today to edit together a revision. I hope by tomorrow, but who knows.
This one’s a fairly major revision, since Wade’s now clearly provided a definition of LFW that has more detail than “determinism isn’t true about human choices”; I’m going to use the best definition I have. But, of course, incorporating this is going to take more time.
-Wm
In order to understand what you’re saying, I need to know what you mean by “to know via doxastic inputs”. The context for your argument makes it sound like you’re trying to provide a counterexample to my argument, but I never used that phrase.
In order to provide a valid counterexample, you’d have to show how, in the absence of the system itself, it’s possible to know the future of the system using only information about the starting state of the system.
Your example here doesn’t do that work — it allows you to know the future state of the system because it requires that the system actually exist prior to you knowing its future.
-Wm
I see your problem. You’re confusing the fact high probability is adequate justification for believing something, with the incorrect idea that high probability makes something true. Another way to put this is that you’re assuming that epistemic probability is the same as ontological probability.
So I believe that you’re not going to murder someone who spills water on Mt. Rushmore; and I’m justified in that belief; but unless it’s actually true that you don’t DO that, I don’t KNOW that. Even if my belief is VERY justified!
I could modify the statement to make it true: I could say “If someone spills a glass of water on Mt. Rushmore, the probability that you’d kill them for that reason is less than 0.00001%.” Now, even without going to Mt. Rushmore, I could prove that I know this statement: it’s objectively true, I believe it, and my belief is justified.
-Wm
It’s going to be a while before I can post again; perhaps next week. I’m doing renovations on my argument; I stand by my previous version (there’s nothing wrong with it), but I’m happy to add requested features. My general approach is to switch Wade’s old definition from the vague “not deterministic” to a somewhat more specific “outcome isn’t a necessary result of previous conditions” (the philosophical definition of the term, rather than the model theoretical one I’d picked). The result is going to be a lot longer; it’s helped a little by some more specific mathematical theories (dynamical systems).
The biggest help I could receive would be an actual definition of LFW from Wade; you asked for a proper definition of “model” from me, but you won’t give either a necessary or sufficient condition for LFW, so I’ve wasted a lot of time posting arguments that don’t contact yours.
I’d also like you guys to interact with what I’ve said, of course; your comments have been very helpful, and I’d like to see you address my responses. (Some of our discussions are leaving me puzzled — it’s becoming apparent that some of your “counterexamples” might possibly be counterexamples to something other than what I thought they were, and I’d like to understand their target better.)
-Wm
Is anyone here? I’ve got my rework partially completed, but stopped while waiting for feedback on my above questions. There are too many directions I could take this, I really need to know.
Michael, I think your objection’s the best one. The answer is simple — quantum theory has a deterministic wavefunction, but a nondeterministic wave collapse. It’s not a question of being able to model a nondeterministic process with a deterministic one; rather, it’s two distinct processes. I won’t be able to incorporate that into my argument, because it’s actually a side issue; but it’s a good, observant one.
-Wm
While researching for my reworked argument, I found a very interesting model of human decisions, which although it doesn’t deal with my Molinism argument at all, it does impinge on our other discussions, and will allow me to develop some ideas I’ve long had but was unable to justify. I think you’ll like how this goes — basically, this explains how nondeterminism might work within the process of human choice, allowing us to validly suppose that human agency CAN be creative (in the sense of producing things that are not the necessary results of prior events), while still allowing it to work in a “adequately deterministic” manner, in theory allowing the human to be in reasonable control of his own choices.
http://www.informationphilosopher.com/freedom/standard_argument.html summarizes adequately.
This provides a nice counterexample against the hard determinist position, which makes me happy. Of course, I still claim that Molinism is hard determinist.
-Wm
Well, it depends. I had originally understood it to mean that knowledge of antecedent conditions were sufficient knowledge to know the outcome. If that is rejected, then the premise is true, but then #4 being true wouldn’t entail determinism (and thus the argument would appear invalid).
It seems that I a post can only be 1000 characters long (argh) so here we go:
Pretty much what it sounds like; doxastic inputs are those inputs in which beliefs are formed in response to (for us humans at least, this would include e.g. sensory perception).
No, that’s not what I’m doing. I’m not claiming that high probability makes something true.
I apologize if this wasn’t clear before, but one necessary condition is that determinism is false with respect to human volition; that is, the results of human choices are indeterministic, and are not the necessary results of previous causes (they could have been otherwise). There are various doctrines of LFW; one is simple indeterminism (where basically human choice is a random event) and the other is agency theory, where an agent (person, self) brings about events.
Regardless, I look forward to your revamped argument.
Okay, “human choice is not deterministic” and “the results … could have been otherwise”. I still need specifics (as I’ve complained, this is a vague and negative definition), but I’m going to hazard some guesses.
May I assume:
1. that every nondeterministic event is caused rather than uncaused (nor can it be its own cause)? (Justification: Kalaam)
2. that, therefore, when we say “could have been otherwise” we require that the results, if different, would have actually been caused in a different way (possibly by different things, certainly by different amounts of things)? (Justification: in a lawful universe, two effects with the same cause seem to be the same effect)
3. SO THAT saying (negatively) “human will is not deterministic” is the same as saying (positively) “human will includes the ability to determine which of the causes of a decision will produce effects”?
Hm… I like it. No events without a cause; simply effects that never exist.
-Wm
My argument breaks out WHY #4 being true entails determinism. Please address the reasoning to explain where it breaks down; simply finding that #4 is not where it breaks down doesn’t prove that it must break down somewhere else.
Edit: my most recent argument appears to be http://www.reclaimingthemind.org/blog/2010/10/why-i-reject-the-arminian-doctrine-of-prevenient-grace-2/comment-page-7/#comment-46978
-Wm
I haven’t been trying to deny that it’s logically possible; rather, I’ve been questioning the idea that it has anything to do with what we’re discussing.
First, for example, where (prior to creation) does God form beliefs about anything? If God doesn’t form beliefs, then He’s not using doxastic inputs.
Second, what does this have to do with any of my arguments? You bring it as a counterexample, but what is it a counterexample against?
I also raised some questions as to its logical possibility with some of the other premises of our arguments (obviously it’s not logically possible to time-travel prior to the creation of time), but this isn’t your point. What…
…is?
-Wm (wow, the counter’s wrong. This is the end of my previous post, truncated by the comment script.)
There are two separate lines of argument in which I see you making that claim. Let me know where I err.
First, when we’re discussing whether I can “know” that you won’t kill me because I spill a glass of water on Mt. Rushmore, you insist that because there’s less than 0.00000…01% probability that you would do that, therefore I can know that you wouldn’t.
Second, http://www.reclaimingthemind.org/blog/2010/10/why-i-reject-the-arminian-doctrine-of-prevenient-grace-2/comment-page-6/#comment-46664 directly pertaining to my argument, you claim that it’s possible to know things as a result of information even when there remain live contradictions to what you know (so long as they’re very, very low probability).
We’ve been going over this for a while. What points were you actually trying to make?
-Wm
Melani, your position is mainstream, but your arguments are completely untenable. The passages you cite do not support the meanings you hang on them.
Biblical regeneration occurs when God sends Jesus Christ into the heart of a man (Eph. 3:17)
False: Eph 3:16-17 says that Christ’s indwelling is the result of the Holy Spirit’s preceding strengthening work.
to give him the gift of eternal life (1John 5:11, 12).
False: the purpose of Christ’s indwelling is in Eph 3:18-19.
False: the claim in 1Jn 5:11-12 (and the entire section) is regarding the distinction between those who have and those who do not have, not anything about the process of conversion.
This is the act of God that saves a man—we are saved by His life (Rom. 5:10),
False: saved *from God’s wrath*, not saved from our unregeneration.
saved by regeneration (Titus 3:5),
True! But this regeneration is God’s work, not ours.
I’m out of space, and don’t see any point in…