I. Calvinism and its discontents
Many theologians and philosophers have serious problems with Calvinism. Typically, Calvinism is presented as a system committed to (some form of) determinism about all of man’s actions. Call this determinism, ‘theological determinism’ (TD). Since Calvinists hold that man is (at least) morally responsible for some of his actions, then it seems that Calvinism is committed to compatibilism. I’ll define ‘compatibilism’ (COMP) as the view that moral responsibility is compatible with TD. That is, there exists a model on which both TD and COMP are true. Call this conjunction, ‘THEOCOMP’.
In our current (philosophical-) theological atmosphere, THEOCOMP is not viewed in a flattering light, to put it mildly. First, it is said that, necessarily, TD rules out moral responsibility; so, COMP is false (which entails THEOCOMP is false). Second, it is said that even if we grant COMP, the conjunction of TD with the thesis that there exists evil in the world entails the conclusion that God is the author of sin—which is supposed to be an anathema. Indeed, such doctrines inspired one contemporary Christian philosopher, Jerry Walls, to write a paper titled, “Why No Classical Theist, Let Alone Orthodox Christian, Should Ever be a Compatibilist” (Philosopia Christi, Vol. 13, No. 1, 2011). The alleged implications of THEOCOMP, therefore, create quite the dialectical burden for the Calvinist (philosophical-) theologian. For some, it’s as if they have an albatross around their necks, and the sooner it can be removed, the better.
Clearly, one way to address the above challenges to THEOCOMP is to address them head-on. Another approach is to find a way to have your cake and eat it too. Roughly, this second way aims to keep their Calvinist bona fides while affirming libertarian freedom, at least for the vast majority of our actions.
Libertarian freedom is the kind of freedom that is, at least currently, the fashionable view of the church—or, at least her intellectuals, especially the more philosophically oriented among them. While contentious, I’ll offer a standard definition of libertarian free will (LFW):
LFW = A person S is libertarian free with respect to some action/choice/omission X if and only if (i) S could have done otherwise than X and (ii) S is the ultimate source of his X-ing.
Let me briefly unpack (i) and (ii). (i) says S is free with respect to X-ing only if S could have done other than X given an identical history leading up to the moment of his X-ing. As an example, let’s think of this condition in terms of possible worlds: You are free with respect to choosing Lucky Charms for breakfast only if there is some possible world with the same history leading up to your choice at which you don’t choose Lucky Charms. This means, everything leading up to your choice is the same. You choose otherwise given your same tastes, preferences, upbringing, reasons for eating Lucky Charms, etc. (N.B. Some libertarians don’t hold to (i) but will hold to something like (i*): there must be alternative possibilities, which doesn’t entail the the agent must be able do otherwise.)
(ii) says you have to be the ultimate source of your choosing Lucky Charms. Suppose you have fostered a Lucky Charms loving nature, such that, given your nature, you will always choose Lucky Charms if given the choice. Libertarians will say you are free here only if you formed your Lucky Charms loving nature. The “origin” is ultimately in you. Many libertarians will say that (i) is needed to truly “form” your character or “set” your will. But the crucial point is that (ii) claims that you can’t be the ultimate source of your will, character, nature, actional spring, etc., if that will etc., has been determined to be that way.
LFW, then, is incompatibilist. This is an important point to remember. COMP is logically (and/or metaphysically) incompatible with LFW.
It is said—though I find it very contentious—that affirming the above definition of freedom is the only way to maintain true ascriptions of moral responsibility and also get God off the hook for all the evils.
III. The mashup
If the Calvinist can keep the Calvinist essentials while incorporating LFW into her overall theory, she can be said to have her cake and eat it too. Of course, TD can’t be essential to Calvinism. Call this position CAKE.
I’m not sure how the argument for CAKE goes. I assume there’s many ways to slice it. One initially promising way to cut CAKE would be to say that most of our everyday choices, e.g., what cereal to eat, what book to read, what church to attend, whether to forgive a slight against us, etc., are all libertarian free. To maintain the Calvinist bona fides, however, our repenting and trusting in Christ is determined. I’ll call this DECISION. Greg Koukl presents something like this view here.
Since CAKE is constituted by lots of propositions, I will list the two crucial propositions we will be concerned with. For our purposes, we will define it thus
Cake = DECISION is determined & LFW is true.
IV. An unsatisfying CAKE
I think such a position is fundamentally flawed. Here’s my brief case for why CAKE should be abandoned. (INC = ‘incompatibilism):
First, I take it as assumed by proponents of CAKE that many of our post-DECISION choices that spring from our redeemed nature are free. This might include going to church, loving neighbor, forgiving a slight, reading the Bible, etc. Now, assume, for reductio, that CAKE is true, then
- (a) DECISION is determined and (b) LFW is true. (def of CAKE)
- Some of our post-DECISION choices that spring from our redeemed nature are free. (premise, Christian theism, common belief)
- If DECISION is free, then it is either compatibilist free or libertarian free. (premise)
- But, DECISION can’t be libertarian free, since it’s determined. (def. of LFW)
- But, DECISION can’t be compatibilist free either, since that entails a contradiction (see EDIT below). (logic)
- So, DECISION is not free. (4-5, 3).
- But, if DECISION is not free, then none of our post-DECISION choices that spring from our redeemed nature are free. (they can’t be compatibilist free, per the reasoning in EDIT; and, they can’t be libertarian free because DECISION itself isn’t free, and this violates the sourcehood condition of LFW).
- None of our post-DECISION choices that spring from our redeemed nature are free. (6, 7)
- Contradiction! (2, 8)
Here’s the basic reasoning: If our decision to turn to Christ, e.g., repent, believe, trust, etc., is determined, which it must be (ex hypothesis) to maintain a Calvinist bona fide, then it can’t be free. For if it were free, it’d have to be compatibilist free. But compatibilism is true, then LFW is (necessarily) false (and vice versa). And so the main motivation of CAKE, viz., to incorporate LFW, is undermined. Okay, so we deny that the decision to repent, believe, trust, etc., is free in any sense. It’s, for lack of a better phrase, “hard determined.” But then if this were so, what of the countless ostensibly free actions/choices we do that stem from our redeemed nature? This might be going to church, loving our neighbor, turning the other cheek, family worship, etc. We’ve already seen that these can’t be compatibilistically free. That leaves libertarian freedom. But given condition (ii) in §II, these actions/choices can’t be libertarian free. So something has to go. Presumably, the Calvinist bona fide is not negotiable. If it were, what’s the point of CAKE? Just drop Calvinism and be done with it. I also assume that denying that our ostensibly free post-conversion actions/choices aren’t free after all, is a dead-end. So it’s the libertarianism that must go, and (this slice of) CAKE fails.
V. A way out
Here’s a way out for the beleaguered Calvinist. He does not need to do anything as rash as affirm LFW which, you will recall, is partly defined by the thesis of incompatibilism. Rather, he can affirm that all of our free actions are not determined. This is not to affirm LFW, however, for indeterminism isn’t incompatibilism, and LFW needs the latter. He could affirm a definition of freedom like this one: S is free with respect to X if and only if S wants to X for the right reason.
I’m not saying this definition is good, but it’s short and sweet and allows me to make my main point, viz., it’s a compatibilist definition of freedom. As indeterminism isn’t sufficient for LFW, indeterminism also doesn’t falsify a compatibilist definition of freedom. That is, compatibilism doesn’t entail the actual truth of determinism. However, since the beleaguered Calvinist wants to maintain our above bona fide, this definition of freedom would possibly allow our post-conversion acts/choices—acts/choices which fail to meet the condition of ultimate sourcehood, and thus can’t be libertarian free—to be free. Moreover, this view would allow our beleaguered Calvinist to claim that the vast majority of evil acts are not determined by God.
This position doesn’t run into the problems brought out in IV. But that doesn’t mean it doesn’t run into its own set of problems. In fact, I don’t think it’s sustainable—at least in giving the beleaguered Calvinist the peace she seeks. But, it’s a more viable strategy than that of the above slice of CAKE. Detailing its problems is beyond the scope of this post.
A friend asked about (4), (5), and (6), namely, why accepting compatibilism for some of our post-DECISION acts that flow from our redeemed nature shows the falsity of CAKE, i.e., ¬(1). Here’s the answer I gave him:
I take it that CAKE is false because CAKE = Calvinist bona fide (CBF) & libertarianism.
Here’s how I define LFW: INC & Someone is free/does a free action.
This entails that CAKE = CBF & INC.
Now consider the latter conjunct, INC. I take it that INC negates a possibility claim, namely, COMP. Roughly, COMP says that possibly, some free act is determined. So, <>(Fx & Dx). (I take it the act isn’t determined but *derivatively* L-free, i.e., traced back to a prior L-free choice to set one’s will, undergo hypnosis, get drunk, etc). Thus, INC = ~<>(Fx & Dx).
So, this entails that CAKE = CBF & ~<>(Fx & Dx)
But to affirm (4a) is to affirm <>(Fx & Dx). So, CAKE would entail that there is a model on which these are true: [CBF & ~<>(Fx & Dx) & <>(Fx & Dx)]. But this conjunct is necessarily false.