The Logical-Possible Chasm of Molinism

Consider counterfactual of creaturely freedom (CCF) p: If person S were in state of affairs C, S would freely A.

C represents the relevant history of the world prior to S freely doing A. Within Molinism, given C, S always As. Therefore, if God wills S would freely A, God need only actualize S in C since S in C never results in ~A. This necessity is not true of Augustinianism given that within divine causal determinism p is an object of God’s free knowledge, which is grounded in God’s creative decree and not his Middle Knowledge of fixed brute facts. Of course, Molinism does offer a way around such sufficient-condition necessity, which requires indexing the set of all might-counterfactuals that do not intersect would-counterfactuals to un-actualizeable infeasible worlds. The aim of this post is to explain why Molinism needs such possible-infeasible worlds and to expose such an escape hatch.

The two-fold ambition of Molinism:

For Molinism to lay claim on the doctrine of God’s exhaustive omniscience, there must be a fixity to future contingencies. This requires that some might-counterfactuals of creaturely freedom also identify as would-counterfactuals of creaturely freedom. Simultaneously, Molinism also seeks to maintain indeterminism, which requires that some might-counterfactuals of creaturely freedom not identify as would-counterfactuals. After all, if all such possible contingencies would always occur, they would be necessary truths. So, Molinism requires within the set of all possible worlds a subset of infeasible worlds that God is incapable of actualizing. That’s where un-actualizeable contingencies that would never occur can be parked.

Entailments of Molinism:

Given S in C, Molinism entails that ~A might happen but never would happen. Moreover, Molinism entails that if p is true, then p exists in all feasible worlds – all worlds God is capable of actualizing. Conversely, p is false in some infeasible worlds (i.e., possible worlds God is incapable of actualizing). {Note: p needn’t be false in all infeasible worlds given that what makes some possible worlds infeasible worlds are future contingents other than p. In other words, ~p is sufficient for an infeasible world, but it is not necessary.}

Molinism entails that CCFs like p are contingent truths, which is to say, the freely chosen A that p contemplates is a future contingent. Given C, S would not necessarily freely A; though necessarily, S would always A in C if p is true.

Molinism has no claim on future contingents of creaturely freedom:

How is p not necessarily true given that God believes ~p is universally false within all possible worlds God is capable of actualizing? What is a true possibility that God cannot make truly actual, after all?

If a CCF like p is contingently true, then it follows that there is a possible world in which p is positively false or at least does not exist as true (depending on one’s take on the principle of bivalence). Given that Molinism allows for infeasible worlds within the set of all possible worlds, Molinists believe they have made room for the actual possibility of p being false (even if it is true that God is incapable of weakly actualizing ~A by strongly actualizing S in C). In other words, Molinism entails the actual possibility of future contingents that God cannot possibly actualize, (e.g, ~p). This invites the question, if God is incapable of actualizing a possible world because of an uncooperative future contingent such as ~p, then in what sense is such a future contingent a meaningful possibility? (Or, if God only believes p is false in infeasible worlds, then in what sense is p truly and possibly false?)

Molinist semantics and the logical-ontological or possible-actual chasm:

These creaturely dependent possibilities exist in the semantic land of possible-infeasible worlds. Therefore, Molinism entails some logical possibilities that are purely theoretical – so much so they are impossible for God to know as actualized realities, unless uninstantiated essences – the ultimate source of such true possibilities – make them so. So, Molinism entails true possibilities that can be actualized a whopping zero number of times, even though there are an “infinite number” of these possibilities. This is all the more striking when we consider the spontaneity of the pure contingency of libertarian freedom.

No matter how any world can be arranged, if it contains the actuality of the potential of what p abstractly contemplates, then <S freely As> always obtains as a concrete reality given the actualization of S in C. Accordingly, p as an abstract entity is true in all feasible worlds whether the concrete reality that p contemplates is actualized or not.

The irony of the quest for divine foreknowledge apart from determinism:

Now regarding logical necessity, if a particular truth exists in all possible worlds, it exists necessarily. Added to this, if something is logically necessary, then there is no possibility of it being other than what it is. This has little impact upon the Molinist position but only because within Molinism possibility is not necessarily God dependent. Molinism includes the claim that CCFs such as p are contingent truths because of a supposed logical possibility of p being false even though ~p is a sufficient condition for an infeasible world, a world which cannot become actual along with ~p! Therefore, for the Molinist some logical possibilities are admittedly impossible for God to actualize, yet those possible impossibilities are supposedly what prevents CCFs from becoming necessary truths. The potential for actualization of might-counterfactuals that are not also would-counterfactuals are supposedly real and creature dependent although God is incapable of actualizing the worlds in which they are true counterfactuals. They are abstract truths about possible realities that God cannot bring into actual existence, which would undermine their actual possibility, which in turn would make their negation (e.g., ~~p —> p) necessary truths. Therefore, the contingency of CCFs and exhaustive divine omniscience Molinism seeks on the basis of indeterminism ends in the brute fact necessity of all CCFs.

As intimidated here, a necessary truth is one that exists in every possible world. And although Molinism upholds a theory of possible worlds that affords room for contingent CCFs, if we maintain that necessary truths are truths that exist in every possible world that can possibly be actualized, then the truth values of CCFs in infeasible worlds are irrelevant in evaluating whether a counterfactual is a necessary truth. At the very least, could God believe such counterfactuals are possible?

Divine Causal Determinism saves future contingencies:

From an Augustinian perspective God freely determines what a person would freely do in any state of affairs. God is capable of actualizing a world in which I freely do not type this post under the same state of affairs in which I freely do type this post. Therefore, from an Augustinian perspective p is a contingent truth. Yet such future contingents are inconsistent with Molinism. The trajectory of Molinism leads to the untenable position that some logical possibilities are impossible for God to know as concrete, actualized realties. Accordingly, Molinism cannot bridge the logical-ontological / possible-actual chasm.

Compatibilist Freedom and Properly Basic Beliefs, an analogy of unlikely bedfellows

It’s interesting that many incompatibilist libertarians subscribe to properly basic beliefs that are formed in us but not strictly by us, which they’d say we are nonetheless morally responsible to live by. But how can such incompatibilists consistently maintain that we can justly be held responsible for such unwilled beliefs if we may not be held responsible for unwilled intentions? After all, wouldn’t unchosen beliefs be causally formed in us beyond our ultimate control no less than any caused intention? From an evangelical libertarian perspective, why would an infidel be responsible for a causally formed belief in God but not a causally formed intention to reject God? In fact, she heartily approves of the latter whereas the former is an inconvenience, which she suppresses because it doesn’t meet with her approval!

Plain and simple, we are responsible for what we believe and what we intend because they are our beliefs and our intentions. I maintain that it’s not the freedom of compatibilist freedom that’s so objectionable to libertarians, but rather it’s more likely to be God’s determination of the intentions of such freedom that they find so distasteful.

Libertarian free will, regress or crickets?

Libertarians and Compatibilists can agree that there are two distinct components when choices come to fruition, (a) an intention to act and (b) a specific act that proceeds from an intention. An actual act of the will comes from an intention to make a willed act.

Intention to act —> act of the will

Without an intention to act there is no act of the will. When an act of the will occurs, the choice is consummated. Both components of the choice obtain. An intention to act gives way to the actual act the intention contemplates. We may safely say the intention of the moral agent causes the act. The act is effected by the agent’s intention.

Examples:

Choice: I choose to eat ice cream.

My intention to eat ice cream —causes—> my actual eating ice cream.

Choice: I choose to dwell on the past.

My intention to dwell on the past —causes—> my actual dwelling on the past.

Both acts – eating ice cream and dwelling on the past – are caused by an intention to do. Therefore, we may say the acts of eating and dwelling are the effects of intention, lest we have un-willed and uncaused acts, which would not be subject to responsibility or moral evaluation.

It is not difficult to grasp what causes the acts we choose. Surely our intentions do. When we freely eat ice cream it’s because we choose to eat ice cream according to an intention to eat ice cream. Simple enough.

This invites the question, if our intentions cause our willed actions, then what causes our intentions? That question gets to the heart of the free will debate.

Infinite regress?

Assume for a moment that the intentions that trigger our acts of the will are themselves chosen acts of the will (just like eating ice cream and dwelling on the past are chosen acts of the will). As chosen acts of the will, intentions would be chosen effects of the will. Accordingly, intentions to act would be the effect of a preceding cause (just like the acts of eating ice cream and dwelling on the past are effects of a preceding intention). So, if an intention to eat ice cream is itself an act of the will, it too must be an effect of some intention. Some intention would have to cause that intention!

Recall that there are two components for a completed choice: an intention to act and the actual act that follows the intention. Now consider again my choice to eat ice cream. My act of eating of ice cream would be caused by my intention to eat ice cream:

My intention to eat ice cream —causes—> my act of eating ice cream.

So, if I not only choose to eat ice cream, but also choose my intention to eat ice cream, then my choosing of the intention to eat ice cream must be the effect of a preceding intention in order that I might have the intention to eat ice cream! (By now you see where this is going.)

An intention to have the intention to eat ice cream would cause the intention to eat ice cream, which in turn would give way to my actually eating ice cream. 😳

Intention to have the intention to eat ice cream —causes—> my intention to eat ice cream —causes—> my actual eating of ice cream.

Now then, what causes the intention to have the intention to eat ice cream? Well, if we choose our intentions, then another intention ad infinitum. We’d have a backward regress for any choice.

Agent causation?

Here’s a libertarian solution to the infinite regress conundrum. It’s called agent causation. Rather than choosing our intentions, the agent simply causes it.

Agent —causes—> intention —causes—> actual eating ice cream.

But what about the agent that is within the agent would cause the intention? We’ve already ruled out chosen intentions for that would lead to a regress conundrum. Well, what then causes intention from a libertarian perspective? What do we hear from the libertarian camp regarding what within the agent causes the intention to eat ice cream, and how is that agent property, whatever it is, an uncaused first mover? If it’s the agent’s will, then what inclines it? Crickets

Anticipated questions addressed:

For a simple explanation of (a) how unchosen intentions can be rational from a compatibilist perspective but not from a libertarian perspective, (even though morally significant intentions are formed within the agent yet not by the agent, being caused from without the agent), and (b) how libertarian freedom would destroy moral accountability, try here.

For a simple explanation of how compatibilist freedom can account for moral responsibility, try here.

For a simple explanation of why incompatibilist-libertarians might subscribe to the philosophical surd of libertarian freedom rather than to intuitive sufficient conditions for moral responsibility that compatibilism has to offer, try here.

For a more advanced treatment dealing with the truth-making of causal relationships as they relate to God’s (non-necessary) free knowledge of contingent truths that God determines to make causally necessary, try here.

For a more advanced treatment of how libertarian freedom when coupled with exhaustive omniscience results in necessary counterfactuals of creaturely freedom, try here.

What drives libertarian freedom, moral responsibility or determinism?

What seems to drive Libertarians to their view of freedom is not the reasonableness of pure contingency. It’s seems intuitive that compatibilist freedom provides the sufficient conditions for moral responsibility. I don’t think many libertarians would have looked any further than to those conditions if determinism wasn’t part of the discussion.

In other words, if we merely summarize the essence of freedom as the possession of certain cognitive capacities that produce different willed acts given different states of affairs, who’d object? Such freedom would seem to entail moral responsibility. Now introduce determinism, and then people feel the need to scramble for something additional to save moral responsibility, but it’s not because compatibilist freedom is lacking. The idea of libertarian freedom is merely an attempt to break the chain of determinism for reasons that don’t impinge upon personal responsibility! After all, isn’t an ultimate cause compatible with a proximate cause? Who killed Saul? (1 Chronicles 10:4,6,14)

Libertarian freedom does nothing to advance the cause of moral responsibility. In fact, such detached freedom would seem to abolish moral responsibility.

From whence come intentions, and how is compatibilism any better in this regard?

This post aims to address how unchosen intentions can be rational and person-relevant from a compatibilist perspective but not from a libertarian perspective. Even though morally significant intentions are formed within the agent, they are not formed by the agent, being caused from without the agent. Secondly, it touches upon how libertarian freedom would undermine moral accountability.

An unhappy choice for libertarians, infinite regress or ex nihilo:

The will is the faculty of choice, or that by which the mind chooses. If the will forms intentions to act, then intentions are a result of the mind choosing. We may add that if an intention to act is produced by the will, then it would have to be a result of a previous intention because definitionally the will cannot produce an unintended act. Yet if the mind chooses intentions intentionally (according to a preceding intention), then intentions would be a product of the will ad infinitum, as argued here. There would be what I’ll call a regress conundrum.

How can libertarian philosophy avoid regress, other than by agent causation? In agent causation the willing agent becomes the first cause. Pure spontaneity of intention saves freedom and moral responsibility, or so it’s said. Yet such autonomous independence would detach influence, reason, and relevant history from intentions and willed actions. We’re asked to believe by implication that the agent rises above all influences, wherefrom a posture of equilibrium forms intentions from a functionally blank past. In other words, given the liberty of indifference that agent causation contemplates, choices would be unmapped to personal history, entailing a radical break from the person doing the choosing.

Libertarianism’s dead end:

So, libertarianism is a project that entails acts of the will that bring into existence intentions – while simultaneously denying chosen intentions because of the regress conundrum. The libertarian commitment to the causal contingency of agent causation leaves libertarianism with unintended intentions mysteriously formed by the will, an internally inconsistent notion for libertarianism that would render unintelligible a libertarian claim on moral responsibility based upon ultimate sourcehood and regulative control.

A challenge to Christian compatibilists:

Augustinians should acknowledge that intentions are not chosen. Contingent beings neither choose nor cause the intentions of the will. However, with that acknowledgment comes significant challenge. What makes intentions any more rational and morally relevant from an Augustinian perspective? In other words, what’s the relevant difference between a conception of an intention that springs from nothing and an Augustinian conception that posits that intentions are caused by unwilled states of affairs that are the consequence of causal influences that don’t originate with us and are outside our regulative control? As the title of this post asks, “From whence come intentions, and how is compatibilism any better in this regard?” After all, neither philosophy adequately accounts for agent willed intentions, though only libertarians try to do so. (*Libertarians need it for moral responsibility. Compatibilists do not!)

What makes unintended intentions sensible?

If you’re Augustinian I would suggest you not read on until you feel the weight of the philosophical problem from the previous paragraph.

Nobody rationally determines intentions in a libertarian construct. There’d be no reason to guard the heart for we’d be able to kick bad habits spontaneously, according to a will that’s impervious to causal influences. Such radical spontaneity would result in pure randomness of choice, destroying moral relevance by detaching choice from person. In a split moment we should expect to see saints behaving like devils, and devils like saints. The implications of pure contingency of choice demand it! And any libertarian appeal to will formation doesn’t comport with the metaphysical contingency of libertarianism. Libertarians may not have their cake and eat it too.

The Augustinian solution:

The problem restated: If we don’t sovereignly instantiate our intentions but rather they are formed in us, then how can intentions be morally relevant to the person?

Although intentions are formed in us, they are not formed by us. Notwithstanding, our intentions are rationally relevant because when God maps the cause of our intentions to providential states of affairs, he determines that our resultant intentions remain consistent with our person. They fall out naturally and by design, even with our approval! This uniformity is not a guarantee for libertarianism since it would not be normative that intentions have any relevance to the person given the contingency entailed by libertarian freedom. So, if the life practicing thief finds a billfold loaded with cash, from an Augustinian perspective the formed intention will likely result in a free act of ditching the wallet and pocketing the cash. However, God could also trigger a childhood memory resulting in an intention to freely do right given identical circumstances. Unlike with the implications of libertarian spontaneity, from an Augustinian standpoint either intention would be causally relevant to the thief’s past. Also, both outcomes could be actualized by God, which is not the case with Molinism and profoundly undermines the contingency of CCFs that Molinism seeks on the basis of indeterminism, in turn exposing Molinism for the brute fact necessity of all CCFs, as argued here.

*For Augustinianism, moral responsibility is sufficiently obtained by other factors whereby agent-willed intentions are unnecessary (not to mention, philosophically inexplicable). As long as intentions are formed within the agent, even though they are are not formed by the agent, they are morally relevant. They are morally relevant because they are the agent’s intentions.

Natural Knowledge or Free Knowledge of CCFs?

Natural Knowledge: God’s knowledge of all necessary truths, including all possibilities logically prior to his creative decree.

Definition from Divine Foreknowledge Four Views, Edited by Beilby & Eddy, page 211.

God knows all possible worlds according to his natural knowledge. Yet many Reformed thinkers tend to extend natural knowledge to the objects of counterfactuals of creaturely freedom (CCFs) within possible worlds. I believe John Frame and Paul Helm are representative:

When God knows possible worlds, does he not also, by virtue of that knowledge, also know all possible creatures and their possible actions? So, from a Reformed point of view, there is no reason why we shouldn’t regard God’s knowledge of contingencies under the category of necessary knowledge.

John Frame, The Doctrine of God, page 503. (By “necessary knowledge” Frame means natural knowledge. He equates them along with knowledge of intellect, page 500.)

Paul Helm is perhaps more precise:

But if God knows what Jones, if placed in circumstances C, would do, then this is surely part of God’s natural knowledge, his knowledge of all necessities and possibilities.

Paul Helm, Shunning Middle knowledge.

It would seem that Frame assumes the premise that Helm asserts. Frame infers God’s necessary knowledge of CCFs from God’s necessary knowledge of all possible worlds. The problem is, CCFs are would-counterfactuals and as such do not merely pertain to all possibilities that God would necessarily know. A contingent aspect is being overlooked. To know what is generally possible is not to know what would be specifically true. That God necessarily knows all possible worlds does not imply that he knows counterfactual particulars within possible worlds other than freely and as contingently true.

By cataloguing CCFs under God’s natural knowledge as have Frame and Helm, such counterfactuals are relegated either to necessary truths or possibilities. CCFs are either like laws of logic that actually exist in every possible world and could not have been false, or they are akin to potential realities that necessarily exist as possible, though might never actually exist (other than as abstract possibilities.)

Although my actual existence is not a necessary truth, that, P, <I would, in this possible / actual world (Wp/a) freely type this post if placed under circumstance C> is true. Given that God believes all truth, God eternally knows P. This particular bit of counterfactual knowledge of my typing this post, X, should be considered transworld by such Augustinians. The transworld object of knowledge can be dropped into any relevant states of affairs, C, in any possible world, Wpn, so that in Wp1, Wp2, Wp3… God would know X would occur under equally similar Cs in any Wpn given the implied intrinsically causal power of C, which in the thinking of some is relegated to an object of Natural Knowledge. (We will table the question of whether X in C could be contingently related to which Wpn is in view, which I hope will become obvious later.)

This sort of intrinsically causal necessity is understandable among Causal (Nomological) Determinists, but it is an unnecessary and improper concession among Causal Divine Determinists. Has Christian determinism been so influenced by secular philosophy? (See James Anderson site for the various stripes of Determinism.)

When Augustinians catalogue such would-counterfactuals under God’s natural knowledge, what is implied is some sort of necessity for CCFs without which counterfactual knowledge could not obtain. What is implied is that CCFs are logically, metaphysically or in some other sense still indeterminately caused. After all, if some sort of necessity for there to be natural knowledge is not maintained, then C need not result in X, my freely typing this post, under C. In which case, the fixity of the result of C (i.e., the free choice of X) would defy truth value and, therefore, could not be an object of natural knowledge. Hence the need for some sort of necessity within the confines of natural knowledge. Yet, if the grounding of the counterfactual is God’s will, which it is(!), then the counterfactual would be contingent truth, an object of God’s free knowledge! (NOTE: This is not to posit the metaphysical contingency of libertarianism, which might be confusing some. True counterfactuals are not necessary truths, otherwise they’d exist necessarily. Notwithstanding, they don’t fall out purely contingently in a metaphysical sense, but rather they become causally necessary by decree, which is not to be confused with something being a necessary truth.)

Like with Molinism, such Augustinians as these, if consistent, are consigned to a view that would entail that any actualizable (truly possible!) world that includes equally similar Cs (i.e., similar relevant states of affairs), always results in X , my freely typing this post. (In passing we might note, even Middle Knowledge entails causality that Molinism cannot avoid. Molinists engage in a type of special pleading when they introduce might-counterfactuals and insist the set of all possible worlds include infeasible worlds!)

Scott Christensen has this to say:

Determinism refers to the idea that all things that occur in our world are necessarily and causally determined by prior conditions. Thus, given specific prior conditions, only one outcome could possibly take place.

Scott Christensen, What About Free Will page 12. (Scott makes a similar error on page 170 and perhaps elsewhere: “God could ordain any variety of outcomes that transpire in the natural world and the human plane of that world. But if he ordained something different to occur, then the preceding conditions would be different as well.” Page 170 (emphasis mine).

What these Augustinians are suggesting is that it’s the relevant states of affairs, circumstances or prior conditions that necessitate free choice. By cataloguing CCFs under natural knowledge, it is (unwittingly?) implied that the effect is ultimately caused by something intrinsic to the nature of C, otherwise God would not know X like he naturally knows all necessary truths and possibilities! Unlike Dabney who wrongly, I argue, attributed this knowledge to “Middle Knowledge” (yet of non-libertarian choices, gratefully!), these Augustinians would like to attribute God’s knowledge of CCFs to his natural knowledge, which would reduce the object of such natural knowledge either to (i) a brute fact or (ii) a reflection of the divine essence (if they’re not freely determined).

Christensen goes on to liken the causality of choice to our living in a “cause-effect universe…” Even offering as an analogy, “When the temperature cools to 32 degrees Fahrenheit, it causes water to freeze.” (Page 13.)

Now clearly Christensen is not a physical determinist when it comes to the mechanics of choosing. He’s a soft-determinist. One of the good guys(!), along with Helm and Frame. Notwithstanding, what is implicitly denied by more than a few is that God pre-interprets the particulars that comprise any C, and in doing so freely determines the causal relationship and truth values of counterfactuals. Therefore, with respect to CCFs, these too are a matter of God’s free knowledge, whereas possible counterfactuals are part of God’s natural knowledge. What must be remembered is that from a consistent Augustinian perspective CCFs are would-counterfactuals, not might-counterfactuals. They have definite truth values (albeit they are contingencies), which presuppose a truth maker. As contingencies, these eternal truths cannot be grounded in God’s ontology or natural omniscience, nor in anything outside of God, which only leaves his will of determination, making Divine Knowledge of CCFs a species of free knowledge.

Take liquid water freezing at 0 degrees C. (No need to get into pressure, additives, purity and nucleation centers etc.) Does God know this according to his natural knowledge? Consider that water at 4 degrees C is at its highest density, which means it will expand whether it is heated or cooled. Must that causal relationship necessarily hold true given all relevantly identical circumstances? Could not God have determined that water continue to become increasingly dense as it is cooled below 4 degrees C? (We could just as easily consider the direct relationship of temperature to gas viscosity and the inverse relationship it has to liquid viscosity.)

Now, of course, there are physical “explanations” for these sorts of phenomenon in this world, but the point should be obvious. “Laws of nature” merely map God’s will, which is to say his pre-interpretation of how new facts introduced into relevant states of affairs, fixed circumstances, or existing conditions would effect outcome. If this is true in the material world, how much more should we expect it to hold true when considering what must be considered pre-interpreted facts that are introduced into fixed circumstances…, which result in free choices? The resultant or subsequent abstract thoughts, motives, desires, intentions etc. are not randomly triggered but rather “caused” – yet according to God’s pre-interpretation of the variable(s). God gives causal facts their interpretive meaning. There are no brute facts. As I’ve noted elsewhere, can’t God determine that the same song introduced into equally similar states of affairs, within different possible worlds, result in different formed intentions, ending in, say, freely writing a letter, making a phone call or something else?

By cataloging CCFs under God’s Free Knowledge we rid ourselves of unnecessary, improper or unintended nods toward brute particulars, while being able to maintain that God is the only eternal propositional truth maker. To maintain what I’ve argued against is to imply that God must know that I would type this post under identical circumstances in any possible world! It would imply that necessarily, ice cubes float under identical circumstances in all possible worlds, and fish must necessarily have a place to live under frozen ponds.

In sum:

I argued that the knowability of CCFs are matter of God’s free knowledge, not God’s natural knowledge. Accordingly, given the exact same state of affairs, it is false that antecedent influences for any intention of the will necessitates the same choice in all possible worlds. The contingency of the outcome would not be due to libertarian freedom or a brute fact but rather a matter of God’s preinterpretation of antecedent particulars, which can vary from possible world to possible world according to God’s will. A non-theistic determinist obviously cannot make that claim. She is consigned to the objects of influence as being brute facts. I find many Christian compatibilists have followed that lead by mapping effects to metaphysical causal influences, overlooking God’s free determination of those relationships. Accordingly, they catalog knowledge of CCFs under natural knowledge. *I am inclined to think this misstep would readily be conceded by those who’ve made it. I tend to think their goal is to remove CCFs from Middle Knowledge. The reason CCFs might have been unwittingly parked in Natural Knowledge is because Free Knowledge is often associated merely with God’s eternal decree, not counterfactuals per se. Yet what tends to be missed is counterfactuals are decretive truths that pertain to possible worlds whether actualized or not.

(*After private interaction with one Augustinian thinker, it has become clear to me that it is believed by some that by virtue of God decreeing a counterfactual true it, therefore, becomes a necessary truth, which in turn makes it an object of natural knowledge. That is simply wrong by definition and entails dualistic implications, not unlike Molinism. Other Augustinian thinkers more steeped in contemporary taxonomy, analytic philosophy and philosophical theology will grasp the error and its implications immediately.)

I alluded to in this post and have developed elsewhere that molinists have no claim on contingent CCFs, whereas compatibilists do in that qualified sense I mention above having to do with God’s giving states of affairs their causal interpretation. There’s somewhat a delicious irony here given the fixity of CCFs in all feasible worlds for the molinist position. Their use of Middle Knowledge requires a fixity of causal influences that compatibilism does not. In other words, Molinism entails an impossibility of contrary choice under identical circumstances once we establish that infeasible worlds (ie, unactualizeable worlds) are statistically irrelevant when considering the possibility of choosing otherwise. Jones freely chooses X 100% of the time in an “infinite number” of actualizeable worlds in which Jones freely chooses between X and ~X given C. That’s a necessity quite foreign to Augustinianism.

Assume I’m a Molinist!

Why are these three points insufficient for human responsibility from an Arminian perspective?

Assume I were a Molinist asking the question!

  1. The possession of certain cognitive capacities that produce different acts given different states of affairs.
  2. Dispositional powers, which is to say the power to try to choose x rather than refrain from x or choose ~x.
  3. A “mesh” of first and second-order desires (desire to act and approval of desire to act) that are both intuitive and particular to choices in contradistinction to brute instincts, perhaps addiction and phobias too.

All three of those points are compatible with God’s eternal decree. Accordingly, how does a Reformed view of divine decree logically contradict moral accountability given that 1-3 would appear sufficient for moral accountability? To point to inability to do otherwise or the settledness of what a divine decree contemplates is a classic example of question begging. It’s merely to say compatibilist freedom is not libertarian freedom.

Molinist Counterfactual Backfires

Christian compatibilists and incompatibilists agree that man is morally responsible for his choices, and God has exhaustive foreknowledge of the same. Therefore, if man has free will, it must be compatible with God’s exhaustive foreknowledge. “It seems to me much clearer(!)” – and to the rest who desire to make sense of God’s knowledge of counterfactuals of creaturely freedom (CCFs) – to maintain that any true CCF must have as its propositional truth-maker God’s free and sovereign determination. The only other option, lest we deny God’s sole eternality by positing ungrounded CCFs, is that CCFs are necessary truths, like laws of logic that are grounded in God. In other words, unless we are willing to accept mysterious propositional dualism, we are consigned to accept some species of determinism with respect to necessarily or contingently true CCFs being reflections of God’s attributes or will respectively, the latter being more theologically sensible. In sum, since God’s foreknowledge is inconsistent with indeterministic freedom, we either are not free at all or else we are free in some other sense, a deterministic sense. If there is to be creaturely freedom, and if CCFs are contingent truths, then God knows them according to his free knowledge.

(Somewhat ironically, and as I’ve argued elsewhere, Middle Knowledge reduces true CCFs to necessary truths – true in every possible world that could be actualized (i.e. all real possible worlds!) – given that might-counterfactuals, which are contrary to would / would not counterfactuals and, therefore, never true, can neither be known nor actualized. [Obviously I reject the Molinist distinction of possible and feasible worlds. Though I entertain the distinction when considering modality of logical vs metaphysical possibility.])

Libertarian free will would destroy moral accountability, for how can pure spontaneity or agent causation (metaphysical concepts that detach influences, reasons and relevant history from willful actions) produce morally relevant choices? (More on that in a moment.)

Molinists like to point to Jesus’ rebuke of the inhabitants of Chorazin and Bethsaida as proof of God’s Middle Knowledge – for had Jesus performed the same miracles in Tyre and Sidon that he had performed in Chorazin and Bethsaida, Tyre and Sidon would have repented. The prima facie interpretation of the parallel passages is not that Jesus was revealing how others would have responded to those same miracles. Rather, the immediate inference is that inhabitants of Israel were even more hardened to revelatory truth than pagans (and will accordingly be counted more culpable on the day of judgment). It was a rebuke, not a nod toward Middle Knowledge. Yet aside from the obvious, let’s run with the Molinist interpretation and see where it gets us.

Consider possible world Wp with the exact same relevant state of affairs as actual world Wa up to time t, which is shared in both worlds. At t in Wp, Jesus performs in Tyre and Sidon the same exact miracles from Wa that he performed in Chorazin and Bethsaida at t. The result in Tyre and Sidon is repentance. If that is not causality, what is? Remove the miracles, no repentance. Introduce the miracles, repentance. Remove the miracles, no repentance. Introduce the miracles, repentance… Like a light being switched on and off, the miracles would have causally triggered repentance. If not, then what? Would the miracles have triggered (nebulous) agent causation? If so, how would that not entail divine causal determinism given exhaustive omniscience? The only escape hatch is that the miracles trigger nothing in Wp, which would only serve to highlight the morally irrelevant nature of libertarian free choices per the passing reference above. For what reason(s) would repentance obtain if not for the causal connection of the miracles?

Now of course, from a Reformed perspective, God could effect repentance and index such to immediate or secondary causes of either ordinary acts of providence or miracles. God freely knows all such counterfactuals. Notwithstanding, given a Molinist use of the alleged counterfactual in view, it proves too much. It either undermines the spontaneity of agent causation Molinism contemplates, or else it underscores the compatibilist premise that libertarian freedom brings to naught the influences, reasons and relevant history that make our choices ours, rendering them morally irrelevant, not unlike purely random movements.

Molinism, Dualism and Omniscience

At the heart of Molinism is Middle Knowledge (MK), God’s knowledge of true counterfactuals of creaturely freedom (CCFs) – i.e. God’s knowledge of what creatures would freely do under all sets of circumstances. Now, of course, Augustinians also believe that if there are CCFs, then an omniscient God must have knowledge of them. However, unlike Molinists, Augustinians maintain that such divine knowledge is only possible if causal determinism is true, which eliminates a need for middle knowledge – a knowledge of truth that cannot be appropriated either under God’s natural or free knowledge. (As I argue here, and more fully developed here, Augustinians should ground such counterfactuals in God’s free knowledge of what he would determine and not in God’s natural knowledge of what is necessary or possible.)

The False God of Molinism

If God (eternally) knows that I would freely type this post under C conditions, then it is true that I would freely type this post under C conditions (otherwise God would and could not know it). For the Molinist, the truth that I would freely type this post exists without any truth-maker – a determining source of the propositional truth bearer: Ron would type Molinism post under C conditions. (Let that sink in, that MK contemplates some truths exist without anything to make them true.) That is because Molinists deny that free choices are causally determined. By denying causal determinism in this way, Molinist are unable to “ground” the eternal truth value of counterfactual freedom. Nothing or nobody makes CCFs eternally true. Molinists deny that God determines the truth value of CCFs and they also (rightly) deny that free moral agents retroactively cause eternal truth values after having chosen in time (or in another possible world). Consequently, Molinists admit that God knows certain truths that they believe are neither necessary nor freely and divinely determined. These truths are simply there, eternally existing alongside God and his divine will, (which, as I argue here and more forcefully here, would make such would-counterfactuals necessarily true in all feasible worlds as opposed to contingently true). Like eternal holiness and the eternal divine will to send the Son, CCFs are uncreated abstract entities, not unlike divine attributes or God’s good pleasure. Obviously, such a philosophical construct denies the historic Christian faith by implicitly denying that God alone is from everlasting and the ultimate source of all things visible and invisible. It smacks at Dualism.

Another problem with Molnism is that it operates under the philosophical notion that creaturely freedom entails libertarian freedom. Libertarian freedom, or as it is often referred – libertarian free will (LFW), is a philosophical position that entails that free will is incompatible with causal determinism. We are asked to believe that for a choice to be free, it truly might not occur under the same exact circumstances in which it truly would occur. Therefore, God would somehow have to know that free choices would occur even though they truly might not occur. The problem with such a musing about metaphysical contingency is that God would know contrary truths, which is logically impossible! As long as it is true that I might (and might not) type this post, it remains false that I in fact would type this post. In which case, God could not know I would type this post, which limits God’s exhaustive omniscience.

Aside from the fact that LFW cannot be derived from Scripture – yet divine causal determinism can, LFW is too ambitious of a Christian position. For to subscribe to LFW is to affirm a species of theological dualism. Also, if taken to its logical conclusion, LFW leads to Open Theism, a heresy that limits God’s divine omniscience.  

Molinism & The Fixity of CCFs

A necessary truth is one that could not have been false. From an Augustinian perspective, a necessary truth is true in all possible worlds. In other words, from this perspective a necessary truth is true in all worlds that could be actualized.

I argue here and more thoroughly here that CCFs (counterfactuals of creaturely freedom) from a divine causal determinism perspective, being contingent truths, are best viewed as a matter of God’s free knowledge (as opposed to his natural knowledge). Whether person S would freely choose A or ~A in state of affairs C is up to God. Both choices as a matter of God’s good pleasure are possible given the same relevant state of affairs C. From within any possible world W, S would choose A in C is true in some possible worlds; S would choose ~A in C is true in all other possible worlds. As I wrote in that previous entry, “Given a state of affairs God could determine different resultant dispositions to act. Given an identical state of affairs, God could determine a fragrance or song from yesteryear to causally produce a disposition either to look at an old photo album, pick up the phone to call someone or something else. These alternative possibilities would not be indeterminate might-counterfactuals of libertarian creaturely freedom but rather intrinsic possibilities, part of God’s natural knowledge, from which God could determine and freely know any true CCF.”

From a Molinist perspective only feasible worlds can be actualized. Now allowing for that concession within Molinism – that some possible worlds are infeasible worlds due to libertarian freedom, then necessarily, true CCFs are true in all feasible worlds (i.e. worlds that can possibly be actualized into a reality within which instantiated essences could make those free moral choices), but are not necessarily true – need not be true in possible worlds that cannot be actualized (i.e. infeasible worlds that are impossible for God to actualize). Molinism also entails that the counterfactual need not be true even in all feasible worlds given the assertion that would-counterfactuals are not contrary to their use of might-counterfactuals.

Given Molinism, if A would be freely chosen by S given C in one feasible world, that counterfactual would be the case for all feasible worlds. Yet that would make the counterfactual a necessary truth, though not the actualization of it a necessary truth. (Such fixity of counterfactuals is not the case from an Augustinian perspective once we catalog CCFs under God’s free knowledge as opposed to God’s natural knowledge.)

Although within Molinism S might (and therefore might not) choose A given state of affairs C, if S would choose A in C, then the knowability of that would-counterfactual must entail that the truth bearing proposition of the counterfactual be a necessary truth – one that would be true in all feasible worlds. After all, if that were not the case, then knowing state of affairs C (and the workings of S) would not provide the certainty God needs in order to know A would be freely chosen by S under circumstance C. 

Within an Augustinian perspective, God knows C would cause S to freely choose A or ~A by sovereign fiat. The outcome would be a contingent truth and as such could vary from W1 to Wn per God’s free choiceWhereas within Molinism, the outcome of S’s free choice given C, being divinely undetermined, is therefore mysteriously fixed as true and, therefore, a necessary truth for all feasible Ws as either always A or else always ~A. Ironic?

Given molinism, although S would A or ~A in C would be purely contingent metaphysically, the proposition that bears the true truth-value of the CCF would be necessary (being the same in all possible actualizeable worlds). From a molinist perspective, in every possible world that is actualizable, the counterfactual is consistently either true or false and, therefore, should be conceded as a necessary truth either of counterfactual S would A or ~A in C. But how can God not be the truth maker of a necessary truth? Brute facts are not facts.

In summary:

A necessary truth is one that exists in every possible world. And although Molinism upholds a theory of possible worlds that affords room for contingent CCFs (i.e. a would-counterfactual is false or true depending on the possible world in view), if we maintain as a redundant compound statement that (a) necessary truths are truths that exist in every possible world (b) that can possibly be actualized (i.e. a correct view of every possible world!), then the truth value of CCFs in infeasible worlds (worlds that are impossible to actualize), should be disregarded when evaluating whether a counterfactual is a necessary truth.

Molinism may not properly lay claim on CCFs being contingent truths. After all, given LFW, some “possible” choices would never occur regardless of the number of trials. Therefore, those counterfactuals should be deemed necessarily false because they are false in every world that could possibly be actualized. Zero possibility of occurrence is semantically contrary to actual possibility that A would occur or that A might occur.

Within Molinism only feasible worlds can be actualized; yet that semantic concession still must come with the price of considering all relevant CCFs – those CCFs that have a non 0% probability in identical circumstances – as necessary truths given LFW and MK.

At the very least, within Molinism the chasm between what is logically consistent (all possible worlds, including infeasible ones) and what is actually metaphysically feasible (true CCFs) is too vast. For a fuller, more technical treatment of this idea, perhaps consider this.