In formal logic and decision theory, “closure” implies that choosing a goal requires accepting all its necessary implications. This works cleanly in abstract logic, where everything is transparent and necessary. Human intentions, however, are more focused and goal-oriented. They track specific reasons for acting rather than every logical consequence, some of which may be unforeseeable.
For example, if one intends to run a marathon, the athlete knows that this entails being physically exhausted afterward. Yet, it would be a mistake to say that the anticipated exhaustion is intended. Rather, the intention is to complete the race, and the inevitable exhaustion is merely a known and unavoidable consequence that follows from that goal. In this scenario, exhaustion is a foreseen consequence not a sought after effect.
We can observe that an intention does not automatically transfer to every known consequence. In other words, we can intend an end while only accepting (but not intending) certain known costs or possible risks. This principle is called “non-closure of intentions under known entailment.” What this means is that one’s goals do not include every side effect one knows will happen. This distinction is useful because it separates what an agent aims at from what an agent merely expects. However, if intentions were “closed,” we would be constrained to say that the married father who intends to earn a promotion by working excessive hours necessarily intends to neglect his wife and family. Obviously, that is false. Therefore, by denying closure of intentions, we can maintain that an agent can expect that a decision will lead to a specific outcome without it being a purposeful target of his will.
Case study:
Sticking with the ambitious employee scenario, assume the man who is seeking a promotion knowingly adopts a plan requiring an 80+ hour work week for two straight years. He certainly foresees this will cause substantial time away from his family, yet accepts this foreseen byproduct without intending it as a goal, for he would gladly pursue his career goals in a less disruptive way. Later, a catastrophic result emerges. Sadly, his continued physical and emotional absence contributes to his spouse filing for divorce.
Common Deflections:
Imagine the husband says that he only intended the promotion and not to neglect his family, let alone destroy his marriage. This much would be true. But suppose he then adds that because he did not intend the harm, his intention for the promotion could not have entailed the neglect. This move is patently fallacious. It confuses intention with foreseeability. While he may not have intended the harm, he foresaw (or should have foreseen) its likelihood, and therefore bears responsibility for the predictable consequences of his plan.
With respect to the unforeseen divorce, evasion is subtler, if not more deviant. The husband can truthfully say he did not foresee the demise of his marriage. Yet this does not absolve him of deliberately choosing an extreme plan that created the conditions for marital collapse. In essence, the husband has weaponized the entailment thesis by sliding from the truth that he did not intend the consequences to the claim that he’s therefore innocent.
Lastly, if the husband is a deviant philosopher, he might gaslight that the state of affairs in conjunction with his choices did not cause strain upon the marriage necessarily. This, of course, is also true. Both emotional stress and seeking a divorce are only contingently true. However, while choices may not logically entail harm across all possible worlds, such a decision as this is tied to predictable consequences in our actual world. Basically, that an action does not guarantee a bad result in the abstract, doesn’t mean one isn’t responsible for the highly predictable damages it causes in reality.
Sadly, this sort of maneuver has too many life applications. In its ultimate inversion, accountability is not only evaded but reversed. The ambitious husband recasts his prolonged absence as a noble sacrifice for his family, even demanding gratitude for the very actions that are destroying his marriage!
Application to Divine Sovereignty and Human Agency:
The same tension between primary intention, foreseen side effects, and moral accountability lies at the heart of the long-standing debates over divine sovereignty, human freedom, and the authorship of sin. The ambitious husband’s deflection in some ways mirrors theological attempts to preserve God’s character while acknowledging a world filled with evil.
Molinism easily handles this challenge. In the Molinist framework, the Counterfactuals of Creaturely Freedom (CCFs), which are truths about how free creatures would freely act in any given scenario, reside in God’s middle knowledge. They are logically prior to the divine decree and independent of God’s will. Accordingly, God does not determine these counterfactuals, rather he merely discovers them as fixed truths.
Intending the Canvas, Not the Stain – Non-Closure in Divine Sovereignty:
In Molinism, God’s decretive intention is purely at the macro-level. To miss this is to miss the relevant dissimilarity with Calvinism. God decree intends the world as a whole outcome that maximizes his glory. The distinguishing factor is that specific sins within that world remain unintended, as they are truly structural byproducts of independent creaturely freedom over which God lacks control. Within a distinctly Molinistic framework, God intends only a complete feasible canvas, and not the darkened brushstrokes that are fixed upon the canvas. Essentially, because God does not determine the internal mechanics of human choices in a decretive sense, the non-closure defense holds intact. Furthermore, because God can neither extract nor attach sin to any set of circumstances as he pleases, there can be no determinative “authorship” of sin.
The short of it is, regarding the decree, the set of all possible worlds are objects of God’s natural knowledge. God then consults his infallible middle knowledge to infer which possible worlds can be actualized. God then freely intends, according to his good pleasure, a complete feasible world to govern for his own purposes. Nowhere in the those three logical moments does God orchestrate sin. Therefore, sin is clearly unintended as it relates to the creative decree.
In response to this many Calvinists will press the point that within a Molinistic framework, God strongly actualizes a set of circumstances in order to weekly actualize a particular sin for his purposes. This strategy aims to place Molinists and Calvinists on level ground when Molinists accuse Calvinists of making God the author of sin. However, this maneuver entails a shift from (a) the macro-teleological intentionality of Molinism at the decretive level to (b) an atomistic, component-conditionality at the level of divine providence. Up at the decretive level, no sin is intended. Down at the providential level, the total decreed description of reality falls out according to God’s foreknowledge. The final upshot is, because in the divine decree God never engineers or causes sin within any of the three logical moments, sin is not philosophically intended in a non-closure sense in the selection of the actual world. This blunts the tu quoque. At the very least, if Molinists can be accused via weak actualization, Calvinists face stronger charges via theological determinism.
By contrast, in Calvinism there is no such pre-volitional buffer because God actively determines the free choices of men. How this works is that within Calvinism, there are infinite possibilities available to God for actualization that are not available under a Molinistic framework. That’s because in Calvinism, specific sins (as well as praiseworthy intentions) are not just located in the decree but positively determined as part of the decree. Necessarily, because God can actualize any possibility, he can extract sinful intentions or add them under identical circumstances according to his good pleasure. (This is not an entailment of Molinism.)
Therefore, Calvinists who wish to employ non-closure must therefore insist that sin functions strictly as an unintended but foreseen byproduct within the decree and not as (i) a positively chosen instrument, (ii) an intended end, or (iii) an intended means to an end. But this belies theological determinism. Yet conversely, if the Calvinist tries to define sin as an intended means to redemption, it would by definition collapse the non-closure defense.
Getting Deeper:
A Calvinist cannot disclaim intention for a necessary step that is deliberately determined and then incorporated into the decree. That sin is intended for a greater good does not make sin unintended in this present context.
It would me helpful to review more precisely that non-closure means that if an agent intends an end (E), and knows that E entails a consequence (C), it does not logically follow that the agent therefore intends C. Again, intention is not closed under known entailment. Accordingly, one can intend E while merely foreseeing, accepting, or permitting C as a foreseen but unintended side effect or byproduct. The immediate takeaway is that intention tracks what the agent positively aims at or sets as a goal, and not with everything they know will (or must) result from that goal. However, if C functions as an intended means to E (e.g., God chooses C in order to achieve E), then God does intend C. Non-closure does not shield this.
The Calvinist Equivocation:
Calvinists often appeal to this principle by saying something like: “God intends the good end (a world that displays His glory and redemption), and although he knows this entails sin, he does not intend the sin itself.” This only “works” if these Calvinists use a weak, Molinist-friendly sense of intention. But this clashes with consistent Calvinism, which requires a strong, determinative sense of intention. When God decrees a possible world within a Calvinistic framework, he does not merely select a world but instead positively determines the truth-values of creaturely counterfactuals as components of the plan. Here, sin is not a mere byproduct that God is required to work with but a positively intended event that serves a greater intended purpose.
This Calvinist defense lies chiefly with using two different meanings of “intention.” Calvinists cannot consistently invoke the weak sense (mere selection + foresight) to claim non-closure protects God, and then rely on the strong sense (meticulous determination all the way down) to remain faithful Calvinists.
At the end of the day, within the context of the divine decree non-closure only applies cleanly when the Agent (God) lacks determinative control over the negative consequence (as in Molinism). Yet if the divine decree includes determination of evil consequences, the “unintended byproduct” claim collapses.
Teleological Asymmetry, a Solution for the Divine Decree and the Problem of Evil:
The Reformed solution is not with non-closure but instead rests on a basic and profound teleological asymmetry within the single decree. God ordains a history that entails transgression for a morally sufficient reason, including the manifestation of his attributes, particularly sovereign grace and redeeming love. One example might be that the elect could never fully appreciate the depth of divine mercy had they not been chosen out of genuine ruin. (Existential understanding supersedes purely abstract understanding.) Or perhaps, reprobation and unconditional election make better worshippers out of the redeemed. Regardless of the morally sufficient reasons of God (who is accountable to no one), the comprehensive determination of all things does not compromise human accountability or implicate God in the malice of sin. As long as fallen agents actively desire, execute, and approve their own transgressions, there is no sound reason to dismiss their culpability. To argue that God’s determination of their sin somehow makes them unaccountable is to argue by way of false disjunction. The incompatibilst must show how willed, responsible intentions are incompatible with causal divine determinism.
God’s intention is teleologically fixed on the glorious redemption he has secured, while the sinner alone bears the moral responsibility for the evil he willingly commits without any morally sufficient reason. We needn’t go further than that.
Closing:
In both the marital analogy and the theological debate, non-closure can protect an agent from blame of intent for true side effects, but not from responsibility for all consequences of a chosen path, especially when those consequences flow from mechanisms the agent has deliberately employed. The key difference lies in moral sufficiency. The husband’s pursuit of a promotion through unintended neglect hardly justifies the destruction of his marriage, whereas God’s wise and holy purposes, displaying the full spectrum of His justice, mercy, and glory, provide a morally sufficient reason for determining the details of a world in which redemption shines most brightly against the backdrop of real evil that humans desire.
Philosophical Theology 