Leibniz on Middle Knowledge and the Best (1 of 2)

Theological Letters

My friend and colleague, “Herman,” came across my work on Leibniz. He noticed two claims that stood out as odd, given his (rather standard) understanding of Leibniz. First, he noticed that I deny that Leibniz advocates middle knowledge (i.e., the medieval theory that God has knowledge of hypothetical futurities that would be under certain conditions). “Herman” didn’t see how I could say this, given that Leibniz talks about God’s knowledge of possible worlds that never come to be. Is this not middle knowledge? The second claim “Herman” noticed is that I advocate the (extreme minority) reading of Leibniz that says he is a libertarian, as opposed to a hard determinist or a compatibilist.¹ “Herman” was curious about my application to Leibniz’s theory of the best. Leibniz famously advocates that, given what God is, God inevitably knows and wills the best. Hence our world, being willed by God, must be the best of all possible worlds. “Herman” noticed that, on my reading, Leibniz believes God freely creates our world, such that he could have not created it. “Herman” didn’t see how this was possible, given Leibniz’s commitments. In the following letter, I explain these rather surprising claims. Please subscribe and support my work!

Dear “Herman,”

It’s good to hear from you. It would be better to see you. But these are strange days. I’m pretty sure it’s the first time in human history that the phrase was uttered, Chrysostom's liturgy is buffering again.²

As for your questions, let me take them in reverse order and begin with whether Leibniz believes in middle knowledge.³ If you’re defining middle knowledge in a broad sense as knowledge of hypothetical futurities from which God chooses what to bring into existence, then Leibniz certainly believes in middle knowledge. Historically speaking, however, “middle knowledge” (scientia media) is a term that refers specifically to the Jesuit model of divine knowledge named for Luis de Molina, namely Mohilism. Molinism was not the only model of hypothetical futurities available to Leibniz, and from what I can tell, Leibniz sees problems with the Molinist model. So, when I say that Leibniz does not advocate middle knowledge, I mean that he does not advocate Molinism. Leibniz certainly believes that God has knowledge of hypothetical futurities, but I read him as presuming a different model of possible worlds than we find in Molina. Here’s why.

The two features of Molinism that Leibniz clearly rejects are (a) its presumption of freedom as equipoise and (b) its claim that future counterfactuals are part of God’s “pre-volitional knowledge” (i.e., God knows these hypotheticals prior to any act of will to know them).⁴ Both presumptions Leibniz sees as problematic. 

Beginning with equipoise, this model of freedom presumes that the will is indifferent to its options prior to making a choice, and the will retains its indifference even after self-determining in favor of one option over others. Opponents of Molinism thought this claim was strange, to say the least. Leibniz sees it as falling into blatant contradiction. To choose is to determine in favor of one option over others. To suggest that the will has self-determined but remains indeterminate is gibberish. (On this point, it is noteworthy that in older Latin dictionaries, determino, from which we get “determinism,” means toward or concerning an end, and this term was actually associated with free self-determination: i.e., the will freely choosing a means (e.g., steak) to its end (e.g., satiating hunger). The term did not indicate a lack of free choice, as it does today.)⁵ Leibniz agrees that the will must be indeterminate in itself to have the power of self-determination. He also agrees that to preserve such indeterminacy, there must be no inner necessity that conjoins the will to its choice. In short, the will is indeterminate concerning its options before choosing. But Leibniz presumes this indeterminacy does not apply after the will chooses since to choose is to self-determine. Such is the very nature of choice.

One further point of note on this, Leibniz also believes that the will has preferences and leanings, so to talk about true equilibrium within the will (i.e., the will being a tabula rasa, void of inclinations) is just delusional.⁶ Now, to be sure, Leibniz does not believe that one’s inclinations determine his will. We can incline toward something and not choose it, so the final indeterminacy of the will remains, despite leanings and preferences. But there is no denying the reality of human inclinations.

As a rather important aside, freedom of equipoise rose to prominence in the 1700s. In my assessment, the rise of equipoise significantly colors discussions about free choice in the 18th century forward. I think the newfound dominance of this model contributes to so many reading Leibniz as a determinist — from his day to the present. Scholastically minded thinkers, such as Leibniz and the Reformed and Lutheran scholastics, rejected determinism, but they also rejected equipoise — primarily because they thought equipoise was a nonsensical chimera, as Leibniz himself argues. Nonetheless, as more and more libertarians embraced equipoise in the 1700s onward, equipoise became the litmus test for libertarian freedom in the minds of many philosophers. To such thinkers, the fact that Leibniz rejects equipoise means that he rejects libertarian freedom in favor of determinism. I think this is wildly anachronistic since scholastic theology (medieval and post-Reformation) includes numerous examples of anti-determinist libertarians who reject equipoise. Be that as it may, the interpretive trend in the period in favor of equipoise is real nonetheless, and its influence on how folks read Leibniz persists to the present day.⁷  

Returning to Molinism, a further consideration against the Molinist reading of Leibniz is this. Leibniz distinguishes between advocates of middle knowledge and the “predeterminators.” Here, Leibniz contrasts the doctrine of middle knowledge — in which God foreknows future counterfactuals and selects which ones to bring into being — with the doctrines of premotion and divine concourse, found amongst the Dominicans, for example. The view of providence that the latter doctrines suggest is that God brings into being creatures; the existence of these creature is perpetually upheld by God who “runs alongside” creation (the literal meaning concursus); and creaturely motion is a product of divine premotion by which God moves the creature in keeping with its nature. The predeterminators harmonize divine providence with free choice by distinguishing physical premotion from moral premotion. The former simply underwrite creaturely movements; the latter underwrite the moral character of those movements. When a creature cooperates with God, it receives both physical and moral premotion from him. But such cooperation is far from inevitable since free creatures have the power of self-determination. When a creature sins, God continues to supply physical premotion but not moral premotion. A common example to explain the distinction here is some variation on the analogy of a man pushing a horse with a broken leg. If we ask why the horse is moving, the answer is that the man is pushing it. If we ask why it is limping, the answer is because its leg is broken. So, in the same way, God is always the efficient cause of creaturely movements, but when the creature sins, the deficient cause of sin is not God but the creature’s own will, or self-determination. 

The relevance of all of this to middle knowledge is that the Molinists define middle knowledge as pre-volitional. In other words, just as God has necessary knowledge of things, such as mathematics, prior to any exercise of will, since these are necessary truths, so God has middle knowledge of hypotheticals, or future counterfactuals, prior to any choice concerning these hypotheticals.⁸ However, several medieval thinkers reject the idea that hypothetical propositions can have a truth-value unless first grounded by an act of divine will. I’ll return to this point in more detail in a moment. But suffice it, for now, to say that Henry of Ghent and John Duns Scotus argue that the divine will must provide ontological grounding to hypothetical futurities (and any other contingencies) if they are to have a meaningful truth-value. For futurities and other such contingents are contingent upon God. Unless God is willing to bring about a hypothetical, that hypothetical is not possible in any substantive way. For example, if God were, for whatever reason, wholly unwilling to create human beings, then human beings, though containing no contradiction in themselves (i.e., being logically possible), would not, in fact, be possible. For the very concept of a human being entails that it is contingent upon God producing it, and God is unwilling to do so. This discussion of the grounding of hypotheticals serves as the backdrop for the so-called “grounding objection” against Molinism. To wit, What is it that grounds the future hypotheticals, or counterfactuals, that God knows prior to any act of will? The predeterminators, as advocates of the view that God must supply premotion and concourse to creatures, tended to think that pre-volitional knowledge of hypotheticals is nonsense. The divine will must underwrite these hypotheticals if they are to be possible in any meaningful way. 

Circling back to Leibniz, he advocates the doctrines of premotion and concourse. To no surprise, he  also sees the grounding objection against Molinism as legitimate. This places him on the predeterminator side of the middle knowledge dispute, a dichotomy Leibniz himself sets up. Were this not enough, I think there is good evidence that Leibniz affirms that the hypothetical futurities God knows (i.e., the possible worlds from which God chooses) are products of God’s will, not sets of pre-volitional hypotheticals. Hence, Leibniz states quite plainly that the possible worlds God knows include God’s own choices to supply premotion and concourse to the creature in that world,⁹ since “possible beings” without these divine underpinnings are no possibility at all. Combined with Leibniz’s rejection of equipoise, I think it is evident that, while Leibniz believes that God has knowledge of hypothetical futurities, his model is not middle knowledge. Rather, if his model has scholastic roots (which I think it does), those roots must stretch back to scholastic thinkers who reject equipoise and advocate post-volitional hypothetical futurities. 

In my dissertation, I make the case that John Duns Scotus is a likely antecedent.¹⁰ Henry of Ghent, contra Thomas Aquinas, developed the view that future contingents must be grounded by the divine will, not the divine intellect. Hence, Henry argued that the divine will generates future contingents, and the divine intellect, knowing the decrees of the divine will, knows what the future holds. Duns Scotus feared that such a model makes divine foreknowledge inferential and discursive, as if God’s intellect infers the future from God’s decrees. Duns Scotus wanted a stronger view of divine knowledge, but he saw the flaw in the Thomist model that the divine intellect, not the divine will, is the foundation of God’s free knowledge.¹¹ Dun Scotus thus sought a middle road. He embraced the intellectualist claim found in Aquinas that it is the nature of the divine intellect to know all things that are knowable, rejecting any type of discursive divine knowledge. But he sided with Henry that contingents are contingent on God, and thus if contingents are knowable, they are knowable because they are grounded by the divine will. Duns Scotus using these premises to develop an interplay between divine intellect and will that yields a divine knowledge of hypothetical futurities. He did not call these possible worlds, but what he describes fits what Leibniz means by the term. The interplay runs roughly as follows.