A Brief History of Predestination (2 of 4)
Theological Letters
A friend, “Moses,” reached out to me in search of a “pithy” explanation of how the Eastern Church fathers understand predestination. I provided an answer that was relatively brief but not quite pithy. I, then, decided to follow that reply with a more thorough treatment of the topic. In a word, I offered a brief history of the doctrine of predestination, in the Christian East and West. My previous post looked at the Latin views prior to Augustine as well as the Augustinian shift. Today, I will cover the medieval developments in the wake of Augustine. Part 3 will look at the Reformation disputes that followed, and part 4 will explore how the Christian East contrasts with this Latin history. Once I’ve posted all three, I’ll supply my not-so-pithy explanation that occasioned this history.
Some readers may notice a change in title. I originally promised that part 1 was 1 of 3, and we now have 1 of 4. Suffice it to say that when planning this series, I underestimated how much I could cram into a single post — originally expecting to put medieval and post-Reformation into one post. Alas, I needed to spread this letter out a bit further. Apologies.
Thank you for reading Theological Letters. To my subscribers, thank you for being here. To my paid subscribers, thank you for sustaining this work. A paid subscription costs about two coffees and unlocks my full archive, weekly letters, the 15-hour Orthodox Foundations lecture series, and monthly live Q&A sessions. If paid support isn’t possible, your likes and restacks truly help expand my reach. I appreciate you all.
Dear “Moses” continued —
Post-Augustinian Medieval Predestination
Augustine’s approach to predestination became formative in the medieval Latin West and definitively so. The reasons are several, but two stand out. The first is that Pelagianism was condemned at the Council of Carthage in 418. Though the council did not carry ecumenical weight,1 the decree nonetheless established a red line for Latin writers.
The second is Augustine’s vast shadow in the Latin West. Augustine was unquestionably a figure of great influence, known for his rhetorical force and prolific pen. But that influence was codified in the medieval era by Lombard’s Sentences,2 a collection of patristic quotes on various topics. Lombard relied heavily on Augustine, far more so than any other Church father. Hence, the work offers a largely Augustinian lens through which to view Christian doctrine. The significance of the work was amplified by Alexander of Hales’ “gloss” (i.e., margin notes and headings) on the Sentences,3 which became the standard work for training “theological doctors” in Paris, England, and elsewhere. As a result, the Sentences played a central role in Western theological education, second to only sacred Scripture.4 In other words, the Augustinian lens offered by Lombard and Hales became the lens through which Latin theologians viewed Christian doctrine, full stop.5
In his gloss, Alexander lays out the essential elements of predestination and reprobation. These elements should sound familiar, given what has been said about Augustine’s post-Pelagius views.
God’s preparation of a creature for giving or withholding grace is based on foreknowledge of the creature’s disposition to accept or reject divine aid.
Concerning those who are predestined, grace is not given on the basis of merits, for the creature cannot perform any meritorious deeds prior to or without grace.
God gives grace for his own glory, not due to the creature’s merit or any obligation to do so.
As for the reprobate, the cause of God passing over a creature is the creature’s sinful resistance that arises from his free choice.
The cause of the creature’s final damnation is the creature’s own sin.
These core elements provide the basic parameters for the medieval discussion in the wake of Augustine.6
The rather standard early medieval take on predestination is articulated by Thomas Aquinas (1225 - 1274 AD). Like Augustine, Aquinas advocates a single-predestination position. That is, God predestines to salvation not to damnation. Or put otherwise, salvation is traceable to God while damnation is traceable to the creature.
Now, Aquinas takes the rather standard position, found in both pagan philosophy and early Christian writers, that God’s Goodness means that he wills the good of every creature. The view presumes the realist position that a creature’s nature entails its end or purpose, and that purpose is the good of the creature. For example, the nature of an eye entails the purpose of sight. Seeing, then, is the good of an eye while blindness is an evil to it.7 Hence, being Good, God wills the good of every creature and being provident, he guides each toward its proper formation and purpose.8 The point naturally raises a question: Does God have an obligation to save every human being? After all, just as the eye is made for seeing, the human person is made for God.9 If God wills the good of every creature, then it seems he must will the salvation of all. To deny it is to deny either that God is the object of the properly formed man or to deny that God is Good.
Aquinas answers this problem by drawing a distinction between natural goods and goods that are above nature (i.e., those that exceed the creature’s natural capacities). Salvation is in the latter category, here echoing the nature-grace divide of Augustine. According to Aquinas, God’s Goodness impels him to will natural goods, but God has no obligation to will goods that are above nature and require extrinsic aid — these being the realm of gift not of duty. For this reason, God can be selective in his decision to give or withhold grace, to save or to pass over.
As in Augustine, predestination refers to God’s ordering of events in the divine mind in eternity (i.e., in his foreknowledge) in a way that leads a creature to eternal life. Such ordering is initiative by God’s choice to gift to the creature infused grace, and this grace results in salvation when the creature makes a right use of it.
Reprobation, by contrast, is not an affirmative choice to orchestrate damnation. Rather, reprobation is a permissive decree. Keep in mind that Thomas, like Augustine, presumes that God grants grace to some who freely cast it off. Like in Augustine’s portrait of Judas, the cause of his apostasy and damnation is his own free choice; there was no deficiency in the grace given him nor necessity in his choice to cast off grace. Yet, God permits this choice and permits him to reap what he has sown. In short, divine permission is not the cause of Judas’ damnation; Judas’ free choice to sin is the cause. Such is how Thomas frames the predestination-reprobation distinction. The former is an affirmative decree while the latter is a permissive decree.
One final point, Thomas insists that reprobation is permissive because, in keeping with earlier writers, he holds that God, being Good, cannot will evil.10 On the point, Thomas invokes the antecedent-consequent will distinction of John of Damascus.11 The distinction, in short, suggests that God wills every good, considered in itself, and repels every evil, considered in itself. Such is the antecedent will of God. Yet, our world is complex. So there are times when God must forego a good that he wills and permit an evil that he wills not. For example, it is good that man lives and an evil that he dies. But if a murderer is a danger to society, threatening to bring about a great many deaths, it is better that he die.12 In the same way, it is good that man is saved and an evil that he is damned. Hence, God wills that all are saved (antecedently), but his providential dealings with man often lead him to permit damnation (consequently).
Aquinas’ contemporary, Bonaventure (1221 - 1274 AD), expands the medieval discussion by introducing a series of distinctions for parsing predestination and reprobation, namely, the eternal purpose, the mediate effect, the meritorious effect, and the ultimate result. The terms do not amount to a substantive change, but they did offer a formal way of parsing the various components of predestination and reprobation.
According to Bonaventure, the “eternal purpose” of predestination is God’s Goodness and glory. In keeping with resistance to Pelagianism, what impels God to predestine some souls to salvation is not their merits, since they have none apart from grace, but God’s own Goodness and glory. The “mediate effect” by which God executes his eternal purpose in space and time is infused grace, which makes it possible for the elect to produce merits. As for those whom God has passed over, the mediate effect of their damnation is that God hardens their heart. The “meritorious effect” is the merit produced by the elect who make a right use of grace. As for the reprobate, the meritorious effect is his demerit resulting from his hardened heart. The “ultimate result” of all of this is salvation for the elect and damnation for the reprobate.
Notice that Bonaventure’s position diverges from Lombard — and arguably from Augustine and Aquinas. To wit, Bonaventure denies that there is a meritorious cause for God withholding grace. That is to say, there is no innate resistance in the reprobate that prompts God to harden the reprobate’s heart. Unlike prior authors who point to reprobates casting off or resisting grace, Bonaventure appears to toy with double predestination. That is, God chooses to not offer grace and even hardens the heart of the one passed over, which is the mediate effect of his damnation.
The shift may be shocking, but there is a sense in which it naturally follows from the Latin resistance to Pelagianism. After all, if one suggests that God grants grace to one because of his innate cooperation and passes over another because of his innate resistance, then one is very close to saying that the elect have natural merits that impel God to give grace while the reprobate has natural demerits that impel God to damn them. Put otherwise, salvation and damnation are within man’s natural capacities of will. Such is the essence of Pelagianism and the very claim Bonaventure wishes to avoid. Hence, he insists that God gives or withholds grace, not based on natural merits or demerits, but due to his eternal purpose — period.
John Duns Scotus (1266 - 1308 AD) introduces a further development with his concept of “instances of nature.” The claim suggests that God is not subject to the succession of time, but there are still logical sequences to divine decrees. We may not be able to talk about what God did from one moment to the next, but we can talk about the logical order from one instance to the next — akin to non-temporal sequences like in math or logic. The claim supplied a way of talking about the dynamic interplay between God and creatures within God’s foreknowledge, described by Augustine and others, and would serve as the foundation for the later post-Reformation discussions about the order of divine decrees.
To illustrate, let’s consider the narrative of Judas and Peter that Augustine first explained in time and then placed in God’s foreknowledge. To re-tell the tale using instances of nature, we might say something like this. Within the divine mind in eternity, Peter and Judas are on equal footing at Instance of Nature (IN) 1. Both are born into the world bound by original sin and therefore incapable of producing meritorious deeds. At IN2, God determines to give grace to both Peter and to Judas. At IN3, God foreknows that Peter and Judas become followers of Christ and make a right use of grace. At IN4, God foreknows that both men face a test and both men fail said test, Peter denying Christ and Judas betraying Christ, not due to any deficiency in the grace given them but due to their own free choice. At IN5, God faces a choice whether to offer further grace and restore these two men or exercise justice and let them reap what they have sown. God, for reasons unknown to us, determines to restore Peter by giving him further grace and to pass over Judas, letting him reap what he has sown. At IN6, God foreknows that Peter makes a right use of grace, leading to his salvation, and Judas hangs himself, leading to his damnation.
As you can see, the theory allows the sort of dynamic give and take between divine freedom and creaturely freedom that Augustine sought in his description. Duns Scotus merely formalizes this dynamic by providing a vocabulary for talking about the logical sequencing of decrees in the divine mind in eternity.
Duns makes one further contribution to the discussion worth noting, namely, his distinction between nolo and non volo. On this point, I’ll quote from my forthcoming Leibniz book:
Prior to Scotus, both intellectualists and voluntarists presume that choice is binary, choose (volo) or reject (nolo). Scotus diagnoses the problem of psychological determinism that follows. If choice is reduced to choose or reject, and if the will desires the good, then choice is mere perception. So long as the object is perceived to be good, the choice follows. For this reason, Scotus expands considerably the dynamics of freedom. Like Ghent, Scotus is a voluntarist. But Scotus says the capacities of the will go beyond rejecting intellectual judgments. The will is able to direct the intellect, press deliberation further, redirect to alternatives, and desist the considered choice altogether. In addition, he makes the case that no perceived good can compel the will into action. And for this reason, one capacity of will is to not act. The will can choose (volo), reject (nolo), or not choose (non volo). The binary, choose or reject, Ghent refers to as freedom of contradiction while the power to refrain Scotus calls freedom of contrariety.13
The distinction is especially relevant to predestination. I would not be surprised if you raised an eyebrow at the suggestion that God does not affirmatively damn the reprobate but passes over them. But Duns Scotus here offers a further nuance to the point. To not will something can be affirmative (nolo) or it can be passive inaction (non volo). The distinction provides clarity and a vocabulary to the claim that damnation is permissive, not active. God does not actively withhold grace and reject the reprobate (nolo). Rather, he permits his actions to run their natural course (non volo).
Peter Aureol (1280 - 1322 AD) builds on these developments, crafting a theory of general election. For context, however, I need to offer an aside on my above description of the Peter-Judas tale. The version I laid out above using “Instances of Nature” is based on Augustine’s telling, where the damnation of Judas takes into account his choice to cast off grace. Duns Scotus, by contrast, places God’s choice at the beginning of the drama: At IN1, Peter and Judas are on equal footing, and at IN 2, God chooses Peter (volo) and not Judas (non volo).
Now, Peter Aureol points out that God wills the salvation of all (1 Tim 2:4). But if Duns Scotus is correct and there is no difference between Peter and Judas at IN1, and if Bonaventure is correct that there is no meritorious cause for the withholding of grace from Judas, and yet, God gives grace to Peter but not to Judas, then it follows that God does not, in fact, will the salvation of all. Hence, Aureol rejects these claims.
According to Aureol, because God wills the salvation of all, we must presume instead that God gives grace to both Peter and Judas, and what differentiates Peter from Judas in election and reprobation is that Peter makes a right use of grace, while Judas does not. Notice that this claim is arguably a tacit premise in both Augustine and Aquinas. Augustine’s illustration of predestination and reprobation chooses Peter and Judas as its sample and grants that Judas was a true recipient of grace who could have made a right use of it until the end and been saved. Likewise, when talking about the reprobate, Aquinas talks about God permitting the reprobate to fall away — that is, from grace. The point raises an important ambiguity in these earlier writers. To wit, does God provide grace to all, such that all could make a right use of it and be saved? Or does God choose, for whatever reason, to never extend grace to some, knowing this means salvation was never truly offered to them?
Aureol makes the question explicit, and he insists that 1 Timothy 2:4 requires the conclusion that God does in fact extend grace to all, and therefore, the reprobate is damned of his own choosing, just as in the case of Judas.
The theory of general predestination that Aureol offers looks very much like what we see in Augustine and what I described above. At IN 1, Peter and Judas are both bound by original sin, incapable of producing meritorious deeds. At IN 2, God supplies grace to both men. At IN 3, Peter makes a right use of grace, cooperating with God, while Judas freely casts off grace, resisting God. At IN 4, Peter attains glory while Judas reaps damnation. (Forgive the truncated summary for the sake of efficiency.) As I said, the view is essentially the same as what we find in Augustine, but Aureol removes any ambiguity about whether God gives grace to all. He insists that, willing all to be saved, God gives grace to all, and what differentiates the elect from the reprobate is cooperation or resistance.
William of Ockham (1285 - 1347 AD) offers the most idiosyncratic view within this medieval discussion. Why? Because Ockham is a nominalist.
I do not wish to disappear down a rabbit hole on nominalism, so I will keep this brief. The realism-nominalism divide concerns the structures of reality — and within medieval thought, the boundaries of omnipotence. Realism maintains that the structures of mind (i.e., categories of genera and species, principles of math and logic, etc.) are abstractions of structures that are “out there” in the world. In short, these structures are real — hence realism. Nominalism, by contrast, suggests that these structures are mere “names” that the mind projects onto reality — hence “nominalism” from the Latin nomen (name). Medieval realists held that these structures cannot be violated, not even by God. Hence, logical contradictions mark the boundary between the possible and the impossible, and “impossible” includes for God. To be clear, the medieval realists did not mean that there are “things” that sit on the wrong side of this line that God lacks the power to make. To the contrary, they insist that contradictions are not things at all; they are non-sensical word combinations. Hence, God cannot make “them” because these words fail to mark a thing for God to make. Or as C. S. Lewis once put it, “nonsense remains non-sense even when we talk it about God.”14 For this reason, medieval realists, like Aquinas, suggest that contradictions are beyond the bounds of omnipotence.15
Nominalists, like Ockham, chafed at the suggestion that God might be limited in such ways. Hence, Ockham drew a distinction between God’s absolute power and his ordained power. The former, so the claim goes, includes a host of things that defy human reason. The latter identifies whatever rules God has put in place. Hence, we may think it impossible that God could create a square-circle, for example, since the claim falls into formal contradiction. But such reasoning is based on the ordained power of God that established the rules of our world. The absolute power of God, Ockham argues, is not bound by such rules per se.
To be sure, Ockham’s position is fully innovative within Christian history. As I’ve shown elsewhere, Christianity has been committed to realism from its founding, and this commitment is not incidental. It is confessional, playing an essential role to the doctrinal commitments of the Christian faith16 — which is why it is no surprise that Ockham was eyed with suspicion in his day. Nonetheless, such is Ockham’s claim.
Applied to predestination, Ockham offers a series of idiosyncratic premises. In his absolute power, God is capable of making creatures with the natural capacity to produce merit. But in his ordained power, God has willed that creatures are incapable of doing so and thus require divine assistance (grace) to produce merits. In his absolute power, God need not reward merit nor punish demerit. But in his ordained power, God chooses to reward merits and punish demerits.
Now, how this plays out in Ockham is rather like what we see in Peter Aureol with one exception: General election describes the ordained nature of predestination and reprobation, but God could have ordained a wholly different dynamic.17 The resulting theory runs roughly as follows: God ordains to make creatures who need grace to produce merits, though he could do otherwise; God ordains to reward merit and punish demerit, though he could do otherwise; God choose to give grace to all, though he need not; some make a right use of grace while others cast it off; God foreknows the merits of those who make a right use of grace and the demerits of those who resist grace; God predestines some to damnation and permits others to be damned based on this foreknowledge.
The view was suspected of Pelagianism, largely because of Ockham’s claim that God could make creatures capable of producing merits without grace. Yet, Ockham plainly grants that God’s ordained will is contrary to Pelaginaism. At least in our world, no creature is capable of doing what Pelagius suggests. Hence, grace is always required for the production of merits, and God has ordained that merits are required for salvation. Thus, the anti-Pelagians are correct, even if God could have ordained the world that Pelagius envisions.
Lastly, we reach Gregory of Rimini (1300 - 1358 AD). Gregory opposes the theory of general election in Aureol and Ockham, arguing that the position is Pelagian and grants too much to free choice. In response, Gregory offers the most overt theory of double predestination found amongst the medievals.
Gregory argues that none are predestined due to foreknowledge of a right use of will, and none are are reprobate due to foreknowledge of resistance to God. God predestines solely based on mercy. As for reprobation, Gregory rejects the passive conception of damnation (non volo), making it an active choice (nolo). The end result is this. Predestination is a unileral act of God by which he chooses to give eternal life to a creature, and the effect of that choice is that God calls the creature, infuses him with grace, and brings about his justification and future glory. Reprobation, by contrast, is a choice by God to withhold mercy from the creature, the effect of which is the creature’s damnation.
Admittedly, the positions of some earlier medievals appear suscceptible to a double predestination reading — a fact that Peter Aureol observed. Yet, these earlier writers equally sought to temper this ramification with the notion of reprobation by permission. In Gregory, however, we find an explicit embrace of double predestination.
This medieval landscape sets the stage for the post-Reformation discussion to follow. As we will see, there’s nothing new under the sun.
To be continued in Part 3.
In part 1 of this letter, I discuss whether or not the condemnation of Pelagius and the Latin advocacy of the Augustinianism that followed has ecumenical weight (see note 17 of that prior post). Rather than sending readers to dig for that note, I will repaste it here: The [Pelagian] dispute was sparked, not by Pelagius, but by his disciple, Celestius. Celestius was seeking ordination to the priesthood when he was challenged by Deacon Paulinus of Milan. From Paulinus’ letter, we get a sense of Celestius’ teachings, which were unquestionably heretical. Six claims stand out: Adam would have died even if he had not sinned; Adam’s sin harmed only himself, not his progeny; newborns enter the world in the same condition as Adam when first created; humanity neither dies in Adam nor rises in Christ; the Mosaic Law is as good a guide to salvation as the gospel of Christ; prior to the advent of Christ, there were men who lived sinlessly. While the Latin condemnation of Pelagianism and the ratification of the Augustinianism developed in its wake does not carry ecumenical weight, the teachings of Celestius were universally condemned by all, East and West. Celestius is [formally] condemned by the ecumenical council of Ephesus in 431 A.D. (canon 4). The condemnation of Pelagius, by contrast, is relegated to the Latin West. The best case for the ecumenical rejection of Pelagianism in favor of Augustinianism is an appeal to the seventh ecumenical council, Nicea II (787 A.D.). Nicea II ratifies the local council of Trullo. One could make a chain argument that Trullo received the African codes, which presume the canons of Carthage. The difficulty with this argument is threefold. First, the links in this chain become increasingly weak the further one moves away from Trullo. The claim that the African codes presume Carthage and that this presumption would have been recognized by Nicea II is unlikely. In other words, Nicea II probably does not understand its ratification of Trullo to also ratify Carthage. Second, the Eastern fathers offer clear statements in support of several positions rejected by Carthage: (a) man is subject to corruption but not guilt through Adam (e.g., Cyril of Alexandria, Commentary on the Epistle to the Romans, PG 74:788-89; cf. Irenaeus, Contra haereses, 3.23.6; 4; 5.27.2; Basil, Homily on Why God is not the Cause of Evils, PG 31:345); (b) fallen man has the freedom to respond to the gospel (e.g., Origen, De oration, 6; De princ. 3.1; Irenaeus, Contra haereses, 4.37.4-6; Justin Martyr, Apologia Prima, 43-44; Gregory of Nyssa, De Virginitate, 12; Oratio cat. 30-31; John Chrysostom, Homiliae XXIV in Epistolam ad Ephesios; John Damascene, Expositio fid. 2.24-27); and (c) infants, having no guilt of their own, are not subject to damnation (e.g., Gregory of Nyssa, De infantibus qui praemature abripiunt). All three of these affirmations are contrary to the canons of Carthage. So the idea that the Eastern fathers intended to ratify Carthage by means of their ratification of Trullo is dubious. Third, we find a general silence on Pelagianism in the Christian East, and in the rare instances when Pelagius is spoken of, he is viewed favorably. For example, Pelagius is not mentioned by Ephesus, Nicea II, or John of Damascus’ extensive catalog of heresies in the 8th century. By contrast, in the 9th century, we find Photios, the Ecumenical Patriarche of Constantinople, commending certain writings of Pelagius. In this light, the most plausible reading of this history is that there was ecumenical agreement that Celestius’ teachings were heretical and worthy of condemnation (hence, Ephesus), but the condemnation of Pelagius and affirmation of the Augustinian alternative was not embraced by the Christian East.
The original Latin text of Lombard’s Sententiarum libri quatuor can be found in PL 192.
Glossa in quatuor libros Sententiarum Petri Lombardi.
S.F. Brown and J.C. Flores, Historical Dictionary of medieval Philosophy and Theology (Scarecrow Press, 2007), 10.
The significance of Augustine in Alexander’s own thought is evident in his Summa universae theologiae — his own summary/commentary of the Sentences — in which he quotes Augustine explicitly 4,814 times and implicitly 1,372 times. See Irena D. Backus, The Reception of the Church Fathers in the West: From the Carolingians to the Maurists (Brill, 1997), 303.
Readers will notice a limited smattering of notes in this section on medieval views. Unfortunately, I do not have handy a series of citations, so I need to go from memory, and time does not permit me to dig up the primary sources. Or to echoe C. S. Lewis’ The Problem of Pain, “As this is not a work of erudition I have taken little pains to trace ideas or quotations to their sources when they were not easily recoverable.”
Here, “evil” is not a moral evil but a physical one. Philosophical literature often differentiates metaphysical evil, physical evil, and moral evil. In the case of blindness, we are discussing a physical evil — that is, a malformation or corruption — not a moral infraction.
On this topic, see my forthcoming book: Leibniz, Classical Theism, and the Problem of Evil: Why Classical Theism Must Affirm That Our World is the Best of All Possible Worlds, chapter 4, which can be found here.
The above phrasing is not quite precise on a technical level. According to the medievals, drawing on Aristotle, man is made for happiness — this being his cheif end — and the object of happiness — that is, the only thing that bring about this end — is God. But for clarity, I’ve chosen to avoid such pedentry.
See John of Damascus, De fide orthodoxa, 2.29 (PG 94:963a-70b). I discuss this distinction at length in several places, including: Why Would God Make the Damned?, To a Woman Who Lost Several Children, and Leibniz, Classical theism, and the Problem of Evil, chapter 5.
This sort of example can lead people to mistakenly read the distinction as utilitarian in nature, such that good ends justify evil means. But such is not the case. On this point, see my Leibniz, Classical Theism, and the Problem of Evil, chapter 6 or Stump, “Providence and the Problem of Evil,” 82-4.
The above quote, with relevant citations, can be found in chapter 4 of my forthcoming Leibniz book, which can be found here.
C. S. Lewis, The Problem of Pain (Samizdat University Press, 2016), 12.
Aquinas, ST, I, q.25, a.4.
See, e.g., Does Christianity Have Metaphysical Commitments? (1 of 2) and (2 of 2).
Admittedly, a fair bit of scholarly controversy surrounds what precisely Ockham thinks on this topic. Older scholarship (e.g., Pannenberg) read him as offering a theory of single-predestination, not terribly unlike other theories of the day. More recent scholarship (e.g., Oberman) accuse Ockham of semi-Pelagianism. For my part, I advocate the reading of James Halverson, which sees Ockham as offering a rather peculiar form of general predestination
As one surrounded by Calvinists (mostly of the Piper persuasion), I thank you for such clear explanation and yet great detail of how we got to the common misconceptions about predestination, providence, and grace. I am seeing that much of the disagreement among lay folks like me comes from a failure to define terms and rectify them with other terms or truths.
I'm so thankful for this series. Please continue to write these letters. I know you have a lot on your plate, however.