The concept of Pelagianism lies at the core of Eric Nelson’s recent book The Theology of Liberalism: Political Philosophy and the Justice of God (2019).[i] Nelson argues that Pelagianism, which was deemed heretical by the early church, formed the “animating conviction” of proto-liberal theorists like John Milton, John Locke, Gottfried Wilhelm Leibniz, Jean-Jacques Rousseau, and Immanuel Kant. Indeed, “early-modern ‘liberalism’ simply was Pelagianism” (xi). In contrast, modern liberalism emerges from John Rawls’s “self-conscious repudiation of the Pelagian theological tradition” and is, in effect, “secularized Augustinianism” (xi, 205, respectively). Thus, Nelson maintains that whether one wishes to understand liberalism’s origins or else its development in the last five decades, the concept of Pelagianism proves central.

Yet what does Nelson mean by Pelagianism? A close reading of the book’s early pages shows that he offers three distinct formulations of the concept. Should we accept them? I argue that we should not. Each conception either overlooks orthodox alternatives to Pelagianism or else rules out such alternatives, based on a view of original sin that is needlessly partisan. These moves lead Nelson to define Pelagianism too broadly; they therefore threaten to sow confusion in what is otherwise an excellent book.

The First Formulation

Nelson first defines Pelagianism as a movement that “inferred the possibility of human freedom and merit from the justice of God” (xi). As a general description of Pelagianism, this may suffice; yet it’s worth comparing with some finer-grained conceptions. The Oxford Dictionary of the Christian Faith, for example, calls Pelagianism “the heresy which holds that man can take the initial and fundamental steps towards salvation by his own efforts, apart from Divine Grace.”[ii] Another definition of Pelagianism comes from John Lennox’s 2017 book Determined to Believe? The Sovereignty of God, Freedom, Faith, & Human Responsibility. Lennox calls Pelagianism “the idea that human beings can initiate their own salvation, or respond to God independent of his grace” (141). The latter two definitions make clear that the “freedom” and the “merit” in question concern salvation. In contrast, Nelson leaves them unspecified.

Consider the merit question. Although orthodox Christianity denies the idea that humans can merit salvation, this is not the only form merit can take. In theory, at least, one could perform meritorious actions that did not achieve salvation but which were nevertheless commendable. Such non-salvific merit seems implied in Jesus’ interaction with the rich young ruler in Mark 10:17-22. Jesus does not contradict the ruler when he claims to have observed the commandments faithfully from his youth (v. 20). Nevertheless, Jesus tells him that he “lack[s] one thing” needed to “inherit eternal life” (vv. 21, 17).[iii] This suggests that not all merit concerns salvation.

Likewise with the freedom question. In theory, at least, one could initiate any number of free actions. The question for Pelagianism is whether we can take the specific action that leads to salvation, apart from (or in advance of) divine help. Even Augustine, whom Nelson rightly pits against Pelagius, maintains a place for free will in non-salvific acts—and not just in his early writing. In Book V of The City of God, Chapters 9-10, Augustine repeatedly emphasizes the reality of human free will in matters of morality. “Consequently, it is not in vain that laws are enacted, and that reproaches, exhortations, praises, and vituperations are had recourse to; for these… are of great avail.”[iv]

These observations matter for Nelson’s thesis, since we need to know what sort of freedom or merit the proto-liberals had in view, for any given passage, before we can conclude that they were espousing Pelagianism. For instance, at one point Nelson summarizes Milton, qua Pelagian, in writing, “The distinctive excellence of human beings is our ability to withstand temptation and choose the good” (11). Yet, a modest version of such claims seems compatible with an insistence on the need for Christ.

The Second Formulation

Nelson’s second formulation of Pelagianism gives rise to similar concerns that the kind of freedom in view is simply too broad. Nelson writes that “a Pelagian, in short, is a rationalist who insists upon the metaphysical freedom of human beings in order to address the theodicy problem” (3). This statement warrants brief unpacking.

By a “rationalist,” Nelson means someone who believes that morality is both objective and independent from the will of God (2). As for the “theodicy problem,” this refers to the need many feel—including the proto-liberals—to, in Milton’s words, “justify the ways of God to men.” That is, to reconcile God’s goodness with evil’s existence. Yet it is with “metaphysical freedom” that my interest lies. I take this last term to be equivalent to what Nelson elsewhere calls “the reality of human freedom,” that is, “our ability freely to choose the right” or “to choose not to sin,” as opposed to some form of determinism (3, my emphasis).

Notice that, as with Nelson’s first formulation, this definition of Pelagianism doesn’t narrow the scope of freedom to matters of salvation but speaks of it in general terms. Thus, the abovementioned cautions continue to apply. One might affirm the reality of human freedom in matters not pertaining to salvation—and might, moreover, think such freedom matters for a theodicy. To the extent any of the proto-liberals meant only this by their reference to freedom, he would not be a Pelagian in the precise sense stated above.

Indeed, we can take this line of reasoning up a notch. For, one might think that not only can we exercise freedom in cases not pertaining to salvation, but we can also exercise freedom in the matter of salvation itself—namely, by freely receiving or rejecting Christ’s atoning death on the cross—provided one insists that the “initial and fundamental steps toward salvation” belong to God alone.[v] And, in the sort of cases I’m considering, this initial and fundamental step arguably would belong to God, since “while we were still sinners, Christ died for us” (Romans 5:8), and since we would have no will with which to receive (or reject) Christ unless God had first given us one.[vi]

Moreover, our refusal to accept Christ has implications for a theodicy. When human beings, who were made for God, refuse Him, we elevate lesser goods instead. This process has generated countless evils, not least the destructive ideologies of the past century. If one can exercise freedom in the sorts of case I’ve just noted—cases that fall short of committing Pelagianism, in the precise sense, but which nevertheless have implications for a theodicy—then Nelson’s second formulation remains too broad.

The Third Formulation

What accounts for Nelson’s linkage between human freedom (broadly conceived) and Pelagianism? We find the answer contained within his third formulation. That formulation reads as follows: “I reserve the label ‘Pelagian’ for those who either deny the doctrine of original sin outright, or accept it in principle while denying that it brought about any effective change in the ability of human beings to avoid sin” (4).

Original sin. Herein lies the culprit—or, rather, a certain interpretation of original sin. To see the problem I am driving at, we should ask ourselves, What would original sin need to mean to warrant the lack of qualification we find in Nelson’s discussion of freedom and merit? In particular, why label any freedom-based theodicy “Pelagian”?

Original sin would have to imply that human beings had lost their freedom altogether; otherwise, Nelson’s formulations remain open to the abovementioned criticisms. Not all Christian bodies understand original sin as implying determinism.[vii] But some do. According to some, especially certain members of the Reformed tradition, original sin implies that “mankind is totally unable to… repent… [or] obey God.”[viii]

It is interesting, then, that Nelson’s summary of original sin in Chapter 1 contains language that closely resembles classic Reformed terminology. He writes that “human nature is depraved” and that “we depend utterly on God’s grace, which comes irresistibly to those he has predestined for salvation” (3, my emphasis). The latter quote seemingly alludes to the Calvinist doctrine known as “irresistible grace”; while the former may be shorthand for the doctrine of “total depravity.” If so, they represent the “T” and the “I” in the five-point summary of Calvinism known as “TULIP.”[ix] If I am right in drawing this connection, it suggests Nelson may be interpreting original sin according to the strong view found in certain contemporary Reformed circles. As just noted, this view leaves no place for human free will—at least among the unregenerate. This interpretation explains an otherwise-puzzling feature of Nelson’s prior formulations: why they are so broad, making no distinction between free will in salvation and free will in other contexts.

Yet this interpretation also has a downside. For, unless we are willing to grant that orthodox Christianity rules out all save the “strong” view of original sin, with its deterministic implication, then it’s clear that Nelson’s conception of Pelagianism excludes positions that should be fair game for orthodox Christians.[x] To put things crisply: All Pelagians affirm free will, but not everyone who affirms free will is a Pelagian. This includes affirmations that play a role in a theodicy. And this, in turn, raises questions about one of Nelson’s central theses. For we might wonder whether each of the proto-liberals truly was a Pelagian in the strict sense, or whether some, at least, might have been operating within the bounds of Christian orthodoxy, at least on the points in question.

Salvaging the Argument

Can Nelson’s argument be salvaged? Indeed, it can. As things stand, I have merely shown that Nelson’s senses of Pelagianism are too broad. I have not shown that the proto-liberals he surveys were not, in fact, Pelagians in the strict sense. Logically speaking, there are three possibilities: either all of them were, or none of them, or some. Further analysis might vindicate Nelson’s universal claim, even given the narrower reading of Pelagianism noted above. Yet if it does not, Nelson can maintain his basic insight by simply softening his central term. Regardless of whether each proto-liberal was Pelagian, Nelson has convincingly shown us that they all sought to defend God’s justice through an appeal to human freedom. This merits scholarly attention in itself.


[i] For a larger summary of that work, readers should consider Daniel Strand’s earlier review in Providence, along with Mark Tooley’s subsequent interview with Nelson.

[ii] See entry on “Pelagianism.”

[iii] All Scripture quotations in this essay comes from the English Standard Version (ESV).

[iv] Augustine, The City of God (New York: Random House, 1950), 157. See also Christian Tornau, “Saint Augustine“, The Stanford Encyclopedia of Philosophy (Summer 2020 Edition), 44, 47, and 50, for Augustine’s mature insistence on human moral agency.

[v] The Oxford Dictionary of the Christian Church, “Pelagianism.”

[vi] Augustine, The City of God, 155-157.

[vii] See, for example, the Catechism of the Catholic Church.

[viii] Erik Thoennes, “Biblical Doctrine: An Overview,” in ESV Study Bible (Wheaton: Crossway, 2008), 2530-31, my emphasis.

[ix] John Lennox, Determined to Believe? The Sovereignty of God, Freedom, Faith, & Human Responsibility (Grand Rapids: Zondervan, 2017), 70-71.

[x] Note that Nelson concludes his discussion of Augustine’s “mature doctrine of original sin” (in which we find the abovementioned language suggestive of Reformed terminology) by noting that it came to “shore up Christian orthodoxy” (3-4). Whether he thinks Augustine’s conception is separable from the distinctively Reformed elements involved in his gloss, I don’t know.