Augustine Before and After — The Development of the Doctor of Grace¶

Augustine of Hippo (354–430) is the hinge of this argument. Every classical Reformed soteriology is an Augustinian plant: the priority of grace, the bondage of the will, the gratuity of election, the perseverance of the saints. A critic who announces that "Calvinism is unbiblical" has not reckoned with the fact that the West's most influential theologian taught something very close to it fifteen hundred years before Geneva.

The question is narrower than whether Augustine taught grace. He did. It is whether he taught the same thing at the end that he taught at the beginning — and Ken Wilson's two books argue that he did not.

The standard narrative¶

The Reformed account is not a caricature. Pelagius made grace a supplement to a healthy nature; Augustine saw that this made the cross decorative and prayer superfluous. Driven back to Romans 9 and to Romans 11:6 — "if by grace, then it is no longer by works" — he concluded that election cannot rest on foreseen merit or even on foreseen faith, because then grace would be wages. The later treatises are the church's answer to a real heresy, and the Reformers recovered what the West had received.

In this telling the development is a deepening, not a reversal: the early Augustine knew grace was necessary, and the later worked out what that entails. Warfield presses exactly this, and Calvin insists he is quoting Augustine, not inventing a theology. That is the position to beat.

The historical case¶

1. The early Augustine on free will (388–395)¶

De libero arbitrio was begun around 388 and finished in the early 390s. Its problem is the origin of evil: if God is good, where does sin come from? Augustine answers that evil is a privation, not a substance, and that sin is a voluntary turning of a will God made good. The will is genuinely the will's own; that is what makes blame meaningful.

Book 3 takes the objection that foreknowledge destroys freedom and answers: foreknowledge is not causation. His commentary on Romans (c. 394–395) still grounded Jacob's election in God's foreknowledge of faith.

2. Ad Simplicianum (396): the turn¶

In 396 Augustine became bishop of Hippo and answered questions from Simplician, the Milanese presbyter who had baptised him, on Romans 7 and 9. His treatment of Romans 9:10–29 in Book 1 is the turning point.

Why did God love Jacob and hate Esau before they were born? Not because of works — Paul says so. But then why? If the answer is "because God foresaw that Jacob would believe," election still rests on something in the creature and grace is a payment. Augustine concludes that this cannot be Paul's meaning: the call does not depend on "him who wills or runs" (Rom 9:16).

He does not yet have a full doctrine of reprobation, but he has the load-bearing claim: election is not conditioned on anything foreseen in the elect, including faith.

3. Wilson's three stages¶

Wilson identifies three successive positions rather than two:

Stage	Position	Roughly
1	Foreknowledge of merit	God elects those whose good works he foresees.
2	Foreknowledge of faith	God elects those whose faith he foresees; grace enables it, but the response is the person's.
3	Unilateral election	God gives faith to whom he will; election causes faith rather than following it.

Wilson argues that stage 3 is not a clarification of stage 2 but a different account of how God and the will relate, and that this third stage — not the New Testament — is the foundation of the later Reformed system.

Two caveats: the thesis is not the consensus of patristics, and the labels simplify a messier story. What is not contested is that something changed at Ad Simplicianum, and that Augustine knew it.

4. The anti-Pelagian works: the mature position¶

The mature position appears in three late treatises. De correptione et gratia (On Rebuke and Grace, 426/427) asks why we rebuke sinners if grace is sovereign; rebuke, Augustine says, is one means by which God gives grace, and perseverance is a gift to the elect.

De praedestinatione sanctorum (428/429) is the most important. Augustine revisits and retracts his earlier view that election was according to foreknown faith: if faith is foreseen as the ground of election, grace is no longer grace. Predestination is the preparation of grace.

De dono perseverantiae (429/430) adds that perseverance is given only to the elect, and irrevocably.

Together: all are born corrupted and cannot will saving good; God chooses individuals unconditionally; he gives them faith and perseverance by efficacious grace; the rest are left in the mass of perdition, and this is just.

5. Continuity and discontinuity¶

There is both. The early Augustine already insists grace is necessary and the will is wounded; he never becomes a fatalist. His later doctrine is not "the will does nothing" but "the will is freed by grace to do what it could not do."

But the early Augustine thinks election ratifies a free response; the later thinks election produces it, and where the early made foreknowledge a kind of knowing, the later makes it the execution of a decree. Wilson's phrase "non-free free will" labels this: free from external coercion, not from the decree or the disposition of sin. Whether that is the best label is arguable; that it differs from De libero arbitrio is hard to deny.

Objections and qualifications¶

Objection 1: There was no real change¶

Some Reformed readers say Augustine held one doctrine throughout and the "development" is only emphasis. That is hard to square with De praedestinatione sanctorum, where he says he formerly held the foreknowledge-of-faith view and now rejects it.

Objection 2: Wilson overstates the discontinuity¶

This has force. Augustine never stopped believing in free choice; he redefined what freedom is for. Concede the point, then restate modestly: the ground of election moved from something foreseen in the creature to the Creator's unconditioned choice.

Objection 3: The thesis is contested, and the case does not depend on it¶

Wilson has been welcomed by critics of Calvinism and received cautiously by patristics scholars who read the development as an organic unfolding of a consistent theology of grace. Brown, TeSelle, and Bonner are not Calvinist apologists; they simply doubt the chronology is that clean. Treat the stages as a thesis, not a fact.

The best Reformed reply¶

The strongest Reformed response is exegetical. Augustine changed his mind because he was interpreting Romans 9, and the Reformed reader says: read Romans 9. If God's purpose stands "not because of works but because of him who calls" (Rom 9:11), and faith, though not a work, is still something the creature does, the question is whether Paul excludes foreseen faith as well as foreseen works. Augustine came to think he does; that it took a controversy to make him say it does not make it false.

A subtler version: if grace is necessary for the beginning of faith, the difference between saved and lost is already located in something God gives, and the later doctrine may be the early doctrine's logical cash value. This casebook must answer that from the text rather than from chronology.

What this does and does not prove¶

It does not prove Augustine wrong, that development is corruption, or that the Reformed reading of Romans 9 fails; Newman's essay on development warns against treating every change as betrayal.

It proves something narrower: the claim that the church has always taught unconditional individual election is false as history. The doctrine has a datable origin in one man's reading of one text under pressure — which does not make it untrue, but means antiquity does not carry the burden of proof.

What would count against this file: showing Ad Simplicianum continuous with De libero arbitrio on the ground of election, or that the pre-Augustinian fathers taught unconditional individual election. The second is the subject of the early-church file.