Patristic source review: what the fathers actually say¶

This appendix reviews the historical claims in 01-augustine-before-and-after.md, 02-manichaeism-and-the-making-of-a-determinist.md, and 08-the-early-church-on-free-will.md. It is written to do two things: to put the sources on the table, and to mark where the casebook's claims are stronger than the evidence.

The rule: the fathers are witnesses, not a magisterium. "The early church did not teach X" is a historical claim, answerable to the sources; it is not by itself an argument that X is false.

The claim under review¶

The casebook argues:

The pre-Augustinian church consistently taught that human beings have genuine free will, that grace is resistible, and that election is according to foreknown faith.
Augustine's later doctrine of unilateral, unconditional election was a development, influenced by his controversy with the Pelagians and (on some accounts) by the residue of his Manichaean years.
The Second Council of Orange (529) rejected any foreordination to evil, and the Eastern church has never accepted double predestination.
Therefore the claim that "the whole church has always taught the Reformed doctrine" is false.

Claims 1-4 are not all of equal strength. This review grades them.

1. The pre-Augustinian witnesses¶

Justin Martyr (c. 100-165)¶

Justin's First Apology 43 argues that human beings are responsible because they have the power to choose; he grounds the judgment in the freedom of the will, not in a decree. His Dialogue with Trypho repeatedly treats unbelief as a culpable choice.

Strength: clear on free will and responsibility. Limit: Justin is not writing about predestination; he does not address the later question of how grace and election relate.

Irenaeus (c. 130-c. 202)¶

Against Heresies 4.37 is the classic patristic statement of human freedom: man is endowed with the power of choosing, and the commandments presuppose that he can obey or disobey. Irenaeus also teaches recapitulation: Christ sums up humanity and reverses Adam's disobedience (3.18-19).

Strength: freedom, responsibility, and a universal recapitulation. Limit: "recapitulation" is not a doctrine of the atonement's extent in the later sense; it is a theology of incarnation and restoration.

Clement of Alexandria (c. 150-c. 215)¶

Stromata 2.4 argues that faith is a free act and that the will is not compelled. Clement is a Platonist and a moralist, and his categories are not Augustine's.

Strength: free response to grace. Limit: Clement's intellectualism makes him an awkward ally for any later system.

Origen (c. 185-c. 254)¶

De Principiis 3.1 is the most sustained pre-Augustinian treatment of free will: the will is free, and the diversity of human outcomes is due to the use of that freedom. Origen also speculates about pre-existent souls and universal restoration — views later condemned.

Strength: free will and moral responsibility. Limit: Origen's cosmology is not a safe foundation, and his own orthodoxy was contested.

Tertullian (c. 155-c. 240)¶

Against Marcion 2.5-6 defends the goodness of the Creator and the freedom of the will against determinism.

Strength: anti-determinism. Limit: Tertullian later joined the Montanists; his authority is uneven.

Methodius (d. c. 311)¶

On Free Will (preserved in fragments) argues against fatalism and for human self-determination.

Strength: explicit anti-determinism. Limit: fragmentary.

John Chrysostom (c. 347-407)¶

Chrysostom's Homilies on John 45-47 treat John 6:44. His reading is that the Father's drawing does not override the will: the drawing is persuasive, and the drawn person comes willingly. He also insists that the Jews' unbelief was their own choice.

Strength: a major Greek father reading the central proof text in a non-determinist way, within living memory of Augustine. Limit: Chrysostom is a preacher and exegete, not a systematic theologian; he does not use the later vocabulary.

John Cassian (c. 360-c. 435)¶

Conferences 13 is the classic statement of the semi-Pelagian position: grace and free will cooperate, and the beginning of faith can be a human initiative. Cassian was condemned in part by the Western church (Orange, 529, and the later "semi-Pelagian" label), and he is not a safe authority for a Protestant casebook.

Strength: shows that the East and the West debated the relation of grace and freedom, and that a non-Augustinian position was live. Limit: his position was rejected by the church that produced him. Do not cite him as "the early church's view"; cite him as evidence of a debate.

Vincent of Lérins (d. c. 445)¶

The Commonitorium 2 gives the famous criterion: that which has been believed "everywhere, always, by all" (quod ubique, quod semper, quod ab omnibus). Vincent wrote in the middle of the Augustinian controversy, and his criterion cuts against novelty of any kind — including, potentially, the later Augustine.

Strength: the church's own instinct against doctrinal novelty. Limit: Vincent's criterion is notoriously hard to apply, and he was not adjudicating the predestination question.

2. Augustine's development¶

The development is not seriously disputed:

Early Augustine (386-395): De libero arbitrio defends free will; De vera religione and the early commentaries treat faith as a human response enabled by grace.
396: Ad Simplicianum marks a turn. Augustine, wrestling with Romans 9, concludes that God's election is not based on foreseen merit or faith. In Retractationes 2.1 he says he had earlier understood the passage differently.
The anti-Pelagian works (412-430): De peccatorum meritis, De spiritu et littera, De gratia et libero arbitrio, De correptione et gratia (426/427), De praedestinatione sanctorum (428/429), De dono perseverantiae (429).
The mature position: grace is prior, unilateral, and infallible in the elect; the number of the elect is fixed; perseverance is a gift; the non-elect are left in their sin.

What is undisputed: Augustine changed his mind, and the mature position is a development. The pre-Augustinian fathers did not teach it.

What is contested: the cause of the change. Ken Wilson's Oxford thesis argues that Augustine's later view borrowed the shape of Manichaean determinism, having rejected its metaphysics. Reformed scholars reply that Augustine was driven by exegesis (especially Rom 9), that his doctrine is a legitimate development, and that the "Manichaean" thesis overstates a polemical connection. The casebook should present Wilson's thesis as a thesis, not a consensus.

What the casebook should not say: that Augustine "was a Manichaean" or that his doctrine is "Manichaean." He was a Manichaean auditor for roughly nine years, converted in 386, and spent his later life opposing Manichaean dualism. The historical argument is about influence and structure, not secret allegiance.

3. Orange (529) and the Western settlement¶

The Second Council of Orange is the most important Western document between Augustine and the medieval period. Its conclusion states:

"We not only do not believe that any are foreordained to evil by the power of God, but even state with utter abhorrence that if there are those who want to believe so evil a thing, they are anathema."

It also affirms:

"The sin of the first man has so impaired and weakened free will that no one thereafter can either love God as he ought or believe in God or do good for God's sake, unless the grace of divine mercy has preceded him."

What Orange does: it rejects Pelagianism (grace is necessary, and the beginning of faith is a gift); it rejects any foreordination to evil; it affirms prevenient grace and human cooperation.

What Orange does not do: it does not adopt Augustine's double predestination; it does not teach irresistible grace in the later Reformed sense; it does not use the later vocabulary of "unconditional election" and "reprobation."

How to use it: Orange is evidence that the Western church's settlement was not double predestination. It is not evidence that Orange was Arminian; it was not. It is a "both/and" document, and the casebook should present it that way.

4. Gottschalk (c. 808-867)¶

Gottschalk taught a double predestination: God predestined the elect to salvation and the reprobate to damnation. He was condemned at Mainz (848) and Quierzy (853), and his teaching was opposed by Hincmar of Reims. The controversy is often cited as proof that the church rejected double predestination.

Nuance: the condemnations were as much about Gottschalk's denial of human freedom and his ecclesial disobedience as about the doctrine itself; the Carolingian theologians were not all Augustinians, and the issues were tangled with politics. The safe claim is: Gottschalk's formulation was condemned, not the church definitively rejected all forms of predestination.

5. The East¶

The Eastern Orthodox churches have never accepted the Reformed doctrine of double predestination or a limited atonement. The Synod of Jerusalem (1672), under Dositheus, explicitly rejected Calvinism. Orthodox theology emphasizes synergia (cooperation) and theosis (deification), and it treats the Augustinian settlement as a Western development.

What this shows: the claim that "the church catholic has always taught the Reformed doctrine" is false as a historical claim. The East is a massive counter-witness.

What it does not show: that the East is right on every point. The East has its own controversies, and its rejection of Calvinism is not an exegesis of Romans 9. It is corroboration, not proof.

6. Where the casebook overreaches¶

"Unanimous." The pre-Augustinian fathers consistently teach free will and resistible grace, but they do not all address predestination, and some are ambiguous. Say "the consistent pre-Augustinian pattern" rather than "unanimous."
"Election according to foreknowledge" as the patristic doctrine. Several fathers say this (e.g., in their readings of Rom 8:29), but it is not a developed doctrine in every writer. Say "the dominant pre-Augustinian pattern" and cite the specific texts.
The Manichaean causal claim. Present it as Wilson's thesis, with the Reformed pushback, and do not build the case on it.
"The church rejected double predestination at Orange." Orange rejected foreordination to evil; it did not address every later formulation. Say exactly what it says.
Using the East as a trump card. The East's rejection is evidence of diversity, not a proof of the non-Calvinist reading.
Cassian as a witness. He was condemned; use him to show the debate, not to settle it.

7. What the history can and cannot prove¶

It can prove: that the Reformed doctrine of unconditional election and reprobation is a post-Augustinian development, not the apostolic deposit; that the pre-Augustinian church and the East did not teach it; that the Western settlement at Orange rejected foreordination to evil; and that the "historic view" claim is false as stated.

It cannot prove: that the Reformed reading of Romans 9 is wrong; that the fathers were Arminians; that Augustine was dishonest; or that novelty is heresy. History is corroboration, not exegesis.

The casebook's historical files should be read as saying: this doctrine has a history, and the history is not what the popular claim says it is. That is enough.

Sources¶

Justin Martyr, First Apology 43; Dialogue with Trypho.
Irenaeus, Against Heresies 3.18-19; 4.37-39.
Clement of Alexandria, Stromata 2.4.
Origen, De Principiis 3.1.
Tertullian, Against Marcion 2.5-6.
Methodius, On Free Will (fragments).
John Chrysostom, Homilies on the Gospel of John 45-47.
John Cassian, Conferences 13.
Vincent of Lérins, Commonitorium 2.
Second Council of Orange (529), canons and conclusion — text at andrews.edu and newadvent.org.
Augustine, Ad Simplicianum; De correptione et gratia; De praedestinatione sanctorum; De dono perseverantiae; Retractationes 2.1.
Ken Wilson, Augustine's Conversion from Traditional Free Choice to "Non-free Free Will" (Mohr Siebeck, 2018); The Foundation of Augustinian-Calvinism (2019).
Richard Muller, Calvin and the Reformed Tradition (2012), chs. on Augustine and Calvin.
Gerald Bonner, Freedom and Necessity: St. Augustine's Teaching on Divine Power and Human Freedom (2007).
Peter Brown, Augustine of Hippo: A Biography (rev. ed., 2000).