There is one sentence in the pamphlet that Harnack does not argue for. It is on his eighteenth page and it is stated as though everybody already agreed: “For my part I need not repeat that M. gravely misunderstood Paul and would have been rejected by him with the deepest indignation, for I have expressed that in the plainest way in my presentation.”
He need not repeat it, and so he does not defend it. That is the whole of the case.
What makes this worth pausing over is the company the sentence keeps. It sits in the middle of a paragraph whose other business is to list what Marcion demonstrably got from Paul, and the list is not short. Marcion “gained his concept of faith, the separatio legis et evangelii, the fateful nature of the νόμος, the significance of the death on the cross and the nova creatura in Christ from the apostle’s letters.” Harnack calls this “incontestable.” Two sentences later he refuses Bauer’s suggestion that Paul became significant for Marcion only secondarily, and refuses it with feeling: “that Paul whose Gospel was for him the Gospel, and whom his adherents saw enthroned at Christ’s right hand!”
The Undefended Sentence
So the position is this. Marcion took his doctrine of faith from Paul, his separation of law and gospel from Paul, his understanding of the law’s fatality from Paul, his valuation of the death on the stake from Paul, and his new creation from Paul; Paul’s gospel simply was the gospel for him; and his followers pictured Paul beside Christ in glory. And he gravely misunderstood him.
I am not going to claim that Marcion read every part of Paul correctly. He did not, and one of the places where he did not is coming up. The narrower point is that Harnack has not earned the sweeping judgment, and that a historian who lists five substantial debts and then asserts a grave misunderstanding owes his reader an argument about which parts were misread and how.
The reason he thinks he does not owe one is that the misunderstanding seemed self-evident from inside his frame. His critics had already set the terms. Bauer had argued on his fifth page that an experience like Paul’s, the recognition of the decisive antithesis of law and grace, is “altogether inconceivable in the Gentile Christian M.,” and inconceivable as a starting point, since Marcion’s protest is not against law as such but only against the Old Testament law. Grejdanus had gone further on the pages before. Marcion does not know grace at all, he says, because in Paul grace is love that forgives guilt, and in Marcion sin is not sin against God. No guilt, therefore no forgiveness, therefore no grace. Both arguments assume that the thing to be found in Paul is a doctrine about the guilty individual. Harnack answered them on that ground and won on that ground, as the third section of this afterword has already argued. He then took the residue (everything in Marcion that would not fit) and called it misunderstanding.
But the residue is Paul’s own language, and it is language about Powers.
The Language of Powers
Consider what has to happen to a handful of sentences before the Marcionite reading of them becomes a misreading. The rulers of this aeon are said not merely to have been mistaken but to have acted in a decisive ignorance with a decisive consequence: “None of the rulers of this aeon understood it; for if they had, they would never have staked the Lord of glory” (1 Corinthians 2:8 TT). The condition before Christ is described as ownership by beings with a claim, and the danger of return is described as re-enslavement: “Formerly, when you did not know God, you were enslaved to those who are gods by nature. But now, having come to know God, or rather to be known by God, how do you turn again to the weak and poor elements, to which you want to be enslaved once more?” (Galatians 4:8–9 TT). And the letter that opens the collection announces the deliverance out of the present evil aeon before it makes any argument at all (Galatians 1:4 TT). That verse has been the hinge of this afterword since the third section and I will not re-argue it; I note only that it is where Paul begins, and that it names an aeon rather than a ledger.
Marcion read those sentences as meaning what they say. Rulers who did not know. Beings who owned us. An aeon we are taken out of. That reading is radical and it is not obviously right; the language can be read as vivid figure for a moral and existential predicament, and a great many careful readers have so read it. What it is not is a failure to understand the words. To call it that, Harnack has to treat the Powers as furniture, which is exactly what the scaffolding footnote on his twenty-fifth page does. Once the Powers are scaffolding, taking them literally looks like naivety. The charge of misunderstanding is not an independent finding. It is what his frame produces.
Reverse the frame and something shifts about what grace is for. If the predicament is guilt, then the remedy is forgiveness, and everything else (release, transfer, new creation) is metaphor decorating a pardon. If the predicament is captivity in an order that produces death, then forgiveness is one thing that happens in the course of a rescue rather than the rescue itself. The third section made the case that grace and faith in this collection describe the manner in which rescue is received, and I will not repeat it. What is worth adding is that the texts themselves put forgiveness inside the transfer rather than beside it. Colossians describes the Son as the one “in whom we have redemption, the forgiveness of sins” (Colossians 1:14 TT), and the sentence immediately before it is about being rescued from one realm and moved into another. Forgiveness is named, and it is named as a feature of what happened when custody changed. That is not a diminishment of forgiveness. It is a location for it.
The same holds of sin, and here the grammar does the arguing. Paul can write of sin as something a person does and is answerable for, and he can write of it as an occupying power with its own agency, and in Romans 7 he does both within a few lines. “Now it is no longer I that act, but sin dwelling in me” (Romans 7:12 TT). That is not a denial of responsibility; the same chapter is full of a first-person voice that knows what it is doing and hates it. It is the recognition that something is at work in the field that is not simply the self’s choosing. The fifth section already set out what the law belongs to, and I take that as read. The point here is only this: sin, law, flesh and death behave in these letters like a regime with jurisdiction rather than like items in an account, and a reader who noticed that was noticing something.
Perishing
Which brings me to the word that the Lutheran vocabulary cannot carry, and that this collection uses without embarrassment.
The condition of those outside the rescue is not, in the first place, guilt. It is perishing. “For the message about the stake is foolishness to those who are perishing, but to those who are being rescued it is the power and wisdom of God” (1 Corinthians 1:18 TT). Both participles are present and both are continuous. Nobody in that sentence is waiting for a verdict; two processes are already under way, and the announcement lands differently depending on which one you are in. Elsewhere Paul can use the same pair almost casually, describing the apostles as an aroma of Christ among those being rescued and among those who perish.
Perishing is a strange word to build a soteriology on, and its strangeness is the point. It is not a sentence handed down but a condition being undergone. It is ongoing, which is why the participle is present. It is embodied, because what perishes is flesh, and the promise addressed to it is not acquittal but a change of nature: “For this perishable must put on incorruption, and this mortal must put on immortality” (1 Corinthians 15:45 TT). It is social, because Paul addresses assemblies of the perishing and not individuals in a confessional. And it is cosmic in scale, because decay is the signature of the whole order, and the aeon we are delivered from is described as evil rather than merely as fallen.
Set the two vocabularies beside each other and the difference in what they diagnose is plain. A debtor needs the debt cancelled. Someone who is perishing needs to be got out. Both are true in these letters, and only one of them is the frame within which the other makes sense.
Where Harnack Is Right
Now the part of this argument where Harnack is right and his critics were wrong, and I want to say it without hedging, because the last two sections have been hard on him.
Grejdanus’s syllogism fails, and Harnack dismantles it properly. On his tenth page he points out that Marcion holds men responsible in two directions at once: to the creator God and to the good God for transgression of the ten commandments, since Marcion “did after all let the catechism-requirement stand as self-evident,” and to the good God for unbelief now that Christ has appeared. He then asks the question that ought to have ended the discussion: “am I to begin enumerating the number of Church Fathers in whom the misery of sin likewise stands more in the foreground than the guilt of sin? How does the matter stand with Athanasius?” If foregrounding misery over guilt disqualifies a man from Christian seriousness, the disqualification will not stop with Marcion. And he notes that Marcion said expressly, with Origen as witness on Romans 5:20, that the law was given so that sin might abound, which is a doctrine about sin, not an evasion of it.
The best of it is on his eleventh page. Marcion’s answer to the taunt “Why do you not simply sin away, if your God does not judge?” was, simply, “God forbid, God forbid.” Harnack sees what that is: an answer “derived from Rom. 6:1 f.,” and therefore attesting “not frivolity or indifference toward sin, nor embarrassment, but the deepest earnestness and a sheltering experience of grace.” A sheltering experience of grace. Whatever one concludes about the Creator, that judgment about Marcion is sound, and it is worth more than the pages of philology around it.
And in a footnote on his twelfth page he assembles the rest of the case: that a μεταβολή out of badness into being good is what redemption yields; that the description of the sinful woman as per paenitentiam ex fide iustificata is in the highest degree probably Marcion’s own; and that given the scope of what the good God frees from, “it is not surprising that in M. everything converged in the one mood: ‘Mercy has befallen me, mercy of which I am not worthy.’” A man in whom everything converges on that sentence has a consciousness of sin. The critics who read Marcionite grace as moral indifference were simply wrong, and Harnack is the reason we can say so.
He is also the reason we have to concede something about the asceticism, and the concession is his own. On his eleventh and twelfth pages he allows that Marcion’s description of the misery of sin rooted in the flesh, and his feeling of disgust at it, “may derive from a way of looking at things originally foreign to Christianity, which had representatives outside Christianity in his age”, adding that among later Christian monks and Neoplatonists one can find testimonies about the flesh and about procreation that “yield nothing to the Marcionite ones in realism and in feeling of disgust.” And in a footnote on his twenty-fifth page he adds the qualifier plainly: “the pressing of this stance toward the world to the point of abolishing sexuality was given by the age.”
That distinction is the right one and it should be accepted rather than argued around. The diagnosis (that flesh decays, that the world is not neutral, that this order produces death) is in the letters. The particular disciplines by which second-century Christians of every party expressed that diagnosis were the disciplines available to second-century Christians, and Marcion’s were not even the most extreme on offer. I take it that a living confession inherits the diagnosis. It does not inherit an obligation to reproduce every practice through which the diagnosis was once expressed, any more than the Great Church considers itself bound by the ascetic athletics of its own desert.
The Hard Case
And now the hard case, which I would rather not raise and which any honest treatment has to.
Paul grieves over his own people. The argument about Israel in Romans is the place where his attachment to the God of that people and to their history is least deniable, and it is the place where a Marcionite reading is under most strain. What can be established here is narrower than the assertions usually made in either direction, and the distinctions matter.
Most of Romans 9–11 is unattested for Marcion, along with Romans 1:17b and 3:31–4:25. That is a statement about what the witnesses report, not about what stood in his text. The step from there to deliberate excision is an inference: Harnack’s, in his reconstruction of the Apostolikon, and, as BeDuhn reports, Nils Dahl’s after him. Against it stands what is attested: Romans 10:3 is in the Apostolikon, cited there for the point that God was not truly known even to the Jews. And BeDuhn has made the argument that tells most heavily against reading the gaps as a programme. If the absence of the scriptural quotations in Galatians 3, Romans 9–11 and Romans 15 proves a Marcionite agenda, he asks, what is to be made of the many more scriptural quotations left standing: Deuteronomy at Galatians 3:10 and 3:13, Habakkuk at 3:11, Genesis at 4:22, Leviticus at 5:14, Isaiah, Jeremiah and Job in 1 Corinthians. All of them are quotations of Israel’s scriptures, and they stand in the Apostolikon as evidence of what the collection contained rather than as authority invoked by this afterword. The True Testament text confirms his point from the inside: the curse of the law in Galatians 3:9 TT still carries its Deuteronomy quotation, as the fifth section had occasion to quote. A censor who removed Israel’s scriptures from Paul made a poor job of it.
BeDuhn’s summary of the whole collection is worth having beside Harnack’s charge. On his account the Marcionites read essentially identical versions of eight of the ten letters, with only minor differences in Galatians and Colossians, and it is Romans alone that would leave a substantially different impression, adding that arguably nothing deserving mention in an overview of Paul’s core ideas would be missing from a summary based on the Apostolikon.
So the blunt claim that Marcion handled Paul’s grief for Israel by cutting it out is not available, and I do not make it. What remains is narrower and still uncomfortable, and it is my inference rather than anyone’s finding. Romans is the one letter that differs substantially, and it is in Romans that Paul’s sorrow over his kin is concentrated. So the Paul of this collection is a Paul whose grief over Israel is harder to hear than the Paul of the canonical letters. Whether that is because of Marcion, because of what he inherited, or because of what his opponents chose to quote, cannot be determined from the evidence I have seen. It is a real cost and it should be entered as one. It does not settle the question of who read Paul better. It does mean that a confession which takes the Apostolikon as it stands should know this about the text it has received.
That leaves the charge in a narrower place than Harnack left it. Marcion may well have strained Paul, selected from him, and mishandled him in places. But Harnack was not entitled to the judgment that he gravely misunderstood him, and he did not try to be. The Marcionite reading takes Paul’s language of hostile Powers, enslavement, transfer and rescue at something close to face value, and that is a reading rather than a blunder. Forgiveness, responsibility and inward transformation are all real inside it; Harnack proved as much, against critics who denied it. What changes is the order. Perishing names the condition. Rescue names the Father’s act. Guilt and pardon are inside that, not around it.
That reading produced not just a theology but a church, and for a century or more it was the Great Church’s principal rival. What happened when the two collided is the last historical question left.