Harnack spends his first seven pages fighting a label, and it is worth being clear about which label and why.
The soft version was Bousset’s: that Marcionite dualism is a variety of Persian dualism. Harnack had rejected it in the monograph, and he notes with some satisfaction on his second page that his critics “have all approved, so far as they have entered into the question, of my rejection of Bousset’s thesis.” He restates the reasons on his seventh page. Marcion has bad matter alongside the world-creator; he does not identify the world-creator with the devil; and he “knows nothing of the dualism of light and darkness.” Whatever Marcion was, he was not a Zoroastrian in Christian dress.
The hard version was Walter Bauer’s, and Harnack takes it far more seriously, because it “returns to the older view (Marcion a Gnostic).” Bauer’s case, as Harnack summarizes it, is cumulative rather than single. The two-God conception with a subordinate Demiurge is borrowed. The “alien” God “does after all belong with the ‘unknown’ God and appears to be no wholly original creation”: Reitzenstein’s point. The anti-Jewish orientation “had various representatives in that age, both in the Great Church and above all among the Gnostics (especially among the Mandaeans — Reitzenstein).” And the world-hating, ascetic strand in Marcion’s piety, which Bauer thinks explains much of his appeal, comes from the same current. On this reading Marcion is not an originator at all. He is a man swept along, “less under the spell of Paul than gripped by a current of the age.” Then: “Thus far Bauer.”
It is worth saying that the ground under that argument has moved. Reitzenstein’s reconstruction of an Iranian mystery of redemption rested substantially on a text he took for a fragment of Zarathustra, which later philology identified as a Manichaean composition, that is, as evidence from a religion that came after Marcion rather than before him. Bauer was appealing in good faith to the best work of his day. The work did not hold.
Harnack’s own answer, when it comes on his nineteenth page, is a single criterion: “where the doctrine of the αἰῶνες and of divine προβολαί is lacking, there is no Gnosticism.” No aeons, no emanations, therefore no Gnosticism. And in a footnote on the following page he throws the challenge back: “The alien God in M. ought by itself to be a warning against this; I await the demonstration that he is no other than the unknown God, or that M. took him over from a predecessor.”
Right Conclusion, Weak Criterion
I think his conclusion is right and his method is not good enough.
The trouble is that a definition built on one feature can be adjusted until it delivers whatever verdict you want. Harnack picked aeon-emanation, which Marcion demonstrably lacks, and declared the question closed. Bauer could have picked world-negation, or the subordinate Demiurge, or the unknown God, and declared it closed the other way. Nothing in the procedure settles which feature is essential, because nothing had established that the category had an essence to begin with. Harnack won an argument by stipulation, and he was fortunate that his stipulation pointed at the truth.
How the Category Came Apart
What has changed since 1923 is not that someone found the missing aeons. It is that the category itself came apart in scholars’ hands. Michael Allen Williams worked through the traits usually said to define “Gnosticism” (inverse exegesis, anticosmism, hatred of the body, the choice between asceticism and libertinism) and found that they do not distribute consistently across the texts assigned to it. Individual clichés may fit this or that source, he concluded, but they “do not at all capture something essential or characteristic about the collection.” Karen King traced how the label had been assembled, showing that modern scholarship had often carried on the heresiologists’ work of marking off a contaminated Other so that normative Christianity could look pure.
It matters to say precisely what they did and did not do. Neither treated the ancient movements as fictions. Williams proposed replacing the label, where appropriate, with a deliberately modern and narrower one: “biblical demiurgical traditions,” meaning those that “ascribe the creation and management of the cosmos to some lower entity or entities, distinct from the highest God” and that also “incorporate or adapt traditions from Jewish or Christian Scripture.” That is a proposal to classify better, not to stop classifying, though he does want the old category abandoned rather than patched. King is more explicit still about the limits of her own case. Having said that no religion called Gnosticism existed in antiquity, she writes, “we still have to account for all the ideas, writings, persons, and practices described by ancient polemicists, not to mention the texts found at Nag Hammadi and elsewhere. If they are not Gnostic, what are they?” And her aim is not abolition: it is “not so much to eliminate the term per se, but to recognize and correct the ways in which reinscribing the discourses of orthodoxy and heresy distort our reading and reconstruction of ancient religion.”
What they exposed, then, was the category’s instability, its dependence on the classifications of men who were arguing rather than describing, and the sheer diversity it had been concealing. The movements gathered under the name were not all teaching the same thing, and the differences among them are frequently larger than the differences between some of them and Marcion.
So the question Harnack and Bauer were arguing about (was Marcion a Gnostic?) is not a question that can be answered well, because it is not precise enough to do historical work. What survives, and what is worth doing, is comparison on specified points. Six of them matter here.
Six Tests, Applied
Divine emanation. Marcion has none. This is Harnack’s point and it holds: no procession of beings out of the Father, no populated upper world.
A fall or rupture within the divine realm. Also absent. There is no Sophia, no error in heaven, no repair of a divine wound. The Creator is not a lapsed portion of the Good Father. He is simply another Power, and the Father owes him nothing.
Salvation through knowledge of one’s heavenly origin. This is the sharpest difference, and it is where the Apostolikon speaks most plainly. We are not told that we already belong above and have merely forgotten it. We are told that we were somewhere else and were taken out: “I still say, speaking in human terms, that while we were children, we were enslaved under the elements of the world” (Galatians 4:3 TT). The change is not recognition but transfer of ownership, and Paul is careful about which direction the knowing runs: “now, having come to know God, or rather to be known by God” (Galatians 4:9 TT). Being known precedes knowing. And the warning is against exactly the sort of speculative confidence a knowledge-of-origins scheme produces: “See to it that no one takes you captive through philosophy and empty deceit, according to human tradition and the elementary principles of the world, and not according to Christ” (Colossians 2:5 TT).
Fixed spiritual classes by nature. Harnack saw this clearly and put it in his list of the ideas that endure. Marcion divides men into bad, righteous and good, “but without the Gnostics’ assumption that these classes are given by constitution and that the individual belongs by nature and forever to one of them.” The Apostolikon agrees, and its grammar does the work. Paul sets out the two men: “The first man is from the earth, of dust; the second is the Lord from heaven” (1 Corinthians 15:39 TT), and then says what follows not as a description of who we already are but as something to be done: “And just as we have borne the image of the one of dust, we should also bear the image of the heavenly one” (1 Corinthians 15:41 TT). Should. Not do. The heavenly image is a destination, and the verdict on the present condition is universal, not sorted by class: “flesh and blood will not inherit the kingdom of God, nor does decay inherit incorruption” (1 Corinthians 15:42 TT).
A revealer who awakens a hidden identity. On the record we have, the rescue does not work by reminding anyone who he really is. The rulers of this aeon do not fail because they lack self-knowledge; they fail because they did not recognize the one they were killing: “if they had, they would never have staked the Lord of glory” (1 Corinthians 2:8 TT). Ignorance is on their side of the ledger, and it is defeated by an event, not by disclosure of a secret self.
The relation between Creator, matter, and evil. Here Marcion is closest to the movements he is compared with, and I am not going to pretend otherwise. He has bad matter. He has a Creator who is not the highest God. He has a judgment on the world so severe that Bauer’s instinct to file him alongside the world-haters of his century is not foolish. This is the point where the comparison genuinely bites, and it is the subject of a later section.
Two Honesties
Two further honesties are owed. At least some Marcionites were apparently willing to use the word, and Harnack quotes the evidence himself in the footnote where he issues his challenge. Clement, describing the ascetic practice of “those of Marcion’s following,” says they proclaim “the alien knowledge, as they say”, τὴν ξένην, ὥς φασι, γνῶσιν εὐαγγελίζονται. Clement’s little qualifier is a way of holding the word at arm’s length, which is itself the sign that ξένη was their term and not his. It should be said that he is reporting his own contemporaries rather than Marcion. And Marcion shared opponents, a vocabulary, and a good deal of world-negating judgment with teachers no one hesitates to place under the old label.
But the direction of that sharing is less obvious than Bauer assumed. Christoph Markschies has asked whether the impulse directed against Marcion may have mattered far more to the emergence and shaping of the Valentinian systems than has been recognized, questioning the habit of reading Valentinian teaching as a reform of some prior classical Gnostic scheme. His larger suggestion is that an inventory of the references to Marcion, hostile and admiring, would let us understand the whole of so-called Gnosis better. If he is right even in part, then Marcion is not downstream of these movements. Some of them are downstream of him.
None of that makes the label useful. It makes the comparisons useful, and the comparisons are what Harnack was reaching for when he settled for a slogan about aeons. Note too that Williams’s replacement term would not exempt Marcion: a Creator distinct from the highest God, reached through Jewish and Christian scripture, is “biblical demiurgical” almost by definition. Which is the right result. Renaming a category was never supposed to hand anyone an acquittal; it was supposed to make people say what they actually meant.
What the tradition we have received is cannot be settled by what it is not. It is defined by what it confesses: a Good Father who was under no obligation to us, and a rescue out of a regime we did not choose. That confession has to be argued from the Evangelion and the Apostolikon, which is where this afterword arrives once Harnack’s own frame has been dealt with. But the ground is clearer once the old label is off the table, and Harnack deserves the credit for getting it there, even by a route that would not survive review today.