stephan huller's observations: “Marcion Is Bullshit”: Adamantius, Maximus, and the Reuse of a Heresiological Debate

All 64 units of Eusebius, Praeparatio evangelica 7.22, compared with the Valentinian episode of the Dialogue of Adamantius, with Greek texts, printed line references, and implications for the independence of the Marcion witnesses.

The comparison supplies a concrete demonstration that heresiological evidence can be made out of inherited literature. Substantial stretches of the philosophical debate attributed by Eusebius to Maximus reappear in the Valentinian discussion of the Dialogue of Adamantius, with changes of speaker, abbreviation, relocation, and explicitly sectarian additions. The opponent's contributions are shared too. What appears to be a named heretic's testimony can therefore be part of the same transmitted argumentative script.

The strongest conclusion: these versions cannot be counted as independent witnesses to a debate or to its participants' doctrine. Under the conventional account of dependence, the extant Dialogue adapts material from Methodius. The precise direction among Methodius, an earlier dialogue, and the material called “Maximus” remains disputed. The collation establishes the literary relationship without needing to settle that dispute.

The polemical title “Marcion Is Bullshit” is most defensible here as a challenge to the practice of turning literary constructions into independent historical data. This passage does not establish that Marcion never existed, that no Marcionite codex was consulted anywhere, or that every biblical citation in the Dialogue is invented. It establishes why those propositions must be tested against the composition and transmission of each passage rather than inferred from the dramatic setting.

What is shared, and what is not yet established

Witness	Location and attribution	What it establishes
Eusebius	Praeparatio evangelica 7.21.5 introduces Maximus's On Matter; 7.22.1–64 gives the excerpt.	A long continuous version of the argument, without the named Droserius–Valens–Eutropius staging of the extant Dialogue.
Methodius	De autexousio / On Free Will 5.1–12.8 overlaps PE 7.22; its earlier and later context also overlaps the Dialogue.	The relationship extends beyond the limits of Eusebius's quotation. The opening opponent is conventionally designated “Valentinian” in the edited and translated tradition (Greek abbreviation ΟΥΑΛ.); the underlying discussion is already a dialogue.
Dialogue of Adamantius	Section IV, especially GCS 4, pp. 146–162; the wider Valentinian episode is pp. 136–168, legacy pagination 840a–849b.	A version distributed among Adamantius, Droserius, Valens, and the judge Eutropius, with a purported reading from Valentinus and a triumphal dismissal.
Philocalia 24	The text is explicitly taken from Eusebius; its closing notice also identifies the material with a dialogue attributed to Origen.	Evidence that readers recognized the overlap. It is not another independent quotation of Maximus, since it identifies its immediate source as Eusebius.
Rufinus's Latin	The Latin translation printed opposite the Greek in Bakhuyzen, here chiefly IV.1–11.	An essential control on Greek wording, speaker allocation, omissions, and editorial supplements. It cannot simply be treated as a literal duplicate of the surviving Greek recension.

The date of Maximus. The user's association with Commodus belongs to an ancient chronological tradition: Jerome, De viris illustribus 46–47, places Heraclitus under Commodus and Severus and Maximus under the same emperors. Eusebius, Historia ecclesiastica 5.27, is less precise and explicitly says that the dates and histories of the writers he lists cannot be fully supplied. Accordingly, “late second or early third century in the ancient notices” is safer than treating a date under Commodus as independently verified. Identifying this Maximus with Maximus of Tyre is a further, speculative proposal.

Direction of dependence. Robinson (1893, pp. xl–xlix) and Bakhuyzen (1901) argued for Methodius's priority and an error in Eusebius's attribution. Barnes's 1979 reconstruction reverses the customary priority, making Maximus the author of an earlier Adamantine dialogue and Methodius a borrower; the further explanation of Eusebius's attribution requires additional hypotheses. Carriker (2003, pp. 225–227) reviews the problem and favors a connection between Eusebius's wording and Methodius. Ramelli (2020, pp. 42–44, 56–57) reopens the case for an earlier dialogue or source behind Methodius. Consequently, “Adamantius copied Eusebius” is more specific than the demonstrated evidence warrants. “The same discussion is transmitted in several differently framed forms” is secure.

Methodius is not a third complete independent Greek manuscript. His work has a composite transmission. Robinson distinguishes the direct Greek fragment first published by Meursius in 1619, covering roughly the first quarter of the Eusebian extract with its preceding context, from Photius's abridged presentation of the whole discussion (Bibliotheca, cod. 236), other excerpts, and the Slavonic version. Modern reconstruction also draws on the parallel witnesses. The appearance of the full passage in an edition of Methodius must therefore not be used circularly as if every word had independent continuous Greek attestation there.

The purported written source: a definition of Valentinus

The most revealing part of the episode begins before Eusebius's excerpt. Droserius enters while Marinus's discussion is in progress, declares himself an adherent of Valentinus, and offers a written definition to be read. The wording makes the documentary claim explicit: ἔχω τὸν ὅρον Οὐαλεντίνου καὶ ἀναγινώσκω—“I have Valentinus's definition, and I read it” (136.18). Eutropius then orders the doctrine of Valentinus to be read (136.24). This is a representation of access to an author's document inside a literary dialogue.

Adamantius location	What the episode does	Control from the related discussion
136.11–24	Introduces Droserius, labels his allegiance, and stages the production of Valentinus's definition.	The named entrance and the order to read create the evidentiary setting; they must be distinguished from the speech that follows.
136.25–138.24	The purported reading begins with a first-person narrator returning home, then witnessing violence, robbery, and sexual wrongdoing.	The same first-person narrative is in Methodius's earlier context, On Free Will 3.1 ff. A dramatic participant's narrative in that version becomes, in Adamantius, a document attributed to the sect's founder. This is not part of PE 7.22 itself.
140.1–21	The narrator asks where evil comes from and refuses to attribute it to a good God.	The same sequence belongs to the related account in Methodius. The identity concerns the construction of the problem, not merely a shared belief that God is good.
140.21–142.7	Coexistent matter explains evil; God orders its better elements and leaves the dregs.	The same solution supplies the opponent's position in Methodius. The later speech of Valens at 156.15–19 appeals back to this definition, making the reused narrative function as an internal documentary authority.
142.8–144.5	Droserius praises the definition, pledges to defend it, and exchanges statements of goodwill with Adamantius.	Methodius's corresponding context has a friendlier conversation. Adamantius retains phrases of welcome while explicitly addressing Droserius and challenging his position.
144.6–26	Droserius invokes Genesis 1:2, the judge orders a reading, and Adamantius exposes his supposed failure to read Genesis 1:1.	This scriptural contest is absent from PE 7.22. It makes the opponent's defeat visible before the long inherited philosophical refutation starts.
144.27–146.14	The separation of God and matter is debated in brief turns, including an intervention by Eutropius.	It anticipates the substance of PE 1–6, immediately before Adamantius restarts the longer form at 146.15. The resulting repetition is a compositional seam.
162.4–168.5	The debate moves to evil as the manner of acting, followed by an explanation of human free choice.	The overlap with Methodius continues beyond the end of Eusebius's extract: Bakhuyzen's introduction, p. xxxviii, maps this stretch to Bonwetsch's 1891 Methodius, pp. 43–48. PE 56–64 are not the ending used here.
168.6–22	Eutropius celebrates the church's inexhaustible truth, dismisses Droserius and Valens, and returns to Marinus's remaining questions.	The restoration of the interrupted discussion frames the Valentinian material as an inserted episode. The judge also acts as a partisan supplier of refutations inside the episode.

Under the conventional dependence model, this is a particularly strong case of an inherited dramatic speech acquiring the status of a founder's written declaration. If priority is reversed, the comparison still prevents the speeches from counting as independent records: it then shows the removal or alteration of their original setting. The verbal relationship alone cannot decide whether any genuine Valentinian material ultimately lies behind the shared source. It does show that the reader's immediate impression of documentary access is not self-authenticating.

The argument changes order and changes mouths

Using Eusebius's numbered paragraphs as a reference, Adamantius's main course is: 1–15 → material corresponding to 35–55 → material corresponding to 23 and 29–34 → further material parallel to Methodius outside PE 7.22. This is an alignment of surviving versions, not a reconstructed archetype. The unmatched sections must remain visible.

Change	Where to check	Why it matters
A conversational recollection becomes a reference to a reading.	PE 7.22.14 / Adamantius 152.4–5.	The text itself adjusts to the documentary frame: something said earlier becomes something read earlier.
A second discussant becomes the named Valens.	PE 7.22.35–38 / Adamantius 152.16–24.	Different answers to a philosophical problem can be staged as disagreement among named followers of one teacher.
The judge supplies part of the defender's argument.	PE 7.22.40 / Adamantius 154.4–13; PE 49–55 / 156.23–158.22.	The appearance of an independent adjudicator does not create a new evidentiary voice when his reasoning belongs to the common literary discussion.
“Your argument” becomes “Valentinus's argument.”	PE 7.22.55 / Adamantius 158.20–22.	A sectarian name can reside in the adaptation while the philosophical sentence remains inherited.
The substance–activity proof changes position.	PE 7.22.29–34 / Adamantius 160.7–162.3.	The same unusual reasoning is placed after the refutation of Valens. This is larger than a commonplace shared metaphor.

Controls against manufacturing the parallel ourselves

The collation uses all 64 numbered paragraphs of PE 7.22 as its units. Some contain more than one sentence or question-and-answer turn. Every unit is accounted for, including those absent from Adamantius. The attached Adamantius references give printed page and line in Bakhuyzen's GCS 4, rather than the PDF's image number. This is a complete comparison by numbered unit, with line references; it does not pretend that Eusebius and Adamantius have identical lineation.

The Eusebian Greek follows the accessible numbered transcription, checked in sense against the public-domain Gifford translation. The Adamantius Greek is extracted from the GCS-based digital transcription and checked at decisive points against the 1901 scan and its apparatus. This is a source-critical working collation, not a new critical edition of either work. Orthographic, accentual, and minor inflectional differences remain visible. No numerical similarity score is offered, since a percentage would conceal the effects of omissions, repeated questions, editorial supplies, and altered speaker divisions.

Bakhuyzen's angle brackets indicate additions and square brackets exclusions. Parentheses mark passages present only in the Greek or only in the Latin; an asterisk records his judgment that the passage belongs to the original Greek. Parentheses therefore do not by themselves mean that he reconstructed the words from Eusebius. In particular, 160.15–20 is Greek manuscript material absent in Rufinus, while the words at 158.16 about evil not having ceased were actually supplied from the Philocalia. The row notes distinguish these cases. The broken occupational example at 160.9 is also a manuscript lacuna, not a demonstrable authorial abbreviation.

The Philocalia notice needs its own check. Its closing statement calls the passage word-for-word material in a dialogue attributed to Origen, with Εὐτροπίου δικάζοντος, Μεγεθίου δὲ ἀντιλέγοντος: “Eutropius judging, and Megethius opposing.” The accessible 1911 English rendering says “defending” for Eutropius; the Greek means judging. Eutropius's role is therefore not the disagreement. The opponent's name is: the extant parallel is spoken against Droserius and Valens, not Megethius. Robinson already noted both the overstatement of word-for-word identity and the mistaken opponent (pp. xlvii–xlviii). Misremembering is a sufficient possible explanation; a different recension is possible but is not established by this notice alone.

Complete comparison: PE 7.22.1–64

Each numbered unit gives the Eusebian argument in English, the location and speaker of its counterpart in Adamantius, and the significant verbal or structural differences. Open the Greek panel to inspect the complete Eusebian unit beside the relevant printed Adamantius lines. Adjacent panels sometimes repeat an Adamantius line because the Eusebian paragraph boundary falls within it. This is overlap in display, not a claim of two independent matches.

Print the concise analysis with panels closed, or open them for the full collation.

PE 7.22.1Close wording

Adamantius: 146.15–19 · Speaker(s): Adamantius

Two uncreated realities must be either separate or inseparable.

The opening is substantially the same, with a different word order and a finite verb where PE has an infinitive. This is the beginning of a continuous verbal parallel, not merely agreement on divine unity. The printed traditions also differ in spelling uncreated with one or two nus.