The Reconstructed Baseline Gospel Text (*Ev)

The UPDV uses *Ev as the abbreviation for a reconstructed baseline gospel text: an early form of the gospel tradition that survives only through later testimony. The technical name behind the abbreviation is Evangelion ("Gospel"), but this article uses the plainer phrase baseline gospel text wherever possible. The asterisk is a warning label. It means the text is reconstructed, not preserved in a surviving manuscript.

This baseline text is often called "Marcion's Gospel" in older scholarship because Marcion's movement preserved and used it, and because his opponents discussed it under that label. That label can mislead. The point now under review is not that Marcion wrote a gospel or created one by cutting up canonical Luke. On the reconstruction the UPDV is considering, Marcion inherited and used an already-existing baseline gospel text. In his canon, it was simply the Evangelion.¹

The ancient witnesses to this baseline text are mainly opponents of Marcionite Christianity: Tertullian's Against Marcion, especially Book 4; Epiphanius's Panarion, especially the material on Marcion in section 42; and the Dialogue of Adamantius. These sources preserve real textual information, but indirectly and polemically. They do not give us a manuscript of *Ev itself.²

🔗Why the Baseline Text Matters

For many centuries the standard explanation was simple: canonical Luke came first, and Marcion shortened it to fit his theology. That is the old "Marcion cut Luke" model.

Recent reconstruction work has challenged that model. Matthias Klinghardt argues for the strongest reversal: canonical Luke is an expansion of the gospel text used by Marcion, and that same pre-canonical gospel text also stands upstream of the other canonical gospels.³ Jason BeDuhn argues that Marcion did not substantially edit Luke into this text, but sanctioned a gospel text already in existence; he treats Marcion's text and canonical Luke as related textual forms whose priority cannot simply be assumed from the later patristic accusation.¹ Dieter Roth reconstructs more cautiously, limiting reconstruction to what the ancient witnesses directly attest or securely imply.⁴

This challenge did not begin with recent scholarship. John Knox argued in 1942 that the evidence does not establish the simple opposite of the old charge — that Marcion merely abridged our Luke — and he described the Gospel section of Luke-Acts as an enlarged form of Marcion's Gospel, while still allowing for a more complex primitive gospel behind both forms.⁵ Markus Vinzent and Klinghardt press the source-history question beyond Luke: Vinzent argues that the authors of Matthew, John, Luke, and Mark took material from Marcion's Gospel, while Klinghardt argues that canonical Luke edited and supplemented *Ev and that all four canonical gospels depend on *Ev in their pre-canonical textual forms. That shared source-history pressure should not collapse their models: Vinzent further argues that Marcion personally authored *Ev, while Klinghardt and the UPDV operate with the inherited-text paradigm stated above.⁶

The UPDV does not need every scholar to state the same model in order for *Ev to matter. The important point is narrower: this reconstructed baseline gospel text may preserve evidence from a stage of gospel transmission earlier than the surviving manuscripts. That is the kind of evidence the UPDV's eclectic method is designed to consider. The UPDV already weighs textual, linguistic, historical, versional, and patristic evidence in order to recover the earliest attainable form of the text. *Ev belongs in that evidence stream when the reconstruction is strong enough. These source-history arguments also mean that canonical Mark is not automatically treated as an untouchable primitive bedrock; however, that is a source-critical judgment, not a claim that every disputed Markan unit is directly omitted by an extant witness or by *Ev itself.

🔗The Policy

*Ev is not a replacement for UBS6, and it does not move the UPDV text by itself. It is a reconstructed upstream witness. It can bear on a UPDV decision only when it converges with another line of evidence, such as internal coherence, versional evidence, patristic testimony, or a source-critical pattern across the gospels.

Convergence is necessary but not sufficient. *Ev bears on the printed text only where two conditions both hold: the material creates a genuine problem — a contradiction with another account, or an inability to cohere with the work it sits in — that ordinary explanations (a known scribal variant, a reasonable harmonization, an ordinary literary feature) do not already resolve; and a post-*Ev editorial layer is the best or only remaining explanation for that problem. Absence from *Ev alone is never a reason to change the text, and a passage being merely harsh, unusual, or without parallel is not, by itself, a problem of the kind meant here.

The first question is always how secure the reconstruction is. A reading supported by multiple independent ancient witnesses carries more weight than a reading inferred from one hostile witness. A structural absence reported consistently across the main witnesses is not the same kind of evidence as a single uncertain word variant. Disagreement among modern reconstructions also matters: where Klinghardt, BeDuhn, and Roth diverge, the UPDV should not write as though *Ev is plain.

Harnack's 1924 data remains useful at this point because it organizes the ancient reports verse by verse. It is a data foundation, not a mechanical verdict.⁷ Harnack's labels help define the evidence, but the UPDV still has to decide whether a given case is direct attestation, qualified silence, source-critical inference, structural tethering, or some combination of these.

A single, late patristic attestation that posits the existence of a verse may be critically discounted based on the facts and circumstances. Such discounting must be stated explicitly in the research record. Silence by itself still does not establish absence, but qualified silence may be weighed when the sources otherwise discuss or attest the surrounding sequence closely enough that noncitation is probative rather than accidental.

The UPDV may therefore recognize a critically inferred structural absence. This category applies when the total evidence indicates that a passage, sequence, or component belongs to a post-*Ev editorial layer even though no surviving ancient witness states the absence in exactly those terms. Relevant factors may include a discounted positive attestation, expected-context silence, disagreement or caution among modern reconstructions, source-critical dependence, internal narrative coherence, chronological or literary disorder, and manuscript or versional disturbance. These are factors to be weighed case by case, not mechanical checkboxes.

Structural tethering and composite-seam analysis are applications of this broader inquiry. The UPDV may evaluate a neighboring component with an attested or inferred absence when the component depends on the larger unit for its function or sense. It may also evaluate a disturbed sequence as a whole when the evidence suggests that layers have been combined, even if every word of the proposed seam is not separately attested absent. The research record must explain why the proposed unit should be judged together, why ordinary explanations do or do not resolve the problem, and how each major reconstruction or witness has been weighed.

For this policy, an "ordinary explanation" resolves the problem only when it explains the passage as coherent within the reconstructed baseline text. It does not resolve the problem merely by describing how a later canonical editor could have combined sources, duplicated a saying, or arranged inherited traditions. If the explanation itself requires a post-*Ev editorial layer, then it belongs on the evidence side of the scale rather than functioning as an objection to the textual decision.

The second question is what kind of evidence *Ev is providing:

• A word-level reading can bear on a word-level textual decision, especially in Luke, where *Ev is normally compared against canonical Luke.
• A structural absence, critically inferred structural absence, or different sequence can bear on larger editorial-layer questions, including where material may have entered the gospel tradition.
• A structural-tether or composite-seam diagnosis can bear on a unit or sequence where the evidence suggests that layers have been joined or rearranged, even if every component does not have the same evidentiary status.
• A source-composition argument can bear on Mark or Matthew where the issue is not a direct *Ev omission but a later canonical layer inferred from the wider history of gospel formation. The research record must label this clearly as source-critical inference upstream of the extant manuscript tradition.
• A weak or uncertain reconstruction may still be worth recording in the research trail without affecting the printed text.

The outcome is therefore case by case. Classification guides the decision, but it should not become a mechanical bar when the evidence is strong enough to decide the question more generally. Weak or uncertain reconstruction may be recorded without changing the text. Conversely, when Tier A structural evidence—an unambiguous structural absence reported consistently across the main ancient witnesses, a critically inferred structural absence satisfying the case-by-case rule above, a structurally tethered absence, or a composite seam—establishes, with independent convergence, that material belongs to a post-*Ev editorial layer, research-record only and footnote-only are not available outcomes. In that case the text itself must carry the decision—through exclusion, brackets only where the UPDV's existing bracket-inclusion criteria already apply, or a larger Gospel-level arrangement as the case requires—and the footnote or article documents the editorial act rather than substituting for it. Material being an evangelist's own composition is not, by itself, a reason to retain it; the policy question is whether the material belongs to the reconstructed baseline gospel text or to a later editorial layer.

🔗How UPDV Footnotes Cite It

Where a UPDV textual decision rests in part on this reconstructed baseline gospel text, the footnote may cite it with the abbreviation *Ev. That citation means only this: the reconstructed baseline text was part of the evidence considered. It does not mean that *Ev alone decided the reading, and it does not mean that every detail of the reconstruction is certain.

Decisions that affect whole verses or larger units are also recorded in Summary of Excluded Passages and Variant Exceptions, as with other major UPDV textual decisions.

🔗Modern Reconstructions and Source-History Works

The UPDV consults the main reconstructions and source-history works in this discussion, while distinguishing their different claims and degrees of caution:

• Adolf von Harnack, Marcion: Das Evangelium vom fremden Gott: Eine Monographie zur Geschichte der Grundlegung der katholischen Kirche, 2., verbesserte und vermehrte Auflage (Leipzig: J. C. Hinrichs'sche Buchhandlung, 1924).
• Adolf von Harnack, Marcion: The Gospel of the Alien God, trans. John E. Steely and Lyle D. Bierma (Durham, NC: Labyrinth Press, 1990; English translation of the 1924 second German edition).
• John Knox, Marcion and the New Testament: An Essay in the Early History of the Canon (Chicago: University of Chicago Press, 1942).
• Jason David BeDuhn, The First New Testament: Marcion's Scriptural Canon (Salem, OR: Polebridge Press, 2013).
• Markus Vinzent, Marcion and the Dating of the Synoptic Gospels (Leuven: Peeters, 2014).
• Dieter T. Roth, The Text of Marcion's Gospel (New Testament Tools, Studies and Documents 49; Leiden: Brill, 2015).
• Matthias Klinghardt, The Oldest Gospel and the Formation of the Canonical Gospels, 2 vols. (Biblical Tools and Studies 41; Leuven: Peeters, 2021).

The policy is deliberately cautious. *Ev matters because it may preserve evidence earlier than the manuscript tradition now available to us. It is also reconstructed, indirect, and contested. The UPDV treats it neither as unusable nor as controlling. It is evidence to be weighed.

---