| Origen (Comm. Matt) | Cited reference in your list | Gospel | Ammonian pericope # | Canon | Ammonian pericope range (from your file) | Why this stays (boundary-only rule) |
|---|---|---|---|---|---|---|
| MAT.COM 12,12 | Mt 6:24 | Matthew | 48 | 5 | Matthew 6:24 | The citation equals the start of an Ammonian unit (6:24). |
| MAT.COM 12,12 | Mt 16:24 | Matthew | 170 | 2 | Matthew 16:24–26 | The citation equals the start of an Ammonian unit (16:24). |
| MAT.COM 12,15 | Mt 16:20 | Matthew | 168 | 2 | Matthew 16:20–21 | The citation equals the start of an Ammonian unit (16:20). |
| MAT.COM 12,15 | Mt 16:17 | Matthew | 167 | 10 | Matthew 16:17–19 | The citation equals the start of an Ammonian unit (16:17). |
| MAT.COM 12,15 | Lk 9:21 | Luke | 95 | 2 | Lk 9:21–22 | The citation equals the start of an Ammonian unit (9:21). |
| MAT.COM 12,15 | Mt 10:5–6 | Matthew | 81 | 10 | Matthew 10:5–6 | The citation equals the start of an Ammonian unit (10:5). |
| MAT.COM 12,15 | Mt 10:17–18 (and the larger “10:17–23” you listed) | Matthew | 87 | 1 | Matthew 10:17–18 | The citation equals the start of an Ammonian unit (10:17). |
What can actually be shown from the Greek is not that Origen ever names “Ammonius,” but that his exegetical procedure repeatedly behaves as if the Gospel text has already been discretized into stable, re-locatable units whose incipits can be cited, whose internal contents can be handled as a block, and whose parallel relationships can be tracked without re-narrating the whole story. That is precisely the kind of reading economy an Ammonian apparatus presupposes and rewards.
The place where Origen comes closest to sounding like someone working with a synopsis is 12.15, where his Greek arranges Matthew, Mark, and Luke in explicit comparative sequence: «ὁ μὲν Ματθαῖος… ὁ δὲ Μᾶρκος… ὁ δὲ Λουκᾶς…». This is “parallel-column” rhetoric. It is then reinforced by a second kind of scholarly signal, textual criticism: «ἰστέον μέντοι ὅτι τινὰ τῶν ἀντιγράφων…». The combination—parallel alignment plus manuscript-aware variation—creates the kind of reading environment in which an Ammonian apparatus would be genuinely useful and naturally employed. Yet the Greek still never does what Letter to Theodore does: it never marks a boundary by “after this clause, and the next until…” nor does it indicate a “here the text adds, word-for-word,” nor does it treat the Gospel as a sequence of bounded pericopes that can be entered and exited by formula.
In Origen’s Commentary on Matthew, the most probative Greek evidence lies in the way the exposition is “parked” on lemma-like boundary phrases and then advanced by formulaic transitions. Origen regularly treats a short clause as a unit header and builds an extended discussion “under” it, rather than merely seasoning his argument with scattered verbal echoes. The key is that the Greek cues that trigger movement are the same cues that function as natural pericope hinges in continuous narration—openers like Ἐν ἐκείναις ταῖς ἡμέραις, Ἀκούσας δέ, Καὶ ἐγένετο ὅτε…, Καὶ ἐξελθὼν ἐκεῖθεν…, or the tightly bounded dominical logion that can serve as a self-contained heading. When these are treated as stable “addresses,” they become indexable. Your own reduction of the citation list to genuine Ammonian starts (e.g., Mt 6:24; Mt 16:17; Mt 16:20; Mt 16:24; and the associated Lukan start at Lk 9:21) is exactly the point: the citations that survive are not mid-stream verbal snips but unit-head verses that can carry a whole block of exposition. That pattern is hard to explain if Origen is navigating only by memory of a continuous roll; it is easy to explain if Origen is navigating with a segmentation logic in view—whether personally devised or inherited—that corresponds to the sort of pericope arithmetic the Ammonian tables formalize.
The Greek of “Theodore,” by contrast, is not merely consistent with unit segmentation; it is explicitly written in the register of apparatus use. The passage you quote foregrounds a sequence of editorial/technical operators: δι᾽ αὐτῶν τοῦ εὐαγγελίου λέξεων frames the procedure as refutation by controlled citation of “the Gospel’s diction,” then Ἀμέλει μετὰ τὸ… καὶ τὰ ἑξῆς ἕως… gives a locating instruction of the sort one writes when one expects a reader to find an insertion point by scanning for a distinctive boundary clause and then moving forward to a second boundary: “after ‘ἦσαν δὲ ἐν τῇ ὁδῷ…’ and the following up to ‘μετὰ τρεῖς ἡμέρας ἀναστήσεται.’” That is not the idiom of casual quotation; it is the idiom of textual address. The next operator makes this even plainer: Ὧδε ἐπιφέρει κατὰ λέξιν…—“here it adds, word for word”—signals that what follows is not a paraphrase but a mechanically reproducible excerpt. Then again, Ἐπὶ μὲν τούτοις ἕπεται… καὶ πᾶσα ἡ περικοπή introduces the decisive piece of technical vocabulary: περικοπή is not a random word for “passage” here; it is a label for a bounded textual unit that can be spoken of as “the whole pericope,” as something that “follows” in sequence, and as something whose extent can be assumed. The passage ends by returning to the same apparatus posture: Μετὰ δὲ τὸ… ἐπάγει μόνον…—again “after X it brings only Y,” i.e., post-boundary control of what the text contains and does not contain.
That difference of register matters for the question “stronger/weaker witness.” On one side, Origen’s evidence is largely inferential: he behaves as if the text is segmented, and in your curated set he cites precisely those verses that are Ammonian unit starts. That is good evidence that his practice is compatible with Ammonian sectioning, and—importantly—compatible in a way that is difficult to attribute to chance once the boundary-only filter is applied. Yet it still stops short of proving dependence on a named apparatus, because a disciplined commentator can independently develop “unit headings” simply by the needs of teaching and memory, especially when the headings are obvious narrative hinges.
On the other side, “Theodore” provides a more direct witness to apparatus-like employment because it speaks the language of textual navigation: μετὰ τὸ…, τὰ ἑξῆς ἕως…, Ὧδε ἐπιφέρει κατὰ λέξιν, πᾶσα ἡ περικοπή. The author is not merely expounding; he is instructing a reader how to locate, delimit, and verify a sequence in a Gospel text, and he does so by naming boundary clauses and treating the intervening material as a unit. If one asks, narrowly, “Which text more plainly presupposes a technology of segmentation?” the Greek of “Theodore” reads more like a user manual for one.
But if the question is instead “Which is the stronger witness to Origen’s use of Ammonius specifically?” the balance shifts. “Theodore” is explicit about pericope handling, but it never supplies Ammonian section numbers, never gestures to a tabular cross-Gospel system, and the locating is purely intra-Markan (a placement after one Markan clause and before another). That can be done without Ammonius: any reader can cite “after X… until Y…” as a practical locating device. Origen, by contrast, sits exactly in the cultural and scholarly niche in which an Ammonian-style apparatus is functionally valuable—continuous commentary that repeatedly needs stable starting points and controlled parallel alignment. When Origen’s citations fall on boundary incipits that match Ammonian starts, and when his exegesis proceeds as if those starts are the appropriate “addresses” for discourse, the testimony is weaker as explicit evidence but potentially stronger as contextual evidence of participation in the same scholarly infrastructure that later becomes visible as numbered sections and canons.
So, in strictly philological terms: “Theodore” is the clearer Greek witness to pericope-addressing behavior (because it actually says περικοπή and uses boundary operators as locating instructions), while Origen is the more plausible witness to systematic scholarly use of segmentation in a cross-Gospel environment (because his commentary practice, especially when your citation list is pruned to Ammonian incipits, aligns with what an Ammonian apparatus is for). “Theodore” looks more overtly “apparatus-driven” on the surface; Origen looks more deeply “apparatus-compatible” in method. Whether that makes Theodore “stronger” depends on what you want to prove: stronger for the existence and intelligibility of pericope-based navigation in the Clementine/Markan discourse; weaker for pinning that navigation to Ammonius rather than to a more general late-antique habit of delimiting and indexing Gospel units.
