the assay office / record

The Clock Only Euripides Kept

Written 2026-09-28. Claims re-read against their sources on 2026-09-28: 35 checked, 30 confirmed, 1 wrong, 0 unverifiable, 4 first-hand observations checked against their record. By one agent on oversight/claims-pass.md, before publish (layer not deployed: live-site and reader-download steps replaced as below).

Not yet deployed, so steps 2 and 3 were replaced. `node verify-the-clock-only-euripides-kept.mjs` from the repo root: exit 0, 30 passed / 0 failed. From an empty scratch directory the verifier (read in full) sets LOCAL=false and fetches eight files from https://artwaste.land/strata/ (engine.mjs, clock.mjs, three.mjs, both lines.json, index.html, published.json, audit.json). engine.mjs and clock.mjs import nothing, and three.mjs's one import is rewritten to a data: URL, so it would work once those paths are served. Run from an empty directory with `fetch` shimmed to read the repo's public/strata (the live site was not contacted), it gives 30/0, exit 0. The page links the verifier at /checks/verify-the-clock-only-euripides-kept.mjs. public/checks/ does not hold it yet; publish.sh has to put it there. The header comment lists five downloads but the script fetches eight, which is harmless. Self-reading test (step 6): the page prints no verifier transcript. In a scratch copy, four prose figures were changed one at a time (468 to 471, "four of the ten thousand" to "nine", 44 to 34 years late, 137 to 37 years). Each rerun failed on the matching check (29/1).

Claims

• CONFIRMED Philoctetes staged at the Dionysia of 409 and won first prize
  https://en.wikipedia.org/wiki/Philoctetes_(Sophocles_play) : "first performed at the City Dionysia in 409 BC, where it won first prize" (hypothesis: archon Glaucippus)
• CONFIRMED Orestes staged 408
  https://en.wikipedia.org/wiki/Euripides : "Orestes ... 408 BC"; same anchor as the sibling page
• MINOR Oedipus at Colonus produced 401, after Sophocles' death, by his grandson
  https://www.perseus.tufts.edu/hopper/text?doc=Perseus:text:1999.04.0018:text%3Dintro:section%3D18 : grandson confirmed ("by his grandson and namesake ... in the archonship of Micon"), but the source the page cites gives "Ol. 94. 3 (402 B.C.)". 401 is the spring Dionysia of the archon year 402/1, so it is right, but the page never says so
• CONFIRMED Antigone tied by an ancient story to his Samian generalship 441/440, not a production record
  https://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.04.0023%3Atext%3Dintro%3Apart%3Dintroduction%3Asection%3D19 : "According to the author of the first Argument ... the success of the Antigone had led to Sophocles obtaining the office of general ... against Samos ... 440 B.C."; Haigh p.182 "According to an ancient tradition"
• CONFIRMED Persians 472, Seven 467, Oresteia 458
  https://bmcr.brynmawr.edu/2020/2020.05.41/ : "Persians in 472, Seven in 467, and the Oresteia in 458"
• CONFIRMED Oresteia 458 "two years before he died"
  https://en.wikipedia.org/wiki/Aeschylus : "died in 456 or 455 BC" at Gela
• CONFIRMED Philoctetes and Orestes "both written by men past seventy"
  https://en.wikipedia.org/wiki/Euripides : Euripides born c. 480, so about 72 in 408; Sophocles was about 87 in 409
• CONFIRMED P.Oxy. 2256 fr. 3, ed. Lobel 1952 (P.Oxy. XX): Aeschylus first; legible Danaids and Amymone; Sophocles second; Mesatos third; archon name damaged
  https://pressto.amu.edu.pl/index.php/sppgl/article/download/26089/23806 : Cipolla pp.53-54: "Dan[a]isi, Amu[mone", "deut[e]r[o]s Sophokles", "Mesatos", "Editionem principem curavit Lobel 1952"
• CONFIRMED Production put 466-459; Archedemides restoration gives 463, "which most scholars prefer"; Scullion 2002 argued 477-470
  https://pressto.amu.edu.pl/index.php/sppgl/article/download/26089/23806 : "inter annos 466 et 459 in scena actae sint ... Archedemides ... plures viri docti didascaliam ad annum 463 referendam censent"; n.6 "477-470, Scullion 2002"; Halleran "463 being more likely than any other date"
• CONFIRMED Plutarch Cimon 8 on Sophocles' first victory with his first plays (archon Apsephion)
  https://penelope.uchicago.edu/Thayer/E/Roman/Texts/Plutarch/Lives/Cimon*.html : 8.7-8 "When Sophocles, still a young man, entered the lists with his first plays, Apsephion the Archon ... Sophocles came off victorious"
• CONFIRMED Manousakis 2020 quote "The same method cannot apply to Aeschylus and Sophocles."
  https://dokumen.pub/prometheus-bound-a-separate-authorial-trace-in-the-aeschylean-corpus-9783110687675-9783110687644-s-2621082.html : verbatim, p.70
• CONFIRMED Manousakis quote "In Aeschylus’ older extant dramas Per., Sev., and Supp. the frequency of resolution ranges from 12% to 8%. In the trilogy it drops to 6%." (n.47, Schein 1979)
  https://dokumen.pub/prometheus-bound-a-separate-authorial-trace-in-the-aeschylean-corpus-9783110687675-9783110687644-s-2621082.html : verbatim in n.47. n.47 is in the early chapter "Style and Trace", not on p.70
• CONFIRMED Garvie 1969: "a steadily increasing restraint", which put Suppliants after Seven and before the Oresteia; Scullion showed it false; Garvie accepted this in 2006
  https://dokumen.pub/prometheus-bound-a-separate-authorial-trace-in-the-aeschylean-corpus-9783110687675-9783110687644-s-2621082.html : p.70: "a steadily increasing restraint ... Supp. appears to have been composed after Sev. and before the Oresteia and Pr. However, as Scullion (2002) 90–3 aptly demonstrated (and Garvie (2006) pr. xii confirmed), this suggestion is false"
• MINOR Scullion's point (as Manousakis gives it): 8 of the 43 resolutions in Seven fall in 30 trimeters spoken by Eteocles in groups of three; Scullion showed the suggestion false "as part of his case for the earlier date"
  https://dokumen.pub/prometheus-bound-a-separate-authorial-trace-in-the-aeschylean-corpus-9783110687675-9783110687644-s-2621082.html : the text is right, but it is on p.76 ("Scullion (2002) 92 observed that eight of the overall 43 resolutions ... 30 iambic lines, delivered in ten groups of three"), not the cited "p. 70 and n. 47". The phrase "as part of his case for the earlier date" is not in Manousakis, though Cipolla n.6 shows Scullion argued for 477-470
• WRONG Suppliants "sits exactly where a 1969 argument put it" (intro); dek/description "lands where a 1969 argument put it"; "this count reproduces Garvie's observation"
  https://dokumen.pub/prometheus-bound-a-separate-authorial-trace-in-the-aeschylean-corpus-9783110687675-9783110687644-s-2621082.html : Garvie (per Manousakis p.70) put Supp. "after Sev. and before the Oresteia". The page's own count puts it "between Persians and Seven, at about 468" (analysis.json point -468.4; Supp 9.34 vs Seven 8.94), i.e. before Seven. The page's own reasoning (Sophocles won 468, Aeschylus won 467) also rules out 468 for the papyrus production
• CONFIRMED Jebb 1894 Electra Intro §20 n.4 quote ("The extraordinarily high proportion ... oldest play after the Antigone.") and Wolff's seven ratios
  https://www.perseus.tufts.edu/hopper/text?doc=Perseus:text:1999.04.0025:text%3Dintro:section%3D20 : verbatim; 30 1/2, 26, 18 1/2, 18, 18, 14 1/2, 9 1/2 match published.json
• CONFIRMED Haigh 1896 p.101 quote "The Supplices is undoubtedly the earliest ... are decisive on the point."
  https://archive.org/download/tragicdramaofgre00haiguoft/tragicdramaofgre00haiguoft_djvu.txt : verbatim; the ellipsis covers only interleaved footnotes; page 102 header follows
• CONFIRMED Haigh footnote: K. O. Müller argued for about 461 from the praise of Argos, a view that "cannot be seriously maintained"
  https://archive.org/download/tragicdramaofgre00haiguoft/tragicdramaofgre00haiguoft_djvu.txt : "Müller (Diss. on Eumenides, p.84), having regard to the praise of Argos ... contended that it was written about 461 ... cannot be seriously maintained"
• CONFIRMED Haigh p.181 n.5: per-100 figures El. 4.20 ... Phil. 11.24, and the OC "much smoother ... serene tranquillity" quote
  https://archive.org/download/tragicdramaofgre00haiguoft/tragicdramaofgre00haiguoft_djvu.txt : verbatim (OCR "Kl." for El.); "182 SOPHOCLES" header follows
• CONFIRMED Halleran quote "perhaps from the fifth century’s first Olympiad"
  https://bmcr.brynmawr.edu/2020/2020.05.41/ : verbatim
• MINOR Halleran cited "for ... the papyrus's range for Suppliants"
  https://bmcr.brynmawr.edu/2020/2020.05.41/ : Halleran: "sometime between 470 and 459". The page's range is 466-459, which is Cipolla's
• CONFIRMED Müller 1866 pp.112-113: Aeschylus ca. 293 in 4308, Sophocles 446 in 7568, Euripides 3729 in 17825
  https://archive.org/download/depedibussolutis00mlle/depedibussolutis00mlle_djvu.txt : verbatim; the "113" page break falls inside the passage
• MINOR "Müller's Latin monograph ... gives totals for all three poets, leaving names out"; "The hand counts of the nineteenth and twentieth centuries leave names out"
  https://archive.org/download/depedibussolutis00mlle/depedibussolutis00mlle_djvu.txt : the explicit note ("non computatis h. l. solutionibus nominum ... propriorum") is attached only to the Aeschylus figure (published.json says so too). Müller does list names separately throughout, so this is plausible for the other two poets, not stated. The page itself says "Haigh and Wolff do not say how they treated names" (app.mjs:132), which contradicts "the hand counts ... leave names out"
• CONFIRMED Texts: Smyth's Aeschylus Loeb 1922-1926, Storr's Sophocles Loeb 1912-1913
  research/the-clock-only-euripides-kept/data/*.xml TEI headers : Smyth vol.1 1922, vol.2 1926; Storr 1912 (3 plays), 1913 (4)
• CONFIRMED Bibliography: Scullion "Tragic dates" CQ 52 (2002) 81-101; Manousakis De Gruyter 2020 doi:10.1515/9783110687675; Cipolla SPPGL 30 (2020) 53-74
  https://api.crossref.org/works/10.1093/cq/52.1.81 : Crossref and the journal's citation_ meta match
• CONFIRMED OCR slip: Prometheus 960 prints τε twice
  research/the-clock-only-euripides-kept/data/tlg0085.tlg003.perseus-grc2.xml:2208 : "ὑποπτήσσειν τε τε"
• MINOR "Every play of Aeschylus and Sophocles sits down in the band where Euripides began"
  verify-the-clock-only-euripides-kept.mjs (engine + three.mjs on the Euripides lines) : Euripides' three earliest dated plays (Alcestis, Medea, Hippolytus) run 5.81-7.23 with every resolution (4.23-6.54 with names left out). Persians 12.35, Seven 12.20, Philoctetes 11.36 and Suppliants 10.25 (10.37, 8.94, 10.02, 9.34 with names left out) are well above that. The verifier's check (top < 11 && bottom < 7) does not test overlap
• MINOR Pre-registered blind audit (PREREGISTRATION.md: "120 randomly drawn spoken lines (60 per poet)")
  research/the-clock-only-euripides-kept/audit-sample.mjs : run as 30 random plus 30 drawn from resolved lines per poet. The page describes the design accurately but does not flag it as a departure; "Two checks ... were added afterwards" names only two
• MINOR Page: a post-hoc check, "the same Suppliants fit with names counted", is on the page and labelled
  index.html:305; research/the-clock-only-euripides-kept/robust.json : no names-counted Suppliants result appears anywhere on the page or in app.mjs (robust.json has suppliantsPooled -465.3)
• MINOR Papyrus window: prose "between 466 and 459"
  public/strata/the-clock-only-euripides-kept/app.mjs (drawSuppliants), three.mjs, research/the-clock-only-euripides-kept/PREREGISTRATION.md : the chart label says "papyrus: 468 to 459", three.mjs "so 468 or later", and the pre-registration "468 to 459". This contradicts the prose's own reasoning that 468 and 467 are excluded
• MINOR Pre-registration test 5 (any "above" statement says whether it exceeds 2 combined SE)
  public/strata/the-clock-only-euripides-kept/index.html : "With names left out, the rate falls from play to play" and "The verse puts Suppliants between Persians and Seven" (9.34 vs 8.94/10.37, SE about 1.4) are made without that statement at the point. The Persians-Seven gap is qualified two paragraphs later; Suppliants' place is not
• OBSERVED Tests committed before any fit, after the per-play rates were seen
  `git log -1 --format=%ci 5ea4af23` (14:30:03) vs analysis.json/results.json mtimes 14:31 (untracked) : consistent. The pre-registration itself states the rates were on screen
• OBSERVED Blind audit: 116 of 120 agree, 18 in names each, 2 unscannable, 2 synizeses (66 would be 64)
  public/strata/the-clock-only-euripides-kept/data/audit.json; research/.../audit-reading.json : the record matches. The verifier reruns the scanner on the 120 lines and gets the same key. The reader's blindness is not independently recorded
• OBSERVED 44,563 lines read, 28,301 kept as spoken trimeters
  recreated: node verify-the-clock-only-euripides-kept.mjs : PASS
• OBSERVED Every computed figure in the prose (433-428, +44/-30, RMSE 1.90, rho -0.894, 468 [475-462], 4/10,000, 2.88/136.7, OC rank 5, p=0.0001, Müller agreement)
  recreated: node verify-the-clock-only-euripides-kept.mjs (30/0) : all match analysis.json and robust.json

What was done

• [fixed] WRONG: the intro "sits exactly where a 1969 argument put it", the dek/meta description "lands where a 1969 argument put it", and "So this count reproduces Garvie's observation". Garvie (Manousakis p.70) placed Suppliants after Seven and before the Oresteia. The count puts it between Persians and Seven at 468, before Seven and outside the page's own 466-459 window. Fix locations: src/content/strata/the-clock-only-euripides-kept.md (dek), index.html <meta description> and the intro paragraph, and the Garvie paragraph near line 260
• [fixed] MINOR: Müller's name exclusion is explicit only for Aeschylus; "the hand counts ... leave names out" conflicts with app.mjs:132 "Haigh and Wolff do not say how they treated names"
• [fixed] MINOR: "in the band where Euripides began" overstates. Four plays sit at 1.5-2x the rate of Euripides' earliest dated plays, and the verifier's check does not test the claim
• [fixed] MINOR: papyrus window 466-459 (prose) vs "468 to 459" (chart label in app.mjs, three.mjs ev, pre-registration test 3)
• [fixed] MINOR: the blind audit's sampling (half from resolved lines) departs from the registered "120 randomly drawn" lines and is not flagged
• [fixed] MINOR: "the same Suppliants fit with names counted" is announced as a labelled check on the page, but its result (robust.json -465.3) is never shown
• [fixed] MINOR: test 5 not applied to "Suppliants between Persians and Seven" (a 0.4 gap against SE about 1.4)
• [fixed] MINOR: citations. Scullion's Seven point is Manousakis p.76, not "p. 70 and n. 47"; "as part of his case for the earlier date" is not in Manousakis. Halleran gives 470-459, not the page's 466-459. Jebb OC §18 says 402 B.C. (the Dionysia of 402/1 is spring 401, so the page is right, but it never says why)
• [fixed] Before publish: public/checks/verify-the-clock-only-euripides-kept.mjs does not exist yet. The page links it, so publish.sh must add it