IV

Register of constraints and sources

6 min read
In-depth:

This register is the control side of the work: it sets every load-bearing constraint — every primary datum that limits the variable space — beside its source and its status. It is ordered by axis, in the order of the chapters. The status is a single mark, which says how the datum was inspected: SEEN, PRIMARY / SEEN FIRSTHAND (firsthand, as page image, own count or calculation) · DOCUMENTED (transmission — translation, summary, OCR) · CALCULATED (derived arithmetically or by automation) · CONSTRUED (a reading, not a measurement) · NOT SEEN (not inspected, and therefore without weight). Where a constraint carries two sides, the second is given in parentheses. It is a curated selection of the load-bearing constraints; the machine-readable base register in the deposit (https://doi.org/10.5281/zenodo.22944959) contains the full, numbered collection.

Axis 1 · Herod's year of death and the counting rule (chapters 2–3)

The four degrees of freedom (investiture year, capture year, counting rule, year-beginning) and the constraints that fix the year 4/3 BC.

ConstraintSourceStatus
Lunar eclipse shortly before the death; only 5, 4 and 1 BC umbral and visibleAJ 17.167; Espenak eclipse canonCALCULATED
34 regnal years since the capture, 37 since the investiture — plainly 4/3 BCBJ 1.665; AJ 17.191 (Niese)SEEN, PRIMARY
Investiture under the consuls of 40 BCAJ 14.389SEEN, PRIMARY
Capture under the consuls of 37, the 185th Olympiad, "27 years after Pompey, on the same day"AJ 14.487SEEN, PRIMARY
SVM itself names the way out "27 instead of 26"Schürer–Vermes–Millar I, 286 n. 11SEEN, PRIMARY
Sabbatical year during the siege; bears toward neither sideAJ 14.475; North 1954, 85–86SEEN, PRIMARY
Josephus's sabbatical-year notices "cannot be reduced to a consistent chronology [...] theorizing"North 1954, 85–86, 88 (reprise of Biblica 34, 1953)SEEN, PRIMARY
Coin gap years 1-3 of Antipas and PhilipRPC I; chance argument in the right unitCALCULATED
Eight knob combinations computed: two reach 1 BC, one the spring of 4automated, own ephemerisCALCULATED
Chain profiles over 20 data: C 0 V/0 E, CA 1 V, SY 3 V/8 Pthe ledgerSEEN FIRSTHAND
Inclusive counting rule retained for Herod, rejected for the Hasmoneansvia Cameron 2018, n. 18; SVM I 200–201 n. 1DOCUMENTED (SVM itself NOT SEEN)
Cameron: AJ 18.106 fits 4/3 untouched; for 3/2 two interventionsCameron 2018, n. 53; AJ 18.106CONSTRUED (Greek NOT SEEN)
SY: three rejections (consuls 40, consuls 37, Olympiad)Steinmann & Young 2020b, 445–447SEEN, PRIMARY
Philip and "Ἡρώδης ὁ ἐκ Μαριάμμης", not PhilipBJ 1.573; AJ 17.53 (Niese)SEEN, PRIMARY

Axis 2 · Matthew 2 — the birth–death interval (chapter 4)

ConstraintSourceStatus
Birth before Herod's death ("until the death", ἕως τῆς τελευτῆς)Matt 2:15 (SBLGNT)SEEN, PRIMARY
"two years and under" — a limit set by the king, not an ageMatt 2:16 (SBLGNT)SEEN, PRIMARY
Smallwood's measure: "at least a year [...] not necessarily two"; Egypt "need not have been lengthy"Smallwood 1976, 570–571 (App. E)SEEN, PRIMARY
Smallwood's own date of death ± February 4 BC, Passover as limitSmallwood 1976, 104 n. 158SEEN, PRIMARY
The τότε is Matthean style, not a temporal markerMatt 2:16 (SBLGNT), countCALCULATED
Egypt without duration; "the estimates are just guesses"Davies & Allison IDOCUMENTED
The Matthew floor is "months", not a full yearown weighing of Matt 2CALCULATED
The star as a report: five properties (appears, out of view, the same on return, goes before, stands over); only a comet satisfies all five among celestial candidates, with "stands over" as an optical impressionMatt 2:1–10 (NA28); own eliminationINTERPRETED
No celestial object can stand still over one place; ἐστάθη ἐπάνω as direction or as an object near at handgeometry, own computationCALCULATED
All five elements of Matt 2:9 stand in the LXX around the guiding light of the exodusExod 13:21; 33:9; Num 9:15; Deut 1:33 (LXX)SEEN, PRIMARY
Selective sight: 2 Kgs 6:17; Num 22:23; Acts 9:7, 22:9, 26:13; Herod must ask (Matt 2:7)LXX; NA28SEEN, PRIMARY · INTERPRETED
Chrysostom: "some invisible power transformed into this appearance"Hom. in Matt. 6.3 (NPNF)TRANSMISSION
Babylon: last dated observation 30 June 10 BC, then only almanacs; Greco-Roman: no comet between 12 BC and AD 9ADART V no. 31; ADART VII; Ramsey 2006SEEN, PRIMARY
The one contemporary reading of no. 63 is Chinese and dynastic (Xia Heliang 5 BC; Wang Mang)Han Shu 26; Dubs 1955, 28–32SEEN, PRIMARY
The interval depends on the part of the year and the season of death: 5 BC under winter 4/3 ample in every part (Nisan 19.5–23.8 months, autumn feasts 13.3–17.6), under spring 4 (± 3 April) only on 15 Nisan (11.4 months; autumn feasts 5.5, under Smallwood's February 4 BC 4.0); 4 BC 7.5–11.7 months on 15 Nisan against 1.7–6.3 at the autumn feasts`db/veeg-seizoenen.py`, run; table §6.2CALCULATED

Axis 3 · The legates of Syria, the coins and Luke 2:2 (chapter 5)

ConstraintSourceStatus
Fasti of Syria: Saturninus, then Varus 7-c. 4 BC, then Piso; Quirinius only in AD 6/7Dąbrowa 1998, 26, 135SEEN, PRIMARY
Varus coin-anchored: "ΑΝΤΙΟΧΕΩΝ ΕΠΙ ΟΥΑΡΟΥ", year 25 (7/6) and year 27 (5/4)RPC I 4242, 4252 (RPC Online)SEEN, PRIMARY
From year 28 (4/3) Augustus himself on the Antiochene bronzes, no legateRPC I 4255 "ΚΑΙΣΑΡΙ ΣΕΒΑΣΤΩ ΑΡΧΙΕΡΕΙ" (RPC Online)SEEN, PRIMARY
Gaius on the consilium at the will; consul in Syria 1 January AD 1BJ 2.25; AJ 17.229; Dio 55.10a.4SEEN, PRIMARY
ILS 918 (titulus Tiburtinus) headless; iterum "far from stabilized"ILS 918; Sherwin-White 1963, 164 (page image)SEEN, PRIMARY
ILS 2683 (Aemilius Secundus) carries no date — does not date QuiriniusILS 2683; EDCS-08000281SEEN FIRSTHAND
A genuine census record dates by imperial year/consuls, not by the governorP.Yadin 16 (AD 127)SEEN, PRIMARY
Luke 2:2: ordinal + genitive absolute = temporal; the article ἡ is secureNA28 on Luke 2; Acts 5:37SEEN, PRIMARY
"before Quirinius" fails: Luke says πρό where he means itcount of πρῶτος + genitive in Luke/LXXCALCULATED

Axis 4 · The season — season of death and season of birth (chapter 6)

ConstraintSourceStatus
No lunar eclipse visible from Judaea between Passover 4 BC and Passover 3 BC (5 Sept 4 BC daylight; 31 Jan 3 BC below horizon; 2 March 3 BC penumbral 0.12)own computation, NASA-calibratedCALCULATED
Day count AJ 17.167–213: window 22–57 days; 13 March → Passover 4 BC 30 days (inside), → Passover 3 BC 414 days (357 unnarrated)AJ 17.191, 17.200, 17.213; BJ 1.665, 2.1; `db/aj17-rekening.py`CALCULATED
Ages at baptism 15 Tishri AD 27: Sukkot 5 BC 31.0, Passover 5 BC 31.5, Sukkot 4 BC 30.0 (only under winter 4/3); exact ages; with a baptism after 20 October 27 (late autumn 27 or winter 27/28) at most 0.4 higher; in completed years no cell of the window changes, in year cells 6 BC drops out if a winter baptism counts as baptism 28Parker & Dubberstein; Luke 3:23; §6.2, §9.3CALCULATED
Antipas's year 43 = AD 39/40 puts year 1 at 4/3 BC; excludes 1 BCRPC I 4934 "ΓΑΙΩ ΚΑΙΣΑΡΙ ΓΕΡΜΑΝΙΚΩ" (RPC Online)SEEN, PRIMARY
Regnal years allow nothing before 1 Nisan 4 BCBJ / AJ, double figureSEEN, PRIMARY
Travel clock: only a death immediately before Passover excludedKushnir-Stein 1995, 76–79SEEN, PRIMARY
Illness clock is a retribution topos and does not clock — withdrawnLadouceur; Africa 1982; ZeichmannCONSTRUED (withdrawal)
Ptolemy sweep: Josephus distinguishes the Ptolemies, prices block 2 without breaking itautomated sweep over Niese (AJ 17 // BJ 2)SEEN, PRIMARY
7 Kislev points to winter but does not choose the yearMegillat Taʿanit (scroll); scholion post-TalmudicSEEN, PRIMARY (scholion DOCUMENTED)
Priestly courses do not choose: Sukkot 5 at +8.9 wk, Passover at +10.3continuous-course model (`koersen.py`)CALCULATED
Comet Ho no. 63 in the sky on 22 April 5 BCHan Shu 天文志; Ho Peng YokeSEEN, PRIMARY (CALCULATED)
Eighth day: 15 Tishri → circumcision on Shemini AtzeretLev 23:36, 39; counting backCALCULATED
No ancient source places the birth in TishriClement, Chronograph of 354, EpiphaniusSEEN, PRIMARY

Axis 5 · Patristics as a witness class (chapter 7)

ConstraintSourceStatus
No Father supplies a year in an independently readable era — all conversionFinegan 1998, Tab. 139SEEN, PRIMARY
The same figure 41 spreads 15 years across the five Augustus epochsFinegan 1998, Tab. 128CALCULATED
SY's "11 witnesses" = Finegan's nine; three rows point elsewhereFinegan 1998, §500SEEN, PRIMARY
Two of the 11 have Herod die after 1 BC (Origen AD 2/3; Epiphanius's 37th year)Finegan 1998, Tab. 143–144; Epiphanius, Pan. 51.10 and 51.22DOCUMENTED (underlying loci NOT SEEN)
Clement's interval "194 years, 1 month, 13 days" before CommodusClement, Strom. 1.21.144–146 (ANF)DOCUMENTED (Greek NOT SEEN)
Most recomputable Eastern figure lands on 25 December 4 BCBar Hebraeus, AG 309SEEN, PRIMARY

Axis 6 · The logigram — run A and run B (chapters 8–9)

ConstraintSourceStatus
Run A (freestanding): C/CA/HV → 8-4 BC; SY → 5-1`logigram.py`, outputCALCULATED
Run B-consistent (coupled): C/CA/HV → 6-4 BC`logigram.py`, outputCALCULATED
Kill matrix: Matthew 2 "after the death" 264, Luke 3:23 364, "≥5 years before" 120`logigram_output.json` (deposit)CALCULATED
Tiberius floor: baptism not before AD 27 (inclusive); mainstream coverageSteinmann 2011, 100; Smallwood 1976, 168 n. 82SEEN, PRIMARY
Coin material matches inclusive reckoning — but measures the pagan framework, not the Jewish oneRPC I 4270, 4330; PhiloSEEN, PRIMARY
"about thirty" on ±2 (Plummer): early edge 6, not 7; 8 measurable ὡσεί passages, 7 at zeroLuke 3:23 (SBLGNT); Plummer, ICCSEEN, PRIMARY (CALCULATED)
Run B uses the Luke 3:23 age band without an imported birth-year floor`logigram.py`CALCULATED

The open "not seen" entries are in the apparatus (appendix D); the full machine-readable base register is in the deposit, https://doi.org/10.5281/zenodo.22944959.