This book has laid its own case against the strictest measure, and built up the strongest rival, AD 30, at his best before weighing him. Honesty now asks one thing more, and it is the heaviest: the same measure against the reigning date itself. For there is one year that counts in the scholarly literature not as a candidate but as the answer — AD 33, with its famous Good Friday of 3 April. If this book charges the tradition with never having weighed AD 31, then it must indeed weigh AD 33, and with exactly the same scales on which it has measured itself. No double standard, in either direction.
So let AD 33 speak first at its strongest, for weak it is not.
Countercasethe full case for AD 33
AD 33 is not the second choice but the first choice of most scholars, and it has real strength. First the language: the dominant New Testament idiom is "on the third day", and whoever takes that as the norm and subordinates the sign of Jonah to it arrives at a Friday — and AD 33 is the best-documented Friday. Second the astronomy: the classic study of Humphreys and Waddington defends Friday 3 April 33 with a lunar eclipse that evening, a "blood moon" over Jerusalem which Peter invokes in Acts 2:20 — the most-cited astronomical argument in the entire debate (Humphreys & Waddington, 1983, 1992). Third the ministry: AD 33 requires no longer career; with a baptism at the end of 29, on the ordinary Tiberius count, it too has three and a half years. And fourth the weight of eighteen centuries: Good Friday is the day of the church, east and west. Coherent, mainstream, with an eclipse of its own. Thus AD 33 deserves to lie on the table, and thus it lies there.
Now the measure. And the first question is the one the whole book asks: is this a measurement or a majority? For the strength of AD 33 is largely a strength of consensus and of comfort — and as soon as one recomputes it astronomically and textually, with the same rigour that charges AD 31 with its Adar II, it proves less hard than it presents itself.
Begin with the most surprising fact, which is seldom mentioned: the famous Good Friday of 3 April does not stand in the standard work of the Babylonian calendar. Parker and Dubberstein — the table on which the whole chronology of this century rests — place 1 Nisanu of AD 33 on 19 April, and thus 14 Nisan on 2 May, a Saturday (Parker & Dubberstein, 1946, p. 46). Not 3 April; not a Friday. The standard authority, consistently applied, does not deliver the Good Friday at all. That is, however, the Babylonian calendar, and the link from Babylon to Judea does not hold of itself; on the same ground the 25 April of 31 cannot rest on Parker and Dubberstein. It arises only when one lets the March moon count after all — the moon that stood that year a day and a half before the spring equinox — that is, inserts no intercalary month and lets Passover fall before the equinox. That is a calendar choice, and it is precisely the kind of choice held against AD 31, only in the other direction.
Expertwhat the Friday of AD 33 hangs on
The March moon of AD 33 gave 1 Nisan around 21 March, a day and a half before the true equinox (22/23 March). Under the strict rule — Nisan begins only on or after the equinox, the rule on which Parker and Dubberstein hang their entire table — AD 33 becomes an intercalary year and 14 Nisan falls on 2 May (Saturday): no Friday. Only under the lenient rule — Passover may fall before the equinox — does the March moon stand, and 14 Nisan comes to Friday 3 April. The Good Friday of AD 33 is therefore no astronomical fact but a consequence of the weak rule; it hangs on a day and a half of indulgence. Compare that with AD 31, and note sharply exactly where the difference lies. There the March moon stood nine days before the equinox, deep in winter. AD 31 too has two options — under the strict rule the winter moon is rejected and 14 Nisan falls on Wednesday 25 April; under the lenient rule it stands and 14 Nisan falls on Tuesday 27 March (ch. 8). But neither is a Friday.There lies the asymmetry, and it is sharper than a robustness claim: the Friday of AD 33 exists under one of the two rules and vanishes under the other, whereas the Friday of AD 31 exists under neither. The consensus thus needs the indulgence to keep its own year standing, and simultaneously denies it to the year that does not even need it to lose the Friday. Which of the two days of AD 31 is the real one — the Wednesday or the Tuesday — is decided not by the equinox but by the barley, and that is the question chapter 8 measures out and which the excursus at the end of this chapter works through with an opponent. And this is a measurement, not a preference.
See what that means for the honesty of the field. The charge that AD 31 carries — that it needs a calendar liberty no one can verify, a Nisan that falls "too late" — applies to AD 33 in equal measure, only its Nisan falls "too early". Both fill in the same missing decision of the Sanhedrin; the difference is that AD 31 does it out loud and names it as an assumption, while AD 33 does it tacitly and lets it pass for a fact. And the liberty that rescues AD 33 is precisely the liberty the consensus denies AD 31: when Sacha Stern writes that one did not always wait for the equinox — "if both crops and fruits were sufficiently ripe, no intercalation would be made, even if this meant Passover occurring before the equinox. This may have happened quite frequently" (Stern, 2001, p. 70) — then that leniency rescues the Friday of AD 33, but it simultaneously strips the consensus of the right to reject the intercalary month of AD 31 "because otherwise there is no Friday". One cannot invoke the calendar liberty to save one year and refuse it in order not to see the other.
Dossierthe same March moon, three distances to the equinox
year
March moon → 1 Nisan
distance to the equinox
14 Nisan under the strict rule
under the lenient rule
Friday possible?
AD 30
25 Mar
+2.1 d after
Friday 7 April
(same) — the rules coincide
yes, under both
AD 31
14 Mar
−9.2 d before
Wednesday 25 April
Tuesday 27 March
no — under neither
AD 33
~21 Mar
−1.7 d before
Saturday 2 May
Friday 3 April
only under the lenient
Read the last column, for there lies the whole point. Not: "AD 31 is robust and AD 33 fragile" — AD 31 itself has two outcomes. But: the Friday of AD 33 hangs on the choice between the two rules, and the Friday of AD 31 exists under neither. The choice of rule that the consensus needs for itself does not help it with AD 31.
The asymmetry is calculated, not a choice (ch. 7). AD 33 is the only one of the three in which the choice of rule overturns the weekday. Parker and Dubberstein, strictly, give 2 May; the Good Friday requires the lenient exception. [CALCULATED]
Sharper still than the arithmetic is one documentary objection, and it is the heaviest the consensus really holds against a late Passover — and thus against the April moon that makes AD 31 a Wednesday. Honesty asks that I present it at its strongest, for weak it is not.
Countercasethe sixteen March dates of Sardica
From the documents of the Council of Sardica (winter AD 343; codex Veronensis 60, discovered in 1905 by Eduard Schwartz) a list of Jewish Passover dates has been preserved for the years 328–343 — and all sixteen of them fall in March, not one in April. The series is moreover reliable: the eastern bishops included it in order to show that they did not celebrate together with the Jews, and "the polemical credibility of that argument depended on the authenticity of these Jewish dates; they are therefore unlikely to be merely invented or falsified" (Stern, 2001, p. 75) — and the list stops precisely in the year of its compilation, so that it records observed dates, not a prediction. To this is added a hostile ancient witness: Peter of Alexandria (early 4th century) argues that the Jews in Jesus' day kept the equinox rule, but after the destruction of the temple "fell into error" and began to celebrate Passover at the wrong time. On this view the base rate for a 25-April Passover is 0 out of 16 — attested, contemporary, and backed by a church father who says the early church knew better. This may not be reasoned away; it must be explained.
And it can be explained — not by me, but by the consensus authority that published these dates itself. For Stern, who weighs the sixteen Sardica dates, places them expressly outside the first century. His own section bears the heading "From the first century to the fourth: a radical change", and his conclusion is the opposite of what the series seems at first sight to say.
Expertwhat Sardica really measures
Stern reconstructs the first-century Passover as late: the celebration of Passover before the equinox, widespread in the fourth century, stands "in contrast with our findings for the first century, when Passover appears to have fallen considerably later, sometimes more than a month after the equinox" (Stern, 2001, p. 71). That reconstruction hangs on two independent first-century data points: AD 37, where the Passover — from Josephus, Antiquitates 18.122–124, with the "fourth day" and the news of Tiberius's death — fell around 19 April; and the Berenice inscription, where Tabernacles fell on 29 April, "well over a month after the spring equinox" (p. 61). The sixteen March dates are thus the fourth-century end state after a regression that Stern explains from the end of the pilgrimage and the ebbing Babylonian influence after 70 — and which the hostile witness Peter of Alexandria himself dates to after the destruction of the temple. "We may endorse the view of Peter of Alexandria that a change in the Jewish calendar reckoning took place after AD 70," Stern writes, "but for entirely different reasons" (p. 71). The 0-out-of-16 thus measures the wrong century: it is the fourth-century calendar, not the first. Applied to the first century, Stern gives the reverse — late is the norm, and 25 April (14 Nisan; 1 Nisan ≈ 12 April) lies within the band he reconstructs. [DOCUMENTED — Stern's analysis §2.3.5 + §2.5.2–3 (pp. 61–76) read at first hand; the sixteen lines of the Sardica document itself not seen line by line, but via Stern's rendering]
One limit belongs with it. A year earlier Stern himself noted a second practice: in Egypt, from Aristobulus to Philo, the Passover fell at or around the equinox, on average a month before the Babylonian date (Stern, 2000, p. 171 n. 69). That is no first-century Judean date, and it does not touch the reconstruction above; but it means that the late Passover is one of two documented practices, and which of them Jerusalem kept is the open question (ch. 8). [DOCUMENTED]
And now the honest limitation, in the same breath, for otherwise I measure with two measures. Stern's reconstruction yields no 25 April: "considerably later, sometimes more than a month" is a range, not a point date, and 25 April lies at its late edge, not in the middle. It rests moreover on two data points plus Babylonian conformity — a reconstruction, not an attested series like the sixteen of Sardica themselves. Symmetry demands that I concede it: the March dates are documented; Stern's "late first century" is inferred. But the decisive thing stands: it is the consensus scholar himself who, on non-numerical evidence, places the sixteen dates outside the first century and reconstructs that century as late. The heaviest documentary objection against a late Passover disarms itself in the standard work from which it comes. As a falsifier of 25 April, Sardica falls away; as evidence for 25 April I do not advance it — it is a rebuttal, not a confirmation. [CONSTRUED where it concerns the point date; DOCUMENTED where it concerns Stern's reconstruction]
There remains the blood moon — the astronomical showpiece of AD 33. And here honesty asks that I apply my own measure, not a stricter and not a more lenient one. Yes, on the evening of 3 April 33 there was a lunar eclipse, and the moon rose eclipsed over Jerusalem. But two things, and both hold just as hard for my own eclipse as for that of the consensus. First, that eclipse is marginal: the eclipse reached its maximum before the moon rose, so that the moon appeared low on the horizon while the umbral phase was already waning — a brief, low blood-colour, not a high, ample eclipse. Second, and decisively: the blood-moon argument is a construal, not a measurement. It links an astronomical phenomenon via Acts 2:20 to the crucifixion — precisely the kind of theological reading that this book expressly treats as an image for its own eclipse of 25 April 31, granting it not a grain of weight (ch. 7). Whoever is consistent does that on both sides. Under my own rule, then, the blood moon of AD 33 does not count as evidence for the year — no more than mine does. It is beautiful; it does not measure.
And then the family tree. AD 33 is, among the ancient writers, the choice of Eusebius — and that choice is one man deep, built on the longest ministry and anchored to the phantom witness Phlegon (ch. 6). The tradition does not remember the year; it computes it, from Luke 3:1 plus a guessed duration, and AD 33 is the outcome of the longest guess. There is, in short, no independent ancient measurement carrying AD 33. What carries it is a computation, an eclipse-as-construal, and a majority — and none of the three is stricter than what carries AD 31; they are only more familiar.
And if the family tree of the year is one man deep, then the largest patristic Easter dossier shows why that is so: the early church did not measure the crucifixion year — it quarrelled over liturgical practice.
Expertthe quartodeciman controversy does not measure the year
The heaviest second-century quarrel over Easter was not about when Jesus died, but about when one should commemorate it: on 14 Nisan itself — the Jewish date, to which the "quartodecimans" (fourteeners) held fast — or on the following Sunday. Around AD 190 Polycrates of Ephesus, in a letter to Victor of Rome, defended the 14-Nisan observance with an appeal to the apostolic tradition (John, Philip); Victor pressed the Sunday through and threatened exclusion, upon which Irenaeus urged moderation. Melito of Sardis, author of Peri Pascha, belonged to that same fourteener tradition. The whole dossier stands in Eusebius, HE 5.23–25. And what it does not contain is decisive for this book: none of these witnesses gives a consulate, a Tiberius year, an Olympiad, or even a weekday of the crucifixion. Polycrates and Melito deliver zero chronology. The struggle is about the calendar practice of the church, not about the year of the cross. [DOCUMENTED that the dispute concerns the liturgical date — Eus. HE 5.23–25, with Beckwith (1996) read at first hand; that it hands down no year is a negative observation, marked here CONSTRUED]
That is cheap but real confirmation of what the family tree already showed: whoever invokes "the tradition" for AD 33 or for the Friday invokes a tradition that hands down nothing about the year and never quarrelled over it either. It computes it, from Luke 3:1 plus an assumed length of ministry; and what it does passionately dispute — the Easter date — is a matter of liturgy, not of measurement.
ExpertAugustine's thirty-six hours, a mirror
Whoever would charge a Wednesday count with "numerology" must know one passage of the tradition itself. Augustine faces honestly the best-known shortfall of the Friday count: from the evening of the burial to the dawn of the resurrection he himself counts thirty-six hours — not three full days. And he rescues the count with a number: thirty-six is six squared, and in the ratio of the single to the double that lies within it (twelve to twenty-four) is hidden "the greatest concord of likeness" (De Trinitate IV.6.10). Matthew 12:40 he does not cite there. This is no stain on Augustine — number symbolism was in his century respectable exegesis, and his inclusive count is a genuine Jewish idiom — and it is no proof against the Friday. It is a measure: the same kind of construal one charges against a Wednesday count sits also on the side of the consensus, at the heart of its own chronology. Whoever disqualifies the one for it disqualifies the other — or marks both as construal and seeks the proof elsewhere. [CONSTRUED — a symmetry point, not an argument for the year]
Expertthe consensus claims more than it measures, on the year-axis too
The same holds beyond the weekday. Steinmann concludes that there is "no evidence for inclusive reckoning" (2022a, p. 115) — a general statement. Egypt's own corpus contradicts it: Bureth's repertory counts the first years of Gaius, Claudius and Nero inclusively, and Nero's fourteenth year — the year of his death — stands there on some thirty ostraca, whereas on the non-inclusive count it cannot exist (ch. 9). In fairness: this does not deliver autumn 27, for Tiberius is precisely the exception in Egypt. It shows only that the general claim is wider than the evidence that carries it — exactly the measure this chapter applies. [DOCUMENTED — Bureth, seen at first hand; CONSTRUED for the weighing]
And the same measure on my own side, for there is a second place where Steinmann lends support and the counterweight usually stays away. For the building years of the temple there is Josephus's statement that the priests completed the courts in eight years (Ant. 15.420) — that is the support. The counterweight stands in Josephus himself, and it belongs in the same breath: Ant. 20.219 says the sanctuary was only finished under Agrippa II — τὸ ἱερὸν ἐτετέλεστο — and that eighteen thousand workmen were then left without work. The building was therefore not finished in the twenties of the first century, and whoever cites the forty-eighth year as an end point cites an interim state. That does not make Steinmann's datum unusable; it makes it narrower than it is usually used — and that is precisely the rule this chapter applies to the consensus: grammar and the statement first, chronology only after. [DOCUMENTED — Ant. 15.420 and 20.219; ⚠ read in translation, not in the Greek]
Excursuswhere AD 33 does stand stronger, honestly conceded
Two things must go on the other side, for without them this chapter would itself be measuring with two measures. First: AD 33 reads the dominant idiom naturally. The whole third-day family — "on the third day", "after three days", "in three days", nineteen forms together, against the single "three days and three nights" — bends effortlessly to a Friday, and AD 33 need not force a single text for it. Whoever reads the sign of Jonah as a manner of speaking rather than as a measure (ch. 3) has in AD 33 a comfortable and coherent year. Second: AD 33 has no need to bend the ordinary Tiberius count; its baptism at the end of 29 follows from the best-attested reading of Luke 3:1 (ch. 9). On those two points AD 31 pays a price that AD 33 does not pay. That is the honest reverse side, and it stands here, not hidden away: AD 33 is astronomically fragile and without independent measurement, but textually comfortable and chronologically conventional. It is not a weak year. It is a conventional year that passes itself off as a measured year.
Excursusa third reconstruction, and where the calendar dispute really lies
There exists a living, published opponent of the April moon of AD 31, and he deserves a place here — not because he decides the matter, but because recomputing his work shows where the decision does fall. Richard Lanser reconstructs an "Exodus Hebrew Calendar" in which the Jewish intercalation series runs one cycle-year ahead of the Babylonian. For AD 31 that yields Nisan 1 on 14 March instead of 12 April, and thus 14 Nisan on Tuesday 27 March — a day that neither the Friday nor the Wednesday can use.
His facts are correct, and that must be said first. I have recomputed his data against the tables of Parker and Dubberstein in machine-readable form: the second Ululu of 313 BC, the intercalation of 311 BC, the two-year interval, the cycle positions — all correct. And his scheme is not ad hoc but relentlessly consistent: his Nisan 1 is each time the thirteenth lunation of the preceding Babylonian intercalary year, in every year without exception. For AD 31 that is P&D's Addaru of 14 March, for AD 33 their Addaru II of 21 March — in both cases exactly his date. He says it himself: the intercalary moons are redefined as Nisan moons.
And there is the seam, for two things he presents as one are not. His justification is a renumbering: the cycle positions of the Babylonian intercalary years, {3, 6, 8, 11, 14, 17, 19}, become {1, 4, 7, 9, 12, 15, 18} with him. His chart does something else: it shifts the whole series by one year. That the two yield the same list does not make them the same, and the difference is exactly what matters: only the shift removes the intercalary month of AD 31. The renumbering by itself does not. What that refutes is not his calendar but the ground given for it: the reason he names does not carry the outcome he records. Anyone who wants to keep his scheme may — but with a different justification than this one. [INTERPRETED]
Three observations, and the third is the only one that counts.
First. His own first step — replacing the second Ululu of AD 30 with an Adar II at the end of that Hebrew year — sets Nisan 1 of AD 31 precisely in April. It is the shift of one cycle-year that brings him back to March. The shift is therefore no ornament of his reconstruction; it is its core.
Second. His calendar diverges from the Babylonian in precisely the years that follow an intercalary year — in the window examined here, AD 31 and AD 33 — and coincides with it in all other years. In AD 33 that moves the Passover from Saturday 2 May to Friday 3 April; in AD 31 from Wednesday 25 April to Tuesday 27 March.
Third, and this is the gain of the whole recomputation.The first crescent cannot settle this dispute. His 14 March and the 12 April of Parker and Dubberstein are both genuine first crescents — two successive lunations. The crescent establishes on which day a month begins; which lunation carries the first month can be established only by the aviv. And on that we agree: he expressly appeals to it.
> The whole calendar dispute thus comes down to one testable question: did the barley in Judea stand in the aviv at the end of March? On his placement the sheaf of firstfruits falls on 29 March; on this book's on 27 April. Not the cycle decides, and not the numbering — the crop.
And that is precisely the question that Chapter 8 treats with Dalman's regional figures, with Milgrom, and with the omer rules of m. Menachot 10:2 and 10:8 — with b. Menachot 85a supporting them rather than carrying them.
The symmetry belongs here, and it is real. This book too uses a scheme as a stand-in for a practice that was observational: that Judea followed the Babylonian intercalation is an assumption and not a record. We are both reconstructing. And the symmetry must go further here than is comfortable, for two things I should like to hold against him I cannot hold against him. Not the Passover of AD 37: his calendar gives for that year the same 19 April as this book's, so that the Vitellius year confirms him just as well as this book (ch. 8). And not Parker's Jerusalem computation either: Parker likewise applies the Babylonian series to Jerusalem, and both parties thereby fill in the same undocumented decision of the Sanhedrin. Parker tests the placing of the crescent; he does not test the intercalation choice, and that choice is precisely what the dispute is about.
What remains, then? One thing, and it is enough: the difference is not rigidity but testability. This book's reconstruction hangs on something given outside the scheme — the crop — and it can fail against it. His fixes the starting point and thereafter lies fixed for seventy-two years, with nothing left against which it could fail. That is the asymmetry that counts, and it is the only one I claim. [DOCUMENTED — his calendar and the P&D tables measured through; CONSTRUED — the weighing]
Excursusthe sharpest published attack on AD 31, and whom it actually hits
Since 2025 there is a second living opponent, and he is better documented than the first. André Reis, a New Testament scholar, rejects AD 31 and arrives at Friday 3 April 33. His conclusion is one sentence: AD 31 could be the year of the crucifixion only "if it had an additional month (ruled out per the vernal equinox rule), Nisan 1 was delayed by two days (not permitted by Jewish calendrical rules) and John's chronology of the Passion is completely ignored in favor of a crucifixion on Passover Day/Nisan 15 (highly improbable)."
Two of those three grounds do not touch this book, and it is accuracy — not courtesy — that has that said here. They are written against two named defenders of AD 31. C. Mervyn Maxwell puts the crucifixion on Nisan 14 and on a Friday, and to do so must shift the beginning of Nisan by three days, from Tuesday to Friday; he concedes himself that AD 31 is a candidate only "granted a few unprovable yet reasonable assumptions." Juarez R. de Oliveira maintains a crucifixion on Nisan 15, shifts the beginning of the month by one day, and advances Passover to Thursday 26 April 31.
This book does neither. It shifts no month-beginning: 1 Nisan stands on the first visible crescent, without a correction day, and the whole computation lies open to recalculation in appendix A. And it puts the crucifixion on Nisan 14, not on 15 — precisely where Reis puts it. On the day of death we agree; the dispute is the weekday and the year. Of his three grounds, one therefore remains over against this book, the intercalary month, and that is a calendar question, computed in H8 and not here.
The paragraph stands here because without it the book inherits a refutation that was not written for it. Whoever reads Reis and then this book must be able to see that two of the three knives pass by — and that the third stays where it belongs: in the calendar. [DOCUMENTED — Reis (2025), pp. 16 and 18; ⚠ Maxwell (God Cares I, 1981, pp. 260–263) and de Oliveira (Chronological Studies, 2004, p. 46) stand here on Reis's report and reference, not consulted at first hand]
Where does the inquiry stand, then, after laying the consensus against its own measure? Not at a refutation — AD 33 is not impossible, and I do not declare it dead. Stern's "quite frequently" makes the pre-equinox Passover genuinely plausible, so that Friday 3 April 33 stays alive along the barley and weak-rule route. But that is precisely the point: AD 33 lives on exactly the same calendar liberty that carries AD 31, and its version of it is the weaker one — a March moon of a day and a half against nine, a Good Friday the standard table does not give, a blood moon that as a construal does not count, a family tree one man deep. AD 33 has no lead in rigour. It has a lead in familiarity. And those two are not the same — that is the whole thesis of this book, applied here to the date that stood in its way the longest.
With that, all the pieces lie on the table. The heaviest verse has been weighed, the strongest rival built up, and the reigning date laid against its own measure. What remains is the weighing itself — the whole field on the scales, with the photo finish shown rather than left out.
HeartHEART
It is not pleasant to test a date that millions of people have kept with reverence and love. Good Friday carries centuries of prayer, and nothing in this chapter wishes to belittle that. But it is precisely reverence for what is holy that asks for precision, not for comfort. If the tradition has somewhere passed too quickly over a measurement, it is no dishonour to show her that — it is a service to the same truth she loves. I lay AD 33 on the scales with the same hand with which I laid AD 31 on them: not to win, but to weigh honestly. And the outcome is not that the consensus errs, but that it never measured what it thought it knew. That is enough. The rest belongs to the scales.
Note on the sources
The astronomical data (14 Nisan AD 33 = 2 May under the strict rule; 3 April under the lenient one; the distances of the March moons to the equinox; the marginal eclipse of 3 April 33) are recomputable and worked out in ch. 7 and appendix A; Parker and Dubberstein have been checked in the source itself. The position of Humphreys and Waddington and that of Stern were consulted at first hand, with location. This chapter does not refute AD 33; it shows that AD 33 is at no point stricter than AD 31 — astronomically more fragile, without independent measurement, and rescued by the same calendar liberty that the consensus denies AD 31. The blood moon of AD 33 is treated by the same measure as the eclipse of AD 31: a construal, not evidence. Two documentary witnesses have also been weighed. The sixteen March dates of Sardica (winter 343) and Peter of Alexandria are drawn from Stern's analysis, read at first hand (§2.3.5, §2.5.2–3, pp. 61–76); the sixteen lines of the Sardica document itself (codex Veronensis 60, ed. Schwartz 1905) have not been inspected line by line — they carry here via Stern's rendering, and that is marked as such. The quartodeciman controversy rests on Eusebius, HE 5.23–25, and on Beckwith's discussion of it (read at first hand); HE 5.23–25 itself has not been reread line by line in this study, but HE 7.32 has (ch. 7). In both cases symmetry holds: the measure applied to AD 31 is applied here to the consensus counter-evidence — and it does not carry as a falsifier of the Wednesday. Sardica measures the fourth-century Easter reckoning, not AD 31; the quartodeciman struggle measures liturgical practice, not the year of the cross. Augustine's thirty-six-hour passage (De Trinitate IV.6.10) is cited in translation; it is marked exclusively as construal and carries no burden of proof — it serves the equal measure, not the year.
On the excursus with the third reconstruction, finally, for there the equal measure applies most sharply. The calendar of Richard Lanser has not been assessed at second hand: his published overview and his spreadsheet have been laid line by line beside the machine-readable tables of Parker and Dubberstein, and his figures have been found correct — including the Old Babylonian calibration points on which he bases his cycle numbering. What is brought against him here is thus not a calculation error but a single parameter choice, and that has been named as such. Two arguments that lay ready to hand have expressly not been used, because on recomputation they do not work against him: the Passover of AD 37, which on his calendar gives the same 19 April, and Parker's Jerusalem table, which makes the same undocumented assumption about Judea as both parties. What remains is the testability against the crop, and that is marked as CONSTRUED. This chapter claims a proof for neither side.
Literature (APA)
Beckwith, R. T. (1996). Calendar and chronology, Jewish and Christian: Biblical, intertestamental and patristic studies. Brill.
Humphreys, C. J., & Waddington, W. G. (1983). Dating the crucifixion. Nature, 306(5945), 743–746.
Humphreys, C. J., & Waddington, W. G. (1992). The Jewish calendar, a lunar eclipse and the date of Christ's crucifixion. Tyndale Bulletin, 43(2), 331–351.
Lanser, R. D. (2026). Reconstructing the original Hebrew calendar (revised overview table, 8 August 2026, with accompanying spreadsheet). Associates for Biblical Research.
Parker, R. A., & Dubberstein, W. H. (1946). Babylonian chronology 626 B.C.–A.D. 45 (Studies in Ancient Oriental Civilization 24, p. 46). University of Chicago Press. (Also in the machine-readable edition of the tables, pp. 25–46.)
Reis, A. (2025). The chronology of the crucifixion: An updated Adventist view (pp. 6, 16, 18). Self-published. [read at first hand, 22 pp.]
Schwartz, E. (1905). Christliche und jüdische Ostertafeln (Abhandlungen der Königlichen Gesellschaft der Wissenschaften zu Göttingen, philologisch-historische Klasse, N.F. 8.6). Weidmann. [edition of the Sardica document, codex Veronensis 60; here cited via Stern, 2001, not inspected at first hand]
Stern, S. (2000). The Babylonian calendar at Elephantine. Zeitschrift für Papyrologie und Epigraphik, 130, 159–171.
Stern, S. (2001). Calendar and community: A history of the Jewish calendar, 2nd century BCE–10th century CE (pp. 56–76, esp. pp. 61, 70, 71, 75). Oxford University Press.
Primary texts: Matthew 16:21 (and the "third day" family) · Acts 2:20 · Joel 2:31 · Augustine, De Trinitate IV.6.10 (in translation) · Eusebius, Chronicon / Demonstratio Evangelica 8.2 · Eusebius, Historia Ecclesiastica 5.23–25 (Polycrates of Ephesus; Melito of Sardis; the quartodeciman controversy) · Peter of Alexandria (fragment on the post-70 calendar change, via Stern) · the Sardica documents (codex Veronensis 60, ed. Schwartz 1905, via Stern) · Josephus, Antiquitates 18.122–124 (the Passover anchor of AD 37). Astronomical reproduction: author's own computation (ephem/Swiss Ephemeris), calibrated; full tables in appendix A.
This book has laid its own case against the strictest measure, and built up the strongest rival, AD 30, at his best before weighing him. Honesty now asks one thing more, and it is the heaviest: the same measure against the reigning date itself. For there is one year that counts in the scholarly literature not as a candidate but as the answer — AD 33, with its famous Good Friday of 3 April. If this book charges the tradition with never having weighed AD 31, then it must indeed weigh AD 33, and with exactly the same scales on which it has measured itself. No double standard, in either direction.
So let AD 33 speak first at its strongest, for weak it is not.
Countercasethe full case for AD 33
AD 33 is not the second choice but the first choice of most scholars, and it has real strength. First the language: the dominant New Testament idiom is "on the third day", and whoever takes that as the norm and subordinates the sign of Jonah to it arrives at a Friday — and AD 33 is the best-documented Friday. Second the astronomy: the classic study of Humphreys and Waddington defends Friday 3 April 33 with a lunar eclipse that evening, a "blood moon" over Jerusalem which Peter invokes in Acts 2:20 — the most-cited astronomical argument in the entire debate (Humphreys & Waddington, 1983, 1992). Third the ministry: AD 33 requires no longer career; with a baptism at the end of 29, on the ordinary Tiberius count, it too has three and a half years. And fourth the weight of eighteen centuries: Good Friday is the day of the church, east and west. Coherent, mainstream, with an eclipse of its own. Thus AD 33 deserves to lie on the table, and thus it lies there.
Now the measure. And the first question is the one the whole book asks: is this a measurement or a majority? For the strength of AD 33 is largely a strength of consensus and of comfort — and as soon as one recomputes it astronomically and textually, with the same rigour that charges AD 31 with its Adar II, it proves less hard than it presents itself.
Begin with the most surprising fact, which is seldom mentioned: the famous Good Friday of 3 April does not stand in the standard work of the Babylonian calendar. Parker and Dubberstein — the table on which the whole chronology of this century rests — place 1 Nisanu of AD 33 on 19 April, and thus 14 Nisan on 2 May, a Saturday (Parker & Dubberstein, 1946, p. 46). Not 3 April; not a Friday. The standard authority, consistently applied, does not deliver the Good Friday at all. That is, however, the Babylonian calendar, and the link from Babylon to Judea does not hold of itself; on the same ground the 25 April of 31 cannot rest on Parker and Dubberstein. It arises only when one lets the March moon count after all — the moon that stood that year a day and a half before the spring equinox — that is, inserts no intercalary month and lets Passover fall before the equinox. That is a calendar choice, and it is precisely the kind of choice held against AD 31, only in the other direction.
Expertwhat the Friday of AD 33 hangs on
The March moon of AD 33 gave 1 Nisan around 21 March, a day and a half before the true equinox (22/23 March). Under the strict rule — Nisan begins only on or after the equinox, the rule on which Parker and Dubberstein hang their entire table — AD 33 becomes an intercalary year and 14 Nisan falls on 2 May (Saturday): no Friday. Only under the lenient rule — Passover may fall before the equinox — does the March moon stand, and 14 Nisan comes to Friday 3 April. The Good Friday of AD 33 is therefore no astronomical fact but a consequence of the weak rule; it hangs on a day and a half of indulgence. Compare that with AD 31, and note sharply exactly where the difference lies. There the March moon stood nine days before the equinox, deep in winter. AD 31 too has two options — under the strict rule the winter moon is rejected and 14 Nisan falls on Wednesday 25 April; under the lenient rule it stands and 14 Nisan falls on Tuesday 27 March (ch. 8). But neither is a Friday. There lies the asymmetry, and it is sharper than a robustness claim: the Friday of AD 33 exists under one of the two rules and vanishes under the other, whereas the Friday of AD 31 exists under neither. The consensus thus needs the indulgence to keep its own year standing, and simultaneously denies it to the year that does not even need it to lose the Friday. Which of the two days of AD 31 is the real one — the Wednesday or the Tuesday — is decided not by the equinox but by the barley, and that is the question chapter 8 measures out and which the excursus at the end of this chapter works through with an opponent. And this is a measurement, not a preference.
See what that means for the honesty of the field. The charge that AD 31 carries — that it needs a calendar liberty no one can verify, a Nisan that falls "too late" — applies to AD 33 in equal measure, only its Nisan falls "too early". Both fill in the same missing decision of the Sanhedrin; the difference is that AD 31 does it out loud and names it as an assumption, while AD 33 does it tacitly and lets it pass for a fact. And the liberty that rescues AD 33 is precisely the liberty the consensus denies AD 31: when Sacha Stern writes that one did not always wait for the equinox — "if both crops and fruits were sufficiently ripe, no intercalation would be made, even if this meant Passover occurring before the equinox. This may have happened quite frequently" (Stern, 2001, p. 70) — then that leniency rescues the Friday of AD 33, but it simultaneously strips the consensus of the right to reject the intercalary month of AD 31 "because otherwise there is no Friday". One cannot invoke the calendar liberty to save one year and refuse it in order not to see the other.
Dossierthe same March moon, three distances to the equinox
Read the last column, for there lies the whole point. Not: "AD 31 is robust and AD 33 fragile" — AD 31 itself has two outcomes. But: the Friday of AD 33 hangs on the choice between the two rules, and the Friday of AD 31 exists under neither. The choice of rule that the consensus needs for itself does not help it with AD 31.
The asymmetry is calculated, not a choice (ch. 7). AD 33 is the only one of the three in which the choice of rule overturns the weekday. Parker and Dubberstein, strictly, give 2 May; the Good Friday requires the lenient exception. [CALCULATED]
Sharper still than the arithmetic is one documentary objection, and it is the heaviest the consensus really holds against a late Passover — and thus against the April moon that makes AD 31 a Wednesday. Honesty asks that I present it at its strongest, for weak it is not.
Countercasethe sixteen March dates of Sardica
From the documents of the Council of Sardica (winter AD 343; codex Veronensis 60, discovered in 1905 by Eduard Schwartz) a list of Jewish Passover dates has been preserved for the years 328–343 — and all sixteen of them fall in March, not one in April. The series is moreover reliable: the eastern bishops included it in order to show that they did not celebrate together with the Jews, and "the polemical credibility of that argument depended on the authenticity of these Jewish dates; they are therefore unlikely to be merely invented or falsified" (Stern, 2001, p. 75) — and the list stops precisely in the year of its compilation, so that it records observed dates, not a prediction. To this is added a hostile ancient witness: Peter of Alexandria (early 4th century) argues that the Jews in Jesus' day kept the equinox rule, but after the destruction of the temple "fell into error" and began to celebrate Passover at the wrong time. On this view the base rate for a 25-April Passover is 0 out of 16 — attested, contemporary, and backed by a church father who says the early church knew better. This may not be reasoned away; it must be explained.
And it can be explained — not by me, but by the consensus authority that published these dates itself. For Stern, who weighs the sixteen Sardica dates, places them expressly outside the first century. His own section bears the heading "From the first century to the fourth: a radical change", and his conclusion is the opposite of what the series seems at first sight to say.
Expertwhat Sardica really measures
Stern reconstructs the first-century Passover as late: the celebration of Passover before the equinox, widespread in the fourth century, stands "in contrast with our findings for the first century, when Passover appears to have fallen considerably later, sometimes more than a month after the equinox" (Stern, 2001, p. 71). That reconstruction hangs on two independent first-century data points: AD 37, where the Passover — from Josephus, Antiquitates 18.122–124, with the "fourth day" and the news of Tiberius's death — fell around 19 April; and the Berenice inscription, where Tabernacles fell on 29 April, "well over a month after the spring equinox" (p. 61). The sixteen March dates are thus the fourth-century end state after a regression that Stern explains from the end of the pilgrimage and the ebbing Babylonian influence after 70 — and which the hostile witness Peter of Alexandria himself dates to after the destruction of the temple. "We may endorse the view of Peter of Alexandria that a change in the Jewish calendar reckoning took place after AD 70," Stern writes, "but for entirely different reasons" (p. 71). The 0-out-of-16 thus measures the wrong century: it is the fourth-century calendar, not the first. Applied to the first century, Stern gives the reverse — late is the norm, and 25 April (14 Nisan; 1 Nisan ≈ 12 April) lies within the band he reconstructs. [DOCUMENTED — Stern's analysis §2.3.5 + §2.5.2–3 (pp. 61–76) read at first hand; the sixteen lines of the Sardica document itself not seen line by line, but via Stern's rendering]
One limit belongs with it. A year earlier Stern himself noted a second practice: in Egypt, from Aristobulus to Philo, the Passover fell at or around the equinox, on average a month before the Babylonian date (Stern, 2000, p. 171 n. 69). That is no first-century Judean date, and it does not touch the reconstruction above; but it means that the late Passover is one of two documented practices, and which of them Jerusalem kept is the open question (ch. 8). [DOCUMENTED]
And now the honest limitation, in the same breath, for otherwise I measure with two measures. Stern's reconstruction yields no 25 April: "considerably later, sometimes more than a month" is a range, not a point date, and 25 April lies at its late edge, not in the middle. It rests moreover on two data points plus Babylonian conformity — a reconstruction, not an attested series like the sixteen of Sardica themselves. Symmetry demands that I concede it: the March dates are documented; Stern's "late first century" is inferred. But the decisive thing stands: it is the consensus scholar himself who, on non-numerical evidence, places the sixteen dates outside the first century and reconstructs that century as late. The heaviest documentary objection against a late Passover disarms itself in the standard work from which it comes. As a falsifier of 25 April, Sardica falls away; as evidence for 25 April I do not advance it — it is a rebuttal, not a confirmation. [CONSTRUED where it concerns the point date; DOCUMENTED where it concerns Stern's reconstruction]
There remains the blood moon — the astronomical showpiece of AD 33. And here honesty asks that I apply my own measure, not a stricter and not a more lenient one. Yes, on the evening of 3 April 33 there was a lunar eclipse, and the moon rose eclipsed over Jerusalem. But two things, and both hold just as hard for my own eclipse as for that of the consensus. First, that eclipse is marginal: the eclipse reached its maximum before the moon rose, so that the moon appeared low on the horizon while the umbral phase was already waning — a brief, low blood-colour, not a high, ample eclipse. Second, and decisively: the blood-moon argument is a construal, not a measurement. It links an astronomical phenomenon via Acts 2:20 to the crucifixion — precisely the kind of theological reading that this book expressly treats as an image for its own eclipse of 25 April 31, granting it not a grain of weight (ch. 7). Whoever is consistent does that on both sides. Under my own rule, then, the blood moon of AD 33 does not count as evidence for the year — no more than mine does. It is beautiful; it does not measure.
And then the family tree. AD 33 is, among the ancient writers, the choice of Eusebius — and that choice is one man deep, built on the longest ministry and anchored to the phantom witness Phlegon (ch. 6). The tradition does not remember the year; it computes it, from Luke 3:1 plus a guessed duration, and AD 33 is the outcome of the longest guess. There is, in short, no independent ancient measurement carrying AD 33. What carries it is a computation, an eclipse-as-construal, and a majority — and none of the three is stricter than what carries AD 31; they are only more familiar.
And if the family tree of the year is one man deep, then the largest patristic Easter dossier shows why that is so: the early church did not measure the crucifixion year — it quarrelled over liturgical practice.
Expertthe quartodeciman controversy does not measure the year
The heaviest second-century quarrel over Easter was not about when Jesus died, but about when one should commemorate it: on 14 Nisan itself — the Jewish date, to which the "quartodecimans" (fourteeners) held fast — or on the following Sunday. Around AD 190 Polycrates of Ephesus, in a letter to Victor of Rome, defended the 14-Nisan observance with an appeal to the apostolic tradition (John, Philip); Victor pressed the Sunday through and threatened exclusion, upon which Irenaeus urged moderation. Melito of Sardis, author of Peri Pascha, belonged to that same fourteener tradition. The whole dossier stands in Eusebius, HE 5.23–25. And what it does not contain is decisive for this book: none of these witnesses gives a consulate, a Tiberius year, an Olympiad, or even a weekday of the crucifixion. Polycrates and Melito deliver zero chronology. The struggle is about the calendar practice of the church, not about the year of the cross. [DOCUMENTED that the dispute concerns the liturgical date — Eus. HE 5.23–25, with Beckwith (1996) read at first hand; that it hands down no year is a negative observation, marked here CONSTRUED]
That is cheap but real confirmation of what the family tree already showed: whoever invokes "the tradition" for AD 33 or for the Friday invokes a tradition that hands down nothing about the year and never quarrelled over it either. It computes it, from Luke 3:1 plus an assumed length of ministry; and what it does passionately dispute — the Easter date — is a matter of liturgy, not of measurement.
ExpertAugustine's thirty-six hours, a mirror
Whoever would charge a Wednesday count with "numerology" must know one passage of the tradition itself. Augustine faces honestly the best-known shortfall of the Friday count: from the evening of the burial to the dawn of the resurrection he himself counts thirty-six hours — not three full days. And he rescues the count with a number: thirty-six is six squared, and in the ratio of the single to the double that lies within it (twelve to twenty-four) is hidden "the greatest concord of likeness" (De Trinitate IV.6.10). Matthew 12:40 he does not cite there. This is no stain on Augustine — number symbolism was in his century respectable exegesis, and his inclusive count is a genuine Jewish idiom — and it is no proof against the Friday. It is a measure: the same kind of construal one charges against a Wednesday count sits also on the side of the consensus, at the heart of its own chronology. Whoever disqualifies the one for it disqualifies the other — or marks both as construal and seeks the proof elsewhere. [CONSTRUED — a symmetry point, not an argument for the year]
Expertthe consensus claims more than it measures, on the year-axis too
The same holds beyond the weekday. Steinmann concludes that there is "no evidence for inclusive reckoning" (2022a, p. 115) — a general statement. Egypt's own corpus contradicts it: Bureth's repertory counts the first years of Gaius, Claudius and Nero inclusively, and Nero's fourteenth year — the year of his death — stands there on some thirty ostraca, whereas on the non-inclusive count it cannot exist (ch. 9). In fairness: this does not deliver autumn 27, for Tiberius is precisely the exception in Egypt. It shows only that the general claim is wider than the evidence that carries it — exactly the measure this chapter applies. [DOCUMENTED — Bureth, seen at first hand; CONSTRUED for the weighing]
And the same measure on my own side, for there is a second place where Steinmann lends support and the counterweight usually stays away. For the building years of the temple there is Josephus's statement that the priests completed the courts in eight years (Ant. 15.420) — that is the support. The counterweight stands in Josephus himself, and it belongs in the same breath: Ant. 20.219 says the sanctuary was only finished under Agrippa II — τὸ ἱερὸν ἐτετέλεστο — and that eighteen thousand workmen were then left without work. The building was therefore not finished in the twenties of the first century, and whoever cites the forty-eighth year as an end point cites an interim state. That does not make Steinmann's datum unusable; it makes it narrower than it is usually used — and that is precisely the rule this chapter applies to the consensus: grammar and the statement first, chronology only after. [DOCUMENTED — Ant. 15.420 and 20.219; ⚠ read in translation, not in the Greek]
Excursuswhere AD 33 does stand stronger, honestly conceded
Two things must go on the other side, for without them this chapter would itself be measuring with two measures. First: AD 33 reads the dominant idiom naturally. The whole third-day family — "on the third day", "after three days", "in three days", nineteen forms together, against the single "three days and three nights" — bends effortlessly to a Friday, and AD 33 need not force a single text for it. Whoever reads the sign of Jonah as a manner of speaking rather than as a measure (ch. 3) has in AD 33 a comfortable and coherent year. Second: AD 33 has no need to bend the ordinary Tiberius count; its baptism at the end of 29 follows from the best-attested reading of Luke 3:1 (ch. 9). On those two points AD 31 pays a price that AD 33 does not pay. That is the honest reverse side, and it stands here, not hidden away: AD 33 is astronomically fragile and without independent measurement, but textually comfortable and chronologically conventional. It is not a weak year. It is a conventional year that passes itself off as a measured year.
Excursusa third reconstruction, and where the calendar dispute really lies
There exists a living, published opponent of the April moon of AD 31, and he deserves a place here — not because he decides the matter, but because recomputing his work shows where the decision does fall. Richard Lanser reconstructs an "Exodus Hebrew Calendar" in which the Jewish intercalation series runs one cycle-year ahead of the Babylonian. For AD 31 that yields Nisan 1 on 14 March instead of 12 April, and thus 14 Nisan on Tuesday 27 March — a day that neither the Friday nor the Wednesday can use.
His facts are correct, and that must be said first. I have recomputed his data against the tables of Parker and Dubberstein in machine-readable form: the second Ululu of 313 BC, the intercalation of 311 BC, the two-year interval, the cycle positions — all correct. And his scheme is not ad hoc but relentlessly consistent: his Nisan 1 is each time the thirteenth lunation of the preceding Babylonian intercalary year, in every year without exception. For AD 31 that is P&D's Addaru of 14 March, for AD 33 their Addaru II of 21 March — in both cases exactly his date. He says it himself: the intercalary moons are redefined as Nisan moons.
And there is the seam, for two things he presents as one are not. His justification is a renumbering: the cycle positions of the Babylonian intercalary years, {3, 6, 8, 11, 14, 17, 19}, become {1, 4, 7, 9, 12, 15, 18} with him. His chart does something else: it shifts the whole series by one year. That the two yield the same list does not make them the same, and the difference is exactly what matters: only the shift removes the intercalary month of AD 31. The renumbering by itself does not. What that refutes is not his calendar but the ground given for it: the reason he names does not carry the outcome he records. Anyone who wants to keep his scheme may — but with a different justification than this one. [INTERPRETED]
Three observations, and the third is the only one that counts.
First. His own first step — replacing the second Ululu of AD 30 with an Adar II at the end of that Hebrew year — sets Nisan 1 of AD 31 precisely in April. It is the shift of one cycle-year that brings him back to March. The shift is therefore no ornament of his reconstruction; it is its core.
Second. His calendar diverges from the Babylonian in precisely the years that follow an intercalary year — in the window examined here, AD 31 and AD 33 — and coincides with it in all other years. In AD 33 that moves the Passover from Saturday 2 May to Friday 3 April; in AD 31 from Wednesday 25 April to Tuesday 27 March.
Third, and this is the gain of the whole recomputation. The first crescent cannot settle this dispute. His 14 March and the 12 April of Parker and Dubberstein are both genuine first crescents — two successive lunations. The crescent establishes on which day a month begins; which lunation carries the first month can be established only by the aviv. And on that we agree: he expressly appeals to it.
> The whole calendar dispute thus comes down to one testable question: did the barley in Judea stand in the aviv at the end of March? On his placement the sheaf of firstfruits falls on 29 March; on this book's on 27 April. Not the cycle decides, and not the numbering — the crop.
And that is precisely the question that Chapter 8 treats with Dalman's regional figures, with Milgrom, and with the omer rules of m. Menachot 10:2 and 10:8 — with b. Menachot 85a supporting them rather than carrying them.
The symmetry belongs here, and it is real. This book too uses a scheme as a stand-in for a practice that was observational: that Judea followed the Babylonian intercalation is an assumption and not a record. We are both reconstructing. And the symmetry must go further here than is comfortable, for two things I should like to hold against him I cannot hold against him. Not the Passover of AD 37: his calendar gives for that year the same 19 April as this book's, so that the Vitellius year confirms him just as well as this book (ch. 8). And not Parker's Jerusalem computation either: Parker likewise applies the Babylonian series to Jerusalem, and both parties thereby fill in the same undocumented decision of the Sanhedrin. Parker tests the placing of the crescent; he does not test the intercalation choice, and that choice is precisely what the dispute is about.
What remains, then? One thing, and it is enough: the difference is not rigidity but testability. This book's reconstruction hangs on something given outside the scheme — the crop — and it can fail against it. His fixes the starting point and thereafter lies fixed for seventy-two years, with nothing left against which it could fail. That is the asymmetry that counts, and it is the only one I claim. [DOCUMENTED — his calendar and the P&D tables measured through; CONSTRUED — the weighing]
Excursusthe sharpest published attack on AD 31, and whom it actually hits
Since 2025 there is a second living opponent, and he is better documented than the first. André Reis, a New Testament scholar, rejects AD 31 and arrives at Friday 3 April 33. His conclusion is one sentence: AD 31 could be the year of the crucifixion only "if it had an additional month (ruled out per the vernal equinox rule), Nisan 1 was delayed by two days (not permitted by Jewish calendrical rules) and John's chronology of the Passion is completely ignored in favor of a crucifixion on Passover Day/Nisan 15 (highly improbable)."
Two of those three grounds do not touch this book, and it is accuracy — not courtesy — that has that said here. They are written against two named defenders of AD 31. C. Mervyn Maxwell puts the crucifixion on Nisan 14 and on a Friday, and to do so must shift the beginning of Nisan by three days, from Tuesday to Friday; he concedes himself that AD 31 is a candidate only "granted a few unprovable yet reasonable assumptions." Juarez R. de Oliveira maintains a crucifixion on Nisan 15, shifts the beginning of the month by one day, and advances Passover to Thursday 26 April 31.
This book does neither. It shifts no month-beginning: 1 Nisan stands on the first visible crescent, without a correction day, and the whole computation lies open to recalculation in appendix A. And it puts the crucifixion on Nisan 14, not on 15 — precisely where Reis puts it. On the day of death we agree; the dispute is the weekday and the year. Of his three grounds, one therefore remains over against this book, the intercalary month, and that is a calendar question, computed in H8 and not here.
The paragraph stands here because without it the book inherits a refutation that was not written for it. Whoever reads Reis and then this book must be able to see that two of the three knives pass by — and that the third stays where it belongs: in the calendar. [DOCUMENTED — Reis (2025), pp. 16 and 18; ⚠ Maxwell (God Cares I, 1981, pp. 260–263) and de Oliveira (Chronological Studies, 2004, p. 46) stand here on Reis's report and reference, not consulted at first hand]
Where does the inquiry stand, then, after laying the consensus against its own measure? Not at a refutation — AD 33 is not impossible, and I do not declare it dead. Stern's "quite frequently" makes the pre-equinox Passover genuinely plausible, so that Friday 3 April 33 stays alive along the barley and weak-rule route. But that is precisely the point: AD 33 lives on exactly the same calendar liberty that carries AD 31, and its version of it is the weaker one — a March moon of a day and a half against nine, a Good Friday the standard table does not give, a blood moon that as a construal does not count, a family tree one man deep. AD 33 has no lead in rigour. It has a lead in familiarity. And those two are not the same — that is the whole thesis of this book, applied here to the date that stood in its way the longest.
With that, all the pieces lie on the table. The heaviest verse has been weighed, the strongest rival built up, and the reigning date laid against its own measure. What remains is the weighing itself — the whole field on the scales, with the photo finish shown rather than left out.
HeartHEART
It is not pleasant to test a date that millions of people have kept with reverence and love. Good Friday carries centuries of prayer, and nothing in this chapter wishes to belittle that. But it is precisely reverence for what is holy that asks for precision, not for comfort. If the tradition has somewhere passed too quickly over a measurement, it is no dishonour to show her that — it is a service to the same truth she loves. I lay AD 33 on the scales with the same hand with which I laid AD 31 on them: not to win, but to weigh honestly. And the outcome is not that the consensus errs, but that it never measured what it thought it knew. That is enough. The rest belongs to the scales.
Note on the sources
The astronomical data (14 Nisan AD 33 = 2 May under the strict rule; 3 April under the lenient one; the distances of the March moons to the equinox; the marginal eclipse of 3 April 33) are recomputable and worked out in ch. 7 and appendix A; Parker and Dubberstein have been checked in the source itself. The position of Humphreys and Waddington and that of Stern were consulted at first hand, with location. This chapter does not refute AD 33; it shows that AD 33 is at no point stricter than AD 31 — astronomically more fragile, without independent measurement, and rescued by the same calendar liberty that the consensus denies AD 31. The blood moon of AD 33 is treated by the same measure as the eclipse of AD 31: a construal, not evidence. Two documentary witnesses have also been weighed. The sixteen March dates of Sardica (winter 343) and Peter of Alexandria are drawn from Stern's analysis, read at first hand (§2.3.5, §2.5.2–3, pp. 61–76); the sixteen lines of the Sardica document itself (codex Veronensis 60, ed. Schwartz 1905) have not been inspected line by line — they carry here via Stern's rendering, and that is marked as such. The quartodeciman controversy rests on Eusebius, HE 5.23–25, and on Beckwith's discussion of it (read at first hand); HE 5.23–25 itself has not been reread line by line in this study, but HE 7.32 has (ch. 7). In both cases symmetry holds: the measure applied to AD 31 is applied here to the consensus counter-evidence — and it does not carry as a falsifier of the Wednesday. Sardica measures the fourth-century Easter reckoning, not AD 31; the quartodeciman struggle measures liturgical practice, not the year of the cross. Augustine's thirty-six-hour passage (De Trinitate IV.6.10) is cited in translation; it is marked exclusively as construal and carries no burden of proof — it serves the equal measure, not the year.
On the excursus with the third reconstruction, finally, for there the equal measure applies most sharply. The calendar of Richard Lanser has not been assessed at second hand: his published overview and his spreadsheet have been laid line by line beside the machine-readable tables of Parker and Dubberstein, and his figures have been found correct — including the Old Babylonian calibration points on which he bases his cycle numbering. What is brought against him here is thus not a calculation error but a single parameter choice, and that has been named as such. Two arguments that lay ready to hand have expressly not been used, because on recomputation they do not work against him: the Passover of AD 37, which on his calendar gives the same 19 April, and Parker's Jerusalem table, which makes the same undocumented assumption about Judea as both parties. What remains is the testability against the crop, and that is marked as CONSTRUED. This chapter claims a proof for neither side.
Literature (APA)
Beckwith, R. T. (1996). Calendar and chronology, Jewish and Christian: Biblical, intertestamental and patristic studies. Brill.
Humphreys, C. J., & Waddington, W. G. (1983). Dating the crucifixion. Nature, 306(5945), 743–746.
Humphreys, C. J., & Waddington, W. G. (1992). The Jewish calendar, a lunar eclipse and the date of Christ's crucifixion. Tyndale Bulletin, 43(2), 331–351.
Lanser, R. D. (2026). Reconstructing the original Hebrew calendar (revised overview table, 8 August 2026, with accompanying spreadsheet). Associates for Biblical Research.
Parker, R. A., & Dubberstein, W. H. (1946). Babylonian chronology 626 B.C.–A.D. 45 (Studies in Ancient Oriental Civilization 24, p. 46). University of Chicago Press. (Also in the machine-readable edition of the tables, pp. 25–46.)
Reis, A. (2025). The chronology of the crucifixion: An updated Adventist view (pp. 6, 16, 18). Self-published. [read at first hand, 22 pp.]
Schwartz, E. (1905). Christliche und jüdische Ostertafeln (Abhandlungen der Königlichen Gesellschaft der Wissenschaften zu Göttingen, philologisch-historische Klasse, N.F. 8.6). Weidmann. [edition of the Sardica document, codex Veronensis 60; here cited via Stern, 2001, not inspected at first hand]
Stern, S. (2000). The Babylonian calendar at Elephantine. Zeitschrift für Papyrologie und Epigraphik, 130, 159–171.
Stern, S. (2001). Calendar and community: A history of the Jewish calendar, 2nd century BCE–10th century CE (pp. 56–76, esp. pp. 61, 70, 71, 75). Oxford University Press.
Primary texts: Matthew 16:21 (and the "third day" family) · Acts 2:20 · Joel 2:31 · Augustine, De Trinitate IV.6.10 (in translation) · Eusebius, Chronicon / Demonstratio Evangelica 8.2 · Eusebius, Historia Ecclesiastica 5.23–25 (Polycrates of Ephesus; Melito of Sardis; the quartodeciman controversy) · Peter of Alexandria (fragment on the post-70 calendar change, via Stern) · the Sardica documents (codex Veronensis 60, ed. Schwartz 1905, via Stern) · Josephus, Antiquitates 18.122–124 (the Passover anchor of AD 37). Astronomical reproduction: author's own computation (ephem/Swiss Ephemeris), calibrated; full tables in appendix A.