The week with two sabbaths

Let the mainstream reading speak here at full strength, for it is not weak. First, the other gospels appear to present the last supper as the Passover meal itself — "on the first day of unleavened bread" (Mark 14:12) — and then only one sabbath is needed: the ordinary weekly one, with the crucifixion on the Friday before it. Second, John 19:31 can also be read within that single sabbath: the weekly sabbath which that year coincided with the feast was for that very reason a "great" day — "great" not because it was a separate festival sabbath, but because an ordinary sabbath falling in the Passover week carried extra weight. Third, the spice sequence can be harmonised without a second sabbath: one assumes that the women bought and prepared on Friday evening, after the sabbath had begun yet, in Mark's recollection, "when the sabbath was past" — a common, though not effortless, reconciliation. In short: the two sabbaths are not compelled by any of these texts. That must be said honestly, and it stands.

The scholar who saw the spice contradiction most sharply is Adalbert Merx. In his commentary he names the crease precisely: Mark has the women buy "when the sabbath was past," while Luke places the purchase before the sabbath begins and then has them rest (Merx, 1905, p. 509). That he establishes the tension independently — without sharing this book's conclusion — is the reason I cite him: as a witness that the crease is real, not as an authority for what lies beneath it. For his own solution — a difference among the evangelists themselves — is itself a reconstruction, exactly as the two-sabbath reading is one. It is not measured but chosen, and it does not exclude the possibility of two feast days or sabbaths; it merely lays something else beneath the crease. The measure this book applies to itself — "this is a reconstruction, not proof" — therefore applies just as strictly to Merx. My argument does not lean on him here: the shared observation is the crease, and the two sabbaths are the most economical answer to it; Merx's different answer is no refutation of that, but a second reconstruction alongside the first.

There was a temptation here, and because it was beautiful and suited me well, the reader ought to know that I went after it and that it does not hold.

Leviticus 23:11 has the sheaf of firstfruits waved "on the day after the sabbath" (מִמָּחֳרַת הַשַּׁבָּת), and over that one word the parties divided. The Sadducees and the Boethusians — and the Samaritans to this day — read "the sabbath" as the weekly sabbath: then the sheaf always falls on a Sunday. The Pharisees read "the sabbath" as the feast day of 15 Nisan: then the sheaf falls on 16 Nisan, whatever weekday that may be. If the sheaf always fell on a Sunday, then the resurrection "on the first day of the week" is the day of the sheaf of firstfruits — and then Paul names Him by the name of precisely that offering: "ἀπαρχὴ τῶν κεκοιμημένων", "the firstfruits of them that slept" (1 Corinthians 15:20). A splendid thought. And the high priests of the year 31 were Sadducees — Luke calls them by that name: "ἡ οὖσα αἵρεσις τῶν Σαδδουκαίων" (Acts 5:17).

But the field is not twofold, and that must be said at once. The standard commentary on Leviticus — Jacob Milgrom in the Anchor Yale Bible — counts four: *"whether the ʿomer is brought on the second day (the Pharisees), the following Sunday (Samaritans), the Sunday after the festival (Qumran), or whenever the grain ripens (conjectured original practice)"* (Milgrom, Leviticus 23–27, p. 1989). [DOCUMENTED — primary source consulted] Chwolson too knew that third position: the Syriac translation of Leviticus 23:11 and the book of Jubilees read "the twenty-second," the first day after the seventh feast day.

And note what Milgrom reconstructs as the presumably original practice: "whenever the grain ripens." No fixed day, no fixed month, no cycle — the sheaf comes when the crop is ripe.

Here is a temptation I expressly leave lying, for it would shoot this book in the foot. It is tempting to say: look, the standard commentary itself formulates the crop-standard against which ch. 8 tests the calendar. But that confuses two things that must be kept apart. Milgrom here reconstructs a pre-exilic origin. In the temple of the first century the day was precisely fixed — m. Pesachim 4:8 speaks in so many words of "the fixed day for the ʿomer offering." And it is that fixing which makes the argument of ch. 8 possible: precisely because the sheaf had to come on a fixed day of the month, the ripeness of the barley becomes a test on the placing of the month — there is no evading it by waiting. Were Milgrom's floating sheaf to hold for AD 31 as well, then an omer at the end of March is no anomaly, for then the farmer simply waits until the field is ready. What ch. 8 tests is the intercalation, not the counting — and that distinction must stand before anyone uses it against this book.

I leave the Sadducean convergence lying, and the reason is not that that counting is refuted — it is not — but that the convergence measures nothing.

And on the linguistic hinge Milgrom is sharper than I dared to be. He closes his treatment of the whole pericope with: *"The resolution, I submit, is that *šabbat* has been wrongly translated. In vv. 11–16, it bears only one meaning, the sabbath-week, not the sabbath day" (p. 2060). And he points to Deuteronomy 16:9 as the oldest form of the counting — "mehahel ḥermeš baqqamah", "when the sickle is first put to the standing grain" — with the remark that "the language of the clause betrays the origin of this hag as a private celebration, dependent on the day when the individual farmer began to harvest his grain"* (p. 2059). [DOCUMENTED]

First the case of the opposing party, for it seems decisive and is not entirely so. Three witnesses point to the Pharisaic date. The Septuagint renders "the day after the sabbath" as "τῇ ἐπαύριον τῆς πρώτης" and "seven sabbaths" as ἑπτὰ ἑβδομάδας ὁλοκλήρους, seven weeks. Philo, a contemporary, dates the sheaf to "ἡ μετὰ τὴν πρώτην εὐθὺς ἡμέραν" (Spec. Leg. 2.162). Josephus, a priest, gives not the weekday but the calendar date — "ἕκτη δ' ἐστὶν αὕτη καὶ δεκάτη", the sixteenth — and does so in the present tense (Antiquitates 3.250). Add to this that shabbat in Hebrew itself can also mean "unit of seven": Leviticus 25:8 counts "seven sabbaths of years" and calls that forty-nine years. [DOCUMENTED]

But Chwolson knew those three witnesses and answered all three, and whoever dismisses him ought to know with what. Josephus holds, on his account, "nur für seine Zeit" — the last decade of the temple, not the year 31; on the Septuagint he points to the marginal readings in Field's Hexapla, which carry the sabbath reading; and his chief argument is historical: the old festival calendar Megillat Taanit sets apart the days from 8 to 22 Nisan in memory of the victory of the Pharisees in this very question, and "ein Sieg setzt einen Kampf voraus" — a victory presupposes a struggle. [DOCUMENTED — Chwolson 1892, pp. 61–62, seen at first hand] I do not take that dispute on here; I let it stand — for it decides nothing in this book, and the next paragraph shows why.

I do not decide that dispute here, and I need not decide it. For two things stand fast, whichever counting one holds.

The first is arithmetic. If the sheaf always fell on a Sunday, then it fell on a Sunday in AD 30 as well, and in AD 33. A quantity that is the same in every year cannot choose between years. Its probative force for this year is exactly nil. [CALCULATED]

The second is the reason I wanted it, and that must be said aloud. On the chronology of this book — 14 Nisan on Wednesday 25 April — 16 Nisan is a Friday. On the Pharisaic counting the sheaf then does not fall on the day of resurrection; only the Sadducean counting saves the convergence. That is precisely the reason not to build on it. It would have been chosen because it comes out right. That is the error this book points out in the consensus; I may not commit it myself here — not even when the chosen side is defensible, and it is.

What remains is an open dispute that does not touch the year. It was still running in the first century: the Mishnah must formulate a rule of conduct for the high priest "so that one may not confirm the words of those who say: Atseret falls after the sabbath" (Chagigah 2:4), and in the Talmud every name that refutes the Boethusians is from after the destruction — after which Rava, in the fourth century, himself demolishes most of those proofs ("כּוּלְּהוּ אִית לְהוּ פִּירְכָא", b. Menachot 66a). There is no witness who says that the temple in the year 31 counted in the Pharisaic way, nor any who says it counted in the Sadducean way. I therefore hold in this book to the ordinary counting — the sheaf on 16 Nisan — because it is the best attested, and on the firstfruits typology I build no date, only what it always already was: an image. [CONSTRUED]