Werner VorstmanInstruments for reading Scripture closely
Instrument VII

The Torah Reading

What is read this week: the portion of the Torah, the haftarah that goes with it — in the Ashkenazi custom and the Sephardi one — and a passage from the new covenant. The first two are old and are copied here. The third is computed, and the instrument shows what it rests on.

Three readings, and who chose them

The division of the Torah into fifty-four portions is not mine. It is old, it is settled, and it covers the five books with no gap and no overlap: Genesis 1:1 through Deuteronomy 34:12, one after another. That is not a claim but a test carried out at build time — the fifty-four ranges are laid against Moria’s own chapter and verse counts, and a single mistyped verse breaks the contiguity and stops the build. 5,852 verses fit inside.

The haftarah, the passage from the prophets read after the Torah, is not mine either. Ashkenazi and Sephardi congregations choose a different passage in a dozen or so places, and sometimes a different book altogether; so they stand side by side here rather than my picking one. Where the two customs agree — in most portions — there is one line.

The passage from the new covenant is mine, or rather the method’s. Lists of New Testament passages to read with each portion circulate widely and contradict one another, because each list is the judgement of whoever drew it up. Here the passage is computed from the links Moria itself has measured, and behind every reading a fold opens showing which verses hold each other, and why.

How that passage is chosen

For every verse of the Torah portion, Moria is asked which New Testament verses are joined to it, along three roads:

Citation through the Septuagint. The New Testament verse laid against the Greek text of the Old. This is citation evidence at the level of the original tongue, and it weighs heaviest here.

Citation through the Statenvertaling. Rare four-word phrases occurring in both verses. Citation evidence in a translation a reader has in front of him.

Classical cross reference. The references out of the tradition, scored by the votes they gathered. That is a judgement and not a measurement, so it counts for about half.

The verses that rise this way are grouped into passages — neighbours no more than three verses apart, inside one book — and the passage with the most evidence wins. Then two corrections come in, and without them the result is nonsense.

The first: a verse thick with proper names is damped. A genealogy looks like a genealogy and a tribal list like a tribal list, on the names alone. That is an artefact of the concordance, not a reading. Moria’s own name register supplies the count.

The second: a passage is given away only once. The cycle is walked in the order the Torah runs, and each portion takes the best passage still free. Without that rule Hebrews 11 becomes the answer to half of Genesis. With it something pleasant happens: where one New Testament chapter retells a long stretch of the Torah — Stephen before the council does exactly that — the successive portions receive successive slices of that chapter.

Nothing is steered by hand afterwards. Where the evidence is thin the instrument says so, and names the runners-up.

Two numberings, and why two are sometimes shown

Jewish sources give these references in the Hebrew numbering. The translations on this site — Statenvertaling, King James, Berean Standard, Reina-Valera 1909, Versión Biblia Libre — count differently in five places in the Torah, and in a handful more among the prophets. Genesis 32:1 in Hebrew is Genesis 31:55 in yours; Leviticus 6:1 is Leviticus 5:20; Numbers 30:1 is Numbers 29:40; Deuteronomy 29:1 is Deuteronomy 28:69.

Everything you see here is in your own Bible’s numbering, because otherwise you would not find it. Where the two part company, the Hebrew numbering follows in brackets, so that you can look it up in a Jewish edition too.

Why the cycle runs differently here

The cycle turns on the eighth day of the seventh month — the day Leviticus 23:36 sets apart at the close of Tabernacles. The last portion is read that day, and the first on the sabbath after it. The sabbaths in between carry the rest.

Not all of them. A sabbath falling inside the week of Unleavened Bread or the week of Tabernacles, or one that is itself a holy convocation, carries no weekly reading: there the reading yields to the feast. That is not a nicety but an arithmetical necessity. A biblical year has twelve or thirteen lunar months and therefore fifty to fifty-five sabbaths, and fifty-four portions do not simply fit.

Whatever difference is left over is absorbed the way it has always been absorbed: by joining portions in pairs. Seven pairs are eligible, in the order in which the received calendar actually joins them — Matot with Masei first, Chukat with Balak last. Above the table stands how many are joined this year. In a long year that is none or one; in a short year five or six. That is exactly the picture the received calendar gives, and it is not copied here but computed.

The sabbath itself runs from sunset to sunset, and the biblical month opens at the new moon as it falls at Jerusalem — the same reckoning the strip at the top of this site, the counters on the front page and the calendar page all use. The three cannot come apart.

What this instrument is not

It is not a halakhic ruling. It follows this site’s calendar and not the rabbinic one, and the two do not coincide every year. Anyone who wants a synagogue’s reading should ask there.

Nor does it know the special sabbaths that carry their own haftarah in the tradition — Shekalim, Zachor, Parah, HaChodesh, HaGadol, Shuvah. In practice those replace the ordinary haftarah on a handful of sabbaths a year. Here the haftarah of the annual cycle itself always stands. The seven haftarot of consolation between Devarim and Nitzavim are the annual cycle, so those are in.

And the passage from the new covenant is an outcome, not a pronouncement. It says: these are the verses lying nearest this portion, measured along citation and reference. Whether that is the reading you want is for you to judge — and the evidence lies open for it.

Moria, where the links come from →  ·  The calendar, for the whole year →  ·  Texts and datasets used →