The Anselm Project Bible · The Crux Registry

Luke 10:1, did Jesus send seventy, or seventy-two?

Luke 10:1 is transmitted with both seventy-two and seventy disciples. The difference is whether the small word δύο appears after ἑβδομήκοντα.

ἑβδομήκοντα δύο
hebdomēkonta dyo, seventy-two
ἑβδομήκοντα
hebdomēkonta, seventy
One small word (δύο) and the manuscripts split almost evenly. Either number carries symbolic weight, which is exactly why neither side would let go.
IThe Text

The APB text at this verse

After this the Lord appointed seventy-two others and sent them ahead of him two by two into every city and place where he himself was about to go.
Luke 10:1 (source numbering) · Anselm Project Bible
IIThe Witnesses

The manuscripts, editions, and versions

Westcott-Hort (1881)
WH · critical edition, 1881
δύο, the “two” present, seventy-two
Westcott-Hort include the word, with very early witnesses; the count recurs in verse 17. Westcott-Hort reads δύο, giving seventy-two; its note says the count recurs in verse 17. SBLGNT apparatus (Faithlife, CC BY 4.0)
Nestle-Aland 28
NA28 · critical edition, 2012
δύο, the “two” present, seventy-two
NA28 prints the word in brackets (seventy[-two]) keeping it while marking the doubt. NA28 retains δύο in brackets, printing seventy[-two] while marking doubt about the word. SBLGNT apparatus (Faithlife, CC BY 4.0)
Tregelles (1857-1879)
Treg · critical edition, 1857-1879
, no “two”, seventy
Tregelles reads seventy, with the majority of manuscripts. Tregelles omits δύο and reads seventy, with the majority of manuscripts. SBLGNT apparatus (Faithlife, CC BY 4.0)
Robinson-Pierpont Byzantine Textform
RP · Byzantine majority edition, 2005
, no “two”, seventy
The Byzantine text counts seventy, the King James's number. The Robinson-Pierpont Byzantine Textform omits δύο and counts seventy. SBLGNT apparatus (Faithlife, CC BY 4.0)
supports if seventy-twosupports if seventy
IIIThe Committee

The committee record at this verse

The APB's translation committee works blind: it is never told it is translating the Bible, never given a book name or a verse number, and it never sees the witnesses above. Since revision 3.5, the places where the manuscripts disagree are decided before translation by three readers, Grok 4.3, GPT-6 Astra, and Claude Fable 5.1, who see the readings and the witnesses by age and class but never a reference or a manuscript name. Whatever was done at this verse was done on the language and the evidence alone, and the record is public.

Textual decisionrelease apb_v3_5three blind readers

“The base text reads καὶ. The copies also carry the omission [early greek manuscript tradition]; the committee kept the base text. The external evidence favors καὶ because of its earlier and more diverse attestation. The internal evidence is nearly neutral but slightly supports καὶ as the harder reading to lose.”

“The base text reads δύο. The copies also carry the omission [greek papyrus; early greek manuscript tradition; later majority tradition]; the committee kept the base text. default chosen because early independent attestation outweighs the later omission”

“The base text reads καὶ. The copies also carry the omission [early greek manuscript tradition]; the committee kept the base text. Default chosen because the external evidence for καὶ is earlier and more widespread, and the internal evidence favors the harder reading.”

“The base text reads ⸀δύο. The copies also carry the omission [greek papyrus; early greek manuscript tradition; later majority tradition]; the committee kept the base text. The committee retains the default, but would want another reader because both readings are early and the nearby repetition of δύο makes addition a credible explanation.”

“The base text reads ⸁δύο. The copies also carry the omission [greek papyrus; early greek manuscript tradition; later majority tradition]; the committee kept the base text. The default wording is preferred. The early support for the omission prevents complete certainty, but the combined external evidence and plausible omission account favor ⸁δύο.”

Committee recordaudit verdict: clean

The APB 3.5 decision retained ⸁δύο in the committee’s printed text, which reads seventy-two. Its final note says, “The default wording is preferred.” It also records that early support for omission prevents complete certainty and that nearby repetition of δύο makes addition a credible explanation. The three blind readers saw readings and witnesses only by class and date, without a reference, book name, or manuscript name. The pericope audit verdict is “clean.”

IVThe Arguments

The argument for each reading

If seventy-two

The case for seventy-two rests on the presence of δύο in WH and NA28. In the apparatus, both entries for the disputed word align these editions on the side that includes it: “⸀δύο WH NA28 ] – Treg RP” and “⸁δύο WH NA28 ] – Treg RP.” WH’s witness note associates its wording with very early witnesses and observes that the count recurs in verse 17. NA28 keeps the word in its printed text, though its brackets signal hesitation rather than an unqualified presentation. On this reading, ἑβδομήκοντα δύο gives the number seventy-two. The symbolic force of a larger numbered mission helps explain why the precise count attracted attention.

If seventy

The case for seventy begins with the omission of δύο in Treg and RP. The same apparatus lines place both on the omission side: “⸀δύο WH NA28 ] – Treg RP” and “⸁δύο WH NA28 ] – Treg RP.” Tregelles is described as reading seventy with the majority of manuscripts, and the Robinson-Pierpont Byzantine Textform likewise counts seventy. This reading leaves ἑβδομήκοντα without the added number word. Since either total can carry symbolic weight, omission need not be treated as a casual simplification; it represents a sustained transmitted count in its own right.

The unresolved point is whether δύο belonged in the wording at Luke 10:1 or entered through transmission. The witnesses listed here divide almost evenly, and the committee record itself preserves uncertainty. The difference changes the stated number from seventy to seventy-two, but not the account’s appointment of others or their being sent two by two. Broader doctrines do not hinge on this count alone; they rest on many passages.

VQuestions

Common questions

Did Jesus send seventy or seventy-two disciples in Luke 10:1?

The transmitted readings give both answers. WH and NA28 include δύο and read seventy-two; Treg and RP omit it and read seventy.

What word creates the difference in Luke 10:1?

The difference is δύο. With it, the wording is ἑβδομήκοντα δύο, seventy-two; without it, the count is ἑβδομήκοντα, seventy.

What does the APB printed text say?

The APB printed text says that the Lord appointed seventy-two others. Its 3.5 decision retained ⸁δύο while noting uncertainty about the omission.

VIThe Registry

The registry gives no verdict

The registry does not say which reading is original. Each page lists the witnesses, shows what the committee did with the same text, and leaves the judgment to the reader.