How many of Arah's offspring returned from Babylon?

What follows is one scholar’s way of reading this disagreement — a critical-historical view that often concludes the texts genuinely disagree. It draws on the peer-reviewed sources listed below and stands as one position among the answers above, presented for transparency, not as GEBible’s position.

How critical scholarship weighs this: Leans toward a genuine contradiction

This summarizes where critical scholarship tends to land on this tension — a report on the field’s center of gravity, not GEBible’s own ruling.

Ezra 2 and Nehemiah 7 are two versions of a single document: a register of the families who returned from Babylonian exile under the Persian authorization associated with Cyrus. The relationship is literary, not merely topical — the two lists share the same family names in essentially the same order, the same structure (laypeople by clan, then priests, Levites, temple servants, etc.), and the same grand total of 42,360 (Ezra 2:64; Nehemiah 7:66). One is copied from the other or both descend from a common archival Vorlage. Nehemiah 7:5 even frames its list as a found document ("I found the register of the genealogy of them which came up at the first"). When two copies of the same record disagree on a number, the disagreement is a transmission phenomenon, and these two chapters disagree on roughly twenty individual family totals — Arah being one, at 775 (Ezra) versus 652 (Nehemiah), a gap of 123.

The apologetic harmonization, advanced by Gleason Archer and elaborated in John W. Haley's nineteenth-century Examination of the Alleged Discrepancies, proposes that the two lists were compiled at different moments: Ezra records an initial registration (perhaps in Babylon, or of those who enrolled to go), while Nehemiah records the count of those who actually arrived, the differences attributable to deaths en route, defectors, or latecomers. The theory is internally coherent and would, if the texts permitted it, dissolve the discrepancies. The obstacle is that both headings describe the same thing. Ezra 2:1 introduces "the children of the province that went up out of the captivity... and came again unto Jerusalem." Nehemiah 7:6 introduces "the children of the province, that went up out of the captivity... and came again to Jerusalem." Both are, on their own self-description, lists of those who completed the return — not a before-and-after pair. The timing distinction is supplied from outside the text to rescue the numbers.

The critical reading, set out by Hugh Williamson in the Word Biblical Commentary and Joseph Blenkinsopp in the Old Testament Library volume, treats the divergences as ordinary numerical corruption between two copies of one source. A decisive internal datum supports this: the round total of 42,360 is identical in both books, yet the individual sub-totals add up to neither that figure nor each other in either book (the laity sum to roughly 24,000-30,000 depending on the version). The shared round total preserves one tradition while the component numbers drifted independently during copying. Loring Batten's ICC commentary tabulated every divergence and showed the differences run in both directions and across a wide range — precisely the random pattern expected from scribal error, not the directional pattern a single attrition narrative (deaths in transit) would produce.

The mechanism is well understood. Hebrew numerals in this period were sometimes written out as words and sometimes, in administrative documents, with numerical signs or letter-abbreviations vulnerable to confusion; a damaged or misread stroke could shift a figure by tens or hundreds. Ralph Klein's standard handbook on Old Testament textual criticism documents how readily numeric data corrupts in transmission, and the Ezra/Nehemiah lists are a textbook case studied precisely because the two witnesses survived side by side in the canon.

The verdict: the harmonization requires a registration-versus-arrival distinction that both headings explicitly contradict, and it would predict a directional drift the data do not show. The transmission-error reading accounts for both the scattered, bidirectional divergences and the preserved-but-inconsistent grand total. The two numbers for Arah cannot both be correct; at least one is a copyist's error.

Where the critical reading lands: genuine numerical discrepancy, best explained as scribal corruption in parallel copies of one source.