Genealogy and Times: From Creation to Christ
Jesus's genealogy is recorded twice (Matthew 1, Luke 3), and Luke's runs all the way back to Adam. That claim is only as strong as the data behind it, so this study does the work directly: it reads Genesis 5 and 11:10-26 — the only two places in the Old Testament that give a father's age at his heir's birth, not just a name — in all three surviving textual witnesses (the Masoretic Text, the Septuagint, and the Samaritan Pentateuch), pulled straight from this site's own primary-source database rather than from secondary comparison tables. The raw data and the reasoning documented here feed docs/data/genealogy's structured files; this page is where the why behind those numbers lives.
Two things are true at once, and this study tries to hold both without collapsing either into the other: the genealogy is a real chronological record, precise enough to argue over and capable of being wrong in transmission — and it is also a theologically shaped document, tracking a single promised line (Genesis 3:15's "seed of the woman") through named individuals whose names themselves carry meaning. Getting the math right and hearing what the names say are not competing projects.
How Genesis actually gives this data
Genesis 5 (Adam to Noah) and Genesis 11:10-26 (Shem to Terah) share a distinctive formula, repeated once per patriarch: he lived [age], and fathered [heir]; he lived [years] more after fathering [heir], and had other sons and daughters; all his days were [total]. That formula is what makes three-way manuscript comparison possible at all — nowhere else in Scripture is this much chronological data given about this many consecutive individuals. It also stops cold after Terah: Genesis 11:26 gives his age, but from Abraham onward the text gives ages at specific named events (Abraham 100 at Isaac's birth, Genesis 21:5; Isaac 60 at Jacob's, Genesis 25:26) rather than a systematic per-generation formula. That's a real change in genre, not a gap in the data — the rest of this study's method (comparing an "age at heir's birth" figure across manuscripts) simply doesn't apply past Terah, because Genesis stops giving one.
The three witnesses
- The Masoretic Text (MT) — the standard Hebrew text underlying most English translations,
and the one this site's own
zadok_yearnumbering already assumes (see The Zadok Calendar for the calendar side of that convention). - The Septuagint (LXX), in the Brenton edition — the pre-Christian Greek translation, whose Genesis 5 and 11 numbers diverge from MT's in a strikingly patterned way (below).
- The Samaritan Pentateuch (SP) — preserved independently by the Samaritan community, and the least commonly consulted of the three, but not the least interesting: it resolves two real problems that MT and LXX both leave open (see Methuselah and Terah below).
All figures below were queried directly from this repo's references/build/bible-text.db
(morphhb-wlc for MT, ebible-grcbrent for LXX, scrollmapper-SP for SP) and cross-checked by
computer against each tradition's own stated total (age at heir's birth + years after = total
lifespan — see references/build/genealogy_chronology.py, which fails loudly if a figure
doesn't add up). Every number here passed that check.
Genesis 5: Adam to Noah
| Patriarch | MT (age / after / total) | LXX (age / after / total) | SP (age / after / total) |
|---|---|---|---|
| Adam | 130 / 800 / 930 | 230 / 700 / 930 | 130 / 800 / 930 |
| Seth | 105 / 807 / 912 | 205 / 707 / 912 | 105 / 807 / 912 |
| Enosh | 90 / 815 / 905 | 190 / 715 / 905 | 90 / 815 / 905 |
| Cainan (Kenan) | 70 / 840 / 910 | 170 / 740 / 910 | 70 / 840 / 910 |
| Mahalalel | 65 / 830 / 895 | 165 / 730 / 895 | 65 / 830 / 895 |
| Jared | 162 / 800 / 962 | 162 / 800 / 962 | 62 / 785 / 847 |
| Enoch | 65 / 300 / 365 (no death) | 165 / 200 / 365 (no death) | 65 / 300 / 365 (no death) |
| Methuselah | 187 / 782 / 969 | 167 / 802 / 969 | 67 / 653 / 720 |
| Lamech | 182 / 595 / 777 | 188 / 565 / 753 | 53 / 600 / 653 |
| Noah (age at Shem/Ham/Japheth) | 500 | 500 | 500 |
The pattern for six of these nine (Adam through Mahalalel, and Enoch) is remarkably clean: LXX adds exactly 100 years to the age-at-heir-birth figure and subtracts the same 100 from years-after, so the total lifespan is identical across MT, LXX, and SP every time. That's a systematic shift, almost certainly deliberate on someone's part, not manuscript noise — noise doesn't reproduce a constant offset nine times running with a matching total.
Jared, Methuselah, and Lamech break that pattern, each differently. Jared is untouched by the LXX shift (MT and LXX agree exactly) but SP shortens both his age and his total. Methuselah gets the standard MT/LXX same-total treatment (969 both) but SP shortens the total, not just the split. Lamech is the strangest: all three traditions give a different total (777 / 753 / 653) — no clean two-agree-one-differs pattern at all.
Genesis 11:10-26: Shem to Terah
| Patriarch | MT | LXX | SP |
|---|---|---|---|
| Shem | 100 / 500 / 600 | 100 / 500 / 600 | 100 / 500 / 600 |
| Arphaxad | 35 / 403 / 438 | 135 / 400 / 535 | 135 / 303 / 438 |
| (Cainan) | — (absent) | 130 / 330 / 460 | — (absent) |
| Shelah | 30 / 403 / 433 | 130 / 330 / 460 | 130 / 303 / 433 |
| Eber | 34 / 430 / 464 | 134 / 270 / 404 | 134 / 270 / 404 |
| Peleg | 30 / 209 / 239 | 130 / 209 / 339 | 130 / 109 / 239 |
| Reu | 32 / 207 / 239 | 132 / 207 / 339 | 132 / 107 / 239 |
| Serug | 30 / 200 / 230 | 130 / 200 / 330 | 130 / 100 / 230 |
| Nahor | 29 / 119 / 148 | 179 / 125 / 304 | 79 / 69 / 148 |
| Terah | 70 / 135 / 205 | 70 / 135 / 205 | 70 / 75 / 145 |
A different, equally consistent pattern shows up from Shelah through Serug: SP takes LXX's higher age-at-heir-birth figure but keeps MT's total, by shortening years-after to compensate. Four consecutive patriarchs do this identically — that's a real editorial signature, not coincidence. Eber breaks it (SP just matches LXX outright, total included). Nahor breaks it a third way (three different ages, though SP's total still matches MT's). And Terah — the last one, and the most consequential — breaks it in the direction that matters most for everything downstream.
Case studies
The Cainan question
Luke 3:36 (ESV) names a Cainan between Arphaxad and Shelah. MT and SP don't have him; LXX does, with his own full entry (130 years to Shelah's birth, 330 more after, 460 total — Genesis 11:13 LXX). Luke is following LXX here, not MT. That's not a contradiction to explain away — it's evidence Luke's genealogy is textually dependent on the Greek tradition at this specific point, which matters for anyone using Luke's list to reconstruct an exact year count: if Cainan is a real generation MT and SP dropped, any MT-based total from Adam to Christ is short by roughly this generation's span; if he's a later Greek-tradition addition, LXX's own total (already the longest of the three by a wide margin) inherits an extra generation MT/SP never had. Neither option is resolved by more counting — it's a text-critical question, not a math one.
Methuselah: the name, the number, and the Flood
Methuselah's name (מְתוּשֶׁלַח) is ambiguous at the lexical level — not "one attested reading and one folk etymology," but two real readings built from real roots. Read as m'tei ("men of") + shelach ("javelin," H7973), it's a plain warrior name with no theological freight. Read as mut ("die," H4191) + shalach ("send," H7971), it becomes a sentence-name: "his death shall send [it]." Both parse correctly; nothing in the lexicon settles which one the name-giver intended.
What tips the scales toward the second reading isn't etymology, it's arithmetic — and this is exactly where the textual traditions disagree most sharply. MT and LXX both total Methuselah's life at 969 years (the longest in the record), but via different splits. Run that 969 forward on either tradition's own numbers for Lamech and Noah, and Methuselah is still alive more than a decade after the Flood begins — a long-noted, real problem, not an invented one. If the name really does mean "his death shall send [judgment]," a man who outlives the judgment his own name predicted is an awkward result. SP alone avoids it: its total for Methuselah is 720, not 969 — short enough that his death lands in the very year of the Flood. If the eschatological reading of the name is right, SP is the one tradition where the name and the math actually agree without further harmonizing.
Terah and Abram: a puzzle two different ways
This is the most consequential single data point in the whole survey, because it's not a disagreement about a name's meaning — it's an internal tension inside the text of Genesis itself, and MT and SP resolve it in two structurally different ways.
Genesis 11:26 states Terah was 70 when he fathered Abram (named first, alongside Nahor and Haran — birth order not stated). But Genesis 12:4 has Abram leaving Haran at 75, and Acts 7:4 (ESV) is explicit that this happened after Terah's death. Under MT/LXX's stated 205-year total for Terah (born 1876 zadok on MT's own numbers), the plain "70 at Abram" reading puts Abram's departure at zadok year 1946+75=2021 — a full 60 years before Terah actually dies at 2081. That's a real contradiction with Acts 7:4, not a rounding issue, and it's why the standard harmonization reinterprets Genesis 11:26: Abram is listed first by covenant significance, not birth order, and was actually the youngest of the three sons, born when Terah was 130 (1876+130=2006; +75=2081 — now it lines up exactly with Terah's actual death year).
SP's total for Terah is 145, not 205. Run the same plain 70-year reading Genesis 11:26 states outright — no reinterpretation, no assumption about birth order — and 70 + 75 = 145 exactly. Terah's death and Abram's departure land on the same year without any harmonizing move at all. Both resolutions work internally. MT/LXX's requires reading past the plain sense of one verse to save the numbers; SP's numbers already match the plain sense of that same verse. That asymmetry is worth sitting with rather than picking whichever number is more familiar.
Word studies: what the names actually say
Popular teaching sometimes strings the Genesis 5 names into a sentence — Seth (appointed), Enosh (mortal man), Kenan (sorrow), Mahalalel (the Blessed God), Jared (shall come down), Enoch (teaching), Methuselah (his death shall bring), Lamech (the despairing), Noah (rest/comfort) — read together as something like: Man is appointed mortal sorrow, but the Blessed God shall come down, teaching that his death shall bring the despairing rest. It's a striking claim, and worth checking against the lexicon rather than repeating on the strength of how well it reads.
Querying this repo's own TWOT root data (references/build/twot_lookup.py) name by name, rather
than trusting the chain as a whole:
| Name | Verdict | What's actually attested |
|---|---|---|
| Seth | Solid | "Appointed" — Genesis 4:25 states the wordplay itself; not reconstructed |
| Enosh | Solid | "Mortal man" — frailty-connoting, distinct from adam/ish, but inferred from usage elsewhere, not an in-text gloss |
| Kenan | Speculative | No TWOT root for the name at all; "sorrow" rests on an unconfirmed link to a different root |
| Mahalalel | Solid | "Praise of God" — transparent theophoric compound |
| Jared | Solid | "Descent" — direct nominal form of "to come down" |
| Enoch | Solid, imprecise popularly | "Dedicated/initiated" (same root as Proverbs 22:6) — "teaching" is a loose paraphrase |
| Methuselah | Ambiguous, both real | See above — not one solid reading and one invented one |
| Lamech | Unknown | No TWOT root exists; standard lexicons mark the derivation uncertain |
| Noah | Solid, needs precision | Genesis 5:29's own wordplay uses nacham ("comfort"), not nuach ("rest") — a real double sound-play, not a simple derivation |
So: six of nine names hold up on their own lexical merits, one (Methuselah) is a real, motivated ambiguity rather than a coin-flip, and two (Kenan, Lamech) have no lexical footing for the reading the popular chain wants from them. That's worth saying plainly rather than smoothing over. It doesn't wreck the pattern — six solid, theologically resonant names in a row (appointed, [frail] man, praise of God, shall come down, dedicated, comfort) is still a real feature of the text, not manufactured. But claiming a complete nine-word sentence requires filling two genuine gaps with unattested glosses, and the temptation to do that — to round "suggestive" up to "complete" — is itself worth naming as a temptation, not just quietly resisting it.
Toward a most probable timeline
Three real manuscript traditions and a fourth option — synthesizing rather than simply picking
one — are laid out in docs/data/genealogy/index.json's timeline_variants. This study's
working position, harmonized_v1, takes MT as the base (matching this site's existing
zadok_year convention and docs/data/events.json) and adopts SP's reading in exactly two
places:
- Methuselah — because SP is the only tradition where his death doesn't survive the Flood his own (possible) name predicts.
- Terah — because SP's total resolves the Abram-departure puzzle without requiring Genesis 11:26 to be read against its own stated birth-order.
Both substitutions share the same justification: they're adopted because they resolve a
demonstrable internal problem in the base reading, not merely because SP is available, older in
places, or shorter. That's a deliberate filter. SP has other divergences — the Shelah-through-
Serug redistribution pattern, the three-way split at Nahor — that this study does not adopt
into harmonized_v1, because nothing about those specific numbers resolves a contradiction the
way Methuselah and Terah's do. Adopting a reading only where it earns its keep, and leaving the
rest alone, is the whole point of naming this a proposed synthesis rather than just crowning
one manuscript the winner.
Computed results (references/build/genealogy_chronology.py, zadok year 0 = 4004 BC = Adam's
creation in every variant):
| Variant | Flood (zadok / Gregorian) | Terah's death (zadok / Gregorian) |
|---|---|---|
| MT | 1656 / 2348 BC | 2081 / 1923 BC |
| LXX | 2242 / 1762 BC | 3547 / 457 BC |
| SP | 1307 / 2697 BC | 2322 / 1682 BC |
| harmonized_v1 | 1536 / 2468 BC | 1901 / 2103 BC |
The gap between MT's and LXX's Flood dates alone is 586 years — a reminder that "the biblical
timeline" is not a single settled number even before archaeology enters the picture. This is
presented as the working answer, not the final one — the state file behind this study
(references/study-state/genealogy-times.yml) tracks it as harmonized_v1, open to revision if
a better-justified case for a different substitution turns up.
One outstanding item, deliberately not smoothed over yet: The Day is Near uses a different creation epoch again (~3925 BC) than any of the four variants above. Reconciling that is future work, tracked but not resolved here.
The Exodus-to-Solomon gap: genealogy as a check on the Judges chronology
Genealogical age-data stops at Terah. From here to Solomon, the chronological evidence changes character entirely: instead of a systematic formula, Scripture gives one summary figure (1 Kings 6:1: 480 years from the Exodus to Solomon's 4th year) and a long list of individual judges' and oppressors' reign-lengths in between. Those two kinds of evidence don't agree with each other on a naive reading, and two different genealogies — one priestly, one royal — turn out to pull in opposite directions on how to resolve it.
The problem: the numbers don't add up to 480
Summing every individually-stated figure in Judges, in the order given:
| Event | Years | Verse |
|---|---|---|
| Cushan-Rishathaim oppression | 8 | Judges 3:8 |
| Othniel / land's rest | 40 | Judges 3:11 |
| Eglon (Moab) oppression | 18 | Judges 3:14 |
| Ehud / land's rest | 80 | Judges 3:30 |
| Jabin (Canaan) oppression | 20 | Judges 4:3 |
| Deborah/Barak / land's rest | 40 | Judges 5:31 |
| Midian oppression | 7 | Judges 6:1 |
| Gideon / land's rest | 40 | Judges 8:28 |
| Abimelech | 3 | Judges 9:22 |
| Tola | 23 | Judges 10:2 |
| Jair | 22 | Judges 10:3 |
| Ammon oppression (east, Gilead) | 18 | Judges 10:8 |
| Jephthah | 6 | Judges 12:7 |
| Ibzan | 7 | Judges 12:9 |
| Elon | 10 | Judges 12:11 |
| Abdon | 8 | Judges 12:14 |
| Philistine oppression (west) | 40 | Judges 13:1 |
| Samson (during the Philistine oppression) | 20 | Judges 15:20; 16:31 |
That's 410 years for Judges proper. Add the wilderness wandering (40, fixed), roughly 7 years for Joshua's conquest (derived from Joshua 14:7, 10: Caleb was 40 at the spies' mission, 85 "45 years" later, and 38 of those 45 were the imposed wilderness delay, leaving ~7 for the conquest itself), Eli's 40 years judging Israel (1 Samuel 4:18), an unspecified stretch of Samuel's own ministry before the monarchy, Saul's reign, David's 40 (2 Samuel 5:4-5), and Solomon's 4 years to the temple, and the sequential total comfortably exceeds 480 before Samuel's own years are even counted — by at least 100 years, likely more. This is a long-recognized problem, not a new one, and it has a name in the scholarly literature: the "Judges chronology problem."
Saul's own reign-length can't even be read off the Hebrew text as it stands: 1 Samuel 13:1 reads, transliterated, "Saul was a son of a year when he began to reign, and two years he reigned over Israel" — a well-known textual lacuna, not a translation choice. A number has dropped out of the Masoretic transmission at Saul's age, and "two years" for his whole reign is implausibly short given everything the text elsewhere attributes to it. The traditional 40-year figure comes not from Samuel but from Acts 13:21 (ESV), where Paul states it plainly: God "gave them Saul... for forty years." Worth knowing when that number is used: it's patching a real gap in the Hebrew manuscript tradition, not resolving an ambiguity within it.
The textual basis for overlap — this isn't invented
The standard resolution treats several of these judgeships as regional rather than national, and therefore overlapping in time rather than strictly sequential. That's not a modern harmonizer's convenience — the text says so directly. Judges 10:7-9 (ESV) states that God "sold them into the hand of the Philistines and into the hand of the Ammonites" in the same breath, with the Ammonite oppression explicitly located "beyond the Jordan... in Gilead" (east) for 18 years, while Philistine pressure came from the west. Jephthah, Ibzan, Elon, and Abdon's combined 31 years belong to the eastern/Gilead side of that same double-oppression; Samson's 20 years explicitly take place "in the days of the Philistines" (Judges 15:20) — he judges during the 40-year Philistine oppression, not after ending it, and 1 Samuel's own narrative shows Philistine dominance continuing well past Samson's death, through Eli's era and into Samuel's early ministry. None of this requires inventing an overlap the text doesn't support; it requires taking the text's own geography seriously instead of defaulting to a single linear national timeline the book of Judges never actually claims to be giving.
Genealogy check #1: the priestly line — broadly consistent with ~480 years
The high priestly line from Aaron to Zadok (David and Solomon's priest) is given twice, independently: 1 Chronicles 6:35-38 and Ezra 7:1-5 (tracing Ezra's own ancestry back to Aaron), and the two lists agree exactly: Aaron → Eleazar → Phinehas → Abishua → Bukki → Uzzi → Zerahiah → Meraioth → Amariah → Ahitub → Zadok. That's 10 generational steps from Aaron (who dies in the wilderness, so effectively at the Exodus end-point) to Zadok (serving at the very end of David's reign and the start of Solomon's). Over roughly 400-480 years, that's 40-48 years per generation — on the high side for a strict father-to-son succession, but not implausible for a priestly office where a man might not become high priest, or father his own heir, particularly young. This genealogy doesn't prove 480 years, but it doesn't strain against it either.
Genealogy check #2: the Davidic line — in real tension with 480 years
Ruth 4:18-22 gives David's own ancestry from Judah's son Perez: Perez → Hezron → Ram → Amminadab → Nahshon → Salmon → Boaz → Obed → Jesse → David. Nahshon isn't a random name — Numbers 1:7 (ESV) names him as the tribal leader of Judah during the wilderness census, firmly placing him in the Exodus generation. From Nahshon to David is only 5 generational gaps (Nahshon-Salmon-Boaz-Obed-Jesse-David). Spread across the same 400-480 year span the priestly line tolerates, that's 80-96 years per generation — not plausible for ordinary human fathering, by a wide margin.
This doesn't mean Ruth's genealogy is wrong; it means it's very likely telescoped — skipping generations the way ancient genealogies regularly do (this site's own Matthew 1:17 discussion above is a directly comparable case: three known kings dropped to hit a structuring number). Telescoping in an official succession line, especially one this short, is a well-attested biblical pattern, not a special plea invented to save this one case.
Where this leaves the reconstruction
Two genealogical checks, pointing in different directions, is the honest state of the evidence — not a failure of the method. The priestly line is compatible with something close to 1 Kings 6:1's 480 years; the royal line, taken at face value, is not, and all but requires accepting that Ruth's list omits names. Put together with the judges-overlap evidence above, the most defensible reading is: 1 Kings 6:1's 480 years is plausible as a real total, achieved by real regional overlap among the judges (textually supported, not invented) rather than strict national sequence, while Ruth's five-generation genealogy for David almost certainly telescopes rather than recording every link — the same kind of compression this study already had to reckon with in Matthew's genealogy, just applied one book earlier. Nothing here fixes an exact year-by-year allocation of which judge overlaps which by how much; that level of precision isn't recoverable from what the text actually states, and claiming otherwise would overshoot the evidence in the same way summing the numbers naively does.
What this means for prophecy and Christ
None of the above is only an arithmetic exercise. Two genealogies of Jesus survive (Matthew 1:1-17, Luke 3:23-38), and they're doing visibly different jobs. Matthew's is explicitly structured — "fourteen generations" three times over (Matthew 1:17) — and to hit that count it compresses the king-list of Judah, skipping three known kings between Joram and Uzziah (compare Matthew 1:8 with 1 Chronicles 3:11-12). That's not sloppiness; ancient genealogies routinely telescoped names for a structuring purpose without being understood as lying about lineage. It also means Matthew's list, unlike Genesis 5 and 11, was never trying to support a year count at all — it's making a royal, covenantal argument (this is David's heir), not a chronological one. Luke's list runs the other direction, all the way back past Abraham to "the son of Adam, the son of God" (Luke 3:38) — and that ending is the argument. Luke is setting up the same connection Paul makes explicitly: Jesus as the second Adam, undoing in obedience what the first Adam did in disobedience (Romans 5:12-21; 1 Corinthians 15:22, 45). The genealogy exists, in Luke's hands, to make a theological claim stick to a real, traceable human line — not in spite of it being real, but because it is.
That's the frame this study's chronological work sits inside. The line from Adam to Christ isn't being tracked because a date is owed; it's being tracked because Genesis 3:15's promise — that the woman's seed would come, and would matter — runs through actual named people whose own names turn out, more often than not, to be saying something true about what's coming. Seth, appointed, in place of a murdered brother. Enoch, dedicated, taken without dying, a preview that death isn't the last word for those who walk with God (Hebrews 11:5). Noah, comfort, the one who carries the appointed line through judgment rather than being consumed by it. Whether or not Methuselah's own name predicted the Flood by its own arithmetic, the pattern around him does the same thing on a larger scale that the whole genealogy does: a real record of real people, shaped by a real author, tracking a promise that is still, twenty-some centuries after its last recorded chapter, being kept.
The specific date question — whether creation was 6540, 5960, or 5308 years before Christ on
this study's three witnesses, or something closer to the 5940-year figure harmonized_v1
implies — stays open, and honestly reported as open above. What doesn't stay open is the shape
of the claim: a single traceable line, named generation by generation, carrying a promise from
Eden to an empty tomb. The math was always in service of that; it was never the point on its
own.
Israel and the Church, and the anchors this study deliberately deferred
This study intentionally did not pull in prophecy-events-times.md's external archaeological
anchors (Qarqar, Sennacherib, the Babylonian and Persian records) while working out the
genealogical reasoning above, so that anchor-based dating wouldn't quietly bias which manuscript
readings looked more "probable." Now that the genealogical case stands on its own, linking the
two — checking where harmonized_v1's numbers land relative to Thiele's Qarqar-anchored
chronology for the divided monarchy, for instance — is the natural next step, tracked as open
work in this study's state file rather than done here.
References & Recommended Reading
- Primary texts (queried directly via
references/build/query.pyagainstreferences/build/bible-text.db):morphhb-wlc(Westminster Leningrad Codex, MT),ebible-grcbrent(Brenton Septuagint, LXX),scrollmapper-SP(Samaritan Pentateuch) - TWOT root/Strong's/gloss data (
twot_strongs_map.json), queried viareferences/build/twot_lookup.py, for every word study above - James C. VanderKam, Calendars in the Dead Sea Scrolls: Measuring Time — on the broader Second Temple textual environment these traditions come from
- The Zadok Calendar, The Day is Near, and Prophecy Events and Times — this site's other chronology studies, including the still-open creation-epoch discrepancy noted above
docs/data/genealogy/index.json,antediluvian.json,patriarchal.json— the structured source data this study explainsreferences/build/genealogy_chronology.py— the validator/generator computing the table above from that source datareferences/study-state/genealogy-times.yml— the full research trail, including items deferred rather than resolved in this draft