what the lines carry when the day axis is thought away
15 min read
In-depth:
1 · One measure, before the count
The previous version of this chapter set 10 signposts side by side, called them independent, and conceded at the close that after deduction of the shared assumptions "roughly two and a half" remained. Both numbers stood in one chapter. This chapter counts once, with one measure, and fixes the measure before the count begins: think Casus 31 and Dating the Birth of Jesus away. What does each line then still deliver of itself?
The result comes first, because it orders the rest. Without the day axis no line delivers the year 31. Daniel 9 gives the baptism in 27, under premises that chapter 3 ("The decree") writes out. Pilate, Caiaphas and the quake layer of Ein Gedi give a window from 26 to 36. The Talmud and the church fathers give an atmosphere around 30. The year itself comes from Casus 31: the weekday of 14 Nisan and the four Passovers of John. Casus 31 says it without detour: "The ministry window is the sieve; the day is the hand that reaches through the sieve" (Casus 31, ch. 9). This chapter describes the sieve.
A line that does not choose the year can still carry it: it can admit 31, lean towards it, or set the fence within which every candidate year must fall. That weight is named per line, in three classes: lines that stand apart from the day axis, lines that hang on it or on a counting knob they share with another line, and lines that only sketch the atmosphere. What in the previous version was called "load-bearing" and on closer inspection hangs on a knob is entered here as a dependency; what rested on one sabbatical chain that the oeuvre no longer uses lapses.
2 · The lines in one image
Workbenchthe count, per line
Line
Ground
Delivers of itself
Class
Hangs on
Daniel 9 from Ezra 7
Tishri 457 + 483 years
baptism 27
independent, under premises
year-weeks; Ezra 7 as the decree; the epoch on the sabbatical boundary (Zuckermann)
Ministry of three and a half years
the Passovers in John (2:13; 5:1; 6:4; 11:55)
27 → 31
support, not compulsion
the reading of John 5:1 as Passover
15th year of Tiberius
Luke 3:1
autumn 27 on the autumn-inclusive count; 28/29 on the documented one
available, not attested
John 2:20 (convergence choice, Casus 31, ch. 9)
Temple, 46 years
Josephus, Ant. 15.380
first Passover 27/28 or 28/29
dependent
the counting rule of Herod's regnal years
"About 30"
Luke 3:23; band 6–4 BC (gv, run A)
gradient baptism 27 > 28 > 29
consistent, conditional
the age band (±2 or ±3); Herod's season of death
Pilate / Caiaphas
inscription of Caesarea; Tacitus; Ant. 18.95-97
fence 26–36
independent
nothing
Quake layer Ein Gedi
Williams, Schwab & Brauer 2012
31 or 32 (±5)
independent, window
the count of the annual layers, with or without year 0
Antipas / Aretas
Ant. 18.116-119
fits 27–31
frame
nothing
40 years before 70
b. Yoma 39b
≈30
atmosphere
a round number
Church fathers
Tertullian, Clement, Africanus, Eusebius
29 to 33, according to the assumed length of ministry
atmosphere
the length of ministry each father assumes
Sabbatical year at Luke 4
Wacholder (27/28)
nothing on Zuckermann
lapsed
the choice of shemitah chain
Independent of the day axis and with a year as result: one line, and it gives the baptism, not the cross. Independent and with a window: two. The rest hangs, leans or colours. [CALCULATED · DOCUMENTED]
3 · Daniel 9: the baptism in 27, under premises
The only line that delivers a year apart from the day is Daniel's prophecy of the 70 weeks. From "the going forth of a word to restore and to build Jerusalem until an anointed one, a prince" run 7 and 62 weeks (Daniel 9:25); "after the 62 weeks an anointed one will be cut off" (9:26). Read as year-weeks, that is 483 years. Counted from Tishri 457 BC, in ordinary solar years and without a year 0, they end in AD 27: the anointing, the baptism. [CALCULATED]
That sum stands, but it stands on three premises, and chapter 3 ("The decree") weighs them. The first: the weeks are year-weeks. The second: the "word" of 9:25 is the rescript of Ezra 7, not the decree of Nehemiah 2. The strongest card of the Nehemiah reading is the wording, "restore and build"; it partly turns round, because Ezra 4:7-23 has the Jews building the city already before Nehemiah, but that that letter falls chronologically after Ezra 7 is a choice and is entered as such. Ezra 9:9 carries nothing here: gader is "fence", not chomah, "city wall". The third premise is the epoch. Ezra travels in Nisan 458; the count opens on the next boundary of the sabbatical cycle, Tishri 457, and that boundary exists on the chain of Zuckermann, not on that of Wacholder. Decisive there is how the accession of Artaxerxes was reckoned: with Xerxes' death in August 465 the autumn count gives 458 for Ezra's seventh year, and 457 requires a late-reckoned accession.
What the line gains is the measure. It comes out with ordinary solar years, whereas the count from 444 (Anderson, Hoehner) needs years of 360 days and then still falls four days short of the entry of 33: 173,884 days against 173,880 (Casus 31, ch. 17). [CALCULATED] Why should Daniel's year count 360 days and every other biblical year 365?
What the line does not give is the year 31. This book reads the 70th week as future; the "cutting off" of 9:26 falls "after" the 69 weeks, at a distance the text does not name. Whoever fills in three and a half years takes that number from the four Passovers of John, not from Daniel. With that the baptism of 27 is the end of what Daniel carries. Dating the Birth of Jesus says the same from the other side: the 69 weeks "close as a sum and carry nothing as an independent clamp: they are by construction the target of the logigram" (Dating the Birth of Jesus, ch. 9). That is no reproach to the sum, but a limit on what it proves.
4 · The temple: a line that hangs on a knob
At the first Passover of the ministry the opponents say: "46 years this temple has been in building, and You will raise it up in three days?" (John 2:20). Josephus has Herod begin the rebuilding in his 18th regnal year (Antiquities 15.380). The previous version of this chapter called that line "load-bearing" and "wholly apart from Daniel". Apart from Daniel it is; apart from the Herod axis it is not.
Which calendar year the 18th regnal year is hangs on the same counting rule as the year of Herod's death. Whoever counts inclusively from the capture of Jerusalem in 37 BC arrives at 20/19 BC, and 46 years later at 27/28: the first Passover in 28, the baptism in the autumn of 27. Whoever counts non-inclusively from 37, or from 36, arrives at 19/18 BC and lands on 28/29. [CALCULATED] The same knob that in Dating the Birth of Jesus gives the consensus its 4 BC gives here the 27/28; whoever turns it shifts both. Casus 31 adds Augustus' visit to Syria as a Roman anchor for the 18th year (Casus 31, ch. 9; Cassius Dio 54.7). That bounds the knob, but the 18th year remains a regnal year of Herod, and which calendar year that is, the counting rule decides.
The reading of the verse itself is reversible too. The majority reading takes "46 years" as the duration of a building that was still going on, and that fits Josephus' broad use of naos (Ant. 11.87; 15.421). An opponent reads the dative as a point in time and the aorist as completed: the sanctuary had stood 46 years, since about 18/17 BC, and then the first Passover falls around 30 and the cross in 33. Casus 31 therefore calls John 2:20 "a reversible anchor" (Casus 31, ch. 9), which under the majority reading fits 31 and is bent towards 33 by a practised opponent. So the line stands entered here: it leans towards 27/28, it hangs on the counting rule of Herod, and it does not carry alone. [DOCUMENTED · INTERPRETED]
5 · The 15th year of Tiberius and the four Passovers
Luke dates the appearance of John the Baptist in the 15th year of the reign of Tiberius (3:1). Counted from the death of Augustus in August 14, on the documented count, that is 28/29. The baptism in the autumn of 27 requires an autumn-inclusive count, in which the months up to the next new year already count as year 1. Casus 31 has tested that count against the coin corpora and against Jewish halakhah and enters it as "unattested and not refuted" (Casus 31, ch. 9): the only doubly dated Tiberius piece, the Antiochene coin RPC I 4270, puts the 15th year at 28/29, and no document can say whether the accession months counted in Judea as year 1. The co-regency from 12 does not take part; no ancient reckoner uses it. [DOCUMENTED]
The autumn count is therefore not chosen because Luke compels it, but because John 2:20 points in the same direction. Casus 31 calls that form by its name: "It is a convergence choice, not a measurement: the count is taken because a second anchor agrees with it, and had John 2:20 pointed to 29, the same reasoning would have chosen the non-inclusive count" (Casus 31, ch. 9). Two lines that support each other like that do not together count as two.
The step from 27 to 31 sets the length of the ministry. John names three Passovers by name (2:13; 6:4; 11:55); the nameless "feast of the Jews" in 5:1 is the fourth, if it is a Passover. Irenaeus counts it so (Adversus Haereses 2.22.3), though he drew from his own count a conclusion nobody follows, a ministry of more than 10 years. Four Passovers from the baptism in 27 give the cross in 31; three give 30. Casus 31 has tested the reading on criteria fixed beforehand, failed two of the three, and carries it as "support and not compulsion" (Casus 31, ch. 9 and appendix E). The same three and a half years also carry the 33 model, with a baptism in 29 on the documented count. The length of ministry therefore does not distinguish the years; it only connects the baptism year with the year of the cross. [INTERPRETED]
6 · The age: a band, not a point
Luke says that Jesus at the beginning of His ministry was "about 30 years" old (3:23). The previous version of this chapter reckoned with a birth in the autumn of 5 BC and arrived at 31 at the baptism and 34 at the cross; the 33 model got, under that same birth, a baptism at 33 years, "beyond about thirty". That comparison was not fair. The point autumn 5 had been chosen in Casus 31 because it joins most smoothly to a baptism in 27, and it was then deployed against a baptism in 29. Each model ought to get its own most favourable cell within the same, independently established window.
That window is supplied by Dating the Birth of Jesus. Herod dies in the spring of 4 BC; the winter of 4/3 remains possible, but is less likely (Dating the Birth of Jesus, ch. 6). Without the baptism axis, in run A, the window stands free at 8 to 4 BC, with 6, 5 and 4 as the cells that pay no price; the band this book carries is 6–4 BC. The age at the baptism is baptism year plus birth year minus one, and "about" gets a width: Plummer ±2 (28 to 32), Hoehner ±3 (27 to 33). Dating the Birth of Jesus keeps to ±2, because ὡσεί with numbers marks a round number with small play and not a wide tolerance (Dating the Birth of Jesus, ch. 9).
Workbenchper baptism year, which birth cell reaches the band
Age in completed years; Plummer's band, 28 to 32; window 6–4 BC (Dating the Birth of Jesus, run A). Recomputed on the days of Dating the Birth of Jesus, ch. 6 and ch. 8, without year 0.
Baptism
Cells within 28–32
Under spring 4
Under winter 4/3
27
6 (32), 5 (31), 4 (30)
6, 5 (4 BC falls after the death)
6, 5, 4
28
5 (32), 4 (31)
5
5, 4
29
4 (32)
none
4, narrow
On Hoehner's ±3 everything shifts one cell: a baptism in 29 then also gets 5 BC (33), a baptism in 27 also 7 BC. Under spring 4 and Plummer's band the 33 model, with its baptism in 29, has no birth cell within 6–4 BC; under winter 4/3 it has one, 4 BC, and that one is narrow: the child is then born 1.6 to 7.5 months before a death in Kislev, with the magi, the flight and the massacre of the children in between. [CALCULATED]
What remains is a gradient: 27 above 28 above 29. It hangs on two named conditions, the band and the season of death, and on ±3 it grows weaker. A clamp on one point it no longer is, and the cross age "34" lapses with it: across the window 6–4 the age on 25 April 31 runs from 33 to 35. The line remains what it was, consistent with 31 and leaning against 29, but it does not become "load-bearing", and it does not choose the year. The full debate on Herod's year of death, the eclipse, the coins of the sons and the counting rule stands in Dating the Birth of Jesus (Dating the Birth of Jesus, ch. 3, ch. 6 and ch. 9); this book does not conduct it again. [CALCULATED · DOCUMENTED]
7 · The fence: Pilate, Caiaphas and the floor of the Dead Sea
Three data choose no year, but set the fence. Pilate was prefect from 26 to 36: the inscription of Caesarea gives his office and his title, Tacitus his role ("Christus, put to death under Tiberius by the procurator Pontius Pilatus", Annals 15.44). Caiaphas was high priest from 18 to 36 (Josephus, Ant. 18.95-97). The intersection is 26 to 36, and 31 lies within it; 30 and 33 too. [DOCUMENTED]
The third line comes out of the ground. In the section at Ein Gedi lies a quake layer from the early first century, and Williams, Schwab and Brauer dated it in 2012 by counting the annual layers back from the earthquake of 31 BC, without any assumption about the date sought. Their result: AD 31, with an accuracy of five years. Their addition, 31 before and 31 after, however presupposes a year 0; without year 0, 62 counted years come out at AD 32. The layer thus gives 31 or 32 (±5), according to the count, and the window 26–36 is largely the term of office of Pilate, not exactly. An anchorless radiocarbon dating (Kagan et al. 2011) gives a wider window, AD 12 to 91 at 1σ, which contains it; Kagan afterwards couples the layer to the catalogue earthquake of 33, and whoever says "33" cites that coupling and not the measurement. The base rate belongs with it: at roughly one correlated earthquake per half-century, an arbitrary window of 11 years contains one roughly one time in four to five (Casus 31, ch. 12). The floor confirms at window level and distinguishes nothing within the window. [DOCUMENTED · CALCULATED]
The political frame joins on to it. Antipas' marriage to Herodias, against which John the Baptist spoke (Mark 6:17-18), led to the war with Aretas IV in 36, which Josephus reads as retribution for the death of John (Ant. 18.116-119). A ministry from 27 to 31, with John killed around 29, fits within it; every other year within the fence fits too. Consistency, not a year.
8 · The atmosphere: 40 years and the church fathers
Two lines sketch the surroundings. The Talmud remembers that 40 years before the destruction the temple signs ceased to be favourable: the lot no longer came up in the right hand, the crimson thread no longer turned white, the doors opened of themselves (b. Yoma 39b). Counting back from 70 gives about 30, and 40 is a round number. It colours the window; it fixes nothing.
The church fathers seemed in the previous version to form a cluster at 29 to 31, against 33. Casus 31 has dissected that cluster (Casus 31, ch. 6): each date follows from the length of ministry the writer assumes. Whoever assumes a ministry of zero years arrives at 29 (Tertullian, Lactantius, Augustine); whoever assumes one year, at 30 or 31 (Clement, Africanus); whoever assumes three, at 32 (Hippolytus, Epiphanius); whoever assumes three and a half to four, at 33 (Eusebius). Clement shows it while he reckons: "fifteen years of Tiberius and fifteen years of Augustus, so the thirty years were completed until He suffered" (Stromata 1.21). That is no testimony, but a sum. Whoever rejects the 33 of Eusebius as derived may not harvest the 29 of Tertullian as support. On one measure what remains of the cluster is an atmosphere, which admits 31 as it admits 33. [DOCUMENTED]
9 · What no longer counts: the sabbatical year, the Jubilee and a kindred model
The previous version let the first year of ministry fall in a sabbatical year, 27/28 on the chain of Wacholder, and read in that an entourage for the proclamation of release in Nazareth. The oeuvre now carries one shemitah chain, that of Zuckermann, because the epoch of chapter 3 rests on it; whoever chooses per event the chain that comes out measures by two standards. On Zuckermann 26/27 is the sabbatical year and Tishri 27 opens a new cycle. That fits, and it carries nothing.
The Jubilee moves too. The previous version let the Jubilee year of this framework fall "a sabbatical cycle later", around Anno Mundi 4000, and read Daniel's 490 years as 10 Jubilee cycles that made the 457 line and the sabbatical year one count. On Wacholder 34/35 is itself a sabbatical year and no Jubilee, for a Jubilee follows the seventh sabbatical year and never coincides with it (Leviticus 25:8-11); and 49 and 50 may never go in one sum. How the Jubilee does and does not count, and why it counts nothing towards 2034, stands in chapter 5 ("The hundred and twentieth Jubilee"). What remains is the language of Isaiah 61, and that is theology.
Expertthe "acceptable year": proclamation, not calendar year
In the synagogue of Nazareth Jesus opens with Isaiah 61: "to proclaim an acceptable year of the Lord" (Luke 4:19). Isaiah 61:1-2 uses the vocabulary of Leviticus 25: deror, "release", the word of the Jubilee year (Lev. 25:10). That the "acceptable year" was read as the eschatological Jubilee is not modern: 11QMelchizedek (11Q13, first century BC) already reads it so and connects it with Daniel's 70 weeks as 10 Jubilee cycles. When Jesus says "today this Scripture is fulfilled in your ears" (Luke 4:21), He announces the coming of that great Jubilee.
So as not to measure by two standards: this is a reading of the proclamation, not of the calendar. The model does not claim that 27 was a Jubilee year, nor that the first year of ministry had to be a sabbatical year. An announcement that would be true in any year distinguishes no year. That is why the proclamation carries no chronological weight: the Jubilee language is theology, not dating. So this work avoids the double standard. [INTERPRETED]
The nearest kindred model, Messiah2030 (cross in 30, second coming in 2030), shares with this work the framework of two days as 2000 years (Hosea 6:2; Psalm 90:4); what of it is usable with a shift of one year takes the year 31 as given and therefore belongs in the eschatological work, not here.
10 · The sum
So the count stands. One line stands apart from the day axis and delivers a year: Daniel 9, and that year is 27, the baptism, under three premises. Two lines stand apart from the day axis and deliver a window: the offices of Pilate and Caiaphas, and the floor of the Dead Sea. Two lines sketch an atmosphere around 30. The others hang: the temple on the counting rule of Herod, Tiberius on the temple, the age on the band and the season of death, the length of ministry on the reading of one verse. Casus 31 arrives, after deduction of that overlap, at "roughly two and a half" effective axes: the day axis as one, the year axis as one, and the chronological envelope of Tiberius, the temple and Daniel as at most a half (Casus 31, ch. 16). This chapter does not change that number; it writes out where the half comes from.
The year 31 therefore does not come from these lines. It comes from Casus 31: of the years the fence admits, one has a Wednesday 14 Nisan within the ministry window, and the four Passovers connect that Wednesday with the baptism of 27. Chapter 1 ("The anchor") has named that result with its price and its two breaking points. What the convergence then does is more modest than the previous version claimed and more precise than "nothing": it admits 31 at a price the rivals also pay in their own coin, and whose price is the better founded stands "honestly counted, close to even" (Casus 31, ch. 9); it leans on the temple and on the age lightly towards the early side; and it joins the baptism of 27 to a decree five centuries earlier. That last is the proper work of this book. Not whether 31 is the year, but what rests on it: the chain from the decree via the anointing and the anchor to the epoch and the terminus, labelled per link. The thesis that belongs with it already stands written out in chapter 5: "the two golden pins — 457 and the year 31 — do not hang on the six thousand".
HeartHEART
There is something liberating about an honest count. 10 witnesses who do not know each other and yet say the same compel reflection; three witnesses who do know each other and seven who only nod compel nothing. Whoever sees the difference loses an argument and gains a foundation, for what remains stands on its own ground. A prophet who dates an anointing five centuries ahead. A prefect and a high priest who set a fence. A layer in the ground that knows nothing of the question put to it.
And that is why the posture remains one of the open hand. A firm hand on what has been measured, a loose hand on what has been interpreted, and the distinction out loud. Whoever sees the lines bend towards the same window may calmly say that it convinces, without saying that it is settled. That distinction preserves the honesty, and it preserves the reverence.
The sieve has been described. Now the first link of the chain: the decree with which the 483 years begin, and the year in which it went out. That is chapter 3, "The decree".
Colophon to chapter 2
Verification status. Checked against the text or against authoritative editions: Daniel 9:25-26; John 2:13,20; 5:1; 6:4; 11:55; Luke 3:1,23; 4:18-21; Josephus Ant. 15.380, 18.95-97, 18.116-119; Tacitus Ann. 15.44; RPC I 4270 (via Casus 31, ch. 9); Williams, Schwab & Brauer 2012 ("In the 62 years counted from 31 BC to 31 AD"; "26–36 AD (31 ± 5 years)") and Kagan et al. 2011 ("no anchors are used"; AD 12–91 at 1σ), both read at the audit of this book; b. Yoma 39b; Clement Strom. 1.21 and the patristic series via Casus 31, ch. 6; 11QMelchizedek [DOCUMENTED]. Counts without year 0: Tishri 457 BC + 483 = 27; 31 BC + 62 = AD 32; 173,884 − 173,880 = 4 days; temple 20/19 + 46 = 27/28 and 19/18 + 46 = 28/29; ages per baptism year on run A of Dating the Birth of Jesus [CALCULATED]. Herod: spring 4 BC; the winter of 4/3 remains possible, but is less likely (Dating the Birth of Jesus, ch. 6).
Style. Main text in the voice of the oeuvre; the apparatus in the Workbench, Expert and Heart boxes. No claim of proof; load-bearing word count. Each line: ground, what it delivers of itself, what it hangs on, and its class.
1 · One measure, before the count
The previous version of this chapter set 10 signposts side by side, called them independent, and conceded at the close that after deduction of the shared assumptions "roughly two and a half" remained. Both numbers stood in one chapter. This chapter counts once, with one measure, and fixes the measure before the count begins: think Casus 31 and Dating the Birth of Jesus away. What does each line then still deliver of itself?
The result comes first, because it orders the rest. Without the day axis no line delivers the year 31. Daniel 9 gives the baptism in 27, under premises that chapter 3 ("The decree") writes out. Pilate, Caiaphas and the quake layer of Ein Gedi give a window from 26 to 36. The Talmud and the church fathers give an atmosphere around 30. The year itself comes from Casus 31: the weekday of 14 Nisan and the four Passovers of John. Casus 31 says it without detour: "The ministry window is the sieve; the day is the hand that reaches through the sieve" (Casus 31, ch. 9). This chapter describes the sieve.
A line that does not choose the year can still carry it: it can admit 31, lean towards it, or set the fence within which every candidate year must fall. That weight is named per line, in three classes: lines that stand apart from the day axis, lines that hang on it or on a counting knob they share with another line, and lines that only sketch the atmosphere. What in the previous version was called "load-bearing" and on closer inspection hangs on a knob is entered here as a dependency; what rested on one sabbatical chain that the oeuvre no longer uses lapses.
2 · The lines in one image
Workbenchthe count, per line
Independent of the day axis and with a year as result: one line, and it gives the baptism, not the cross. Independent and with a window: two. The rest hangs, leans or colours. [CALCULATED · DOCUMENTED]
3 · Daniel 9: the baptism in 27, under premises
The only line that delivers a year apart from the day is Daniel's prophecy of the 70 weeks. From "the going forth of a word to restore and to build Jerusalem until an anointed one, a prince" run 7 and 62 weeks (Daniel 9:25); "after the 62 weeks an anointed one will be cut off" (9:26). Read as year-weeks, that is 483 years. Counted from Tishri 457 BC, in ordinary solar years and without a year 0, they end in AD 27: the anointing, the baptism. [CALCULATED]
That sum stands, but it stands on three premises, and chapter 3 ("The decree") weighs them. The first: the weeks are year-weeks. The second: the "word" of 9:25 is the rescript of Ezra 7, not the decree of Nehemiah 2. The strongest card of the Nehemiah reading is the wording, "restore and build"; it partly turns round, because Ezra 4:7-23 has the Jews building the city already before Nehemiah, but that that letter falls chronologically after Ezra 7 is a choice and is entered as such. Ezra 9:9 carries nothing here: gader is "fence", not chomah, "city wall". The third premise is the epoch. Ezra travels in Nisan 458; the count opens on the next boundary of the sabbatical cycle, Tishri 457, and that boundary exists on the chain of Zuckermann, not on that of Wacholder. Decisive there is how the accession of Artaxerxes was reckoned: with Xerxes' death in August 465 the autumn count gives 458 for Ezra's seventh year, and 457 requires a late-reckoned accession.
What the line gains is the measure. It comes out with ordinary solar years, whereas the count from 444 (Anderson, Hoehner) needs years of 360 days and then still falls four days short of the entry of 33: 173,884 days against 173,880 (Casus 31, ch. 17). [CALCULATED] Why should Daniel's year count 360 days and every other biblical year 365?
What the line does not give is the year 31. This book reads the 70th week as future; the "cutting off" of 9:26 falls "after" the 69 weeks, at a distance the text does not name. Whoever fills in three and a half years takes that number from the four Passovers of John, not from Daniel. With that the baptism of 27 is the end of what Daniel carries. Dating the Birth of Jesus says the same from the other side: the 69 weeks "close as a sum and carry nothing as an independent clamp: they are by construction the target of the logigram" (Dating the Birth of Jesus, ch. 9). That is no reproach to the sum, but a limit on what it proves.
4 · The temple: a line that hangs on a knob
At the first Passover of the ministry the opponents say: "46 years this temple has been in building, and You will raise it up in three days?" (John 2:20). Josephus has Herod begin the rebuilding in his 18th regnal year (Antiquities 15.380). The previous version of this chapter called that line "load-bearing" and "wholly apart from Daniel". Apart from Daniel it is; apart from the Herod axis it is not.
Which calendar year the 18th regnal year is hangs on the same counting rule as the year of Herod's death. Whoever counts inclusively from the capture of Jerusalem in 37 BC arrives at 20/19 BC, and 46 years later at 27/28: the first Passover in 28, the baptism in the autumn of 27. Whoever counts non-inclusively from 37, or from 36, arrives at 19/18 BC and lands on 28/29. [CALCULATED] The same knob that in Dating the Birth of Jesus gives the consensus its 4 BC gives here the 27/28; whoever turns it shifts both. Casus 31 adds Augustus' visit to Syria as a Roman anchor for the 18th year (Casus 31, ch. 9; Cassius Dio 54.7). That bounds the knob, but the 18th year remains a regnal year of Herod, and which calendar year that is, the counting rule decides.
The reading of the verse itself is reversible too. The majority reading takes "46 years" as the duration of a building that was still going on, and that fits Josephus' broad use of naos (Ant. 11.87; 15.421). An opponent reads the dative as a point in time and the aorist as completed: the sanctuary had stood 46 years, since about 18/17 BC, and then the first Passover falls around 30 and the cross in 33. Casus 31 therefore calls John 2:20 "a reversible anchor" (Casus 31, ch. 9), which under the majority reading fits 31 and is bent towards 33 by a practised opponent. So the line stands entered here: it leans towards 27/28, it hangs on the counting rule of Herod, and it does not carry alone. [DOCUMENTED · INTERPRETED]
5 · The 15th year of Tiberius and the four Passovers
Luke dates the appearance of John the Baptist in the 15th year of the reign of Tiberius (3:1). Counted from the death of Augustus in August 14, on the documented count, that is 28/29. The baptism in the autumn of 27 requires an autumn-inclusive count, in which the months up to the next new year already count as year 1. Casus 31 has tested that count against the coin corpora and against Jewish halakhah and enters it as "unattested and not refuted" (Casus 31, ch. 9): the only doubly dated Tiberius piece, the Antiochene coin RPC I 4270, puts the 15th year at 28/29, and no document can say whether the accession months counted in Judea as year 1. The co-regency from 12 does not take part; no ancient reckoner uses it. [DOCUMENTED]
The autumn count is therefore not chosen because Luke compels it, but because John 2:20 points in the same direction. Casus 31 calls that form by its name: "It is a convergence choice, not a measurement: the count is taken because a second anchor agrees with it, and had John 2:20 pointed to 29, the same reasoning would have chosen the non-inclusive count" (Casus 31, ch. 9). Two lines that support each other like that do not together count as two.
The step from 27 to 31 sets the length of the ministry. John names three Passovers by name (2:13; 6:4; 11:55); the nameless "feast of the Jews" in 5:1 is the fourth, if it is a Passover. Irenaeus counts it so (Adversus Haereses 2.22.3), though he drew from his own count a conclusion nobody follows, a ministry of more than 10 years. Four Passovers from the baptism in 27 give the cross in 31; three give 30. Casus 31 has tested the reading on criteria fixed beforehand, failed two of the three, and carries it as "support and not compulsion" (Casus 31, ch. 9 and appendix E). The same three and a half years also carry the 33 model, with a baptism in 29 on the documented count. The length of ministry therefore does not distinguish the years; it only connects the baptism year with the year of the cross. [INTERPRETED]
6 · The age: a band, not a point
Luke says that Jesus at the beginning of His ministry was "about 30 years" old (3:23). The previous version of this chapter reckoned with a birth in the autumn of 5 BC and arrived at 31 at the baptism and 34 at the cross; the 33 model got, under that same birth, a baptism at 33 years, "beyond about thirty". That comparison was not fair. The point autumn 5 had been chosen in Casus 31 because it joins most smoothly to a baptism in 27, and it was then deployed against a baptism in 29. Each model ought to get its own most favourable cell within the same, independently established window.
That window is supplied by Dating the Birth of Jesus. Herod dies in the spring of 4 BC; the winter of 4/3 remains possible, but is less likely (Dating the Birth of Jesus, ch. 6). Without the baptism axis, in run A, the window stands free at 8 to 4 BC, with 6, 5 and 4 as the cells that pay no price; the band this book carries is 6–4 BC. The age at the baptism is baptism year plus birth year minus one, and "about" gets a width: Plummer ±2 (28 to 32), Hoehner ±3 (27 to 33). Dating the Birth of Jesus keeps to ±2, because ὡσεί with numbers marks a round number with small play and not a wide tolerance (Dating the Birth of Jesus, ch. 9).
Workbenchper baptism year, which birth cell reaches the band
Age in completed years; Plummer's band, 28 to 32; window 6–4 BC (Dating the Birth of Jesus, run A). Recomputed on the days of Dating the Birth of Jesus, ch. 6 and ch. 8, without year 0.
On Hoehner's ±3 everything shifts one cell: a baptism in 29 then also gets 5 BC (33), a baptism in 27 also 7 BC. Under spring 4 and Plummer's band the 33 model, with its baptism in 29, has no birth cell within 6–4 BC; under winter 4/3 it has one, 4 BC, and that one is narrow: the child is then born 1.6 to 7.5 months before a death in Kislev, with the magi, the flight and the massacre of the children in between. [CALCULATED]
What remains is a gradient: 27 above 28 above 29. It hangs on two named conditions, the band and the season of death, and on ±3 it grows weaker. A clamp on one point it no longer is, and the cross age "34" lapses with it: across the window 6–4 the age on 25 April 31 runs from 33 to 35. The line remains what it was, consistent with 31 and leaning against 29, but it does not become "load-bearing", and it does not choose the year. The full debate on Herod's year of death, the eclipse, the coins of the sons and the counting rule stands in Dating the Birth of Jesus (Dating the Birth of Jesus, ch. 3, ch. 6 and ch. 9); this book does not conduct it again. [CALCULATED · DOCUMENTED]
7 · The fence: Pilate, Caiaphas and the floor of the Dead Sea
Three data choose no year, but set the fence. Pilate was prefect from 26 to 36: the inscription of Caesarea gives his office and his title, Tacitus his role ("Christus, put to death under Tiberius by the procurator Pontius Pilatus", Annals 15.44). Caiaphas was high priest from 18 to 36 (Josephus, Ant. 18.95-97). The intersection is 26 to 36, and 31 lies within it; 30 and 33 too. [DOCUMENTED]
The third line comes out of the ground. In the section at Ein Gedi lies a quake layer from the early first century, and Williams, Schwab and Brauer dated it in 2012 by counting the annual layers back from the earthquake of 31 BC, without any assumption about the date sought. Their result: AD 31, with an accuracy of five years. Their addition, 31 before and 31 after, however presupposes a year 0; without year 0, 62 counted years come out at AD 32. The layer thus gives 31 or 32 (±5), according to the count, and the window 26–36 is largely the term of office of Pilate, not exactly. An anchorless radiocarbon dating (Kagan et al. 2011) gives a wider window, AD 12 to 91 at 1σ, which contains it; Kagan afterwards couples the layer to the catalogue earthquake of 33, and whoever says "33" cites that coupling and not the measurement. The base rate belongs with it: at roughly one correlated earthquake per half-century, an arbitrary window of 11 years contains one roughly one time in four to five (Casus 31, ch. 12). The floor confirms at window level and distinguishes nothing within the window. [DOCUMENTED · CALCULATED]
The political frame joins on to it. Antipas' marriage to Herodias, against which John the Baptist spoke (Mark 6:17-18), led to the war with Aretas IV in 36, which Josephus reads as retribution for the death of John (Ant. 18.116-119). A ministry from 27 to 31, with John killed around 29, fits within it; every other year within the fence fits too. Consistency, not a year.
8 · The atmosphere: 40 years and the church fathers
Two lines sketch the surroundings. The Talmud remembers that 40 years before the destruction the temple signs ceased to be favourable: the lot no longer came up in the right hand, the crimson thread no longer turned white, the doors opened of themselves (b. Yoma 39b). Counting back from 70 gives about 30, and 40 is a round number. It colours the window; it fixes nothing.
The church fathers seemed in the previous version to form a cluster at 29 to 31, against 33. Casus 31 has dissected that cluster (Casus 31, ch. 6): each date follows from the length of ministry the writer assumes. Whoever assumes a ministry of zero years arrives at 29 (Tertullian, Lactantius, Augustine); whoever assumes one year, at 30 or 31 (Clement, Africanus); whoever assumes three, at 32 (Hippolytus, Epiphanius); whoever assumes three and a half to four, at 33 (Eusebius). Clement shows it while he reckons: "fifteen years of Tiberius and fifteen years of Augustus, so the thirty years were completed until He suffered" (Stromata 1.21). That is no testimony, but a sum. Whoever rejects the 33 of Eusebius as derived may not harvest the 29 of Tertullian as support. On one measure what remains of the cluster is an atmosphere, which admits 31 as it admits 33. [DOCUMENTED]
9 · What no longer counts: the sabbatical year, the Jubilee and a kindred model
The previous version let the first year of ministry fall in a sabbatical year, 27/28 on the chain of Wacholder, and read in that an entourage for the proclamation of release in Nazareth. The oeuvre now carries one shemitah chain, that of Zuckermann, because the epoch of chapter 3 rests on it; whoever chooses per event the chain that comes out measures by two standards. On Zuckermann 26/27 is the sabbatical year and Tishri 27 opens a new cycle. That fits, and it carries nothing.
The Jubilee moves too. The previous version let the Jubilee year of this framework fall "a sabbatical cycle later", around Anno Mundi 4000, and read Daniel's 490 years as 10 Jubilee cycles that made the 457 line and the sabbatical year one count. On Wacholder 34/35 is itself a sabbatical year and no Jubilee, for a Jubilee follows the seventh sabbatical year and never coincides with it (Leviticus 25:8-11); and 49 and 50 may never go in one sum. How the Jubilee does and does not count, and why it counts nothing towards 2034, stands in chapter 5 ("The hundred and twentieth Jubilee"). What remains is the language of Isaiah 61, and that is theology.
Expertthe "acceptable year": proclamation, not calendar year
In the synagogue of Nazareth Jesus opens with Isaiah 61: "to proclaim an acceptable year of the Lord" (Luke 4:19). Isaiah 61:1-2 uses the vocabulary of Leviticus 25: deror, "release", the word of the Jubilee year (Lev. 25:10). That the "acceptable year" was read as the eschatological Jubilee is not modern: 11QMelchizedek (11Q13, first century BC) already reads it so and connects it with Daniel's 70 weeks as 10 Jubilee cycles. When Jesus says "today this Scripture is fulfilled in your ears" (Luke 4:21), He announces the coming of that great Jubilee.
So as not to measure by two standards: this is a reading of the proclamation, not of the calendar. The model does not claim that 27 was a Jubilee year, nor that the first year of ministry had to be a sabbatical year. An announcement that would be true in any year distinguishes no year. That is why the proclamation carries no chronological weight: the Jubilee language is theology, not dating. So this work avoids the double standard. [INTERPRETED]
The nearest kindred model, Messiah2030 (cross in 30, second coming in 2030), shares with this work the framework of two days as 2000 years (Hosea 6:2; Psalm 90:4); what of it is usable with a shift of one year takes the year 31 as given and therefore belongs in the eschatological work, not here.
10 · The sum
So the count stands. One line stands apart from the day axis and delivers a year: Daniel 9, and that year is 27, the baptism, under three premises. Two lines stand apart from the day axis and deliver a window: the offices of Pilate and Caiaphas, and the floor of the Dead Sea. Two lines sketch an atmosphere around 30. The others hang: the temple on the counting rule of Herod, Tiberius on the temple, the age on the band and the season of death, the length of ministry on the reading of one verse. Casus 31 arrives, after deduction of that overlap, at "roughly two and a half" effective axes: the day axis as one, the year axis as one, and the chronological envelope of Tiberius, the temple and Daniel as at most a half (Casus 31, ch. 16). This chapter does not change that number; it writes out where the half comes from.
The year 31 therefore does not come from these lines. It comes from Casus 31: of the years the fence admits, one has a Wednesday 14 Nisan within the ministry window, and the four Passovers connect that Wednesday with the baptism of 27. Chapter 1 ("The anchor") has named that result with its price and its two breaking points. What the convergence then does is more modest than the previous version claimed and more precise than "nothing": it admits 31 at a price the rivals also pay in their own coin, and whose price is the better founded stands "honestly counted, close to even" (Casus 31, ch. 9); it leans on the temple and on the age lightly towards the early side; and it joins the baptism of 27 to a decree five centuries earlier. That last is the proper work of this book. Not whether 31 is the year, but what rests on it: the chain from the decree via the anointing and the anchor to the epoch and the terminus, labelled per link. The thesis that belongs with it already stands written out in chapter 5: "the two golden pins — 457 and the year 31 — do not hang on the six thousand".
HeartHEART
There is something liberating about an honest count. 10 witnesses who do not know each other and yet say the same compel reflection; three witnesses who do know each other and seven who only nod compel nothing. Whoever sees the difference loses an argument and gains a foundation, for what remains stands on its own ground. A prophet who dates an anointing five centuries ahead. A prefect and a high priest who set a fence. A layer in the ground that knows nothing of the question put to it.
And that is why the posture remains one of the open hand. A firm hand on what has been measured, a loose hand on what has been interpreted, and the distinction out loud. Whoever sees the lines bend towards the same window may calmly say that it convinces, without saying that it is settled. That distinction preserves the honesty, and it preserves the reverence.
The sieve has been described. Now the first link of the chain: the decree with which the 483 years begin, and the year in which it went out. That is chapter 3, "The decree".
Colophon to chapter 2
Verification status. Checked against the text or against authoritative editions: Daniel 9:25-26; John 2:13,20; 5:1; 6:4; 11:55; Luke 3:1,23; 4:18-21; Josephus Ant. 15.380, 18.95-97, 18.116-119; Tacitus Ann. 15.44; RPC I 4270 (via Casus 31, ch. 9); Williams, Schwab & Brauer 2012 ("In the 62 years counted from 31 BC to 31 AD"; "26–36 AD (31 ± 5 years)") and Kagan et al. 2011 ("no anchors are used"; AD 12–91 at 1σ), both read at the audit of this book; b. Yoma 39b; Clement Strom. 1.21 and the patristic series via Casus 31, ch. 6; 11QMelchizedek [DOCUMENTED]. Counts without year 0: Tishri 457 BC + 483 = 27; 31 BC + 62 = AD 32; 173,884 − 173,880 = 4 days; temple 20/19 + 46 = 27/28 and 19/18 + 46 = 28/29; ages per baptism year on run A of Dating the Birth of Jesus [CALCULATED]. Herod: spring 4 BC; the winter of 4/3 remains possible, but is less likely (Dating the Birth of Jesus, ch. 6).
Style. Main text in the voice of the oeuvre; the apparatus in the Workbench, Expert and Heart boxes. No claim of proof; load-bearing word count. Each line: ground, what it delivers of itself, what it hangs on, and its class.