The previous chapter made a reproach: the field excluded AD 31 without measuring it, on an axiom. Such a reproach lays an obligation on the one who makes it. If I charge others with judging without a measure, then I must lay my own on the table — before anything is weighed, so that you can inspect it and, if need be, use it against me. This chapter is that measure. It contains not a single argument for the Wednesday. It only fixes how weighing is done in this book, and those rules bind me as hard as they bind the consensus.
The first rule is the most important, and it is a prohibition. This book nowhere claims a proof — nowhere does it say that something is established. Not out of false modesty, but because it does not belong here. Historical dating yields no proofs of the kind mathematics knows; it yields, at its best, convergence — the concurrence of independent indications pointing the same way. Convergence is strong; it is not certain. And the difference is no play on words. Whoever claims a proof closes the conversation; whoever says "convergence" keeps it open, and invites the reader to weigh for himself. This book therefore ends not with a proof but with a weighed surmise, and it will show you at every turn how much that surmise weighs — no more, and no less.
To make that weighing checkable, every claim in this book carries a mark. There are three, and they say precisely what a claim is and is not. A calculated fact is reproducible: you can recompute it yourself, and the route to it stands in a Workbench or an appendix. A documented fact has been seen in the source itself: there is a primary text, with its location, that carries it. A construed fact is an interpretation: it weighs as much as you allow the construal, and not a grain more. These three must never be allowed to run into one another. A calculation is no proof of a construal; a construal, however beautiful, may not take the place of a measurement. Where this book paints a picture — the partly darkened moon over Jerusalem, the Son in the deep like Jonah — that is construed, and it is marked as such, lest any reader mistake it for an argument.
Expertthe three marks, and the question beneath them
In the working idiom of the research behind this book the three marks are called calculated, documented and construed. In the running text I translate them each time: "reproducible, see the Workbench" for the first; "seen in the source itself, location such-and-such" for the second; "a construal — it weighs as much as you allow it" for the third. And beneath all three lies one test question, which this book puts to every claim, its own as well as those of the consensus: is this a measurement or a majority? "Most scholars think X" is not a measurement; it is a count of opinions. A measurement has a figure, a base rate, or a falsifier. Where the consensus has only a majority, it counts in this book as construed — not as documented, and certainly not as established.
The second rule forces honesty against the temptation of the hit. Every "hit" comes with a base rate. If a date falls on a Wednesday and that seems significant, the first question is not "what does it mean?" but "how often would that happen by itself?". A Wednesday 14 Nisan occurs in the window of candidate years roughly once in seven; on its own, then, a Wednesday is chance-level, and I say so every time. A pattern that is not set against its probability of chance is not a pattern but an impression. This rule is merciless, for it cuts off the loveliest finds as soon as they come too cheap — and it holds, once again, on both sides of the table.
The third rule is the pivot of the whole book, and it is the hardest to sustain. Every measure I lay on my own thesis, I lay in the same breath on the consensus. The temptation is to treat claims that sound like pious number-speculation with severity, and to grant claims that sound like solid scholarship a free presumption of correctness. That is not rigour; that is a genre reflex, and it destroys the measure. If an opponent demands a falsifier of me — a concrete datum that would refute my thesis — that demand is just, and I shall give one every time. But then I ask the same in return: what would refute the prevailing reading? And often the consensus turns out to have less on offer there than it seems. "The prevailing dating places the matter in AD 33" — does it have a p-value, a base rate, a falsifier? If not, it is no stricter than my surmise; it is only more comfortable. The bar goes up on the side of the consensus; it does not come down on mine.
Dossierthe scale, the base rate, the falsifier
The scale by which this book weighs a signal (wherever a figure is possible): a p-value below 0.01 is called a strong signal; between 0.01 and 0.05, significant; between 0.05 and 0.10, marginal; above 0.10, no signal. This scale is not stretched to save an outcome; p = 0.21 remains p = 0.21, however welcome the outcome may be.
What a base rate does: it changes the question from "how striking is this?" into "how often does this happen by itself?". Only after the chance probability has been deducted does there remain what genuinely calls for explanation.
What a falsifier looks like on both sides. For this book: one clear pre-modern use of the double expression "three days and three nights" — in the corpus of the idiom itself, not in a later exegetical harmonisation about it — for demonstrably less than three full twenty-four-hour days would break the day-axis (ch. 3; Appendix B §B.3c weighs the candidates that have been put forward, and why they do not pass that test); an excavation or a tablet ruling out an intercalary month in AD 31 would break the year-axis (chs. 7–8). For the consensus: what would refute it? For a dating that rests on no calculation and no base rate, the answer is often "nothing concrete" — and an unfalsifiable proposition, on whichever side, is not a measurement but a conviction.
The fourth rule draws a wall straight through the book, and it protects the case against its own most attractive argument. There exists, alongside the historical and astronomical measuring core, a second and wholly different indication for the year 31: the prophetic clock of Daniel, the seventy weeks. It has two guises, and this book keeps them strictly apart. The one is a sum — 457 BC plus the foretold years comes out at the ministry, and three and a half years later at the crucifixion (chs. 9, 17) — and that sum is recomputable, a measurable line pointing the same way as the core. It may therefore count as convergence, but always earmarked as what it is: dependent on an assumed starting point and an assumed reading. For that reason the weighing computes it separately, with and without, and in the strictest weighing it falls away — so that whoever rejects its premises sees the measuring core standing untouched (ch. 16). The other guise is the full prophetic structure — the architecture of the seventy weeks and the wider clock that lies behind it. That is of an entirely different order from a lunar crescent or a textual citation — a construal within an assumed framework — and it therefore stands behind a firewall: the full prophetic structure comes up only after the weighing (ch. 17), and its elaboration belongs not in this book but in the wider work (God's Algorithm). What of the clock counts in the weighing is only the arithmetical 457-line — and that counts, as said, separately earmarked and premise-dependent: the measuring core can exist without it, and Appendix F shows what happens to the outcome when it is removed (ch. 16). This is no trick to hide the clock; it is the only honest place for it.
The fifth rule concerns something a reader has a right to before the first Greek word falls: which text this book actually cites. Whoever disputes over days and nights disputes, in the end, over words, and those words do not stand alike in every edition. This book cites the received text — the printed tradition of Stephanus (1550) and Scrivener (1894) — and not the eclectic text of the modern critical editions. That is a choice, not a matter of course, and it obliges three things.
- First: every Greek reading on which anything rests has been collated against eight printed editions — the two just named, the Byzantine majority text, the text of the Greek Church of 1904, Westcott-Hort, Tischendorf, Nestle 1904, and the modern critical text. Where they differ, the difference stands in the text of this book, not in a footnote no one reads. - Second, and this is the real price: the choice holds even where it tells against us. The heaviest counter-verse of the whole book is Luke 24:21, and precisely there the received text reads one word more — sêmeron, "today" — which the modern editions omit. That reading makes the objection against this book sharper, not softer. It stands there nonetheless, and it is written out where it belongs (ch. 13). Whoever chooses a text-form because he holds it to be the transmitted one may not trade it away at the place where it hurts; that would invalidate the measure of the third rule in a single sentence. - Third, and just as honestly: the received text is not everywhere the majority text. The two are easily confused. In this whole investigation exactly one place has been found where the printed received text stands against the Byzantine majority and where that difference weighs anywhere: Acts 12:25, where the majority and the oldest witnesses read eis and Stephanus and Scrivener ex. That verse carries no conclusion in this book; but whoever raises a text-form to a rule ought to name the place where his own rule does not run smooth, and not wait until another finds it.
Countercasethe case against this whole enterprise, at its strongest
Let the sharpest objection sound here in full, for it is just and it deserves an answer, not an evasion. It runs: this is an outsider reopening a question the discipline has settled. The safeguards of scholarship — peer review, pre-registration of hypotheses, the discipline of a research community — are absent here. A pattern found in the data after the fact cannot prove of itself that it is no accident; that is precisely what pre-registration exists for. Without those safeguards every tale of "convergence" is suspect: man sees faces in the clouds, and the believer sees them sharpest. This is the strongest case against this book, and I do not talk my way around it.
The answer is not that the objection is nonsense — it is not. The answer is that its true core is arithmetic, not an institution. What distinguishes a found pattern from a fluke is not a membership card but a test on data not yet seen: a blind list, a control population, an out-of-sample corpus, a prospective prediction. And that requires no university and no title. Wherever this book makes a load-bearing claim, it therefore tests it on material the hypothesis could not have coloured — as in Chapter 3, where the prediction about "three days and three nights" was first written down and only then tested on corpora that had not been used in the original research. Pre-registration as a class barrier shuts out the lone worker; pre-registration as arithmetic stands open to everyone, and this book submits itself to the second. That is the only honest rebuttal — and it binds me more than any opponent, not less.
There remains one instrument I must name separately, because it returns again and again and because it touches the consensus at its weakest point: the distinction between a rule that held in the first century and a rule that was later projected back onto it. Much of what the prevailing dating calls "the Jewish calendar rule" is in fact a later rule — the calculated calendar of after the fourth century, the fixed equinox limit — laid over the first century as though it had been there all along. To catch that in the act I borrow an instrument from an old master, David Chwolson.
ExcursusChwolson's instrument: "forgotten and restored"
Chwolson observed how a later ordinance could acquire the appearance of high antiquity. Of institutions "of which it was positively known that they were not customary in an earlier time", he wrote, it was often said that they were ancient but had been "in the course of time forgotten and brought back to remembrance by this or that old teacher" — "von denen man aber positiv wusste, dass sie in einer früheren Zeit nicht üblich waren, sagte man oft: sie seien uralt, wären aber im Laufe der Zeit vergessen und von diesem oder jenem alten Lehrer in Erinnerung gebracht worden", with the Hebrew formula shechachum vechazru viyasdum alongside (Chwolson, 1908, p. 24, the author's translation; seen primarily in the anastatically identical printing of 1892). The formula "it was always there, it was only forgotten and then restored" is therefore no proof of antiquity — it is often precisely the sign that a later rule is being antedated retroactively. This book uses this instrument every time the consensus applies a rule to the first century: did that rule truly hold then, or has it been projected back? The question is not rhetorical. The calculated molad calendar is demonstrably of later date; to loose it upon AD 31 is exactly the error Chwolson's instrument lays bare (chs. 7–8). And the measure cuts honestly both ways: where this book cites an early source, it must test just as hard whether that is not a later reading in an older garment.
Expertthe measurement and the verdict, and why that difference must be a rule
There is an operation this book performs again and again, and it deserves a name, for without one it looks like convenience. An author supplies a measurement and draws a conclusion from it. This book takes over the measurement and not the conclusion. That is permitted on one condition and otherwise not: that the two demonstrably rest on different material, so that the conclusion does not follow from the measurement but is set beside it.
The clearest case stands in chapter 9. Andrew Steinmann counts coins, papyri and inscriptions and concludes that there is no trace of inclusive reckoning of Tiberius's reign. That is the measurement, and this book uses it. In the same closing paragraph he writes that the evidence "calls into question AD 30 or 31 [...] as possible years of Jesus's crucifixion" — the year of this book, by name. That is the verdict, and it rests on something else: on a claim about the length of the ministry, which enters his own sentence as a premise — "since the Gospels' testimony [...] indicates a ministry for Jesus of over two years" — and not as a result of his survey. Whoever accepts the first is not bound to the second. Anyone with the article open can check it.
And this is no isolated move. The same operation underlies half a dozen places in this book: Strashny's barley series is taken over together with his own limitations; Stern's pp. 159–160 supply a calculated scheme and his judgement that it "would hardly have been functional"; Schaefer's R values are taken over together with his own R/DR ratio, which says when he is certain and when he is not; Milgrom's sentence about the barley is taken with the sentence behind it; and Newton's exclusion of AD 31 is read with the half-sentence that precedes it (ch. 1). Six times the same movement, and in two of those six it falls out against this book.
### ⛔ And that is why it belongs here as a rule and not as a find, for a rule binds in both directions. Whoever grants himself the right to use an opponent's measurement without his verdict gives that same right away. A reader may take over Strashny's dataset and leave the reading this book gives it. He may use Parker and Dubberstein's tables and reject this book's choice of intercalation. He may accept the crescent computation of appendix C and reject the calendar rule of chapter 8. That is precisely what is permitted here, and the book cannot complain when it happens.
Written out, then, the rule runs thus.Take the measurement where it rests on its own material, and leave the verdict where it rests on other material — name both, name the difference, and accept that another does the same with this book.The price of that freedom is that the book must make its own measurements separable from its own conclusions. That is what the three marks above do, and that is why they are there. [CONSTRUED]
That is the measure, entire: convergence instead of proof; three marks that never run together; a base rate with every hit; the same severity on both sides; a firewall around the construed clock; one text form, even where it tells against it; and two instruments for catching self-deception — the question "measurement or majority?", and Chwolson's "forgotten and restored". With this rod in hand the weighing can begin. And the first scale is not the hardest but the purest, for there lies the only true measurement of the whole case: not the year, but the day. To the sign the Lord Himself gave, and in which He counted the nights.
HeartHEART
A measuring-rod is a humble thing. It makes nothing greater than it is; it only keeps honest what is there. There is a temptation, precisely with what is holy, to want to say more than one can know — to claim a proof where a surmise is more honest, to take a beautiful number for a sign of God before it has been checked. This book resists that temptation, not out of coldness but out of reverence. For the truth has no need of exaggeration. If the date is right, it is right under the strictest measure too; and if it is not right, I want to know. Whoever truly believes that the earth bears its Maker can afford to be exact.
Note on the sources
This chapter advances no argument for the thesis; it fixes the rules of weighing. The illustration of "measurement or majority" (the year assumed eleven times without argument) was checked against the full text of the monograph concerned. The instrument "forgotten and restored" is drawn from the source itself, with its location. The three marks, the base-rate requirement, the falsifier requirement on both sides, and the firewall are the standing rules to which the rest of this book lets itself be held; where the book breaks them, you may call it to account. Hereafter the book claims a proof nowhere.
Literature (APA)
Chwolson, D. A. (1908). Das letzte Passamahl Christi und der Tag seines Todes (2nd, enlarged ed., p. 24). Leipzig. [1st ed.: Mémoires de l'Académie Impériale des Sciences de St.-Pétersbourg, 1892.]
Schmidt, T. C. (2025). Josephus and Jesus: New evidence for the one called Christ. Oxford University Press.
The previous chapter made a reproach: the field excluded AD 31 without measuring it, on an axiom. Such a reproach lays an obligation on the one who makes it. If I charge others with judging without a measure, then I must lay my own on the table — before anything is weighed, so that you can inspect it and, if need be, use it against me. This chapter is that measure. It contains not a single argument for the Wednesday. It only fixes how weighing is done in this book, and those rules bind me as hard as they bind the consensus.
The first rule is the most important, and it is a prohibition. This book nowhere claims a proof — nowhere does it say that something is established. Not out of false modesty, but because it does not belong here. Historical dating yields no proofs of the kind mathematics knows; it yields, at its best, convergence — the concurrence of independent indications pointing the same way. Convergence is strong; it is not certain. And the difference is no play on words. Whoever claims a proof closes the conversation; whoever says "convergence" keeps it open, and invites the reader to weigh for himself. This book therefore ends not with a proof but with a weighed surmise, and it will show you at every turn how much that surmise weighs — no more, and no less.
To make that weighing checkable, every claim in this book carries a mark. There are three, and they say precisely what a claim is and is not. A calculated fact is reproducible: you can recompute it yourself, and the route to it stands in a Workbench or an appendix. A documented fact has been seen in the source itself: there is a primary text, with its location, that carries it. A construed fact is an interpretation: it weighs as much as you allow the construal, and not a grain more. These three must never be allowed to run into one another. A calculation is no proof of a construal; a construal, however beautiful, may not take the place of a measurement. Where this book paints a picture — the partly darkened moon over Jerusalem, the Son in the deep like Jonah — that is construed, and it is marked as such, lest any reader mistake it for an argument.
Expertthe three marks, and the question beneath them
In the working idiom of the research behind this book the three marks are called calculated, documented and construed. In the running text I translate them each time: "reproducible, see the Workbench" for the first; "seen in the source itself, location such-and-such" for the second; "a construal — it weighs as much as you allow it" for the third. And beneath all three lies one test question, which this book puts to every claim, its own as well as those of the consensus: is this a measurement or a majority? "Most scholars think X" is not a measurement; it is a count of opinions. A measurement has a figure, a base rate, or a falsifier. Where the consensus has only a majority, it counts in this book as construed — not as documented, and certainly not as established.
The second rule forces honesty against the temptation of the hit. Every "hit" comes with a base rate. If a date falls on a Wednesday and that seems significant, the first question is not "what does it mean?" but "how often would that happen by itself?". A Wednesday 14 Nisan occurs in the window of candidate years roughly once in seven; on its own, then, a Wednesday is chance-level, and I say so every time. A pattern that is not set against its probability of chance is not a pattern but an impression. This rule is merciless, for it cuts off the loveliest finds as soon as they come too cheap — and it holds, once again, on both sides of the table.
The third rule is the pivot of the whole book, and it is the hardest to sustain. Every measure I lay on my own thesis, I lay in the same breath on the consensus. The temptation is to treat claims that sound like pious number-speculation with severity, and to grant claims that sound like solid scholarship a free presumption of correctness. That is not rigour; that is a genre reflex, and it destroys the measure. If an opponent demands a falsifier of me — a concrete datum that would refute my thesis — that demand is just, and I shall give one every time. But then I ask the same in return: what would refute the prevailing reading? And often the consensus turns out to have less on offer there than it seems. "The prevailing dating places the matter in AD 33" — does it have a p-value, a base rate, a falsifier? If not, it is no stricter than my surmise; it is only more comfortable. The bar goes up on the side of the consensus; it does not come down on mine.
Dossierthe scale, the base rate, the falsifier
The scale by which this book weighs a signal (wherever a figure is possible): a p-value below 0.01 is called a strong signal; between 0.01 and 0.05, significant; between 0.05 and 0.10, marginal; above 0.10, no signal. This scale is not stretched to save an outcome; p = 0.21 remains p = 0.21, however welcome the outcome may be.
What a base rate does: it changes the question from "how striking is this?" into "how often does this happen by itself?". Only after the chance probability has been deducted does there remain what genuinely calls for explanation.
What a falsifier looks like on both sides. For this book: one clear pre-modern use of the double expression "three days and three nights" — in the corpus of the idiom itself, not in a later exegetical harmonisation about it — for demonstrably less than three full twenty-four-hour days would break the day-axis (ch. 3; Appendix B §B.3c weighs the candidates that have been put forward, and why they do not pass that test); an excavation or a tablet ruling out an intercalary month in AD 31 would break the year-axis (chs. 7–8). For the consensus: what would refute it? For a dating that rests on no calculation and no base rate, the answer is often "nothing concrete" — and an unfalsifiable proposition, on whichever side, is not a measurement but a conviction.
The fourth rule draws a wall straight through the book, and it protects the case against its own most attractive argument. There exists, alongside the historical and astronomical measuring core, a second and wholly different indication for the year 31: the prophetic clock of Daniel, the seventy weeks. It has two guises, and this book keeps them strictly apart. The one is a sum — 457 BC plus the foretold years comes out at the ministry, and three and a half years later at the crucifixion (chs. 9, 17) — and that sum is recomputable, a measurable line pointing the same way as the core. It may therefore count as convergence, but always earmarked as what it is: dependent on an assumed starting point and an assumed reading. For that reason the weighing computes it separately, with and without, and in the strictest weighing it falls away — so that whoever rejects its premises sees the measuring core standing untouched (ch. 16). The other guise is the full prophetic structure — the architecture of the seventy weeks and the wider clock that lies behind it. That is of an entirely different order from a lunar crescent or a textual citation — a construal within an assumed framework — and it therefore stands behind a firewall: the full prophetic structure comes up only after the weighing (ch. 17), and its elaboration belongs not in this book but in the wider work (God's Algorithm). What of the clock counts in the weighing is only the arithmetical 457-line — and that counts, as said, separately earmarked and premise-dependent: the measuring core can exist without it, and Appendix F shows what happens to the outcome when it is removed (ch. 16). This is no trick to hide the clock; it is the only honest place for it.
The fifth rule concerns something a reader has a right to before the first Greek word falls: which text this book actually cites. Whoever disputes over days and nights disputes, in the end, over words, and those words do not stand alike in every edition. This book cites the received text — the printed tradition of Stephanus (1550) and Scrivener (1894) — and not the eclectic text of the modern critical editions. That is a choice, not a matter of course, and it obliges three things.
- First: every Greek reading on which anything rests has been collated against eight printed editions — the two just named, the Byzantine majority text, the text of the Greek Church of 1904, Westcott-Hort, Tischendorf, Nestle 1904, and the modern critical text. Where they differ, the difference stands in the text of this book, not in a footnote no one reads. - Second, and this is the real price: the choice holds even where it tells against us. The heaviest counter-verse of the whole book is Luke 24:21, and precisely there the received text reads one word more — sêmeron, "today" — which the modern editions omit. That reading makes the objection against this book sharper, not softer. It stands there nonetheless, and it is written out where it belongs (ch. 13). Whoever chooses a text-form because he holds it to be the transmitted one may not trade it away at the place where it hurts; that would invalidate the measure of the third rule in a single sentence. - Third, and just as honestly: the received text is not everywhere the majority text. The two are easily confused. In this whole investigation exactly one place has been found where the printed received text stands against the Byzantine majority and where that difference weighs anywhere: Acts 12:25, where the majority and the oldest witnesses read eis and Stephanus and Scrivener ex. That verse carries no conclusion in this book; but whoever raises a text-form to a rule ought to name the place where his own rule does not run smooth, and not wait until another finds it.
Countercasethe case against this whole enterprise, at its strongest
Let the sharpest objection sound here in full, for it is just and it deserves an answer, not an evasion. It runs: this is an outsider reopening a question the discipline has settled. The safeguards of scholarship — peer review, pre-registration of hypotheses, the discipline of a research community — are absent here. A pattern found in the data after the fact cannot prove of itself that it is no accident; that is precisely what pre-registration exists for. Without those safeguards every tale of "convergence" is suspect: man sees faces in the clouds, and the believer sees them sharpest. This is the strongest case against this book, and I do not talk my way around it.
The answer is not that the objection is nonsense — it is not. The answer is that its true core is arithmetic, not an institution. What distinguishes a found pattern from a fluke is not a membership card but a test on data not yet seen: a blind list, a control population, an out-of-sample corpus, a prospective prediction. And that requires no university and no title. Wherever this book makes a load-bearing claim, it therefore tests it on material the hypothesis could not have coloured — as in Chapter 3, where the prediction about "three days and three nights" was first written down and only then tested on corpora that had not been used in the original research. Pre-registration as a class barrier shuts out the lone worker; pre-registration as arithmetic stands open to everyone, and this book submits itself to the second. That is the only honest rebuttal — and it binds me more than any opponent, not less.
There remains one instrument I must name separately, because it returns again and again and because it touches the consensus at its weakest point: the distinction between a rule that held in the first century and a rule that was later projected back onto it. Much of what the prevailing dating calls "the Jewish calendar rule" is in fact a later rule — the calculated calendar of after the fourth century, the fixed equinox limit — laid over the first century as though it had been there all along. To catch that in the act I borrow an instrument from an old master, David Chwolson.
ExcursusChwolson's instrument: "forgotten and restored"
Chwolson observed how a later ordinance could acquire the appearance of high antiquity. Of institutions "of which it was positively known that they were not customary in an earlier time", he wrote, it was often said that they were ancient but had been "in the course of time forgotten and brought back to remembrance by this or that old teacher" — "von denen man aber positiv wusste, dass sie in einer früheren Zeit nicht üblich waren, sagte man oft: sie seien uralt, wären aber im Laufe der Zeit vergessen und von diesem oder jenem alten Lehrer in Erinnerung gebracht worden", with the Hebrew formula shechachum vechazru viyasdum alongside (Chwolson, 1908, p. 24, the author's translation; seen primarily in the anastatically identical printing of 1892). The formula "it was always there, it was only forgotten and then restored" is therefore no proof of antiquity — it is often precisely the sign that a later rule is being antedated retroactively. This book uses this instrument every time the consensus applies a rule to the first century: did that rule truly hold then, or has it been projected back? The question is not rhetorical. The calculated molad calendar is demonstrably of later date; to loose it upon AD 31 is exactly the error Chwolson's instrument lays bare (chs. 7–8). And the measure cuts honestly both ways: where this book cites an early source, it must test just as hard whether that is not a later reading in an older garment.
Expertthe measurement and the verdict, and why that difference must be a rule
There is an operation this book performs again and again, and it deserves a name, for without one it looks like convenience. An author supplies a measurement and draws a conclusion from it. This book takes over the measurement and not the conclusion. That is permitted on one condition and otherwise not: that the two demonstrably rest on different material, so that the conclusion does not follow from the measurement but is set beside it.
The clearest case stands in chapter 9. Andrew Steinmann counts coins, papyri and inscriptions and concludes that there is no trace of inclusive reckoning of Tiberius's reign. That is the measurement, and this book uses it. In the same closing paragraph he writes that the evidence "calls into question AD 30 or 31 [...] as possible years of Jesus's crucifixion" — the year of this book, by name. That is the verdict, and it rests on something else: on a claim about the length of the ministry, which enters his own sentence as a premise — "since the Gospels' testimony [...] indicates a ministry for Jesus of over two years" — and not as a result of his survey. Whoever accepts the first is not bound to the second. Anyone with the article open can check it.
And this is no isolated move. The same operation underlies half a dozen places in this book: Strashny's barley series is taken over together with his own limitations; Stern's pp. 159–160 supply a calculated scheme and his judgement that it "would hardly have been functional"; Schaefer's R values are taken over together with his own R/DR ratio, which says when he is certain and when he is not; Milgrom's sentence about the barley is taken with the sentence behind it; and Newton's exclusion of AD 31 is read with the half-sentence that precedes it (ch. 1). Six times the same movement, and in two of those six it falls out against this book.
### ⛔ And that is why it belongs here as a rule and not as a find, for a rule binds in both directions. Whoever grants himself the right to use an opponent's measurement without his verdict gives that same right away. A reader may take over Strashny's dataset and leave the reading this book gives it. He may use Parker and Dubberstein's tables and reject this book's choice of intercalation. He may accept the crescent computation of appendix C and reject the calendar rule of chapter 8. That is precisely what is permitted here, and the book cannot complain when it happens.
Written out, then, the rule runs thus. Take the measurement where it rests on its own material, and leave the verdict where it rests on other material — name both, name the difference, and accept that another does the same with this book. The price of that freedom is that the book must make its own measurements separable from its own conclusions. That is what the three marks above do, and that is why they are there. [CONSTRUED]
That is the measure, entire: convergence instead of proof; three marks that never run together; a base rate with every hit; the same severity on both sides; a firewall around the construed clock; one text form, even where it tells against it; and two instruments for catching self-deception — the question "measurement or majority?", and Chwolson's "forgotten and restored". With this rod in hand the weighing can begin. And the first scale is not the hardest but the purest, for there lies the only true measurement of the whole case: not the year, but the day. To the sign the Lord Himself gave, and in which He counted the nights.
HeartHEART
A measuring-rod is a humble thing. It makes nothing greater than it is; it only keeps honest what is there. There is a temptation, precisely with what is holy, to want to say more than one can know — to claim a proof where a surmise is more honest, to take a beautiful number for a sign of God before it has been checked. This book resists that temptation, not out of coldness but out of reverence. For the truth has no need of exaggeration. If the date is right, it is right under the strictest measure too; and if it is not right, I want to know. Whoever truly believes that the earth bears its Maker can afford to be exact.
Note on the sources
This chapter advances no argument for the thesis; it fixes the rules of weighing. The illustration of "measurement or majority" (the year assumed eleven times without argument) was checked against the full text of the monograph concerned. The instrument "forgotten and restored" is drawn from the source itself, with its location. The three marks, the base-rate requirement, the falsifier requirement on both sides, and the firewall are the standing rules to which the rest of this book lets itself be held; where the book breaks them, you may call it to account. Hereafter the book claims a proof nowhere.
Literature (APA)
Chwolson, D. A. (1908). Das letzte Passamahl Christi und der Tag seines Todes (2nd, enlarged ed., p. 24). Leipzig. [1st ed.: Mémoires de l'Académie Impériale des Sciences de St.-Pétersbourg, 1892.]
Schmidt, T. C. (2025). Josephus and Jesus: New evidence for the one called Christ. Oxford University Press.