The whole record

Every text that mentions him, side by side.

This page does not produce a harmony. Merging the four gospels into one itinerary is a genre with eighteen centuries behind it, and it works by silently picking whose order to keep whenever two accounts disagree — the choice is never marked, so the result reads as though the sources agree. They do not. What is computed here instead is the disagreement: every pair of events that two gospels put in opposite order, found by comparing the sequences rather than by anyone listing the ones they happen to remember.

And around it, the whole surviving record — 17 texts, Christian and not — with what each one individually attests, and which of them are actually independent of each other.

Two things this page refuses to do. It will not rank the gospels' orderings against each other: nothing computable adjudicates whether John or Mark has the Temple incident in the right place, and a number pretending to would be invention dressed as method. And it will not merge the counts into a score for "how well attested Jesus is" — the sources attest different claims, at different removes, with different textual security, and flattening that into one figure is the move that serves polemic in both directions.

What each source actually says

One row per text, one column per claim. The shaded columns are the ones that carry any chronological weight at all — and how few of them are populated outside the Christian texts is the substance of the next section.

Source	A person existed	Was executed	Specifically by crucifixion	Under Pontius Pilate	In the reign of Tiberius	In Judaea	At Passover	On the eve of Passover (14 Nisan)	Had a brother James	Was worshipped by followers	Darkness at the death
Markc. AD 65–75	●	●	●	●	·	●	●	·	·	·	●
Matthewc. AD 80–90 · draws on mark	●	●	●	●	·	●	●	·	·	·	●
Lukec. AD 80–90 · draws on mark	●	●	●	●	●	●	●	·	·	·	●
Johnc. AD 90–110	●	●	●	●	·	●	●	●	·	·	·
Paul (the undisputed letters)c. AD 50–60	●	●	●	·	·	·	·	·	●	●	·
Actsc. AD 80–90 · draws on luke	●	●	●	●	·	●	·	·	·	·	·
Gospel of Thomasc. AD 100–150 (disputed)	●	·	·	·	·	·	·	·	·	·	·
Tacitus, Annals XV.44c. AD 116	●	●	·	●	●	●	·	·	·	·	·
Josephus, Antiquities XVIII.63–64 · disputedc. AD 93	●	●	·	●	·	·	·	·	·	·	·
Josephus, Antiquities XX.200c. AD 93	●	·	·	·	·	·	·	·	●	·	·
Pliny the Younger, Letters X.96c. AD 112	·	·	·	·	·	·	·	·	·	●	·
Suetonius, Claudius 25 · disputedc. AD 121	●	·	·	·	·	·	·	·	·	·	·
Mara bar Serapion, letter to his son · disputedAD 73 – 3rd c. (uncertain)	●	●	·	·	·	·	·	·	·	·	·
Babylonian Talmud, Sanhedrin 43a · disputedcompiled c. AD 500, earlier strata	●	●	·	·	·	·	●	●	·	·	·
Lucian, The Passing of Peregrinusc. AD 165	●	●	●	·	·	·	·	·	·	·	·
Celsus, via Origen, Contra Celsumc. AD 178	●	·	●	·	·	·	·	·	·	·	·
Thallus, via Julius Africanus, via Syncellus · disputedlost; cited 3rd c., quoted 9th c.	·	·	·	·	·	·	·	·	·	·	●

What the record outside the Christian texts actually pins down

This is the number worth having, and it is computed rather than asserted, because it is claimed wrongly in both directions — inflated by people arguing that the external record corroborates the gospels in detail, and dismissed by people arguing it does not exist.

Chronological claim	Narrows the date	Christian sources	Non-Christian	Non-Christian, contested ones removed
Under Pontius Pilate	to AD 26–36	5	2 (Tacitus, Annals XV.44; Josephus, Antiquities XVIII.63–64)	1
In the reign of Tiberius	to AD 14–37	1	1 (Tacitus, Annals XV.44)	1
At Passover	to one week a year	4	1 (Babylonian Talmud, Sanhedrin 43a)	0
On the eve of Passover (14 Nisan)	to one day a year	1	1 (Babylonian Talmud, Sanhedrin 43a)	0

Read the last two columns together. The entire non-Christian record narrows the death to the ten years of Pilate's prefecture and no further — and once the textually contested passages are set aside, that rests on Tacitus alone. The Testimonium Flavianum is the other witness and it is the one everybody argues about. Nothing outside the Christian texts speaks to the year.

One cell is stranger than the rest. The only external source that names a day is the Babylonian Talmud, which puts the execution on the eve of Passover — John's chronology, against the synoptics'. It is the single most precise external claim in the entire record, and it sides with the minority reading this site's own astronomy also favours.

That cell is the most fragile thing on this page, and it is worth being blunt about why. It is the one a reader will want to quote, and it carries four separate problems that have to be stated together or it flatters itself. It is late — the Babylonian Talmud is compiled around AD 500. It is polemical, not reportage. Its independence is doubtful: a source writing centuries after Christians were making this claim may simply be repeating it back. And the passage is absent from many printed editions, having been removed by censors, so which text you consult decides whether you see it at all. Worst of all for the use we are making of it, whether its "Yeshu" is Jesus of Nazareth is genuinely disputed — some rabbinic traditions place that figure a century earlier. A striking agreement between a hostile witness and John is exactly the kind of result that deserves more suspicion than a dull one, not less.

Why the counts are smaller than they look

Seven sources attest the execution under Pilate. They are not seven witnesses. Matthew and Luke both draw on Mark, and Acts shares Luke's author, so the raw count of 7 collapses to 4 independent streams — and to 3 once the contested Testimonium is set aside. Copies are not corroboration. Every count on this page is reported both ways for that reason.

What the archaeology contributes

The question put to a project like this more often than any other is what the physical evidence shows. The answer is unusually clean, and it is worth stating before the table rather than after it: a great deal about the world he lived in, and nothing whatever about when he died.

Find	Status	What it establishes	Narrows the date?
The Pilate Stone Caesarea Maritima, 1961	accepted	That Pontius Pilate existed and held the office the gospels give him.	no — It fixes the man and the office. It carries no date of its own, and Pilate's ten years were never the disputed part.
The Caiaphas ossuary North Talpiot, Jerusalem, 1990	accepted	That the high-priestly family named in the trial narratives existed.	no — A burial box is not dated, and a family being real does not date an event in its lifetime.
Yehohanan's heel bone Giv'at ha-Mivtar, Jerusalem, 1968	accepted	That crucifixion was practised in first-century Judaea, and how.	no — It corroborates the method and the era. It is a different man.
The James ossuary Antiquities market, surfaced 2002	contested	Nothing securely. Its authenticity remains unresolved.	no — Even taken as genuine, an ossuary inscription names people. It does not date them, and all three names were among the commonest of the period.
The Shroud of Turin Documented from the 14th century	contested	Nothing bearing on this question.	no — If the radiocarbon stands, it is medieval and irrelevant. If it falls, it is undated and still irrelevant to which YEAR someone died.
The Talpiot "Jesus family tomb" East Talpiot, Jerusalem, 1980	rejected	Nothing. The identification is rejected by most working archaeologists.	no — Even if the identification were accepted, ossuaries are not dated to the year.
Nazareth and Capernaum excavations Ongoing since the 19th century	accepted	That both settlements were inhabited in the first century.	no — A village existing for centuries does not date an event within them.

0 of 7. 4 of these finds are securely accepted, 2 are contested and 1 is rejected outright — and the column that matters reads the same all the way down regardless. Pilate's own dedication stone confirms the prefect and his title. A high priest's ossuary confirms the family. A nail through a heel bone confirms how crucifixion was actually done in first-century Judaea. Not one of them moves the year by a day.

Why the honest version cuts both ways

A list like this is easy to bend. Read one way — Pilate's stone, Caiaphas's ossuary, a crucified man's ankle — it looks like mounting physical corroboration of the narrative. Read the other way, weighted toward the Shroud and the Talpiot tomb, it looks like a record of wishful archaeology. Both readings dodge the same fact: authentic and contested alike, every object on the list establishes context and none establishes chronology. The genuine finds are genuine. They are also silent on the only question this site is asking.

Where the gospels contradict each other

Each column is one gospel's narrative order. A line joins the same event across accounts, so a crossing is a disagreement — two gospels putting the same pair of events in opposite order. Nothing here is annotated as a conflict that the geometry does not already show; the 4 conflicts below were found by comparing the sequences, not by listing the famous ones.

The 4 conflicts, computed

These two events	are ordered one way by	and the other way by
Cleansing of the Temple / Feeding of the five thousand	John	Mark, Matthew, Luke
Cleansing of the Temple / Peter's confession	John	Mark, Matthew, Luke
Cleansing of the Temple / Entry into Jerusalem	John	Mark, Matthew, Luke
Anointing at Bethany / Entry into Jerusalem	John	Mark, Matthew

Every one of them is John against the synoptics, and Cleansing of the Temple is in 3 of the 4. John puts it at the very start of the ministry; Mark, Matthew and Luke put it in the final week. The usual reconciliation is that it happened twice. That is possible, and it is also exactly the move a harmony makes when the alternative is admitting the accounts differ — so it is worth noticing that no gospel narrates two.

The second is quieter and harder to explain away: the anointing at Bethany falls before the entry into Jerusalem in John and after it in Mark and Matthew. Nobody proposes two anointings.

What this does and doesn't bear on the date

Almost nothing, directly — and that is worth stating, because it would be easy to let a page this size imply otherwise. Narrative order is not chronology, none of these events carries a date, and the accounts could disagree about all of them while agreeing on the year. The one conflict that does touch the date is the one this whole site is built around: whether the death fell on 14 or 15 Nisan. That single disagreement moves the surviving years more than every parameter in the astronomical sweep put together.

Read this before quoting anything above. One: all 17 sources are marked unverified — dates, attributions and contents are recorded from secondary knowledge and the editions have not been opened. This page has more citations than any other on the site and exactly the same amount of checking behind them, which is none. Two: which claims a source attests is a reading, not a measurement; reasonable people place several of these dots differently, and the Testimonium, Suetonius's "Chrestus" and Thallus are contested enough to be flagged in the table itself. Three: the dependency map behind the independent-stream counts assumes Markan priority. It is the majority view, not a fact, and a different source theory gives different numbers. Four: a text being late, hostile or derivative does not make it worthless, and a text being early does not make it right — the columns are there to be weighed, not totted up.

The rival reconstructions → · The textual arguments, enumerated → · Sources →