Almost nothing about this figure can be discussed without someone’s whole framework being at stake, so it is worth stating the method at the outset. What follows is the historical record and the teaching as the sources present it. Whether the claims made about him are true is not a question a page like this can settle, and it will not pretend to.
Three kinds of statement appear below, and they are kept apart throughout. There is what a source says, which is a fact about the source and not yet a fact about the world. There is what historians of every persuasion accept, which is a short list and is marked as such. And there is what is disputed, where the dispute is reported rather than settled, with names attached to the positions. A page that blurs those three reads more smoothly and is worth less.
Two things are agreed by essentially every historian who works on the period, believer or not: a Galilean Jew named Yeshua taught briefly in the late 20s or early 30s CE, and the Roman prefect Pontius Pilate had him crucified.
Everything else is argument.
The sources, and their problem
He wrote nothing. Neither did anyone who heard him, for at least a generation.
The earliest Christian documents are Paul’s letters, from around 50 to 60 CE - twenty to thirty years after the execution, written in Greek by a man who never met him, to congregations around the Mediterranean, about their disputes. Paul refers to a handful of biographical facts and almost never to the teaching. He is writing to people who already know the story, so he does not tell it.
The gospels come later. Mark around 70 CE, at or just after the Roman destruction of the Temple. Matthew and Luke in the 80s, both using Mark as a framework and both drawing on a shared collection of sayings that Mark does not have - the hypothetical source called Q. John, around 90 to 110, different in nearly every respect.
They are written in Greek. He taught in Aramaic. So even the sayings that go back to him have been through a translation before they reach a page.
They are also not neutral, and do not claim to be. They are proclamations, arranged to persuade. Matthew shapes the material to show a new Moses; Luke writes for a Greek readership and is interested in the poor and in women; John is doing theology in narrative form.
Historians therefore use criteria. Multiple attestation: material appearing in independent sources is more likely early. Dissimilarity: material that fits neither the Judaism of the time nor the concerns of the later church is unlikely to have been invented by either. Embarrassment: material the early church would have preferred not to have to explain is unlikely to be fabricated - his baptism by John, which implies subordination; the crucifixion itself; his own followers’ failure to understand him; the women who found the empty tomb, whose testimony carried little legal weight in that world.
Outside the Christian sources there is very little, and what there is matters. Josephus, writing around 93 CE, mentions him twice: once in a passage plainly tampered with by later Christian copyists but with a probable authentic core, and once, almost in passing, identifying James as the brother of Jesus who was called Christ - a reference nobody had a motive to insert. Tacitus, around 116 CE, explaining Nero’s persecution, records that the founder of the sect, Christus, had been executed during Tiberius’ reign by Pontius Pilate, whom he calls a procurator and who was in fact a prefect. He is not friendly; he calls the movement a pernicious superstition. That is precisely why the notice is useful.
Between the events and the first written gospel lies a period of oral transmission of about forty years, and what happened in it is the oldest quarrel in the discipline. Rudolf Bultmann and the form critics of the 1920s held that the sayings and stories circulated as detached units shaped by the needs of the communities that used them, so that the setting a gospel supplies is usually the evangelist’s and not history’s. Richard Bauckham argued the opposite case in Jesus and the Eyewitnesses (2006): that the tradition was not anonymous community property but was attached to named individuals, and that the minor characters Mark bothers to name are named because they were the informants. The proposal has been influential and is not the consensus. Dale Allison and others have replied that memory research gives no warrant for the confidence, and that the recall of gist is robust while the recall of wording is not.
James Dunn reframed the question in Jesus Remembered (2003) in a way that has been more widely accepted than either extreme. The earliest recoverable layer of tradition is not raw event but event already interpreted by people to whom it mattered. There is no uninterpreted deposit underneath waiting to be uncovered, and a method that assumes there is will spend itself digging for something that was never there.
The physical evidence has its own limits. No original of any New Testament book survives. The earliest fragment is a scrap of John in Manchester, conventionally dated to the first half of the second century, though Brent Nongbri has argued that the palaeographic range honestly extends decades later. The earliest substantial copies of the gospels date from around 200, and the first complete Bibles from the fourth century. Roughly 5,800 Greek manuscripts are known, carrying several hundred thousand recorded variants between them, the overwhelming majority of which concern spelling and word order and change nothing. A handful, treated below, change a great deal.
When the gospels about Jesus were written, and how anonymous books acquired their names
None of the four gospels names its author. Not one contains a sentence in which a writer identifies himself. Luke addresses a patron called Theophilus and says that others have written before him, without giving a name of his own. John ends with a reference to the disciple which testifieth of these things, who is never identified in the text.
The names arrived later, and the process is partly visible.
The titles take the form euangelion kata plus a name, the gospel according to Mark and so on, which is the phrasing used to distinguish copies sitting side by side in a collection rather than the phrasing an author uses on his own book. The earliest manuscripts carrying such titles are from around 200: the Bodmer papyrus of John and the Bodmer papyrus of Luke and John both have them. Papias of Hierapolis, writing perhaps around 110 to 130 and preserved only in quotation by later writers, already speaks of a Mark who was Peter’s interpreter and set down what he remembered without order, and of a Matthew who compiled the logia in the Hebrew dialect. The four names appear as a settled, closed set with Irenaeus of Lyon, in Against Heresies 3.1.1, around 180.
The argument about what that means is live. Martin Hengel held that the titles must go back nearly to the beginning, since a scroll released into circulation needs identification and the title forms show almost no variation across the manuscript tradition, which is not what happens when names are invented independently in different places. Against him: Papias’s description of Matthew, a collection of sayings in Hebrew, does not fit the Greek Matthew that exists, which is a Greek expansion of Greek Mark; and the ascriptions attach the two most important gospels to non-apostles, Mark and Luke, which is an odd choice for pure invention but not an impossible one.
The dates rest on internal evidence. Mark 13 describes the temple’s destruction in terms specific enough that most scholars read it as written at or just after 70 CE, which is why Mark is normally placed around then. Matthew and Luke are placed in the 80s because they use Mark and appear to know the war is over. John is placed between 90 and 110 on the strength of its developed theology and its references to expulsion from the synagogue. A minority argues for earlier dates throughout: John A. T. Robinson’s Redating the New Testament (1976) put the entire New Testament before 70, chiefly on the argument that no book mentions the fall of Jerusalem as an accomplished fact; Maurice Casey and James Crossley have both argued for an early Mark on other grounds. The majority is not unanimous and does not pretend to precision beyond a decade.
Three things about the relationship between the first three gospels are near enough to settled that the rest of the field is built on them.
Markan priority. Mark is the shortest, and about ninety per cent of it reappears in Matthew. Where Matthew and Luke agree in the order of events, they agree with Mark; where Mark is absent, their order diverges. Both consistently improve Mark’s rough Greek and both soften passages that create difficulty. Mark 6:5 says that at Nazareth he could there do no mighty work; Matthew 13:58 changes this to he did not many mighty works there because of their unbelief. Mark 10:18 has him deflect the word good; Matthew rewrites the exchange so that the question is about what is good rather than about who is. Editing runs in one direction, and the direction is away from Mark.
The double tradition. Matthew and Luke share around two hundred and thirty verses absent from Mark, mostly sayings, often in a similar sequence and sometimes word for word. The two-source hypothesis explains this by a lost sayings collection, labelled Q from the German Quelle, source. It has never been found and no ancient writer mentions it.
The dissent. The Farrer hypothesis dispenses with Q by supposing that Luke read Matthew as well as Mark. Mark Goodacre made the fullest modern case in The Case Against Q (2002), resting heavily on the minor agreements: the places where Matthew and Luke agree against Mark in small details, which on the two-source hypothesis have to be coincidences and on the Farrer hypothesis are simply Luke copying Matthew. The position is a growing minority rather than a consensus. A smaller group revives the older Griesbach hypothesis, in which Matthew came first.
The stakes are not academic housekeeping. John Kloppenborg’s Excavating Q (2000) reconstructed a compositional history for the document in which the earliest layer is wisdom material and the apocalyptic material is a later addition, and that stratification became one of the main supports for the non-apocalyptic portraits of Jesus. Kloppenborg himself has objected repeatedly to that use of it: a claim about the literary composition of a document is not a claim about the chronology of a man’s preaching, and treating the two as equivalent stacks one hypothesis on another.
What language Jesus spoke, and what survives of it in the gospels
Aramaic, and the point is not trivia. Every word attributed to him in every source is a translation made by somebody else.
Aramaic had been the common language of the Near East since the Persian empire, and in first-century Jewish Palestine the vernacular was a western dialect of it. Hebrew was the language of scripture and liturgy and remained in live use in some circles, as the Qumran scrolls and the later Bar Kokhba letters show. Greek was the administrative and commercial language, current in the cities and among anyone who traded, and the two Greek-style cities of Galilee were within a day’s walk of Nazareth. Whether Jesus himself spoke Greek is disputed and unresolvable; Stanley Porter and others have argued for some competence, while Meier and Casey doubt it mattered.
Mark keeps a small number of Aramaic words untranslated and then translates them, which is a habit of someone reproducing something heard.
- Talitha koum, to the daughter of Jairus, at Mark 5:41.
- Ephphatha, to the deaf man, at Mark 7:34.
- Abba, in Gethsemane, at Mark 14:36.
- Eloi, Eloi, lama sabachthani, from the cross, at Mark 15:34.
- Corban, a technical term for a dedicated offering, at Mark 7:11.
- Boanerges, the nickname given to the sons of Zebedee, at Mark 3:17.
Others survive elsewhere: raca at Matthew 5:22, mammon, Golgotha, Gethsemane, Cephas, and the patronymics Bartimaeus and Bar-Jonah.
Maurice Casey went further in Aramaic Sources of Mark’s Gospel (1998), reconstructing Aramaic originals behind selected passages and arguing that the awkwardness of Mark’s Greek in those places is the awkwardness of literal translation. The method has admirers and sceptics in roughly equal measure; the objection is that a reconstruction of a lost text from a translation of it can be made to support whatever the reconstructor already believed.
Where the translation problem bites hardest is in wordplay. Matthew 23:24 has the image of straining out a gnat and swallowing a camel, which in Aramaic sets qalma against gamla and in Greek is merely a contrast of sizes. Matthew 16:18 plays Peter against rock, which works in Greek and works in Aramaic with Kepha and not in Hebrew. Such observations are suggestive rather than probative, and Matthew Black’s An Aramaic Approach to the Gospels and Acts collected the serious candidates long before the popular ones appeared.
Whether he could read is a separate question and is genuinely open. Luke 4:16 to 20 shows him taking the scroll of Isaiah in the synagogue at Nazareth and reading from it, in a scene with no parallel in Mark. Catherine Hezser’s study of Jewish literacy in Roman Palestine puts the general rate well below ten per cent. Chris Keith argued in Jesus’ Literacy (2011) that the gospels themselves disagree on the point, and that the disagreement is the interesting datum: some communities remembered a teacher who handled texts, and some remembered one who did not, and both were making a claim about status.
Whether Jesus existed, and what the mythicist case rests on
Among people who hold research posts in New Testament studies, ancient history or classics, the position that Jesus never existed has effectively no support. That is a statement about the state of a discipline, not an argument, and the arguments are worth setting out because the question is asked constantly.
The non-Christian notices, in order of weight:
Tacitus, Annals 15.44, written around 116, explaining how Nero found a scapegoat for the fire of 64. The Latin runs auctor nominis eius Christus Tiberio imperitante per procuratorem Pontium Pilatum supplicio adfectus erat.
Christus, from whom the name had its origin, suffered the extreme penalty during the reign of Tiberius at the hands of one of our procurators, Pontius Pilatus, and a most mischievous superstition, thus checked for the moment, again broke out not only in Judaea, the first source of the evil, but even in Rome. Tacitus, Annals XV.44, translated by Church and Brodribb, 1876
Two details are argued over. The single surviving eleventh-century manuscript now reads Christianos, but the first vowel has been altered, and ultraviolet examination shows an erased e beneath it, so what Tacitus wrote was Chrestianos; and Tacitus gives Pilate the title procurator, when the inscription found at Caesarea in 1961 shows he was prefect. Neither difficulty has persuaded Tacitean specialists that the passage is an interpolation, and the near-unanimous view among classicists is that it is genuine.
Josephus, twice, treated in the next section.
Pliny the Younger, Letters 10.96, around 112, asking Trajan how to handle Christians in Bithynia and reporting that they met before dawn and sang a hymn to Christ as to a god. This is evidence about Christians, not directly about Jesus.
Suetonius, Claudius 25.4, reporting that Claudius expelled Jews from Rome who were rioting at the instigation of Chrestus. Whether this is a garbled reference to disputes about Jesus or to an otherwise unknown agitator is unresolved.
Mara bar Serapion, a Syriac letter of uncertain date referring to the execution of the wise king of the Jews. The dating ranges across two centuries and the value depends on it.
The rabbinic references, principally in tractate Sanhedrin, are late and polemical and add nothing a historian can use.
The heaviest evidence is not any of these. It is Paul. In Galatians 1:18 to 19 he says that he went up to Jerusalem to see Cephas, stayed fifteen days, and saw no other apostle save James the Lord’s brother. That places Paul in a room with the brother of a man he says was crucified, within a few years of the crucifixion, and it is a passing remark in an argument about something else. Paul mentions the brothers of the Lord again at 1 Corinthians 9:5, says at Romans 1:3 that he was descended from David according to the flesh, and at 1 Corinthians 15:3 to 8 passes on a list of appearances he says he received rather than composed.
The mythicist case, in its serious forms, has been made by G. A. Wells, who later changed his position, by Earl Doherty, by Robert M. Price, and most fully by Richard Carrier in On the Historicity of Jesus (2014). It holds that the earliest layer of Christian belief concerned a celestial figure crucified in a heavenly realm, and that the earthly biography is a later fiction. Its strongest single move is against the Galatians evidence: that the brother of the Lord may be a cultic title for an ordinary believer rather than a statement of kinship, in which case the one apparently unanswerable datum dissolves. The standard reply is that Paul elsewhere distinguishes the brothers of the Lord from apostles and from believers generally, which a title covering all believers could not do. Ehrman’s Did Jesus Exist? (2012) makes that case at length from within the field. Carrier’s doctorate is in ancient history and Price’s is in New Testament, so the claim sometimes made that no qualified person holds the position is overstated; the accurate statement is that two or three do, out of a field of thousands, and that no departmental consensus anywhere has moved.
One offshoot deserves a sentence because it circulates. René Salm’s The Myth of Nazareth (2008) argued that the site was uninhabited in the first century and therefore that the hometown is fiction. The Nazareth Archaeological Project, directed by Ken Dark since 2006 and published as Archaeology of Jesus’ Nazareth (Oxford, 2023), found first-century Jewish domestic occupation and revised the settlement’s likely population upward, toward a thousand rather than the two or three hundred usually assumed.
What the outside sources establish is narrow and should be stated narrowly. By the early second century, Roman and Jewish writers who had no sympathy for the movement treated the execution of its founder under Pilate as a known fact. None of them conducted an investigation. The weight is carried by Paul and by the multiplicity of early Christian sources, not by Tacitus.
What Josephus wrote about Jesus, and which parts of it he did not
The passage at Antiquities of the Jews 18.63 to 64, called the Testimonium Flavianum, is the most examined paragraph in ancient historiography outside the New Testament. As it stands in the Greek manuscripts it reads, in the wording most commonly reproduced in English:
About this time there lived Jesus, a wise man, if indeed one ought to call him a man. For he was one who performed surprising deeds and was a teacher of such people as accept the truth gladly. He won over many Jews and many of the Greeks. He was the Christ. And when, upon the accusation of the principal men among us, Pilate had condemned him to a cross, those who had first come to love him did not cease. He appeared to them spending a third day restored to life, for the prophets of God had foretold these things and a thousand other marvels about him. And the tribe of the Christians, so called after him, has still to this day not disappeared. Josephus, Antiquities XVIII.63–64
Josephus was a Jewish priest who wrote as a Jew and died as one. Three clauses cannot have come from him: the aside if indeed one ought to call him a man, the flat assertion He was the Christ, and the report of the resurrection on the third day as fact. No serious scholar of any persuasion defends those three in that form.
The question is what remains once they are removed, and there are three answers in the field.
Partial authenticity, the majority position. A shorter, neutral notice stood in the original and was embellished by a Christian copyist. The strongest evidence for it is Origen, who complains in the third century that Josephus did not accept Jesus as the Christ, which means Origen’s copy did not contain the sentence that says he did, while apparently containing something. Meier, Vermes, Dunn, Van Voorst, Craig Evans and Alice Whealey all hold versions of this.
Total interpolation, a minority that has grown. Ken Olson has argued at length that the vocabulary and cadence of the passage are characteristic of Eusebius, who is the first writer to quote it in the received form in the early fourth century. Louis Feldman, the leading Josephus specialist of his generation, moved to this position late in his career. Paul Hopper argued the case on narrative-linguistic grounds.
Substantial authenticity, defended more recently by G. J. Goldberg and T. C. Schmidt, who argue the passage fits Josephus’s paraphrasing habits better than critics allow.
Two non-Greek versions complicate matters and are frequently cited. The tenth-century Arabic chronicle of Agapius of Hierapolis has a form of the passage saying that he was perhaps the Messiah, and the twelfth-century Syriac chronicle of Michael the Syrian has he was thought to be the Messiah. These were once treated as windows onto the original. Whealey’s work on the Syriac tradition indicates they descend from a translation of Eusebius rather than from an independent line, which weakens the argument considerably.
The second reference is short, incidental and far less disputed. At Antiquities 20.200, describing how the high priest Ananus convened a court during the interval between two Roman governors in 62 CE, Josephus identifies the man condemned as the brother of Jesus who was called Christ, whose name was James. There is no motive for a Christian to insert an identification of this kind, which subordinates Jesus to the description rather than the reverse, and it is accepted by nearly everyone.
What Josephus establishes: a Jewish historian writing in Rome in the 90s knew of a Jesus executed under Pilate, knew he had a brother named James who was killed in Jerusalem in 62, and knew that the movement had not died. He establishes nothing whatever about the content of the preaching.
The bedrock: the baptism of Jesus by John, and the crucifixion under Pilate
The baptism by John, and the crucifixion under Pilate. Those two are the ones cited when the field is asked what is beyond argument, and the reason is the same in both cases: each was a problem for the people transmitting it.
The baptism. John the Baptist is independently attested outside the gospels. Josephus devotes more space to him than to Jesus, at Antiquities 18.116 to 119, describing a preacher of righteousness whom Herod Antipas executed because his influence over the crowds looked dangerous. To be baptised by John is to stand under John, and to receive a baptism the sources describe as being for repentance. The gospels register the difficulty and manage it, and the management is datable. Mark 1:9 reports the event plainly. Matthew 3:14 to 15 inserts a conversation in which John protests that the roles should be reversed. Luke mentions John’s imprisonment first and then refers to the baptism in a subordinate clause without saying who performed it. John’s gospel has no baptism scene at all, only the Baptist pointing him out. The trajectory runs in one direction and it starts from an event nobody would have invented.
The crucifixion. Attested by Paul, by all four gospels, by Josephus and by Tacitus. More to the point, it was the hardest fact the first preachers had to carry. Crucifixion was a servile and political punishment; Deuteronomy 21:23 pronounces a curse on anyone hanged on a tree, a passage Paul has to address directly at Galatians 3:13. Paul’s own summary of the problem is unimprovable:
But we preach Christ crucified, unto the Jews a stumblingblock, and unto the Greeks foolishness. 1 Corinthians 1:23, King James Version
Beyond those two, agreement thins out but does not vanish. E. P. Sanders set out eight facts in Jesus and Judaism (1985) that he called almost indisputable, and the list has held up as well as anything in the field: that Jesus was baptised by John; that he was a Galilean who preached and healed; that he called disciples and spoke of there being twelve; that he confined his activity to Israel; that he was involved in a controversy about the temple; that he was crucified outside Jerusalem by the Roman authorities; that after his death his followers continued as an identifiable movement; and that at least some Jews persecuted parts of that movement.
One further item commands wide assent and is regularly misreported. Historians of every persuasion accept that he had a reputation in his own lifetime as a healer and an exorcist: Meier, Sanders, Vermes, Dunn, Ehrman, Crossan and Wright all say so, from very different starting points. That is a claim about what people said of him and what he was known for. It is not a claim that anything supernatural occurred, and it is not a claim that nothing did. The evidence reaches reputation and stops.
John Meier described his own method as an unpapal conclave: a Catholic, a Protestant, a Jew and an agnostic, all competent historians, locked in the basement of the Harvard Divinity School on short rations and forbidden to emerge until they had produced a consensus document on Jesus using nothing but the methods of ancient history. The five volumes of A Marginal Jew are an attempt to write what would come out of that room. The image is useful because it makes the standard explicit: not what is true, but what four people who disagree about everything else could be compelled to sign.
There is a complication, and it is recent and serious. The criteria that produced these lists are now under sustained attack from within the field. The collection Jesus, Criteria, and the Demise of Authenticity (2012), edited by Chris Keith and Anthony Le Donne, gathered a group of specialists arguing that the criteria are a legacy of form criticism that outlived the theory it belonged to. The objections are concrete: the criteria can only ever argue for authenticity and never against it, so a saying that fails them has not been shown to be inauthentic; dissimilarity, applied strictly, guarantees a Jesus unlike both his own religion and the movement he started, which is the least likely result rather than the most; and embarrassment requires knowing what would have embarrassed people whose sensibilities are exactly what is being reconstructed. Allison, who used the criteria for years, abandoned them in Constructing Jesus (2010) in favour of looking for patterns that recur across every source and every literary form.
The bedrock has survived the demolition of the method that quarried it, which is a curious position for a discipline to be in and an honest one to report.
Galilee
He was born around 4 BCE - the traditional dating is off by a few years, since Herod the Great died in 4 BCE - probably in Nazareth, an insignificant village in Lower Galilee whose population is usually put at two or three hundred, and which the recent excavation revises upward toward a thousand. Matthew and Luke both place the birth at Bethlehem, by different and incompatible routes; historians generally regard the Bethlehem tradition as theological, satisfying a prophecy about a Davidic origin.
His father Joseph was a tektōn - a builder in wood and stone, a craftsman rather than a carpenter in the modern joinery sense. He had brothers, named in Mark as James, Joses, Judas and Simon, and unnamed sisters. He probably had some Hebrew and worked in Aramaic; whether he could read is disputed.
Galilee in the 20s CE was under Herod Antipas, a client ruler, in a countryside being reorganised by the building of two Greek-style cities, Sepphoris and Tiberias, and by the taxation that paid for them. Sepphoris was an hour’s walk from Nazareth. The economic pressure on small landholders is visible in the parables: debt, tenancy, absentee landlords, day labourers waiting to be hired.
Both cities are worth a further note, because both are absent from the gospels. Sepphoris was destroyed by the legate Varus in 4 BCE in the disturbances following Herod’s death and rebuilt by Antipas as his capital; Tiberias was founded around 19 or 20 CE on the lake shore, over a burial ground, which made it problematic for observant Jews to enter. A teacher who grew up an hour from one and worked on the shore beside the other is never once placed in either by any source. Whether that reflects the man’s own practice or the interests of the people who remembered him is not decidable, and both explanations are current.
The base of operations was Capernaum, a fishing village on the north shore of the lake, where Mark puts the first day of the ministry and where the house of Simon and Andrew serves as a fixed point in the narrative. The economy there was fishing under licence, with tolls collected on the road, which is why a tax booth appears at Mark 2:14. Crossan and Richard Horsley have read the whole ministry against the commercialisation of that economy under Antipas. Sanders cautioned against overstating it: Galilee was not exceptionally poor by the standards of the empire, and the evidence for a countryside in crisis is thinner than the social readings require.
The family is better attested than most things in the biography, because it creates a problem for later doctrine and survives anyway.
Is not this the carpenter, the son of Mary, the brother of James, and Joses, and of Juda, and Simon? and are not his sisters here with us? Mark 6:3, King James Version
Paul refers to James the Lord’s brother at Galatians 1:19 and to the brothers of the Lord at 1 Corinthians 9:5, and Josephus at Antiquities 20.200 calls James the brother of Jesus. Three later interpretations of the word developed once the perpetual virginity of Mary became doctrine: that they were sons of Joseph by an earlier marriage, the position of Epiphanius and the usual one in Eastern Christianity; that they were cousins, the position Jerome argued in Against Helvidius in 383 and the usual one in the Latin West; and that they were full brothers, the position Helvidius held and the one most historians take, on the ground that the Greek word for brother means brother and the alternatives arrive centuries later for reasons unconnected to the text.
He begins with John the Baptist, an apocalyptic preacher at the Jordan announcing imminent judgement and calling for repentance. Jesus was baptised by him - a fact the later church found awkward enough that Matthew adds a conversation explaining it and John’s gospel omits it. Then John was arrested and executed by Antipas, and Jesus began teaching on his own.
The two accounts of that execution do not match and both may be right. Mark tells the story of Herodias, the birthday feast and the head on a platter, a moral tale about a ruler trapped by an oath. Josephus gives a political reason: Antipas moved against John because a preacher able to assemble crowds is a preacher able to start something, and it was safer to act first. Rulers generally have both kinds of motive, and the sources report the kind each was interested in.
When Jesus was born, and why the date is 25 December
The year is uncertain within about a decade, and the discrepancy is inside the sources rather than between the sources and the historians.
Matthew places the birth in the reign of Herod the Great. Josephus dates Herod’s death to 4 BCE, which puts the birth at 6 to 4 BCE and makes the conventional era begin a few years after the event it counts from. That gap is the work of Dionysius Exiguus, the sixth-century monk who computed the years from the incarnation in 525 and came out short.
Luke places the birth at the census conducted by Quirinius, legate of Syria. Josephus dates that census to the deposition of Herod’s son Archelaus and the annexation of Judaea, in 6 CE. The two datings are ten years apart and cannot both stand. Most historians conclude that Luke was mistaken about the census, a conclusion Emil Schürer already regarded as forced on the evidence in the nineteenth century. Harmonising proposals exist: an earlier and otherwise unattested Quirinian command, or a reading of the Greek that makes this the census before the one Quirinius conducted. They are a minority. A further difficulty is independent of the dating: no Roman census is known to have required anyone to travel to an ancestral town, since censuses assessed property where property was.
The day has no evidential basis at all. No gospel gives one, and the earliest Christians appear not to have marked it. Clement of Alexandria, writing around 200, records that various chronologists had proposed dates, among them a day in May and days in April, and treats the exercise as curiosity rather than commemoration.
The first secure attestation of 25 December is the Chronograph of 354, a Roman almanac compiled by Furius Dionysius Philocalus, which records under that day natus Christus in Betleem Iudeae and is generally thought to reflect a Roman observance already in place by 336.
Two explanations compete, and neither has won.
The history-of-religions explanation holds that the date was taken over from the imperial festival of Sol Invictus, whose celebration on 25 December is attested in the same document. It is the older account and the one most often repeated.
The calculation explanation holds that the date was computed rather than borrowed. An ancient conviction held that a prophet died on the same day of the year he was conceived. A crucifixion dated to 25 March therefore fixes the conception at 25 March and the birth nine months later. Sextus Julius Africanus, writing around 221, appears already to be working with 25 March as the date of the conception, which is earlier than any secure evidence for a Sol Invictus festival on the twenty-fifth. Thomas Talley’s The Origins of the Liturgical Year (1986) is the standard statement of the case.
The honest summary is that the borrowing thesis is popular, the calculation thesis is currently better supported at the level of dated evidence, and neither is established.
What Jesus looked like
Nothing in any first-century source describes him. Not his height, his build, his colouring, his face or his hair. The silence is complete and it is not accidental: the texts are interested in what he said and what happened to him.
The earliest surviving images are third-century and they show a beardless young man. The house church at Dura-Europos on the Euphrates, decorated before the town fell in 256, has him short-haired and clean-shaven; the Roman catacomb paintings put him in the conventional Greco-Roman shepherd type, youthful and unbearded. The bearded, long-haired, centre-parted figure that became standard is a fourth-century development, and it borrows the visual conventions used for enthroned Zeus and Serapis and for the philosopher portrait.
Early writers who speculated were reading prophecy rather than memory. Justin, Tertullian and Origen all took Isaiah 53:2 to describe him:
he hath no form nor comeliness; and when we shall see him, there is no beauty that we should desire him. Isaiah 53:2, King James Version
Others argued the opposite from Psalm 45:2, which praises the king as fairer than the children of men. The disagreement tells us about exegesis and nothing about the face.
Two modern approaches produce the only answers the evidence supports, and both are answers about a population rather than a person. In 2001 the medical artist Richard Neave led a team that built a face from first-century Galilean skulls and forensic soft-tissue modelling. The result was widely reported as the face of Jesus; it is the face of an average Galilean man, which is what it was designed to be and all it can be. Joan Taylor’s What Did Jesus Look Like? (T&T Clark, 2018) reached the same kind of conclusion from a wider base of skeletal remains, texts and Egyptian funerary portraits: olive-brown skin, dark brown to black hair, brown eyes, hair and beard kept short in the ordinary Jewish manner of the period rather than long, and a height of around five feet five inches, which is the average for male skeletons from that place and time.
The Turin Shroud is not part of this. Radiocarbon samples tested independently at Oxford, Arizona and Zurich in 1988 returned a date range of 1260 to 1390. The result has been disputed on the ground that the sample came from a repaired section, and the dispute continues, but no historian of the first century uses the object as evidence.
What he taught
The kingdom
Everything turns on a phrase: basileia tou theou, the kingdom or reign of God. It is the subject of nearly every parable and the content of the proclamation in Mark’s opening: the time is fulfilled, and the kingdom of God is at hand.
It is not heaven, and it is not primarily about what happens after death. It is God’s rule arriving, and arriving here, into the world of Roman occupation, tax collectors, debt and the Temple hierarchy. The Lord’s Prayer asks for it to come on earth.
Albert Schweitzer’s argument in 1906 still frames the field: that Jesus was an apocalyptic prophet who expected the end of the age imminently, and that every attempt to modernise him into an ethical teacher had produced a portrait of the scholar writing it - each researcher looking down the well of history and seeing their own face. The strongest version of that case rests on sayings in which he says the present generation will not pass away before it comes.
The strongest case against it rests on the sayings where the kingdom is already present: among you, like a seed already sown, like leaven already in the dough. The most defensible reading holds both: something has begun that is not yet complete, and the tension is in the material rather than in the scholarship.
Three details of vocabulary matter and are usually skipped.
Matthew almost always writes hē basileia tōn ouranōn, the kingdom of the heavens, where Mark and Luke write the kingdom of God. The substitution is reverential, a way of not saying the divine name, and it tells us about Matthew’s Jewish readership rather than about the preaching.
The proclamation itself, in the earliest gospel, is short enough to quote whole:
The time is fulfilled, and the kingdom of God is at hand: repent ye, and believe the gospel. Mark 1:15, King James Version
And the most quoted present-tense saying rests on a Greek phrase that will carry two opposite meanings. Luke 17:21 has entos hymōn, which the King James renders the kingdom of God is within you and most modern versions render among you or in the midst of you. Interiority and presence are different claims, and the whole modern reception of the saying, from Tolstoy onward, is built on a translator’s choice that the Greek does not settle.
Against those sit the sayings that press hardest in the other direction:
Verily I say unto you, That there be some of them that stand here, which shall not taste of death, till they have seen the kingdom of God come with power. Mark 9:1, King James Version
Verily I say unto you, that this generation shall not pass, till all these things be done. Mark 13:30, King James Version
What those sayings meant is the central quarrel in the field, and it is taken up below. On the plainest reading the generation did pass. The New Testament itself registers the difficulty at 2 Peter 3, where the writer answers people asking where the promise of his coming has got to. A movement that had invented the expectation would not have needed to explain it away within two generations.
The parables
Roughly forty survive, and they are not what they are usually taken to be. A parable is not an illustration of a point that could be stated more plainly. It is a device that gets a listener to assent to something before they see what it commits them to.
The mechanism is visible in the Good Samaritan. A lawyer asks who his neighbour is. Jesus tells a story in which a man is robbed on the Jericho road and left half dead; a priest passes on the other side, then a Levite, then a Samaritan - a member of a group his listeners despised on religious grounds - who stops, treats the wounds, and pays for the man’s care. Then the question is turned around: not who is my neighbour, but which of these was a neighbour to the man? The lawyer cannot even say the word Samaritan; he says the one who showed mercy.
The same structure appears again and again. The Prodigal Son gives us a younger brother who wastes an inheritance and an elder brother who did everything right and is furious at the celebration - and the story ends with the father still outside pleading with the elder brother, and no resolution. The Workers in the Vineyard have men hired at dawn paid the same as men hired an hour before sunset, and the complaint of the early workers is entirely reasonable, and is dismissed. The Mustard Seed compares the kingdom to a weed that takes over a garden.
They are frequently unfair by ordinary standards, and the unfairness is the content.
The scholarly history of reading them is short and consequential. Until the end of the nineteenth century they were read as allegories with a decodable meaning for every element, a habit the gospels themselves already display: Mark 4:13 to 20 supplies a point-by-point allegorical key to the sower in the vocabulary of the later Christian mission, which most scholars take to be an expansion rather than part of the original. Adolf Jülicher’s Die Gleichnisreden Jesu, published between 1888 and 1899, broke the allegorical tradition by arguing that each parable makes one point and that the decoding is a later imposition. C. H. Dodd and Joachim Jeremias then set them inside the eschatological frame. Crossan and others have since read them as deliberately open, closer to poetry than to argument.
Two honest qualifications belong here.
The best-known parables are among the least well attested. The Good Samaritan appears only at Luke 10:25 to 37; the Prodigal Son only at Luke 15:11 to 32. Sole attestation in the latest of the synoptic gospels is a weakness by the standards historians normally apply, and the fame of a passage has no bearing on its provenance.
And the most careful sustained examination anyone has made of the question came out startlingly minimalist. Meier devoted the fifth volume of A Marginal Jew (2016) to the parables and concluded that only four can be judged authentic with fair probability: the Mustard Seed, the Evil Tenants, the Talents and the Great Supper. He said plainly that this was the last conclusion he had expected to reach when he began the project. He did not claim the rest were invented, only that they cannot be demonstrated, which is a different and more uncomfortable finding.
The Sermon on the Mount
Matthew gathers the ethical teaching into three chapters. Luke has much of the same material scattered and delivered on level ground, which suggests the sermon is an editorial construction - but the material inside it has strong claims to be early.
It opens with the Beatitudes, which are not instructions but announcements: blessed are the poor, the mourning, the meek, the hungry, the persecuted. Luke’s version is blunter than Matthew’s - blessed are you who are poor, not poor in spirit - and adds a matching set of woes to the rich.
Then the antitheses: you have heard that it was said… but I say to you. Each intensifies rather than relaxes the existing law: anger is placed under the prohibition on murder, and the lustful look under adultery. This is not the abolition of the law; it is an escalation of it into the interior, where no legal system can follow.
Matthew states the point directly a few verses earlier:
Think not that I am come to destroy the law, or the prophets: I am not come to destroy, but to fulfil. Matthew 5:17, King James Version
Then the instructions that Christianity has spent two thousand years finding ways around. Not to resist an evildoer, but to turn the other cheek. To love enemies and pray for persecutors. To give to everyone who asks. Not to store up treasure on earth. Not to judge. To forgive not seven times but seventy-seven.
The turning of the cheek is worth a note, because the usual reading of it as passivity may be wrong. A backhanded blow to the right cheek was how a superior struck an inferior; offering the left forces the striker either to stop or to use a fist, which was how one struck an equal. On that reading the instruction is neither retaliation nor submission but a third thing, which is harder to do than either.
That reading belongs to Walter Wink, who set it out in Engaging the Powers (1992) and built a wider account of nonviolent resistance on it, extending the same logic to the cloak and to the second mile under the Roman law of compulsory transport. Its standing is worth stating accurately, since it is often repeated as though it were established. It is not the consensus. John Nolland has objected that the sayings offer a limited set of examples precisely in order to make a point sharply, and are not a social-political programme in code. Others have argued that Wink sets up passivity as a straw man and that his third way, viewed from outside, looks identical to the second. The reading is admired more widely as constructive theology than it is accepted as exegesis, and both facts are worth having.
Who he ate with
The most historically distinctive thing about him may not be a teaching at all. It is a practice: he ate with people whom the purity system placed outside - tax collectors, who were collaborators; prostitutes; the ritually unclean. He touched lepers. He talked with women in public, and had women among his followers and funders, which the sources report without emphasis and which was not ordinary.
The charge in Matthew and Luke, quoted as if from his critics, is that he is a glutton and a drunkard, a friend of tax collectors and sinners. It has an excellent claim to authenticity: nobody invents an insult about their founder.
The King James renders the same charge as a man gluttonous, and a winebibber, a friend of publicans and sinners, at Matthew 11:19, and in slightly different word order at Luke 7:34, which is two independent attestations of a hostile description in the surviving material.
The funders are named. Luke 8:2 to 3 lists Mary called Magdalene, out of whom seven demons had gone; Joanna, the wife of Herod’s steward Chuza; and Susanna, with others, who ministered unto him of their substance. A wandering group with no visible means of support is a group with patrons, and Luke says who they were. The presence of a senior Herodian official’s wife in that list is the kind of detail an author with a tidy agenda would not construct.
There is a live disagreement about what the scandal actually was, and it changes the picture. The older reading, still the common one, takes the objection to be ritual: he ate with the impure and disregarded purity boundaries. Sanders argued that this misreads the vocabulary. The word rendered sinners in the gospels translates a Hebrew and Aramaic term for the genuinely wicked, the people who had put themselves outside the covenant by how they lived, rather than for those in a temporary state of ritual impurity, which was ordinary and remediable and affected everyone. On Sanders’s reading the offence is sharper rather than milder: not that he was careless about ritual, but that he offered a place in the coming kingdom to people who had not repented and made restitution, which is what the law required of them. The dispute is unresolved and it is worth knowing which version is being told.
Whether Jesus was an apocalyptic prophet, and why it is the central quarrel
Every reconstruction of Jesus turns on this question, and it is the clearest example on the roster of a division that scholarship has not closed.
The position, in its classic form, is Albert Schweitzer’s. In The Quest of the Historical Jesus (1906) he argued for what he called thoroughgoing eschatology: that Jesus shared the apocalyptic expectation of his time, believed the kingdom would arrive within the lifetime of his hearers, sent his disciples out expecting it to break in before they had finished their circuit, and finally went to Jerusalem to force the issue. Schweitzer’s Jesus is not a Victorian moral teacher but a first-century Jewish prophet whose central expectation was disappointed, and the book closed the nineteenth-century project by pointing out that every liberal life of Jesus had discovered a liberal.
The arguments now advanced for the apocalyptic reading are these.
He began with John the Baptist, who was unambiguously an apocalyptic preacher of imminent judgement. The movement that followed him was unambiguously apocalyptic: Paul, writing in the early 50s at 1 Thessalonians 4, expects to be alive at the coming. A non-apocalyptic Jesus has to be inserted between two apocalyptic termini and then explained. Ehrman treats this as the strongest single consideration.
The Twelve presuppose a restored Israel of twelve tribes, which is an eschatological expectation and not an administrative arrangement. The saying about the twelve thrones survives at Matthew 19:28 and Luke 22:30 despite the obvious embarrassment of promising one of them to Judas.
The sayings about a near end are present in Mark, in the double tradition, and in Paul, and they are the kind of material a later movement would prune rather than plant.
But when they persecute you in this city, flee ye into another: for verily I say unto you, Ye shall not have gone over the cities of Israel, till the Son of man be come. Matthew 10:23, King James Version
Those holding the position, with variations: Sanders, Ehrman, Allison, Fredriksen, Vermes, Casey. Meier holds a qualified version in which the kingdom is both present and imminent. Wright holds it while redefining the terms, discussed below.
The case against was made most visibly by the Jesus Seminar and by Marcus Borg, John Dominic Crossan, Burton Mack and Stephen Patterson. The four-cornered debate was published as The Apocalyptic Jesus: A Debate (2001), with Allison defending the apocalyptic reading against the other three. Their arguments:
The earliest reconstructed layer of Q is wisdom material rather than apocalyptic, on Kloppenborg’s stratification, which would make the apocalyptic frame a secondary development.
The Gospel of Thomas has sayings without any apocalyptic frame at all, and if it is early and independent that matters.
There are present-tense kingdom sayings that sit badly with an exclusively future expectation. Luke 11:20 is the clearest: if I with the finger of God cast out devils, no doubt the kingdom of God is come upon you.
And the apocalyptic material may be the church’s own, retrojected after Easter and hardened by the catastrophe of 70.
Two honest qualifications belong on this side of the ledger. Kloppenborg has objected repeatedly that his stratification of Q is a hypothesis about how a document was composed and not a chronology of a man’s teaching, and that using it as a historical-Jesus argument stacks one conjecture on another. And the early dating of Thomas, on which a good deal rests, is itself among the most contested questions in the field.
Allison’s reply in Constructing Jesus is the strongest counter and it is structural rather than piecemeal. The apocalyptic material is not a removable layer. It appears in Mark, in the double tradition, in material peculiar to Matthew, in material peculiar to Luke, in Paul, in sayings, in parables and in narrative. A reconstruction that strips it out has to strip out so much that what remains is too thin to have produced anything.
N. T. Wright makes a different move in Jesus and the Victory of God (1996). He accepts that the language is apocalyptic and denies that apocalyptic language in Second Temple Judaism meant the physical end of the world. It was, he argues, a metaphorical register for describing concrete political and covenantal events, and reading it literally is a modern error. On his account Jesus expected the end of exile and the vindication of Israel, both of which he takes to have occurred in a sense. Allison and Ehrman both reject the argument, holding that first-century apocalyptic texts mean what they appear to mean.
What hangs on it is not small. If the apocalyptic reading is right, the ethics of the Sermon on the Mount were framed for a short interval before an intervention that did not come, and the founding expectation of the movement failed. If the sapiential reading is right, the ethics are a standing programme for living and the apocalyptic frame is somebody else’s addition. The majority of specialists hold the apocalyptic reading. The minority is a serious minority with serious arguments. This page reports the division and does not settle it.
The Lord’s Prayer, and the word in it nobody can translate
Two versions survive. Matthew 6:9 to 13 is longer, Luke 11:2 to 4 shorter, and the usual explanation is that Matthew’s is the liturgical form of his community. The King James text of Matthew reads:
Our Father which art in heaven, Hallowed be thy name. Thy kingdom come. Thy will be done in earth, as it is in heaven. Give us this day our daily bread. And forgive us our debts, as we forgive our debtors. And lead us not into temptation, but deliver us from evil: For thine is the kingdom, and the power, and the glory, for ever. Amen. Matthew 6:9–13, King James Version
The closing doxology is not in the earliest manuscripts of Matthew. It is a liturgical ending that entered the text through the Byzantine tradition, which is what the King James translators were working from; a form of it appears in the Didache at 8:2, an early Christian church manual. Most modern editions relegate it to a footnote, which is why congregations reciting different translations arrive at the end at different times.
Three petitions repay attention.
The bread. Epiousios, the word translated daily, occurs nowhere else in Greek literature. It is not merely rare; outside these two verses and the Didache it is absent from the entire surviving corpus. Origen, writing in the third century in On Prayer, says he searched and could not find it in any Greek writer, in common speech or in philosophy, and concluded that the evangelists had coined it. Jerome translated it two different ways in the same Bible: supersubstantialem at Matthew 6:11 and quotidianum at Luke 11:3. A supposed documentary occurrence in a fifth-century Egyptian shopping list was re-examined at the Beinecke Library in 1998 and turned out to read elaiou, oil.
The candidates are: for the coming day; necessary for existence; and supersubstantial, meaning the bread of the age to come. A majority of scholars now favour the first, which places the petition inside the apocalyptic frame: bread for tomorrow, asked for today, because tomorrow is the day the kingdom arrives.
The debts. Matthew uses opheilēmata, a commercial word for what is owed, and asks for release from debts on the same terms the petitioner grants release. Luke has sins in the first clause and everyone that is indebted to us in the second, which reads like an explanation of a Semitic idiom for a Greek audience. Whether the original petition was about money or about sin is exactly the question, and the social readings of Galilee take the commercial sense literally.
The temptation. Peirasmos means a trial or a testing rather than an enticement to do wrong, and the petition asks not to be brought to the test. The theological difficulty of a God who leads anyone into temptation is old, and in 2018 the Italian bishops approved a revised rendering that asks instead not to be abandoned to temptation, brought into use in 2020. That is a translation decision, not a textual one: the Greek did not change.
What Christianity has done with the Sermon on the Mount
The interpretive history is not a footnote to the text; it is the clearest evidence available of how hard the text is.
Taken literally and universally, Matthew 5 to 7 dissolves law courts, oaths, accumulated property, armed force and self-defence. Almost every Christian tradition has therefore developed an account of the text’s scope, and the accounts can be named.
Two levels. Patristic and medieval exegesis distinguished precepts, binding on everyone, from counsels of perfection, binding on those pursuing the higher life. In practice that meant the monastery kept the sermon and the world kept the precepts. Aquinas systematised the distinction.
Two kingdoms. Luther argued that a Christian occupies two spheres at once, and that the sermon governs a person and not the office they hold. The same man may not retaliate in his own cause and must punish in the magistrate’s. The device preserves the text intact and removes it from politics.
The impossible ideal. Also Luther, and hardened by later Lutheran orthodoxy: the demands are pitched deliberately beyond human capacity, in order to destroy the hearer’s confidence in their own righteousness and drive them toward grace. On this reading the sermon is not a programme but a mirror.
The interim ethic. Johannes Weiss and then Schweitzer held that the instructions were framed for the short interval before the end of the age, and were never intended as a standing social arrangement. This is the apocalyptic reading’s consequence for ethics, and it is why the two questions cannot be separated.
Literally and universally. The Anabaptists of the sixteenth century, the Quakers, and Tolstoy. The Kingdom of God Is Within You, published in 1894 after being banned in Russia, argues that the churches had spent eighteen centuries constructing exemptions, and takes the prohibition on resisting evil as a plain prohibition of state violence, courts, oaths and military service. The book reached a young lawyer in South Africa; Gandhi named it as one of the decisive influences on his life, and the line of descent from it runs through the Indian independence movement to the American civil rights movement.
A future dispensation. Nineteenth-century dispensationalism, and the Scofield Reference Bible that carried it, placed the sermon in the law of a future millennial kingdom rather than the present age, which removes the obligation entirely.
Refusal of all the devices. Dietrich Bonhoeffer’s Nachfolge (1937), known in English as The Cost of Discipleship, rejects every softening move as a way of hearing the words without doing them. The book’s argument is that the sermon means what it says and costs what it costs. Its author joined a conspiracy to kill Hitler and was hanged at Flossenbürg on 9 April 1945, which is the sharpest illustration available of how the problem behaves under pressure, including for people who have thought hardest about it.
The historical observation underneath all of this is simple and worth stating flatly. The text does not indicate which of these it is. It contains no instruction about its own scope, no marker separating counsel from command, and no clause about circumstances. Whatever else it is, it is a document engineered to remain an argument.
What Jesus said about wealth, and the eye of the needle
The sayings on wealth are among the best attested in the tradition, appearing in Mark, in the double tradition, in material peculiar to Luke and in Thomas, and they are consistently harder than their reception.
The centrepiece is the encounter at Mark 10:17 to 27. A man who has kept the commandments is told to sell what he has and give it to the poor, and goes away grieved. Then:
It is easier for a camel to go through the eye of a needle, than for a rich man to enter into the kingdom of God. Mark 10:25, King James Version
There was no gate in Jerusalem called the Needle’s Eye. The claim is medieval and its route has been traced. The earliest form of it appears as a gloss attributed to Anselm of Canterbury, preserved in the thirteenth-century Catena Aurea of Thomas Aquinas. Theophylact of Ohrid, the eleventh-century commentator usually blamed for it, says no such thing: he records only that some read the Greek word as kamilos, a ship’s cable, rather than kamēlos, camel. The misattribution appears to have been spread by a marginal note in the sixteenth-century Geneva Bible. Agnieszka Ziemińska set out the chain in New Testament Studies in 2022.
The internal evidence points the same way. If the saying described a low gate that a camel could squeeze through on its knees, the hearers would not have reacted as they do:
And they were astonished out of measure, saying among themselves, Who then can be saved? Mark 10:26, King James Version
The cable reading is a variant in some manuscripts and is a softening of the same kind, a few centuries earlier.
Luke is harsher throughout than Matthew or Mark. His Beatitudes bless the poor without qualification and add woes to the rich. He alone has the rich man and Lazarus, and the instruction to sell possessions and give alms. Whether Luke sharpened an existing emphasis or preserved one that Matthew softened is disputed, and both are argued.
Two things frequently attributed to Jesus on this subject are not his.
Money is the root of all evil is not a saying of Jesus and, in that form, is not a quotation of anything. The verse is 1 Timothy 6:10, from a letter most scholars regard as written in Paul’s name after his death, and it reads the love of money is the root of all evil. The two versions license entirely different conclusions.
God helps those who help themselves is not in the Bible at all. Its familiar English form comes through Algernon Sidney and then Benjamin Franklin’s Poor Richard’s Almanack, and its sentiment is close to the reverse of the passages above.
One counterweight inside Luke is worth naming, because it is the text every argument for a moderate reading has to use. Zacchaeus, at Luke 19:1 to 10, is a rich tax collector who gives half his goods to the poor and repays fourfold what he has defrauded, and is told that salvation has come to his house. He is not told to give everything. Whether Luke intends that as a limit on the demand or as an example of what the demand produces has been argued both ways for a very long time.
What Jesus said about hell, and what Gehenna actually was
The English word hell does work that three different Greek words do in the sources, and collapsing them has consequences.
Hadēs is the realm of the dead, the word the Greek Old Testament uses to render the Hebrew Sheol. It is a location for everyone who has died, not a place of punishment for some. Tartaroō occurs once, at 2 Peter 2:4, and is borrowed from Greek mythology. Gehenna is the word on Jesus’s lips in almost every passage that has been read as teaching about hell, and it is not a mythological term at all. It is a place name.
Gehenna transliterates the Hebrew gê-hinnōm, the Valley of Hinnom, a ravine running along the south and west of Jerusalem. In 2 Kings 23:10 and Jeremiah 7:31 it is where children were burned as offerings to Molech, and Josiah defiles the site to end the practice; Jeremiah 7:32 announces that it will be renamed the valley of slaughter. By the Second Temple period the name had become an image for the place of final judgement, and it functions that way in 1 Enoch and in later rabbinic literature.
The explanation almost universally offered, that Gehenna was the city’s perpetually burning rubbish dump and that the imagery is drawn from smouldering refuse, has one source. It appears in a commentary on Psalm 27 written by Rabbi David Kimhi in the thirteenth century, more than a thousand years after the period it describes. There is no supporting literary evidence from the intertestamental or rabbinic periods and no archaeological evidence of any kind, as Lloyd Bailey set out in the Biblical Archaeologist in 1986. Kimhi himself offered it as an analogy.
The word occurs twelve times in the New Testament, eleven of them attributed to Jesus and concentrated in Matthew, with one occurrence in the letter of James. The most vivid passage quotes Isaiah:
And if thine eye offend thee, pluck it out: it is better for thee to enter into the kingdom of God with one eye, than having two eyes to be cast into hell fire: Where their worm dieth not, and the fire is not quenched. Mark 9:47–48, King James Version
The closing line is Isaiah 66:24, describing corpses outside Jerusalem, and Mark reproduces the Greek of it almost word for word, shifting the tenses into the present.
What the sayings do and do not settle is a matter of long dispute inside Christianity, and the dispute is largely about one adjective. Matthew 25:46 has these shall go away into everlasting punishment: but the righteous into life eternal, and the Greek word rendered both times is aiōnios, which can mean everlasting or can mean belonging to the age to come. Three positions have contended for the whole history of the church, each citing the same verses: eternal conscious punishment; annihilation, in which the condemned cease to exist; and universal reconciliation, held by Origen under the name apokatastasis and rejected by Augustine in book 21 of the City of God.
The historian’s observation is narrower than any of them. The imagery is inherited Jewish apocalyptic language, deployed with the intensity characteristic of that literature, in which extreme statements about judgement are a genre convention rather than a systematic doctrine. The developed doctrines of hell are later, and the fathers who framed them were already disagreeing with each other by the third century.
Whether Jesus claimed to be God
The answer depends entirely on which gospel is being read, and that fact is the substance of the question.
In Mark, Matthew and Luke there is no passage in which he identifies himself as God. There are passages that point the other way. At Mark 10:18, answering a man who calls him good master, he replies:
Why callest thou me good? there is none good but one, that is, God. Mark 10:18, King James Version
Matthew, writing later and using Mark, rewrites the exchange so that the question concerns what is good rather than who is, which is precisely the kind of editorial correction that indicates the earlier text was felt as a problem.
At Mark 13:32, on the timing of the end:
But of that day and that hour knoweth no man, no, not the angels which are in heaven, neither the Son, but the Father. Mark 13:32, King James Version
When Peter identifies him as the Christ at Mark 8:29, the response is a command to tell no one. The pattern of enjoined silence runs through Mark and is one of the oldest puzzles in the study of the gospel, named the messianic secret by William Wrede in 1901.
The one direct exchange, at the hearing before the high priest, shows the later evangelists softening as they copy. Mark 14:62 has him answer I am. Matthew 26:64 changes this to Thou hast said. Luke 22:70 changes it again to Ye say that I am. Editing in the direction of ambiguity is not what a tradition inventing a claim would produce.
Son of Man is the phrase he uses of himself more than any other in the synoptics, and almost nobody else in the New Testament uses it of him. The sayings fall into three groups: present activity, coming suffering, and future coming in glory, the last of which draws on the figure in Daniel 7:13. What the phrase meant is disputed. Vermes argued that Aramaic bar (e)nasha was an ordinary circumlocution meaning something like a man in my position, and therefore carried no titular weight; Casey developed the argument at length. Hurtado and others reject it, on the ground that an ordinary idiom does not explain the peculiar Greek double article that every gospel uses, and that the phrase is never applied to him as a title by anyone else, which is odd if the church created it.
In John the identity is stated outright and repeatedly.
Verily, verily, I say unto you, Before Abraham was, I am. John 8:58, King James Version
I and my Father are one. John 10:30, King James Version
The gospel opens with In the beginning was the Word, and the Word was with God, and the Word was God, and in the narrative the hearers respond by picking up stones, which is the author’s way of telling a reader what has just been claimed.
The near-unanimous historical judgement is that the Johannine discourses are the evangelist’s composition and not a record of speech: they are in a single style, that style is the style of the narrator, and the content has no parallel in the three earlier gospels. That is a statement about a document. It is not a statement about whether the claim is true, and nothing in the historical method bears on that.
What historians argue about instead is when devotion to Jesus as divine began, and here the division is sharp and the positions are named. Ehrman’s How Jesus Became God (2014) traces a development from an exaltation christology, in which a man was raised to divine status, toward an incarnation christology in which a divine being descended. Larry Hurtado, in Lord Jesus Christ (2003), and Richard Bauckham argue that worship directed to Jesus alongside God is already present in the earliest Pauline material, within two decades of the crucifixion, and that it is therefore a Jewish development and not a slow Hellenistic drift. James Dunn’s Christology in the Making (1980) had argued that incarnation proper is not present before John. The area of agreement is larger than the polemic suggests: Ehrman and Hurtado both hold that very high claims appear very early, and disagree about their shape and their source.
The pre-Pauline hymn at Philippians 2:6 to 11 is the passage every side has to handle, and the argument turns on a single Greek word, harpagmon, which will support either a being who did not grasp at equality he already had or one who did not seize equality he lacked.
One frequently repeated claim can be closed. The Council of Nicaea in 325 did not invent the divinity of Jesus, and the idea that it did, popularised by The Da Vinci Code, is rejected across the field, by sceptics and confessional scholars alike. Nicaea debated whether the Son was of the same substance as the Father or of a like substance, a dispute between parties who both held him to be divine in some sense. The vote was not close and the question on the table was not the one the popular version supposes.
Whether Jesus was married
No source of any kind says that he was.
The argument from silence cuts in an unusual way here, because it is genuinely two-edged. Most Jewish men of the period married, and celibacy was unusual enough that some scholars think it needs an explanation rather than being the default assumption. The gospels supply a possible one at Matthew 19:12, with the saying about those who have made themselves eunuchs for the sake of the kingdom, and a comparable pattern of celibacy is attested among some at Qumran.
Against a marriage, the decisive consideration is Paul. Arguing at 1 Corinthians 9:5 for his own right to be accompanied on his travels, he writes:
Have we not power to lead about a sister, a wife, as well as other apostles, and as the brethren of the Lord, and Cephas? 1 Corinthians 9:5, King James Version
Paul cites Peter and the brothers of Jesus as precedents. He does not cite Jesus, in a passage where citing Jesus would have ended the argument at once.
Mary Magdalene is named in all four gospels, appears at the crucifixion and at the tomb in every account, and is never described as a prostitute anywhere in the New Testament. Luke 8:2 says seven demons had gone out of her and nothing more. The identification of her with the anonymous sinful woman of Luke 7 and with Mary of Bethany was made by Gregory the Great in a homily preached in 591 and became standard in the Latin West; the Eastern churches never adopted it, and the Roman Catholic calendar separated the figures again in 1969.
The marriage claim in the popular literature rests on two documents.
The Gospel of Philip, a third-century Coptic text from Nag Hammadi, calls Mary Magdalene his koinōnos, companion or partner, and contains a damaged sentence saying that he used to kiss her often on the [gap]. The manuscript has a hole exactly where the noun would be. Every reconstruction of that word is a guess, and the text is in any case two centuries later than the events and belongs to a tradition that used the language of union symbolically.
The Gospel of Jesus’s Wife was announced by Karen King of Harvard in 2012: a business-card-sized papyrus fragment in Coptic containing the words Jesus said to them, my wife. King published on it in the Harvard Theological Review in 2014. In 2016 Ariel Sabar traced the fragment’s owner for The Atlantic to a man in Florida named Walter Fritz, and the provenance story collapsed. King then said the new information tipped the balance toward forgery. Sabar’s account of the whole episode is Veritas (2020), and the fragment is now universally regarded as a modern fake.
The historian’s conclusion is dull and should be stated as such: unmarried is the more likely reading of the evidence, it is not proven, and nothing else in the reconstruction depends on it.
The last week
He went up to Jerusalem for Passover, when the population multiplied and the Roman garrison reinforced. He entered the city in a way the gospels present as a deliberate enactment of a prophecy about a king coming on a donkey.
Then the Temple. He overturned the tables of the money-changers and the seats of those selling doves. Whatever its scale - a symbolic act in one corner, or something larger - it was an action against the economic operation of the central institution of Judaism, taken at the most crowded moment of the year, in a city where the Roman prefect had come up from Caesarea specifically because Passover was when trouble happened.
He was arrested at night, questioned by the high priest Caiaphas, and handed to Pilate. The gospels increasingly shift responsibility from the Romans to the Jewish authorities as they get later, which correlates suspiciously well with a movement trying to survive within the empire and to distinguish itself from a Judaism that had just fought a catastrophic war. The historical facts are that crucifixion was a Roman punishment administered by Roman soldiers on Roman authority, reserved for sedition and for slaves, and that the placard read King of the Jews.
He was crucified outside the walls, probably in 30 or 33 CE, and died within hours - fast, for crucifixion. Mark, the earliest account, gives him one sentence from the cross, and it is a quotation of Psalm 22: My God, my God, why have you forsaken me? The later gospels add other last words. That Mark’s version is the harshest, and the earliest, is the kind of detail on which historical judgement turns.
The empty tomb and the resurrection appearances are where history stops being able to adjudicate. What can be said historically is that within a very short time his followers - who had scattered - were proclaiming in Jerusalem that he had been raised, and that a number of them were killed for continuing to say so.
Several details of that week repay a closer look.
The chronology does not agree with itself. The three synoptic gospels present the last supper as a Passover meal, with the crucifixion on the day of Passover itself. John places the crucifixion on the day of Preparation, before the Passover meal, so that Jesus dies at the hour the lambs are slaughtered. The two cannot both be right. Many historians prefer John’s chronology on the practical ground that a hearing and an execution conducted on the festival day are improbable; others hold that John has adjusted the calendar for theological effect, which is demonstrably something John does. The question is open.
The year narrows to two candidates. A crucifixion on a Friday, on 14 or 15 Nisan, during Pilate’s prefecture, is astronomically constrained. Colin Humphreys and W. G. Waddington set out the calculation in Nature in 1983, and the surviving possibilities are 7 April 30 CE and 3 April 33 CE. Both are defended and neither can be eliminated.
Pilate. Prefect of Judaea from 26 to 36 CE. The inscription found at Caesarea Maritima in 1961, reused as a step in a theatre staircase, gives his title as praefectus Iudaeae, which is what corrects Tacitus. Philo, in the Embassy to Gaius, describes him as inflexible and merciless, and Josephus records incidents of provocation and massacre. That figure is difficult to reconcile with the reluctant administrator of the gospel trial scenes, and the reluctance increases as the gospels get later, reaching its extreme in Matthew, where Pilate washes his hands and the crowd answers:
Then answered all the people, and said, His blood be on us, and on our children. Matthew 27:25, King James Version
The verse is unique to Matthew. It has no parallel in Mark, in Luke or in John, which makes it evidence about the community Matthew wrote for and its quarrel with the synagogue, rather than testimony about a crowd. It was nonetheless read for centuries as a self-imposed collective guilt, and used to justify violence against Jewish communities across Europe. The Second Vatican Council repudiated the collective charge in Nostra Aetate in 1965. That is a fact about the reception of a text and not about the first century, and the two are worth keeping apart.
Barabbas. There is no evidence outside the gospels for a Passover custom of releasing a prisoner, and no parallel in Roman provincial practice. Most historians regard the episode as unhistorical, and the scene as it stands performs the same transfer of responsibility.
The execution itself. The men crucified alongside him are called lēstai, a word meaning brigands, which Josephus routinely uses of insurgents. Crucifixion in Judaea was not rare: Varus crucified two thousand after the disturbances of 4 BCE, and Josephus reports that during the siege of Jerusalem the soldiers ran out of wood and space for crosses. Only one crucified skeleton has ever been recovered, from a tomb at Giv’at ha-Mivtar north of Jerusalem excavated in 1968: a man named on his ossuary as Jehohanan, with an iron nail eleven and a half centimetres long still fixed through his right heel bone, bent where it had struck a knot in the wood so that it could not be pulled out. Joseph Zias and Eliezer Sekeles re-examined the remains in 1985 and concluded that the two heels had been nailed separately, on either side of the upright.
One of the sayings from the cross is textually insecure. The prayer at Luke 23:34, Father, forgive them; for they know not what they do, is absent from the third-century Bodmer papyrus of Luke, from Codex Vaticanus, from the first hand of Codex Bezae and from the Sahidic version, and present in Sinaiticus and the great majority of later witnesses. Ehrman has argued that it is original to Luke and was removed by second-century scribes unwilling to have Jesus pray for the forgiveness of Jews. Others hold it to be an early addition. Either way, the most quoted line of Christian forgiveness rests on a divided manuscript tradition.
The burial is disputed. Mark has Joseph of Arimathea, a member of the council, ask for the body and place it in a rock-cut tomb. Crossan argued that the ordinary Roman practice was to leave the crucified on the cross or throw them into a common pit, that the disciples had fled and would not have known what happened, and that the burial narrative is a later construction. Against him: Jewish law required a body taken down before nightfall, Josephus states in the Jewish War that Jews did take down and bury the crucified before sunset, and the Jehohanan ossuary is direct evidence that at least one crucified man in first-century Jerusalem received a family burial. Byron McCane proposed a third position in 1998: a real burial, but a dishonourable one in a plot reserved for the executed, which fits both the law and the shame and would explain why the tradition needed a named benefactor.
Why Jesus was executed, and what the charge on the cross says
Crucifixion was not a general-purpose penalty. Romans used it on slaves, pirates, rebels and provincial subjects, and its purpose was public deterrence rather than punishment in any narrow sense. Roman citizens were exempt except in cases of treason. The single most important consequence of that fact is that a crucified man was, by the act itself, being declared a political danger.
Then there is the notice fixed above him:
And the superscription of his accusation was written over, THE KING OF THE JEWS. Mark 15:26, King James Version
This is among the strongest data in the whole reconstruction. All four gospels preserve it, with small variations. No Christian text uses the phrase as a confession of faith; the titles the movement adopted were Christ, Lord and Son of God, not king of the Jews. A movement inventing a charge would not invent one it never afterward used, and one that reads as a Roman sneer. Whatever it means, it comes from outside the tradition that transmitted it.
What it does not settle is why. The competing reconstructions are worth setting out with their proponents, because the disagreement here is genuine and productive.
Sanders put the temple at the centre. The overturning of the tables was a prophetic sign act announcing the destruction of the existing temple and the arrival of a new one in the coming kingdom, and combined with any implied claim to a role in that kingdom it gave the temple authorities both a motive and a case. The chief priests, whose position depended on the temple and on Roman tolerance, acted, and Pilate did what a prefect does at Passover.
Fredriksen, in Jesus of Nazareth, King of the Jews (1999), builds her whole reconstruction on the detail nobody else foregrounds: Pilate executed Jesus as an insurrectionist and arrested none of his followers. If Rome had thought it was facing a movement, it would have taken the movement. She doubts the temple incident happened as described, on the ground that a disturbance in the temple court at Passover, with troops stationed in the adjacent fortress precisely against that possibility, would have produced an immediate arrest rather than several more days of public teaching. Her answer is the crowd: pilgrims acclaiming a figure as king, in a city at its most volatile, in front of a prefect whose job that week was to prevent exactly that. Pilate killed the man the crowd was shouting about, publicly, and left the followers alone, because he knew the acclamation was the danger and the man was not.
Crossan reads him as a peasant social revolutionary whose execution was one among many, and takes much of the trial narrative to be prophecy historicised, composed from the Psalms rather than from memory.
Wright holds that the messianic claim was real and openly made, and that the execution follows directly and unmysteriously from it.
Ehrman argues that the operative charge was kingship: he was executed for proclaiming a coming kingdom in which he expected to rule, a reading for which he takes the saying about the twelve thrones as decisive, since it survives despite promising a throne to the man who betrayed him.
One ancient parallel throws the whole question into relief. Josephus records, in the Jewish War at 6.300 to 309, that around 62 CE a peasant named Jesus son of Ananias began crying out at a feast against the city and the sanctuary, would answer no questions, was beaten by the leading citizens, was handed over to the Roman procurator Albinus, was flogged to the bone, was judged to be mad, and was released. He kept it up for more than seven years and died in the siege. The parallel establishes two things at once. A man denouncing the temple at a festival was a recognised category of nuisance and did provoke exactly this chain of custody. And the normal outcome was a flogging. What needs explaining about Jesus of Nazareth is not the arrest; it is the cross.
One further matter is worth stating plainly, because the alternative is to leave it implied. The sentence was Roman, the method was Roman, and the soldiers were Roman. The gospels transfer weight from Pilate to the Jewish authorities progressively across the four accounts, and the earlier a source is, the less of that transfer it contains. The verse at Matthew 27:25, discussed above, is the end point of that movement and not a report from the beginning of it.
Whether Jesus said “let him who is without sin cast the first stone”
The story of the woman taken in adultery stands at John 7:53 to 8:11 in most printed Bibles, usually inside brackets or with a footnote. It contains one of the most famous sentences in the language:
He that is without sin among you, let him first cast a stone at her. John 8:7, King James Version
It is almost certainly not part of the Gospel of John, and the manuscript evidence is unusually clear.
The passage is absent from the two earliest substantial papyri of John, from Codex Sinaiticus and Codex Vaticanus, and from the earliest Syriac, Coptic and Armenian versions. The Greek fathers who comment on John verse by verse, Origen among them and later Chrysostom and Cyril of Alexandria, pass from 7:52 to 8:12 without noticing a gap. It first appears in a Greek manuscript in Codex Bezae in the fifth century, and it is in the Latin tradition earlier than that. Later scribes who copied it frequently marked it with obeli, the symbols used to flag doubtful text, which shows they were transmitting something they knew was contested. Most tellingly, the passage floats: manuscripts place it after John 7:36, after John 21:25, and in one family of manuscripts after Luke 21:38. Text that moves around is text that was inserted.
Jerome knew it and said it stood in many Greek and Latin copies. Augustine suggested it had been deleted by men who feared it made adultery look forgivable. Didymus the Blind knew a version of the story. Its Greek vocabulary and narrative manner are closer to Luke than to John.
Two conclusions follow, and they are different conclusions. The first, on which the field is effectively unanimous, is that the passage was not in the Gospel of John as written. The second, which many scholars accept and which does not follow from the first, is that the story itself may nonetheless be old. It is exactly the kind of free-floating tradition that oral transmission preserves and canonical editing loses: nobody would gain by inventing an episode in which Jesus declines to enforce the law of Moses, and later Christian writers found it awkward enough that Augustine had to defend its retention.
The same textual situation applies to the other famous passage most readers do not know is disputed. Mark 16:9 to 20, the longer ending, is absent from Sinaiticus and Vaticanus and from the earliest Syriac, Sahidic, Armenian and Georgian versions; its vocabulary is unlike Mark’s throughout; and several manuscripts carry alternative shorter endings instead, which is what happens when copyists find a text that stops abruptly and try to fix it. On the best evidence, the earliest gospel ends here:
And they went out quickly, and fled from the sepulchre; for they trembled and were amazed: neither said they any thing to any man; for they were afraid. Mark 16:8, King James Version
An empty tomb, three frightened women, no appearance, and silence.
The resurrection of Jesus, and what history cannot reach
This is the point at which the discipline reaches its own boundary, and the useful thing a page like this can do is describe the boundary accurately rather than pretend to stand on either side of it.
What the sources say, and how they differ. Paul’s list at 1 Corinthians 15:3 to 8 names appearances to Cephas, to the twelve, to more than five hundred at once, to James, to all the apostles, and last to Paul himself. It mentions no empty tomb and no woman. Mark, as best attested, ends with an empty tomb and no appearance at all. Matthew has appearances in Jerusalem and in Galilee. Luke keeps everything in and around Jerusalem and has the ascension on the same day, then places it forty days later at the opening of Acts. John has Mary Magdalene at the tomb, then the disciples, then Thomas, then a Galilean epilogue. The accounts agree that the tomb was found empty by women and that appearances followed, and agree about almost nothing else in sequence or geography.
What is early. The formula Paul quotes is introduced with the language of received tradition, is structured as a set piece, and uses the Aramaic-derived name Cephas rather than Peter. Most scholars, across the confessional range, date it to within a few years of the crucifixion, which makes it the earliest datable Christian statement about anything.
What historians of every persuasion accept. That people who had scattered were, within a short time, publicly claiming in Jerusalem that they had seen him. That Paul, who had been persecuting the movement, joined it on the strength of something he describes as a seeing. That some of them accepted death rather than withdraw the claim. These are facts about what people believed and what it cost them, and they are not in dispute.
What is disputed and can be argued historically. Whether the tomb was empty, and if so why. Whether Joseph of Arimathea is a historical figure. Whether the appearances derive from group experience, from individual visionary experience, or from later apologetic composition. Whether the divergences between the accounts are the ordinary noise of independent testimony or the signature of construction.
What cannot be reached, and why. Historical method works by comparing candidate explanations on how probable each is given everything else known. An explanation invoking an event with no analogue anywhere else in the record will always rank last on that comparison, however strong the testimony, because the method is built to prefer the explanation that requires the fewest departures from the known. Ehrman states this explicitly as a limitation of the discipline rather than as a verdict on the event: a historian, as a historian, cannot affirm a miracle, and that is a statement about the tool. N. T. Wright’s reply in The Resurrection of the Son of God (2003) is that a method which excludes a conclusion before examining the evidence is not neutral but question-begging, and he argues at length that bodily resurrection is the best available explanation of both the empty tomb and the appearances. Gerd Lüdemann and Crossan argue the other way, treating the appearances as visionary experiences of a familiar human kind.
Where the most careful recent treatment ends up. Dale Allison spent four hundred pages on the question in The Resurrection of Jesus (2021), including a long comparison with the modern literature on apparitions of the dead and on bereavement experiences, a body of evidence almost nobody on either side had taken seriously. His conclusion is that the purely historical evidence is not good enough to make disbelief unreasonable and not bad enough to make faith untenable. He declines to decide. Coming from the scholar who has done most to dismantle the criteria that other historians rely on, that refusal has been more widely quoted than any positive finding in the book.
The statement of the limit is the finding, and it points in neither direction. History can establish that a group of people believed something with enough conviction to die for it, and can describe with some precision when and where and at what cost. Whether what they believed occurred is a question the method is not constructed to answer. Saying so is not a diplomatic evasion of a difficult subject. It is a description of what the instrument measures.
Where to start reading about the historical Jesus
The field’s disagreements are invisible to anyone reading only one book, because every one of them is written with the confidence of a settled case. The following is the order in which the arguments become legible.
Sanders, The Historical Figure of Jesus (1993). The entry point. Short, plainly written, sceptical without relish, by the scholar whose Jesus and Judaism (1985) reset the field. Anyone reading only one book should read this one.
Vermes, Jesus the Jew (1973). The book that made it impossible to discuss him outside first-century Judaism, and which reads him against the Galilean charismatic holy men of the rabbinic sources. Its importance is partly historical: Susannah Heschel’s The Aryan Jesus (2008) documents what the discipline had looked like when it tried to detach him from Judaism, which is why the Jewishness of Jesus is not a neutral scholarly preference.
Fredriksen, Jesus of Nazareth, King of the Jews (1999). For Jerusalem, for Rome, and for the sharpest statement of the puzzle about why the followers survived.
Ehrman, Jesus: Apocalyptic Prophet of the New Millennium (1999) for the apocalyptic case at general-reader length, and Did Jesus Exist? (2012) for the mythicist question handled at length by someone with no confessional stake in the answer.
Allison, Constructing Jesus (2010) for the methodological crisis, and The Resurrection of Jesus (2021) for the most careful treatment of the hardest question. Allison is the most useful writer in the field for a reader who wants to watch a scholar refuse to overstate.
Meier, A Marginal Jew (five volumes, 1991 to 2016). The reference work, not a first read. Volume 2 covers the miracle traditions, volume 5 the parables, and the whole is written to the standard of the unpapal conclave.
Crossan, The Historical Jesus (1991) for the social and economic reading. Widely contested and worth reading for what it notices about Galilee that the theological accounts pass over.
Wright, Jesus and the Victory of God (1996) as the strongest confessional counterweight, argued at full length and on historical grounds rather than around them.
Dunn, Jesus Remembered (2003) for the memory turn, and Keith and Le Donne, Jesus, Criteria, and the Demise of Authenticity (2012) for the attack on the machinery that produced most of the conclusions above.
On the texts themselves, Bruce Metzger’s A Textual Commentary on the Greek New Testament is the standard reference for why particular verses are bracketed, and Ehrman’s Misquoting Jesus is the readable version of the same material.
The one instruction that matters is structural rather than bibliographic: any single book in this list will read as though the question is closed. Two of them will not.
Why he is at this table
Seat seven, in the right wing, between Leonardo and Buddha.
His function is the one nobody else here can perform: he is the only counsellor who addresses the person rather than the problem. Every other seat offers an analysis - of power, of nature, of the self, of the species, of the state. He offers a relation, and the difference shows up immediately when a question is put to the table. Where Marcus asks what is within a person’s control and Machiavelli asks what will actually happen, Jesus asks what is owed to whoever is standing in front of them, and the answer is generally more than seems reasonable.
He is also, at this table, the one whose teaching most obviously does not work - in the sense Machiavelli means. A state run on the Sermon on the Mount would not survive a decade. Machiavelli says so directly in the Discourses, and he is not sneering; he thinks it is true and that the consequences are visible in Italian history. The reply is that the teaching was never a proposal for running a state.
Nietzsche, across the room, traced the whole of slave morality to this seat - and then made the distinction that makes him worth listening to on it: in The Antichrist he writes that there was only ever one Christian and he died on the cross, and that everything after is Paul. He attacks the institution and treats the man as a case he cannot quite dismiss.
Buddha, beside him, arrived at renunciation and compassion by a completely different route and offers no God, no self and no salvation from outside. The two are often assimilated and should not be: one says the self is a process to be seen through, the other says the self is beloved. They agree about almost every practical instruction and about nothing underneath it.
Dostoevsky, across the empty chair, wrote him a scene - the Grand Inquisitor - and could not give him a single line, because any line would have made him smaller. He gets a kiss and a silence, and the silence is Dostoevsky’s considered judgement about what an answer would cost.
And Diogenes, at the far end, would notice something the theologians rarely mention: that of everyone at this table, the man from Nazareth is the one who most resembles him. No property, no house, no book, no institution, a habit of saying the unwelcome thing in public, and an early execution by the authorities for doing it.
The works
Nothing written
He left no text. The single scene in which he writes is John 8, where he writes in the dust with his finger and the passage does not say what - and that passage is absent from the earliest manuscripts.
c. 70 CE
The Gospel of Mark
The earliest gospel and the source Matthew and Luke both used. Urgent, compressed, and the least theologically elaborated: its Jesus is hurried, sometimes angry, and repeatedly tells people to say nothing.
c. 80–90 CE
The Gospels of Matthew and Luke
Both expand Mark and both draw on a shared body of sayings not in Mark - the hypothetical source scholars call Q. Matthew writes for a Jewish audience, Luke for a Greek one, and the differences are visible in what each keeps.
c. 90–110 CE
The Gospel of John
A different order of events, a different chronology, a different voice. Where the other three have short sayings and parables, John has long discourses in which Jesus states his own identity directly. The one historians trust least as a record of speech, and the one that shaped Christian theology most.
c. 50–60 CE
The letters of Paul
Earlier than any gospel by fifteen to twenty years, and the earliest Christian documents in existence - written by a man who never met him and shows almost no interest in his teaching.
c. 60–140 CE
The Gospel of Thomas
One hundred and fourteen sayings, with no narrative, no trial and no crucifixion. A Coptic copy surfaced at Nag Hammadi in 1945; Greek fragments had been at Oxyrhynchus since 1897. Whether it is an independent early collection or a second-century compilation drawn from the gospels is the argument, and much of the case for a non-apocalyptic Jesus turns on the answer.
c. 93–116 CE
The non-Christian references
Josephus mentions him twice (one passage certainly tampered with, one widely accepted), and Tacitus records that "Christus" was executed under Pilate. Thin, hostile or indifferent, and decisive against the claim that he never existed.
Where the answers part
The same question, read differently by the seats nearby.
Niccolò Machiavelli
Whether the meek inherit anything
Machiavelli’s Discourses diagnose Christianity as a political weakness: it made men humble and contemplative, glorified suffering, and left republics undefended against those who had not been trained in it. He is not sneering; he thinks it is true and consequential. Jesus is the source of the ethic being diagnosed, and the disagreement is not about whether the teaching is beautiful but about who pays for it.
Friedrich Nietzsche
Where the trouble started
Nietzsche traced slave morality to precisely this teaching - the blessing of the poor, the meek and the persecuted read as ressentiment given a liturgy. He is careful, though, to separate the man from what followed: in The Antichrist he writes that there was only ever one Christian, and he died on the cross. The quarrel is with Paul; the man himself he treats almost as a lost case study.
Marcus Aurelius
Whether the world is ending or only turning
The preaching that most historians reconstruct expects a decisive intervention, soon, after which the present arrangement is not repaired but replaced. Marcus, writing privately on campaign, holds the opposite shape: things dissolve and recur, nothing is new, and the only available work is the correct disposition of one person toward an order that will not change. One waits for an end and finds the interval short. The other assumes there is no end and finds the interval sufficient.
Contents
Did Jesus exist?
Effectively every historian of the period, religious or not, accepts that he did. The evidence is the multiple independent early Christian sources, the letters of Paul written within two decades and referring to people who had known him, and the non-Christian notices in Josephus and Tacitus. The strongest argument is the criterion of embarrassment: a movement inventing a messiah would not have invented one executed by Rome as a criminal, which was the single hardest fact its earliest preachers had to explain.
What did Jesus actually teach?
The arrival of the kingdom of God - not a place after death but God’s rule breaking into the present and overturning the existing order. Around that: love of enemies, forgiveness without limit, the reversal of status so that the poor and the last come first, and an unusual disregard for purity boundaries. He taught mostly in parables, which are not illustrations but traps: short stories that lead a listener into agreeing with something before they see what they have agreed to.
Why was Jesus crucified?
Crucifixion was a Roman penalty for sedition and rebellion, applied to non-citizens. The charge on the cross - King of the Jews - is preserved in all four gospels and is not something Christians would have invented. He entered Jerusalem at Passover, when the city was full and the authorities nervous, and caused a disturbance in the Temple. He was executed as a political threat, and the historical question is how much of one he was.
How reliable are the gospels as history?
They are theological documents written thirty-five to eighty years after the events, in a different language from the one he spoke, by authors who did not witness them and who arrange material for their own purposes. They also preserve material that is early, embarrassing, and unlikely to be invented. Historians use them the way they use any partisan ancient source: carefully, with criteria, and without expecting a transcript.
What is the Sermon on the Mount?
Three chapters in Matthew collecting his teaching: the Beatitudes, the antitheses that cite a received rule and then override it, the Lord’s Prayer, and the instructions on wealth and judgement. Whether it was one sermon is doubtful - Luke has much of the same material scattered and delivered on a plain. As a document it is the most demanding ethical text in the Western canon, and the history of Christianity is largely a history of arguing about how much of it was meant literally.
Did Jesus claim to be God?
In Mark, Matthew and Luke he does not say so. He refuses the word good for himself at Mark 10:18, and says the Son does not know the hour at Mark 13:32. In John he says it repeatedly, and John is the latest gospel and the least reliable as a record of speech. Scholars divide over when devotion to him as divine began: Bart Ehrman traces a development, while Larry Hurtado and Richard Bauckham find it already in the earliest Pauline material.
When were the gospels written?
Mark around 70 CE, Matthew and Luke in the 80s, John somewhere between 90 and 110. The dates are estimates from internal evidence, chiefly Mark 13, which describes the destruction of the temple. A minority of critical scholars, among them John A. T. Robinson and James Crossley, argue for earlier dates. All four were written anonymously in Greek, and the four names appear as a settled set with Irenaeus around 180.
What language did Jesus speak?
Aramaic, the everyday language of first-century Galilee. He would have known Hebrew as the language of scripture, and Greek was common in the region, though whether he spoke it is disputed. The gospels are in Greek, so every saying reaches a reader through a translation made by someone else. Mark preserves a handful of Aramaic words untranslated, among them talitha koum at 5:41 and Eloi, Eloi, lama sabachthani at 15:34.
Is there historical evidence for the resurrection?
There is evidence about people. Within a few years of the crucifixion, followers who had scattered were saying in Jerusalem that they had seen him, and Paul preserves a list of appearances at 1 Corinthians 15:3 to 8 that is older than his own letters. Whether the event itself occurred is not a question historical method is built to answer, in either direction. Dale Allison, after a book-length examination, declines to decide.
Was Jesus married?
No source says so. Paul, arguing at 1 Corinthians 9:5 for his own right to travel with a wife as Cephas and the brothers of the Lord did, would have cited Jesus if he could. Mary Magdalene is never called his wife in any first-century text. The so-called Gospel of Jesus’s Wife, announced in 2012, was exposed as a modern forgery in 2016. Unmarried is likely, and unproven.
Is the eye of the needle a gate in Jerusalem?
No such gate existed. The claim first appears as a medieval gloss attributed to Anselm, travelled through Aquinas and then through a marginal note in the Geneva Bible, and was traced by Agnieszka Ziemińska in New Testament Studies in 2022. Theophylact, usually blamed for it, said only that some read the Greek word as cable rather than camel. At Mark 10:26 the hearers ask who then can be saved, which suggests the saying was meant to be impossible.
Sources and further reading
- The Historical Figure of Jesus1993The standard sober account of what can be established, by one of the historians who reset the field.
- Jesus and Judaism1985The book that opened the Third Quest by starting from the actions rather than the sayings, and that set out the eight almost indisputable facts.
- Jesus the Jew: A Historian’s Reading of the Gospels1973The book that made it impossible to discuss him outside first-century Judaism.
- A Marginal Jew: Rethinking the Historical Jesus (5 vols., 1991–2016)The most exhaustive treatment in any language; a Catholic priest writing under strict historical-critical method. Volume 2 covers the miracle traditions, volume 5 the parables.
- Did Jesus Exist? The Historical Argument for Jesus of Nazareth2012
- Jesus: Apocalyptic Prophet of the New Millennium1999
- How Jesus Became God: The Exaltation of a Jewish Preacher from Galilee2014
- The Historical Jesus: The Life of a Mediterranean Jewish Peasant1991The social-radical reading; contested, and useful for its attention to the economics of Galilee.
- Jesus of Nazareth, King of the Jews: A Jewish Life and the Emergence of Christianity1999Builds the whole reconstruction on one puzzle: Rome executed him as an insurrectionist and arrested none of his followers.
- Constructing Jesus: Memory, Imagination, and History2010Abandons the criteria in favour of recurrent patterns across the sources, and argues that the apocalyptic material cannot be removed without removing most of the tradition.
- The Resurrection of Jesus: Apologetics, Polemics, History2021Four hundred pages that end without a verdict, on the ground that the evidence is neither good enough to make disbelief unreasonable nor bad enough to make faith untenable.
- Jesus and the Victory of God1996The main confessional counterweight to the sceptical readings, arguing the eschatology through.
- The Resurrection of the Son of God2003
- Jesus Remembered2003The memory turn: the earliest tradition is already an interpretation, and there is no uninterpreted layer beneath it to recover.
- Jesus, Criteria, and the Demise of Authenticity2012The collective attack on the criteria of authenticity, with contributions from Allison, Morna Hooker, Mark Goodacre and others.
- Lord Jesus Christ: Devotion to Jesus in Earliest Christianity2003
- Jesus and the Eyewitnesses: The Gospels as Eyewitness Testimony2006
- Aramaic Sources of Mark’s Gospel1998
- The Case Against Q: Studies in Markan Priority and the Synoptic Problem2002
- Excavating Q: The History and Setting of the Sayings Gospel2000
- The Four Gospels and the One Gospel of Jesus Christ2000The case that the gospel titles are early, on the ground that a scroll circulating in a collection needs a label and the title forms vary almost not at all.
- The Quest of the Historical Jesus1906Von Reimarus zu Wrede. The book that put thoroughgoing eschatology at the centre and observed that each researcher had found his own face at the bottom of the well.
- What Did Jesus Look Like?2018
- Archaeology of Jesus’ Nazareth2023The Nazareth Archaeological Project, running since 2006. First-century Jewish occupation, and a population revised upward toward a thousand.
- A Textual Commentary on the Greek New Testament1994Second edition. The standard reference for the manuscript evidence behind the pericope adulterae, Luke 23:34 and the ending of Mark.
- Engaging the Powers: Discernment and Resistance in a World of Domination1992The source of the third-way reading of turning the other cheek. Widely admired as constructive theology and widely doubted as exegesis.
- The Origin of the “Needle’s Eye Gate” Myth: Theophylact or Anselm?2022
- Gehenna: The Topography of Hell1986Establishes that the burning rubbish dump has one source, a thirteenth-century commentary by David Kimhi, and no support in any earlier literary or archaeological evidence.
- The Origins of the Liturgical Year1986The standard statement of the calculation hypothesis for 25 December, against the older derivation from Sol Invictus.
- The Crucified Man from Giv’at ha-Mivtar: A Reappraisal1985The re-examination of the only skeletal evidence for crucifixion yet recovered, excavated by Vassilios Tzaferis in 1968.
- Dating the Crucifixion1983
- Veritas: A Harvard Professor, a Con Man and the Gospel of Jesus’s Wife2020
- The Aryan Jesus: Christian Theologians and the Bible in Nazi Germany2008What the field looked like when it tried to detach him from Judaism, and why the Jewishness of Jesus is not a neutral scholarly preference.
- Antiquities of the Jews, XVIII.63–64 and XX.200c. 93 CE
- The Jewish War, VI.300–309c. 75 CEJesus son of Ananias, the peasant who cried out against the temple at a feast, was flogged by the procurator Albinus, judged mad and released.
- Annals, XV.44c. 116 CEQuoted here in the translation of Alfred John Church and William Jackson Brodribb, 1876.
Last reviewed:
