The questions, with the concessions left in.

“Aren’t you just using the Bible to prove the Bible?”

If that were what I was doing, you’d be right to stop reading. It isn’t.

The circular version goes: the Bible is God’s word, because the Bible says it is. That’s not an argument, it’s a loop, and a philosopher takes it apart in one sentence. If you’ve heard it put that way, you were handed a bad version.

Here’s the claim I’m actually making. It’s a different shape. The Bible generates the evidence by which it can be tested — and it doesn’t control that evidence.

It names kings. It names cities. It names Roman officials with specific job titles in specific provinces in specific years. Those people left records in archives nobody in the Bible had any access to, written by people who wanted Israel destroyed. That isn’t self-witness. That’s a text handing you a checklist and inviting you to go look.

And it goes further than that. It stakes itself. Paul writes to Corinth that if Christ has not been raised, “our preaching is in vain, and your faith also is in vain” (1 Corinthians 15:14, WEB). He doesn’t say the faith would be weakened. He says it would be empty. That’s a religion writing its own kill switch into its own scriptures.

So the claim isn’t “trust the book because the book says to.” It’s this: this book can be wrong in ways you can find out about, it has been tested that way for two hundred years by people who wanted to break it, and here’s how it did.

That’s a falsifiability claim, not a circular one. It also puts an obligation on me — I have to tell you what would actually falsify it. That’s question 6, and I’d hold me to it.

“Doesn’t science rule this out before we even start?”

No. But I’m not going to try to prove God to you with physics either, and you should be suspicious of anyone who does.

This one matters more than it looks, because it runs before everything else. If miracles are impossible in principle, then every piece of historical evidence on this site is just an interesting fact about an old document. The filter gets installed before the evidence arrives.

So I need one thing here, and it’s a low bar: that the question of whether there’s a mind behind nature is open, not closed. Not likely. Not proven. Open.

One — the beginning. In 2003, three cosmologists — Arvind Borde, Alan Guth and Alexander Vilenkin — published a four-page paper in Physical Review Letters titled “Inflationary spacetimes are not past-complete.” Their result is remarkable for what it doesn’t assume: no energy condition, no Einstein field equations, no assumption that the universe is smooth. It needs one thing — that the region in question has, on average, been expanding. Under that condition, the past history can’t be extended indefinitely. Trace it back and you reach a boundary.

Now the part that usually gets left out. The paper’s own conclusion is that inflation alone can’t describe what’s at that boundary — “some new physics is necessary,” in their words. The theorem is stated in classical spacetime, and near that boundary physicists expect you’d need a quantum theory of gravity, which nobody has. Alan Guth, one of the three authors, publicly thinks the universe is probably eternal. And Vilenkin has written, under his own name, that in his view the theorem “does not tell us anything about the existence of God” (Inference, 2015).

I’m quoting that against myself on purpose. He’s right, and my argument doesn’t need him to be wrong. A theorem doesn’t oblige its own authors to draw the conclusion its users draw, and I’d rather you heard that from me than from someone correcting me.

Two — the constants. The equations of physics contain constants the equations themselves don’t fix. Change several of them and you get universes with no stars, no chemistry, no structure. That’s not a fringe observation; it’s the subject of review articles in Physics Reports and in the Publications of the Astronomical Society of Australia.

And I am not going to give you a number. You’ll see “one chance in ten to the whatever” all over apologetics. To state a probability, you need a way of measuring the space of possible values, and there is no agreed way to do it — a point pressed hardest by three Christian philosophers, in the journal Mind, against their own side (McGrew, McGrew and Vestrup, 2001). Meanwhile Fred Adams’s parameter surveys find that viable universes exist across a range of parameter space — roughly a quarter of the region he surveyed permits working stars. I’ll take that number too, and it cuts against me.

Sean Carroll, who is about as far from my conclusions as a physicist gets, has written that the fine-tuning “is real” — while arguing it sits in the universe’s initial conditions rather than the constants. I take both halves of that.

Here’s where that leaves us, and it’s less than you might want. None of this proves a mind. Carroll’s deepest objection is that every explanation ends in a brute fact, including mine — “God exists” is as much a stopping point as “the universe exists.” I think that’s a serious argument and I’m not going to answer it here.

What I claim is only this: physics has not closed the question. Which means when I hand you a historical claim about a man walking out of a tomb, you are not entitled to rule it out before reading it. You’re entitled to weigh it. That’s all I wanted from this section, and it’s the last time I’ll mention physics.

“Does the archaeology actually back it up?”

In places, remarkably. In other places, no — and the places where it doesn’t are the famous ones.

Bad news first, because you’ll find it anyway and I’d rather you got it from me.

The patriarchs. Abraham, Isaac, Jacob, Joseph. There is no external attestation of any kind. Not one inscription, not one record, not one mention outside the text. What gets argued for instead is verisimilitude — that the customs, names and travel routes fit a second-millennium world — and even that is contested. Anyone who tells you archaeology confirms Abraham is telling you something that isn’t so.

Jericho. You’ve heard the walls came down. Kathleen Kenyon dated that destruction to roughly 1550 BC. Bryant Wood re-dated it to about 1400, which would fit an early Exodus — but radiocarbon dating of charred grain from the destruction layer supports Kenyon, and the single 1400 BC date Wood leaned on came from a batch the British Museum retracted after its equipment turned out to be out of calibration. On the current evidence, Jericho was not a walled, occupied city in either Exodus window.

I hold the early date, around 1446 BC. I hold it in spite of the archaeology at Jericho, not because of it. The late date, around 1250, is a live position held by serious people, and the split is real.

The conquest. The destruction layers at the key sites don’t line up in one period. Worth noticing, though: Judges 1 already describes a slow, incomplete settlement that undercuts the sweeping summary in Joshua. The Bible kept that tension instead of smoothing it out. That’s not what a tidied-up story looks like.

Now the other half. Attestation rises as the story moves forward. From the divided monarchy onward, the Bible’s cast list starts turning up in archives written by people who wanted Israel dead.

• An Assyrian monolith of Shalmaneser III — British Museum 118884 — lists “Ahab the Israelite” among enemy kings at the battle of Qarqar in 853 BC. The Bible never mentions that battle at all.
• The Black Obelisk — British Museum 118885 — records tribute from “Jehu, son of Omri.” Jehu was not Omri’s son. He exterminated Omri’s family to take the throne. Assyria simply filed Israel under its standing label for the place, regardless of who was sitting on it. That is not the mistake a forger makes copying somebody else’s story. It’s the mistake a foreign scribe makes writing down what he actually knows.
• Sennacherib’s Taylor Prism — British Museum 91032 — describes his 701 BC campaign against Judah and boasts that he shut Hezekiah up “like a bird in a cage” in Jerusalem. He does not claim to have taken the city.
• Merneptah’s stele in Cairo — JE 31408, around 1208 BC — names “Israel” as a people in Canaan.
• The tunnel Hezekiah cut under Jerusalem carries an inscription describing two work crews meeting underground (KAI 189). In the 1990s, two respected scholars argued the whole thing was five hundred years later than the Bible places it. That claim got tested: radiocarbon and uranium-thorium dating of the tunnel’s own mineral deposits, published in Nature in 2003, put it right around 700 BC. I mention the challenge and not just the answer, because a book that only shows you the tests it passed isn’t showing you a test at all.

Lawrence Mykytiuk has published criteria for identifying a biblical person in an inscription — name, father’s name, title, all three matching — and counted 53 individuals meeting them as of his 2017 update.

And here’s what I won’t use. The Jezebel seal. The Ahaz bulla. The so-called Isaiah bulla. The James Ossuary. The “Pilate ring.” Every one of them either came off the antiquities market with no excavation record, or had its headline identification collapse under later review. Buying an artifact instead of digging one up is exactly the blank check a forger needs. I’d rather have fewer exhibits I can defend than more I can’t.

The honest limit on all of it: naming real kings proves the writers knew real history. It does not prove the theology. What it proves is that this is not a book of timeless sayings floating free of the world. It has its feet on ground you can dig — and that means it could have been caught, and in the places it matters most, it hasn’t been.

“You picked the books. Of course it reads like one story.”

Partly true, and I’ll concede it. But the people who preserved the front half didn’t believe the back half — and they kept the passages that hurt them most.

This is the strongest objection anybody makes to me. So let me state it at full strength instead of the weak version. The weak version is “Nicaea picked the books and burned the rest.” Nicaea didn’t discuss the canon. Skip it.

Here’s the strong version, and it’s genuinely hard. The Hebrew corpus is the product of centuries of collection, editing and arrangement by one religious tradition with a continuous theological programme. Then the New Testament writers, saturated in those texts, composed their narratives against them — Matthew doesn’t discover that Jesus fulfils Hosea; Matthew writes the flight into Egypt because of Hosea. Then the church selected, out of a much larger pool, the writings that best matched the reading it already held. Bruce Metzger — a Christian scholar, not a critic — lists conformity to the rule of faith as one of the criteria they used. So coherence at the end of that pipeline isn’t surprising. It’s the pipeline’s output specification.

That’s a good argument, and I’m going to concede a real piece of it. Some of the coherence is curatorial. At the New Testament end, the selectors were selecting partly for fit, and I’m not going to pretend otherwise.

Here is what curation doesn’t explain.

The selectors only controlled one end. The coherence I’m pointing at runs Old Testament to New. The Hebrew texts were in the custody of a community that rejected the Christian reading of them and had every motive to quietly lose the passages Christians leaned on hardest. They didn’t. The texts most useful to my side sat in the hands of people who did not want my conclusion, and they preserved them anyway. Physical pre-Christian copies exist. You can go look at the Great Isaiah Scroll.

The canon kept material that works against the people carrying it. Ecclesiastes. Job’s unanswered question. Lamentations. Judges 1 undercutting Joshua’s summary. Four resurrection accounts that don’t line up on the details. An editor optimising for a smooth story would have fixed every one of those, and it would have been easy.

Marcion is the control experiment. In the second century, Marcion did exactly what this objection describes — he curated for coherence. He cut the entire Old Testament. He cut down to one gospel. He cut ten of Paul’s letters and edited those. That’s what selection-for-fit actually produces: subtraction. The mainstream canon went the opposite direction on every axis, keeping the whole Jewish scriptures and four gospels that don’t agree with each other.

And the arc doesn’t change across canons. I said up top that 66 is the Protestant count. Catholics have 73, Orthodox churches more, Ethiopian Tewahedo more still. What the additions add is wisdom literature, martyr history, a narrative novella, and apocalyptic expansion — all inside the same storyline. The story doesn’t change shape depending on whose Bible you pick up.

Where that leaves me is a step back from where I’d like to be, and I’m going to say so out loud. I can’t call the coherence proof of supernatural coordination. Curation explains part of it. What I can say is that it’s a striking structural feature that curation does not fully explain. That’s a demotion. I’d rather take it in public than defend a bigger claim I can’t carry.

“Why does it all come down to the resurrection — and how far does the evidence actually go?”

Further than most skeptics expect, and nowhere near as far as most Christians claim. Here’s exactly where it stops.

Everything else on this page is an argument about a book. This one is an argument about a body.

And the book itself says this is where the weight sits. Paul tells Corinth that if Christ has not been raised, the preaching is empty and the faith is empty. Not diminished — empty. I’m not going to defend a version of Christianity that survives a negative answer here, because the New Testament doesn’t offer me one.

First, something uncomfortable about where this sits. Everywhere else, I’ve been showing you a text that submits to outside checking — kings named by their enemies, cities in datable layers, officials whose job titles turn up on inscriptions. Now I arrive at the claim the whole thing rests on, and I have to tell you that here the outside evidence is at its thinnest.

The setting is superbly attested. Pilate, the high priest, the city, Roman crucifixion practice — we even have the heel bone of a crucified man recovered from a first-century tomb in Jerusalem, the only crucifixion victim ever recovered by excavation anywhere. The event is not attested by anybody who was not, or did not become, a Christian. No non-Christian source reports a resurrection as an event. That is the position, and I’d rather hand it to you myself than have you find it in a review.

What the evidence does carry. The oldest thing in the New Testament about the resurrection isn’t a story. It’s a formula. Paul quotes it in 1 Corinthians 15 and flags it as a quotation — he says he received it and passed it on, which is the standard language for handing over fixed material. It has non-Pauline vocabulary, a balanced structure, and it uses Peter’s Aramaic name. Scholars across the spectrum agree Paul didn’t write it.

Work backward from a date we can actually anchor — an inscription at Delphi fixing when Gallio was proconsul of Achaia, and therefore when Paul was in Corinth — and that formula was already circulating within a few years of the crucifixion. Years, not generations. That’s a real result, and it closes off most of the room a slow legend needs.

Now the limits, in the same breath. Early is not the same as true. A false claim made in AD 33 is still a false claim. And it’s worth knowing that the scholar most often quoted for how early the creed is, Gerd Lüdemann, does not believe Jesus rose — the early date isn’t a problem for his theory, it’s a requirement of it. When you see that quotation presented as a hostile witness reluctantly admitting something, you’re being handled.

I’m not going to tell you nobody dies for a lie. It isn’t true. Around AD 112 a Roman governor named Pliny wrote to the emperor describing exactly how he handled Christians: he offered them release if they’d curse Christ, and he reports that some of them took it. Some early Christians recanted, and we know it because a hostile administrator wrote it down. What I can say is narrower — the specific people who claimed they personally saw Jesus alive took beatings, prison, and in several cases execution rather than stop saying it. That rules out one thing: that they were knowingly making it up. It does not rule out that they were sincerely wrong. Sincerely wrong is the serious position, not the silly one.

I’m also not going to use the number. You’ll hear that 75% of scholars accept the empty tomb. That comes from one researcher’s bibliography that has never been published for anyone to check, and its own author says the figure covers only the publications that address that specific question — a count of articles by people who chose to write about it, not a survey of the field. It may well be pointing at something real. It is not a measurement, and I won’t present it as one. What I can tell you is who holds what. N. T. Wright, Dale Allison, Raymond Brown and James Dunn think the tomb was found empty. Rudolf Bultmann, John Dominic Crossan, Gerd Lüdemann, Bart Ehrman and Michael Goulder don’t.

The strongest single card is a man’s handwriting against himself. Paul says, in letters nobody disputes he wrote, that he violently persecuted the church and tried to destroy it. He isn’t being accused of it — he volunteers it, to people who knew him, about twenty years on. Two things follow: the movement existed and was already worth attacking within a very few years of the crucifixion, and we know it from the man who attacked it; and that man changed sides and said the reason was that he saw Jesus.

Here’s the cost of that card. Paul’s own word for what happened to him is revelation, and elsewhere he describes being caught up into the third heaven. He was a man who had visionary experiences and says so himself. If you want to explain him as a visionary, the text hands you the vocabulary. I’m not hiding that.

The serious alternative, at full strength. It isn’t that they lied, and it isn’t that Jesus never existed — almost nobody with a doctorate in the field holds either. It’s that grief and guilt produced experiences that felt exactly like encounters. That’s not a slur on the disciples and it isn’t a fringe idea; experiences of the recently dead are common and they are convincing to the people who have them. What can be said in reply is narrower than apologists usually claim: encounters with the recently dead don’t normally produce a claim that the grave is empty, and they don’t normally found a religion. That’s a difference in what follows, not a difference in what was felt.

So here’s exactly what I think the evidence carries, and then I’ll stop. That within a few years of the crucifixion — years, not generations — a fixed formula was circulating naming specific living people who said they’d seen Jesus alive. That those people accepted beating, prison and in several cases death rather than stop saying it, which rules out knowing fabrication but not sincere error. That one man who’d been violently attacking the movement joined it and said the reason was that he saw Jesus. And that most specialists who’ve examined the question think the tomb was found empty.

It does not carry that any of them saw what they thought they saw. A natural explanation is available. It isn’t stupid, it isn’t desperate, and it’s held by serious people who’ve read more of this than I have.

And here’s where the history runs out. History can get you to: something happened to these people, it happened almost immediately, and they were not lying. History cannot get you to: and therefore God raised him. That gap is not a gap in the evidence. It’s a gap in what the tool can do. The structure makes the claim live. It cannot make it true.

“What would change your mind?”

Three things, named specifically. And I’ll tell you which popular “defeaters” don’t count, because a test that can’t fail isn’t a test.

One boundary first, because I said it in the last answer and I’m not going to quietly walk it back here. History can’t prove God raised Jesus. That cuts both directions — it can’t disprove it either. So nothing below is a claim that some discovery could disprove the resurrection outright. What a discovery can do is remove the specific historical reasons I currently think the claim is live rather than merely asserted. That’s a real test. It’s just not the one the question might make you expect, and I’d rather tell you than have you catch it.

One. The depth turns out to be fake. My whole case assumes a genuinely ancient corpus written across centuries by many hands. If it were shown that the material anchored earliest is deliberate archaizing by a much later single hand — and that the pattern held broadly enough that no independent early layer survived anywhere — the coordination argument would collapse, and part of everything else would go with it. Hebrew linguistic dating is a live, unsettled field. No working scholar currently argues the full version of that claim. But it’s a real position in a real field, not a hypothetical I invented to look brave.

Two. An early document that contradicts the creed. A papyrus, ostracon or inscription — provenanced, securely dated to the first century or very close — either directly disputing the empty-tomb or appearance claims from a source close enough to the events to know, or showing that the formula I’ve built on postdates the timeline I’ve argued for. That is not science fiction. Papyri turn up on a normal schedule. It simply hasn’t happened yet, and if it does, I want you to know in advance that I said so.

Three. Comparative parity. I’ve claimed the kind of historical claim the Bible stakes — named, checkable people, places and dates — isn’t matched by the other books it gets compared to. If a documented pre-Columbian find matched Book of Mormon geography at a controlled excavation, I’d have to give up the distinction the whole comparison rests on.

Now what would not count. Finding a bone and calling it Jesus’s would not count. There’s no reference sample on earth to compare it against, and anyone claiming otherwise is selling something. A dispute about the age of Genesis’s early chapters would not count, because I told you at the top those chapters are outside what I’m arguing. A scribal variant, or a debate over who actually wrote a given letter, would not count — because I also told you at the top that my view of Scripture absorbs exactly that without losing anything.

I name those because a defeater that can’t in principle occur is worse than no defeater at all, and I don’t want to be caught offering one.

And I’d point at the record. I’ve already given up the popular Jericho case. The traditional sixth-century date for Daniel. The seventy weeks as a chronological prediction. The prophecy against Tyre, on the text’s own later admission. The 75% statistic. “Nobody dies for a lie.” Every one of those cost me something I would rather have kept.

That’s the only thing that makes a stated defeater worth anything. Anybody can list what would change their mind. Fewer people can show you the last six times it did.

Where this shows up in the course →