A historian deciding whether the Gospels are reliable does not start with faith or with suspicion. She asks a short list of questions that she would ask of any ancient document. Does the writer know the place he describes? Does the name attached to the book make sense? How far is the writing from the events? Do separate accounts fit together in ways nobody arranged? Can we tell whether the text was changed later? This guide teaches those questions one at a time, using the evidence Peter Williams lays out in his Jude 3 Project conversation with Lisa Fields. Then it hands the questions to you, so you can run them on the Gospels yourself.
You can work through the guide alone in one long sitting. A group can split it into two meetings, with the tests in the first and the practice and discussion in the second. Each section ends with a short task that uses a Bible and nothing else.
Three distinctions to settle first
Most arguments about the Gospels go wrong before they start, because people mean different things by “reliable.” Three distinctions will keep the work clear.
Reliable is not the same as error-free. A historian calls a source reliable when it gives us good access to what happened. That judgment comes in degrees. Asking whether the Gospels are inerrant is a different question, and it is a theological one. This guide is about the historical question.
Reliable history is not the same as a proven miracle. The tests can show that a document is early and well informed. They cannot, by themselves, force anyone to believe Jesus rose from the dead. What they can do is remove the excuse of dismissing the sources before reading them.
A scholar’s argument is not the same as an established fact. Williams is a careful scholar who leads Tyndale House, a research center devoted to the Bible. When he says something about the evidence, this guide reports it as his argument. Part of your task is to check whether the argument holds.
Lisa Fields knows why these distinctions matter. She says she came into apologetics because Bart Ehrman’s work was the textbook in a course she took, and the question of the Gospels’ reliability was one she wrestled with deeply. Ehrman is a prominent skeptic, and he appears often below. Keep in mind that Williams treats him as a serious opponent to answer, not a villain to mock.
The rule behind every test: one standard for all documents
Before the individual tests, learn the rule that governs all of them. A historian must apply the same standard to every ancient source. If she accepts a classical author’s work on a certain kind of evidence, she cannot demand much more from the Gospels simply because they make religious claims.
Williams presses exactly this point. He compares the evidence that John wrote John’s Gospel with the evidence that Plato wrote his dialogues, and argues the case for John is stronger. Nobody loses sleep over Plato. Williams says that if you set aside the culture war between belief and skepticism and “look at it like you would a classical or historical work, you would not be having this conversation.”
The rule also works in the other direction. A Christian cannot ask the world to lower its standards for the Bible. The request is fairness, not a discount.
Quick task: Write down one ancient figure whose writings you have never doubted. Then ask yourself what evidence you actually have that this person wrote them. Most of us discover we have never asked.
Test one: does the writer know the ground?
A writer working far away and generations later tends to slip. He gets distances wrong, misplaces towns, or describes customs that belong to his own time. A writer close to the events tends to get such details right without effort.
Williams treats this test as nearly settled. He says that scholars across the spectrum, Ehrman included, agree that “the gospel writers really knew an awful lot about the place” where the events happen, naming Jerusalem and Galilee. This one concession disposes of a popular idea that Williams says ordinary skeptics often hold: that an emperor or a pope made the whole story up hundreds of years later.
Notice the limit. Knowing a setting well is necessary for reliable history, but it does not guarantee it. A good novelist can know a city street by street and still invent the plot. This test narrows the options. It does not finish the case.
Test two: does the name on the document make sense?
This test has two parts. The first is motive. When people forge a document to give it authority, they attach the most impressive name they can. So a historian asks whether the name on a book is the kind a forger would choose.
Williams applies this to Mark and Luke. Neither man was one of the twelve, and both would be forgotten if not for their Gospels. He concludes: “So there’s no motive you can have for putting the names Mark and Luke on, because they’re not even one of the 12 disciples.”
The second part is internal fit. Does the content of the book match what we know about its supposed author? Matthew was a tax collector, and Williams observes that “there’s a huge amount of financial interest in Matthew’s Gospel.” He points to the gifts of the wise men, the money paid to the guards at the tomb, and the money given to Judas. His verdict: “It really fits.”
A fit like this does not prove authorship. It means the internal evidence agrees with the tradition instead of pulling against it.
Side-by-side task: Read Mark 14:10-11 and then Matthew 26:14-16. Both tell of Judas going to the chief priests. Mark says they promised him money. Matthew records Judas asking what they will give him and names the amount, thirty pieces of silver. Then read Matthew 28:11-15, the bribing of the soldiers, and look for it in the other Gospels’ resurrection chapters. You will not find it there. Ask: is this the attention to money you would expect from a tax collector, or could another explanation account for it equally well?
Test three: how close is the writing to the events?
The farther a document stands from what it reports, the more room there is for legend. A historian wants to know whether the writing falls within living memory, when people who saw the events could still correct a false account.
Williams does something worth noticing here. He does not fight over exact dates. He prints Ehrman’s dates in his book, says he thinks they are too late, and accepts them for the sake of argument. “Even then, the gospels are within a plausible memory time.” He adds that his own case is that the Gospels are first-generation documents, without fixing particular years.
He admits Ehrman would answer that memory corrupts quickly. Williams replies with his grandmother, who is 98 and remembers the Second World War as if it were yesterday. Then he makes an honest concession: “you can play the length of time either way.” Time alone settles nothing. That is why test three must work alongside the others.
Luke’s Gospel speaks directly to this test. In his preface Luke says he is passing on accounts “Even as they delivered them unto us, which from the beginning were eyewitnesses, and ministers of the word” (Luke 1:2). That is a claim, not a proof. But it is the kind of claim a historian can test against the rest of the evidence.
Test four: do separate accounts fit without being arranged?
When two sources describe the same event, a historian watches for two warning signs. Accounts that match word for word may simply be copying one another. Accounts that clash at every turn suggest someone was careless or inventing. Independent witnesses to a real event usually produce a third pattern: small agreements nobody planned, often in details too minor for anyone to bother faking.
Williams argues the Gospels show this third pattern: “we can explain one gospel and another gospel, and how they have subtle agreements with each other best by assuming that they’re both speaking the truth.” He adds that Ehrman’s approach has no explanation for these agreements. Williams does not walk through specific cases in this conversation, so the next task lets you test one yourself.
Comparison task: Read John 6:5, where Jesus asks Philip where they can buy bread for the crowd. John gives no reason for asking Philip in particular. Now read Luke 9:10, which places this feeding near a city called Bethsaida. Then read John 1:44 and John 12:21, which say Philip came from Bethsaida. Neither writer connects these dots. Decide for yourself: is this a real unplanned fit, a coincidence, or something else? One case proves very little. The argument depends on whether such fits keep turning up.
This test also covers sources outside the Bible. Williams notes that the Jewish Talmud, not only the Gospels, says Jesus died on the eve of Passover. Outside agreement strengthens a source. Outside disagreement raises a real question, which brings us to the hardest case.
Test five: can we tell whether the text was changed?
A reliable original is useless if someone later rewrote it in secret. So a historian asks how a text was copied and whether changes would show.
Williams takes the hardest example head on: the account of the woman caught in adultery (John 7:53-8:11). He believes it is an early passage but not part of the earliest form of John, while making clear he is not saying the event never happened. The passage is still printed in the King James Version, and many Bibles add a note about it. Williams argues that this passage, along with the ending of Mark, actually helps the case. Where the passage appears, it shows up in different places in different manuscripts. That disorder shows “there’s been no central power controlling all of these manuscripts.”
He calls the two passages a test case for the whole text: “those two passages provide a very good control sample.” In other words, they show what tampering looks like. When someone does alter the text, “we get a bit of a chaotic pattern in manuscripts.” Williams says Erasmus already noted the problem with the adultery passage in his printed Greek edition of 1516, and five hundred years of manuscript discovery have turned up nothing comparable anywhere else.
He also uses a plain comparison for anyone who thinks one doubtful passage infects the rest: “If you miss a bit of a film, it doesn’t make all of the rest of the film uncertain.” The comparison has a limit. A film has one source, and the Gospels survive in many copies. That difference actually favors the Gospels, because many copies let us see where they differ.
The mistake most people make with hard cases
Here is the error to avoid. People often assume that one unsolved problem sinks a source. Historians do not work that way with any ancient author.
Williams names what he considers possibly the worst problem in the Gospels: the census in Luke 2. The King James Version reads, “(And this taxing was first made when Cyrenius was governor of Syria.)” (Luke 2:2). Cyrenius is the older spelling of Quirinius. Luke and the Jewish historian Josephus do not obviously agree on the timing. Williams admits “we don’t know enough.” Then he asks a fair question: “if one of them has to be wrong, why do I have to say, Luke got it wrong?” He argues Luke wrote earlier and closer to the events, and allows that both writers may turn out to be correct.
A second form of the same mistake is the long list of alleged contradictions. Williams compares the Gospels to Dickens opening a novel with the best of times and the worst of times. No reader quits the book there. He asks readers to come with the attitude, “I wonder what’s meant by that, rather than, oh, is there something I can try and catch out a Christian on?”
Watch for one more presupposition. Some date the Gospels late because Jesus predicts the destruction of the temple, and prediction is assumed impossible. Williams answers that this decides the question before looking at it: “That’s not an argument that’s just starting with your presuppositions.” A historian may choose not to explain events by miracles. But that choice comes from philosophy, not from the documents.
Practice: run the five tests yourself
Choose one chapter. Luke 1, Mark 14 or John 19 all work well. Read it slowly with a sheet of paper divided into five columns, one for each test.
- Under ground, list every place, custom, office or name the writer mentions without explaining.
- Under name, note anything in the chapter that fits or strains against the traditional author.
- Under distance, note any claim about where the information came from.
- Under fit, look up one detail in another Gospel and record whether it agrees, adds, or seems to conflict.
- Under transmission, check whether your Bible has a note on any verse in the chapter, and what it says.
Leave a column empty if the chapter gives you nothing for it. An empty column is information too. If you are leading a group, assign each person or pair one column and have them report what they found before anyone draws a conclusion. The aim is to practice observation before judgment.
Questions for discussion
- Before this guide, how did you decide whether a historical source deserved trust? Which of the five tests were you already using without a name for it?
- The Mark and Luke argument rests on motive. What evidence, if it turned up, would weaken it?
- Williams grants that the length of time can be argued either way. What features of an account would make you trust a memory more or less?
- Where did the Bethsaida comparison leave you? What would you need to see before calling a pattern of such fits convincing?
- Is it fair for a historian to rule out miracles from the start? What would a genuinely neutral starting point look like?
- Williams answers long lists of objections by asking only two questions: “what’s your most difficult question? And what’s your most important question?” If you are willing to share, how would you answer them about the Gospels? What would count as a good answer?
- How should a footnote about a disputed passage affect your confidence in the rest of your Bible?
On question six, no one should feel pressed to answer aloud. Some people’s most important question is personal, and writing it down privately is a complete way to take part.
What the tests leave in your hands
The five tests can tell you whether the Gospels deserve a serious reading. They cannot read the Gospels for you. Williams makes the point practically: “It only takes about nine hours to read all of the gospels.” Once the documents pass the tests a historian would use on any ancient source, the open question shifts from the books to the person they describe.
So carry this question into your next reading: if these are early, informed accounts from people close to the events, what do I do with the Jesus they report?