What Is the Bible?

The Bible as Mirror

“There are more differences among our manuscripts than there are words in the New Testament.”

Bart D. Ehrman, Misquoting Jesus

The sentence above is true, and it is the single most effective thing ever written against the reliability of the Bible. It is also, taken by itself, seriously misleading — and the man who wrote it says so in an appendix to the same book, where he grants that no essential Christian belief depends on a disputed reading. Both halves of that need explaining, and this page tries to.

The question in the title is deliberately not “can we trust the Bible?” That question smuggles in an assumption — that the book is the sort of thing that stands or falls as a single object, like a witness whose testimony is either sound throughout or worthless. The prior question is what kind of thing it is: a library rather than a book, written across roughly a thousand years in three languages and a dozen genres, by authors who were not attempting the same task and did not know they were contributing to one collection.

Much of the argument between believer and skeptic turns out to be a disagreement about that, rather than about the evidence. Read the opening of Genesis as a laboratory report and it is false; read the book of Job as a court transcript and it is absurd. The Church has said for a very long time — and said it formally in 1943, before most of the modern controversy — that the first task of the interpreter is to determine what the author was actually doing.

Let’s take the text first, then the canon, then the harder question of inspiration, and ask …

  • If we have hundreds of thousands of variants, how much of what Christians actually believe turns out to be in doubt?
  • And if the Church settled which books belonged in the Bible, what does that imply about which of the two came first?

I invite your engagement with me on these questions. You may leave comments below, but please be sure to read our policy on commenting before doing so.

Atheist/Skeptic View

The skeptic holds that the Bible is what it looks like: an anthology of ancient religious writing, produced over centuries by many hands, copied by hand for fourteen hundred years before printing, selected by committees with interests of their own, and containing history that fails where it can be checked and morality that would be prosecuted today. It is a document of considerable human interest. It is not the word of God.

What They Affirm

We do not have the originals, and the copies disagree. Not one autograph survives. What survives are copies of copies, and among them there are somewhere between 200,000 and 400,000 variants — more differences than there are words in the New Testament. Scribes were tired, careless, and sometimes had views; passages were added, including the ending of Mark and the woman taken in adultery, which appear in Bibles still today.1

The Gospels are anonymous, late, and not by eyewitnesses. None names its author; the titles were attached later. They were written in Greek, forty to seventy years after the crucifixion, probably outside Palestine, by people who did not witness what they describe. Mark came first and Matthew and Luke copied him, which is not how independent testimony behaves.2

They contradict each other on details they ought to agree on. The two nativity accounts cannot both be right about the journey to Bethlehem, and the genealogies of Christ differ from each other. At the tomb the number of women, the number of angels, and what the women did next vary from Gospel to Gospel. These are not deep mysteries; they are the ordinary discrepancies of stories that grew.

The canon was decided, not discovered. There were other gospels, other apocalypses, other letters, and Christians disagreed about them for centuries. Athanasius’s list of 367 is the first to match the present New Testament exactly. A collection assembled by fourth-century bishops, from a field of competitors, is the product of a process with winners and losers — and the winners wrote the account of it.3

Where the history can be tested, it often fails. Archaeology has found no evidence of two million people wandering Sinai for forty years, and the pattern of settlement in Canaan does not look like the conquest of Joshua. Luke dates the nativity by a census under Quirinius that took place a decade after the death of Herod, in whose reign Matthew places the same birth.4

And parts of it are morally indefensible. The conquest narratives command the killing of entire populations, including children. Slavery is regulated rather than forbidden, in both Testaments. Women are property in the legal codes. A book may be an interesting historical artifact and still be a poor guide to conduct, and the usual reply — that these passages must be read in context — is available to the defenders of any text whatever.

Endnotes
  1. Bart D. Ehrman, Misquoting Jesus: The Story Behind Who Changed the Bible and Why (HarperOne, 2005), from which the phrase about more differences than words is taken; on the longer ending of Mark (16:9–20) and the pericope adulterae (John 7:53–8:11), see Bruce M. Metzger and Ehrman, The Text of the New Testament, 4th ed. (Oxford, 2005). ↩
  2. On anonymity, dating and Markan priority, see Ehrman, Jesus, Interrupted (HarperOne, 2009), chs. 3–4, and the standard critical introductions. ↩
  3. Athanasius, Thirty-Ninth Festal Letter (A.D. 367). On the disputed books and the competing literature, see Ehrman, Lost Christianities (Oxford, 2003). ↩
  4. On the exodus and conquest, see Israel Finkelstein and Neil Asher Silberman, The Bible Unearthed (Free Press, 2001). On the census, compare Luke 2:1–2 with Matthew 2:1; Quirinius became legate of Syria in A.D. 6, and Herod died in 4 B.C. ↩

Catholic View

The Catholic answer is not that the objections are baseless. Most of the facts in the left panel are facts, and a defense that denies them will not survive contact with anyone who has read the field. The answer is that the facts have been fitted to the wrong picture — a picture in which the Bible is a single book dropped complete out of heaven, whose credibility fails if any part of it can be questioned. That is not the Catholic doctrine of Scripture, and it never was.

What Is Granted

No autographs survive. The variants are real and numerous, and the longer ending of Mark and the woman taken in adultery are almost certainly later additions — which is why modern Catholic Bibles print them with a note saying so. The Gospels do not name their authors in the text. Matthew and Luke did use Mark. The resurrection accounts differ in their details. The conquest narratives are morally appalling on their face, and saying so is not a concession wrung out of a Catholic; it is where any honest reading has to start.

The Variants, Counted Properly

The famous figure needs its denominator. There are more variants than words in the New Testament because there are more than 5,800 Greek manuscripts, plus about 10,000 Latin and thousands more in other languages, spanning fifteen centuries. Every spelling slip in every copy counts as a variant. The number is a measure of how much evidence survives, not of how corrupt the text is — a work preserved in two manuscripts has almost no variants, and we know far less about what it originally said.

Sort them and the picture changes. The overwhelming majority are spelling, word order, and slips no translation could register. Textual critics classify only a small fraction as both meaningful and viable — capable of changing the sense and with a real chance of being original. And this is precisely where the author of the sentence at the top of this page agrees: in the appendix to Misquoting Jesus, Ehrman states that essential Christian beliefs are not affected by textual variants. The scholarship that produced the alarming number also produced that verdict.

Compare Like With Like

Ancient history is not done by demanding autographs, because nobody has any. Caesar’s Gallic War survives in perhaps ten good manuscripts, the earliest some nine hundred years after he wrote. The Annals of Tacitus depend on two. Nobody proposes that we therefore know nothing of Roman Gaul. Against that standard the New Testament is an embarrassment of riches: thousands of witnesses, fragments within a century of composition, and a spread of independent transmission lines that lets scholars triangulate what the earliest text said. Applying to this one book a standard no ancient document could meet is not rigor. It is a special rule.

The Canon Was Recognized, Not Invented

The skeptical account of the canon is nearer the Catholic one than it is to the Protestant one, and it proves something other than it intends. Yes: the Church decided. There is no table of contents in Scripture, no inspired list of inspired books, and no way to get the canon out of the Bible without already having the canon.

But the decision was a recognition of what the churches were already reading. The four Gospels and the Pauline letters are being used as authoritative long before any council; the Muratorian fragment, around 170, already lists most of the present New Testament, and Irenaeus argues for exactly four Gospels in the 180s. The books eventually disputed were the short ones at the edges, not the substance. The apocryphal gospels were not suppressed rivals of equal standing; most are second century or later and read like it.

What follows is uncomfortable for one side of Christianity rather than for the faith itself. If the Church was competent to identify the Scriptures, she is not a body established by them, and the appeal to the Bible against the Church saws off the branch it stands on. The Scriptures are the Church’s book, and this is the oldest fact about them.

What Inspiration Actually Claims

Here the objection usually attacks a doctrine Catholics do not hold. The Church does not teach that the Bible is a divinely dictated encyclopedia of history and science whose every sentence is a literal report. The Second Vatican Council put the claim carefully: the Scriptures teach faithfully and without error that truth which God wanted put into the sacred writings for the sake of our salvation. The qualification is the whole point, and it is not a modern retreat — Augustine wrote that the Spirit did not intend to teach men things of no profit for salvation, and Leo XIII and Pius XII made the principle formal in 1893 and 1943.

With that in hand, the reader’s first duty is to determine what a given author was doing. Genesis 1 is a liturgical poem about who made the world and why, not a chronology; Job is a drama; the Psalms are songs; Revelation is apocalyptic, a genre with its own conventions its first readers knew. Treating each as a police report and then declaring it false is a mistake about the text, not a discovery about it.

Differences That Look Like Testimony

The variations among the resurrection accounts are the usual exhibit, and they are real. But consider what a fabricated, coordinated story looks like: it agrees on details and diverges on nothing. What we have instead is four accounts that agree on the core — the tomb was empty, women found it first, the disciples did not expect it — and differ on peripheral detail. That the first witnesses are women is itself hard to explain as invention, in a culture where a woman’s testimony carried little legal weight; it is the last detail anyone would choose. Divergence in the margins with agreement at the center is what real testimony looks like, and every police officer knows it.

The Passages That Should Trouble Us

The conquest narratives cannot be tidied away, and attempts to do so are worse than the problem. Several things can be said honestly. The literary conventions of ancient Near Eastern war reporting were hyperbolic, and the same book that says everything was destroyed has the population alive a chapter later. Revelation was progressive: God met a people where they were and led them, and the trajectory runs toward the Sermon on the Mount rather than away from it. And the Christian reads the whole collection through its center, which is Christ — who takes the ancient law and says you have heard it said, but I say to you, which is not the language of a man endorsing everything that came before.

That does not dissolve the difficulty, and this page will not pretend that it does. It locates it. The question is not whether a Christian may be troubled by Joshua. It is whether being troubled by Joshua is a reason to think the whole collection uninspired — and it is not, unless one first assumes the flat, dictated model of inspiration that the Church rejects.

The Reply in Brief

The variants are numerous because the manuscripts are numerous, and the man who made the number famous concedes in the same book that no essential belief hangs on a disputed reading. Judged by the standard applied to every other ancient text, the New Testament is the best-attested document of the ancient world by a wide margin. The canon was indeed settled by the Church, which is an argument against the Bible standing alone rather than against the Bible.

And inspiration has never meant dictation. Scripture teaches without error the truth God willed for our salvation, in the genres its authors were actually writing; read that way, the poetry is not failed physics and the ancient war reporting is not a divine endorsement. What the skeptic has usually refuted, carefully and at length, is a doctrine of Scripture the Church does not teach.

The full essay works through the manuscript evidence and how textual criticism actually reaches its conclusions, sets out the formation of the canon from Irenaeus to Carthage, examines Dei Verbum on inerrancy and the salvation clause alongside Divino Afflante Spiritu on genre, takes up the census of Quirinius and the exodus at their hardest, and treats the conquest narratives at the length they deserve.

Summary

The case against the Bible is largely a case against a doctrine of Scripture the Catholic Church does not hold — that the book is a divinely dictated encyclopedia which fails entire if any part of it can be questioned. What the Church actually teaches survives the objections, and several of them turn out to point the other way.

The variants are numerous because the evidence is. There are more differences among the manuscripts than words in the New Testament — because there are over 5,800 Greek manuscripts and thousands more in other languages, spanning fifteen centuries, and every spelling slip counts. A text preserved in two copies has almost no variants, and we know far less about what it said. The overwhelming majority are spelling and word order; only a small fraction are both meaningful and viable.

And the scholar who made that fact famous concedes the point. In the appendix to Misquoting Jesus, Bart Ehrman grants that essential Christian beliefs are not affected by textual variants. The same scholarship that produced the alarming number produced that verdict, and honesty requires quoting both.

By the standard applied to any other ancient text, this one is exceptional. Caesar’s Gallic War survives in about ten good manuscripts, the earliest some nine centuries after writing; the Annals of Tacitus depend on two. Nobody concludes that Roman Gaul is unknowable. Demanding of the New Testament a standard no ancient document could meet is a special rule, not a rigorous one.

The canon was settled by the Church — which proves something other than intended. There is no inspired table of contents; the collection was recognized by the body already reading it, with the four Gospels and Paul functioning as authoritative long before any council. If the Church was competent to identify the Scriptures, then an appeal to the Bible against the Church cuts the branch it sits on. The argument tells against Scripture standing alone, not against Scripture.

Inspiration has never meant dictation. The Second Vatican Council teaches that the Scriptures convey without error “that truth which God wanted put into the sacred writings for the sake of our salvation.” That qualification is not a modern retreat — Augustine said the Spirit did not intend to teach what does not profit salvation, and Leo XIII and Pius XII made the principle formal in 1893 and 1943. Genesis 1 is a liturgical poem, Job a drama, Revelation an apocalypse. Reading each as a police report and declaring it false is an error about the text.

The differences look like testimony, not fabrication. A coordinated invention agrees on details. The resurrection accounts agree at the center — empty tomb, women first, disciples not expecting it — and diverge at the margins, which is what real witnesses do. That women are the first witnesses is the last detail anyone inventing a story in that culture would have chosen.

And the hard passages stay hard. The conquest narratives are appalling on their face. Ancient war reporting was hyperbolic, and the same books describe as annihilated populations that reappear chapters later; revelation was progressive, running toward the Sermon on the Mount; and the Christian reads the whole through Christ, who said “you have heard it said, but I say to you.” That locates the difficulty rather than dissolving it, and pretending otherwise would be worse than the problem.

In short: the Bible is a library, not a monolith, and the Church that assembled it is older than it. Read for what its authors were actually doing, it is the best-attested document of the ancient world and the book of a living community — not a code dropped from the sky, and not refuted by the discovery that it isn’t one.

For a more complete essay on this topic, download What Is the Bible? (.docx)