How the Bible was written

The first adjustment is one of scale. The Bible gathers historical narrative, legal code, poetry, proverb, prophetic oracle, ancient biography, personal letter, circular letter and visionary literature. These are genres that in any other library would sit on different shelves — and that were bound together here by a decision later than the texts themselves.

The distance in time between the two ends is wide enough to separate worlds. The oldest material of the Old Testament and the last writings of the New are separated by more than a thousand years, and a good share of reading difficulties comes from exactly there: someone who reads an eighth-century-BC oracle and a first-century letter with the same expectation of language will find both strange.

It is worth naming the practical consequence, because it runs through the rest of this article: there is no “the author of the Bible” in the singular, and there is no single process of composition. There are processes, plural, and they differ from book to book.

Most of the Old Testament was written in Hebrew. Some passages — chapters of Daniel and Ezra, and smaller stretches — are in Aramaic, the lingua franca of the Middle East after the exile. The whole New Testament reached us in Koine Greek, the common Greek of the first-century Mediterranean, which is not the literary Greek of the classics but the language of trade, administration and conversation.

Some of the deuterocanonical books were written in Greek from the start; of others, such as Sirach, textual evidence shows there was a Semitic original, and the book’s own prologue describes the work of rendering it. In other words: translation is not something that happened after the Bible was finished. It is part of the composition history of some of the books.

This explains a phenomenon that confuses many readers: expressions that sound strange in English are not badly translated, they are translated. A language carries a way of thinking, and there are Hebrew constructions no European language reproduces without losing something. Every edition you open is the result of thousands of choices about what to preserve and what to give up.

Much of the biblical material circulated by word of mouth before being fixed in writing, and that holds both for ancient Old Testament narratives and for the sayings of Jesus, which were handed on for decades before the first Gospel was drafted. It is one of the few points on which scholars of very different positions agree.

It is worth undoing the most common misunderstanding. Oral transmission, in a culture that depended on it, is not the game of telephone at a children’s party. It was a craft, with memorisation techniques, fixed formulas, rhythm, repetition and communal control — the group corrected anyone who told it wrong, because that was their archive. Societies like that retained volumes of text that would look impossible to hold today.

And it is worth not exaggerating in the other direction either. Oral transmission is not recording. It preserves the core well and is more flexible in the frame — the order of episodes, scenic detail, the join between one scene and the next. That is one reason the Gospels tell the same scenes in different sequences without any of them being in error: they are working with material that already came organised by criteria that are not chronological.

The image of the prophet writing his own book alone is rare in the text. What the text describes, when it describes anything, is work in pairs: one dictates, another writes, and the second is not an accessory. Scribes were trained professionals, and in the ancient world the secretary who drafted a letter had, in varying degrees, a hand in its formulation.

This is not specialist conjecture: it is recorded inside the Bible, in at least two places where the text stops and reports how it was done. And it is the fact that best explains why books attributed to the same author can vary in style — the hand that writes is not always the same one, even when the voice is.

Before the bound book there was the scroll. A papyrus scroll has a physical limit of length, and that limit conditioned the size of texts: long works were divided into separate volumes, which helps explain why certain books reached us split in two. Consulting a scroll is also slow — you do not open it at an arbitrary point, you unroll to it.

The turning point is the codex: folded and sewn leaves, the format of the modern book. Christians adopted the codex very early and in far higher proportion than the rest of the ancient world, and the consequence is enormous: for the first time it was possible to bind into a single object texts that had previously filled several scrolls, and to locate a passage without running through everything.

The divisions you use to cite a verse are more recent still, and none of them come from the authors. The division into chapters is attributed to work from the early thirteenth century; the verse numbering of the New Testament comes from a printed edition of the mid-sixteenth. They are addresses, created to make reference easier — and that is why a sentence sometimes begins in one verse and ends in the next.

No biblical author wrote knowing he was producing a chapter of the Bible. Paul wrote letters to specific communities with specific problems; the Gospels were written for concrete readers; the prophets spoke to dated situations. The collection came afterwards, and it came first through use: the texts communities read in their gatherings, copied and exchanged among themselves were the ones that consolidated.

There are documented landmarks along the way. A festal letter of Athanasius, in 367, lists the twenty-seven books of the New Testament in the form we know them. Regional councils in North Africa, at the end of the fourth century, approve lists. Centuries later, the Council of Trent, in 1546, formally defines the canon for the Catholic Church, in a context of controversy with the Reformation. Another ancient document frequently cited, the so-called Muratorian fragment, carries a partial list — but its dating is disputed among scholars, with proposals ranging from the second century to the fourth, and it therefore supports less than it is usually credited with.

What those landmarks record, in most cases, is the recognition of something already established by practice, and not a choice made from scratch in a room. That holds for the New Testament, whose core circulated uncontested long before any official list. Where there was real and persistent divergence was at the boundary of the Old Testament — a subject that has its own article on this site.

No original survived. There is no scroll written by Baruch and no sheet on which Tertius took down the letter to the Romans, and this is not a peculiarity of the Bible: there is no autograph of practically any work from antiquity. What exists are copies, and in the biblical case they are numerous — thousands for the New Testament, in Greek and in ancient translations.

The discovery that most changed the picture for the Old Testament was the manuscripts found at Qumran, from 1947 onwards. They pushed the available evidence for several books back by around a thousand years and allowed direct comparison between copies separated by ten centuries. The overall result was considerable stability — with variations, and with cases where the finds agree with the Greek textual tradition rather than the medieval Hebrew one.

Comparing copies is the work of textual criticism, and it is public: critical editions record the variants in notes, in plain sight of anyone who wants to check. The vast majority are irrelevant to meaning — spelling, word order, synonyms. A small set affects the sense of a passage, and a smaller set still involves longer stretches, which modern editions usually flag in a note. None of this is a specialist’s secret, and no Christian tradition depends on denying that it exists.

An honest article on this subject has to end by saying where there is no consensus — and there is plenty of such ground. The authorship and date of composition of the Pentateuch have been debated for more than two centuries, with positions ranging from the traditional attribution to Moses to models of layered drafting across centuries. The unity of the book of Isaiah is the object of a similar debate. The dating of Daniel divides scholars between a much earlier composition and one in the second century BC.

In the New Testament, the dating of the Gospels varies by decades depending on which author you consult, and the authorship of some letters attributed to Paul is debated on grounds of vocabulary and style. There is also debate about which Gospel was written first and what the others used as a source.

Two honest observations close the point. First: these disputes are about the history of composition, not about the content of the text — nobody disputes what Isaiah says, the dispute is about when and through how many hands it was said. Second: anyone presenting any of these questions as settled, for either side, is announcing a certainty the available evidence does not support.