Field Notes: How Humans Shaped Sacred Texts

Enough to follow the conversation, and to know when someone's overselling it.

Field Notes are short guides to subjects I've either worked in or simply couldn't stop reading about. Each one gets updated as the subject moves, and the full collection lives here.

Last updated: 9 November 2026. First edition.

Why I got curious

Many years ago I set myself a personal project for the year: figure out where the Christian Bible actually came from. I expected a dry history of councils and manuscripts. What I found instead were translation errors, some accidental and some not, gospels that didn't make the cut, and a surprising amount of church politics. I was hooked. So I did the same exploration for Judaism, Islam, Hinduism, and Buddhism, and a couple of years ago that work became the Sacred Editors series.

This Field Note is the short version: what I learned across all five traditions, and the patterns that surprised me most.

What does it mean to edit scripture?

I should say up front what I mean by editing, because the word can sound like an accusation, and I mean it more neutrally than that. For believers in every tradition here, a sacred text originates with God or with an enlightened teacher. My interest is the human side of the story: the people who memorized, wrote down, copied, selected, translated, and standardized those texts so that they reached us at all.

Across all five traditions, people have shaped sacred texts in roughly five ways:

From memory to writing. Most scripture lived first as oral recitation, sometimes for centuries. Writing it down meant deciding which version to record.
Choosing the canon. Someone had to decide which texts were in and which were out. A canon is that official list.
Copying. Before printing, every copy was made by hand. Scribes made accidental errors, and occasionally deliberate "corrections."
Translation. Every translation is a series of decisions about meaning, and some of those decisions shaped doctrine for centuries.
Standardization. Eventually communities settled on one authoritative version, often helped by a new technology like vowel marks or the printing press.

Every tradition used all five. What differs, and what I find most interesting, is when each one happened and how much room it left for variation.

How did we get here?

Each tradition took its own path. The dates below are approximate, and scholars debate many of them.

Hinduism

When the text stabilized: the sound of the Vedas very early; the epics and Puranas kept growing for centuries.

Hinduism holds the most remarkable preservation story I came across. For more than two thousand years, students memorized the Vedas using recitation patterns that repeat the words in different orders, so that a single slip would be caught. The goal was to preserve the exact sound, pitch accent included, because the sound itself was considered sacred. Medieval sources name hundreds of Vedic recitation schools, each with its own branch of the tradition, and most of them are now lost.

The epics went the other direction. The Mahabharata and Ramayana grew and branched over centuries into regional versions, and when scholars in Pune produced a critical edition of the Mahabharata in the 20th century, they had to work through hundreds of manuscripts. In Hinduism, then, the oldest core was fixed in sound while later texts kept expanding.

Judaism

When the text stabilized: with the Masoretes, roughly the 6th to 10th centuries CE.

Judaism went from plurality to precision. The Dead Sea Scrolls, found starting in 1947, showed that around the time of Jesus several forms of the biblical text circulated at once. The Septuagint, a Greek translation made for Jews in Alexandria, made its own interpretive choices that Christians would later inherit.

Then, between roughly the 6th and 10th centuries, scholars called the Masoretes added vowel marks, chanting marks, and marginal counts of words to the consonantal Hebrew text. The Leningrad Codex, completed in 1008, is still the base text for modern scholarly Hebrew Bibles.

Buddhism

When the text stabilized: the Pali Canon was written down in Sri Lanka in the 1st century BCE; the Chinese and Tibetan canons came later.

Buddhism never settled on a single canon, and the tradition largely made peace with that. The teachings were recited in monastic communities for roughly four centuries before monks in Sri Lanka wrote the Pali Canon on palm leaves in the 1st century BCE.

As Buddhism spread, Chinese and Tibetan translators built their own large canons, which include texts the Pali tradition doesn't recognize. A sealed cave library at Dunhuang in western China, opened around 1900, held tens of thousands of manuscripts, including a copy of the Diamond Sutra dated 868, the oldest dated printed book.

Christianity

When the text stabilized: the canon was broadly settled by the late 4th century and fixed for Catholics at the Council of Trent in 1546; the printing press standardized the text.

Christianity is where my own curiosity started, and it's the tradition where translation did the most work. Early Christians read the Hebrew scriptures mostly in Greek, so when Isaiah 7:14 described a "young woman" ('almah), Matthew quoted the Septuagint's "virgin" (parthenos), and Jerome kept "virgin" (virgo) in his Latin Vulgate even though he knew the Hebrew. Jerome also rendered metanoia, a change of mind, as paenitentia, a word that leaned toward acts of penance.

Copyists added passages too. The story of the woman caught in adultery and the longer ending of Mark are missing from the earliest manuscripts, and a verse about the Trinity, the Comma Johanneum, entered printed Greek New Testaments in the 1520s. The upside of all that copying is evidence: thousands of Greek manuscripts survive, so scholars can trace these changes in detail.

Islam

When the text stabilized: Uthman's standard codex, around 650 CE, about two decades after the Prophet's death.

Islam's story differs in timing more than in kind, and I want to be careful here. Muslims believe the Qur'an is the word of God, revealed to the Prophet Muhammad and preserved without alteration. The historical record shows a text that was gathered and standardized remarkably early.

According to Islamic tradition, the first caliph, Abu Bakr, had the written portions collected after many reciters died in battle, and around 650 CE the third caliph, Uthman, issued a standard codex and ordered other copies destroyed. Since then the consonantal text has stayed remarkably consistent for nearly fourteen centuries, and Sunni and Shi'a Muslims today recite the same text.

The human choices continued after Uthman. Early Arabic script lacked most vowel and dot marks, so readers depended on trained reciters, and scholars came to recognize seven, and later ten, canonical qira'at, or readings, with small differences in wording and pronunciation. The Sana'a palimpsest, found in Yemen in 1972, has an erased lower text from outside the Uthmanic standard, with many variants, most of them minor in meaning.

Two leaves at the University of Birmingham were radiocarbon dated to between 568 and 645 CE, and their text closely matches today's Qur'an. In 1924 a printed edition in Cairo, based on the Hafs reading, became the global norm. My Islam volume calls the result curation rather than corruption, and I still think that's the fairest summary.

Comparing the five

Side by side, the clearest pattern is timing. Islam did its gathering and standardizing within about two decades. Judaism's text was fixed by the Masoretes many centuries after its books were written. Christianity's canon and text settled gradually over centuries, with translation playing an unusually large role. Hinduism fixed the Vedas in sound while other texts stayed fluid, and Buddhism accepted several canons. I don't read any of this as a ranking. Each tradition made choices that fit its own needs, and each preserved its texts with real care.

What was lost along the way?

What didn't survive is part of the story too. Fire and war destroyed collections in every tradition. Nalanda, one of the great Buddhist monastic universities in India, was sacked around 1200, and its library was lost. Other texts were simply set aside. The Nag Hammadi codices, buried in Egypt and found by farmers in 1945, preserved Christian writings like the Gospel of Thomas that never made the canon. One of the more surprising threads in my research was how often texts survived in another tradition's hands: Muslim scholars, and Christians living under Islamic rule, kept copies of some Christian writings that were lost or neglected in the Latin West.

Women's voices are the quietest loss. Every tradition had women who taught, composed, and preserved scripture: the Buddhist nuns whose poems survive in the Therigatha, Aisha as a key early transmitter of the Prophet's sayings, Miriam and Huldah as prophets in the Hebrew Bible, and Mary Magdalene in the gospels. Some were later diminished. A sermon by Pope Gregory I in 591 blended Mary Magdalene with an unnamed sinful woman, and her image as a reformed prostitute stuck for well over a thousand years until the Catholic Church separated the figures again in 1969. My Lost Texts and Lost Women volumes go much deeper on both.

Where does it stand today? (November 2026)

Manuscript scholarship is in the middle of a technology shift. Multispectral imaging recovers erased and faded writing, which is how scholars read the lower text of the Sana'a palimpsest. Radiocarbon dating has tightened the timelines for early manuscripts in several traditions. In 2025, a team led by the University of Groningen published Enoch, a model that combines new radiocarbon dates with AI analysis of handwriting. It suggested that many Dead Sea Scrolls are older than handwriting experts had assumed, and it can estimate dates without cutting into the scrolls.

Digitization has also put the evidence online. Projects like the Leon Levy Dead Sea Scrolls Digital Library and Corpus Coranicum let anyone look at manuscripts that once required a trip to a vault.

There's a twist here I find fascinating. Technology is uncovering diversity in the old texts, but it's also standardizing what ordinary readers see. As my Islam volume notes, most Qur'an apps default to the Cairo edition and the Hafs reading, so the variety scholars are recovering often never reaches the person reading on a phone.

What's hype and what's real?

Sacred texts attract more confident misinformation than almost any subject I've studied, from believers and skeptics alike. Here are the claims I hear most often, one or more from each tradition.

"The Council of Nicaea chose the books of the Bible." It didn't. The council in 325 dealt mainly with the nature of Christ and the date of Easter. The New Testament canon took shape gradually, and the first list matching today's 27 books appears in a letter by Athanasius in 367. The Nicaea version owes a lot to popular novels like The Da Vinci Code.

"The church hid the real gospels." Some non-canonical gospels were discouraged or destroyed, and buried copies like those at Nag Hammadi survived. But most were left out because they were later, more regional, or outside the theological mainstream, and scholars have read them openly for decades. Be wary of new "discoveries," too. The "Gospel of Jesus' Wife," a papyrus fragment announced at Harvard in 2012, turned out in 2016 to be very likely a modern forgery.

"The Qur'an is unchanged to the letter" versus "the Qur'an was corrupted." Both extremes miss the evidence. The consonantal text has been remarkably stable since around 650, and early manuscripts like the Birmingham leaves support that. At the same time, Islamic scholarship itself has always recognized multiple canonical readings and early variant codices, so the history includes human choices. Relatedly, the Gospel of Barnabas, sometimes cited as a suppressed original gospel that agrees with Islam, survives only in manuscripts from many centuries after Jesus and is not considered an early text by historians.

"The Dead Sea Scrolls were suppressed because they'd overturn Christianity." Publication was painfully slow, which fed decades of conspiracy theories. When the full set of images became available in 1991, they contained nothing of the kind. Most are Jewish religious texts, including the oldest known copies of the Hebrew Bible.

"The Vedas describe ancient airplanes." The popular claims trace largely to the Vaimanika Shastra, a text that appeared in the early 20th century. The Vedas themselves are hymns and ritual texts, and their real achievement, preservation by sound across millennia, is impressive enough without the embellishment.

"The Buddha said..." Many of the Buddha quotes that circulate online don't appear in any Buddhist canon. Since there are several canons, the honest check is whether a quote can be traced to a specific sutta or sutra.

What happens next?

This field moves slower than quantum computing, but a few things are worth watching. When one of them changes, I'll update this section.

AI on damaged and erased texts. Tools like Enoch for dating, and imaging methods for erased writing, are improving quickly, and there are plenty of unread palimpsests and fragile manuscripts left.
New finds. The big discoveries have mostly come by accident: a shepherd at Qumran, a mosque renovation in Sana'a, farmers at Nag Hammadi. There's no reason to think the accidents are over.
Translation debates. Arguments over gendered language and how to render key terms continue in every tradition that reads scripture in translation.
Digital standardization. Whether apps and online editions preserve the variety scholars have recovered, or flatten it further, may matter more to ordinary readers than any new discovery.

What does this mean for me?

Mostly, it means reading a little more carefully. If you read scripture in translation, which most people do, the translation you choose matters, and comparing two is often eye-opening. A study Bible with textual notes, or a Qur'an translation that explains alternate readings, shows you where the human choices were made.

For believers, I've found this history tends to deepen respect for the texts. People in every tradition spent their lives memorizing, copying, and protecting these words, often at real risk. For skeptics, it's a caution against easy dismissal, because many popular claims about hidden or corrupted texts don't hold up once you look at the actual manuscripts.

It also helps in conversations across faiths. Every tradition faced the same basic problems of memory and language, and each solved them differently. Knowing that makes it easier to talk about the differences with curiosity instead of suspicion.

How to sound smart at dinner

Nicaea didn't pick the books of the Bible. It argued about the nature of Christ and the date of Easter.
"Virgin" in Isaiah was a translation choice. The Hebrew says "young woman," the Greek translation said "virgin," and Matthew quoted the Greek.
Islam standardized early. The Qur'an's standard text dates to about 650, roughly two decades after the Prophet's death, though Muslims also recognize several canonical readings.
The Vedas were preserved by sound. Memorization techniques kept them accurate for well over two thousand years before anyone relied on writing them down.
Buddhism has more than one canon. The Pali, Chinese, and Tibetan collections overlap but differ.
Your Hebrew Bible rests on a manuscript from 1008. The Leningrad Codex is still the base for modern scholarly editions.

Glossary

Canon: the official list of texts a tradition accepts as scripture.
Codex: a bound book of pages, as opposed to a scroll.
Recension: a distinct edited version of a text, like a regional version of the Mahabharata.
Palimpsest: a manuscript page that was scraped clean and reused, sometimes with the earlier text still recoverable.
Textual criticism: the scholarly work of comparing manuscripts to reconstruct the earliest recoverable text.
Septuagint: the Greek translation of the Hebrew scriptures, begun in the 3rd century BCE, which early Christians used as their Old Testament.
Masoretes: Jewish scholars of roughly the 6th to 10th centuries who added vowel and chanting marks to the Hebrew text and standardized it.
Qira'at: the canonical readings of the Qur'an, which differ slightly in pronunciation and wording within the same consonantal text.
Shruti and smriti: in Hinduism, "what was heard" (revealed texts like the Vedas) and "what was remembered" (later traditional texts like the epics).
Apocrypha: texts outside a tradition's canon, sometimes respected and sometimes rejected.