How Ancient Texts Are Dated

What does it mean to date a text?

Each Writings entry keeps three dates apart: the date of the text itself, the date of a surviving copy, and the date the copy was found. The text's date is given as its tradition describes it, such as revealed, heard, or written, and as historians date it. Muslim tradition holds that the Quran was revealed over about 23 years, from 610 CE to Muhammad's death in 632 CE.¹² Radiocarbon tests place the parchment of the Tubingen Quran codex between 649 and 675 CE, with a probability of over 95 percent. The Tubingen library acquired it in 1864, and its age stayed unknown for over a century.²

How does radiocarbon dating work, and why does it give a range?

Radiocarbon dating measures the carbon-14 left in organic material, such as leather, paper, or plant remains, and turns it into a range of calendar years. The method destroys the sample.³ A range is normal: every measurement carries an uncertainty, and because the amount of carbon-14 in nature has varied over time, labs convert it with a calibration curve built from tree rings and other records.³⁴

The test dates the material, not the writing. Oxford's radiocarbon unit placed the parchment of the Birmingham Quran leaves between 568 and 645 CE at 95.4 percent confidence. The library notes that this dates the death of the animal whose skin made the parchment, and that no scientifically reliable way yet exists to date ink.⁵ A Buddhist birch-bark scroll from Gandhara, held by the Library of Congress, was dated to between the first century BCE and the first century CE.⁶

In 1995, a University of Arizona team dated 14 parchment and 4 papyrus samples from caves in the Judean Desert, among them Dead Sea Scrolls. Where handwriting estimates existed, the results were in reasonable agreement with them, apart from one first sample that came out too young and was redone. Of three papyri with written dates, two matched, and the third fell about ten years short.⁷ See Islam, Buddhism, and Judaism.

How does handwriting date a manuscript?

Paleography, the study of old handwriting, dates a manuscript by comparing its handwriting with that of other manuscripts.⁸ The first editor of Papyrus 52, a scrap of John's Gospel in Manchester, dated it to 100 to 150 CE, and the library says recent research points nearer to 200 CE.⁸ J. K. Elliott wrote in 2010, in words Nongbri quotes, that experts are generally agreed that it was written by the mid-second century.⁹ Brent Nongbri notes that the usual date, about 125 CE, rests entirely on handwriting, and argues that the range extends into the third century.⁹ Radiocarbon tests on Dead Sea Scrolls were in reasonable agreement with handwriting estimates where those existed (see above).⁷ See Papyrus 52.

A stone slab from Veal Kantel, Cambodia, records a gift of the Mahabharata, the Ramayana, and a Purana. The surviving text names no year, but its editors judge the lettering characteristic of the seventh century CE.¹⁰ See Hinduism.

What can the physical make-up of a book show?

Codicology is the study of the physical construction and make-up of a book, such as the quires, or groups of folded sheets, that form a codex, a book of sheets sewn along one edge.¹¹ A Quran papyrus in Hamburg is one quire of seven folded sheets, making 28 pages. Double rows of sewing holes along its fold suggest it was a stand-alone booklet.¹²

At least thirty-six folios of one early Mishnah codex are scattered across libraries in at least six cities. Judith Olszowy-Schlanger reconstructs its last quire from those scattered leaves.¹³ See Islam and Judaism.

How does the place where a text was found help?

When an item comes from a dig, the layer, the objects beside it, and the site's history can limit its date. A Greek Gospel fragment excavated at Dura-Europos in 1933 dates to before 256 or 257 CE, according to a Yale manuscript checklist.¹⁴ The city was abandoned after a Sasanian siege and sack in 256 CE.¹⁵ See the Dura-Europos Gospel fragment.

Gabriel Barkay found two silver amulets with biblical Hebrew text in 1979, in a burial cave at Ketef Hinnom, Jerusalem.¹⁶ Erik Waaler reports that the pottery from the burial repository runs from the seventh to the early fifth century BCE and that one amulet lay near the bottom of the deposit. Barkay placed the plaques before the Babylonian exile, and Waaler reports that most scholars date them to the late seventh or early sixth century BCE. Waaler himself dates them to 725 to 650 BCE, and Nadav Na'aman to the early Second Temple period.¹⁷¹⁶

Richard Salomon reports that the British Library's Gandharan scrolls have no known provenance, or place of origin. Thermoluminescence tests on the associated clay pots gave only a broad span, the first to the eighth century CE, while handwriting and language point to the first century or the first half of the second.¹⁸ See Judaism and Buddhism.

How do clues inside a text, and mentions outside it, fix a date?

Internal evidence is what a text or its copy says about itself; external evidence is what other sources say about it. The cases below show each at work.

The Muratorian fragment, a list of New Testament books, says the Shepherd of Hermas was written very recently, in the writer's own time, while Pius was bishop of Rome.¹⁹ Scholars have long dated the list to the end of the second century CE or the start of the third, while Albert Sundberg and Geoffrey Hahneman argued for the fourth.²⁰²¹ See the Muratorian fragment.

Here a mention elsewhere sets a latest date. A slab in a temple of the goddess Dadhimati gives its year as 289 of the Gupta era, which its editor equates with 608 CE. He identifies one of its verses as quoted from the Devi Mahatmya. If he is right, that text existed by 608 CE.²²²³

Oskar von Hinuber considers it likely that some titles in an inscription of Ashoka, an Indian king of the third century BCE, refer to texts now in the Suttanipata, a Pali collection of short discourses. If correct, he writes, this is the oldest Indian inscription evidence for extant Buddhist texts.²⁴ See Hinduism and Buddhism.

How do scholars compare copies to rebuild a text's history?

Textual criticism compares the readings found in surviving manuscripts and ancient translations of one text, with the aim of establishing the text itself.²⁵ The Samaritan Pentateuch agrees with the Greek Septuagint against the Masoretic Hebrew text in about 2,000 places, and most authorities hold that it existed in the third century BCE.²⁶

For the Paippalada Atharvaveda, Michael Witzel studied copyists' errors and pronunciation mistakes from oral transmission. He proposed a lost written ancestor, or archetype, from about the ninth century CE behind both the Kashmir codex and the Odisha palm-leaf copies.²⁷

Papyrus 75 agrees closely with Codex Vaticanus. Handwriting dated it to about 175 to 225 CE, and some scholars took the agreement as a sign that the Vaticanus text existed by the second century. Nongbri argues that handwriting and codicology fit the fourth century, as does Vaticanus, widely acknowledged as a fourth-century codex.²⁸ Tommy Wasserman replies that the closeness to Vaticanus is no reason to date the codex to the fourth century, because the scribes of Vaticanus had access to older copies.²⁹ For the Book of Mormon, the editors of the Joseph Smith Papers, published by the Church Historian's Press of The Church of Jesus Christ of Latter-day Saints, report textual evidence that the editors of the 1837 edition used the printer's manuscript and first edition, not the original manuscript.³⁰³¹ See Papyrus 75 and Latter-day Saints.

How do imaging tools read faded or unopened texts?

Multispectral imaging photographs a manuscript under many wavelengths of light, from ultraviolet to near-infrared, because parchment and inks reflect each wavelength differently. The Sinai Palimpsests Project at St. Catherine's Monastery uses it to recover erased writing from palimpsests, manuscripts that were erased and written over.³² In 2018 the Birmingham library used multispectral imaging with X-ray fluorescence to identify the inks on its Quran leaves, iron gall and red lead.⁵

Virtual unwrapping uses scans to read a scroll that cannot be opened. The En-Gedi scroll, found in a synagogue excavated in 1970, was too burned to open, and Brent Seales's team flattened its layers digitally. The text turned out to be Leviticus, and Emanuel Tov notes that the scroll has been dated to the first centuries CE.³³

How are texts dated that were passed on by memory?

Dates here come indirectly: from a text's language, from references inside it, and from the earliest inscription or manuscript that names it.

Muslim tradition holds that the Quran was both memorized and written down in Muhammad's lifetime, and that every generation since has included many who memorized all of it.¹ A Tubingen University article adds that historians of Islam agree on its oral and written transmission then, though scholars debate when the writing began.²

Vedic tradition holds that the Rigveda is a revealed text, whose hymns were not composed but seen by sages. The Bhandarkar Oriental Research Institute in Pune adds that it was orally transmitted through the ages with little change.³⁴ Jan Heesterman writes that no definite dates can be assigned to the Veda, though shared words and rites with the Iranian Avesta suggest some verse may go back to the twelfth century BCE or earlier.³⁵ The oldest dated manuscript in the institute's set of 30 Rigveda copies is from 1464 CE.³⁴

Theravada tradition holds that the Pali canon was first written down in Sri Lanka in the first century BCE, and holds that Pali is the language the Buddha spoke. Oskar von Hinuber holds that early Buddhist literature arose in a time of pure orality, before writing reached India under Ashoka. He holds that a comparison with the languages of Ashoka's inscriptions points to a western Middle Indic base for Pali, and that Pali is younger than the Buddha's time.²⁴

Rabbinic tradition regards the Oral Law as given to Moses on Sinai.³⁶ Stephen Wald writes that Judah ha-Nasi edited the Mishnah at the beginning of the third century CE, from traditions that began long before.³⁷ Olszowy-Schlanger reconstructs the colophon, or scribe's closing note, of a Toronto leaf as 840 or 841 CE and argues that this makes the codex the leaf belongs to the earliest dated Mishnah manuscript.¹³

Catholic teaching holds that the four Gospels are of apostolic origin: what the apostles preached, they and apostolic men afterward wrote down, selecting from what had been handed on by word of mouth or in writing. The Second Vatican Council's Dei Verbum gives no year for any Gospel.³⁸ The Orthodox Church in America's catechism likewise names the four evangelists as the writers.³⁹ Dale Martin of Yale holds that stories and sayings of Jesus circulated aloud before the Gospels, and that Paul's letters, the earliest about 50 CE, were written before any Gospel. He dates Mark to about 70 CE and John to maybe a year in the 90s, and adds that modern scholars regard all four Gospels as anonymous.⁴⁰⁴¹ See Hinduism, Buddhism, Judaism, and Christianity.

How are modern documents dated?

The Book of Mormon has two kinds of dates: the text's date, which the church and some historians describe differently, and the manuscripts' dates.

These manuscripts carry the English text, not the ancient record the church holds Smith translated. The printer's manuscript, the copy prepared for the printer, was made between about August 1829 and January 1830. The Joseph Smith Papers editors assign 84 percent of the manuscript to Oliver Cowdery and about 15 percent to an unidentified scribe, and count eight paper types. The Reorganized Church of Jesus Christ of Latter Day Saints, now Community of Christ, bought it in 1903, and the Latter-day Saint church acquired it in 2017.³⁰³¹

The editors report that Joseph Smith placed the original dictated manuscript in the Nauvoo House cornerstone on 2 October 1841, so it existed by then.⁴⁴ See Latter-day Saints.

What do these dates show, and what do they leave open?

The date a copy was written is a latest date for its text, not the text's age. Where methods overlap, such as handwriting checked against radiocarbon, scholars can test them against each other.⁷ Where they disagree, the entries say so, with the tradition's view and the critical view side by side.