Feature

AI pegs Dead Sea Scrolls' age at 2,300 years

AI pegs Dead Sea Scrolls' age at 2,300 years

The Dead Sea Scrolls, among the most celebrated archaeological discoveries in history, have long captivated scholars and the public alike for their profound insights into ancient Judaism, early Christianity, and the origins of the Bible. 

A ground-breaking new study, published in the journal PLOS One, leverages cutting-edge artificial intelligence (AI) and advanced radiocarbon dating to reveal that many of these iconic manuscripts may be up to 100 years older than previously estimated. 

This shift, pushing some scrolls back to around 2,300 years ago, aligns them closer to the time of their presumed authors and opens transformative possibilities for understanding the cultural, religious, and historical contexts of these ancient texts. Led by Mladen Popović, dean of the Faculty of Religion, Culture and Society at the University of Groningen, this research marks a pivotal moment in biblical scholarship and manuscript analysis, reports CNN.

Dead Sea Scrolls: A window into the past

Discovered by chance in 1947 by Bedouin shepherds in the caves near Khirbat Qumran in the Judaean Desert (modern-day West Bank), the Dead Sea Scrolls consist of thousands of fragments from approximately 1,000 manuscripts, recovered from 11 caves. 

Over 200 of these manuscripts contain the earliest known copies of the Hebrew Bible (Old Testament), alongside other religious, legal, and literary texts written primarily in Hebrew, with some in Aramaic and Greek, on parchment and papyrus. 

These documents, dating roughly from the third century BCE to the second century CE based on earlier palaeographic estimates, have reshaped our understanding of ancient religious traditions.

“The Dead Sea Scrolls were a game-changer when they were found,” Popović explains. “They gave us the oldest physical evidence of the Hebrew Bible, revealing what the text looked like 2,000 years ago and offering a snapshot of what people were reading, writing, and thinking during a formative period for Judaism and Christianity.” 

He describes the scrolls as a “time machine,” providing tangible evidence of a critical historical juncture that shaped one of the world’s most influential texts, the Bible. This makes them invaluable not only for religious scholars but also for anyone interested in the cultural evolution of Western civilization.

Challenge of dating the scrolls

Unlike modern documents, the Dead Sea Scrolls rarely include explicit dates, leaving scholars to rely on palaeography—the study of ancient handwriting styles—to estimate their age. Palaeographic analysis examines letter forms, writing techniques, and script evolution to place manuscripts within broad timeframes, previously pegging the scrolls between the third century BCE and the second century CE. However, palaeography is subjective and often yields wide date ranges, limiting precision.

Radiocarbon dating, pioneered by chemist Willard Libby in the 1940s, offered a scientific alternative by measuring the decay of carbon-14, a radioactive isotope absorbed by all living organisms. As carbon-14 decays at a predictable rate after death, this method can date organic materials up to 60,000 years old. Early radiocarbon tests in the 1990s provided some clarity but were flawed due to contamination from castor oil, applied in the 1950s to enhance text legibility. This modern contaminant skewed results toward more recent dates, undermining their reliability.

A dual approach: Modern radiocarbon and AI innovation

Popović’s team tackled these limitations with a two-pronged strategy. First, they conducted new radiocarbon dating on 30 manuscripts using updated techniques to account for contaminants like castor oil. The results were striking: most manuscripts were older than previously thought, with some dating to the late fourth century BCE—potentially 100 years earlier than palaeographic estimates. Only two manuscripts were found to be younger.

To extend these findings non-destructively, the researchers developed an AI model named Enoch, after the biblical figure associated with wisdom and longevity. Enoch was trained on high-resolution images of the newly dated manuscripts, learning to identify subtle visual patterns—such as ink density, stroke thickness, and material degradation—that correlate with age. When tested on manuscripts with known radiocarbon dates (but withheld from Enoch), the AI correctly estimated their age 85 per cent of the time, often providing narrower date ranges than radiocarbon alone.

The team then applied Enoch to 135 undated Dead Sea Scrolls. Scholars evaluated the AI’s estimates against palaeographic expertise, rating 79 per cent as “realistic.” 

For example, a scroll containing verses from the Book of Daniel, previously dated to the second century BCE (a generation after its presumed author), was reassigned to the author’s era. Similarly, an Ecclesiastes manuscript, once thought to date from 175–125 BCE, was re-dated to 300–240 BCE. 

These shifts suggest that some scrolls were created closer to the lifetimes of their authors, offering new insights into their composition and circulation.

Implications for history and scholarship

The revised dates have far-reaching implications. By placing the scrolls closer to their authors’ time, researchers can better explore how these texts were written, shared, and interpreted within and beyond their original communities. 

“This opens up new possibilities to think about how these texts moved to other users and readers,” Popović notes, highlighting potential networks of scribes, scholars, and early religious groups. The findings could reshape historical reconstructions of Second Temple Judaism (circa 516 BCE–70 CE) and the emergence of Christianity, shedding light on the Bible’s development as a cultural and religious cornerstone.

Moreover, the study underscores the scrolls’ role as a bridge across faiths and disciplines. “Whether you’re Christian, Jewish, or don’t believe at all, the Bible is one of the most influential books in history,” Popović says. 

“The scrolls let us study it as a form of cultural evolution, showing how ideas were transmitted across generations.”

AI as a game-changer for manuscript analysis

Enoch’s success points to a future where AI could transform the study of ancient manuscripts. Unlike radiocarbon dating, which requires destroying a small sample (about 7 milligrams), Enoch analyses high-resolution images non-invasively, preserving fragile artefacts. “With Enoch, you don’t have to cut off a piece of a Dead Sea Scroll,” Popović emphasises. “This is a first step, with all sorts of possibilities to improve further.”

The AI’s potential extends beyond the scrolls. Popović envisions adapting Enoch to analyse other ancient scripts, such as Syriac, Arabic, Greek, and Latin, potentially revolutionizing fields like papyrology and epigraphy. By training on larger datasets, Enoch could achieve even greater accuracy, possibly supplanting radiocarbon dating for certain applications.

Expert reactions and future directions

Scholars unaffiliated with the study have praised its innovative approach. Charlotte Hempel, a professor at the University of Birmingham, highlights the synergy between AI and enhanced radiocarbon dating, noting that AI’s narrower date ranges suggest unprecedented precision. 

Lawrence H Schiffman of New York University calls the study a “pioneering” effort to extend scientific dating to undated manuscripts, though he cautions that further validation is needed for broader application. 

Brent Seales, a computer science professor at the University of Kentucky, commends the rigorous methodology, likening AI’s potential to a fine wine that improves with more data. However, he warns that replacing radiocarbon entirely may be premature, as AI relies on patterns from limited samples.

The study’s small sample size—30 carbon-dated manuscripts for training—remains a limitation, but its success with 135 undated scrolls demonstrates scalability. 

Future research could expand Enoch’s training dataset, incorporate additional imaging technologies (e.g., multispectral imaging), or integrate other AI models to cross-reference palaeographic and material data.

A new era for ancient texts

This study marks a “massive step forward” in dating ancient manuscripts, as Seales puts it, blending data-driven science with historical inquiry. By refining the timeline of the Dead Sea Scrolls, it not only deepens our understanding of their historical context but also showcases AI’s transformative potential in archaeology. 

As Enoch evolves, it could unlock new secrets from the past, preserving fragile artefacts while illuminating the origins of texts that have shaped human history.

Source: CNN