In a climate-controlled laboratory in the Netherlands, a machine-learning algorithm named Enoch is quietly redefining the intellectual landscape of ancient Judea.
Developed by researchers at the University of Groningen, this AI model, trained on the microscopic curves and strokes of ancient Hebrew script, has begun to challenge decades of scholarly consensus about the age and origins of the Dead Sea Scrolls.
Discovered in 1947 in the caves of Qumran near the Dead Sea, the scrolls have long been hailed as one of the most significant archaeological finds of the 20th century. Comprising nearly a thousand manuscripts, including the earliest known copies of several Hebrew Bible texts, the scrolls offer a rare glimpse into the religious and cultural life of Second Temple Judaism. But for all their importance, their precise dating has remained elusive until now.
A New Chronology Emerges
Traditionally, scholars relied on palaeography, the study of ancient handwriting, to estimate the age of the scrolls. This method, while valuable, is inherently subjective and limited by the absence of securely dated comparative material. Radiocarbon dating, introduced in the 1990s, offered a more empirical approach but was hampered by contamination concerns and the limited number of fragments evaluated.
Enter Enoch. Named after the apocalyptic figure whose writings appear among the scrolls, the AI model combines radiocarbon dating with advanced handwriting analysis to produce more accurate, empirically grounded estimates. Using a Bayesian ridge regression method, Enoch was trained on 24 scrolls with known carbon-14 dates and then applied to 135 previously undated manuscripts. The results were striking.
According to the Groningen team, led by Professor Mladen Popović and data scientist Maruf Dhali, many scrolls are significantly older than previously believed. The Hasmonaean script, once thought to have emerged around 150 BCE, may date back to 200 BCE. Herodian script styles, similarly, appear to have originated decades earlier than assumed. These findings suggest that different scribal traditions overlapped for longer periods, complicating the linear models of script evolution that have dominated scholarly discourse.
Implications for Authorship and Transmission
Perhaps most provocatively, the study found that two biblical scroll fragments, one from the Book of Daniel and another from Ecclesiastes, may have been copied during the lifetimes of their traditionally attributed authors. Fragment 4Q114, for instance, was previously dated to the late 2nd century BCE. Enoch’s analysis places it between 230 BCE and 160 BCE, potentially aligning it with the historical Daniel’s era.
This revelation challenges long-held assumptions about the transmission of sacred texts. If compositions were being copied within a single generation, it implies a level of literacy and textual infrastructure previously unacknowledged in ancient Judea. It also suggests that scripture was not solely the domain of religious elites but may have circulated more widely among lay communities.
“The idea that these texts were copied so close to their time of composition changes how we think about authorship, authority, and access,” said Popović in a recent interview. “It points to a more dynamic and decentralised textual culture.”
Reassessing the Qumran Community
The new chronology also casts fresh light on the enigmatic Qumran community, often associated with the Essenes, a Jewish sect known for its asceticism and apocalyptic beliefs. If nearly a third of radiocarbon dates diverge from palaeographic assessments, as the Groningen study suggests, then the timeline of the community’s activity may need revision.
This has broader implications for understanding the political and religious landscape of Judea in the centuries leading up to the destruction of the Second Temple in 70 CE. The overlap of scribal styles, for instance, may reflect periods of cultural hybridity rather than clear-cut transitions. It also raises questions about the role of Qumran as a centre of textual production versus a repository of diverse traditions.
AI as a Scholarly Tool
While the use of AI in archaeology and textual studies is still in its infancy, the success of Enoch marks a turning point. By analysing ink-trace patterns at both the microlevel curvature and texture and macrolevel character shapes , the model offers a quantitative basis for dating that traditional palaeography cannot match.
The researchers employed BiNet, a deep neural network developed in-house, to digitise and analyse the manuscripts. This allowed for geometric shape analysis of the handwriting, providing a level of precision previously unattainable. Combined with radiocarbon dating, the model bridges the chronological gap between the fourth century BCE and the second century CE, a span that encompasses the entire Dead Sea Scrolls corpus.
“This technology is like a time machine to the ancient world,” Popović remarked. “It allows us to see patterns and connections that were invisible before.”
A New Chapter in Jewish History
The implications of this research extend beyond the academic. By redating the scrolls, scholars are not merely adjusting footnotes, they are rethinking the foundations of Jewish intellectual history. The proximity between composition and copying suggests a more immediate engagement with sacred texts, one that blurs the lines between author, scribe, and reader.
It also invites a reconsideration of cultural identity and authority. If scripture was accessible outside formal religious institutions, then the dynamics of power and literacy in ancient Judea were more complex than previously thought. The scrolls, long seen as relics of a vanished world, now speak more directly to the lived experience of their creators.
As AI continues to reshape the humanities, the Dead Sea Scrolls stand as a testament to the power of interdisciplinary collaboration. Science and tradition, far from being adversaries, have joined forces to illuminate the past, thereby opening new pathways to understanding the origins of faith, culture, and community.
References
1. University of Groningen Official Announcement
- Title: Dead Sea Scrolls older than previously thought
- Summary: Details the development of the AI model “Enoch” and its integration of radiocarbon dating with palaeographic analysis. Highlights the re-dating of key biblical fragments and the implications for Jewish textual history.
2. SciTechDaily Coverage
- Title: New AI Model Reveals Biblical Dead Sea Scrolls Are Far Older Than We Thought
- Summary: Offers a digestible overview of the research, including technical details about the BiNet neural network and Bayesian ridge regression used in the Enoch model.
3. Archaeology Magazine Feature
- Title: AI reveals Dead Sea Scrolls may be older than previously thought
- Summary: Discusses the broader implications of the findings, including the reassessment of Qumran community timelines and the potential for AI to reshape historical narratives.
Be First to Comment