A groundbreaking study led by researchers from Duke University has leveraged artificial intelligence (AI) to explore the authorship of some of the oldest texts in the Bible. By examining the first nine books of the Hebrew Bible, known as the Enneateuch, the team uncovered distinct language patterns that suggest multiple authors or scribal groups contributed to these sacred writings.
The research team, which includes mathematicians, linguists, archaeologists, and computer scientists, employed advanced linguistic analysis alongside a bespoke AI model. Their findings reveal three unique writing styles within the texts, indicating that different authors influenced the composition of the Bible.
Thomas Römer, a professor at Collège de France, noted the significance of their findings: “We found that each group of authors has a different style—surprisingly, even regarding simple and common words such as ‘no,’ ‘which,’ or ‘king.’ Our method accurately identifies these differences.” This revelation may offer fresh insights into longstanding debates regarding the authorship of biblical texts.
The project, spearheaded by Shira Faigenbaum-Golovin, began in 2010. The initial focus was on analysing handwriting from ancient pottery fragments dated to around 600 BC. Faigenbaum-Golovin explained, “We concluded that the findings in those inscriptions could offer valuable clues for dating texts from the Old Testament.” With the formation of a collaborative team, the researchers aimed to merge scientific methodology with traditional biblical scholarship.
Through their AI-based analysis, the team identified three distinct literary styles: the Priestly source, the Deuteronomistic History, and the Book of Deuteronomy. While many chapters aligned clearly with these categories, the researchers were surprised to find that certain sections of the Ark Narrative in 1 Samuel did not fit any of the identified styles. This discrepancy raises new questions about the Bible’s composition and suggests further layers of complexity in its authorship.
The researchers are optimistic that their methodology can be applied to other historical manuscripts. Faigenbaum-Golovin remarked, “If you’re looking at document fragments to find out if they were written by Abraham Lincoln, for example, this method can help determine if they are real or just a forgery.”
The team articulated their findings in a paper published in the journal PLOS One, stating that their work provides “interpretable, statistically significant evidence” of linguistic characteristics that can be identified automatically. This new approach offers a valuable tool for addressing contentious issues in biblical studies.
As the researchers continue to explore the potential applications of their findings, Faigenbaum-Golovin expressed her enthusiasm for the collaboration between science and the humanities: “It’s such a unique collaboration; it’s a surprising symbiosis, and I’m lucky to work with people who use innovative research to push boundaries.” The implications of this study have the potential to reshape our understanding of ancient texts and their authorship, opening new avenues for research and discussion in the field of biblical scholarship.
