A groundbreaking study led by researchers from Duke University has harnessed artificial intelligence (AI) to analyze the authorship of several books in the Hebrew Bible. This innovative approach aims to illuminate the complex authorship debates surrounding some of the Bible’s oldest texts, providing new insights into the Enneateuch, which comprises the first nine books of the Old Testament.
The research team employed advanced linguistic analysis and a bespoke AI model to identify distinct writing styles within these biblical texts. Their findings revealed three separate literary styles, suggesting that multiple authors or scribal groups contributed to these ancient writings. This evidence adds a new dimension to long-standing discussions regarding the authorship of these sacred texts.
Thomas Römer, a professor at the Collège de France, remarked on the significance of the study, stating, “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 meticulous analysis highlights how AI can uncover subtle linguistic features that may have gone unnoticed in traditional scholarship.
The project, spearheaded by mathematician Shira Faigenbaum-Golovin, commenced in 2010 with the examination of ancient pottery fragments dating back to 600 BC. Faigenbaum-Golovin explained the broader implications of their findings, emphasizing that the inscriptions could provide valuable clues for dating texts from the Old Testament. “When we started putting together our current team, we aimed to analyze these biblical texts using a scientific methodology,” she added.
The AI system categorized the literary styles into three distinct sources: the Priestly source, the Deuteronomistic History, and the Book of Deuteronomy. While certain chapters fit neatly into these classifications, the researchers encountered an unexpected challenge. Portions of the Ark Narrative in 1 Samuel did not align with any of the identified styles, suggesting the presence of an additional, unidentified author or scribal group.
As the research progresses, the team believes that their approach could be applied to other historical manuscripts. Faigenbaum-Golovin noted, “If you’re looking at document fragments to determine whether they were written by Abraham Lincoln, for example, this method can help assess their authenticity.”
In their published study in the PLOS One journal, the researchers asserted, “Our paper sheds new light on the question of authorship of biblical texts by offering interpretable, statistically significant evidence of the existence of linguistic characteristics in the writing of biblical authors/redactors, that can be identified automatically.” This methodology provides a new tool for addressing contentious issues within biblical studies.
Looking ahead, the academic team is eager to explore how this innovative approach can yield fresh insights into other ancient manuscripts. Faigenbaum-Golovin expressed her enthusiasm for the collaboration, stating, “It’s such a unique collaboration between science and the humanities. It’s a surprising symbiosis, and I’m lucky to work with people who use innovative research to push boundaries.”
This research not only enhances the understanding of biblical authorship but also illustrates the potential of AI in uncovering the complexities of historical texts.
