Artificial intelligence is reshaping biblical scholarship by untangling centuries of editorial layering to reveal distinct writing styles within ancient texts, according to a multidisciplinary team of researchers. An international group combining computer science, archaeology, and biblical studies applied machine learning and statistical modeling to the first nine books of the Hebrew Bible, successfully separating distinct textual strata that traditional computational methods struggled to parse.
Overcoming Centuries of Editorial Complexity
Traditional algorithms historically struggled with biblical analysis because ancient texts underwent extensive editing, copying, and redacting over generations. According to the research team, this extreme fragmentation and linguistic layering made standard machine learning models ineffective. To solve this, an interdisciplinary group led by Duke University mathematics adjunct Shira Faigenbaum-Golovin developed a specialized artificial intelligence model capable of detecting subtle linguistic patterns even within short text segments.
Did you know? The foundational research began back in 2010 when Shira Faigenbaum-Golovin started examining Hebrew inscriptions on ceramic fragments from the sixth century BCE in collaboration with Israel Finkelstein from the University of Haifa, ultimately leading to the 2025 AI analysis release reported by Radio ZET and PLOS.
How AI Analyzed Textual Styles in the Hebrew Bible
The study focused on three primary groups of texts: the priestly books of the Torah, the Book of Deuteronomy, and the Deuteronomistic History, which includes Joshua, Judges, Samuel, and Kings. According to the findings, the artificial intelligence model isolated three distinct writing styles by identifying specific recurring words and phrases. The model determined that the Book of Deuteronomy and the Deuteronomistic History share closer linguistic similarities with each other than they do with the priestly texts. In 84 percent of the test cases, the automated AI classifications matched the historical evaluations made by expert biblical scholars.
Future Applications Beyond Biblical Studies
The quantitative model developed for this research holds broad utility outside of theology, according to Shira Faigenbaum-Golovin. The AI framework can identify structural differences and execute rigorous quantitative analysis in fields requiring forensic medicine, archaeology, and general history. Researchers plan to deploy the same statistical model to evaluate other high-value ancient artifacts, including the Dead Sea Scrolls.
Frequently Asked Questions
Who led the artificial intelligence study on biblical texts?
The interdisciplinary research team was led by Shira Faigenbaum-Golovin, an adjunct in mathematics at Duke University.
Which specific books of the Hebrew Bible were analyzed by the AI model?
The analysis covered the first nine books of the Hebrew Bible, specifically focusing on the priestly books of the Torah, the Book of Deuteronomy, and the Deuteronomistic History spanning Joshua, Judges, Samuel, and Kings.
How accurate was the artificial intelligence compared to human scholars?
According to the study results, the automated AI stylistic assignments matched the independent evaluations of expert biblical scholars in 84 percent of the cases.
Join the Discussion
How do you think machine learning will impact future historical discoveries? Share your thoughts in the comments below or subscribe to our newsletter for the latest updates on science and history.