What if a computer could read the Bible — not for faith, but for authorship clues hidden deep in ancient word patterns? That’s exactly what a bold team of researchers set out to do.
Key Points at a Glance
- AI and statistics reveal distinct writing styles across the Hebrew Bible’s earliest books
- Model distinguishes authorship using variations in common word usage
- The approach works with short, ancient text samples where other methods fail
- Study reshapes how we understand biblical composition and historical authorship
For centuries, scholars have debated who truly wrote the earliest texts of the Hebrew Bible. Now, artificial intelligence is entering the fray — and it has answers.
A cross-disciplinary team of mathematicians, biblical scholars, physicists, and computer scientists has combined modern AI and linguistic analytics to identify distinct scribal traditions across the first nine books of the Hebrew Bible, also known as the Enneateuch. At the heart of this effort is Shira Faigenbaum-Golovin, assistant research professor of Mathematics at Duke University.
Her journey began in 2010, when she worked with archaeologist Israel Finkelstein to identify the authorship of ancient Hebrew inscriptions on pottery fragments. The success of that project laid the foundation for something much bigger: using mathematics and statistical modeling to unlock mysteries hidden in one of humanity’s most scrutinized texts.
The team analyzed language patterns in Deuteronomy, the historical books from Joshua to Kings, and the priestly writings in the Torah. Their custom-designed AI model compared usage of common words — including terms as simple as “no,” “king,” and “which” — across these texts. Despite how subtle these differences are, the model distinguished three clear authorial styles.
When the model was tested on chapters with already-agreed authorship, it confirmed traditional scholarly conclusions. But then it did more. It evaluated disputed passages, assigning them to one of the three scribal styles — and crucially, it explained how it reached its conclusions. The model could pinpoint which words or phrases drove its decisions, offering rare transparency in AI-driven analysis.
One compelling case involved the Ark Narrative in the Books of Samuel. Though these passages are often considered parts of a unified story, the AI showed they were likely written by different authors — a subtle but powerful insight that could reshape how scholars interpret these books.
Traditional machine learning methods faltered when analyzing ancient texts because of the limited data — some segments were only a few verses long. Instead, the team crafted a bespoke model that relies on direct comparisons of sentence structures and root-word frequencies (lemmas), sidestepping the need for massive training datasets.
“It’s a new paradigm for analyzing ancient texts,” said Finkelstein. The project’s implications stretch far beyond biblical studies. Faigenbaum-Golovin suggests the method could be used to authenticate historical documents or even investigate literary forgeries. “If you have letters supposedly from Abraham Lincoln, this can help verify if they’re genuine.”
The fusion of AI and humanities — often seen as opposites — is yielding unexpected insights. As Faigenbaum-Golovin put it, “It’s a surprising symbiosis. I’m lucky to work with people who use innovative research to push boundaries.”
With plans to apply the same techniques to texts like the Dead Sea Scrolls, this is just the beginning. The Bible, it turns out, still has many secrets to reveal — and now, it has an algorithm to help.
Source: Duke University
Enjoying our articles?
We don’t show ads — so you can focus entirely on the story, without pop-ups or distractions. We don’t do sponsored content either, because we want to stay objective and only write about what truly fascinates us. If you’d like to help us keep going — buy us a coffee. It’s a small gesture that means a lot. Click here – Thank You!
