Artificial intelligence tools allow researchers to ‘shake hands’ with former scribes

The Dead Sea Scrolls, discovered about seventy years ago, are reputed to contain the oldest manuscripts of the Hebrew Bible (Old Testament) and many ancient Jewish texts unknown until now. But the people behind the scrolls have escaped scientists because the scribes are anonymous. Now, by combining the sciences and the humanities, researchers at the University of Groningen have cracked the code, which allows them to uncover the scribes behind the scrolls. They presented their results in the newspaper PLOS ONE April 21.

The scribes who created the scrolls did not sign their work. Researchers have suggested that some manuscripts should be attributed to a single scribe based on the handwriting. “They would try to find a ‘smoking gun’ in the handwriting, for example, a very specific trait in a letter that would identify a scribe,” says Mladen Popovic, professor of Hebrew Bible and ancient Judaism at the Faculty of Theology and religion. Studies at the University of Groningen. He is also director of the university’s Qumran Institute, dedicated to the study of the Dead Sea Scrolls. However, these identifications are somewhat subjective and often hotly debated.


This is why Popovic, in his project The Hands that Wrote the Bible funded by the European Research Council, joined forces with his colleague Lambert Schomaker, professor of computer science and artificial intelligence at the Faculty of Science and engineering. Schomaker has long worked on techniques that allow computers to read handwriting, often from historical materials. He also conducted studies to investigate how biomechanical traits, such as the way a person holds a pen or stylus, would affect writing.

In this study, with doctoral student Maruf Dhali, they focused on one parchment in particular: the famous large scroll of Isaiah (1QIsaa) from Qumran Cave 1. The writing on this scroll appears almost uniform, but it has been suggested that it was done by two scribes sharing a similar writing style. So how could this be decided? Schomaker: “This parchment contains the letter aleph, or ‘a’, at least five thousand times. It is impossible to compare them all with the naked eye. Computers are well suited for analyzing large sets of data, such as 5,000,000 manuscripts. Digital imaging enables all kinds of computer calculations, at the micro-level of characters, such as measuring curvature (called textural), as well as whole characters (called allographic).

Neural network

“The human eye is amazing and presumably takes these levels into account as well. This allows experts to “see” the hands of different perpetrators, but this decision is often not made by a transparent process, ”Popovic says. “In addition, it is virtually impossible for these experts to process the large amounts of data provided by the rolls.” This is why their results are often inconclusive.

The first obstacle was to train an algorithm to separate the text (ink) from its background (leather or papyrus). For this separation, or “binarization,” Dhali has developed an advanced artificial neural network that can be trained using deep learning. This neural network retains intact the original ink traces made by the scribe over 2000 years ago as they appear in digital images. “This is important because the old ink traces are directly related to a person’s muscle movements and are specific to the person,” says Schomaker.


Dhali performed the first analytical test of this study. His analysis of textural and allographic features showed that the 54 columns of text in the Great Scroll of Isaiah fell into two different groups that were not randomly distributed in the scroll, but were clustered together, with a transition around the mid -path.

With the remark that there might be more than one writer, Dhali then handed the data to Schomaker who then recalculated the similarities between the columns, now using the letter fragment patterns. This second analytical step confirmed the presence of two different ones. Several other checks and controls were carried out. Schomaker: “When we added extra noise to the data, the result didn’t change. We have also succeeded in demonstrating that the second scribe shows more variation in his handwriting than the first, although their handwriting is very similar.


In the third step, Popovic, Dhali and Schomaker produced a visual analysis. They created “heat maps” that incorporate all of the variations of a character across the scroll. Then they produced an average version of that character for the first 27 columns and the last 27 columns. Comparing these two medium letters with the naked eye shows that they are different. This links computerized and statistical analysis with human interpretation of data by approximation, as heat maps are neither dependent nor produced from primary and secondary analyzes.

Certain aspects of the scroll and the positioning of the text had led some scholars to suggest that after Column 27 a new scribe had started, but this was not generally accepted. Popovic: “Now we can confirm this with quantitative analysis of handwriting as well as robust statistical analyzes. Instead of basing our judgment on more or less impressionistic evidence, with the intelligent help of the computer, we can show that the separation is statistically significant.

New window

In addition to transforming the paleography of the scrolls – and potentially other corpus of ancient manuscripts – this study of the great scroll of Isaiah opens up a whole new way of analyzing the texts of Qumran in terms of physical characteristics. Now researchers can go to the micro-level of individual scribes and closely observe how they worked on these manuscripts.

Popovic: “It’s very exciting, because it opens a new window into the ancient world that can reveal much more complex connections between the scribes who produced the scrolls. In this study, we found evidence of a very similar writing style shared by the two scribes of the Great Parchment of Isaiah, suggesting a common formation or origin. Our next step is to study other scrolls, where we can find different origins or formations for the scribes.

In this way, it will be possible to learn more about the communities that produced the Dead Sea Scrolls. “We are now able to identify different scribes,” Popovic concludes. “We will never know their names. But after seventy years of study, it’s like we can finally shake their hands with their writing.

Agriculture Lifestyle political