Home https://server7.kproxy.com/servlet/redirect.srv/sruj/smyrwpoii/p2/ Science https://server7.kproxy.com/servlet/redirect.srv/sruj/smyrwpoii/p2/ Artificial intelligence helps break the code of the Dead Sea Scrolls

Artificial intelligence helps break the code of the Dead Sea Scrolls

Handwriting analysis of the Dead Sea Scrolls

Dead Sea scroll text analysis collage. Credit: Mladen Popovic

The Dead Sea Scrolls, discovered seventy years ago, are known to contain the oldest manuscripts of the Hebrew Bible (Old Testament) and many hitherto unknown ancient Hebrew texts. But the individuals behind the scrolls have escaped the scientists because the scribes are anonymous. Now, combining science and the humanities, researchers at the University of Groningen have broken the code that allows them to find the scribes behind the scrolls. They presented their results in the magazine PLOS ONE on April 21

, 2021.

The scribes who created the scrolls do not sign their work. Scholars suggest that some manuscripts be attributed to a scribe based on handwriting. “They would try to find a ‘smoking gun’ in handwriting, for example, a very specific feature in a letter that would identify the scribe,” explains Mladen Popovic, a professor of Hebrew Bible and Ancient Judaism at the Faculty of Theology and Religion at the University of Groningen. . He is also the director of the Qumran University Institute, dedicated to the study of the Dead Sea Scrolls. However, these identifications are somewhat subjective and are often hotly debated.


That’s why Popovich teamed up with his colleague Lambert Schomaker, a professor of computer science and artificial intelligence at the Faculty of Science and Engineering, in his Handwriting Bible project, funded by the European Research Council. Schomaker has long worked on techniques that allow computers to read handwriting, often from historical materials. He is also conducting research to investigate how biomechanical characteristics, such as the way someone holds a pen or stylus, will affect handwriting.

Analyzing the letters scrolls from the Dead Sea

Two 12 × 12 Kohonen cards (blue colored cards) with a full character character and a bet from the Dead Sea Scrolls collection. Each of the characters in Kohonen’s maps is formed by multiple instances of similar characters (shown with an enlarged field with red lines). These maps are useful for analyzing chronological style. In the present study identifying writers, Fraglets (fragmented forms of symbols) were used instead of complete forms of symbols to achieve more accurate (stable) results. Credit: Maruf A. Dhali, University of Groningen

In this study, together with PhD candidate Maruf Dali, they focused in particular on one scroll: the famous Great Scroll of Isaiah (1QIsaa) from Qumran Cave 1. The handwriting in this scroll looks almost identical, but is thought to have been made by two scribes sharing a similar style of writing. So how can this be solved? Schomaker: “This scroll contains the letter aleph, or ‘a,’ at least five thousand times. It is impossible to compare them only by eye. “Computers are very suitable for analyzing large arrays of data, such as 5,000 handwritten a. The digital image makes possible all kinds of computer calculations at the micro level of symbols, such as measuring curvature (called textural), as well as whole characters (called allographic).

Neural network

“The human eye is amazing and probably takes these levels into account. “This allows experts to ‘see’ the hands of different authors, but this solution is often not reached through a transparent process,” Popovic said. “Furthermore, it is virtually impossible for these experts to process the large amount of data that the scrolls provide.” Therefore, their results are often not convincing.

Binarization of the Dead Sea Scrolls

(Left to right) Gray image of column 15 of the Great Scroll of Isaiah, the corresponding binary image using BiNet, and the cleared corrected image. From the red fields of the last two images you can see how the rotation and geometric transformation are corrected to get a better image for further processing. Credit: Reprinted by Lim TH, Alexander PS. Volume 1. In: Electronic Library of the Dead Sea Scrolls. Brill; 1995 under license CC BY, licensed by Brill, original copyright 1995

The first obstacle was to train an algorithm to separate the text (ink) from its background (skin or papyrus). For this separation, or “binarization,” Dhali developed a modern artificial neural network that can be trained through deep learning. This neural network keeps the original traces of ink made by the clerk more than 2,000 years ago intact as they appear in digital images. “This is important because the ancient traces of ink are directly related to the movement of human muscles and are specific to humans,” explains Schomaker.


Whether he 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 are divided into two different groups, which are not randomly distributed throughout the scroll, but are grouped, with a transition of about half a mark.

Dead Sea Scrolls Aleph

Illustration of how to generate heat maps of normalized mean symbol shapes for individual letters (in this example: aleph). Credit: Maruf A. Dhali, University of Groningen

Noting that there may be more than one writer, Dalí then passed the data on to Schomaker, who then recalculated the similarities between the columns, now using the templates of letter fragments. This second analytical step confirmed the existence of two different ones. Several additional inspections and controls were carried out. Schomaker: “When we added extra noise to the data, the result didn’t change. We have also been able to show that the second scribe shows more variation in his writing than the first, although their writing is very similar. “


In the third step, Popovic, Dali and Schomaker made a visual analysis. They created “heat cards” that included all variants of a symbol in the scroll. They then created an averaged version of this sign for the first 27 columns and the last 27 columns. A comparison of these two middle letters by eye shows that they are different. This links computer and statistical analysis to human interpretation of data by convergence, as heat maps are neither dependent nor produced by primary and secondary analysis.

Some aspects of the scroll and the positioning of the text have led some scholars to suggest that a new scribe began after column 27, but this is not generally accepted. Popovic: “Now we can confirm this with a quantitative analysis of handwriting, as well as with stable statistical analyzes. Instead of basing our judgment on more or less impressionistic evidence, with the intelligent help of the computer, we can prove that the division is statistically significant. “

New window

In addition to transforming the palaeography of the scrolls – and potentially other ancient manuscripts – this study of the Great Scroll of Isaiah opens up a whole new way of analyzing Qumran texts based on physical characteristics. Researchers can now access the micro level of individual scribes and observe closely how they worked on these manuscripts.

Popovic: “This is very exciting because it opens a new window to the ancient world, which can reveal much more complex connections between the scribes who created the scrolls. In this study, we found evidence of a very similar writing style shared by the two great scribes of Isaiah, which suggests a common teaching or origin. Our next step is to explore other scrolls where we can find different origins or training for scribes. “

In this way, it will be possible to learn more about the communities that created the Dead Sea Scrolls. “We can now identify different clerks,” Popovic concluded. “We will never know their names. But after seventy years of training, it’s a feeling as if we can finally shake hands with them through their handwriting. “

Reference: “The identification of a writer based on artificial intelligence generates new evidence for the unknown scribes of the Dead Sea Scrolls, illustrated by the Great Scroll of Isaiah (1QIsaa) ”By Mladen Popovic, Maruf A. Dali and Lambert Schomaker, April 21, 2021, PLOS ONE.
DOI: 10.1371 / journal.pone.0249769

Digital images of the Dead Sea Scrolls and the Great Scroll of Isaiah were kindly provided by publishers Brill and the Israel Antiquities Authority (Leon Levy’s Digital Library of the Dead Sea Scrolls).

Source link