Periodic Reporting for period 4 - HandsandBible (The Hands that Wrote the Bible: Digital Palaeography and Scribal Culture of the Dead Sea Scrolls)
Période du rapport: 2020-05-01 au 2022-02-28
The Dead Sea Scrolls are ancient Jewish manuscripts from the third century BCE to the second century CE. They come from eleven caves near the site of Qumran on the north-western shore of the Dead Sea and from other Judaean Desert sites such as Masada. The scrolls contain the oldest manuscripts of the Hebrew Bible (Old Testament) and many hitherto unknown ancient Jewish texts.
The discovery of the Dead Sea Scrolls in the 1940s–1960s has fundamentally transformed our knowledge of Jewish and Christian origins. The manuscripts provide a unique vantage point for studying the dynamic and creative engagement with authoritative scriptures that were to become the Bible. Before the discovery of the scrolls, our oldest, complete manuscript in Hebrew of the Old Testament was from around the year 1000. The scrolls allow us to jump back 2000 years and more in time and observe what people wrote, copied, collected, read, and studied. The caves from the Judaean Desert with the scrolls are a time machine as it were.
The Dead Sea Scrolls offer evidence for a scribal culture “in action”. Palaeography can provide access to this scribal culture, showing the human hand behind what came to be regarded as holy texts. The exciting aspect of this project is that it will, through the innovative and unconventional digital palaeographic analysis that we use, bring these scribal identities back to life as it were.
The discovery of the Dead Sea Scrolls in the 1940s–1960s has fundamentally transformed our knowledge of Jewish and Christian origins. The manuscripts provide a unique vantage point for studying the dynamic and creative engagement with authoritative scriptures that were to become the Bible. Before the discovery of the scrolls, our oldest, complete manuscript in Hebrew of the Old Testament was from around the year 1000. The scrolls allow us to jump back 2000 years and more in time and observe what people wrote, copied, collected, read, and studied. The caves from the Judaean Desert with the scrolls are a time machine as it were.
The Dead Sea Scrolls offer evidence for a scribal culture “in action”. Palaeography can provide access to this scribal culture, showing the human hand behind what came to be regarded as holy texts. The exciting aspect of this project is that it will, through the innovative and unconventional digital palaeographic analysis that we use, bring these scribal identities back to life as it were.
The ERC project The Hands that Wrote the Bible was set up to tackle two fundamental problems in the palaeography of the Dead Sea Scrolls, in the belief that solving these problems would have important implications for understanding the Dead Sea Scrolls as a collection, and also to better understand text production and collection practices in ancient Judaea. Those two problems were: the lack of a method to identify the anonymous scribes and to date manuscripts based on their writing style, as none carries an internal date.
The project used state-of-the art image processing and pattern recognition tools developed by Lambert Schomaker’s research group at Artificial Intelligence in Groningen in order to analyse high-resolution multispectral images of the Dead Sea Scrolls made by the Israel Antiquities Authority and also on digitized images of the scrolls by Brill Publishers. Maruf Dhali, PhD candidate in the project, developed these tools further to apply them to the Dead Sea Scrolls, and he also created a robust binarization tool that is necessary to successfully perform writer identification and date prediction analyses, BiNet (https://arxiv.org/pdf/1911.07930v1).
Combining the humanities and the sciences, we have succeeded in establishing an empirical, quantitative basis for the palaeography of the Dead Sea Scrolls, which can also be applied to other historical corpora. Our digital palaeography allows us to better determine the identity of scribes or the difference between them, without bias. Additionally, our digital palaeography allows us to date manuscripts, without it being a self-fulling prophecy or cherry picking.
The international online conference in April 2021 was important for knowledge and technology transfer and to engage and test our novel tools and results for the community of both traditional palaeographers and AI experts. See the introduction and YouTube for all presentations: https://www.youtube.com/watch?v=u97DA-s5Vz0&t=29s https://www.youtube.com/watch?v=Nqc77F-xGN4&t=3s.
Dr Drew Longacre, postdoc in the project, has made significant contributions to the stylistic classification of Hebrew scripts and the interpretation of the diverse Psalms manuscripts from the Judean Desert. See, in addition to his publications, already his presentation at the April 2021 conference: https://www.youtube.com/watch?v=uFaWsB7aiag&t=9s.
Gemma Hayes, PhD candidate in the project, focused on writer identification of a Qumran scribe using AI and digital tools, with discussions on how digital tools support the work of traditional palaeographers. Through the lens of the identified scribe and the manuscripts he copied, her forthcoming dissertation explores the handwriting practices (palaeography), spelling practices (orthography and morphology), codicological features and literary content of the Qumran scrolls. From the collated data, Hayes offers assessments of the dominant models in the field used to categorise and classify texts. See already her presentation at the 2021 conference: https://www.youtube.com/watch?v=zSxlKNi2lyY&t=1s.
The combination of AI, statistics, and post-hoc visual inspection makes for a robust, new method for writer identification, and is an advancement for the field. See our article in PLoS ONE, April 2021: doi.org/10.1371/journal.pone.0249769. This study was widely covered in the media see, e.g. BBC, Euronews, Haaretz, ABC, El País, New Scientist. Our method changes how we do palaeography by benefitting from the combination of different disciplines, combining quantitative and qualitative approaches. Also, it represents a major step forward in interdisciplinary method and communication between researchers in these fields, having already an impact on digital palaeography beyond Dead Sea Scrolls studies. Furthermore, our approach makes it possible to unlock the microlevel of individual scribes.
To tackle the problem of a lack of absolute dates for the Qumran scrolls across the time period, we performed radiocarbon testing on 30 samples, kindly provided to us by the Israel Antiquities Authority. Hans van der Plicht (University of Groningen) has conducted the C14 analyses for the project, and Kaare Rasmussen (University of Southern Denmark) and Perla Colombini and her team (University of Pisa) have performed cleaning procedures and chemical analyses to assure the cleaning treatment. The C14 data was used to train the algorithm for date predictions. The integration of these two disciplines, AI and radiocarbon dating, is an innovation of the project. For the work in progress on our date prediction model using AI and C14 datings, see the 2021 conference, YouTube, e.g.: https://www.youtube.com/watch?v=P2QKWUZXfa8 https://www.youtube.com/watch?v=JZeyWBbWawM&t=22s https://www.youtube.com/watch?v=Z0osmcXwggY&t=12s.
The project used state-of-the art image processing and pattern recognition tools developed by Lambert Schomaker’s research group at Artificial Intelligence in Groningen in order to analyse high-resolution multispectral images of the Dead Sea Scrolls made by the Israel Antiquities Authority and also on digitized images of the scrolls by Brill Publishers. Maruf Dhali, PhD candidate in the project, developed these tools further to apply them to the Dead Sea Scrolls, and he also created a robust binarization tool that is necessary to successfully perform writer identification and date prediction analyses, BiNet (https://arxiv.org/pdf/1911.07930v1).
Combining the humanities and the sciences, we have succeeded in establishing an empirical, quantitative basis for the palaeography of the Dead Sea Scrolls, which can also be applied to other historical corpora. Our digital palaeography allows us to better determine the identity of scribes or the difference between them, without bias. Additionally, our digital palaeography allows us to date manuscripts, without it being a self-fulling prophecy or cherry picking.
The international online conference in April 2021 was important for knowledge and technology transfer and to engage and test our novel tools and results for the community of both traditional palaeographers and AI experts. See the introduction and YouTube for all presentations: https://www.youtube.com/watch?v=u97DA-s5Vz0&t=29s https://www.youtube.com/watch?v=Nqc77F-xGN4&t=3s.
Dr Drew Longacre, postdoc in the project, has made significant contributions to the stylistic classification of Hebrew scripts and the interpretation of the diverse Psalms manuscripts from the Judean Desert. See, in addition to his publications, already his presentation at the April 2021 conference: https://www.youtube.com/watch?v=uFaWsB7aiag&t=9s.
Gemma Hayes, PhD candidate in the project, focused on writer identification of a Qumran scribe using AI and digital tools, with discussions on how digital tools support the work of traditional palaeographers. Through the lens of the identified scribe and the manuscripts he copied, her forthcoming dissertation explores the handwriting practices (palaeography), spelling practices (orthography and morphology), codicological features and literary content of the Qumran scrolls. From the collated data, Hayes offers assessments of the dominant models in the field used to categorise and classify texts. See already her presentation at the 2021 conference: https://www.youtube.com/watch?v=zSxlKNi2lyY&t=1s.
The combination of AI, statistics, and post-hoc visual inspection makes for a robust, new method for writer identification, and is an advancement for the field. See our article in PLoS ONE, April 2021: doi.org/10.1371/journal.pone.0249769. This study was widely covered in the media see, e.g. BBC, Euronews, Haaretz, ABC, El País, New Scientist. Our method changes how we do palaeography by benefitting from the combination of different disciplines, combining quantitative and qualitative approaches. Also, it represents a major step forward in interdisciplinary method and communication between researchers in these fields, having already an impact on digital palaeography beyond Dead Sea Scrolls studies. Furthermore, our approach makes it possible to unlock the microlevel of individual scribes.
To tackle the problem of a lack of absolute dates for the Qumran scrolls across the time period, we performed radiocarbon testing on 30 samples, kindly provided to us by the Israel Antiquities Authority. Hans van der Plicht (University of Groningen) has conducted the C14 analyses for the project, and Kaare Rasmussen (University of Southern Denmark) and Perla Colombini and her team (University of Pisa) have performed cleaning procedures and chemical analyses to assure the cleaning treatment. The C14 data was used to train the algorithm for date predictions. The integration of these two disciplines, AI and radiocarbon dating, is an innovation of the project. For the work in progress on our date prediction model using AI and C14 datings, see the 2021 conference, YouTube, e.g.: https://www.youtube.com/watch?v=P2QKWUZXfa8 https://www.youtube.com/watch?v=JZeyWBbWawM&t=22s https://www.youtube.com/watch?v=Z0osmcXwggY&t=12s.
Palaeographers who say they can date Dead Sea Scrolls with a precision of, say, a 25-50 years date range have yet to substantiate their model. Pending new data in the future, we dare say we have a model that works consistently and is able to date manuscripts with an empirically based precision that was not possible before. This is a huge advance for the field. Furthermore, there are also cases where scholars disagree on the date of a manuscript, ranging a century or more apart sometimes. Our model may help to decide on such disputes about very divergent date suggestions by presenting scholars with empirically based probability options.
Regarding writer identification, our breakthrough in scroll paleography is that we can now work confidently on the microlevel of individual scribes and carefully observe how they worked on their manuscripts. Applying our digital method to other groups of scrolls may reveal relations between manuscripts and texts that have previously gone unnoticed. The ability to identify scribes can contribute to a finer understanding of the different groupings and collections of manuscripts within the Dead Sea Scrolls, contributing to a better understanding of the dynamics of manuscript production and collection.
Regarding writer identification, our breakthrough in scroll paleography is that we can now work confidently on the microlevel of individual scribes and carefully observe how they worked on their manuscripts. Applying our digital method to other groups of scrolls may reveal relations between manuscripts and texts that have previously gone unnoticed. The ability to identify scribes can contribute to a finer understanding of the different groupings and collections of manuscripts within the Dead Sea Scrolls, contributing to a better understanding of the dynamics of manuscript production and collection.