AI recreates ancient Babylonian hymn from 3000-year-old cuneiform script
In a groundbreaking development, artificial intelligence (AI) has played a pivotal role in the reconstruction of a 3,000-year-old Babylonian hymn, once fragmented into hundreds of tiny clay pieces. The project, a collaboration between Ludwig-Maximilians-Universität Munich and the University of Baghdad, has successfully restored approximately two-thirds of the hymn, shedding light on a classical text that was previously obscured from scholars despite its popularity in antiquity.
The researchers used the Electronic Babylonian Library, a vast repository containing 1,402 manuscripts, as their primary source. The AI software, trained on thousands of photographed fragments, identified overlaps that had been missed by human experts for over a century, effectively stitching the fragments together virtually.
The AI employed in Babylonian text reconstruction relies heavily on pattern recognition and image analysis of cuneiform tablets. The system can handle large datasets of fragment images to detect overlaps and textual continuities that are extremely difficult for humans due to fragmentary preservation. This technique leverages machine learning algorithms trained on massive libraries of known cuneiform signs and variants to predict the placement and content of missing sections, enabling scholars to virtually reunite broken manuscripts.
Similar AI-driven methods have transformed the study of the Dead Sea Scrolls, ancient Hebrew Bible manuscripts dating from the third century BCE to the first century CE. AI programs like Enoch analyse handwriting styles using high-resolution images, producing new radiocarbon dating estimates and pushing back the scrolls' ages by up to 50-150 years in some cases, providing a sharper chronology and closer connection to original authors.
These AI methods combine handwriting analysis with improved physical dating techniques, such as radiocarbon dating corrected for earlier contamination, to offer a holistic reconstruction of both text and historical context.
AI-assisted reconstruction is opening new windows into ancient civilizations, allowing researchers to recover long-lost texts, understand linguistic evolution, and gain cultural insights that were formerly inaccessible due to manuscript fragmentation and decay. The integration of AI with traditional philology and archaeology suggests a future where virtual libraries and digital reconstructions will make ancient writings widely accessible and more comprehensible, enabling continuous refinement as more fragments and data become available.
Professor Enrique Jiménez, Professor of Ancient Oriental Languages at LMU, led the team behind the rediscovery of the Babylonian hymn. He emphasised that while languages like Akkadian and Sumerian are underrepresented in large language models, efforts are being made to improve computational tools for ancient Near Eastern studies.
The passage from the Hymn of Babylon reads, "The foreigners among them they do not humiliate. The humble they protect, the weak they support. Under their care, the poor and destitute can thrive. To the orphan they offer succour and favour." The hymn praises Babylon and the god Marduk, and includes passages praising Babylon's acceptance of foreigners and support for the poor, suggesting the author was likely a member of Babylon's priestly class.
AI is becoming increasingly common in the reconstruction of ancient texts among scholars. As more manuscripts are digitised and AI tools continue to evolve, we can expect to see many more ancient texts being brought to light, offering invaluable insights into the history and culture of civilisations long past.
The AI software, harnessing pattern recognition and image analysis, was trained to identify overlaps in the Electronic Babylonian Library, aiding in the reconstruction of the 3,000-year-old Babylonian hymn. In the realm of science, technology, and artificial intelligence, this technique plays a significant role in the Eth (initial coin offering) of knowledge archaeology, transforming the study of ancient texts like the Babylonian hymn and the Dead Sea Scrolls.
