A team of researchers has used artificial intelligence to restore a 3,000-year-old Babylonian hymn inscribed on cuneiform tablets, according to newly published research.
Using machine learning, the team behind he achievement identified 30 manuscripts with distinct similarities, then reconstructed and interpreted the full text, providing new insights into daily life and society in ancient Babylon.
Revealing new information about ancient Mesopotamian culture, the rediscovered text now offers a rare, detailed view of daily life and shows how scribes in the early first millennium viewed the natural world. The text also highlights social themes such as women’s religious roles and the value placed on hospitality toward foreigners.
The findings were published in Iraq, the journal of the British Institute for the Study of Iraq.
A Challenge Beyond Manual Reconstruction
The reconstruction was led by Professor Enrique Jiménez of Ludwig-Maximilians-Universität München, a specialist in Mesopotamian literature. Cuneiform tablets often survive in fragments, with many pieces ending up broken, eroded, and dispersed through excavations, sales, and museum transfers over centuries. On their own, these fragments are nearly impossible to piece together.
While cataloging tablets with colleagues from the University of Baghdad, Jiménez found a text with unusual phrasing. It appeared to be a hymn, but did not match any previously seen in Babylonian records. In the search for similar tablets, the team used the Electronic Babylonian Library Platform, a digital tool that collects and compares cuneiform fragments from global collections.
How the AI Found the Missing Lines
Cuneiform, the writing system of ancient Mesopotamia, was made by pressing wedge-shaped marks into clay tablets. The Electronic Babylonian Library Platform uses machine learning models trained on thousands of cuneiform texts. The models analyze patterns and writing styles to identify fragments that probably came from the same source.
Using this method, the system found 30 manuscripts with similar lines, words, or structure. By linking these fragments, the researchers filled in missing parts and restored the poem to its original form. The finished hymn is about 250 lines long, making it one of the largest literary works recovered from ancient Babylon.
Jiménez noted that artificial intelligence was essential to the discovery, revealing connections that experts had previously overlooked.
A Lost Hymn That Once Echoed Throughout Babylon
The poem provides a detailed view of Babylon during its cultural peak. It describes monumental buildings, busy temple districts, and the agricultural abundance supported by the seasonal floods of the Euphrates River. The focus on environmental imagery is unusual, as most Mesopotamian literature centers on gods, rulers, and rituals rather than the natural world.
The hymn also describes a society that valued hospitality toward outsiders, a theme rarely found in other existing texts. It also includes one of the few surviving references to women’s priestly roles in Babylon.
In ancient times, the text was likely well-known. The number of copies identified suggests that the hymn was a standard exercise for students learning to scribe cuneiform. The fragmentation and global dispersal of the surviving tablets led to the hymn’s disappearance for thousands of years.
Rebuilding a Civilization’s Written Memory
This discovery is part of a project to digitize and preserve the tablets from the Sippar Library, an archive of the ancient city of Sippar. While legends suggest that Noah also hid tablets at this location before the flood, historians generally view this as symbolic and primarily regard the site as an important center of Mesopotamian scholarship.
By digitizing these cuneiform tablets, researchers aim to reconstruct additional texts to better understand the culture and intellectual environment of Babylon and nearby ancient cities.
a Forgotten Voice Revived
Today, the ruins of Babylon are located about 50 miles south of Baghdad. Recognized as a UNESCO World Heritage Site, the remains of its walls, palaces, and temples offer a window into Iraq’s former landscape.
Artificial intelligence now allows researchers to read the voice of a Babylonian scribe lost for three millennia. The reconstructed hymn enriches the literary record of ancient Mesopotamia and illustrates how digital tools can restore stories nearly erased by time.
Austin Burgess is a writer and researcher with a background in sales, marketing, and data analytics. He holds a Master of Business Administration and a Bachelor of Science in Business Administration, as well as a certification in Data Analytics. His work combines analytical training with a focus on emerging science, aerospace, and astronomical research.