This article records tradition as it has been passed down and reported. Its sources are not yet part of the atlas's verified catalogue.
Sumerian was not the first ancient script scholars learned to read. It was, in a real sense, a discovery made inside another one. Through the first half of the nineteenth century a handful of European scholars worked to crack cuneiform using the Behistun inscription, a cliff face monument in western Iran carved for the Persian king Darius I in three languages side by side. Henry Rawlinson, a British officer, spent years copying the trilingual text from a narrow ledge hundreds of feet above the ground, and by the mid 1840s he and others had worked out the Old Persian portion. The harder problem was the inscription's third language, Babylonian, written in the same wedge shaped cuneiform signs but representing an entirely different, Semitic tongue: Akkadian. In 1857, with the decipherment still disputed, the Royal Asiatic Society organized a test. Four scholars, Rawlinson, the Irish clergyman Edward Hincks, the French Assyriologist Julius Oppert and the amateur Henry Fox Talbot, were each given the same newly found cuneiform inscription, one commemorating the Assyrian king Tiglath Pileser I, and asked to translate it independently and in sealed submissions. When the results were compared, Rawlinson, Hincks and Oppert agreed almost exactly on the substance of the text, while Talbot's attempt fell short. Cuneiform, and Akkadian with it, could genuinely be read. That breakthrough is what made Sumerian legible at all. Later Assyriologists working through Akkadian scribal libraries found bilingual dictionaries and word lists in which Akkadian speaking scribes had glossed an older, unrelated language, the tongue of the very first cuneiform tablets. Sumerian turned out to survive only inside the record keeping of the civilization that had inherited its script, a language recovered because its own students, three thousand years earlier, had refused to let it go untranslated.