Long before cuneiform, people placed memory outside any single body: in images, notches, tokens, monuments, paths, objects, and altered landscapes. Those systems were not lesser because they required trained participants. Cuneiform also required scribes, conventions, institutions, and readers.

What changed at Uruk was not a clean escape from embodiment. Clay tablets could hold many counted relationships in a durable, rearrangeable form: quantities beside commodities, offices beside obligations, marks that could travel through an administration and survive its maker. Memory gained a new surface, then built a profession around learning how to read it.

The earliest surviving Uruk tablets are overwhelmingly administrative: grain, animals, labour, offices, rations. Their signs are incised and impressed rather than the later script's regular wedges; some resemble objects, while many are abstract.

Accounting is therefore central to the evidence, but 'writing was invented for debt' is too neat. Tokens, seals, numerical tablets, labels, and lexical lists formed an evolving administrative ecology. Literature appears later in the surviving record. The wonder is not that poetry lost a race to paperwork, but that a counted mark could gradually learn to carry a name, a sound, a law, a lament.

An Uruk III tablet from about 3200–3000 BCE records roughly 135,000 litres of barley over 37 months and ends with signs read as Kushim or Kushin. CDLI describes the term as a government official responsible for a brewery account. Scholars have also asked whether it names a person, a title, or an institution.

That uncertainty changes the encounter. We may be near one of the earliest written individuals, or looking at an office made durable enough to outlast everyone who held it. Clay preserves the signs with startling clarity; it does not preserve the speaker who could settle their grammar.

Around 1750 BCE, Nanni sent Ea-Nasir a letter complaining that the wrong grade of copper had been delivered after a Gulf voyage, and objecting to delays and the treatment of his messengers. The British Museum catalogues the tablet as a business letter; its fame as the 'oldest customer complaint' is a modern superlative, safer phrased as one of the earliest surviving complaints of its kind.

Its anger crosses the translation with unusual force, but the resemblance to an email is ours, not Nanni's. What the clay genuinely preserves is more interesting than sameness: a Bronze Age commercial relationship in which quality, obligation, distance, and dignity could all be argued in writing.

Cuneiform is a family of scripts, not a language. Across more than three millennia, related wedge-shaped sign systems were adapted to write Sumerian, Akkadian, Hittite, Elamite, Hurrian, Urartian, Old Persian, and other languages. The Latin-alphabet comparison helps only at that level: each adaptation changed the work signs had to do.

Nor did pictures become sounds in one sudden leap around 3200 BCE. Early proto-cuneiform combined numerical and word signs, many of them abstract. A few rebus-like phonetic uses appear between about 3200 and 3000 BCE; the Met notes that consistent phonetic writing is clear only after about 2600 BCE.

That gradual shift matters. Sound did not make clay infinitely expressive overnight. Generations of scribes altered signs, trained readers, and fitted the system to new languages. Accounting and literature are not opposite shores of one miracle; they are different achievements built through a long social technology.

By the nineteenth century, cuneiform no longer had a living reading community. At Behistun, about 60 metres above the plain, Darius I had ordered versions of one royal account in Old Persian, Elamite, and Babylonian. Henry Rawlinson's hazardous copying and work on Old Persian were crucial.

But no lone man unlocked 'cuneiform.' Grotefend had made earlier progress on Old Persian; Edward Hincks, Rawlinson, Jules Oppert, William Henry Fox Talbot, and others advanced different languages and sign values. In 1857, four independent translations of an Assyrian inscription were compared as a public test. Decipherment was a relay: cliff, copy, rivalry, collaboration, and correction.

Hundreds of thousands of cuneiform tablets and fragments are dispersed across excavated sites, storerooms, museums, and digital catalogues. A count changes with joins, new records, and the boundary between a tablet and a fragment; 'unread' can mean uncatalogued, unpublished, untranslated, or simply not yet reinterpreted.

Clay's survival is uneven. Some tablets were deliberately fired, many dried in the sun, and some were hardened when buildings burned. At Nineveh, destruction buried and in places baked parts of Ashurbanipal's library. Fire preserved by damaging: an ending became one condition of the archive's survival.

The romance of unread shelves is real, but so is the labour around them — conservation, provenance research, digitisation, sign identification, joining fragments, and learning languages whose meanings continue to be revised.