Invention of Cuneiform: How Sumer Created the First Writing

Cuneiform was the first full writing system in the world, created in southern Mesopotamia, in what is now Iraq, around 3400-3200 BC. It began as a way for officials of the city of Uruk to record grain, livestock and labour on tablets of clay, and over many centuries became a script able to write any kind of text, from laws to poetry. Used for about 3,000 years and for more than a dozen languages, cuneiform gave humanity its first written history.

Before writing: counting with tokens

People in the Near East kept records long before they wrote words. From about 8000 BC, villages across the region used small clay tokens in shapes such as cones, spheres, discs and cylinders. The archaeologist Denise Schmandt-Besserat argued from the 1970s onward that each shape stood for a fixed amount of a commodity, such as a measure of grain or one sheep.

By the middle of the fourth millennium BC, tokens were sometimes sealed inside hollow clay balls, known as bullae or envelopes, as a guarantee of a transaction. Some envelopes carry impressions of the tokens on their outside, so the contents could be read without breaking them. In time, Schmandt-Besserat suggested, officials realised that the impressions alone were enough and the tokens became unnecessary.

Her theory remains influential, but many specialists consider it too neat. The link between tokens and particular written signs is clear for some numerical signs and much weaker for most others. What is widely accepted is that writing came out of accounting: people needed lasting records of who owed or received what.

Numerical tablets and proto-cuneiform at Uruk

The next stage was the numerical tablet: small flat pieces of clay marked only with numbers and sometimes a cylinder seal impression. Such tablets are known from Uruk, from Susa in south-west Iran and from sites in Syria, and date to roughly 3500-3300 BC.

Then, in the Eanna temple district at Uruk, the earliest true texts appear. German excavators working at Uruk (modern Warka) from 1928 onward found around 5,000 archaic tablets and fragments, most of them thrown away as rubbish and reused as building fill. The oldest belong to the archaeological level called Uruk IV, usually dated to about 3350-3200 BC; a later group belongs to Uruk III, also called the Jemdet Nasr period, about 3200-3000 BC.

These tablets combine number signs with word signs drawn in the clay. Scholars call this script proto-cuneiform. It contained roughly 900 to 1,200 different signs, depending on how variants are counted. Many began as pictures: a head, a bowl, an ear of barley, a sheep pen. The famous sign for “ration” seems to combine a human head with the bevelled-rim bowl used to hand out food.

What the first texts recorded

About 85 per cent of the archaic Uruk tablets are administrative. They record harvests and grain rations, beer and bread, herds of sheep and goats, textiles, fields and work teams. Some tablets list labourers, both men and women, by category.

Scribes counted different things with different number systems. Researchers led by Peter Damerow and Robert Englund identified around a dozen systems, including a sexagesimal (base-60) system for counting discrete objects and separate systems for grain capacity, area and time. This variety later fed into the Mesopotamian base-60 mathematics that still survives in our hours and degrees.

The other 15 per cent or so are lexical lists: standardised lists of words for professions, officials, animals, vessels, place names and other categories. The best known, the list of professions called “Lu A” by modern scholars, was copied by scribes for many centuries afterwards. These lists show that writing was already being taught and that scribal training was organised.

The creators of cuneiform are anonymous. They were administrators attached to the great temple households of Uruk and other cities of Sumer, the land of the Sumerians in the far south of Mesopotamia. Most scholars think the language behind proto-cuneiform was Sumerian, a language unrelated to any known living language, though the earliest tablets record so little grammar that this cannot be proved.

Uruk was then the largest settlement in the world, and its complex economy helps explain why writing appeared there. A related script, proto-Elamite, appeared at Susa around 3100 BC, perhaps inspired by Mesopotamian practice; it remains largely undeciphered.

From pictures to wedges and sounds

The first signs were drawn or incised with a pointed tool. Over the third millennium BC scribes increasingly pressed the triangular end of a reed stylus into the clay, producing wedge-shaped strokes. The pictures became groups of wedges that no longer looked like the objects they stood for.

The decisive change was the use of signs for sounds. Through the rebus principle, a sign for a word could be borrowed to write another word, or a syllable, that sounded the same. The Sumerian word for arrow, ti, sounded like the word for life, so the arrow sign could write “life”. Scribes also added unspoken determinatives, signs that classify a word as a god, a place, a wooden object and so on.

This made it possible to write personal names, grammatical endings and eventually whole sentences. By about 2600 BC, tablets from Fara (ancient Shuruppak) and Abu Salabikh include literary works and proverbs. Writing had become a tool not only for accounts but for religion, royal display, law and literature.

Spread and long life

From Sumer, cuneiform spread with trade, diplomacy and conquest. It was adapted for Akkadian, a Semitic language, by the mid-third millennium BC and became the main script of the Akkadian, Babylonian and Assyrian states. It was also used for Eblaite in Syria, Elamite in Iran, Hittite and other languages in Anatolia, Hurrian and Urartian.

In the second millennium BC, Akkadian cuneiform was the language of international diplomacy, used even in letters between Egypt and the rulers of the Near East. At Ugarit on the Syrian coast, scribes created a separate alphabet of about 30 wedge-shaped signs, described in the article on Ugarit and its alphabet. The Achaemenid Persian kings also commissioned a simplified cuneiform for Old Persian.

Cuneiform carried the works now regarded as the first great literature, including the Epic of Gilgamesh, as well as law codes, letters, contracts, medical texts and astronomical observations. After Aramaic and Greek replaced it in daily life, a small circle of temple scholars kept using it. The latest securely dated tablet is an astronomical text from about AD 75.

Decipherment

Knowledge of how to read cuneiform was lost for nearly 2,000 years. European travellers began copying inscriptions at Persepolis in the 17th and 18th centuries; the Danish explorer Carsten Niebuhr published accurate copies in the 1770s and noticed that they contained three different scripts.

In 1802 the German schoolteacher Georg Friedrich Grotefend correctly identified the names of the Persian kings Darius and Xerxes in the simplest of the three scripts, Old Persian. The British officer Henry Rawlinson copied the great trilingual Behistun Inscription in Iran between 1835 and 1847 and worked out the Old Persian text, which offered a key to the Babylonian version beside it.

The Irish clergyman Edward Hincks made crucial discoveries about how Akkadian signs worked, including that one sign could have several readings. In 1857 the Royal Asiatic Society in London gave copies of a newly found inscription of the Assyrian king Tiglath-Pileser I to four scholars, Rawlinson, Hincks, Jules Oppert and William Henry Fox Talbot. Working separately, they produced translations that largely agreed, convincing the public that Akkadian could really be read. Sumerian was recognised as a separate language later in the 19th century.

By the numbers

Myths and facts

A common idea is that writing was invented to record myths or royal deeds. In fact the earliest texts are almost entirely bookkeeping; literature and royal inscriptions came centuries later.

Another myth is that cuneiform is a language. It is a script, like the Latin alphabet, and was used to write many unrelated languages. It is also often said to be purely pictographic, but from early on it mixed word signs, sound signs and classifying signs.

Timeline

• c. 8000 BC: Small clay counting tokens come into use across the Near East.
• c. 3500-3300 BC: Clay envelopes with token impressions and numerical tablets appear at Uruk, Susa and elsewhere.
• c. 3350-3200 BC: The earliest proto-cuneiform tablets are written in the Eanna district of Uruk.
• c. 3200-3000 BC: Uruk III (Jemdet Nasr) tablets show a growing, more standardised script.
• c. 2600 BC: Literary texts from Fara and Abu Salabikh show Sumerian written in full.
• c. 2350 BC: Cuneiform is used widely for Akkadian under the Akkadian kings.
• c. AD 75: The latest securely dated cuneiform tablet, an astronomical text, is written.
• 1802: Grotefend identifies Persian royal names in Old Persian cuneiform.
• 1857: Four scholars’ independent translations in London confirm the decipherment of Akkadian.

Why it matters

Cuneiform turned spoken language and economic memory into something that could be stored, checked and passed on. It allowed states to tax and administer, priests to preserve rituals, and poets and scholars to hand down their work. Because clay survives fire and burial, it has left us one of the richest written records of any ancient civilisation, beginning at Uruk more than 5,000 years ago.