What is cuneiform? The first writing, explained

Cuneiform is the writing system of ancient Mesopotamia, in what is now Iraq, made by pressing a cut reed into wet clay so that each mark is a small wedge; the name comes from the Latin cuneus, "wedge." It began in the city of Uruk around 3300 BC as a way of keeping accounts, and by scholarly consensus it is the oldest known writing. For about 3,000 years it was used to write a dozen or so languages, from Sumerian to Persian.

In short

• Cuneiform grew step by step out of bookkeeping: clay tokens for goods, then marks of tokens on clay, then drawn signs, then signs for sounds.
• Its great leap was the rebus: using a picture for its sound, so that names, grammar and eventually whole literatures could be written.
• It went unread for almost 1,800 years and was read again in the 19th century, with a famous test in 1857 to prove the reading was right.

How did writing begin in Mesopotamia?

Not with stories, but with receipts. The archaeologist Denise Schmandt-Besserat traced cuneiform back to small clay tokens of different shapes, cones, spheres, disks, that people in the Near East used from the eighth millennium BC to count goods: one token per jar of oil or measure of grain. Her account of the evolution of writing lays out the stages.

Approximate date	Stage	What changed
From about 8000 BC	Clay tokens	Each shape stands for a unit of a commodity, counted one for one
About 3500 BC	Sealed clay envelopes	Tokens for a debt are sealed in a hollow clay ball; their shapes are pressed on the outside so the contents can be seen without breaking it
About 3200 BC	Impressed tablets	The tokens are dropped; the impressions alone, on a flat tablet, carry the record
About 3100 BC	Pictographs	Signs are drawn with a stylus, and signs for numbers separate from signs for things: "5" and "sheep" instead of five sheep-marks
About 3000 BC	Sound signs	Signs begin to stand for syllables, first to write people's names

The first clear documents come from the Uruk IV period, from about 3300 BC, according to the overview of cuneiform, and most of them are accounts. Dates for each stage are approximate, and scholars still debate how direct the line from tokens to writing was, but the overall direction, from counting to recording speech, is well supported.

How did a picture become a sound?

Follow one sign. Early scribes drew an arrow for the Sumerian word for arrow, ti. The word for "life" in Sumerian sounded almost the same, til, but life is hard to draw. So scribes used the arrow sign for "life" too, because of its sound. That move, using a sign for what it sounds like rather than what it shows, is the rebus principle, and it turned a bookkeeping tool into writing. English does the same in a puzzle that draws an eye, a heart and a ewe for "I love you."

Once signs could stand for sounds, they could be strung together to spell any word, including names, foreign words and the small grammatical endings that pictures can't show. Cuneiform ended up with three kinds of sign:

• Word signs, standing for a whole word, like the star sign for "sky" or "god."
• Sound signs, standing for a syllable such as ti, ba or an.
• Determinatives, unspoken markers that tell the reader what kind of word follows: a god, a city, a wooden object.

Meanwhile the look of the signs changed. Scribes pressed a reed with a wedge-shaped tip into the clay instead of drawing curves, which was faster, and the pictures became patterns of wedges that no longer looked like anything. The whole script was also turned 90 degrees. The number of signs fell from about 1,500 in the earliest phase to about 600 in regular use.

Which languages were written in cuneiform?

Cuneiform was a script, not a language, much as the Latin alphabet writes English, Turkish and Vietnamese today. It was first used for Sumerian, a language isolate with no known relatives. It was then adapted for Akkadian, the Semitic language of Babylon and Assyria, which became the language of diplomacy across the Near East; even letters between Egypt's pharaohs and other kings in the 14th century BC were written in Akkadian cuneiform. Hittite, Elamite, Hurrian and Urartian were written in it too, and the Persian kings had a simpler cuneiform made for Old Persian.

Scribes trained for years in schools where they copied word lists and model texts, and they wrote everything: contracts, letters, laws, medical recipes, mathematics, astronomy, prayers and literature, including the Epic of Gilgamesh. Babylonian scribes counted in base 60 with a place-value system, a story told in the note on who invented zero. An estimated half a million tablets are held in museums around the world, and comparatively few have been published.

The script faded as Aramaic, written with a short alphabet, spread across the region. The last tablet that can be firmly dated, an astronomical text, comes from about AD 79 or 80. After that, no one could read cuneiform for almost 1,800 years.

How was cuneiform deciphered?

The key was a trilingual inscription, much as the Rosetta Stone was for Egyptian. Darius I of Persia, who ruled from 522 to 486 BC, had an account of his reign carved on a cliff at Behistun, in western Iran, about 100 meters above an old road, in three languages, all in cuneiform: Old Persian, Elamite and Babylonian Akkadian, as described in the account of the Behistun inscription.

1. Old Persian first. It had only a few dozen signs, so it was likely a near-alphabet. In 1802 the German teacher Georg Friedrich Grotefend guessed that repeated groups were the names and titles of kings and identified several signs.
2. Copying the cliff. From 1835 the British officer Henry Rawlinson climbed the rock face to copy the inscription, and by 1847 he had sent a full copy to Europe.
3. Into Akkadian. With the Old Persian text read, the Babylonian version could be matched to it name by name. The Irish clergyman Edward Hincks worked out that Akkadian signs could stand for syllables and that one sign could have several values.
4. The test. Many people doubted that a script with so many values could really be read. 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, Julius Oppert and William Henry Fox Talbot, to translate separately. A jury compared their sealed versions and found them in close agreement on all essential points, with Hincks's and Rawlinson's virtually identical. The decipherment was accepted.

Sumerian, which is related to no known language, was worked out later, largely through the bilingual word lists that Babylonian scribes had made to learn it.

Was cuneiform really the first writing?

It is the earliest known, but it was not the only beginning. Egyptian hieroglyphs appear around 3200 to 3100 BC, so close in time that scholars still debate whether Egypt borrowed the idea from Mesopotamia or invented it independently; the evidence has not settled it. Writing arose on its own in China by about 1200 BC and in Mesoamerica in the first millennium BC. Each followed a similar path, from pictures and numbers to signs for sounds, which suggests the rebus is a discovery people make wherever records grow complex enough.

Cuneiform's own line ended, but its influence didn't. The idea of signs for sounds, refined into one sign per consonant, is part of the story of where the alphabet came from. And writing as a whole is a late arrival compared with speech, as the note on what language is describes: people had been talking for tens of thousands of years before the first accountant in Uruk pressed a reed into clay.

Check yourself

• Without looking back: describe the steps from clay tokens to a written word, and say which step made true writing possible.
• How does the arrow sign for "life" show the rebus principle at work?
• Why did the Behistun inscription matter, and what did the 1857 test prove?

Sources

• The evolution of writing — Denise Schmandt-Besserat, University of Texas at Austin, 2014
• Cuneiform — Wikipedia
• Behistun Inscription — Wikipedia
• Edward Hincks — Wikipedia