A lost ancient script reveals how writing as we know it really began

Early writing is a tale of two scripts. Egyptian hieroglyphs and Mesopotamian cuneiform both emerged independently about 5300 years ago. The political powers of ancient Egypt and Mesopotamia flourished in the centuries to come, partly because writing helped states control the flow of goods and consolidate power. The pen (or ancient stylus) was mightier than the sword.

Or so the conventional story goes. But there is a glaring omission here because, at the dawn of writing, there weren’t two scripts. There were three. That third, mysterious script, called proto-Elamite, appeared in ancient Iran while cuneiform and hieroglyphs were both in their infancy – and has been shockingly overlooked by all but a handful of scholars since its discovery 125 years ago.

That is beginning to change, with far-reaching consequences. Although proto-Elamite remains largely undeciphered, there is tantalising evidence that it became by far the most advanced of the three scripts in operation about 5000 years ago. What we now know about the script’s story is so surprising and counterintuitive that we might need to rewrite the early history of writing.

Remarkably, this obscure writing system could represent a giant leap forward in how we represent speech in written form. Spoken language might be 1.7 million years old, but it wasn’t until proto-Elamite that we may finally have been able to start writing down exactly what we were saying. So why, then, did this incredible script vanish not long after it was invented?

Proto-Elamite writing tablets have been turning up at archaeological sites across the Iranian plateau since 1899. Most were found at the ancient city of Susa, which is associated with the Elam culture that appeared about 4500 years ago. But the tablets predate the rise of Elam, which is why the writing system has been named proto-Elamite. The latest thinking is that the oldest tablets are about 5200 years old, suggesting they slightly postdate the earliest texts written using Egyptian hieroglyphs or an early version of cuneiform dubbed proto-cuneiform.

Proto-Elamite was probably inspired by proto-cuneiform, according to Jacob Dahl at the University of Oxford. This is hardly surprising, given that Susa – now long abandoned – is just a few hundred kilometres from the ancient Mesopotamian city of Uruk, which was a major centre of proto-cuneiform writing. Just like the Mesopotamian script, proto-Elamite was inscribed into wet clay using a stylus, and some signs are almost identical, such as the one for “sheep” – a cross inside a circle. The ancient scripts were used in a similar way, too, primarily to keep economic records.

There are, however, other ways to interpret the origins of proto-Elamite. Dating ancient clay tablets can be tricky, partly because many were dug up more than a century ago during less-than-meticulous excavations. As a result, some researchers think it is possible that proto-Elamite is just as old as the other two writing systems, with all three emerging independently. Here, the similarities with proto-cuneiform are explained by both scripts borrowing signs and conventions from earlier, pre-writing systems used across south-west Asia. “I would struggle to say one writing system follows the other,” says Amy Richardson at the University of Reading, UK.

Regardless of its exact origins, proto-Elamite is a far more obscure and mysterious writing system than proto-cuneiform. Even today, it remains largely undeciphered. Although we know how numbers were written (see “Complex counting“), it isn’t clear what most of its non-numerical signs represented. This is partly because of the choices ancient Iranian scribes made when inventing the script. While many proto-cuneiform signs are clearly pictures that hint at each sign’s meaning – a human hand to represent “give” or a spiked stem to represent “barley” – proto-Elamite signs are typically more abstract, so it is far less obvious what they represent.

This feature does, however, give proto-Elamite a surprisingly modern appearance, given that the signs and letters in most of today’s writing systems are also abstract. This air of modernity is enhanced by the fact that ancient Iranian scribes wrote in lines, which were read from right to left. Mesopotamian scribes had a more complex writing process in which information was encoded in boxes, which makes proto-cuneiform tablets look a little like the output of a spreadsheet app rather than a word-processing program.

Dahl’s work over the past 25 years has transformed our understanding of proto-Elamite. In the early 2000s, he and his then-doctoral supervisor, the late Robert Englund, began a project to digitise all 1700 known proto-Elamite texts and make them freely available online for study.

Through careful analysis of the original clay tablets, Dahl has also produced and refined a list of proto-Elamite’s non-numerical signs. Their exact number isn’t clear, because it is difficult to determine whether two slightly different signs are truly distinct or whether they simply show allowable variation within a single sign.

A proto-Elamite dictionary

Dahl’s current estimate is that proto-Elamite signs numbered in the hundreds to low thousands. The hope is that patterns in the way these signs appear in the texts will help us define a great many of them, effectively giving us a proto-Elamite dictionary that we can then use to read the tablets.

Initial work is encouraging. For instance, one challenge has been to identify the sign for “cow”, an animal we know was important to the ancient Iranian economy because archaeologists have found cattle bones at many sites. One sign has been tentatively identified as a “cow” sign – but, strikingly, it never appears on tablets bearing a sign that we know means “plough”. This might suggest the “cow” sign proposal is wrong.

But over the past five years, a team including Kathryn Kelley at Uppsala University in Sweden and Logan Born, formerly at Simon Fraser University in Canada, has analysed Dahl’s online archive of proto-Elamite texts using computer software. The team’s work has identified a strong but hidden connection between the “plough” and proposed “cow” signs. Despite never co-occurring, both are members of a broader group of signs that do appear together – signs that presumably are linked to the world of farming.

The software analysis revealed other features, too. The ancient scribes sometimes placed one sign inside another – a little like placing a letter “A” inside a letter “O”. The exact meaning of these combined signs is unclear, but seems to indicate an overlooked “grammar” in the way combinations of characters were made. These combinations are found on tablets at sites across what is now Iran, suggesting there was a degree of standardisation in how different scribes followed proto-Elamite’s rules.

Despite such work, progress towards decipherment is slow. Even so, the researchers who have studied the script largely agree on one point: proto-Elamite was by far the most advanced writing system in operation 5000 years ago.

Writing at this early stage was incredibly simple. With little more than a collection of signs representing ideas to work with, scribes could record information only in note form – for instance, using the signs “man”, “goat” and “50” to document that a particular person had a herd of 50 goats. But there is evidence that proto-Elamite had escaped this limitation, and that ancient Iranian scribes had begun to use signs to encode spoken language.

It is difficult to overstate what a significant breakthrough this was. Spoken language may be up to 1.7 million years old, and it has evolved into a complex and nuanced communication system. When writing began encoding speech, it instantly gained most of that complexity. “It piggybacked on the amazing functionality of language to communicate,” says Piers Kelly at the University of New England in Australia. Today, we take for granted that writing can be used to persuade, delight or anger a reader – but it can do so only because it encodes speech.

In fact, Kelly and many other researchers argue that encoding speech isn’t just an important feature of writing, but its defining one. This would mean that scripts like proto-cuneiform that don’t encode speech aren’t really writing at all. Accept that argument and proto-Elamite – if it really did encode spoken language – was the world’s first true writing system.

Evidence for this language encoding comes from proto-Elamite tablets in which non-numerical signs occur in curious sequences between four and 12 signs long. These sequences are difficult to explain if the signs represent objects. But they would make more sense if the signs instead represented syllables in long, multisyllabic words – almost certainly the names of important people.

Additional support for the proposal comes from Dahl’s work on the proto-Elamite sign list. It turns out that ancient scribes used a subset of about 100 signs to write the curious sequences. This number is significant because many spoken languages are constructed from roughly this number of distinct syllable sounds, so writing systems that record a language’s syllabic speech in full often contain between 40 and 100 signs (see “Three ways to write”).

“If those signs really are an early syllabary, that would be so exciting,” says Dahl, given that Egyptian hieroglyphs and Mesopotamian cuneiform didn’t encode syllabic speech in full for another 500 years. He cautions, however, that it is just an idea, albeit a popular one.

But if the ancient Iranians did invent the most advanced writing system of their time, what happened next? Over the past few years, two very different scenarios have emerged.

The first is both exciting and profound. It suggests the ancient Iranians began a long-lasting relationship with writing similar to that seen in Mesopotamia and ancient Egypt. “We are dealing with three cradles of writing,” says François Desset at the University of Liège in Belgium.

Desset reached this conclusion partly through his work on another ancient Iranian script that was discovered during excavations at Susa 125 years ago. This script, known as Linear Elamite, was in use in ancient Iran about 4100 years ago.

An Elamite Rosetta stone

Linear Elamite has long been considered just as mysterious as the proto-Elamite script. Then, in 2020, Desset announced that he had successfully deciphered Linear Elamite. He realised that the Linear Elamite texts inscribed onto a set of silver goblets were prayers – and that we already knew the contents of those prayers because they also appear on another set of goblets, but written using a readable script. By comparing the two sets of goblets, Desset worked out how to read Linear Elamite. The approach is similar to the way in which Egyptian hieroglyphs were deciphered using copies of the same text written in different scripts on the Rosetta stone.

Desset worked out that the Linear Elamite signs encode syllables, and he began assigning sound values – “ha”, “pe”, “su” and so on – to each one. “I can read about 96 per cent of the signs in the inscriptions,” he says.

He and his colleagues gave details of this decipherment in a 2022 study. But the researchers made an additional claim in their work. They highlighted similarities in the appearance of some Linear Elamite and proto-Elamite signs, arguing that this is evidence that the two scripts are actually the same writing system at different stages in its development. “There is a continuity of the scribal tradition on the Iranian plateau, which was mostly overlooked up to now,” says Desset.

Other researchers, including Dahl, are sceptical of this claim, instead arguing that a different scenario played out all those years ago. They think that shortly after proto-Elamite became potentially the most advanced script of its day, the ancient Iranians abandoned it and gave up writing. “They simply rejected it,” says Dahl.

He says this scenario fits with the evidence – or rather, the lack of evidence. There is very little in the archaeological record to suggest that the ancient Iranians wrote down anything between 4900 and 4100 years ago. One or two disputed examples aside, we know of no proto-Elamite or Linear Elamite texts from this 800-year period.

This suggests to Dahl that Linear Elamite wasn’t a continuation of proto-Elamite, but a distinct and independent invention of writing – perhaps recycling some of the signs found on long-discarded tablets. This scenario is arguably far more surprising than the one outlined by Desset and his colleagues, because it doesn’t fit with modern thinking on literacy. It is difficult for us to imagine a society willingly discarding writing – particularly a script that had potentially made an enormous technological leap by encoding speech. But such a rejection actually fits well with an ongoing reassessment of human history.

Over the past five years, archaeologists have begun questioning long-held assumptions. Where once they traced a simple path from ancient hunter-gatherers to early farmers to the rise of civilisations, they now accept that the story was more complicated. For instance, while farming was practised in Britain when Stonehenge was built 5100 years ago, populations then reverted to hunter-gathering. Societies don’t all make the same choices, and even two societies that follow the same path don’t necessarily continue to develop in the same way.

Ditching writing

Dahl suspects this explains why writing failed to take hold in ancient Iran. In recent years, he and one of his colleagues at the University of Oxford, Parsa Daneshmand, have both published articles suggesting that ancient Iranians chose to reject writing because words are often used to control populations. “Proto-Elamite was a repressive system, a way to keep track of goods and allow those in charge to say: ‘You didn’t bring in enough,’” says Dahl. It was probably extremely unpopular.

The same was presumably true of proto-cuneiform and early Egyptian hieroglyphs, which means the real surprise isn’t that proto-Elamite failed, but that the other two writing systems succeeded. Perhaps, says Dahl, they did so because the elites who benefited from writing lost power in ancient Iran, but maintained control in Mesopotamia and ancient Egypt.

It is worth considering both Desset’s and Dahl’s scenarios for the fate of proto-Elamite, says Kelley. “With an undeciphered script, you should always explore every option.” But she and many other researchers find Dahl’s scenario more plausible. This isn’t just because there seems to be an 800-year writing gap in the ancient Iranian archaeological record. It is also because the ancient Iranians were apparently unenthusiastic about literacy even when they did write.

For instance, almost from the moment Mesopotamians began writing, there is evidence they invested heavily in the idea: archaeologists have found plenty of proto-cuneiform tablets that were clearly teaching aids, designed to train more scribes and embed a writing tradition in Mesopotamia’s urban centres. In contrast, there are no known proto-Elamite teaching texts.

“Proto-Elamite was never particularly committed to in the same way that we possibly see with proto-cuneiform,” says Richardson.

This lack of commitment is also reflected in the number of texts. Although there are 1700 known proto-Elamite tablets, this is a small number compared with the 8000 proto-cuneiform tablets. Linear Elamite was used even more sparingly: only about 40 short inscriptions have been discovered.

But in a final twist, ancient Iran’s aversion to writing didn’t hold it back. In fact, the illiterate ancient Iranians grew more powerful than their literate Mesopotamian neighbours.

“If you go down to the Sukkalmah period [around 3800 years ago], there’s no doubt that Mesopotamians look at Iran and say: ‘They are far more important than we are. They’re stronger and richer than we’ll ever be,’” says Dahl. “They actually call the king of Elam ‘our father’.”

But, says Dahl, the king of Elam had little interest in writing. Even in an increasingly literate world, he – and his subjects – knew that writing wasn’t essential for success.

Scientific pioneers of the ancient world, Cairo and Alexandria: Egypt

Embark on an unforgettable journey through Egypt’s two most iconic cities, Cairo and Alexandria, where ancient history meets modern charm.