Current Biology, Volume 22, Issue 23, R981-R984, 4 December 2012
doi:10.1016/j.cub.2012.11.032
The evolution of writingSummary
- Decoding the world’s oldest as yet undeciphered writing system could help to improve our understanding of the origins of writing and of how this crucial cultural progress spread, branched out, and in some cultures died out.Michael Gross reports.
MAIN TEXT
Just over 5,000 years ago, the earliest cities flourished in Mesopotamia and western Iran. One explanation that has been considered is that making these areas with their seasonal alternations of flooding and drought amenable to agriculture required a complex infrastructure for water retention and irrigation. Thus, only a critical mass of workforce and a certain level of organisation could make these lands habitable, leading to urban settlements like Uruk in Mesopotamia and, a few hundred kilometres to the east, Susa in Iran. On the other hand, humans adapt well to diverse climatic conditions, and southern Mesopotamia may have lent itself to the growth of incipient complex societies without the need for large-scale hydraulic management as more current archaeological theories suggest.
Whatever their origin, the complex societies that built these early cities soon found they needed to keep records of their stocks, property, and commerce. The people of Uruk came up with proto-cuneiform, which evolved into cuneiform. This script encoding words and syllables was eventually supplanted in the Middle East by alphabetic writing systems that spread across the Mediterranean.
A few centuries later, the people of Susa picked up the idea of keeping written records and apparently borrowed a few signs from proto-cuneiform, but mostly they invented their own system of record-keeping, which is now known as proto-Elamite. The writing quickly spread across Iran, even though other archaeological evidence doesn’t suggest that the country had a unified culture at that time. Early excavations in Susa found more than 1,500 clay tablets with proto-Elamite script, and hundreds more have been found since.
As the name indicates, researchers were initially hoping to find a more advanced ‘Elamite’ script that might have followed up on this early version, like cuneiform evolved out of proto-cuneiform. However, systematic archaeological investigation has shown that this next step never happened. After a few centuries of using proto-Elamite, the people of Susa and other towns in Iran stopped writing altogether. For a period of 500 years, there is no trace of writing in Iran, until the introduction of cuneiform from Mesopotamia around 2300–2200 BC and, concurrently, the development as the Susan royal court of a second indigenous script known as Linear Elamite, which is unrelated to proto-Elamite.
In anthropological terms, the proto-Elamite script is the Neanderthal of writing systems. It branched off from our line of descent early on, spread for a while, then became extinct for mysterious reasons. Understanding it better might help us to understand our own cultural evolution. It would help, obviously, if we could decipher it.
Deciphering challenge
Jacob L. Dahl, of the Faculty of Oriental Studies at the University of Oxford, started out studying cuneiform and then became more interested in the path less travelled, the branch that became extinct. Dahl, the world’s leading expert in this writing system heads a research team dedicated to understanding it.
Many have considered the script undecipherable, and linguists are frustrated by the lack of any parallel documents like the Rosetta Stone, which could help, and by the lack of proper prose, as the tablets seem to be listing quantities of commodities owned by specific households. The writing uses 17 numerical and around 1,400 non-numerical signs. Of the latter, Dahl believes, around 100 may have been used as syllables to code for names. These occur where names of owners are listed, are poorly standardised, and have no obvious pictorial meaning.
“An additional challenge,” says Dahl, “is the fact that they use no signs depicting body parts. They must have had a taboo forbidding that. The only exceptions are two pictographically constructed signs for female and male workers, which they took over from proto-cuneiform, obviously without regard to their pictorial associations.”
Early excavations produced large numbers of texts and other artefacts but failed to contextualise these, and, whereas the early publishers worked fervently to quickly publish the results, the published copies cannot necessarily be trusted.
In the late 1970s, the Swedish mathematician Jöran Friberg managed to work out the numerical system used in archaic Iran, based on the observation that the tablets usually contain sums of the quantities listed in each line. Later, Peter Damerow at the Max-Planck Institute for the History of Science and Robert K. Englund at the Free University, both at Berlin, proposed identifications for some of the signs and deciphered the content of some tablets. Thus, signs for containers for weakly fermented beer, dairy products, and grains are understood. Dahl, who worked with Englund at UCLA and with Damerow at Berlin, deciphered a number of signs relating to sheep and goat herding, but still scratches his head over how the proto-Elamite scribes may have referred to cattle, for example.
Dahl has started producing high-quality images of the tablets using a novel device, the Reflectance Transformation Imaging (RTI) system (building on technology developed originally by Tom Malzbender and others at the HP labs in Palo Alto). This uses a black plastic dome lined with 76 separate LED lights and fitted with a 25 megapixel camera. Each artefact is photographed 76 times, each time illuminated from a different angle by one of these LEDs. A dedicated software package combines the 76 shots into a single image file, with which the users can then create different viewing experiences, as if they were looking at the original tablet and shining a torch at it from different sides and angles to get a feeling for the shapes, the textures, and the depths of the grooves.
With these complex images, Dahl now hopes to launch an internet-based crowdsourcing project to help complete the decipherment, following the example of success stories such as Folding@Home (see Curr. Biol. (2012) 22, R35–R38). Combining the ideas and skills of many different people, and including perspectives from mathematics, linguistics, art, and so on, may be the clue to cracking the remainder of the code. Some images have already been made available online, with more and in increasingly higher quality over the coming six months (seehttp://cdli.ox.ac.uk/wiki/proto-elamite for an introduction to proto-Elamite with links to images of the tablets andhttp://cdli.ucla.edu for more on online cuneiform).
Rise and fall
Taken together with the available evidence from the Middle East, a better understanding of proto-Elamite would be a big step towards a complete overview of the evolution of writing, including its birth, branching out, diffusion, and dying out.
Contrary to what one might expect based on today’s notions of literacy, writing did very clearly not arise from a move to record spoken language. Instead, it evolved out of primitive accounting methods, more closely related to abacuses than to storytelling.
“At first, there were tokens used to represent quantities of commodities such as grain,” explains Dahl. “Most tokens were made of clay, a few examples exist of stone tokens. We recently made some clay tokens for our class and were able to replicate the calculations — additions only — in rather complex texts without any use of abstract numbers, or even number words.”
In a second stage, people turned the very same tokens into a permanent recording of the relevant quantities by keeping them in clay envelopes. “These consisted of a hollow clay ball, also called a bulla,” Dahl explains. “It is very likely that tokens continued to be used for centuries.”
Finally, clay tablets bearing the impressions of tokens, and then, similar shapes produced with a stylus became the record keeping, making the tokens redundant. The stylus was usually made of reed, but perhaps also of hard wood or metal: in fact one metal tool that may have been used as a stylus was found at Tepe Yahya close to some proto-Elamite tablets.
The third stage can be pinned down exactly. “Writing is a technology. Regardless of whatever mnemonic devices may have been used in different parts of the ancient Near East, the earliest crystallisation of that technology occurred at the great southern Mesopotamian metropolis of Uruk in the specific context of a large institution dedicated to the city goddess Inanna,” explains Dan Potts from the University of Sydney, Australia. “The earliest texts served to document the incomings and outgoings of those commodities (naturalia, realia) that served to sustain the institution, which some would call a temple complex, and its personnel (not just priests and scribes but agricultural and craft labourers as well). Lexical texts, lists of words classified by domain (names of different categories of animals, trees, plants, professions), represent concrete expressions of early scribal training. It took many more centuries before writing was used to record royal inscriptions, literature, letters and other types of texts.”
Initial proto-writing systems could represent only a certain repertoire of relevant objects — their users could not write down current events or stories they may have told each other. Gradually, the proto-writing systems evolved into complete writing systems, allowing people to write down whatever they could express in words of their language.
The invention of writing happened at least twice and no more than four times in the history of mankind. The two clear cases are cuneiform and the Mayan scripts, both clearly independent inventions that went on to become complete writing systems. Deciphering of Mayan writing has made rapid progress since the 1970s, following the realisation that it is a phonetic representation of a language related to the one still spoken in the area today. It might have been easier if 16th century colonialism had not actively sought to eradicate knowledge of the script that still existed at the time.
Cuneiform was decoded in the second half of the 19th century using trilingual inscriptions written in Old Persian cuneiform, Akkadian cuneiform and Elamite cuneiform. Since the content of some of these texts was known (ruler names and titles of the kings of Persia) this was in fact a very complex linguistic puzzle of replacing signs with sounds and speculate language affiliation.
Some people still question whether the Egyptian hieroglyphs are an equally pristine invention. While there is no similarity in the signs used, Egyptians and, as some have speculated, even ancient Chinese may or may not have gleaned the idea of writing things down for accountancy from Mesopotamia.
Proto-Elamite by contrast, clearly got the idea and a small number of signs from Mesopotamia, and then went on to add a whole range of new signs to the repertoire. It is the earliest writing system that we know to be a derived one. This early branching point is thus the equivalent of speciation in biological evolution. The separate writing systems of these neighbouring regions must have been mutually incomprehensible.
Over a short time span — three centuries at most, but probably much less — the proto-Elamite script spread across Iran, offering a prime example of cultural diffusion. There is no archaeological evidence suggesting a mechanism for this spreading, such as central government or long-distance trading, so the rapid expansion remains one of the mysteries of proto-Elamite.
As the proto-Elamite script spread and developed further, it became richer in its sign repertoire, but Dahl notes that it also ran into problems. “There was an inflation of signs in proto-Elamite,” says Dahl, “and even in high-level accounts, such as those for the household of the ruler of Susa, you see systematic errors and bad practice.” For instance, scribes would cram in information at the end of a line, rather than planning for the space available, like their colleagues in Mesopotamia would have done. And they made elementary mistakes in the bundling of numbers, as it would be a mistake in Roman numerals to write IIIII instead of V.
The key cultural difference is that cuneiform was backed up by a lexical tradition from early on, says Dahl. In Uruk, lists of standardised signs were used for reference. No such lists have ever been found for proto-Elamite. Dahl can’t resist the temptation to speculate that it may have been the failure to invest in the quality of proto-Elamite writing culture that led to its deterioration and ultimately to its downfall.
The ensuing period of five centuries without writing makes Europe’s descent into the Dark Ages pale in comparison. Prophets of linguistic doom who worry about youth slang and text speak will delight in this example of cultural downfall that was possibly triggered or accelerated by bad writing practice. Seeing writing as a trait that has evolved in human populations, it is only natural that it can not only arise, diversify and spread, but also die out. That’s just life.
http://www.cell.com/current-biology/fulltext/S0960-9822(12)01376-0?large_figure=true
Michael Gross