Clay Tokens: The Neolithic Seeds of Mesopotamian Writing

The Monopoly Pieces of the Ancient Writing Past

Clay Tokens, Uruk Period, Excavated from Susa, Iran
Clay Tokens, Uruk Period, Excavated from Susa, Iran. Louvre Museum (Department of Near Eastern Antiquities). Marie-Lan Nguyen

Writing in Mesopotamia--if you define writing as recording information in a symbolic manner--took an important step forward with the domestication of plants and animals, during the Neolithic period of at least as long ago as 7500 BC. Beginning then, people recorded information about their agricultural goods--including domestic animals and plants--in the form of small clay tokens. Scholars believe that the written form of language that I use to pass this information along today evolved out of this simple accounting technique.


Mesopotamian clay tokens were not the first accounting method used: by 20,000 years ago, Upper Paleolithic people were leaving tally marks on cave walls and cutting hash marks into portable sticks. Clay tokens, however, contained information about what was being counted, an important step forward in communication storage and retrieval.

Neolithic Clay Tokens

Neolithic clay tokens were made very simply: a small piece of clay was worked into one of about a dozen different shapes, and then perhaps incised with lines or dots or embellished with pellets of clay. These were then sun-dried or baked in a hearth. The tokens ranged in size from 1-3 centimeters (about 1/3 to one inch), and about 8,000 of them dated between 7500-3000 BC have been found so far.

The earliest shapes were simple: cones, spheres, cylinders, ovoids, disks, and tetrahedrons (three-dimensional triangles). The premier researcher of clay tokens Denise Schmandt-Besserat argues that these shapes are representations of cups, baskets and granaries.

The cones, spheres and flat disks, she said, represented small, medium and large measures of grain; ovoids were jars of oil; cylinders a sheep or goat; tetrahedrons a person-day of work. She based her interpretations on similarities of the forms to shapes used in the later Mesopotamian proto-cuneiform written language and, while that theory has yet to be confirmed, she may very well be right.

Tokens were non-lingual, meaning that no matter what language you spoke, if both parties understood that a cone meant a measure of grain, you were in business. Whatever they represent, the same dozen or so token shapes were used for some 4,000 years throughout the Near East.

The Sumerian Take Off: Uruk Period Mesopotamia

But, during the Uruk period in Mesopotamia [4000-3000 BC], urban cities blossomed and administrative needs for accounting expanded. Production of what Andrew Sherratt and VG Childe called "secondary products"--wool, clothing, metals, honey, bread, oil, beer, textiles, garments, rope, mats, carpets, furniture, jewelry, tools, perfume--all of these things and many more needed to be accounted for, and the number of types of tokens in use ballooned to 250 by 3300 BC.

In addition, during the Late Uruk period [3500-3100 BC], tokens began to be kept in sealed globular clay envelopes called "bulla" (illustrated on page 2). Bulla are hollow clay balls about 5-9 cm (2-4 in) in diameter: the tokens were placed inside and the opening pinched shut. The exterior of the ball was stamped, sometimes all over the surface, and then the bulla were fired. About 150 of these clay envelopes have been recovered from Mesopotamian sites.

Scholars believe that the envelopes were meant for security purposes: that the information needed to be protected from being changed at some point.

Eventually, people would impress the token forms into the clay on the outside, to mark what was inside. Apparently, by about 3100 BC, bulla were replaced by puffy tablets covered with the impressions of the tokens and there, says Schmandt-Besserat, you have the beginning of real writing, a three-dimensional object represented in two dimensions: proto-cuneiform.

Persistence of Clay Token Use

Although Schmandt-Besserat argued that with the dawn of written forms of communication, tokens stopped being used, MacGinnis et al. have noted that, although they did decrease, tokens continued in use well into the first millennium BC. Ziyaret Tepe is a tell in southeastern Turkey, first occupied during the Uruk period; the Late Assyrian period levels are dated between 882–611 BC.

A total of 462 baked clay tokens have been recovered from those levels to date, in eight basic shapes: spheres, triangles, disks, tetrahedrons, cylinders, cones, oxhides (squares within indented sides) and squares.

Ziyaret Tepe is only one of several later Mesopotamian site where tokens were used, although tokens do seem to drop completely out of use before the Neo-Babylonian period about 625 BC. Why did the use of tokens persist some 2200 years after the invention of writing? MacGinnis and colleagues suggest that it was a simplified, para-literate system of recording that allowed more flexibility than the use of tablets alone.


Near Eastern Neolithic clay tokens were recognized and studied first in the 1960s by Pierre Amiet and Maurice Lambert; but the major investigator of clay tokens is Denise Schmandt-Besserat, who in the 1970s began studying the curated corpus of tokens dated between the 8th and 4th millennium BC.


This article is a part of the guide to the Mesopotamia, and the Dictionary of Archaeology.

Algaze G. 2013. The end of prehistory and the Uruk period. In: Crawford H, editor. The Sumerian World. London: Routledge. p 68-94.

MacGinnis J, Willis Monroe M, Wicke D, and Matney T. 2014. Artefacts of Cognition: the Use of Clay Tokens in a Neo-Assyrian Provincial Administration. Cambridge Archaeological Journal 24(2):289-306. doi:10.1017/S0959774314000432

Schmandt-Besserat D. 2012. Tokens as precursors of writing. In: Grigorenko EL, Mambrino E, and Preiss DD, editors. Writing: A Mosaic of New Perspectives. New York: Psychology Press, Taylor & Francis. p 3-10.

Schmandt-Besserat D. 1983. Decipherment of the Earliest Tablets. Science 211:283-285.

Schmandt-Besserat D. 1978. The earliest precursors of writing. Scientific American 238(6):50-59.

Woods C. 2010. The Earliest Mesopotamian Writing. In: Woods C, Emberling G, and Teeter E, editors. Visible Language: Inventions of Writing in the Ancient Middle East and Beyond.

Chicago: The Oriental Institute of the University of Chicago. p 28-98.

Woods C, Emberling G, and Teeter E. 2010. Visible Language: Inventions of Writing in the Ancient Middle East and Beyond. Chicago: The Oriental Institute of the University of Chicago.