From counting to language: How writing evolved
The earliest version of cuneiform wasn't used to write language at all—it was used to count! And that Sumerian system of counting still influences our counting systems today. Here's the story of Sumerian numerals.
Contents
Prefer to watch or listen to this article instead? Here's a video version:
Introduction
The earliest known writing system is cuneiform, which was first used to write the Sumerian language c. 3300 BCE, and later the Akkadian, Babylonian, Assyrian, Elamite, Hurrian, Hittite, Old Persian, and Ugaritic languages as well (Figure 1).
But the earliest version of cuneiform wasn’t used to write language at all—it was used to count! And that Sumerian system of counting still influences our counting systems today.
This is the story of Sumerian numerals.
Proto-Cuneiform
Sumer is the earliest known civilization (c. 5500–1800 BCE). As Sumerian society grew in complexity and established rich trade networks throughout Mesopotamia, the development of systems of record keeping and accounting became paramount. Previously Neolithic or tribal societies had less use for extensive systems of record keeping or even counting, and this is still true today. The Ös language of Central Siberia, for example, doesn’t have a word for thousand, which one speaker explained by saying, “In the olden days, our people never need to count a thousand things … so there’s no word for it.” (Harrison 2007: 189). The Yanoama language of the Amazon lacks words for numbers higher than 3 (Harrison 2007: 187). The Pirahã language of Brazil is famously claimed by linguist Daniel Everett to lack numbers entirely. Many languages spoken by small communities do have rich and complex systems of counting that reach very high numbers, however. Such complex systems just aren’t always necessary.
Daniel W. Hieber, Ph.D.
The growth of complex societies like Sumer, on the other hand, necessitates the use of extensive counting and tallying for the purposes of trade and administration, which is precisely why cuneiform evolved. Starting around 8000 BCE, clay “tokens” appear in the archaeological record in the Middle East—small, nondescript clay objects bearing tally marks (Figures 2–3). Differently shaped tokens were used to tally different kinds of objects. For example, the crossed token seems to have been used to record numbers of sheep.
When shipping goods, merchants would enclose tokens in small spherical clay containers called bullae (Latin ‘bubble, blob’; see Figure 4). The recipient would then open the container to verify that the correct quantity of goods had been received.
Merchants later began impressing the token or a drawing of the token on the outside of the bulla, which obviated the need for the bulla in the first place. The merchant could simply send impressions of the symbols instead, leading to the earliest Sumerian pictographs. At least 30 Sumerian signs correspond closely to the shape of a specific token (Figure 5). Though 30 is a small number in comparison to the ~800 known proto-cuneiform signs, the correspondences nonetheless strongly suggest a connection (Coulmas 1996: 506–509; Gnanadesikan 2009: 15).
But while the use of bullae became more restricted, the pictographs grew in importance. Around 3300 BCE, the first proto-cuneiform tablets appear in the Sumerian city of Uruk. I say proto-cuneiform and not simply cuneiform because it was still not yet at a stage where it represented language. The difference between writing and proto-writing is that writing is a static representation of language, whereas proto-writing is a static representation of information that isn’t systematically related to a specific language. Tally marks, property/name marks, Incan quipu, certain types of cave paintings, and pictorial signs all generally fall under the term proto-writing (Coulmas 1996: 421).
