Writing
The early texts, the proto-cuneiform tablets of Uruk from the end of the fourth millennium BCE, are almost entirely economic: about 85 percent of them are bookkeeping records of temple estates. Writing began not for literature but for counting grain, livestock and rations; signs for numbers appeared before signs for words.
A written procedure no longer depends on the memory of its bearer: it can be accumulated, checked and contested.
Why this date
archaeological, range 3350-3100 BCE: proto-cuneiform tablets of Uruk IV (CDLI 3350-3200, Englund proposes c. 3350), labels from tomb U-j at Abydos (Egypt) dated c. 3200 by context and radiocarbon
What it grew from
-
Counting and recording quantity
influenced · Confidence: medium
Writing grew out of accounting: per Schmandt-Besserat, clay counting tokens (8000-3000 BCE) gave rise to the first signs, and the oldest Uruk texts are bookkeeping records; sources do not assert a direct link to Palaeolithic tallies.
Dispute: The token theory is contested: Bennison-Chapman (Cambridge Archaeological Journal 29, 2019), on 3000 objects from 56 Neolithic sites, concludes that "Clay objects were not created to administer agricultural produce and there is no evidence to suggest that in the Neolithic they formed part of a unified symbolic system"; earlier doubts from Zimansky (1993), Michalowski (1993), Englund (1998). Schmandt-Besserat herself starts the chain with tokens c. 7500 BCE and writes that "Palaeolithic hunters had no need for counting to redistribute meat one piece at a time", so the tallies of c. 20,000 BCE are linked to writing only through the general idea of recording quantity.
Link sources
- Schmandt-Besserat D. «The Evolution of Writing» (2014) Verified 20 Sep 2026
The development from tokens to script reveals that writing emerged from counting and accounting.
- Schmandt-Besserat D. «The Evolution of Writing» Verified 20 Sep 2026
After four millennia, the token system led to writing.
- Spar I. «The Origins of Writing», Metropolitan Museum of Art (2004)
The site was dominated by large temple estates whose need for accounting and disbursing of revenues led to the recording of economic data on clay tablets.
- Schmandt-Besserat D. «Tokens: their Significance for the Origin of Counting and Writing» (2014) Verified 20 Sep 2026
Writing brought abstraction to data processing: the signs abstracting tokens were no longer tangible; abstract numerals such as '1' '10' '60' replaced one-to-one correspondence.
- Schmandt-Besserat D. «The Invention of Tokens» Verified 20 Sep 2026
Palaeolithic hunters had no need for counting to redistribute meat one piece at a time. But computing the cereal reserves of farming communities went far beyond subitising, the minimal innate numerical capacities humans are born with.
"Verified" means that on that day the quotation was found at the address given, word for word. No date means we have not been able to open the source yet, not that the quotation is wrong.
- Schmandt-Besserat D. «The Evolution of Writing» (2014) Verified 20 Sep 2026
What grew out of it
-
Positional decimal system and zero
influenced · Confidence: low
Positional decimal notation exists as a written notation: Indian numerals are signs of the Brahmi script, and the zero dot and positional numbers are known from inscriptions and manuscripts; but no source asserting this link directly has been found.
Dispute: Chrisomalis (Numerical Notation, 2010) treats numerical notations as a separate class of sign systems rather than a part of writing, and stresses that before modern times arithmetic was done on a counting board while notation only recorded results; the positional principle is possible on a board as well (Chinese rod numerals), so writing is not a physical prerequisite but the medium in which place value became a notation. Britannica ignores the Cambodian inscription of 683 with a zero symbol (Coedès 1931; Swetz).
Link sources
- Britannica, «Numerals and numeral systems», «The Hindu-Arabic system» (2024) Verified 24 Sep 2026
None of these early Indian inscriptions gives evidence of place value or of a zero that would make modern place value possible. Hindu literature gives evidence that the zero may have been known earlier, but there is no inscription with such a symbol before the 9th century.
- Chrisomalis S. «Re-counting the Cognitive History of Numerals», The MIT Press Reader (2021)
arithmetic was an embodied and material practice (e.g., on a counting board or abacus) whereas numerical notation was used for writing and recording.
"Verified" means that on that day the quotation was found at the address given, word for word. No date means we have not been able to open the source yet, not that the quotation is wrong.
- Britannica, «Numerals and numeral systems», «The Hindu-Arabic system» (2024) Verified 24 Sep 2026
The share of bookkeeping records comes from a summary sentence in the Wikipedia article. Englund himself («Accounting in Proto-Cuneiform», The Oxford Handbook of Cuneiform Culture, p. 35; the chapter is openly available on CDLI) puts it differently, as an average across periods with a caveat about Uruk IV and Uruk III. A verbatim quotation from his chapter is needed.
Node sources
- Schmandt-Besserat D. «The Evolution of Writing» (2014) Verified 20 Sep 2026, quotation assembled from fragments
About 3200 BC, once the system of impressed signs was understood, clay tablets...replaced the envelopes filled with tokens.
- Spar I. «The Origins of Writing», Heilbrunn Timeline of Art History, Metropolitan Museum of Art (2004)
The site was dominated by large temple estates whose need for accounting and disbursing of revenues led to the recording of economic data on clay tablets.
- Archaeology Magazine, «Earliest Egyptian Glyphs» (1999) Verified 20 Sep 2026
The tags and ink-inscribed pottery vessels have been dated to 3200 B.C. based upon contextual and radiocarbon analysis.
- Wikipedia «Proto-cuneiform», Englund R. K. «Accounting in Proto-Cuneiform», The Oxford Handbook of Cuneiform Culture, p. 35., CDLI, further_reading (2011), 35 Verified 20 Sep 2026
The vast majority of proto-cuneiform texts, approximately 85% of the total (about 2000 from the Uruk IV period and 3600 from Uruk III), are administrative, bookkeeping records.
"Verified" means that on that day the quotation was found at the address given, word for word. No date means we have not been able to open the source yet, not that the quotation is wrong. "Assembled from fragments" means the words do not stand together on the page: we found each fragment separately and joined them with an ellipsis.
Further reading
Not evidence: these carry no quotation and support nothing we wrote. They are books and scans to read next.