Before Writing

From P2P Foundation
Jump to navigation Jump to search

* Book: Before Writing: Volume 1: From Counting to Cuneiform. By Denise Schmandt-Besserat

Key argument: writing emerged from a quirky Sumerian accounting system involving clay tokens

See also: How Writing Came About


Clay Spinuzzi:

"Tokens, Schmandt-Besserat tells us, were the earliest uses of clay - predating pottery and architecture (p.29). (Occasionally, they were also made of stone, ocher, or plaster; p.30). She notes that when the token system was being developed, it started with simple tokens, then developed additional complex tokens. We know that these tokens comprised the same system because they "were found together in the same sites and the same hoards and were enclosed in the same envelopes"; they became perforated at the same time; and they are "the prototypes of pictographs representing basic commodities in the Sumerian script" (p.29). And whereas in the popular book Schmandt-Besserat described how the token system led to writing, here she also points out that once writing emerged, the token system dwindled, reverting to a few basic shapes (p.29).

Tokens varied in care and manufacturing from site to site. She notes only one possible piece of evidence for a mold; most were clearly made by hand, some even retaining fingerprints (pp.29-30).

Schmandt-Besserat emphasizes that these tokens were found in disturbed contexts: they were used, then discarded. Furthermore, 88.5% of tokens were from sacred precincts (p.62). In fact, tokens were associated with major public buildings decorated with clay cone mosaics (p.70). Complex tokens evolved in tune with the evolution of archaic Eama, she notes (p.73). In comparison, in Susa, "complex tokens do not appear during the fluorescence of the Susa temple but rather after its destruction by fire" (p.84) - a fact that suggests that the complex token system was introduced by conquerors (p.182).

- Who were using these tokens? Most were found in storage units and dumps, although a few were found in houses, swept into hearths (p.95). Very occasionally they are found in tombs.

In the early 4th millennium BC, tokens began to be stored using two related methods. One was clay envelopes - basically hollow balls that were baked with the tokens inside. The other was strings that held perforated tokens; these strings were sealed with bullae that closely resembled small clay envelopes, but were solid, with the ends holding the strings (pp.108-109). Generally, envelopes held simple tokens, while strings held more complex ones (p.110). Counters held in envelopes tended to be smaller and more casually made; they crumble more easily and were likely not fired (p.123).

Marked envelopes were quite rare, but fell into at least three types, which I'll list along with the number of examples: sinking counters into the envelope surface (1); impressing counters onto the envelope surface (14); scratching the counter impression after the envelope dried (2); and possibly perforating the envelope to affix a matching string (7) (pp.127-128). The impressions were usually made with the counters themselves, but sometimes with a stick, stylus, or thumbnail. Some of these techniques were dead ends, but direct impressions and stylus markings were the beginning of writing (p.128).


Remarkably, by the time they began using pictographs, our early scribes had made the leap of abstracting numbers. Whereas counters were repeated in 1:1 correspondence, pictographs never were (p.153).

Now we get to the analysis. Since much of this was covered in my review of her other book, I'll just hit the highlights.

One, Schmandt-Besserat firmly associates developments in writing with developments in civilization. Before agriculture, we have sparse evidence of writing-like activity, mostly notched bones, which carried quantitative data about which we can only speculate (p.160). Agriculture began "a new economy based on sedentariness, new settlement patterns in open air villages, new technologies such as ground and polished stone, and the use of new raw materials such as clay" - and also "generated new symbols," which "were different in form and content from anything used previously": the entirely manmade clay tokens (p.161). Schmandt-Besserat speculates that these tokens were themselves the evolution of a previous system using natural materials (twigs, sticks, grains); but the token system comprised unique shapes "for the unique purpose of record keeping" (p.161). The greatest novelty was that these comprised a system, one that could be expanded (p.161) and that represented both quantitative and qualitative information (ex: three sheep tokens) (p.162).

Intriguingly, this new reckoning technology did not seem to have anything to do with the exchange of goods, but rather for administration in the newly hierarchical context of agriculture (pp.167-168). Indeed, social organization determined the function of writing: egalitarian (tribal) societies only needed to tally, while rank (institutional) societies had to develop accounting to support "an elite overseeing a redistributionist economy" (p.170).

Evidence suggests that counting developed as a status symbol: "writing, therefore, bestowed on the ruler the full control over the input, as well as the output, of the community properties" (p.172). With the rise of the State, complex tokens emerged (along with temples, "monumental architecture, the monopoly of force, and bureaucracy, which point to new strategies in pooling communal resources," p.178). In this context, envelopes and bullae were used to represent unpaid taxes (p.181). In this reading, complex tokens in distant countries represented tribute; recall that in Susa, "complex tokens, envelopes, and impressed tablets appear after the destruction of the temple, when the monumental buildings were replaced by modest structures" (p.182).

Schmandt-Besserat then discusses counting and the emergence of writing. Counting, she points out, is not innate but learned (p.184); in some groups, in fact, counting does not involve "finding out how many items there were in a set" so much as "comparing or verifying a collection" (p.185). If you represent five sheep with five tokens, and you verify your collection by matching each sheep with a token, that's comparing, not counting as we might think of it. The token system, then, didn't initially provide for abstract counting: it didn't separate the item from the count. Bear in mind that "the human brain has not evolved since the appearance of Homo sapiens-sapiens about seventy thousand years ago"; they had the same capacity we do, but the hunting-and-gathering way of life did not challenge them to develop abstract counting (p.189). That changed with tokens, which introduced cardinality tied to object specificity (pp.189-190). Tokens did not express abstract numbers; they were the numbers. But at some point, the increasing complexity of the system led to abstract numbers, which appear on the first pictographic tablets (p.191). Abstract numbers, she speculates, were invented by one nameless individual (p.192), but spread rapidly.

The first abstract numerals were impressed signs, formerly representing quantities of commodities, but now abstracted. They retained a dual meaning for a while, until the concrete meaning was replaced with other symbols (p.193). True pictography, Schmandt-Besserat argues, resulted from abstract counting. Eventually, pictographs began to be used phonetically - a change that, Schmandt-Besserat says, was not socioeconomically triggered as the others were (p.194).

Like her other book, this one really impressed me (pun intended) with the scope of history, the pressures that led to the invention of writing, and the fact that it could have gone so much differently (and in other parts of the world, did, of course)." (

More Information