Two precursors of writing: plain and complex tokens
From Escola Finaly
| Two Precursors of Writing: Plain and Complex Tokens
| by Denise Schmandt-Besserat
| Collection: Denise Schmandt-Besserat 2 - 1st edition: Barcelona, 8 December of 2008
License: © Denise Schmandt-Besserat
(The Origins of Writing / edited by Wayne M. Senner. 1991: 27-41.)
- 1. Plain and Complex Tokens
- 2. The assemblages
- 3. Chronology
- 4. Geographic distribution
- 5. Function
- 6. Interpretation
- 7. The Introduction of Two Kinds of Signs
- 8. The invention of numerals
- 9. Summary
- 10. Further Readings
- 11. Notes
- 12. See also
1. Plain and Complex Tokens
Before the Sumerian writing system—the first script ever developed—was invented at the end of the fourth millennium B.C., accounting was practiced in the ancient Middle East by means of small counters.1 These were small tokens modeled in clay in different shapes, each symbolizing a particular commodity. The token system consisted of two kinds of tokens—"plain" and "complex." The two types were similar in many ways, but each managed to develop a life of its own. They had, for example, different appearances, chronologies, geographic extensions, meanings, and functions (see map). As a result, each type of token gave rise to a different type of sign in the Sumerian script and can be considered as a separate precursor of writing.
2. The assemblages
Plain and complex tokens were of identical manufacture but can be distinguished either by their shapes or surface treatment (figs. 1 and 2). Plain tokens are characterized both by a simple geometric form and a plain, smooth surface, devoid of any markings. Their shapes include spheres, flat and lenticular disks, cones, tetrahedrons, and cylinders. These forms seem fully arbitrary and to be dictated only by the concern for making, with the least effort, shapes easy to identify and to duplicate. There is no way of knowing, however, whether they suggested daily life commodities. It is conceivable, for example, that cones depicted pointed vases.
Assemblages of complex tokens include a greater repertory of forms and markings.2 Among the shapes of complex tokens are biconoids, ovoids, bent coils, rhomboids, parabolas, quadrangles, and triangles, as well as miniature representations of tools, utensils, containers, and animals. Some of these, such as a series of small vessels, required skill to manufacture.
The application of markings to the face of tokens is another distinctive feature of the complex counters. These markings consisted of linear patterns, notches and punctations, traced or impressed with a stylus, with rare examples of appliqué pellets. Markings were applied to the traditional plain token shapes, such as spheres, disks, cones, tetrahedrons, and cylinders (thus converting them into complex tokens) as well as to the typical complex forms, such as biconoids, ovoids, bent coils, triangles, parabolas, rhomboids, and quadrangles.
The fact that both plain and complex tokens belong to the same accounting device is undeniable for several reasons. First, all specimens bear obvious family resemblances, sharing the same size, material, color, and method of manufacture. Second, both plain and complex tokens occur in the same basic shapes, namely, spheres, disks, cones, tetrahedrons, ovoids, and quadrangles, either plain faced or covered with markings. Third, examples of both categories of artifacts start to be perforated at the same time in order to be strung. Fourth, plain and complex types are found together in hoards and, furthermore, may be enclosed together in the same envelope.3 Fifth, plain and complex tokens were perpetuated by pictographs of the Sumerian script expressing common commodities.
Plain tokens made their appearance with the beginning of agriculture; complex tokens, not until the rise of cities. The earliest assemblages of plain tokens have been recovered in the remains of villages of the Fertile Crescent dating to 8000-7500 B.C. These villages, built with round huts typical of the period of transition between hunting-gathering and farming cultures, relied upon grain consumption; they show no obvious evidence for animal domestication. They participated in a trade network, attested by the presence of obsidian tools at each site except one. Among these villages, Tell Aswad I, Tell Mureybet III, and Cheikh Hassan in Syria were fully sedentary settlements and show direct evidence for the cultivation of cereals. On the other hand, Tepe Asiab and Ganj Dareh Tepe E were perhaps no more than semipermanent encampments of hunters and gatherers.4
The first occurrence of tokens at Tell Mureybet, in the third level of occupation of the site, is particularly revealing. There were no tokens in Mureybet I and II, when the village economy was based on hunting and gathering but already traded obsidian. Tokens coincide in Mureybet III with such new features as a quantum jump in quantity of cereal pollen in the soil, the first evidence for the cultivation of grain in fields around the site; the construction of rectangular silos; and a substantial increase in the population, which implies a new social structure.5 Accordingly, the invention of a record keeping device in the ancient Middle East appears to have little to do with animal domestication and herding. The correlation with trade is also not convincing. Instead, the need for counting and accounting seems to be related, in that part of the world, with an economy based on hoarding and cultivating cereals and the socioeconomic changes that followed agriculture.6 It seems indeed logical that an economy involving the planning of subsistence over the seasons would require record keeping.
Complex tokens belong to the later part of the fourth millennium B.C., which is characterized in the ancient Middle East by the urban phenomenon and the rise of the Sumerian temple institution, viewed as the origin of state formation. The first occurrence of complex tokens is best documented in the Sumerian metropolis of Uruk. There, the earliest group was recovered in the ruins of Eanna, the major temple precinct dedicated to the goddess of love, Inanna. They belonged to level VI of the temple, dated to about 3350 B.C., which is also the level when buildings decorated with colorful clay cone mosaics were introduced into the precinct.7 It is particularly significant that complex tokens coincide with these architectural features because they bring the evidence for the first monumental public buildings, which in turn mark the emergence of Eanna as a predominant economic institution in the ancient Middle East. We have some insights into the economy implemented by the Sumerian temple between 3350 and 3100 B.C. corresponding to levels VI-IV. It was based on the pooling together, management and redistribution, of a substantial surplus produced by the community. Sumerian art has preserved for posterity the representation of processions of individuals delivering their dues to the temple in the form of goods in kind. The En, or chief administrators, are often seen leading the procession, recognizable by such status symbols as a beard, a special headdress and a long garment (fig. 3). Levels VI-IV are also characterized by a profusion of typical vessels such as beveled rim bowls and nose-lugged jars, which are believed to have served as standardized measures for the deliveries of dues to the temple.8 There is also evidence for a strengthening in the administration, with an increase in the use of seals, and in particular, with the introduction of cylinder seals. Interestingly, some of these have carvings showing the En presiding over scenes of torture, such as beating, probably inflicted on the first tax delinquents.9 In this perspective, the quantum jump in the number of token shapes coincides with the establishment of a coercive redistribution economy. The imposition of taxation required an authority and administration to implement it, a system of measures and a precise reckoning device for record keeping, large storage facilities, and a system of penalties for noncompliance. These needs explain, in levels VI-IV of Eanna the first evidence for the En, the cylinder seals, beveled rim bowls, complex tokens, monumental architecture, and the scenes of tortures depicted on seals. The complex tokens can be viewed, therefore, as fulfilling an important function in the collection of taxes, which is crucial to state formation.
Tokens bearing simple markings such as one or two strokes or notches were already present in the earliest token assemblages of the early eighth millennium B.C.10 Such tokens, however, remained exceedingly rare until the remarkable increase in the number and variety of markings that coincided with the multiplication of token shapes characteristic of the complex tokens. Throughout the fourth millennium B.C., plain tokens continued to exist unchanged. The complex tokens never supplanted the plain ones but, rather, complemented them.
The evolution of a reckoning device can logically be assumed to mirror the socioeconomic development of a society. It is therefore not surprising that the two major events in the development of the token system correspond to the two major economic transformations that occurred in the ancient Middle East: the invention of the counting device coincides with the transition to agriculture, and a quantum jump in the complexity of the system occurs at the rise of the Sumerian temple, which was to lead to state formation.
4. Geographic distribution
Geographic distribution is a third major difference between plain and complex tokens. Plain counters pervaded small and large settlements of the ancient Middle East, but the complex assemblages occur only in selected fourth millennium sites.
Plain tokens are reported in practically every site of the eighth millennium to the fourth millennium B.C., excavated from Anatolia to Palestine and from Syria to Iran, showing that during this period of some four thousand years plain tokens were ubiquitous in the region. From settlement to settlement the collections vary only in number, some areas producing a handful of the artifacts and others yielding several hundred and, in one site—Jarmo, Iraq—more than one thousand.11 Plain tokens seem, therefore, to know no boundaries in the ancient Middle East.
The complex counters, in contrast, have a limited extension; for example, none has been recovered, so far, in Turkey or in Palestine. They seem to be, in fact, a southern Mesopotamian phenomenon, extending only sporadically to isolated sites in the north, no farther than the adjacent Susiana plain toward the east, and to rare sites along the Euphrates River toward the West. In Sumer, complex tokens were used at Uruk, Girsu, Ur, Nippur, and Ubaid. In the north, Tell Billa has yielded a few, but Tepe Gawra none. In Susiana, complex tokens have been found only in Susa, Chogha Mish, Moussian, and KS 54.12 In Syria they are included in the assemblages of Habuba Kabira, Tell Kannas, and Jebel Aruda. The vast discrepancy in the number of complex tokens recovered at each of these sites may reflect more than archeologists' luck. For example, the main metropolis of Uruk and Susa both yielded large assemblages of about eight hundred tokens each, compared with a single example at Ubaid or Jebel Aruda.
Whereas plain tokens were used in all possible settings, including cities, towns, villages, and even cave dwellings, complex tokens occur mostly in urban centers. Moreover, the sites which produced complex specimens usually share a very particular assemblage consisting of clay cone mosaics for the decoration of public buildings; cylinder seals, with examples carved with such motifs as the En in his typical attire; and pottery vessels among which are beveled rim bowls and incised nose-lugged jars. These features, which are characteristic of Eanna levels VI-IV, were a foreign intrusion in Susiana, Syria, and northern Mesopotamia.13 The distribution of complex tokens seems, therefore, to identify centers directly under the influence of the Sumerian temple.
Both kinds of counters served the same purpose, namely, they were part of the same mnemonic device used to organize and store economic data. There is indication, however, that each type of token was handled by different hands and, in particular, was stored in different ways. Plain counters were enclosed in hollow spherical envelopes, whereas complex tokens were strung to a solid oblong bulla (figs. 4 and 5).
It is not hard to conceive the help provided by tokens which translated economic data into symbols easy to manipulate. One can visualize how counters, representing units of real goods in one-to-one correspondence, were lined up in front of accountants who organized them according to types of goods, producers or recipients, entries or expenditures, or any other criteria. Furthermore, the tokens could be arranged into visual patterns, facilitating the estimation and counting of quantities of items at a glance.
Tokens were also used for the reliable storage of data. For example, sets of counters could serve as permanent records for transactions to be completed in the future. This seems to be the case for groups of tokens found enclosed in hollow globular cases covered with seal impressions which, most likely, represented formal agreements—IOUS—kept in temple archives of the fourth millennium B.C. Interestingly, the envelopes predominantly held plain tokens and rarely complex types.
On the other hand, 16 percent of the Uruk counters and 55 percent of those from Susa were perforated, suggesting that some of the complex tokens were strung on a fine tie. This finding, in turn, spread light on the use of the oblong bullae, solid oblong blobs of clay impressed with seals. It is likely that the oblong bullae, which at both ends reveal the imprint of a string around which they were originally wrapped, served to secure the knot and loose ends of the strings holding tokens. More important, they identified the accounts in question by displaying the pertinent seals.
The two devices used to group tokens, envelopes and oblong bullae, were both made of clay and bore seal impressions—in some cases from the same seal. That they held different kinds of tokens, in different ways, suggests that they fulfilled similar, but not identical, functions. There are also multiple examples of perforated plain tokens, and like complex tokens, these too were occasionally kept in envelopes, thus showing a crossover between the storage devices.
The key to cracking the code of the tokens is provided by the Sumerian script which derived from them. It seems that plain and complex tokens referred to different types of goods: the first stood for products of the country, whereas the second represented goods manufactured in urban temple centers.
Bodies of symbols systematized by communication devices exhibit a unique capacity to survive for millennia. Writing systems have the flexibility, for example, of adjusting to new technologies and new cultural needs by modifying the form of signs without altering the meaning. For instance, the letters of our Latin script have retained, for the most part, the value they had in the former Greek and Phoenician alphabets of 2,500 and 3,500 years ago. The Egyptian and Chinese writing systems are other notorious examples of the persistence of symbols through the ages. Some Egyptian signs can be followed, carrying the same meaning, from pre-dynastic three-dimensional amulets to hieroglyphs carved in stone and the cursive hieratic and demotic scripts traced with a brush on papyrus. As I describe below, the same is true for cuneiform writing, which can be traced back in time from the first and second millennium Assyrian and Babylonian syllabaries to Sumerian ideographs of the third millennium B.C. and their token prototypes in the fourth millennium B.C. Only a few tokens can be decoded by following their evolution step by step to the well-understood cuneiform characters of the second millennium B.C. The few scores of examples which can be satisfactorily identified provide an insight, however, into yet another major difference between plain and complex tokens, namely, the kinds of goods represented by each.
The cuneiform and ideographic signs of the second and third millennium B.C. expressing cereals and domesticated animals—the two traditional staples of the ancient Middle East—point to plain tokens as their progenitors. Namely, cones and spheres referred to the two most common Sumerian grain measures, the ban and the bariga, approximately equivalent to our liter and bushel. Large cones, large spheres, and flat disks were larger capacity units of grain. Furthermore, cylinders and lenticular disks were used as units of animal count with the cylinder standing for one and the lenticular disk for a collection of animals—a flock, perhaps ten.14
Complex tokens also find counterparts among Sumerian pictographs indicating animals, but in this case, they include precise reference on sex and age of the creatures. Such disks with various patterns stand, for example, for "one male sheep," "one ewe," "one lamb."15
More typically, however, complex tokens can be matched with signs representing finished products.16 For example, incised cones, ovoids, and rhomboids represented processed foods such as bread, oil, and beer. Biconoids and triangles indicate luxury goods such as perfume and metal. One series of tokens seems particularly significant because the tokens refer to items of the textile industry, important in the early Mesopotamian temple economy. Among them are disks and parabolas with linear markings which signify types of fibers, cloths, and garments; incised cylinders and rectangles stood for strings and mats. Finally, examples of naturalistic tokens clearly represented processed foods, such as trussed ducks, as well as manufactured products, such as tools, weapons, pieces of furniture, and a variety of vessels.
It becomes apparent, therefore, that the token system was used solely for keeping records of commodities. Each category of tokens referred, however, to fundamentally different kinds of goods: plain tokens counted the basic products of the farm, quantities of grain and animals in particular; complex forms were used mostly to calculate goods produced by workshops. This different usage explains the duality of the token system, elucidating, especially, the discrepancy in chronology and geographic distribution. It becomes obvious why the first assemblages of plain tokens coincide with the beginning of farming whereas complex tokens start occurring in the urban period. The different usage also explains why plain tokens were ubiquitous and complex tokens occurred only in selected areas: staples are consumed or accumulated in every possible setting, but industry flourishes in particular circumstances. In the latter case, the typical assemblages of clay cone mosaics, seals, and vessels present in sites yielding complex tokens, envelopes, and oblong bullae make it obvious that the workshops where accounting took place were developed under the auspices of the Sumerian temple. The distribution of complex tokens far afield in Susiana and Syria shows centers of activity of the Sumerian temple outside Sumer. It can be postulated that the system of record keeping with complex and plain tokens, envelopes, oblong bullae, and seals was related to a coercive redistributive economy, as it was in the homeland of Sumer. In this case, however, the amounts of goods in kind required from the foreign centers should be viewed as tribute.
Finally, the different kinds of goods dealt with by the two kinds of tokens explains their different storage in temple archives. In fact they were handled by different hands in different services: the one group belonged to the pens and granaries, whereas the other served the superintendents of workshops.
7. The Introduction of Two Kinds of Signs
The duality of the token system was perpetuated in writing when each kind of counter gave way to different types of signs in the Sumerian script: plain tokens were replaced by impressed markings, whereas complex tokens were reproduced by pictographic signs incised with a stylus.
As I have explained elsewhere, plain tokens can be traced through the various stages of their evolution to writing.17 In short, the metamorphosis was triggered by the fact that plain tokens, contained in envelopes, were hidden by the thick clay walls of the cases. This led clerks to mark the surface of the envelopes by impressing each token prior to enclosing them inside, thus making visible at all times the number and shapes of the tokens included. The marked envelopes, in turn, promptly led to a further improvement, namely, solid clay tablets in the form of a small cushion, bearing token impressions. In other words, plain tokens were replaced by markings consisting of their negative imprint on a clay tablet (fig. 6).
The repertory of impressed markings on envelopes, as well as those on tablets is limited to a dozen signs, eight of them deriving from plain tokens.18 They include circular markings of various diameter and depth, standing for the former small and large spheres, the flat and lenticular disks. They also include wedges of different lengths and widths corresponding to the former cylinders, cones and large cones.
The merging of complex tokens with writing was bound to be different, since they were generally not held in envelopes and, furthermore, the method of imprinting tokens was not well suited for rendering the outline of the complex token shapes and especially the linear and punched markings they bore. Complex tokens were therefore perpetuated on the tablets by signs written with a stylus.19 This technique is not surprising, since it made use of the same pointed stylus as was used to trace the markings on their face.
As is the case at each step of the development of the token system, there is some overlap between the methods of rendering the tokens in a graphic form. Tetrahedrons, one of the most usual plain tokens, led, seemingly, to an incised pictograph.20 On the other hand, there are examples of incised ovoids being impressed on at least two envelopes.21 Finally there is a unique tablet displaying a composite impressed-incised technique with incised markings applied over the impressions of spheres and triangles.22
The new system of notations was infinitely more practical, since the clay tablets displaying neatly aligned signs were far less cumbersome than the loose tokens. It was also more expedient, since impressing or tracing markings was quicker than modeling each individual token. The new formula was so satisfactory that tablets remained in use during the next three millennia in the ancient Middle East, to be displaced only when the Aramaic script, written with a flowing hand on papyrus, provided a yet more efficient system of handling information.
8. The invention of numerals
Differences far greater than technique existed between the two categories of signs, however. The most significant distinction between the impressed and incised signs lay in the way they expressed plurality, which led to the impressed sign expressing numbers and the incised sign indicating the nature of the items counted.
Like the former tokens, the impressed signs continued to show the number of items counted by repeating the marking in one-to-one correspondence: one, two, or three small measures of grain were indicated by one, two, or three small wedges, and one, two, or three bushels of grain were indicated by one, two, or three circular markings. The same is true for the impressed markings indicating units of animal counts.
On the other hand, incised signs are never found repeated in one-to-one correspondence. Thirty-three jars of oil, for example, were no longer expressed by repeating the sign for a jar of oil thirty-three times. Instead, the pictograph for a jar of oil was preceded by numerals—special signs expressing a number (fig. 7).
In fact, the new signs to express abstract numbers were nothing else than the impressed signs for measures of grain used in a novel fashion. An impressed wedge and an impressed circular sign standing for a small and a large measure of grain came also to signify "one" and "ten." This seems confusing to us but it did not seem to trouble the ancient clerks, who could decide from the context which reading was appropriate. In fact, the system of using identical signs for numbers and measures of grain was perpetuated throughout the Sumerian period without causing any confusion, apparently, among scribes.
Whereas plain and complex tokens differed in the products they represented, the one rural and the other urban, the impressed and incised signs differed in their usage to a considerably greater degree. The impressed signs conferred a notion of quantity; the incised signs indicated the nature of the item counted. These concepts of quantity and quality, which were fused together in the token system, were abstracted from each other for the first time in writing. This is why an ovoid stood for "one jar of oil" whereas it took two signs to give the same information on the tablets—an impressed sign for "one" and an incised sign for "jar of oil."
The invention of zero and place notation has been heralded as a major accomplishment of the civilized world, but the literature does not treat the advent of abstract numerals because of the common but erroneous assumption that abstract numbers are intuitive to humans. The token system is one piece of artifactual evidence proving that counting, like anything else, is not spontaneous. Instead, counting is cultural and has to be learned. It seems logical to assume that a reckoning device must reflect the various modes of counting of the culture using it, and consequently, two steps of the evolution of counting can be identified in the ancient Middle East. The first step was taken about 8000 B.C., when tokens of various shapes were used to count different merchandise in a one-to-one correspondence. The second major step was the introduction of abstract numerals when impressed signs showing units of grain measures came to indicate, alternatively, abstract numbers.23 The extraordinary invention of abstract numerals amounted to a revolution in accounting and communication, since it provided, for the first time, a reckoning system applicable to any and every item under the sun. Each numeral stood for the concept of oneness, twoness, threeness, and so on, abstracted from the item counted. This put an end to the cumbersome system necessitating particular symbols for counting different goods. From there on, jars of oil as well as measures of grain or sheep of a flock could be counted with the same symbols. The system did not fully depart at once from the traditional one-to-one correspondence since 1, 2, 3, and so on were expressed by one, two, three wedges and 10, 20, 30, by one, two, three circular signs. Abstract numerals brought about, however, a tremendous economy of notations by replacing ten wedges by one sign for the number ten. As a result, for example, ten jars of oil could be shown by two signs only: "ten" and "jar of oil."
The system of plain tokens which originated at the beginning of agriculture in the ancient Middle East was supplemented by complex tokens at the rise of the Sumerian temple. Plain and complex tokens were counters of the same reckoning device, but each served one branch of the Sumerian economy: plain tokens referred to products of the farm, whereas complex tokens stood for goods manufactured in workshops. For this reason, the two kinds of tokens belonged to different services of the temple administration, where they were kept in a different manner. Plain tokens were stored in globular hollow clay envelopes, while complex tokens were strung together on a tie held by a solid bulla. This, in turn, had major consequences for the origins of Sumerian script. The plain tokens were replaced by impressed markings, but the complex counters gave rise to incised pictographs. The duality of the token system was thus carried over in writing where the split between the two kinds of symbols grew ever wider. The impressed signs evolved to express the quantities of items counted, whereas the incised pictographs indicated the nature of the items counted. The duality of our own writing system which uses numerals (ideographs) and letters (phonetic signs) was presaged in the first reckoning device using tokens. Plain tokens and impressed signs brought about the use of abstract numerals, whereas the complex tokens and, as M. W. Green demonstrates, in the following chapter on Sumerian cuneiform, incised pictographs slowly evolved to the acquisition of phonetic values.
10. Further Readings
- Amiet, Pierre. Elam, pp. 70-71. Auvers sur Oise: Archée Editeur, 1966.
- —. Glyptique Susienne: Mémoires de la délégation archéologique en Iran, no. 43, 1: 69-70. Paris: Librairie Orientaliste Paul Geuthner, 1972.
- Goody, Jack. The Domestication of the Savage Mind. Cambridge: Cambridge University Press, 1978.
- Schmandt-Besserat, Denise. "Before Numerals." Visible Language 18 (1984): 48-60.
- —. "The Decipherment of the Earliest Tablets." Science 211 (1981): 283-85.
- —. "The Earliest Precursor of Writing." Scientific American, (June 1978), 50-59.
- —. "The Envelopes That Bear the First Writing." Technology and Culture 21, no. 3 (1980): 357-85.
- —. "From Tokens to Tablets: A Re-Evaluation of the So-called Numerical Tablets." Visible Language 15 (1981): 321-44.
- —. "The Origins of Writing." Written Communication 3, no. 1 (January 1986): 31-45.
- —. "Tokens and Counting." Biblical Archaeologist (1983): 31-45.
1. Denise Schmandt-Besserat, "The Origins of Writing," Written Communication 3. no. 1 (January 1986): 31-45.
2. Denise Schmandt-Besserat, "An Archaic Recording System in the Uruk-Jemdet Nasr Period," American Journal of Archaeology 83 (1979): 19-48.
3. Julius Jordan, Vorläufiger Bericht über die von der Deutschen Forschungs-gemeinschaft in Uruk-Warka unternommenen Ausgrabungen, Abhandlungen der Preussischen Akademie der Wissenschaften 2 (Berlin, 1931), 47-48, fig. 41; Denise Schmandt-Besserat, "The Envelopes That Bear the First Writing," Technology and Culture 21, no. 3 (1980): 369, fig. 4, Sb 1938.
4. On Tell Aswad I, see Henri de Contenson, "Recherches sur le Néolithique de Syrie (1967-1976)" Comptes Rendus des Scéances de l'année 1978, Proceedings of the Académie des Inscriptions et Belles-Lettres (Paris, 1979), 821; Jacques Cauvin, Les premiers villages de Syrie-Palestine du IXéme au VIIéme millénaire avant J.C.: Collection de la Maison de l'Orient Méditerraneen Ancien 4, Série Archéologique 3 (Lyons: Maison de L'Orient, 1978), 74; Robert J. Braidwood, Bruce Howe, and Charles A. Reed, "The Iranian Prehistoric Project," Science 133 (1961): 2008; Phillip E. L. Smith, "Garij Dareh Tepe," Paleorient 1 (1974): 207-8.
5. Cauvin, Les premiers villages, 74, 43, 75; Olivier Aurenche et al., "Chronologie et organization de l'espace dans le Proche-Orient," in Préhistoire du Levant, Proceedings of the Colloque CNRS, no. 598 (Lyons: CNRS, 1980), 7-8.
6. Denise Schmandt-Besserat, "The Emergence of Recording," American Anthropologist 84, no. 4. (1982): 871-78.
7. Julius Jordan, Vorläufiger Bericht über die von der Deutschen Forschungs-gemeinschaft in Uruk-Warka unternommenen Ausgrabungen, Abhandlungen der Preussischen Akademie der Wissenschaften 3 (Berlin, 1932), 19. Denise Schmandt-Besserat "Tokens at Uruk," Baghdader Mitteilungen 19, (1988): 1-175.
8. Thomas W. Beale, "Bevelled Rim Bowls and Their Implications for Change and Economic Organization in the Later Fourth Millennium B.C.," Journal of Near Eastern Studies 37 (October 1978): 311-12.
9. Mark A. Brandes, Siegelabrollungen aus den Archaischen Bauschichten in Uruk-Warka, Freiburger Altorientalische Studien 3 (Wiesbaden: Frank Steiner, 1979), 17-166.
10. Schmandt-Besserat, "Emergence of Recording," 872.
11. Vivian Broman-Morales, "Jarmo Figurines and other Clay Objects," in Prehistoric Archeology along the Zagros Flanks, ed. Linda S. Braidwood et al., Oriental Institute Publications 105, (Chicago: University of Chicago Oriental Institute, 1983), 369-426.
12. Denise Schmandt-Besserat, "Tokens at Susa," Oriens Antiquus 25, no. 1-2 (1986).
13. Pierre Amiet, "Alternance et Dualité: Essai d'interpretation de l'histoire élamite,"Akkadica 15 (1979): 6; Eva Strommenger, "Ausgrabungen der Deutschen Orient-Gesellschaft in Habuba Kabira," in Archaeological Reports from Tabqa Dam Project-Euphrates Valley, Syria, ed. David Noel Freedman, Annual of the American School of Oriental Research (Cambridge: American School of Oriental Research, 1979), 79.
14. Schmandt-Besserat, "Envelopes," 370-75.
15. Schmandt-Besserat, "An Archaic Recording System," 42; Schmandt-Besserat, "Envelopes," 374-75.
16. Schmandt-Besserat, "An Archaic Recording System," 41-48.
17. Schmandt-Besserat, "Envelopes," 382-85.
18. Denise Schmandt-Besserat, "From Tokens to Tablets: A Re-Evaluation of the So-called Numerical Tablets," Visible Language 15 (1981): 331-33.
19. Schmandt-Besserat, "An Archaic Recording System," 41-48.
20. Schmandt-Besserat, "Envelopes," 375.
21. The two artifacts were excavated at Habuba Kabira. Eva Strommenger, "Ausgrabungen in Habuba Kabira und Mumbaqat," Archiv für Orientforschung 24 (1973): 170-71.
22. Schmandt-Besserat, "From Tokens to Tablets," 328, fig. 4b.
23. Denise Schmandt-Besserat, "Before Numerals," Visible Language 17, (1984): 55-58.
12. See also
- Schmandt-Besserat, Denise. «The Earliest Precursor of Writing», in Scientific American, June 1977, Vol. 238, No. 6, p. 50-58.