Writing is a mechanism that permits us to change the format of our creative endeavors, the shape of our knowledge, our understanding of the world, and our activities within it.
In this volume I examine the interface between writing and art during the early urban period in the ancient Near East. I propose that an exchange between the two media took place in two installments: in the late fourth millennium B.C., writing had an impact of great consequence on art, and, reciprocally, in the third millennium B.C., art had an impact of no less significance on writing. I will show that through this mutual exchange both writing and art multiplied their capacity to communicate information. Art became narrative and writing went beyond accounting to become a comprehensive medium of communication.
Part One is devoted to the impact of writing on art. I hold writing responsible for the fundamental changes that took place ca. 3500 B.C. in Near Eastern art composition—the way designs were organized. To make this case I will compare and contrast compositions of preliterate vs. literate pottery paintings, wall paintings, seals, and stone reliefs. I will argue that before the invention of writing, compositions typically consisted of geometric or animal motifs that were juxtaposed, dovetailed together, or placed in rotating arrangements that symbolized an idea or evoked a story. But after writing, as the process of copying the dispositions of signs on a tablet was introduced, Near Eastern art became linear and thus could tell a story. Parallel lines, used as an organizing principle in a scene, caused the figures within that scene to be arranged in the same upright position upon a ground line. The ground line meant that the individuals pictured shared the same space at the same time. Accordingly, their relative size, location, position, order, and direction could be used to signify hierarchy, rank, intention, action, and interaction. For example, gods were taller than kings, and kings were taller than their fellow citizens. Figures facing one another were interacting, and those facing in the same direction were proceeding toward a common destination. Because the figures in a picture, like the signs of a text, were "read" in succession according to their disposition, the literate compositions were no longer apprehended globally but analytically. Images thus became an established pictorial language with its own conventions, rules, and vocabulary that could tell complex stories involving multiple characters.
Part Two addresses the impact of art on writing. I will show that ca. 2700-2600 B.C. written characters, deserting the mundane clay tablets, were inscribed on lavish funerary artifacts such as gold vessels and lapis lazuli seals. I will analyze how the funerary inscriptions borrowed the principle of writing personal names phonetically from previous economic and lexical lists. But they soon took on a life of their own, becoming increasingly phonetic and acquiring syntactical structures. The earliest funerary inscriptions, consisting of a name—"Meskalamdug," for example—constituted the first non-economic or lexical texts. Their intent was to transcribe the sounds of an individual's name into script to satisfy the Sumerian belief that pronouncing the name of the dead secured eternal life in the underworld. Later statues were inscribed with longer texts that included the deceased's name, title, patronymics, the name of the god or the temple to which the statue was dedicated, and a plea for a "long" life. Because, following Ruth Mayer-Opificius, I interpret the statues as serving a funerary function, I propose that this statement refers to a long life in the hereafter. The funerary texts that transcribed several names were primarily phonetic and, more importantly, they introduced syntax to build sentences that contained subjects, verbs, and complements. I will demonstrate that the funerary inscriptions bridged two phases of Mesopotamian writing, namely the stage when it was used exclusively as an accounting device, and that when it became a means of broad and proficient communication. I contend that by emulating the sound and structure of spoken language, the funerary inscriptions prompted the takeoff of writing. I conclude with an analysis of the second millennium B.C. stele of Hammurabi, which I will present as the ultimate outcome or epitome of the interface between writing and art.
As an introduction, I will define the use of the word "art" and review the background of the interface of writing and art, presenting the early phases of writing—impressed and pictographic writing—during the short but critical period when the first interface transformed art, ca. 3500-3300 B.C. I will also place my study in context.
Art has many definitions, none of which apply universally. In this book I label as "art" a wide range of artifacts—pottery and wall paintings, seals, stone reliefs, precious metal vessels, and statues in the round—that had a domestic, administrative, ritual, funerary, votive, or historical function in the ancient Near East. The aesthetic quality and conscious artistry shared by the objects justify the designation from our own society's point of view.
Art was age-old when writing began. In Anatolia, images of a bison and a deer engraved at the entrance of the cave of Beldibi and carved stones and antlers from Karain attest that in 15,000 B.C. Paleolithic humans produced meaning with pictures. In the following Neolithic period, Near Eastern farmers of the eighth and seventh millennium B.C., from Ain Ghazal in Jordan to Çatal Hüyük in Turkey, decorated either house walls or potteries with paintings, carved seals, statues, and reliefs in stone, plaster, and clay. When writing came about some three millennia later, ca. 3500-3300 B.C., it was a latecomer.
The interface between writing and art had a prehistory. Starting at about 7500 B.C., accounting in the Neolithic farming communities was performed by means of clay tokens of various forms (fig. 0.1). Each token shape was a symbol for one particular unit of merchandise. For example, a cone, a sphere, and a disk represented different measures of grain; a cylinder stood for an animal of the flock; and a tetrahedron was a unit of labor. The clay counters were teamed with seals immediately after these were invented, some 1,000 years after the tokens, around 6500 B.C. From then on, the two types of artifacts formed an administrative tandem: tokens recorded goods, and seals identified the people responsible for these goods, such as temple officials, donors, recipients, or debtors. Together, tokens and seals allowed the early farmers to manipulate and store information on an unlimited number of goods and individuals, making it possible to keep track of present, past, or future transactions such as debts or pledges. Over their three millennia of close association, between 6500-3500 B.C., each medium preserved its symbolic integrity. Seals displayed a great variety of geometric, animal, and, more rarely, human designs to symbolize offices and individuals, and tokens took the shape of three-dimensional geometric and occasionally naturalistic symbols representing units of goods. No token form derived from seals, and exceedingly few seals feature designs in the shape of tokens.
Tokens and seals merged physically on the round, hollow clay balls called envelopes that were devised to keep together for a length of time the tokens of particular accounts in an office archive. The envelopes held the tokens inside and displayed sealings—the impression of seals—on the outside (fig. 0.2). This interface between tokens and seals resulted in no less than the birth of writing. There can be no doubt that later, when accountants had the idea to imprint the tokens themselves on the surface of the envelopes in order to make visible the type and number of tokens held inside, they were copying the practice of pressing seals on the envelopes (fig. 0.3). These negative impressions of tokens on the surface of envelopes were the first signs of writing. The appearance of this revolutionary technique, which reduced tokens to two-dimensional signs, dates to about 3515-3375 B.C. It is the point of departure for the interface between writing and art discussed in this study.
The Impressed Texts
The texts responsible for writing's first interface with art consisted of markings impressed upon the clay envelopes holding tokens, described above; on oblong solid bullae securing strings of tokens; and on the earliest tablets, made of a solid clay lump (fig. 0.4). These are the so-called "impressed texts" that represent the third and fourth stages in the long process of the development of cuneiform writing in the Near East. The stages can be summarized as follows:
- Tokens (fig. 0.1)
- Plain envelopes holding tokens (fig. 0.2)6
- Impressed envelopes (fig. 0.3)
- Impressed tablets (fig. 0.4)
- Pictographic tablets (fig. 0.11)
- Cuneiform script
The corpus of impressed texts, which includes about two dozen envelopes and solid bullae and 250 tablets, forms a cohesive assemblage. The impressed tablets were modeled in clay in a quadrangular or oval shape about 5 cm long and 4 cm wide. Similarly, the globular envelopes and oblong solid bullae measured about 5 to 7 cm. The function of these first written documents was exclusively for accounting. Notations recorded quantities of goods, most often grain or animals, delivered to or supplied by temple or palace granaries and sheepfold as part of a strict system of control over entries or expenditures. And the texts impressed on envelopes and tablets were by no means a local or esoteric phenomenon. On the contrary, the practice extended from Syria to Mesopotamia and Western Iran over four centuries.
The technique that characterizes this earliest form of writing consisted of impressing the signs. At first, actual tokens were sunk into the clay of the envelope, bulla, or tablet, leaving images in negative. For example, a cone left a wedge-shaped indentation, and a sphere a circular depression (fig. 0.4). Evidence that the impressions were made with actual tokens is visible on several artifacts, one of which is an envelope from Habuba Kabira holding incised ovoids representing jars of oil that fit perfectly in the impressions on the envelope's exterior (fig. 0.5). A second is a Susa tablet with three large wedges that distinctly show the entire outline of the cone used to impress them (fig. 0.4), and a third is another Susa tablet showing impressions of triangle tokens complete with an incised line to depict the original incised counter (fig. 0.6). In other cases, circular signs and wedges were already impressed with a stylus, presaging the cuneiform script (fig. 0.7).
For the present study, the significance of the impressed texts is in their layout. It is quite remarkable that the 300 Syrian, Mesopotamian, and Elamite impressed texts share the same format, with exceedingly rare exceptions. This demonstrates that between 3500-3100 B.C., scribes of the Uruk temple administration, though located in distant places, followed identical rules prescribing the direction of script and where and how signs were to be organized on a tablet. Generally, the impressed texts abided by the following set of conventions.
1. Each marking represented a unit of goods.
Among the most frequent markings (fig. 0.8):
- A wedge represented one small unit of grain.
- A circular marking represented one medium unit of grain.
- A long wedge represented one domesticated animal.
2. Size was semantic.
Wedges existed in two sizes, small and large (figs. 0.4, 0.7, and 0.8):
- A small wedge corresponded to a ban, a small measure of grain (equivalent to a liter).
- A large wedge stood for a very large measure of grain.
Circular markings also existed in two sizes, small and large:
- A small circular marking represented a bariga, a medium unit of grain (equivalent to a bushel).
- A large circular marking stood for a large measure of grain.
3. The number of units of goods was shown in one-to-one correspondence (figs. 0.3 and 0.7).
- One small wedge = 1 small unit of grain.
- Two small wedges = 2 small units of grain.
4. Writing was linear (figs. 0.3 and 0.7).
The impressed signs were inscribed in straight lines (figs. 0.4, 0.6, and 0.7). Signs of different types were usually not mixed on the same line. Instead, each line featured only one kind of sign, repeated as many times as needed. For example, on fig. 0.7, a line of four circular signs is followed by a line of four wedges.
Since the orientation of the signs on the impressed tablets is disputed, I take this opportunity to explain my views. I consider symmetry to be a fundamental aesthetic characteristic of Mesopotamian cultures. In other words, as illustrated by the popular master of animals or mirror effect compositions in Mesopotamian art, a feature to the right calls for a similar feature to the left; when a composition consists of an uneven number of figures, one is placed in the center and the others are distributed evenly on either side; and when the composition holds a single figure, it occupies the center of the field. My study of tokens showed that symmetry also governed the disposition of markings on tokens—lines and punctuations were consistently organized in symmetrical sets on either side of a central feature. For example, disks displayed a set of three lines on either side of a median line or a blank space. I propose that the scribes had the same propensity for symmetry when establishing the format of impressed tablets. Accordingly, the scribes arranged the signs in parallel, horizontal lines centered on the middle of the long side of a tablet. This is supported by the fact that when a line features a single sign, it is usually placed in the center of the long side of the tablet (fig. 0.9). When there is an uneven number of signs, one occupies the center, and the remaining signs are divided in equal number on either side; when there is an even number, the signs are displayed symmetrically across the center (fig. 0.7).
I should add that cases—lines separating signs referring to two or more transactions—do not occur on impressed tablets.
5. Location (above/below) on the tablet was semantic.
The lines of signs were by no means placed at random, but were arranged in strict hierarchical order. The signs representing the largest units of merchandise were placed on the uppermost line, followed by lines of lesser and lesser units. For instance, in fig. 0.7, circular signs are placed above the wedges, which represent smaller units. The standardized logical structure ensured easy, efficient reading. The tablet in fig. 0.7 could be read at a glance: "Four bariga and four ban of cereals."
Signs of the same shape but different size, such as small and large wedges, were easily distinguishable because the large wedges, standing for large measures of grain, were placed in the upper part of the tablet (fig. 0.4), while the small wedges, representing small units of grain, were set along the lower edge (fig. 0.7). These signs were placed above and below the circular signs, which represented intermediary units of capacity of grain.
6. Order (right/left) on a line was semantic.
The layout of the impressed tablets again reflects the Mesopotamian love for symmetry and order in that, as mentioned above, when there was a single sign on a line, it was placed in the center of the tablet rather than on the right or left side (fig. 0.9). When, exceptionally, a line included different signs (fig. 0.4, second line), the larger unit was shown on the right and those of lesser value on the left.
7. The direction of the script was boustrophedon.
Notations were inscribed starting from the right and proceeding toward the left. But when all the signs did not fit on one line, the next line continued in the opposite direction, from left to right—in other words, boustrophedon. Both the right-to-left initial direction and the following boustrophedon reversal are well illustrated in several tablets, such as in fig. 0.10.
In sum, when impressed envelopes and tablets began to be used, their efficient format was already well established and perhaps even age-old. The concepts may well have had their roots in prehistory and been elaborated over centuries by generations of village chiefs who arranged tokens in neat lines according to types of merchandise and unit sizes in order to visualize the resources available for the next festival. The custom of organizing tokens according to goods and unit sizes was then transposed onto the markings on envelopes and, subsequently, tablets. Finally, as I will demonstrate, the efficient system of organizing economic notations was adapted to art to create complex visual narratives.
The important impressed texts that reveal the very beginning of writing and, as I will show, play a significant role in the history of the art of the Near East have been mostly ignored by archaeologists and philologists alike. This is especially true for the impressed tablets, although they have received attention in excavation reports. In particular, the impressed texts from Susa, Godin Tepe, and Jebel Aruda have been the subject of special studies, and my two volumes Before Writing (1992) and How Writing Came About (1996) gave impressed texts a full treatment. Nevertheless, though the secondary literature shows a fascination for token envelopes and the later pictographic tablets, it typically skips the impressed tablets. The fact that the impressed tablets are continuously referred to as "numerical tablets," as they were erroneously dubbed when first excavated at Uruk in 1930-1931, is the best evidence for their endemic misinterpretation. I recently heard a noted Assyriologist, lecturing on the origin of writing, commiserating that "the 'numerical tablets' must have been confusing for the Uruk accountants, since they recorded numbers only." The statement was wrong; it deceived the audience on two counts. First and foremost, at the time of the impressed tablets, abstract numbers had not yet been invented. Instead, according to the system of concrete counting, each impressed marking expressed one unit of merchandise. Repeating the sign in a one-to-one correspondence indicated the number of units. The lecturer's object of commiseration was, in fact, a standard account of grain stipulating different measures of barley in various numbers—not a series of nonsensical numbers. Second, the lecturer misrepresented the Uruk accountants, who as early as the mid-fourth millennium B.C. kept precise records of several hundred types of goods with complex tokens. The impressed tablets, far from being a step backward, represented another great innovation by replacing loose tokens with permanent impressions bundled in clear lines. These lines became the fundamental structure of an efficient visual language; they made it possible to bestow meaning to the location, order, and direction of signs on the face of a tablet.
About 3100 B.C., the impressed technique was superseded by a new method of writing involving a reed stylus that ended in a prismatic tip. The so-called "pictographic" economic texts transcribed the more complex tokens via traces of their outlines and markings on the clay tablets. These signs—traced rather than impressed—are referred to as "incised signs." Incised signs were used to depict the tokens for oil, metal, wool, and various textiles, for example, while the signs corresponding to the most frequently transacted goods (rations, grain, animals, and area measures, among others) continued to be impressed. The predominant method of writing began to change from impressed to incised, but the layout continued to be linear. The size of signs still indicated different units of the same commodity, and the location of signs on the tablets remained semantic. In particular, when the state administration required entering the name of the donors or recipients of the goods, donors were featured on the right side of the tablet and the recipients on the left (fig. 0.11). These positions communicated the verbs "given" or "received" (or the prepositions "to" and "from") before verbs and grammatical forms were expressed by the script.
I will show that the pictographic script created new types of signs that served as models to communicate information in narrative art. Among them were determinatives: signs modifying meaning. For example, one of the earliest determinatives was the star-shaped sign ], which when placed next to a name indicated that the name referred to a deity (fig. 0.11). New logograms no longer based on tokens consisted of pictures that evoked a concept. "Female slave," for example, was depicted by a pubic triangle and a scale pattern representing mountains—literally, "woman from the mountains."
In the following four chapters I will show that these conventions—namely linearity, one-to-one correspondence, and the relative size, location, and order of the signs—as well as determinatives and logograms were put to work for art, thus transforming the preliterate evocative symbolic compositions into a pictorial language able to convey a story.
Inscribed art monuments have received much attention in ancient Near Eastern scholarship by both philologists and art historians. Traditionally, the former considered only the text and the latter only the images, but this has changed. Recently, philologists such as Black and Green, Ellis, Finkel and Geller, Hallo, Lambert, Postgate, and Wiggermann have successfully identified the images of legendary heroes and monsters represented in art. Among art historians, Azarpay first recognized the value of inscribed monuments in creating chronological frameworks for the study of style. Bahrani, Barrelet, Suter, and Winter have done remarkable work linking Mesopotamian artworks with inventories and the rhetoric of third- to first-millennium texts.
When Writing Met Art differs from all of the above, as it examines earlier stages of writing and focuses on the interdependence and reciprocal exchange between the two early visual media of communication. This study builds upon my previous work on the origins of writing, elucidating how writing finally shook off the yoke of accounting inherited from the token system and showing how the structures of meaning built to manipulate the token system made an indelible mark in human cognitive development.