Use of Seals

The earliest use of cylinder seals dates to the mid-fourth millennium, the beginning of the Uruk period, when seals were first rolled on hollow clay balls inside which numerical tokens were kept. Tokens representing the number, and possibly the type, of goods being counted were thus protected within the clay, which was then sealed all over its surface with a cylinder seal (Nissen et al 1993: 12-13). After sealing, the balls were sometimes impressed with signs on the outer surface referencing the tokens within. This means of storing and protecting information soon gave way to the earliest tablets, all of them numerical in nature. These tablets were sealed before being inscribed, a practice not so different from that of impressing signs over sealed clay bullae. At the same time, seals were being used to mark the clay stoppers that covered storage containers, or the bulla that hung from strings attached to jars, baskets, bundles, or even store-room doors (Collon 2005: 13).

In the Uruk period, it seems clear that sealing was essentially an institutional practice. Individuals sealed on behalf of the administration. Even if one particular official could be chased down through his seal, it seems clear he owned and used this seal in his capacity as a representative of a certain office, not as a private individual.

It is uncertain when seals began to represent the individual as individual, not merely in terms of his relation to an institution. The first inscriptions of personal names (along with information about parentage and/or profession) appear on cylinder seals in the Early Dynastic period, during the Fara period (Nissen 1993: 17). This might imply a more individualised sense of the relationship of owner to seal. But seals in the Early Dynastic period seem also to have been heavily linked to institutions, the temple and, from ED IIIa onwards, the palace. This is supported by Rathje’s analysis of the seals found in the royal cemetery at Ur. He was interested in the possible relationship between iconography and the owner’s role or profession, and his evidence suggests that there was some such correlation (1977). Thus, at least for the individuals buried in the royal cemetery there, it seems that seals and office were linked. This might have been by choice, but it is also likely that there was some official sense of what seal was suitable for what office. Additionally, in the Sargonic Period we find evidence that the so-called arad2-zu seals (‘your servant’) mighthave been given out by kings, perhaps as rewards. Zettler hypothesises that ‘such seals are seals of office’, whose flow was restricted by the king (1979: 33).
Personal and purely institutional seals existed alongside one another in all but the very earliest periods. In the Ur III period, for example, we find suggestions that official seals could be rolled in an official’s absence (that is, that it was the mark of his office, not his personal involvement that mattered), but at the same time we find a seal belonging to a slave (NATN 679). The latter was certainly an individual with no authority stemming from an office or an institution who nonetheless had his own personal seal (Steinkeller 1979: 44-45).
Although we have seen that seals were mbeing used on tablets already in the Uruk period, after the end of this period seals disappear from tablets except in very rare instances. Fewer than ten sealed tablets dating from the Sargonic period have been identified. These are all administrative in nature and come from a scattered geographical area. Two sealed but blank tablets were found at Tello (AOT b 404, 406), and another one at Umm el-Jerab (Ashm 1932.344). Zettler suggest that these may have been trial rollings or may have been carried by an agent ‘to prove the authority of a verbal message’ (1979: 37). Sealed but uninscribed pieces of clay like this are found in small numbers in all periods and their function in not known (Collon 2005: 119).

Seals were still used extensively in the Early Dynastic and Akkadian periods, on jar sealings or on bullae. These bullae were simply flat lumps of clay which apparently hung loose around the necks of jars or baskets (see for example VA 6298, Ashm 1939.332). We frequently find impressions of string, jar lips, or basket edges on these bullae. On rare occasions a brief cuneiform inscription may be found beside or over the sealing, giving details of ownership or contents (for example, VAT 7187). Because of the interests of museum curators and cuneiformists, sealings on tablets and envelopes tend to occupy the greatest part of our attention. However, as we continue examining sealing practices, the continued use of seal on containers, bullae, or even buildings (Larsen 1977: 94-95) should not be overlooked. From the Ur III period onwards, the full possibilities of sealing begin to be realised. No longer is sealing used only for economic and administrative matters. A seal is now a guarantee in any sense: economic, yes, but also legal or personal. Seals reappear on tablets, but more commonly on envelopes around a tablet (Steinkeller 1979: 45). In legal documents, an envelope generally contained a summary of the tablet (sometimes repeating the entire contents) and an indication of persons involved together with seal impressions (Renger 1977: 75). The envelope fell out of favour for the most part by the first millennium. Instead, legal and administrative texts were often written in duplicate (Greengus 1995: 475).

Although sealing a tablet before use was the norm in the earliest periods, by the Ur III period tablets were sealed after being written. This is perhaps a reflection of the changing and expanding nature of sealing practices. To make a modern comparison, we might think on the one hand of letterheaded stationary and on the other of an individual’s signature. The former can be used by anyone legitimately associated with an institution and represents an authority stemming from the institution. Only an institution that feels legitimately confident in its ability to prevent unauthorised use of its authority and to make good on any claims legitimately made in its name can confidently produce such a thing. The practice of sealing before inscribing rather than after can thus be seen as an illustration of the characteristic strong temple institutions of the Uruk period, a world in which essentially only temple administration was involved in the sealing business, and only in limited areas. In the more varied landscape of the Ur III period and onwards, this was no longer the case.

There is evidence that the presence of a seal on a tablet could be invoked later on with genuine force (Renger 1977: 76, 79). Whether that force was always legal, or merely social is not entirely clear (i.e., whether a sealed promise was legally enforceable because of its sealing). A sealing was designed to hold the sealer to account, as is indicated by the fact that the party required to seal seems generally to have been the one with the most reason to protest later on—the recipient of goods in economic texts, the seller of items in sale documents (Steinkeller 1977: 45; Renger 1977: 76).