Proceed with Caution! Identifying Structures using Royal Inscriptions (Part 2)

I ended my last post with the assertion that inscribed objects are found in the ‘wrong’ building – that is, buildings other than those commemorated in their inscriptions – with some frequency. In a most un-scholarly manner, I provided precisely no evidence for that assertion…so, in this post, I will provide an outline of my methodology, and evidence to support my assertion. For those readers who don’t feel like re-reading that post, let me refresh your memory as to the topic of conversation:

Using a relational database, I have linked inscribed objects, the text of their inscriptions, and the locations in which these objects were found. This endeavor has allowed me to draw connections between characteristics such as object type, inscriptional elements, and place of deposition. The objects in question are those inscribed with royal inscriptions of Mesopotamian kings, dating from the Early Dynastic to Old Babylonian periods (c. 1750-1595 B.C.E.). The database has also allowed me to challenge the assumption that one can identify the archaeological remains of a building based on the inscriptions found within it. As stated above, many of these inscribed objects are found in buildings that are not commemorated in the inscriptions they bear, suggesting that using them to identify archaeological remains is problematic.

For practical reasons that will be explained below, this study is restricted to the southern Mesopotamian city of Ur. Ur is located approximately 220 miles south of Baghdad and 10 miles west of the Euphrates river. It has a long occupation history, from c. 6000 BCE to 539 BCE (Crawford 2015: 135). The first excavations at Ur were carried out in 1854, and then again from 1918-19. The most substantial excavation was an expedition led by Sir Leonard Woolley, which commenced in 1922 and carried out fieldwork for 12 seasons. Due to the less-than-desirable excavation practices of the 19thand early 20thcenturies, it can be difficult to determine with any certainty exactly where many inscribed objects were found. However, excavations at Ur were unusually rigorous, providing a large amount of information relating to the find spots of inscribed objects. This information has been extensively digitized via the Ur Online Project, making it easily-accessible for both scholar and layperson.

Using the Ur Online project, I determined find spots for 537 objects inscribed with building inscriptions that were excavated at Ur. The find spot information varies from a specific location in a room, to a structure, trench, or a general area of the site. Due to this variation, I use the phrase ‘found in association’ to indicate that an object was found within or close to a particular structure or part of the site. This phrase encompasses objects that were discarded in refuse pits, used as fill for building projects, as well as objects that were found ‘in use’, by which I mean that they were found in what appears to be their primary place of deposition – for example, inscribed bricks located in walls.

While more detailed data is preferred, objects can be identified as being in the ‘wrong place’ when they were not found in the structure named in their inscription, even if no additional information is available. For example, an object found in a residential or mortuary area but commemorating a religious structure was clearly found outside the commemorated structure.  Such is the case with 39 objects that bear an inscription of Rim-Sin I (1822-1763 BCE) commemorating construction of the temple ‘Ešutumku’. While the physical temple remains unidentified, as the inscribed objects were found in places such as a ziggurat terrace, the royal cemetery, and a residential area, they were discovered outside their commemorated structure.

Find spots themselves are tagged with the following descriptors:

  1. In use – object found in its primary place of deposition; e.g., a foundation box or built into a wall.
  2. Reuse – object found in location of use but in later archaeological context.
  3. Filling – object used to level ground prior to the erection of a new structure.
  4. Rubbish – object that discarded in rubbish heaps or pits.
  5. Surface – object found on the surface of the site.
  6. Unknown – object whose exact position is unknown.



The 537 inscribed objects entered in my database are objects that were typically physically and permanently incorporated into the architecture of a structure, usually a temple. Of the 537 objects, 341 of them (64%), were definitively found in association with a structure not commemorated by their inscription. Accordingly, only 195 objects (39%) were potentially found in the commemorated structure, and this maximal number is probably too high. A breakdown of the number of objects found in each find spot type can be seen in table 1.

Find spot categorization Possibly commemorated structure Not commemorated structure
In-use 13 16
Reuse 0 4
Filling 5 72
Rubbish 0 34
Surface 0 8
Unknown 176 209

Table 1. Table showing the distribution of inscribed object by find spot category.


Clearly, using the presence of inscribed objects alone to identify a structure is problematic, and this result suggests the benefit of a more cautious. For instance, earlier, I mentioned one scholar’s argument that a single foundation tablet at Ur commissioned by the king Šulgi (2094-47 BCE) proves the existence of a cult and temple to the goddess Ninsar, otherwise unattested, at Ur.  My database demonstrates the real possibility of inscriptions being reused.  Since Šulgi is known to have renovated a temple to Ninsar at a different city, Nippur (Frayne 1997: 112), it is very possible that the tablet commemorates Šulgi’s construction of her temple at Nippur and was reused at Ur, instead of being the only evidence for Ninsar’s cult and temple at Ur.

Similarly, the existence of a temple at Ur called the Ebabbar could be postulated on the basis of two inscriptions with multiple exemplars commemorating its construction (Excavation numbers U 187, 2637, 15071, 16047, 1360, 13085, 13682, 2900, 7798, 16536, and 10136.). While Ebabbar temples are attested at several ancient cities, including Sippar, Larsa, Girsu, and Aššur, no other evidence for one at Ur exists (George 1993: 70-71). Moreover, no exemplar of these two inscriptions was found in use, and there is no obvious concentration that may suggest the location of a now-ruined temple (fig. 1). Again, this distribution could be the result of rulers’ reusing inscriptions that commemorate structures at other cities.


Figure 1. Locations of inscribed objects referencing the Ebabbar temple.

If, as highlighted by these two examples, my database demonstrates the danger of uncritically assuming that the presence of a building inscription identifies the structure in which it was found, royal inscriptions can still aid in the process of identifying structures, if used with caution. For instance, the database sometimes reveals a high concentration of a single inscription in a single location, as illustrated by the multiple exemplars of king Warad-Sin’s inscription commemorating the construction of Etemeniguru, a ziggurat terrace (RIM inscription number E4.2.13.16). 41 clay cones bearing this inscription have known find spots, and 23 of those were found in the excavated terrace of a ziggurat (fig. 2). As 56% of the exemplars were found in a structure of the same type described in the inscription, we may be cautiously optimistic in identifying the structure commemorated in the text with the structure that housed the majority of the exemplars. If, however, there were many fewer copies of this inscription or if a much smaller percentage had been found in a ziggurat terrace, then identification would remain tentative and require additional confirmation.


Figure 2. Locations of exemplars of Warad-Sin’s inscription commemorating the construction of the Etemeniguru.

Finally, the database indicates that some types of inscribed objects may be a more reliable means of identification than others. For example, canephors and foundation tablets were never found in structures definitively not commemorated, only in structures that may be commemorated, making them a good place to start the process of identifying a structure. Door-sockets, on the other hand, account for 12 of the 20 objects found in structures definitively not commemorated, suggesting that door-sockets are a less-reliable means of identifying structures.

This kind of information can and should be combined with other archaeological data – for example, if an archive relating to a particular building is located within a structure that also has a high concentration of inscriptions naming that building, then the identification can be that more secure. Conversely, if a structure houses archives related to a different building(s) than that named in associated inscriptions, then more consideration may be necessary before an identification is made.



These observations raise questions for modern researchers as to how we identify ancient structures. The presence of a foundation tablet in the fill of one structure was enough to lead one scholar to suggest the presence of a cult not otherwise attested at Ur. Based on the study presented here, it seems just as likely that the tablet originally commemorated the construction of a temple at another city and was reused at Ur. Likewise, the spatial distribution of 11 cones commemorating the construction of an Ebabbar temple by Sin-iddinam (1849-43 BCE) more likely derives from the cones’ reuse than from the existence of an otherwise-unattested Ebabbar at Ur. If it is possible for an object to be found in a structure other than the one its inscription commemorates, as my research clearly shows is the case, then we must proceed with caution when identifying ancient structures in the archaeological record. Of course, these results are based only on a small fraction of inscribed objects, and eventually I plan on expanding this study to include all objects for which there is a known find spot…though this may not increase the sample size substantially due to incomplete archaeological records.



Charpin, D. 1986. Le Clergé d’Ur au Siècle d’Hammurabi. Paris.

Clayden, T. 1995. ‘The Date of the Foundation Deposit in the Temple of Ningal at Ur’, in Iraq, vol. 57 pp. 61-70.

Crawford, F. 2015.Ur: City of the Moon God. London.

Fitzgerald, M. 2010. ‘Temple Building in the Old Babylonian Period’, in M. J. Boda and J. Novotny (eds.), From the Foundations to the Crenellations: Essays on Temple Building in the Ancient Near East and Hebrew Bible.Münster. 35-48.

Frayne, D. 1990. ‘Old Babylonian Period (2003-1595 BC)’, Royal Inscriptions of Mesopotamia Early Periods, vol. 4. Toronto.

Frayne, D. 1997. ‘The Ur III Period (2112-2004BC)’, Royal Inscriptions of Mesopotamia Early Periods, vol. 3/2. Toronto.

George, A. 1993. House Most High. The Temples of Ancient Mesopotamia. Winona Lake, IN.

Woolley, L. 1927. Ur Excavations VII: The Old Babylonian Period. London, and Philadelphia, PA.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s