In his book on the proper name LUM-ma (Marchesi 2006), Gianni Marchesi offers a bold interpretation of the often discussed passage (v 9-19) in the Boulder A inscription of Eanatum (RIME 1.09.03.05):
u₄-ba, e₂-an-na-tum₂-ma, e₂-an-na-tum₂, mu u₂-rum-˹ma˺-ni, mu GIR₃.GIR₃-ni, ˹lum˺-ma-a, ^dnin-˹gir₂˺-su-˹ra˺, i_x(A) gibil, mu-na-dun

On pp. 123-125, he summarizes in seven points the problems that cast doubt on the traditional interpretations of the passage in his view. This short note is concerned with his problems nos. 2, 3, and 5.

Marchesi's problem no. 2 relates to the interpretation of ll. v 10-12; about which he states:

“É-an-na-túm-ma É-an-na-túm mu ú-rum-ma-né, lit. ‘of E'annatum, E'annatum is his own name', represents a unique departure from the standard Sumerian syntactic construction of a nominal clause opening with an anticipatory genitive. If the commoninterpretation of the line were correct, we would expect it to be worded *É-an-na-túm¬ma mu ú-rum-ma-né É-an-na-túm (‘of E'annatum, his own name [that is, E'annatum's own name] is E'annatum'), instead.” (p. 123)

Following Gerd Steiner (1975: 15-17), Marchesi then assumes that the name of Eanatum is mistakenly placed in l. 11, and emends the text accordingly. Unfortunately, Marchesi does not quote any evidence on which his expectation concerning the structure of v 10-12 is based. As a matter of fact, there is overwhelming evidence that his expectation is unfounded. Consider, for example, ll. 12-14 of Ur-Ningirsu I 4 (RIME 3/1.01.01.04):
maš-da-ri-a-ba / nin-gu₁₀ he₂-ma-zi-zi / mu-bi
Of this votive gift, “May my lady raise him for me!” is its name.

This construction is used commonly for referring to name of entities in Sumerian royal inscriptions. Its structure consists of at least three parts in this order (it may or may not be followed by a copula depending on the syntactic context): (a) ENTITY=ak (b) “NAME” (c) mu=bi/ani [(d) =copula], and never in the order *ENTITY=ak mu=bi/ani “NAME” (as suggested by Marchesi in the passage quoted above). An example from an Ur III sale document (FAOS 17, 93: 2) is:
sag-ba ha-ba-lu₅-ke₄ mu-ni
(a) (b) (c) “the slave with the name of Habaluke”

An example from an Ur III royal inscription is Ur-Namma 28 (RIME 3/2.01.01.28) ii 10-13:
i₇-da, ^d˹nanna^˺-gu₂-gal mu-bi, i₇ ki-sur-ra-kam, mu-ba-al
(a) (b) (c) “He (= Ur-Namma) dug the canal with the name of Nanna-gugal, a border-canal.” (lit. “The canal with the name of Nanna-gugal is a border-canal, he dug it.”)

These constructions follow the pattern of structures that are used to identify an entity with a proper name, and in which the description of the entity consists of a genitive construction. The reason for the use of the anticipatory genitive appears to be that the topic (= the anticipated possessor) and the subject (= the proper name) of the copular clause is different. An example from an Ur III royal inscription is Šu-Suen 1 iv 44-46 (RIME 3/2.01.04.01):
iri^ki-ba, ^dšu-^dsuen, dingir-bi-im
(a) (b) (c)+(d) “Their town's god is Šu-Suen”

The passage in Ean. 5 v 10-12 is another example of this common construction: (a) Eanatum=ak (b) “Eanatum” (c) mu urum=ani. Its oddity is due to the fact that here the person whose name is specified with this formula is referred to in part (a) by the very name which is then mentioned in part (b). The reason for this is clear: v 10-12 contrast with v 13-14, the former gives Eanantum's usual name, while the later must give another name of his. It seems to me that the proper analysis of v 10-12 in fact supports that traditional understanding of the structure of Ean. 2 v 10-14 and makes Marchesi's questionable, whatever the correct reading and interpretation of the graphemes GIR₃.GIR₃ in v 13 may be.

Marchesi's problem no. 3 is concerned with the apparently chiastic structure of v 11-14. He asserts that “examples of this type of chiasm (syntactic chiasm) are very rare in Sumerian and are confined to literary texts. To find one example in an otherwise prosaic dedicatory inscription is wholly unexpected.” (p. 124). This is an argument that can be countered with an example from another dedicatory inscription, Lugalzagesi 1 (RIME 1.14.20.1), ii 21-22:
bara₂-bara₂ ki-en-gi, ensi₂ kur-kur-ra
“the sovereigns of Sumer, the rulers of all the foreign countries“

Here the reduplicated nouns are in a chiastic arrangement (cf. Wilcke 1990: 478).

Marchesi's 5th problem relates to the analysis of the morpheme after the name LUM-ma. He states:

“The final -a of LUM-ma in line 14 remains unexplained. All the analyses that have been proposed (genitive, locative or ergative postposition; morpheme of the ‘relative sentence'; enclitic copula) are either ungrammatical or in contrast with the scribal spelling conventions in use in the Pre-sargonic period” (p. 124).

Without evidence it is difficult to see which of his statements relate to the ergative postposition. It seems to me that none. If the referent of the complex construction in v 10-14 is Eanatum, then he must be the subject of the transitive verb in the clause, an ergative case-marker is thus expected at the end of the construction. The writing of the ergative case-marker as -a after a word ending in /a/ is well attested in later periods, and Attinger (1993: 211) quotes two examples in administrative texts from the period of Eanatum's inscription (Nik 1, 148 rev. ii 5 and Nik 1, 149 rev. ii 1). Both of his examples contain the suffix -/'a/ before an ergative case-marker, and this may well also apply to the form written as LUM-ma-a.