Reed and twigs were bundled into bundles (sa) and bound into bales (gu-nigin2) of various numbers of bundles for transport. I adopt the translation “bale” from A. Sallaberger (“Zum Schilfrohr als Rohstoff in Babylonien,” in B. Scholz, ed., Der orientalische Mensch und seine Beziehungen zur Umwelt: Beiträge zum 2. Grazer morgenländischen Symposion [Graz 1989] 326, note 63). It is current convention to posit two words for bale, gu-nigin2 and gu-kilib. E. Sollberger, in his glossary to Business and Administrative Correspondence under the Kings of Ur (=TCS 1; Locust Valley, New York, 1966) 122, sub voce “gu” states: “These two words are not only synonyms but homographs and it is only when they are followed by a suffix that the actual reading can be ascertained: the context does not help.” This view has been generally accepted. H. Waetzoldt, “Rohr und dessen Verwendungsweisen,” BSA 6 (1992) 126, observed “Den Ausdruck gu-kilib(-ba) benutzte man in Umma, während in den anderen Provinzen eher gu-nigin2(-na) üblich war.”
In M. Sigrist, SAT 2, 506 and 586, is found the form gu-NIGIN2-bi. The suffix -bi must be the possessive suffix. This shows that the suffix -ba in gu-NIGIN2-ba could be a locative added to the possessive suffix. The contrast between gu-NIGIN2-na and gu-NIGIN2-ba could accordingly consist of a form with and a form without the possessive suffix. That would allow positing a single word, gu-nigin2, “bale.”
This solution is confirmed by the fact that the reading kilib of the sign LAGAB is not attested. It is missing among the entries 27-33 in Proto-Ea (MSL 14, 31) and makes only a half-hearted appearance in the form of ke-el = LAGAB = napharu in Ea I, 42 (MSL 14, 178) and ki-li = [LAGAB] = napharu in Aa 2 (MSL 14, 211).
The formula n sa gi gu-nigin2-na n sa-ta (i3-gal2) means “n bundles reed. (Contained) in a bale (are) n bundles each” and the formula n sa gi gu-nigin2-ba a sa-ta (i3-gal2) “n bundles reed. (Contained) in its bale (are) n bundles each.”