Fortunately, in the case of a large number of names occurring on business documents as the interested parties or as scribes or as witnesses - and it is through these documents that we obtain the majority of the Babylonian-Assyrian proper names - we have variant readings, the same name being written phonetically in whole or part in one instance and ideographically in another. Certain classes of names being explained in this way, legitimate and fairly reliable conclusions can be drawn for many others belonging to the same class or group. The proper names of the numerous business documents of the Khammurabi period, when phonetic writing was the fashion, have been of special value in resolving doubts as to the correct reading of names written ideographically. Thus names like Sin-na-di-in-shu-mi and Bel-na-di-in-shu-mi, i.e. "Sin is the giver of a name" (i.e. offspring), and "Bel is the giver of a name," form the model for names with deities as the first element followed by MU-MU, even though the model may not be consistently followed in all cases. In historical texts also variant readings occur in considerable number.
Thus, to take a classic example, the name of the famous king Nebuchadrezzar occurs written in the following different manners: - (a) Na-bi-um-ku-du-ur-ri-u-ṣu-ur, (b) AK-DU-u-ṣu-ur, (c) AK-ku-dur-ri-SHES, and (d) PA-GAR-DU-SHES, from which we are permitted to conclude that PA or AK (with the determinative for deity AN) = Na-bi-um or Nebo, that GAR-DU or DU alone = kudurri, and that SHES = uṣṣur. The second element signifies "boundary" or "territory"; the third element is the imperative of nasaru, "protect"; so that the whole name signifies, "O, Nebo! protect my boundary" (or "my territory").
It is not the purpose of this note to set forth the principles underlying the formation of proper names among the Babylonians and Assyrians, but it may not be out of place to indicate that by the side of such full names, containing three elements (or even more), we have already at an early period the reduction of these elements to two through the combination of the name of a deity with a verbal form merely, or through the omission of the name of the deity. From such names it is only a step to names of one element, a characteristic feature of which is the frequent addition of an ending -tum (feminine), ān, ā, um, atum, atija, sha, etc., most of these being "hypocoristic affixes," corresponding in a measure to modern pet-names.
Lastly, a word about genuine or pseudo-Sumerian names. In the case of texts from the oldest historical periods we encounter hundreds of names that are genuinely Sumerian, and here in view of the multiplicity of the phonetic values attaching to the signs used it is frequently difficult definitely to determine the reading of the names. Our knowledge of the ancient Sumerian language is still quite imperfect, despite the considerable progress made, more particularly during recent years. It is therefore not surprising that scholars should differ considerably in the reading of Sumerian names, where we have not helps at our command as for Babylonian and Assyrian names. Changes in the manner of reading the Sumerian names are frequent. Thus the name of a king of Ur, generally read Ur-Bau until quite recently, is now read Ur-Engur; for Lugal-zaggisi, a king of Erech, some scholars still prefer to read Ungal-zaggisi; the name of a famous political and religious centre generally read Shir-pur-la is more probably to be read Shir-gul-la; and so forth. There is reason, however, to believe that the uncertainty in regard to many of these names will eventually be resolved into reasonable certainty.
A doubt also still exists in regard to a number of names of the older period because of the uncertainty whether their bearers were Sumerians or Semites. If the former, then their names are surely to be read as Sumerian, while, if they were Semites, the signs with which the names are written are probably to be read according to their Semitic equivalents, though we may also expect to encounter Semites bearing genuine Sumerian names. At times too a doubt may exist in regard to a name whose bearer was a Semite, whether the signs composing his name represent a phonetic reading or an ideographic compound. Thus, e.g. when inscriptions of a Semitic ruler of Kish, whose name was written Uru-mu-ush, were first deciphered, there was a disposition to regard this as an ideographic form and to read phonetically Alu-usharshid ("he founded a city," with the omission of the name of the deity), but scholarly opinion finally accepted Uru-mu-ush (Urumush) as the correct designation.
For further details regarding the formation of Sumerian and Babylonian-Assyrian proper names, as well as for an indication of the problems involved and the difficulties still existing, especially in the case of Sumerian names, see the three excellent works now at our disposal for the Sumerian, the old Babylonian, and the neo-Babylonian period respectively, by Huber, Die Personennamen in den Keilschrifturkunden aus der Zeit der Könige von Ur und Nisin (Leipzig, 1907); Ranke, Early Babylonian Proper Names (Philadelphia, 1905); and Tallqvist, Neu-Babylonisches Namenbuch (Helsingfors, 1905).
 For a survey of the chronological systems adopted by different modern scholars, see below, section viii. "Chronological Systems."
 The compiler of the more complete one seems to have allowed himself liberties. At all events he gives 30 years of reign to Sin-muballidh instead of the 20 assigned to him in a list of dates drawn up at the time of Ammi-zadok's accession, 55 years to Khammurabi instead of 43, and 35 years to Samsu-iluna instead of 38, while he omits altogether the seven years' reign of the Assyrian king Tukulti-In-aristi at Babylon.
 They are also called high-priests of Gunammidē and a contract-tablet speaks of "Tē in Babylon," but this was probably not the Tē of the seal. It must be remembered that the reading of most of the early Sumerian proper names is merely provisional, as we do not know how the ideographs of which they are composed were pronounced in either Sumerian or Assyrian.