A distinctive feature of Babylonian onomastics in the first millennium BCE is the use of family names at most cities by a segment of the population that can be described as the urban notable class. These family names are common and the conventions for their usage are well established in the abundant legal and administrative tablets that date from the so-called ‘long sixth century’: the period stretching from Nabopolassar’s first regnal year in 625 to Xerxes I’s suppression of the Babylonian revolts in 484 (Reference Jursa, Hackl, Janković and KleberJursa et al. 2010, 2–5). The use of family names emerged during the preceding eighth and seventh centuries, and the antecedents of some families and family names can be traced even further back in time to the early first millennium or even the latter part of the second millennium. Furthermore, some of these families persisted into the latter half of the first millennium BCE, as demonstrated by the continued presence of family names in Seleucid-era tablets.

Usage of family names at all times appears to have been restricted. Non-Babylonians never had family names, and only Babylonians of a certain social status were identified in texts with family names. Where the line of social demarcation lay is difficult to determine. Slaves and people of servile status, such as temple oblates, did not have family names, but neither did some men who had sufficient wealth to purchase land associated with the temple (Reference NielsenNielsen 2015b, 101), suggesting that an element of familial pedigree was involved. One could not simply adopt a family name. As a consequence, an understanding of the norms of family-name usage and an ability to identify them in Neo-Babylonian texts is essential for comprehending how individuals from the urban notable class functioned politically, economically, and socially.

After a discussion of the origins of family names in Babylonian society, we will present an overview of the types of family names that were in existence and then outline the different ways in which family names were recorded in texts, before concluding with some comments on the geographical distribution of family names throughout Babylonia.

Origins

Family names first became popular in the cities of Babylon, Borsippa, and Dilbat in the eighth and seventh centuries. They probably served as a means of projecting social cohesion and marking identity among urban notables at a time when the Babylonian state was weak and decentralised. For much of this period, Assyrians and Chaldeans occupied the Babylonian throne, and the urban notables would have had an interest in communicating their local identities to these non-Babylonians in order to ensure that their traditional prerogatives were respected. The practice may have become widespread in imitation of Aramean and Chaldean tribal groups, whose members were distinguished as sons of the eponymous ancestor for which their tribe was named.

Whatever caused the practice to gain popularity, it is evident that it had antecedents in the earliest centuries of the first millennium and even the latter second millennium. The family name Arad-Ea stands out as having belonged to a prominent family from Babylon whose members often held the office of governor (bēl pīḫati) in the royal administration beginning in the Kassite Dynasty (Reference LambertLambert 1957, 2). One member of the family could even trace an incomplete lineage back to the Kassite-era scribe Arad-Ea, from whom the family claimed descent. A Kassite-era cylinder seal from the late fourteenth century bearing the inscription of ‘Uballissu-Marduk, šatammu … of Kurigalzu, king of the world, son of Arad-Ea, the ummiān nikkassi’ is echoed in an inscription on a stele (kudurru) from the second quarter of the twelfth century in which a governor named Marduk-zākir-šumi was called ‘son of Nabû-nādin-aḫḫē, whose grandfather was Rēmanni-Marduk the liplippu of Uballissu-Marduk, descendant of Arad-Ea’ (Reference BrinkmanBrinkman 1993).

The term liplippu, which the Chicago Assyrian Dictionary defines as ‘offspring, descendant’, was not used in administrative texts but did appear in inscriptions, royal genealogies, and colophons on literary and scholarly texts, and typically expressed descent from a more distant ancestor. There are a few instances of genealogies similar to the example from the Arad-Ea family in which liplippu was used to indicate that possessors of family names could claim a genuine, or at least a multi-generational yet fictitious, descent from an ancestor who could be traced to the second or early first millennia. Colophons on tablets from the Epic of Gilgamesh identified members of the Sîn-leqe-unninnī family from Uruk – who often were kalû priests, just as descendants of Arad-Ea frequently held the title bēl pīḫati – as liplippu of Sîn-leqe-unninnī. Members of the Sîn-leqe-unninnī family either wishfully or legitimately claimed descent from a figure who was credited in later Babylonian tradition as having composed the epic and who may have been responsible for editorial undertakings in the second millennium that resulted in the version of Gilgamesh as it was known in the first millennium (Reference Beaulieu, Dietrich and LoretzBeaulieu 2000, 1–16; Reference GeorgeGeorge 2003, 28–33).

It is possible that these lines of descent included multiple ancestors whose names became family names. A stele (kudurru) written in the early ninth century at Borsippa concerns an ērib bīti priest named Nabû-aplu-iddin, son of (DUMU) Abunāya and liplippu of Aqar-Nabû (BBSt. 28). Aqar-Nabû was the family name of the chief administrator (šatammu) of the Ezida temple and ērib bīti priest of Nabû at Borsippa a century later, so it is certain that Nabû-aplu-iddin was an early member of this family. However, Nabû-aplu-iddin was petitioning the king for the restoration of his paternal estate (bīt abi), land that had belonged to his father, Abunāya. Abunāya is also attested as a family name in seventh-century texts, and this attachment to the ‘house of the father’ may have led to familial segmentation in which one branch of the Aqar-Nabû family became known as the Abunāya family (Reference NielsenNielsen 2011, 74–8).

Finally, there are the antecedents of the Šangû-Sippar family found in the Sun God Tablet from Sippar (BBSt. 36). In the waning years of the eleventh century, Ekur-šumu-ušarši, the šangû priest of Šamaš, petitioned the kings Simbar-Šīḫu (1025–1008 BCE) and Eulmaš-šākin-šumi for help maintaining the cult of Šamaš at Sippar following the destruction of the cult statue of Šamaš by Sutean raiders. More than a century later, during the second quarter of the ninth century, Nabû-nādin-šumi, who was the šangû priest of Šamaš at the time, discovered an image of Šamaš and petitioned the king for aid to remake the statue of Šamaš and fully reinstitute his cult at Sippar. Nabû-nādin-šumi had been able to recount Ekur-šumu-ušarši’s earlier efforts to the king and claimed to be a descendant of that earlier šangû priest. He did not call himself a liplippu of Ekur-šumu-ušarši, but instead described himself as ‘from the seed’ (ina zēri). In spite of the difference in terminology, the sentiment embodied in both terms is the same. Furthermore, even though šangû priest of Sippar was only used as a title in the text, it is very likely that a familial attitude towards the title was held by Nabû-nādin-šumi and that he was an early member of what would become the Šangû-Sippar family (Reference Bongenaar and BongenaarBongenaar 2000, 77–8).

Types

Family names can be grouped into two basic categories: ancestral names and occupational names. Ancestral names had fallen out of favour as given names in the first millennium and were practically never used as personal names by living persons. These family names referenced an eponymous ancestor from whom the family claimed descent. In most cases the historicity of this ancestor is unverifiable, but, as the discussion of liplippu demonstrated, there are a few cases where it is possible to identify the historical ancestor from whom the family took its name. As a result, we cannot discount the possibility that any ancestral family name actually referenced a formerly living person, though it is likely that many such family names were based on fictive descent. The overwhelming majority of ancestral family names were masculine names preceded by a so-called Personenkeil, the single vertical wedge that served as a determinative before a masculine personal name in the cuneiform writing system (see Chapter 1). Interestingly, there are a few examples of feminine personal names that were in use as family names (e.g., Arrabtu ‘(female) Dormouse’ or Maqartu ‘Precious’). These names were initially preceded by the sign MUNUS, the feminine determinative in texts. With the passage of time, however, scribes began to ‘masculinise’ these names by replacing MUNUS with the masculine personal name determinative (Reference Wunsch, del Olmo Lete, Feliu and Millet-AlbàWunsch 2006, 459–69).

Unlike ancestral family names, occupational family names are not marked by a personal name determinative in texts, but rather by the occupational determinative LÚ. Many of these names were derived from titles associated with the temples and represented the full extent of the priestly hierarchy. Names taken from both high-ranking temple-enterer priesthoods (e.g., Šangû-DN ‘Priest of DN’ or Kutimmu ‘Goldsmith’) and the lower-ranking purveying priesthoods (e.g., Ṭābiḫu ‘Butcher’, Rēˀi-alpi ‘Oxherd’, or Atkuppu ‘Reed-worker’) were used by families. While these families often had close associations with the temples, there are other occupational family names that may reflect association with the state or military apparatus (e.g., Lāsimu ‘Scout’ or Rēˀi-sisê ‘Horse herder’; Reference StillStill 2019, 82–3). And while it is not always the case that an individual with an occupational family name held that office or title, there are examples of families that had a strong association with or even monopolised the role: the Rēˀi-alpi family, for example, dominated the ox-herder prebends at Borsippa (Reference JursaJursa 2005, 93–4).

Usage

Family names were typically communicated in texts using the language of filiation and descent. They originally replaced the name of the referent’s father in a simple two-tier genealogy in which the individual (PN₁) was called the ‘son of’ (DUMU or A) the family name (PN₂). This practice has the benefit of allowing the reader of the tablet to differentiate between an individual who had an occupational family name and one who belonged to the occupation: the former would be called ‘son of’ the occupation (e.g., Bēl-ibni the son of the Potter [family]), while the occupational title simply followed the name of the latter (e.g., Bēl-ibni the potter).

The use of two-tier genealogies to express family names, however, poses some challenges for modern readers. The first challenge is the occasional appearance of individuals from Chaldean or Aramean tribes in legal and administrative tablets. Tribal affiliation could be expressed in two-tier genealogies, as a sale of a house located at Uruk in 673 BCE at the Chaldean city of Šapīya reveals. The first witness was Ea-zēru-iqīša, the chief of the Chaldean tribe of Bīt-Amukāni, who was identified as the ‘son’ of Amukānu (wr. ^Idé-a-NUMUN-BA-šá A ^Ia-muk-a-nu). However, the second witness, Naˀid-bēlanu, son of Aya-rimî, was probably a Chaldean as well; Naˀid-bēlanu had a Babylonian name, but his patronym, Aya-rimî, was West Semitic (Reference FrameFrame 2013 no. 4). The other witnesses had two-tier genealogies written in the same way as Ea-zēru-iqīša’s, but their patronyms refer to their father’s names and not to a family or tribal name, with the possible exception of the sixth witness, Nabû-zēru-ibni. His patronym, Nabûnnāya, was probably a family name (Reference NielsenNielsen 2015a, 256).

Nabû-zēru-ibni’s example brings us to the second challenge: it can be unclear whether a patronym in a two-tier genealogy is a family name or the father’s name, particularly if the family name is infrequently attested. It is doubtful that this was a problem in antiquity; the corpus of names in use as family names probably would have sounded quaint or old-fashioned to a Babylonian if one had been used as a personal name. The modern reader has to either develop familiarity with the corpus of personal names and family names or consult personal name lists. Nevertheless, the use of two-tier genealogies to express both family affiliation and paternity may still have led to some confusion. One solution to this problem was the appending of -šú šá to DUMU or A in genealogies, resulting in a writing of PN₁ DUMU-šú šá/A-šú šá PN₂. In the latter half of the first millennium the -šú was dropped but the šá was retained. This appended writing made it clear that the patronym was the father’s name and not a family name. The appearance of appended two-tier genealogies did not mean that the writings DUMU or A only preceded family names; there are examples of tablets in which these writings preceded the name of an individual’s father.Footnote ¹ However, if the scribe used both appended and unappended two-tier genealogies in a witness list it could be an indication that the witnesses with unappended genealogies had family names while those with appended writings did not.

The other solution was the introduction of an additional tier to genealogies. In the seventh century, three-tier genealogies in which the father’s name was expressed with an appended writing in the second tier and the family name was recorded in the third tier with an unappended writing (i.e., PN₁ DUMU-šú šá/A-šú šá PN₂ DUMU/A PN₃) became more common in texts. This practice had the benefit of preserving the name of the referent’s father as well as his family name. As a result, it becomes easier to identify brothers, uncles, and even cousins from the same family. A further elaboration of the three-tier genealogy occurred in the Seleucid period; in tablets from Uruk a fourth tier appears in many genealogies. It is unclear why this change occurred, but one possible explanation could be the strong preference for names featuring the god Anu as a theophoric element that had emerged, and the fact that most of the individuals appearing in the cuneiform texts from Seleucid Uruk came from the limited circle of endogamous families that dominated temple affairs (Reference Beaulieu, Waerzeggers and SeireBeaulieu 2018, 202–3). Specifying a man textually may have necessitated the addition of a fourth tier. Furthermore, women, when they do appear in texts, would also be identified by a variant of the three-tier genealogy. The patriarchal nature of Babylonian society meant that women were never affiliated directly with their family names as a ‘daughter of’ the family name. Instead, women were associated with their family on the basis of their relationship to a male family member. A woman was usually ^fPN₁ ‘daughter of/wife of’ ^IPN₂ DUMU/A PN₃, meaning that marriage effectively aligned her with a new family name.

Geographic Distribution

Family names were not ubiquitous throughout Babylonia. Although family names can be found on tablets dated at every Babylonian city in the Neo-Babylonian period, their usage did not become conventional everywhere. Greater population size and density in urban areas may have made family names a useful means for differentiating individuals in texts, and economic and cultural networks between cities probably contributed to the spread of the practice. They were used earliest and with greatest frequency at cities in northern Babylonia, at Babylon and the nearby cities of Borsippa, Dilbat, and Kish. Further north, it is also possible to observe that family names became used more frequently at Sippar. Some of these families, most notably the Šangû-Šamaš or Šangû-Sippar family, had a long presence at Sippar that may have extended back to the eleventh century and the events commemorated in BBSt. 36. Still others, such as the Ša-nāšišu family (Reference BongenaarBongenaar 1997, 470–5; Reference Jursa, Hackl, Janković and KleberJursa et al. 2010, 71–2), had relocated to Sippar from Babylon. At Nippur, however, there seems to have been an almost conscientious rejection of the use of family names (Reference NielsenNielsen 2011, 163–5, 177–80). This was in spite of textual evidence indicating the presence of the same cultural sentiments and practices relating to revered scholars (Reference RubinRubin 2022) and prebendary functions (Reference Joannès and EllisJoannès 1992, 90; Reference BeaulieuBeaulieu 1995, 88–9) at Nippur that were the basis for ancestral and occupational family names elsewhere. Those few family names that are attested in documents dated at Nippur may have belonged to non-Nippureans. Family names were nearly as uncommon in tablets dated at Uruk and Ur in southern Babylonia as they were at Nippur, but there are indications that the practice was taking hold during the seventh century (Reference NielsenNielsen 2011, 217–20). Prosopographical analysis reveals that individuals who were identified in texts with family names appeared in other texts without such names. Furthermore, the names of other male kin to these individuals were also recorded without mention of their family name, with a few exceptions in which it was clear they shared the same family name. Family identity was present among some of the population even if there was no compulsion to record it in texts.

Not only was there an uneven geographic distribution of family-name usage throughout Babylonia, it is also evident that some family names originated at or were strongly associated with specific cities. For example, the Ea-ilūtu-bāni, Aqar-Nabû, and Iddin-Papsukkal families had ties to Borsippa; the Šangû-Dilbat and Salāmu families were from Dilbat; and the Ekur-zakir, Ḫunzû, and Sîn-leqe-unninnī were predominantly from Uruk (Reference Wunsch, Krebernik and NeumannWunsch 2014, 289–314). Furthermore, branches of these families spread to other cities after the relocation of members. The Ṣāḫit-ginê family at Sippar was descended from a man from Babylon named Dayyān-Marduk (Reference WaerzeggersWaerzeggers 2014, 29–30), and it may be possible to trace the Iddin-Papsukkal family at Ur and Uruk back to Borsippa, where the family appears to have had its origins (Reference NielsenNielsen 2009, 171–82). An awareness of the associations that some families had with certain cities and the movements of certain families over time can provide context for understanding the social networks present in a tablet. Furthermore, family names can provide useful clues when damage to an unprovenanced tablet results in the loss of the name of the city at which the tablet was dated.