Applying Heterarchy Theory to Ancient Mesopotamian Religions A

Heterarchy theory is a valuable tool for analyzing complex and changing relationships between elements in a syste m. It has been employed in anthropology, archaeology, and recently in religious studies. Its utility has not yet been exploited for religions that are studied through textual evidence, such as Mesopotamian religions. As Mesopotamian religions were polytheistic and the texts represent multiple genres from a broad time - frame, relationships between system actors such as gods, temples, and cities defy static and lineal arrangements. Heterarchies are well suited for untangling these relation - ships, showing how they change depending on the measuring criteri a. Using the case of the city of Assur, which housed many deities and was both the religious center and a political capital, heterarchy theory shows how the same elements – temples and cities – reveal different rankings that coexisted simultaneously. Heterarchies produc - tively complicate our understanding of these religious relationships and expose the multimodality of each element in the syste m.


Heterarchy Theory
Heterarchies organize elements in a system such that "each element possesses the potential of being unranked (relative to other elements) or ranked in a number of different ways, depending on systemic requirements" (Crumley 1979: 144).1 Crumley first introduced heterarchy theory, adapted from its original application to cognitive structures of the human brain (McCulloch 1945), into anthropology and archaeology (Crumley 1979(Crumley , 1995(Crumley , 2015)).Since then, heterarchical approaches have gained traction in archaeological and anthropological studies as a rejection of viewing sociopolitical complexity in terms of hierarchical organization (for instance, Rautman 1998, with additional bibliography).
In contrast to heterarchies, a hierarchical system "involves three assumptions regarding the organization of the constituent elements of a system: that a lineal ranking of constituent elements is in fact present; that this ranking is permanent (that is, the system of ranking has temporal stability); and the ranking of elements according to different criteria will result in the same overall ranking (that is, the relationships of elements is pervasive and integral to the system, and not situational)" (Rautman 1998: 327).Using heterarchies to understand organizational structures means a rejection of this framework and, as Rautman points out, "forces us to specify more clearly the context and temporal duration of the relationships that we are describing" (327)(328).At the same time, heterarchies exist in dialog with hierarchies, and Crumley explains the relationship between the two as follows: This distinction between hierarchical and heterarchical structure can be portrayed in three-dimensional space: connections among elements in a hierarchical structure are most frequently perceived as being vertical … whereas heterarchical structure is most easily envisioned as lateral, 1 A more detailed description of the forms that heterarchies may take can be found in Rautman 1998: 328 as a mixed paraphrase and quotation of a system originally laid out by Brumfiel: 1. Several elements may operate more or less independently of one another in a single system; 2. The elements at issue may belong to 'many different unranked interaction systems'; 3. The 'participation of each element in the overall system may be determined by the "needs of each element"' (and which presumably may vary independently of the needs of other elements or of the system as an aggregate); 4. Elements may be members in 'many different systems of ranking' such that the same element might occupy a different rank in each different system; 5.There may be 'two or more functionally discrete but unranked systems that interact as equals'; and finally, 6.The overall system may include 'two or more discrete hierarchies that interact as equals.' Brumfiel 1995: 125 Downloaded from Brill.com09/21/2023 11:26:07AM via free access

Zaia
Numen 70 (2023) 598-624 emphasizing the number and variety of connections among elements and the varying circumstantial importance of any single element.Crumley 1979: 144 One system may contain both hierarchies and heterarchies, which variously appear depending upon the filter applied to the elements of the system.2Heterarchy theory has thus far been only occasionally applied in Assyriological studies, and mostly in archaeological or anthropological works, such as E. Stone and P. Zimansky's (2004) work on Maškan-šapir and C. Meyers's (2006) study of Iron Age Israelite society, its popularity in other anthropological and archaeological fields likely facilitating its application to ancient material culture remains.In contrast, heterarchy theory has not received much attention in research based on philological evidence, with the most notable (and perhaps first) foray in this direction in Assyriology being S. Svärd's (2012Svärd's ( , 2015) ) studies of royal women and power relations during the Neo-Assyrian period.She writes: "power is always performed and secured in complex ways that 'hierarchical' just does not describe well enough" (2012: 510), demonstrating how women in the court were able to negotiate and exert power in ways that do not correlate to hierarchical ranks.As she argues, using heterarchies as an analytical tool allows us to view lateral and flexible relationships that existed alongside the hierarchical ones that scholars have traditionally explored (2012: 515; 2015: 147-170).
This article further expands the application of this method for both philological studies and Assyriology in general by employing heterarchy theory in the study of ancient Mesopotamian religion, which modern scholars access primarily through textual evidence.Heterarchy theory has significant potential in its application to religious studies in general, and has already been fruitful in the subfield of urban religion in other disciplines.3Religious studies in Assyriology has much to gain from the theory as well, as the volume of relevant texts, the variety of represented genres, and the broad chronological scope 2 Crumley uses as her example an automobile company: an automobile company may be seen as hierarchically organized in terms of corporate decision making, and heterarchically organized in terms of the production of an automobile: into the final product goes the expertise of administrative, research and design, assembly and sales departments.If the unit of study is the automobile, all aspects are equally important.If the study has as its focus departmental efficiency or an interdepartmental softball tournament, however, the departments might be variously ranked.results in a complexity of relationships between system actors -whether individual gods, temples, or cities -that defies the static and lineal arrangements necessary for hierarchies.Ancient Mesopotamian religions were polytheistic as a rule, the multiplicity of deities and other divine manifestations presenting a chaotic cultic landscape that entices scholars -both ancient and modern -into attempting to untangle the multifaceted relationships and organize them into neat hierarchies and groupings.One can turn to "god lists," texts that essentially function as scribal inventories of deities, to observe the desire to collect and order the pantheon(s).Even these lists, which were highly and linearly organized and in many cases standardized, demonstrate the challenges that the ancient scribes faced when trying to create a hierarchical order based around lexical and/or "theological" principles.4The longest known god list, An = Anum, which contained almost 2,000 names and which was transmitted for generations, is a telling example of how, rather than presenting a simple ranking based on "importance" or "status," even these lists employed blended organizational principles and overlapping hierarchies.To wit, the list began with the oldest and most senior deities, but each of these deities was accompanied by his or her "household," that is, spouse, children, courtiers, and servants, where applicable, favoring groupings based on relationships rather than straightforwardly listing deities from most to least important or oldest to youngest.5In addition, there may have been macro-scale spatial qualities taken into consideration when ordering the list, progressing from above to below earth by beginning with the heavens (with An, the deified heavens, as the first deity of the first tablet) and ending with the netherworld (with Nergal, king of the netherworld, as the first deity of the sixth tablet).6These organizing principles were sometimes flexible, especially in the latter tablets of the series (tablets five through seven).7While the god lists demonstrate some stable patterns of ordering deities, especially among the most consistently attested gods, there was no one single way in which to order all of the known deities of the pantheon(s), much less one that was fully fixed and permanent across time and space, even in a genre that had the organization of the cosmos as one of its core concerns.8The varying relationships of deities outside of the list tradition were even less orderly.In general, Mesopotamian religions were complicated, shifting patchworks whose actors related to one another differently depending on time, place, and text genre; a god's age or characteristics; and scribal and theological traditions, among other considerations.These multifaceted relationships are thus better accessed using an approach in which rankings between elements in a system change based on different criteria and multiple ranking systems exist simultaneously, hence the potential of heterarchy theory.

Applying Heterarchies to Assur
To put this analytical tool into practice, this paper discusses the city of Assur (modern Qalʾat Sherqat, Iraq).Assur was the patron city of the god Aššur, from whom the city took its name (or vice versa), and remained the cultic heart of what would later become the sprawling Neo-Assyrian Empire (the "land of Assur"), even when the seat of government had moved elsewhere.9But, while Aššur was the patron of his eponymous city, and his main (and only major) temple was located there, Assur was far from being the exclusive domain of one deity.Illustrative is an episode from long before the imperial heyday, when Assur was part of the kingdom of Upper Mesopotamia, in which King Samsī-Addu I (ca.1809-1776 bce)10 wrote to his son Yasmaḫ-Addu that Mari and Assur "are full of gods" (Mari u Assurma ša ilānī malû; Ziegler 2019: 55).In his letter, Samsī-Addu I berated Yasmaḫ-Addu for installing so many additional gods in Mari (Tell Hariri, Syria), a capital city that already housed 8 Even if these lists presented fully realized hierarchies for all known deities, their utility for understanding the hierarchies of the living cult as practiced in a certain time or place is unfortunately limited.These lists, much like lexical lists, functioned as traditional collections and pedagogical tools for scribal contexts and, while some longstanding associations and relationships between gods remained relevant in cult practice, the scribes sometimes added new tablets to a series that comprised one or more lists of names from other sources without much modification or attempt to update the entire series according to existing practices (Lambert 1957(Lambert -1971;;1975: 195).Moreover, given the long timespan over which some of these lists survived and the inherent conservatism of the compositions (Lambert and Winters, forthcoming), the cults of many named deities, especially lesser-known ones, would have been inactive or even long defunct by the time of copying (Lambert 1975: 195), or may not have been venerated in Assyria at all.On the use of god lists in scribal curricula, see Shibata 2009: 35;Veldhuis 1997: 59, 130;Peterson 2009: 5-8;Lambert 1957Lambert -1971: 474. 9 : 474. 9 Although Assur, Aššur, and Assyria were all spelled and pronounced essentially the same, I distinguish between the city and the god as Assur and Aššur, respectively.10 Elsewhere in this article as Šamšī-Adad I, following the RIMA 1 conventions.several divinities, that the number of sheep and oxen needed for the regular sacrifices -the most important duty of a temple -now exceeded the available livestock (54-56).11An overabundance of gods concentrated in one space was not a trivial matter.Each deity, represented by a cult image (ṣalmu), meant an investment by the state, which was ultimately responsible for the construction and upkeep of shrines, cellas, and larger temple complexes; for providing regular food and drink sacrifices; for holding festivals; and for appointing staff, including priests and specialists (Ziegler 2019;Zaia 2021).12Multiple gods in one city was not in itself remarkable: while every city in Mesopotamia had at least a patron god, whose responsibility was to protect the city and its inhabitants, the patron's temple complex typically housed shrines to other deities in his or her retinue, such as consorts, children, and officials (Meinhold 2013: 325-334).13Nonetheless, the statements about Mari and Assur suggest that these cities were models of spaces overpopulated by deities, at least within Samsī-Addu I's kingdom.Assur had been home to several major temples since its earliest occupation phases and others were added over time, which makes it a good candidate for heterarchical analysis.Moreover, Assur itself was not the only important religious center in Assyria.Thus, heterarchy theory can reveal the nuances of Assur's religious relationships in two ways: internally, with regard to its many local temples and shrines; and externally, among the other political and religious hubs in the empire.This study examines first the intertemple relationships within Assur and then Assur's own status within the broader urban topography of Assyria.
For heterarchy theory to work as an analytical tool, it has to measure something -such as how Svärd (2012Svärd ( , 2015) ) measured power in relationshipsand one should apply different, ideally equivalent lenses or filters to see how the ranking and relations between comparanda change.14For Assur, both within and without, many of the same filters apply: economic power, religious qualities, political features, age, scope/size, amenities, geography, demography, significance to the state, importance within the pantheon(s), to suggest a few.These are all observable in the combined textual and archaeological data.As will be seen, the statuses that are revealed change dramatically depending on the measuring criteria and heterarchies at play.

Heterarchies in Assur: The Diversity of Intertemple Relations
Due to the divine demographics of the city, attempting to apply a strictly hierarchical approach to untangling the local religious topography of Assur would quickly fail.Other than the primacy of Aššur -whose status as the highest-ranking deity in the Assyrians' official religion and the primary patron of kings was unchallenged, and whose physical manifestation as a cliff in the city meant that he was always the dominant presence there -the other deities living in Assur cannot be comfortably or consistently ranked in relation to each other.15Moreover, Assur, not exclusively but more so than many other cities, represents the nexus of several layers of religious networks (local, regional, imperial) that collide into a messy, seemingly incoherent urban theology.The focus of this paper is on a few sample metrics by which one can observe a diversity of relations between temples in Assur, each metric representing a different "systemic requirement" that results in different rankings between the same elements within a system (that is, temples in the city of Assur).By viewing how the temples' rankings change in relation to each other depending on the applied criteria, heterarchy theory reveals the multiplicity of relationships between temples that existed simultaneously, enriching our perception of the role of each temple in its lived context.

Age and Royal Patronage
One lens by which rankings can be observed is age, which is necessarily intertwined with that of direct state support for temples since the textual evidence for existing temples comes primarily from kings recording the building or renovating of individual temples in their royal inscriptions.A diachronic view from the Old Assyrian period through the fall of the empire shows an evolving religious topography, with kings selectively reconstructing existing temples 15 It should be specified that official (that is, state-sponsored) religion was itself not necessarily representative of all forms of religious practice in the urban environment, such as familial and personal beliefs, but most of the sources are royal in nature.The complex relationships between Aššur and other deities have also been fruitfully examined through social network analysis (Alstola et al. 2019).Indeed, network analysis can be employed effectively as a method by which heterarchical relationships can be exposed and explored, as networks are ideally suited for viewing lateral relationships and how certain circumstances or genres might tangibly alter them.The Middle Assyrian period saw numerous changes to the religious topography through the addition of new temples or modifications to existing ones, with royal inscriptions mentioning many of the same temples in Assur -to Aššur, Ištar, Anu and Adad, Sîn and Šamaš, and Bēl-(l)abrīya (Greenwood 2008: 231-267) -as well as other, smaller shrines that may have been part of the Aššur temple complex; namely, those of Bēlat-šamê, Amurru, and the Divine Decad (264-267; Tiglath-pileser I RIMA 2 A.0.87.1).19In the Neo-Assyrian period, construction or renovation projects were recorded in the royal inscriptions for the following temples: Adad and Anu (Shalmaneser III RIMA 3 A.0 This group comprises the most significant temples in Assur from the royal perspective, since they were important enough for kings to (re)build them, though even here we might postulate that they may have organized the data in additional, different ways: we could argue that they ranked temples from oldest to newest, regardless of attested longevity, on the grounds that Assyrian kings valued tradition highly.Or, we could highlight the temples that appear in all three periods on the basis that their longevity meant the Assyrian kings deemed them worthy of consistent patronage.After all, many temples chronicled in one period (or even by one king) are not attested in others; for instance, the Ereškigal temple is not mentioned again after the Old Assyrian period, not even in the list of cult spaces in Assur.21The Ḫaya and Zababa temples are even shorter lived, appearing briefly under Sennacherib and never again, likely purposefully so (Zaia, forthcoming).We could observe how many individual kings recorded work on a temple, regardless of period, which perhaps indicated that these temples transcended a king's personal preferences as part of a broader royal tradition.There are also hints for the opposite situation; that is, that some shrines were neglected, suggesting that they had less immediate import to the reigning king.For instance, in the late Neo-Assyrian period, a letter to the king reported that Amurru's temple collapsed and the god (and his retinue) moved into the nearby Anu temple until it was rebuilt (SAA 10 21).Amurru's temple in Assur was not mentioned in royal inscriptions after the Middle Assyrian period, and it is only from this letter that we know of its continued, if dilapidated, existence.Did the temples that were no longer attested cease to exist, or did they continue silently in the background?How and why did kings discontinue or deconsecrate some temples?22 Did this have implica-20 These are a sample and are not comprehensive lists of texts in which these projects are mentioned.

21
A tākultu ritual for Aššur-etel-ilānī invokes her, suggesting that she was in Assur, but this is tenuous and the long gap between attestations is problematic (SAA 20 42).22 The gods of Kār-Tukultī-Ninurta were relocated to Assur at some point, and the temple was shut down and sealed with stone slabs (Schaudig 2010: 156;Eickhoff 1985: 34-35, 45, 51).Perhaps inverting the consecration of the temple via installing the divine statues, tions for the affected deity or deities in the larger pantheon?The sources are largely silent on these questions, but, if the temples were placed in a hierarchy based on age and royal patronage, these cases would comprise the lower rungs.
At the same time, we could suggest that there was a privileging of rarely or once-attested temples, which would be the newest or the shortest lived, with the reasoning that they were special projects driven by a king's individual preference or initiative, rendering them more significant than the temples that were maintained out of a longstanding sense of tradition.An example would be Sîn-šarru-iškun's Nabû temple.He claims in his building account that the previous temple had wasted away over time until it no longer existed, and that Nabû and his consort Tašmētu moved to the Ištar temple.23However, this may be a historical fiction, and the Ištar temple in Assur, which had long housed Tašmētu, an originally "single" Assyrian deity who was later given Nabû as a consort (Meinhold 2009: 80-81), may have been identified as the "earlier" residence of Nabû on this basis; the god did have a cult space in the local Ištar temple at least in the Neo-Assyrian period.24There is not yet any earlier archaeological or textual evidence for an independent Nabû temple (84).Thus, if Sîn-šarru-iškun had built an entirely new Nabû temple in Assur, this was a meaningful act of promotion within the official pantheon, one that even retrojected Nabû's importance by claiming the preexistence of an ancient cult space in the city.25 Therefore, even for those temples that received royal patronage, there may have been a distinction between those that were consistent beneficiaries and those that were the subject of especial veneration by a particular king, and each perspective results in a separate ranking of temples.Indeed, what complicates using age as a metric is that Assyrian kings both valued ancient institutions but also expressed their personal preferences via (re)construction projects, so a very ancient temple could be considered as significant in the urban landscape as a brand-new temple introducing a deity into the religious heart of Assyria.Thus, if ages, specifically those at the extremes of oldest and newest, are taken as the organizing principle for hierarchies, the deconsecration was achieved simply by removing the images.Reconstructing temples sometimes required tearing their structures down, which may have required deconsecration first (cf.RINAP

3.2
Individual Perspectives Another set of relationships is offered by professional or personal connections to individual cults, which would not necessarily align with state priorities.Essentially, the nature of these religious relationships within the city depends on whose perspective is taken.Cultic personnel of one temple would naturally have ranked their own deities as more important than others, as evidenced by intertemple rivalries and theft.26Personal and family gods are significant as well: for example, Pān-Aššur-lāmur, the governor (šakin māti) of Assur and later governor of Arbela (active between 776 and 759 bce, at least; PNA 3/1: 983-984), presumably had "professional" hierarchies comprising the state gods, then patron deities of the cities for which he was responsible (Aššur and Ištar of Arbela, respectively).Nonetheless, he dedicated a cylinder seal, a highly personalized object, to Gula, the healing goddess and patroness of the medical profession (Adad-nērārī III RIMA 3 A.0.014.2016).Nothing is known about this official's family or background, leaving open questions such as: did he have an ancestor who was a physician, and Gula was thus a family deity?Did he have personal ties to the Gula temple in Assur?27The seal may have been looted in antiquity from said temple, where it was probably offered as a votive object (Tadmor and Tadmor 1995: 354-355).Similarly, Bēl-tarṣi-ilumma, the governor (šakin māti) of Calah, dedicated an inscribed anthropomorphic statue to Nabû's temple in Calah, prompting the suggestion that Bēl-tarṣi-ilumma saw Nabû as a personal god (Robson 2019: 59-62, 67).Nonetheless, Bēl-tarṣi-ilumma, as governor of Calah, would certainly have venerated Ninurta as the patron of his city and head of the local pantheon in "institutional" contexts.28Personal venera-

26
For example, the Ištar of Arbela temple was robbed by a priest from a nearby shrine to Ea (SAA 13 138).For theological and cultic rivalries in Babylonia, see George 1997. 27 It is tempting to suggest that donations to Gula were made when the dedicant was ill, but it should be noted that dedicated objects inscribed with the hope that the deity would grant the dedicant a long and healthy life are conventional and not restricted to Gula offerings.28 The invocation of "institutional" or "professional" deities can be seen perhaps most clearly in the letter corpus; for instance, the šandabakku (governor) of Nippur, writing to Esarhaddon, blessed him by Enlil, Ninurta, and Nuska, prominent deities of the Nippurean city pantheon (SAA 18 70).Urad-Nanāya, the chief physician, regularly blessed Esarhaddon with good health by Gula, the patron goddess of physicians, and Ninurta, her consort (e.g., SAA 10 315).tion could create rankings of the city cults that were completely separate from "official" ones, and an inhabitant's perspective of personal or familial religion potentially changed their view of the divine relationships in the city in almost unpredictable ways.Naturally, these state, city, and personal hierarchies were not mutually exclusive but were constantly negotiated against one another.

3.3
Topography and Physical Space Topography is a complex lens that takes into account access, visibility, and function.Temples were not the only dwellings of gods, as many deities were embedded into the structural fabric of the city, including in the city walls and gates, as well as on terraces and in other spaces that ranged from fully public to highly restricted.In Assyrian cities, temples were typically clustered in the walled citadel alongside "secular" buildings, particularly the palaces but also residential areas (see Figure 1).The distribution of various social classes in this space further suggests the utility of heterarchical approaches, which cut across rigid categories of "religious" and "political," to understand the lived experience of religion in the city.29In particular, differing levels of access to shrines may have affected the relationships between local humans and deities.
Assyrian scholars were themselves interested in understanding and preserving the religious topography of the empire, and especially of Assur.The most relevant text is the so-called Götteradressbuch, which contains several sections in list form: the temples in Assur; the city's gates and its defensive structures (walls, moats, protective deities); temples around the empire, including in Babylonia; the ziggurats in Assur; the ceremonial names of four significant Assyrian cities (including Assur); the names of Ešarra's gates; and a summary section (George 1992: 167-184;SAA 20 49).Because of the list format, it is tempting to assume a hierarchical order, but it is not so straightforwardwe must consider who made these lists, why, and what their organizational priorities may have been, especially since no single recension of this text includes all sections and may combine independent traditions.30We should invoke deities important to their respective occupations, especially as they were writing in a professional capacity.certainly not read these lists hierarchically, from top to bottom, in order of importance or status.The first three sections at least are exclusive to Assur, and have a spatial component, since the text starts with Ešarra's "holy of holies" (bīt papāḫi) and the gods that lived there.As mentioned, Ešarra was the most prominent temple complex, but it was not Aššur's alone.There are several deities in the inner sanctum and gods stationed in various parts of the complex, including integrated temples to other deities, such as that of Mullissu and her retinue (van Driel 1969: 37-45).These sections thus seem to take into account that there are different types of divine spaces, even within one temple complex.Size is an additional potential metric, delineating between "stand alone" temples and shrines that were located within temple complexes, for example, with the assumption that being the primary or sole deity of a temple reflected a higher ranking than did occupying a shrine within the temple complex of another deity.What follows Ešarra in the text are other Assur-based temples and their occupants, again with some suggestion of spatial relationships, such as placing the Anu and Adad groups next to each other in sequence, as well as the Sîn and Šamaš groups, as these were shared temples (Anu-Adad and Sîn-Šamaš).31The gates and other features like the walls and ramparts are geographically farther from the "core" of the city but are often dedicated to or represented by deities.Gates could be named after specific gods (i.e., the Šamaš gate), and the walls were guarded by "divine sentinels," generally represented by a cult image in a niche in the wall or gate (George 1993: 177).If presence within Ešarra's inner sanctum -perhaps to be regarded as the most elite, restricted space -was considered to be the gauge by which a deity's status is determined, then the deities of the city walls were peripheral or even liminal.On the other hand, if the ranking was based on deities who have important roles protecting the city from attack, on deities to whom someone might routinely pay homage while entering and exiting the city, or on levels of public accessibility and visibility, then the relationships look quite different.Texts like the Götteradressbuch can demonstrate some of the many overlapping organizational principles at stake in Assur.some extent, but they were still subject to power relations within the city and its local pantheon.Conversely, a locally important deity may not have been ranked as highly when the larger state pantheon was taken into consideration.For instance, Gula, who had a temple in Assur, was not a significant deity in the overall state pantheon and was attested only rarely in royal inscriptions, while deities like her consort Ninurta, who did not have an independent temple in Assur, figured more prominently in imperial ideology.Ninurta, in turn, was more visible in the royal inscriptions of Ashurnaṣirpal II (883-859 bce) from Calah, Ninurta's patron city and Ashurnaṣirpal II's new capital, than in inscriptions from other cities (or kings' reigns).While this indicates that Ninurta ranked at the top of his city's hierarchy, and the king promoting Calah as the capital may have also boosted his status in the state pantheon somewhat, Ninurta was still positioned in relation to other deities in the official pantheon, within and outside of Calah.32 Heterarchies better represent the phenomenon in which an individual deity had or took on a more visible role in certain locations, regulating this status not as a change in absolute rank (to the extent that one existed) but in relation to other deities within that context, with the implicit acknowledgment that the lateral relationships were negotiable and subject to certain conditions.In fact, many deities who had only a minor (or no) temple or shrine in Assur were patrons of other cities; that is, they were the most powerful deity of that city's pantheon.The Assyrian kings patronized many of those temples as well, sometimes preferring to support the flagship temple rather than that same god's shrine in Assur (if that deity had one).For instance, Sennacherib (re)built the temple of Nergal in Tarbiṣu, his cult city in Assyria (e.g., RINAP 3/2 213,214,215), but never mentions the shrine of Nergal in the Ešarra slaughterhouse, which is only attested in a list of Ešarra shrines (George 1992: 187).Similarly, Sargon, Esarhaddon, and Ashurbanipal invested in Sîn of Ḫarrān's cult (Novotny 2003: 49-194) rather than in the ancient Sîn and Šamaš temple in Assur.33A deity could also have an independent temple while being a member of another deity's retinue in the same city.Ištar, for example, had her own temple in Assur that housed almost twenty other deities, but she was also listed as a resident in Gula's temple (100)(101)(102)(103)(104)(105)(106)(107)(108).Ištar was moreover patron goddess of two cities, Nineveh and Arbela.Location and local status of a deity thus create two vectors that could each be given different weight in ranking, resulting in several possible configurations: (1) by proximity to Assur, regardless of a deity's local status; (2) by the deity's local status, regardless of location; (3) a combination of proximity to Assur and local status.These three arrangements could be further modified by taking into account deities who lived in multiple temples -whether in the same city or in multiple cities -versus those that are only known from one.Considering the status of the location itself could create an entirely new set of rankings built on similar factors as those discussed above, such as a city's age, its political importance, its size, its geographical location, and so on.These various configurations create a complex arrangement of relationships, many of which functioned almost independently (especially outside of a uniting force like a state pantheon), and all of which the inhabitants of a given city may have understood and perceived simultaneously.

Assur in Heterarchies: Relationships to Other Urban Centers
Moving from Assur to the broader Assyrian urban landscape, one can observe that individual cities were placed into various, shifting rankings with one another as well.Religion is one of the possible criteria by which to evaluate inter-urban relationships, since several cities were important to the state and its people primarily for cultic reasons.This is certainly the case with Assur, which, as the traditional cultic heart of Assyria, consistently dominated the religious hierarchy of Assyrian cities.Nonetheless, assessing Assur's importance on purely religious grounds does not provide an authentic perspective of the city's role in Assyria and its relationships to other urban centers.Indeed, a city's overall profile is a heterarchical composite of unstable rankings and separate but overlapping hierarchies based on criteria including not only religious significance but also size, geography, economic benefits, political status, demographics, natural resources, and so forth.Situating religion among these other features better clarifies its impact in shaping urban networks, and applying these lenses paints a portrait of Assur and its position within Assyria's changing urban landscapes that is more nuanced than its typically straightforward characterization as a cultic center.As the following discussion demonstrates, Assur's relationships to other urban centers appear dramatically different based on which filters are applied and its religious importance did not automatically translate to high rankings in Assyria's urban heterarchies.Spatially, Assur was a minor settlement.By the time Nineveh was the imperial capital, Assur was the smallest city in which the kings had ever based their primary residence.The city was about 80 hectares inside its walls, which bounded the Inner City (including temples and palaces) and the New Town, a narrow residential expansion from the southeast corner of the citadel (Harmanşah 2012: 63-64; see Figure 1).In comparison to Assur, Kār-Tukultī-Ninurta was 240 hectares, Dūr-Šarrukīn was 315-320, Calah was 380, and Nineveh was 750.34 Assur was strategically protected by the Tigris to the north and east, but this limited its potential to expand outwards, as did the fact that the city sat on a rocky outcrop forty meters high, complicating the habitable landscape and water access (Harmanşah 2012: 63;Russell 2017: 424).Thus, for kings who wanted to enrich their city (and, thereby, their everlasting reputation) with building projects, Assur offered only its citadel, which was already crowded with historically important buildings.As a result, many of these structures were built, rebuilt, and sometimes replaced in a patchwork of construction projects over the generations.35In contrast, other cities could be expanded and shaped into royal capitals more freely, allowing kings more personalization.Dūr-Šarrukīn represents the apex of this desire for a customized capital, one that fully represents the fame and successes of its resident king.Sargon II built Dūr-Šarrukīn as his capital ex novo, shaping its every feature and naming the city after himself ("fortress of Sargon").Even existing cities were not limiting in the way that Assur was: Nineveh, for example, was an ancient center but still afforded Sennacherib a relatively flexible canvas.He was able to expand the city, adding numerous buildings (including his elaborate new palace), canals, massive city walls, an arsenal, and several temples (Frahm 2008).
While Assur remained geographically within the Assyrian core, the empire's expansion outwards slowly marginalized Assur from the perspective of agricultural and urban development.Assur's natural properties had never been strong, as its location -which had some irrigable areas nearby but was not in the rainfall zone -had a limited capacity for sustaining communities, whereas the areas of Nineveh and Arbela provided sufficient cultivatable areas and access to resources such as stone quarries, timber, and metal (Harmanşah 2012: 61;Russell 2017: 424;Radner 2011: 321).Outward expansion and organizing new 34 Van de Mieroop 1997: 95;Pedersén et al. 2010: 123;Radner 2011: 324-325, 327 territories into Assyrian provinces meant that these became the focus of state administration and development (Radner 2011: 327-329).With the changing shape of the empire, many other cities became positioned on trade routes that bypassed those connected to Assur.36 Assur's economic potential seems never to have been a central feature, and it is clear that economic benefits were a factor in moving the capital from Assur to Calah (Harmanşah 2012: 67).Other cities were better outfitted for military and defense, and the kings increasingly left on campaign from Calah, Nineveh, and Arbela rather than from Assur (Kessler 1997;Altaweel 2003;Yamada 2000: 259).
From the view of state religion, Assur remained consistently high in rank.Shifting focus to other Assyrian cities and newly conquered areas that were better provisioned and positioned, however, meant that Assur's most relevant features were progressively reduced to those that could not be substituted by other places: its antiquity, its traditions, and its historical significance as the location with the most ancient temples and the Old Palace, which housed the kings' tombs.37Kings routinely stayed in or traveled to Assur because of duties related to these institutions, such as ritual obligations, rather than for political or economic reasons.38Even when kings had religious responsibilities in other cities, Assur was still seen as the most prestigious in this respect.For instance, despite Ashurbanipal's particular veneration of Ištar of Nineveh and Ištar of Arbela, Aššur's role in kingship ideology was essentially unchanged (Porter 2004).
While Assur may have remained prominent and relevant within the local core, other cities became more important to empire making.As such, the most dramatic changes in rankings relate to political status: Assur was the political capital for most of Assyria's existence, with relocation of the king's main palace to other cities constituting brief interludes until more definitive moves away took place during the Neo-Assyrian period.Nonetheless, Assur had been in heterarchical relationships with other political and religious centers since its inception -for example, in the Old Assyrian period with the trading colony in Kārum Kaneš as well as with Mari, Šubat-Enlil, and ideological rival/model Nippur.39The network of relationships is particularly significant in the Middle and Neo-Assyrian periods when the designation of royal capital was moved back and forth between Assur and first Kār-Tukultī-Ninurta and then Calah, then with finality to Dūr-Šarrukīn, after which point it never returned to Assur, relocating permanently to Nineveh under Sennacherib.40This last change may have resulted in an exceptional circumstance in which Esarhaddon purposefully took the throne in Nineveh, not Assur, to establish his legitimacy after his father's assassination and the resultant succession struggle, so specific situations could also change longstanding priorities.41 The intensity of royal investment necessary to turn a city into a capital meant that significant resources were diverted for this purpose.While Assur probably did not suffer like other cities did -the royal correspondence under Sargon II reveals how other cities were made to contribute significant material and laborers to Dūr-Šarrukīn (Parpola 1995) -the capital city was often a higher priority when it came to active royal patronage.In addition, the more time a king spent in any one city, the more it affected the networks of people and communication within the empire.Moving the capital had profound ramifications for demography, especially in Assur, which lost its place as the primary residence of the king and his administrative apparatus and its status as where the old, elite families lived (Radner 2011: 323-325).Anyone seeking an audience had to travel to the city in which the king resided, and many members of the royal family's retinue (for example, advisors, bodyguards, doctors, courtiers, scribes, musicians) would have moved in tandem with the family member with whom they were associated.The consistent presence of the king and his family in a city necessitated infrastructure such as extensive palatial complexes (Kertai 2015), proper fortifications, garrisons, arsenals, and institutions such as living and social spaces in or near the palace,42 including those for educating crown princes (the "House of Succession") and other royal family members (Zamazalová 2011).When Ashurnaṣirpal II made Calah the political capital, the city even became the location where deceased queens were buried, with tombs of contemporary and later queens (ninth-eighth c. bce) located under the Northwest Palace that he built (Hussein 2016;Kertai 2013) members (Kertai 2013: 114).Thus, changes to demography, institutions, infrastructure, and even traditions influenced a city's position in the larger urban networks of the empire.The addition of Babylonia to the Neo-Assyrian Empire meant that its capital at Babylon further expanded the complicated network of urban spaces under Assyria's aegis.43Simultaneously a political and religious center, Babylon was the top of the urban and cultic hierarchies of southern Mesopotamia during the first millennium bce (Beaulieu 2018: 161-164).Thus, when Babylon was integrated within Assyria, it added another significant axis to the urban network.Starting with Sargon II, Assyrian kings attempted to bring Babylon into Assyria's political and religious relationships.Babylon's patron, Marduk, and his son Nabû were more clearly incorporated into the state pantheon, the latter possibly acting as the patron of Dūr-Šarrukīn.In turn, a temple to Ištar of Nineveh, an Assyrian goddess particularly important to kingship, was established in Babylon (Da Riva andFrahm 1999/2000;Esarhaddon RINAP 4 48: rev. 92-93).Because kings held the office of "king of Babylon" in addition to that of Assyria, Babylon became a secondary capital: kings established their royal residence in the local palace; invested in the city through building projects; patronized its temples; and performed state rituals, including the most prominent public ritual, the akītu (Porter 1993).Assyrian royal inscriptions in Babylonia greatly reduced Assur's status, fronting the king's commitment to Babylon and Borsippa in an effort to win over the Babylonians, for whom Assur was unimportant across every metric.44Indeed, when Sennacherib sought to lower Babylon's standing in Mesopotamia, especially in relation to Assur, he targeted exactly the features that made Babylon eminent: he destroyed the walls, temples, and monumental buildings; he adapted and promoted the akītu for Assur, rewriting the Enūma eliš to substitute Aššur as the hero instead of Marduk; he discontinued the akītu in Babylon; he removed Marduk and Nabû from his royal ideology; and he refused to take the throne, leaving Babylonia kingless (Beaulieu 2018: 206-207;Pongratz-Leisten 2015: 416-426;Frahm 2010).During Sennacherib's reign, Babylon disappeared from the rankings of cities from the Assyrian royal perspective, de facto elevating several otherwise minor Assyrian cities.Nonetheless, because Sennacherib neglected all other Babylonian cities equally, Babylon remained at the top of its regional hierarchy, which did Zaia Numen 70 (2023) 598-624 not change from the local perspective.The ranked relationships between Assur and Babylon were modified yet again under Sennacherib's successor, Esarhaddon, who made Ešarra the birthplace of the Babylonian deities, tying the two centers (and their gods) closely together from a religious perspective, though Assur maintained a slight superiority.45

Conclusions
As demonstrated above, heterarchy theory can reveal the nuances of Assur's religious relationships both internally (those contained within the city) and externally (those in which the city was contained) by isolating the diverse possible configurations generated when applying various organizing principles to the elements in the system.Each set of rankings produced by the criteria presented resulted in a different image of Assur's religious landscapes and the ways in which its inhabitants may have perceived them.This diversity of relationships can easily remain invisible in the sources when using traditional philological methods, and heterarchy theory helps us not only to navigate the complex primary data but also, and perhaps most importantly, to accept that these various rankings were all simultaneously applicable and valid within a city's living profile.The overlapping ranking systems assembled above create a robust composite image of individually studied sets of relationships that might otherwise look like a nonsensical and conflicting jumble of characteristics relating to topography, demographics, political and religious status, resources, size, age, and so forth.For Assyrian religion, therefore, heterarchy theory provides a way to access a more comprehensive understanding of the complicated and intersecting relationships between gods, cities, institutions, and the people who lived in and among them.Overall, heterarchies can be a powerful tool for embracing the complexity of ancient Near Eastern culture and society and discouraging the tendency to seek out relationships that are linear, static, organized, or mutually exclusive.It is not that we no longer observe the existence of hierarchies or rankings, but that we acknowledge that they are relational, based in their contexts and in conversation with other, coexisting hierarchies.Multiple deities, temples, and cities could occupy the highest tiers of separate, simultaneously existing rankings, and this would not have caused any issues for the ancient peoples living within these systems.
. After the move away from Calah, the Northwest Palace became primarily a storage facility and administrative center, but not one actively used by royal family 42 See Svärd 2015: 109-127 for female royalty, musicians, and other professionals, and Kertai 2013: 117 for the queens' suites that were not residential but "used for banquets, receptions and state activity related to the queen's considerable economic holdings and related administration."Downloaded from Brill.com09/21/2023 11:26:07AM via free access Numen 70 (2023) 598-624