• 1 Sthran 2015, 35-42; Lanz 2014.
• 2 Rüpke 2016.
• 3 Zuiderhoek 2017, 20-36.
• 4 Lefebvre 1991 (1974), 2003 (1970), 15; Soja 1989.
• 5 The project is directed by Susanne Rau and Jörg Rüpke and funded as FOR 2779 by the German Science (...)

1To date, religion has been seen as cause for dramatic developments in the history of cities: foundations, waves of immigration, transformations, ghettoisation, and destruction; it has been a decisive factor in forming the concept of citizenship as well as in justifying the expulsion of large groups, and has contributed to the monumentalisation of centres and or has given importance to ex-centric places. Very recently, sociologists and anthropologists have been discovering religion in the contemporary and supposedly ‘secular’ global city.1 But still awaiting historical investigation is the specific urban character of religious ideas, practices, and institutions and the role of urban space shaping this very ‘religion’. As the time-span from the Hellenistic age to Late Antiquity has proven to be crucial in the establishment of concepts and institutions of ‘religion’, on the one hand,2 and is a period of re-newed and more extended waves of urbanization, on the other,3 the Mediterranean basin and the Roman Empire offer a rich and comparatively well-documented space for pursuing such an investigation. Taking seriously the proposition that space is condition, medium, and outcome of social relations,4 the development of ‘religion’ in lived urban space as ‘urban religion’ acknowledges processes of religious change that have been neglected by both the history of religion and the study of ancient urbanism. Our key thesis is that city-space engineered the major changes that revolutionised Mediterranean religions, where this space is understood as the array of urban spatial and structural forms, forces, processes, and agents. Evidently, such a claim is beyond proof in the form of an article. Thus we limit ourselves to plausibilizing this claim as a basis for future investigation against the background of actual research and to developing the conceptual tools for such an enterprise. Thus, we can only indicate lines of research, some of which we will pursue in the near future within the larger comparative framework of a newly established Centre of Advanced Studies in the Humanities on ‘Religion and Urbanity’, based at the Max-Weber-Kolleg of the University of Erfurt.5

• 6 See Bowersock 1990; Markus 1990; Schwartz 2010. A new journal, edited at the Max Weber Centre, is p (...)
• 7 McDannell 1995; Orsi 1999, 2005; McGuire 2008.
• 8 Rüpke 2011c, 2016; Albrecht et al. 2018.

2The history of ancient Mediterranean religion is undergoing a major re-evaluation. All in all, the range of sources regarded as important for reconstructing ancient religion and the developments leading to the formation of those ‘late ancient religions’ that still are major factors in today’s world, such as Judaism, Christianity, and Islam,6 all in their many varieties, has been enormously enlarged. For understanding the complexity of religious interaction, the concept of ‘lived religion’7 as developed for contemporary religion has been adapted and enlarged by the ERC-Advanced Grant project ‘Lived Ancient Religion’ for the study of pre-modern and specifically ancient Mediterranean religious practices and ideas.8 It has showed that funerary ritual and domestic religion, the social and ritual practices of voluntary associations, and the unanticipated, non-civic political use of religion by administrators and political elites are neither independent strands of religious practice nor replications of, or counter-models to, ‘civic religion’. Religion is best conceptualised as a single field of action and forces with many loci of religious authority in permanent fluctuation.

• 9 E.g. Degelmann 2018; Patzelt 2018; Rüpke 2018.
• 10 See the studies in Yoffee 2015a – except Sinopoli focusing on religion in cities dominated by compe (...)
• 11 Orlin 2010; Woolf 2011; Mattingly 2011; Eshleman 2012; Ando 2013; Bricault, Bonnet 2013; Collar 201 (...)

3The relationships of religion and space, in general, and between religion and urban space, in particular, are still under-theorized and under-researched. Although practices and discourses previously framed in terms of ‘polis religion’ have been successfully integrated into the description of ‘lived ancient religion’,9 the city itself as a living habitat crucial to the religious practices of antiquity and as the driving force of religious change has never been a major concern of research. In almost all research on cities in the deep past (i.e., prior to the late medieval and early modern period), it is mainly assumed that the task is to illustrate how the viability of the city is grounded in a religious identity that is by the same token also a political one.10 As scholars’ focus has been on political and civic identities, in the history of ancient religion space has been addressed rather reluctantly and above all as imaginary realm. Only very recently has attention been paid to the Roman Empire as a conceptual space of identities and a lived space that informed actual behaviour.11 Yet the dominant models remain synchronic and functional with regard to religion or diachronic and hermeneutic in relation to urban space.

• 12 Soja 2000, 8; 2006.
• 13 Sassen 2010; Berking, Löw 2005.
• 14 Soja 2000, 7-10. On this link see Berking 2008, 15.
• 15 de Certeau 1984 (1980). 91 ff.
• 16 Löw 2011, 19-21. Also Roskamm 2011; Berking 2008; Nassehi 2002.

4It is our contention that the empirical settings of ancient cities are to be viewed and studied specifically as city-spaces,12 that is, not only as a built environment as a physical condition, but as a space apprehended as challenging and inspiring, enabling and constraining, featured by particular configurations of social relations, built forms, and human activities which distinguish it from other spatial formations. A paradigmatic site for examining processes of social change at the time of the founding fathers of sociology and urban studies (Weber, Simmel, and the Chicago School), the early 21^st-century re-emergence of the city as a distinct scientific object of social research13 has largely benefited from critical spatial theory (aka spatial turn) and its emphasis on the ‘spatial specificity of urbanism’.14 Different disciplinary insights now concur in saying the cities are peculiar spatial constellations. Since Michel de Certeau’s seminal ‘Walking in the City’,15 urban sociologists, who rarely take anything into account prior to the modern period, have emphasised ‘lived’ urban space as a praxeological and thus ultimately dialectical relation between the urban space and the human agent, who perceives and appropriates urban space on an individual basis. Scholars of urban planning, social ecology, geography and (political) sociology preferably define cities as phenomena of density with regard to the concentration of people and buildings at certain spots in broader landscapes and to the social consequences of an extreme increase of contact zones.16

• 17 See Soja 1996, Foucault 1986 (1967), and Lefebvre 1991 (1974), respectively.

5Drawing critically upon these theoretical insights, our project surveys ancient city-spaces by concentrating on both real and imagined loci of religious experiences and agency. Imagination beyond concrete, measurable space, of course, is no privilege of the subalterns and agency is no prerogative of the dominants: against the counter-cultural and utopian accent of the typical use of ‘Thirdspace’ and ‘Heterotopy’, we prefer the less normatively charged notion of ‘lived space’,17 without denying that the very nature of religious practices implies the imaginative construction of alternative space beyond any given physical location.

• 18 See Knott 2005; van der Veer 2015; Knott, Krech, Meyer 2016. More literature will be provided below
• 19 Day 2017, 3.

6The very recent rise of the study of religion in contemporary cities focuses on the development of new forms of religious practices, i.e. religious change, and the appropriation of urban space by non-elites.18 Our research combines these insights with recent developments in the study of ancient lived religion that acknowledge the plurality of religious agencies and the complexity of religious interaction and appropriations. It aims to demonstrate that many religious phenomena, and especially major religious changes, can be better understood by viewing them in spatial terms, that is, as a result of a spatially informed dialectic of ‘co-production’19 of urban life and religious communication in the ancient Mediterranean world from the Hellenistic period to the late imperial period. Focusing on the impact of cities on religion and how the interaction with city-space changed religion, we call this side of the dialectic ‘urban religion’.

• 20 See most recently Norenzayan 2013.
• 21 See most recently Yoffee 2015b.
• 22 E.g. Sourvinou-Inwood, C. 1990; Cole 1995; Hägg 1996; Sartre 2006; Kindt 2009; Bruun 2009; Evans 20 (...)
• 23 Crumley 1995.

9Planetary urbanisation is one of the key developments of the present world that cause immense changes in people’s daily lives as well as to the landscape of global networks. Praised as the solution to, or cursed at the embodiment of, the major problems of humankind, the ‘urban question’ is one of the grand challenges of today. Research on Mediterranean Antiquity cannot escape but rather contribute to this discussion by setting ancient religion into a wider framework and by showing that developments in antiquity are relevant even for contemporary concerns. By changing the approach, the outcome of such an agenda will affect claims about the evolutionary role of religion, which is too often viewed in terms of key provider of social cohesion20 or examined in ways largely neglectful of individual agency.21 Instead of focusing on political integration and social stratification as functional foci of a ‘polis’/‘civic religion’,22 we aim to stress urban aspirations, the appropriation of urban space, and the creation of urban diversity. Aside of hierarchy, we attempt to emphasize ‘heterarchy’.23 Instead of urban trapping and central administration, we highlight network activities. As research on ‘Lived ancient religion’ has demonstrated, such results are highly rele­vant to contemporary societal challenges.

• 24 Lanz 2014, 26. For a similar definition, see Orsi 1999, 43. For a contextualization, see Garbin, St (...)
• 25 See Orsi 1999; AlSayyas, Massoumi 2011; Garbin 2012 and 2013; Becci, Burchardt, Casanova 2013; Stra (...)

10The critical use and the establishment of the concept of ‘urban religion’ are paramount for this task. Defining a ‘continual process in which the urban and the religious reciprocally interact, mutually interlace, producing, defining, and transforming each other’,24 urban religion is the flagship notion of an cross-disciplinary approach to the study of religion in contemporary and preferably global cities that has eventually proven the role of religion in shaping allegedly secularized urban city-spaces and vice versa. Moreover, by significantly enlarging the range of the agents and the motivations involved in the creative co-constitution of religion and urbanity, this body of researches has increasingly made aware of the thoroughness of this interaction.25

• 26 Albrecht et al. 2018.

11As demonstrated by the previous research project on ‘lived ancient religion’, religion is always ‘religion in the making’26 and thus ‘urban religion’, too, is a dynamic concept focusing on change. It is inclusive and, when tested on antiquity, it permits analysis of the development of specific religious agency and practices (neighbourhood shrines, theatrical processions; text production and supply of religious services), specific forms of religious knowledge and imaginaries (imaginative places, imagined communities, heavenly cities) and societal phenomena such as civic ritual or religious communities in the appropriation, modification, and formation of urban space. Religious change is always investigated in the ongoing interaction between space and different agents: temporary inhabitants and voluntary or involuntary immigrants, residents and administrators, people living off religion or employing religion for realising their urban hopes.

• 27 Cancik, Rüpke 2009; Rüpke 2011a, 2011b, 2014. On translocality see Freitag, Von Oppen 2010.
• 28 Garbin 2012, 2013.
• 29 Holleran 2011; Woolf 2015; Tacoma 2016; Tacoma, Lo Cascio 2017; Lo Cascio 2017.

12Yet, in its dominant use on contemporary capitalist urbanism, the object of urban religion is not religion but globalization – with religion rather serving as a lens onto globalization. Therefore, the enquired interlacing of globalization, cities, and religion has hardly opened up a real space for historical research, even though historians of religion have been willing to employ this concept for pre-modern phenomena of translocality, universalization, regionalization, localisation and their interactions.27 By critically applying the notion of urban religion to ancient Mediterranean cities and religions, our ambition is precisely to thoroughly historicize the analysis of the range of phenomena illuminated by this conceptual tool. For instance, delving into a typical subject matter of the urban-religious research such as ‘migrant faiths’ and ‘diasporic identities’,28 we aim at understanding the role of religion for migrants and immigrants into cities29 by asking whether participation in specific religious activities or networks was a means of stabilising former identities within a new urban space, or rather a strategy for pursuing urban aspirations and developing new identities. This is only possible by strengthening the diachronic scope and reach of urban religion.

• 30 On the contemporary explosion of urban/rural divide the literature is legion, but see, e.g., Soja, (...)
• 31 See Jacobs 1969, 18.

14a) Human Geography in general and studies in ancient urbanism in particular30 have demonstrated that there is no easy dividing-line between ‘city’ and ‘the rural’ – the latter can even be seen as a response to the need for an antithesis to the urban as a result of urbanisation processes. Many religious practices have been conclusively shown to antedate urbanisation, from the basic forms of communicating with divine others in dance, voice, images or gifts to complex rituals like animal sacrifice and moving in groups (processions). Our approach addresses this problem by a basic research strategy that does not attempt to develop criteria for infallibly distinguishing between urban versus non-urban practices but asks whether changes over time in a set of religious phenomena well attested in cities can fruitfully be interpreted as consequences of their contextualisation in and interaction with city-space (what we propose to call ‘citification’ rather than ‘urbanization’ of religion). Moreover, instead of drawing an unfoundedly clear-cut religious border, we are interested in monitoring and interpreting the ongoing traffic of religious signs, carriers, practices, and institutions across a more or less externally demarcated and discernible city border, thus testing their endurance and changes under different environmental conditions. Inspired by Jane Jacobs’ unrivalled urban imagination, we are looking for cases where rural forms of religion can be explained ‘as urban [religion] transplanted’.31

• 32 Allen et al. 2016.
• 33 Spickermann 2008.
• 34 See Robinson 2017 against the so-called ‘Urban Thesis’ (especially, 14-23).

15Additionally, some well documented rural areas like the ‘dead villages’ in Syria, rural Britain32 or the region of the Ubii33 will be used as control cases. Lastly, since the formation of Christianity has been long central to debates about the supposedly ‘rural’ or ‘urban’ character of a religious tradition, these still most recent discussions34 offer an important touchstone to sharpen our methodology.

• 35 These questions can be mapped onto a more general ambiguity of ‘urbanity’/’urbanism’: see Wüst 64-6 (...)
• 36 See, e.g., Lee 1997; Lindner 2003 (habitus); Berking 2008 (doxa).
• 37 See Berking, Löw 2008; Löw 2012.

16b) A second major risk is the potential diversity of local trajectories. Metropoleis such as Rome, Alexandria and Antioch on the Orontes cannot be ignored due to the wealth of material and textual sources they offer, but they are as exceptional as Baghdad, Cairo or Kaifeng in later periods. Were the crucial religious changes that are often associated with these centres, when seen through the lens of spatiality, singular local events? Were they rather connected to the metropolitan size of the city-spaces? Or were they building on more widespread patterns of ‘urban religion’ as defined above?35 Moreover, drawing on Bourdieu’s pivotal categories of practical sense (habitus, doxa),36 recent sociological research has developed the concept of ‘intrinsic logic of cities’37 to explain why reaction modes and times to analogous societal challenges and impulses can differ considerably from city to city, even under identical conditions in terms of financial possibilities and institutional frameworks. The possibility to transfer such contemporary explanatory repertoire to ancient city-spaces, and thus eventually to look for ‘elective affinities’ between certain religious innovations and some specific cities – was a successful religious implementation and religious-branding also the result of a good entrepreneurial sense of the right urban atmosphere? – is fairly limited by the general quality and the uneven quantity of the documentation at hand.

• 38 See critically Raja, Lichtenberger 2015.
• 39 This was the topic of one of the preparatory conferences of the research project entitled ‘Religion (...)
• 40 Fuhrer, Mundt, Stenger 2015; also partly Kemezis 2015.

17In order to circumvent or limit this risk, our intention is to bring together local case studies selected for historical analyses of a broad range of religious practices, imaginaries, institutions, which pay close attention to the push and pull effects of urban contexts, spatial as well as social and cultural. Granted that many ancient cities have been only partially excavated and even these results only partially published,38 we consider that sufficient reliable information has been accumulated to render the project feasible. Research needs to investigate religious change in cross-regional comparisons of cities of different size and character. For instance, in accordance with the concept of ‘lived city-space’, a line of investigation should focus on the agency of very different types of urban space, namely, on the one hand, dense living quarters,39 and, on the other, venues that were built in order to accommodate and communicate with large crowds within a framework of religious rituals, namely circuses, theatres and amphitheatres. This is to be combined with analyses of how human agents adopted antagonistic strategies of appropriating urban space for purposes of attracting (and keeping) religious clients or followers, namely by firmly localising religious practices or developing movable forms of religious practices and imaginaries adaptable to a multifarious urban space. In addition, the interplay between ancient imagination of cities (‘cityscaping’)40 and spatial conceptualisation of religion as a trans-local literary discourse needs to be investigated. A number of cities in the Western and Eastern parts of the Roman Empire offer sufficiently well published material (topographical data, architecture, coins, images), epigraphic, and occasionally even literary evidence to allow for comparison with regard to specific questions.

• 41 Including incipient phenomena of de-urbanisation in Late Antiquity. See Rizos 2017.

18c) If the geographical extent of exchange relevant for religious signs, practices, ideas and even people, and likewise the impact of urbanisation triggered by the Mediterranean centres, militate against any narrow territorial limitation, the chronological range likewise matters. On the one hand, the bold claims of our research project require that we transcend narrow periods and periodisations. On the other hand, urbanisation (and religion) in the Mediterranean basin and the ‘hilly flanks’ dates as back as the late, if not the mid-Neolithic period; the empires of the Ancient Near East gave religion a central role in their cities; monumentalised temples and religious festivals (and other religious phenomena) figure prominently in the archaic and classical Greek and Italic poleis. We cannot endeavour to cover this entire range of potential topics in the detail that would be required. In effect, we decided to limit ourselves to the time-span from the Hellenistic age to Late Antiquity, which has proved to be crucial in the establishment of more developed concepts and more institutionalised forms of ‘religion’. At the same time this is a period of renewed and more extended waves of urbanisation.41 The residual risk of overstretch will be reduced by addressing only phenomena essen­ti­al from the point of view of spatial theory as well as history of religion as outlined above and substantiated below. The focus is on religious rather than urbanistic change.

• 42 Already in 1950, the Australian Marxist archaeologist Vere Gordon Childe opened one of the most hea (...)

21For the purposes of the project and against the background of the state-of-the-art,42 we prefer to avoid clear-cut and highly technical definitions of city valid only for specific cultures. Very large conceptualizations are no heuristically serviceable either. Nor is our intention to draw on existing typologies or construct new ideal types to be added to an already thriving taxonomic panorama. Rather, we opt for a polythetic approach that deliberately selects out characteristics which are of prime importance for our research questions and thus understands cities as places with the following features :

• 43 Rüpke 2015.
• 44 Knott 2005.
• 45 For this tension, see Tweed 2006.
• 46 In this sense, the trans-local references inherent to religious communication by way of agency clai (...)
• 47 Insoll 2009; Droogan 2013; Raja, Rüpke 2015.
• 48 On this see most recently Lander 2017.

22For the purposes of the project, we theorise religion as communication with special agents (sometimes including objects) with properties different from everyday human, that is, dead (ancestors) or unborn (angels), just (demon) or fully superhuman (gods) agents to whom agency is accorded in a not unquestionably plausible way. Communication with or concerning such divine agents might reinforce or reduce human agency, create or modify social relationships, and change power relationships.43 This is of particular interest with regard to the frequent encounters and dense networks but also fluid and exchangeable relationships (‘weak ties’) typical of cities. Fur­ther­more, like any other cultural practice, religious communication is a spatio-temporal practice induced and shaped by the spatial organization and, in turn, re-creating space.44 At the same time, our conceptualization of religion suggests that there is a specific spatial character of religious communication, a conceptual relationship not likewise valid for other cultural practices. If place-making can be equated with ‘dwelling’ and is frequently achieved with religious practices, religious communication is inherently also a practice of ‘crossing’.45 ‘Religion’ as used here is defined as an action transcending (in a very simple sense) the immediate and unquestionably given situation, that is, temporarily and situationally enlarging the environment judged as relevant by one or several actors.46 In the an­cient Mediter­ra­n­ean world, just as today, the spatial character of religious communication was reinforced by sacralising objects or places and was thus manifest in material form even in other uses of space.47 Creat­ing religious space was part of an ongoing process of claiming and appropriating urban space which implies specific competitive dynamics of place making involving both physical and rhetorical strategies.48

• 49 Gosden 2005, Hodder 2012.
• 50 Joas 1996, Emirbayer, Mische 1998.

24a) Agency and aspiration. Agency can be afforded to objects. By their form or very presence they make humans re-act, ‘afford’ certain behaviour.49 We use the concept of ‘agency of objects’ and places but likewise focus on human agency and creativity within social contexts and situations. We conceive agency as the constant interaction between people that creates the structures and traditions that constrain and limit the subsequent exercise of agency, which in turn may alter or even challenge those same structures and traditions.50 As outlined above, religion enlarges the field of agency. By invoking in specific situations agents or authorities held to be divine, religious agents acquire extended possibilities for imag­in­ing and acting. Yet the converse is also possible: the same mechanism can also trigger an abjuration of personal agency, resulting in impotence and passivity, with agency being reserved for the divine agents.

• 51 Appadurai 2004.
• 52 van der Veer 2015; Goh, van der Veer 2016.

25A specific aspect of agency related to city-space is captured by a further term. Recent urban studies have taken up the term ‘aspirations’ from studies of social mobility51 to designate specific driving motifs and attitudes of immigrants as well as homegrown inhabitants: namely, ‘forward-looking hopes of achievement’ instantiated in ideas and behaviours connect­ed with urban life and mobilized in ways that ‘engage and overcome’ urban issues. The situations in which religiously infused motivations stir urban creativity or create alternative urban worlds, or urban concerns are conducive to religious innovations, are described by the dialectical notion of ‘urban religious aspirations’.52 One of the aims of this project is to build on and develop this rather vague concept on the basis of historical research, at the same time exploring the possibility of a mutually exclusive relationship between religious and urban concerns. The socially uneven, though universal, cross-cultural, and –temporal capacity to engage agentically with the urban environment and to aspire in relation with it is to be used to nuance the analysis and description of individual and collective agents in structured situations.

• 53 See already Lynch 196.
• 54 Lefebvre 1991, 38-39.
• 55 Blum 2003, 13-20.
• 56 Lefebvre 1991, passim.

26b) Spatial imagination. Lefebvre’s notion of ‘lived’ space has taught us that agents do not merely ‘perceive’ or ‘conceive’ urban space.53 Whereas perceived space describes spatial practices that reproduce a spatial order (in the case of the urban, by mapping daily routines onto the established blueprints of everyday urban reality), and conceived space refers to the intellectually worked out dominant conceptions of space (e.g., that of the urban planners, social engineers, and administrative authorities), lived space highlights the human aspirational capacity to imagine space differently, to overlay it with unanticipated systems of symbols and signs, in a word: to change its use and appropriate it.54 Accordingly, urban-based human agents develop alternative ‘representational spaces’, new ways of using the city-space and, moreover, new forms of memorising, symbolising, imagining, and transcending urban spaces beyond their commonsensically perceived and professionally designed spatial layouts – including fictional spaces, theoretical spaces, or dreamed/utopian spaces as modelled by philosophers, poets and religious specialists. The city, thus, reveals itself as ‘an imaginative object’, an ever-changing assemblage of views that in turn conditions particular spatial practices.55 Already highlighted by Lefebvre,56 the role played by religion in producing lived space is foregrounded by considering religious imagination and practice as spatial imagination and practice that do not just reproduce or legitimize existing knowledge already attached to a location but actually creates new interpretations, meanings, and uses by drawing upon extraneous religious knowledge. At the same time, the users’ practice and the knowledge of the city produces and affects lived religion. Hence the concept of spatial imagination helps to elucidate key dynamics of the dialectic between urban space and religious practice and knowledge in our evidence.

• 57 Ashley, Plesch 2002, 6.
• 58 de Certeau 1984 (1980), XIV.
• 59 Carrier 1995, 107-126.
• 60 de Certeau 1984 (1980).
• 61 Carrier 1995, 117-120 (on the gendered quality of appropriations).

27c) Appropriation. Over the last decades, postcolonial and subaltern studies have rehabilitated the hitherto negatively charged notion of appropriation (i.e., to make one’s own) by explaining it as ‘a potentially two-way process’ where ‘exchange and creative response may take place’.57 In its far less controversial sociological use, which we mainly refer to, the term describes a specific transformation process of the relations between human agents and human-made products (language, things, space) whereby the former agentically adopt and adapt the latter by gaining power and/or stressing identity via and over them. Appropriation can thus designate the ‘innumerable and infinitesimal transformations’58 of a language by the speakers, the creative transition of a thing from being an anonymous commodity to a highly personal good,59 the tactical interpretation of the city-space by the footsteps of its walkers. In de Certeau’s seminal praxeological analyses, the frequent use of the term always implies a focus on the acts and practices of the consumer rather than on the producer, or better: on the proactive and productive character of consumption.60 This logic shuffles (and, for its critics, blurs) the hierarchical directionality of cultural processes (designer/public; active/passive; etc.) in a manner that resonates with the nowadays’ fashionable business term of ‘prosumer’. Successful appropriations emphasise the multidimensionality and polysemous nature of signs, things or spaces, but they are conditional upon the decisions and requirements of the agents themselves in concrete situations, as well as upon their social roles and the related expectations.61