Goats and Goddesses. Digital Approach to the Religioscapes of Atargatis and Allat

1 Introduction

Last decades showed the necessity of ordering in a database the large epigraphical evidence, very often dispersed in the publications (PHI database of Greek inscriptions: https://inscriptions.packhum.org, database CGRN of the Greek ritual norms: http://cgrn.ulg.ac.be, database of MAP Project: https://base-map-polytheisms.huma-num.fr). The studies of diffusion and networks within the history of ancient religions (Brughmans, 2013; Collar, 2007, 2013; Glomb, 2021; Mazzilli 2022, pp. 166–171) paid attention to the aspects of mobility of religions, cultural transfers of concepts, and mobility of people and visualized them on maps and graphs for a better understanding of humanity in the past. These studies showed how the ancient world is intertwined and connected in tempo-spatial dimensions (Barker et al., 2024). Gathering the epigraphical and papyrological data in a relational database allows us to see better the patterns of the mobility of the people and the gods (Bonnet et al., 2018, p. 585; Collar, 2013; Williamson, 2021). People on the move transport their gods, adopt new cultural trends, and within the polytheistic religions, honour also the gods of others (Bonnet et al., 2018, p. 568; Parker, 2017, pp. 75–76). The entanglement between the space and social realm is the background of the concept of the religioscape (Hayden & Walker, 2013, p. 407). “Religioscape is understood as the relationship between physical spaces, buildings, societies, groups, individuals and geographical locations and diffusion over time and throughout the regions” (Kilde, 2019, pp. 165–166). This concept is essentially based on the spatial and temporal distribution of religious traditions and the people who created them. Religioscape is more than the religious landscape, identified mostly with the spatial distribution of the sanctuaries. Religioscape is compliant to the changes as people bring their gods and religious practices with them, moving to new places. The research proposed in this work, within the project Al-At. People of the Gods. Worshippers of Allat and Atargatis in the Near East and beyond from the fourth century BC to the fourth century AD, focuses on the social and geographical spaces of worship of two goddesses of Eastern origins. A database of 171 inscriptions was created to study the mechanisms of diffusion and the social strata involved in the cult. Concerning the cult of Atargatis, we counted 146 inscriptions. The cult of Allat is represented only by 25 inscriptions, collected within 2 years of the duration of the project.^[1] The chosen diachronic timespan, from the fourth century BCE to the fourth century CE, provides an insight into the most vibrant period of change in the Near East and the Mediterranean. In this period, we can trace the development, transformations, and spread of cults in the multicultural setup. The lower limit of fourth century BCE is explained by the earliest appearance of the Aramaic name ˀtrˁth of the goddess on the legend on the coins from Hierapolis-Bambyke (modern Manbij in Syria), despite the numismatic sources do not enter to the database. The upper limit of the fourth century CE is justified by the engagement of Roman soldiers in rebuilding the temple of Allat in Palmyra (Gawlikowski, 2017, p. 155).

Within the research described in this work, we map the ethnicities of the worshippers, the involvement of “foreigners” in the cult, and looking at the networks and gender of the people engaged in the cult to provide in the end the comprehensive analysis of the religioscape of the two goddesses. The research is oriented on the inscriptions, people, and places with the help of the digital environment offered by the NodeGoat program.

2 Digital Approach to the People of the Gods: Methodology

2.2 Structure and Content of the Database

The database is composed of two main “types”: inscriptions and people, and two additional ones: sanctuaries and media, the latter will not be discussed within this work, because it serves only as a storage for the pictures of the inscriptions and inscribed objects (Figure 1).^[2] The types are the tables where the data are gathered. They contain “objects,” e.g. in the type “People,” these are names of the persons mentioned in the gathered inscriptions. The NodeGoat space gives also the insight to the type “city,” so all the places with the geo-coordinates are inserted within this and other projects depending on the NodeGoat system. Each “object” is an own spreadsheet, but by looking at the data, there is a relation between them. Just to provide an example, our first object named Inscriptions is in relation with the names of the actors by mentioning them in the text of the inscription and cities by location and place mentioned in the text.

Figure 1

Structure and relations of the database (graphic: NodeGoat, modelled by A. Kubiak-Scheider and S. Mazurek).

The objects are divided into the sub-objects, which permit us to insert the spatial and chronological data, necessary for visualizing the collected material on the map and on the chronological curve.

Our database also contains five classifications, i.e. categories which can establish the connection to the object and label it. Hence, in this context, we created among others the category “capacity” to define the profession and role of the person known from the inscription and category “deity,” which permits us to establish the relation between the inscription, actor, and the gods. The classifications appear in the social network analysis graphs as connections to the people and inscriptions. Another modelled classification is gender to look closely at the scale of engagement of man and woman in the cults.

2.2.1 Inscriptions

The type “Inscriptions” (Figure 2) is the basic source of the information about the people, their identity, the location of the inscription, deities, and chronology. This table is the centre of our data model and the starting point for networking. It gathers epigraphical evidence concerning the cult of Allat and Atargatis as well as related deities, where the relation is established by the place of discovery pointing to the sanctuaries of the goddesses. The objects inserted in this type are given the Working ID and the references to the corpora and databases containing the editions of the inscriptions. The objects are described by the type, language, text of the inscription, and human-made object (i.e. the artefact where the inscription was placed). The object description fields such as: “place of discovery” refers to the classification category “location,” “concerning” refers to the object type “people” and “is related to” connects to the classification category “deity.” Thanks to this modelling, we can make a visual analysis, particularly, it is helpful in the social network graphs (Figure 3).

Figure 2

An object within the type Inscriptions (graphic: NodeGoat, modelled by A. Kubiak-Scheider and S. Mazurek).

Figure 3

Social Network Analysis graph based on type Inscriptions (graphic: NodeGoat, modelled by A. Kubiak-Scheider and S. Mazurek).

“Inscriptions” also contain four sub-objects: datation with a point or period, site where the inscription was found, place mentioned in the text, which allows us to pinpoint all the places noted explicitly in the inscription, and deity which permits us to establish connection between the theonym and a place. The sub-objects site and place mentioned in the text help to visualize the data on the map, by the modelling using the geodata (latitude and longitude). The chronological data show when the cult of Atargatis or Allat appears in the written sources in a giving geographical location. The NodeGoat platform not only provides an opportunity to use the geodata once introduced, but also permits us to insert new city or GeoJSON geometry (good for regions, routes, etc.). For precise locating, references were also included through the Pleiades ID (an online gazetteer for places of the ancient world) and GeoNames ID (an online gazetteer for places of the modern world). Such an added location was the Temple of Atargatis, since the Pleiades provide the coordinates for this building thanks to the urban gazetteer of the IDEA project on Dura-Europos (https://duraeuroposarchive.org/duras-digital-gazetteer/).

2.2.2 People

This table consists of information on the people mentioned in the inscriptions, their gender, families, and professions in connection to the goddesses and other connected deities mentioned together with the goddess or in the sanctuaries dedicated to the goddesses. Here the description of the objects contains information, among others, about the family of the person (genealogical affiliation, optionally names of the siblings, children, when possible, in the field “family”), his or her profession (field “capacity”) which is referenced to the classification category “capacity,” ethnicity, and gender. The field such as “worships” is using the information from the classification category “deity.” The other description field “connected to” is related to the objects from the type “people.” These two relational fields are helpful to establish and visualize the relations on the network analysis graphs. We need to be aware that it is very rare to find the networks based on people due to the recurrent names in the epigraphy and often to the unique mention of someone. In this case, when we encountered several persons of the same name, e.g. Seleukos, we labelled them with a number in the brackets – Seleukos (1), Seleukos (2), Seleukos (3) to differentiate the people. The object type People has three sub-objects: location, connected to the actual find-spot of the inscription; ethnicity, the identity or identities of the person; and period related to the datation of the inscription. Here the data create thanks to the modelling of the sub-objects a map of people and their connections to the physical places. We often cannot estimate the life duration of the people in the ancient epigraphic material, that is why we chose as only criterion the datation of the inscription (Figure 4).

Figure 4

An object within the type People (graphic: NodeGoat, modelled by A. Kubiak-Scheider and S. Mazurek).

2.2.3 Sanctuaries

The third type concerns temples (Figure 5). These data are rather secondary; however, not less important. Two fields establish relations with the other object types and classifications: cities (for location) and deities who are worshipped within. Within this table, it is possible to insert photographs of the places.

Figure 5

Object-type: Sanctuaries (graphic: NodeGoat, modelled by A. Kubiak-Scheider and S. Mazurek).

When it concerns the cult of Allat, it provides a very important result for the location of the temples on the desert and arid zones only, pinpointing the meaning of the physical set-ups for the places of cult (Kubiak-Schneider, 2023, Mazzilli, 2022, pp. 174–175). Furthermore, it serves for mapping sanctuaries and can be potentially useful for the future studies of the cartography of the ancient temples.

3 Exploratory Approach to the Epigraphic, Religious, and Social Data: Preliminary Results

The preliminary results of the project described here picture the observation by Robert M. Hayden and Timothy D. Walker that the sites are not isolated but connected in a network by the communities of people in various settings (Hayden & Walker, 2013, p. 409). The number of inscriptions dedicated to Atargatis and those dedicated to Allat show a striking difference. Atargatis-Dea Syria counts 146 texts and Allat only 25. The inscriptions for Atargatis and associated deities are mostly written in Greek, while for Allat they are mainly in the various dialects and scripts (Palmyrene, Hatrene, Nabatean) of Aramaic. In the epigraphic evidence dominate dedicatory inscriptions, we also count four epitaphs and two tituli honorarii as well as sacred norms. The database provides information on the diverse context of these inscriptions which are not always attached to the sanctuaries of the goddesses. The big clusters concerning the inscriptional data to Atargatis comes from the sanctuaries in Delos (22 inscriptions) and in Dura-Europos (35 inscriptions). For Allat, it is Palmyra (12 inscriptions) and Wadi Ramm with Ain esh-Shallalah (19 inscriptions). It is interesting that Manbij, Syrian Hierapolis, noted at Lucian’s description on the Syrian Goddess as the cultic centre of Atargatis, is underrepresented in the epigraphic material as a direct findspot (IGLS 1.232 dated on the first century BCE). The reason for the poor epigraphic evidence from Syrian Hierapolis is that there is very little preserved of the sanctuary and that this area was never a subject of systematic archaeological excavations.

There are four epitaphs found in Brundisium, Rome, Niha, Pontes (Drobeta-Turnu in Romania), which attest the cultic personnel such as priests and a virgin priestess and prophetess of Atargatis/Dea Syria, but in three cases with the associated gods: Isis and Sarapis, Jupiter Dolichenus, and Magna Mater.

The tituli honorarii are assigned by different commissioners. One from Palmyra (nr 22 in the database) is commissioned by the city council (Boule), while another from Hatra is ordered by the individual person, labelled as priest of Atargatis to honour his wife.

The evidence concerning Allat shows that it stayed regional, attached to the arid zones and people of the East and rather did not go beyond this big region, at least what is visible in the epigraphical material. The texts mentioning Allat concentrate in the Near East, especially on the particularly defined places by the topography: desert and rocky areas (Kubiak-Schneider, 2023).

The cult of Atargatis starts to be diffused from the beginning of the Hellenistic Period – the earliest points on the map are Astypalaia, Pelusium and Fayum, Delos, and Beroia in Macedonia. They are dated on the Hellenistic period. The evidence from Beroia and Astypalaia is from the fourth century BCE. The earliest sanctuary dedicated to Atargatis in the Greek Macedonia (from the fourth century BCE) and of the longest function as the cultic place (still in the Roman Period used as the temple of Atargatis) is the sanctuary in Beroia. The chronological curve (Figure 6) shows that the inscriptions cover the period between the fourth century BCE and the fourth century CE, so it meets the project chronological objectives (Figure 6).

Figure 6

Chronological curve of the inscriptions for Atargatis and Allat (graphic: NodeGoat, modelled by A. Kubiak-Scheider and S. Mazurek).

Atargatis is widespread into different corners of the empires: first Hellenistic than Parthian and Roman. There are no inscriptions with Dea Syria from Northern Africa (territories to the West from Egypt), even though this region attests the cult of Syrian gods, e.g. Yarhibol, Malakbel, and Jupiter Dolichenus.

The database allows researchers to verify some information on the people and their professions in the orderly and graphic way. The type “people” concerns 218 objects constituted by the names from the inscriptions, some of them are fragmentary preserved like “-na daughter of Lysippos” from Dura-Europos. The data gathered in this table show a large number (190) of male actors connected to the goddesses. Female worshippers count only 28 attestations. This disproportion is fairly a common trend in the religious epigraphy in general.

From the gathered material there are no women among the worshippers of Allat. That can be a result of shifts in the position of women in different cultures and places in the ancient Mediterranean (Klaver, 2019; Kubiak-Schneider, 2021; Yon & Aliquot, 2018). From the data concerning people, we can follow the engagement of certain families, such as in the case of Ammonios and the members of his family in Dura-Europos (Figure 7). He and his descendants were active from 31 CE to 92 CE according to the epigraphic material. They are associated with the cult of Atargatis as responsible for the construction works on the temple and female members of the association gathering in a stepped room. The epigraphic sources are sileny about more precise information like on the profession of these women.

Figure 7

Network analysis graph of Ammonios from Dura-Europos (graphic: NodeGoat, modelled by A. Kubiak-Scheider and S. Mazurek).

Concerning the professions (here called capacities), we find in this context not only priests or other ritual specialists (e.g. afkals) but also soldiers and craftsmen (builders, sculptors) as well as kings and emperors. The epigraphic material concerning cult of Allat from Palmyra, beyond Safaitic graffiti, which are not included in the database, does not provide any mention of a priest of the goddess. The information about the priests of Allat comes from the sources from Hatra, Wadi Ramm, and Salkhad. The absence in Palmyra is strange though, because of the remains of the sanctuary of Allat, well attested archaeologically.

Among the worshippers of Atargatis, the evidence from Dura-Europos attests a male and a female cultic association, but their duties are not further described. The female association had within the temple a particular room with stepped seats where they were gathering. There are about 30 inscriptions on the steps of this room mentioning the female owners of the seats (Cumont, 1926). The male association was responsible for building some room for the goddess what indicates the inscription 46 (SEG 68.1496, see also https://www.wikidata.org/wiki/Q126737382 by A. Kubiak-Schneider).

Within the cult of Atargatis, we find a confirmation of engagement of eunuchs. The inscription nr 57 (CIL VI, 32462) is an epitaph of Caius Julius Abdera who is defined as gallus priest of Diasura (i.e. Dea Syria).

The map of the ethnicities presents also an interesting pattern. There is a big concentration of the nodes in the Near East and in Greece and very little in the Western part of the Roman Empire, ending on the Italian Peninsula.

Spatial variations between different regions in the cults of Allat and Atargatis religion illustrate well the religioscape of the goddesses pointing the cultural diversity and diffusion of the latter to the distant places beyond East, such as Rome, Dacia, Macedonia, Egypt, and Greece with islands and mainland. The role of Syrian emigrants seems significant in the implantation of cult in such distant places, especially on Delos where they are the cultic instructors for the Athenian agents (Figure 8). The ethnika of the priests who were from Hierapolis, Laodicea and Antiochia confirm the Syrian religious importation of the cult, gradually opening to the training of the priests of Greek origins with ethnika pointing out on Athens and the surroundings (Sounion). It was also remarked earlier by Nathanael Andrade (Andrade, 2022, p. 3), but the spatial visualization based on the ethnicities and location of the people pictures the network.

Figure 8

Map of worship of the goddess Atargatis, case of Delos (graphic: NodeGoat, modelled by A. Kubiak-Scheider and S. Mazurek).

The map of the worshippers of Atargatis in the Hellenistic Period poses questions about the drift of the cult of Atargatis outside of Syria and being later, in the Roman Imperial Period, adapted into the worship of a Dea Syria. It is interesting that the Dea or Thea Syria is a designation of the goddess used rather outside of the Near East and it is used in majority of the cases by the people of non-Syrian origins.

We encountered also some difficulties in the exact chronology of the inscriptions – not every text contains a time reference and datation on palaeographic criteria was not helpful in establishing a precise date of a text. Propitiously, the NodeGoat permits us to insert periods in such a case.

4 Conclusion and Perspectives

This approach and the use of the digital tools, is complimentary to the traditional approach of describing inscriptional data in publications, providing a presentation of the information from the epigraphic evidence for better understanding the human and geographic impact on the ancient religious systems. The database offers a mosaic of people placed in diverse territories linked by cult of the deities in spatial and social set-ups. It presents graphically a space of interactions and communications between the members of the same and different communities. The analysis of networks, especially on the example of Delos as well as spread of the cult of Atargatis in the Hellenistic Period through the Syrians on the distant places, shows the vertical transmission of religious information (Collar, 2007, p. 155). The ties indicated on the maps display the lively contacts between the places and people and the entangled state of the ancient world cultures and people.

The first results of the proposed NodeGoat database face the problem of religious and divine pluralism in the ancient polytheistic systems (Bonnet et al. 2018, pp. 585–589). Only through the means of connections, we can provide a comprehensive insight into the ancient polytheistic religions and understand the roles of society in creating the worship (Gasparini et al., 2020, pp. 1–8; Rüpke, 2015, pp. 344–346). The research presented within this work confirms the further potential of the studies of polytheistic religions through the prism of geography and social scope. The NodeGoat seems to be a relevant tool for seamless workflow in the domain of epigraphy and religions of the ancient world. The use of the NodeGoat in similar research has high potential in future. It will help to establish a relevant database for the study of the people of the gods, extending the topic on the other deities of the ancient world.