The spread of Egyptian cults across the early Hellenistic Aegean Sea

During the reign of the first six kings of the Ptolemaic dynasty, which is the period between the end of the 4th century BCE and the middle of the 2nd century BCE, Egyptian cults spread successfully from Egypt, particularly from Alexandria, to ports in the ancient Mediterranean. These cults were formed almost exclusively around the divine couple of Isis and Sarapis.

In the subsequent centuries when the Roman Empire gained a significant influence in Egypt, the cults spread further to the mainland reaching all the Roman provinces.

A thorough consideration of the evidence concerning the spread of the Egyptians cults of Isis and Sarapis opens the following questions: what were the most important factors responsible for the observed spatial patterns of the cults’ distributions and which of these factors had more weight than others? The original, most influential, and at the same time divergent hypotheses in academic discussion emphasized either Ptolemaic political propaganda or the maritime trade network as the key factors in the spread of these cults. Available historical evidence partially supports both hypotheses:

A) Ptolemaic rulers were politically engaged in many affairs across the ancient Mediterranean and the cult of Isis and Sarapis was closely tied to the royal dynasty;

B) Ptolemaic Egypt was one of the main exporters of grain by sea and Isis was a patron goddess of sailors. Research has progressed extensively over the last hundred years and historians admit that the combination of factors standing behind the successful spread of Egyptian cults during the Ptolemaic era is more complex than was originally argued.

The topic of the spread of religious innovations in the ancient Mediterranean has been evaluated mainly by established historiographical methods such as the collection and critical analysis of archaeological and literary sources. Hypotheses and conclusions derived from these methods are, however, often unable to reflect the complexity of historical processes. It is, in other words, too difficult for the human brain to follow and evaluate all the interactions among the many variables involved. A possible solution can be found in supplementing this established methodological apparatus by formalized methods, e.g. the coding of relevant datasets, statistics, computational modeling, geospatial modeling, and network analysis. The application of these methods in the study of complex historical processes is inherent to the theoretical and methodological framework of spatio-populational modeling in cliodynamics, generative social science, and network theory.

We focus on the spread of the Egyptian cults of Isis and Sarapis during the period between the end of the 4th century BCE and the middle of the 2nd century BCE in the area of the Aegean Sea and especially on the Aegean Islands. There are several reasons for selecting this area and timespan. The main trading routes between the Egyptian port of Alexandria and continental Greece traversed the Aegean Sea and the first Ptolemies perceived the Aegean Islands as strategic locations for their military and diplomatic operations. In terms of time, the beginnings of the spread of the cults of Isis and Sarapis and other related Egyptian deities (e.g. Anubis) outside Egypt are tied with the first three Ptolemaic kings, who made these cults more accessible for the Greek audience. Although the veneration of Sarapis in its typical form most probably started with Ptolemy I, who assumed the title of pharaoh in 304 BCE, the expansion of Egyptian cults outside Egypt intensified later in the time of Ptolemy II and even more so during the reign of Ptolemy III. Around 167 BCE, the crucial nodes for Egyptian exports were restructured according to Roman influence and the conditions shaping the early spread of Egyptian cults changed significantly.

Because of the geographical layout of the Mediterranean basin, maritime trade was an indispensable part of ancient economies. Hellenistic Egypt was one of the main exporters of grain, a staple food in antiquity. Egyptian grain was channelled to the Aegean Sea and finally to continental Greece mainly through the island of Rhodos, Egypt’s main partner in maritime trade affairs.

When we seek to characterise this broad network of commercial relationships between ancient Egypt and the Aegean Sea region and determine which islands in the Aegean Sea imported Egyptian grain we are hampered by the lack of detail in ancient literary and archaeological evidence. This is a serious problem when considering merchants travelling from Egypt as potential carriers of Egyptian cults. This issue, however, could be partially resolved by analysing pieces of data which are not primarily considered as archaeological or literary evidence but which indicate the islands in the Aegean Sea that were, for example, fertile or barren, thus answering the question of which islands were potentially more dependent on Egyptian grain imports than others.

The Hellenistic trade route connecting Egypt, Rhodos, and Athens is attested by Demosthenes (Speeches 56), Diodorus (Library 20.81, 20.88, 20.96, 20.98, 20.99), and Polybius (Histories 5.88.1–90.4). Proof of Athenian import of Ptolemaic grain during the 3rd century BCE can be also found in some texts from the Inscriptiones Graecae collections (IG). One of them is the Phaidros of Sphettos’ honorific decree (IG II/III3 1, 985; IG II2 682) which describes the embassy to Ptolemy I Soter securing the grain for Athens. The embassy is dated between the years 294–287 BCE. The decree for Kallias of Sphettos describes a donation of 20 000 medimnoi of Ptolemaic wheat received by Athenian representatives on the island of Delos during the reign of Ptolemy II Philadelphus. This text is particularly relevant because it puts the island of Delos on the map of the Ptolemaic grain trade. The strong trading relationship between Rhodos and Egypt is also attested by a large number of Rhodian amphorae found in Alexandria. Moreover, Egypt had an advantage over the Black Sea with respect to grain export due to the weather conditions. While ships traveling from the Black Sea to the Aegean Sea could have been trapped north of the Dardanelles due to early winter, Egypt was able to dispatch ships continuously throughout the year (see e.g. Demosthenes, Speeches 35.13 and 56.30).

The Ptolemies were very active politically during the Hellenistic period in the area of the ancient Mediterranean. They were engaged in several military conflicts with the Antigonid and Seleucid dynasties, and spread their influence on the Aegean Islands and the coastal cities of Asia Minor.

Probably most relevant for the purposes of the quantitative analysis are the Ptolemaic garrisons dispersed across the Aegean Sea during our period of interest because

A) they are easily traceable both in time and space and

B) many of them were stationed for long periods of time, thus increasing the chances that cults were spread by Ptolemaic soldiers.

The Chremonidean War (267–261 BCE) is a particularly relevant conflict in this context. In the 270s Antigonos Gonatas defeated Pyrrhos of Epirus and strengthened the Macedonian position in Greece. This political development was inconvenient for the Ptolemaic diplomatic activities on the Aegean Islands. At the request of the Athenian statesman Chremonides in 268 BCE, Sparta, Athens, and Egypt formed an alliance with the purpose of defeating Antigonos Gonatas. Patroklos, the general of the Ptolemaic forces, led the Ptolemaic fleet to Attica and during the conflict he garrisoned some of the key ports in the Aegean Sea which remained under Ptolemaic control for decades.

To be able to compare the possible impacts of different factors on the spread of Egyptian cults on the ancient transportation network in the Aegean Sea region, we

1) constructed a model of the ancient maritime transportation network as a platform for the quantitative analysis,

2) transformed selected factors of possible influence on the spread of Egyptian cults into georeferenced parameters of the network, and

3) defined a mathematical model which allowed us to determine which parameters of the network explain the spatial dissemination of archaeological evidence connected to Egyptian cults.

Our model can be divided into three parts. The first part of the model rates the production potential of each spatial unit (pixel) on the basis of the local climate, soil quality, and terrain slope. Climate data are normalized on the basis of the minimal requirements for barley production. The bulk density of soil was used as a proxy for soil quality due to its relationships with other properties (porosity, soil moisture, hydraulic conductivity, etc.). The maximum limiting value for terrain slope for agriculture was selected and validated via the analysis of CORINE Land Cover data–land cover categories related to agriculture were extracted; then correlations for particular islands between model outputs and the real situation were identified in order to select the most suitable value. Areas above this threshold were then filtered out as inappropriate for agricultural production and areas below it were rated by measuring the difference from this limit. The theoretical average yield for barley was 680 kg/ha. In the case of our model, this value was assigned to areas (pixels) with a score of 1; other values were assigned proportionally to their score. Finally, to estimate the yield for each island, areas were spatially aggregated and their values summed.

The second part of this model estimates the hypothetical food consumption on these islands. In order to obtain the consumption numbers, the model needs demographical data–specifically, the population sizes on the Aegean Islands in antiquity. Indeed, modern literature provides some population estimates or methods for estimating the populations of the ancient Aegean Islands; however, the data are unreliable and consensus among researchers has not been achieved. As population growth factors such as the environment, technology, agriculture, and health care remained substantially unchanged in this region up to the beginning of the 20th century CE, we can assume that the population distribution in our time of interest should correlate closely to numbers from the first modern population censi. That said, it is necessary to add that we are aware of the very hypothetical character of this approach. However, for research purposes and also for the purposes of the model’s goal, which was to compare the food production/consumption ratio between selected islands, we were more interested in the relative population sizes of particular islands than in exact historical numbers, the latter probably being beyond our grasp.

After the extraction of these population estimates we were able to implement food consumption in our model. We used the average intake of a person in ancient Greece estimated by Foxhall and Forbes to be 212 kilograms of grain per year. In the third part of the model, the difference between food production and consumption was relativised by the number of inhabitants to get a nondimensional coefficient–islands with values under 1 were marked as potentially vulnerable to food shortages; islands with a higher number were considered as self-sufficient. It should be noted that this model has no ambition to serve as the only tool for assessing the question of which islands in the Aegean Sea region imported grain from Egypt. Rather, its purpose is to offer a basic comparative perspective on the islands’ agricultural potential.

Statistical analysis was used to evaluate and compare the possible impact of the factors quantified and transformed into parameters in previous steps on the early spread of Egyptian cults in the Aegean Sea region.

The mathematical models applied in this study revealed statistically significant patterns and correlations with respect to the different factors involved in the process of the early spread of the Egyptian cults of Isis and Sarapis across the Hellenistic Aegean Sea.

One such correlation is the strong connection between the placement of Ptolemaic garrisons and the spatial dissemination of archaeological evidence relating to Egyptian cults. More specifically, and in mathematical language, the presence of Ptolemaic garrisons explains 48% of the variance of the dissemination of Egyptian cults. This means

A) that people situated on an island garrisoned by the Ptolemies often had an Egyptian temple on the same island or in their proximity (e.g. in a near port on the network); or

B) that there is a high probability that people on an island with a Ptolemaic garrison would have had an Egyptian temple in their proximity in the future.

Ptolemaic garrisons remained, in many cases, for decades and one interpretation could be that Ptolemaic soldiers residing on islands contributed significantly to the successful spread of Egyptian cults.

Another statistically significant result is that agriculturally self-sufficient islands were spatially related to the weak presence of Egyptian cults (explaining 11% of the variance). This outcome suggests that islands which were agriculturally self-sufficient were not, in some cases, as attractive to merchants exporting grain from Egypt as those suffering from food shortages. However, the situation is more complicated. More specifically, there is indication that the agricultural factor gained significance in areas further away from Ptolemaic garrisons. This highlights the possibility that Ptolemaic garrisons had more weight with regard to the early spread of Egyptian cults in the Aegean Sea region than the agricultural factor.

A further result explaining 9% of the variance, is that higher levels of centrality intensified the presence of Egyptian cults. Here, the interpretation is that the strategic position of a port could attract the attention of the Ptolemaic dynasty and people travelling from Egypt in general.

The results show that factors from different levels such as trade, politics, network position, and distance probably all worked together in a complex manner that allowed the successful spread of ancient Egyptian cults in the Aegean Sea region. This finding is in accord with Bricault’s hypothesis claiming that the first wave of cults connected to Isis and Sarapis spread successfully because of a combination of commercial, economic, political, and social factors. However, the main aim of this study was to compare the possible impacts of individual factors on the process of the spread. Statistical results from the mathematical models suggest that the spatial distribution of Ptolemaic garrisons from the 3rd and 2nd centuries BCE probably had a more significant impact on the dissemination of Egyptian cults in the Aegean Sea region than other factors considered in the analysis. However, it should be noted that the analysis is tied to a specific region and that the factors discussed could have had different weights in other parts of the ancient Mediterranean. With respect to temporal dynamics, a relatively broad and uncertain dating (often connected only to a certain century or to its part) of the archaeological evidence related to Egyptian cults does not allow us to statistically uncover different stages of the spread of these cults during 3rd and 2nd century BCE.

(Source: “Ptolemaic military operations were a dominant factor in the spread of Egyptian cults across the early Hellenistic Aegean Sea”, by Tomáš Glomb et al.)

Research-Selection for NovoScriptorium: Philaretus Homerides