Guidelines for using the BDEG database

Download this documentation


.1.   A bit of history…


.2.   Functionalities


.3.   User ID and account management


.4.   Display of text in ancient Greek


.5.   Input in ancient Greek




.1. A bit of history…


            The project of creating a Greek Cult-Epithets Data Base (in French: Banque de Données sur les Épiclèses Grecques – BDEG) started as the idea to construct a tool which allowed researchers to study Greek polytheism and, more broadly, Greek religion; and also allowed them to move a step forward in quantitative terms, thus opening new perspectives in their quest to understand this complex human construction. Putting together an extensive meta-source with the available tools at that moment implied a documentary task in itself. However, it led to new questions and innovative ways to answer them. One of these ways is moving from the  polis scale to the region scale and to the Greek world and viceversa; or studying the links between certain cult-epithets and certain deities. It is worth mentioning from now, that here the quantitative aspect often has a positive impact on the qualitative one.   

            In 1997, Pierre Brulé provided the bottomline for this reflection during the CIERGA seminar in Liège: “Le langage des épiclèses dans le polythéisme hellénique [1], which lays the basis for a theoretical reflection about the cult-epithets and outlines the contribution of the diverse tools to explore the divine landscape just as the semantical richness of the Greek  cult-epithets describes it. A long time, as well as patience and work were needed to take that reflection from that point to its current completion and diffusion. First of all, there was a conjunction. Pierre Brulé submitted a project to the recently created Crescam which consisted of doing an in-depth research on the names of gods. Such research would have to be supported by a new instrument: The BDEG. What allowed Pierre Brulé to submit his project was the conjunction of two fields of common interest among the researchers team at the University of Rennes 2 (including Marie-Françoise Baslez, Nicole Belayche, Pierre Brulé, Patrick Le Roux and Jacques Oulhen): The first was onomastics, and the second was ancient religions.

            Since 1997, the intellectual basis that led the construction of this database has been complemented by Robert Parker’s “The problem of the Greek Cult Epithet” [2] and the statements gathered from the acts of the  Nommer les dieux  conference, published in 2005 [3]. In the latter, the contributions from Nicole Belayche, Madeleine Jost, Robert Parker and Vinciane Pirenne-Delforge shed some light  on the nature of the cult-epithets. Nonetheless, its definition was still difficult to figure out. New theoretical problems emerged as researchers were faced with challenges such as the data input in a database along with the technical difficulties it implies; the effect of confronting linguistic features; and what can be inferred about cult realities throughout the document. Sylvain Lebreton and Pierre Brulé tried to give an answer to some of these conceptual and methodological questions in the article “La Banque de données sur les épiclèses divines (BDDE) du Crescam: sa philosophie[4].

            Our objective is simple: We aim to gather a corpus of information about the names of Greek gods. More precisely, we seek to gather and make available all the data about Greek divine epithets, regardless of the source, time and region they come from. Thus, the entire ancient Greek history is our chronological framework, from the first Mycenaean stone tablets ever registered to the last ones which date from the  4th century of our era. From a spatial point of view, we could say that no defined geographical context is convenient. Instead, our field of research is formed by all Greek-speaking regions from the Antiquity. Sometimes, even the dedication of a Greek in a non-Greek-speaking region is evidence of his presence and loyalty to his gods. We seek to work thoroughly in each framework, chronological and geographical.

            Returning to the essential, what is, in brief, the definition of epithet taken into account here? The binomial naming system combines the name of a god, usually in the first place, with a complement which is commonly an adjective (there are also participles and prepositions) into a single nominal group. This complement, named cult-epithet, describes a particular feature of its respective god. Cult context and acts in ancient Greece mainly refer to gods so precisely defined, that it is possible to talk about a naming system. For those complements to be stored in our BDEG as cult-epithets, the nominal group should refer to a divine entity worshipped by ancient Greeks.

            There is an unavoidable stage of interpretation by the editor when storing an entry on the BDEG, since the source hardly ever specifies the quality of a cult-epithet next to a particular god’s name. The editor must then decide if the source refers to a cult practice and what part of the “divine nominal group” is the cult-epithet. In some few cases, it was difficult to make such decision. Generally, these unclear entries were stored to enable future verifications.

      The BDEG went through several stages. The first stage was the construction, completed in 2002 by Laurent Piolot, using FileMakerPro. The Database worked with this program until 2006. During this first stage, the base was enhanced in terms of information, but underwent little technical improvement. Under the direction of Pierre Brulé, this work was again developed by Laurent Piolot (2002 – 2003), who is again responsible for setting the intellectual and system bases  of Hélène Bectarte(2003-2004), Céline Dubois and Pierre Tandé (2004-2005) and Sylvain Lebreton (2004 onwards).

      The second stage was placing the database online. After a test phase (2006), it became available on the web in 2007, thanks to the initiative of Pierre Brulé, Sylvain Lebreton and Yves Kernaleguen (2006 – 2007). Pascal Gouéry, from the System Resources Centre at University of Rennes 2, completed the technical aspects of the database (data transfer, well elaborated layout, inclusion of sorting and search tools). We estimate that in spite of not being complete, the BDEG with about 6700 entries can already be considered as useful for the scientific community.  


       We were, nonetheless, fully aware of the imperfections of the BDEG, from which  incompleteness was the main one. There was a big amount of data to check thoroughly – there is, still – and many corrections and general harmonization had to be applied. Above all, there were some improvements to be made in order for the database to offer all the desired functionalities. The most important was offering the option to sort entries by chronological order. For technical and institutional reasons (lack of regular financial support in the first place), these changes did not happen for several years (2008 – 2013). During this lapse, Sylvain Lebreton was the only editor of the entries and continued to enrich and improve the Database (it has over 9200 entries today).        

       The reform of the Crescam into the LAHM and the CReAAH (UMR 6566) helped the BDEG to regain momentum and to keep evolving on its technical aspects. So far, it had been slowed down by administrative and technical difficulties. Since 2013, thanks to the remarkable job of Jean-Baptiste Barreau (studies engineer at the UMR 6566 CReAAH), responsible for its technical functionality, the BDEG is again available online on a new website. It has now a module of chronological sorting and a map tool; both of which must still be improved. The eventful history of the BDEG, the reality of the documentary data, the progress made on understanding quantitative treatment of the information, and the evolution of the digital tools explain that certain features of the database are very different from what they were at the time of its conception. Initially an entry was composed by a cult-epithet associated to a god’s name in a given location (usually a  polis). All the  occurrences were not meant to be registered:  the earliest and the latest seemed enough. As the base was used, a different approach was needed. In order not to write off some data, the editors  have progressively registered all the  occurrences of a cult-epithet for a given location, which in turn created more entries for a single locality. This actually means that the structure of the database was modified (by empirically entering the entity “attestation d’une épiclèse dans un source” i.e. evidence of an epithet in a source) without changing its theoretical structure, which made it impossible to treat information quantitatively, yet it was useful for the functionalities of chronological sorting and map referencing. Changes in the structure of the BDEG were necessary to ensure it could continue working. Karine Karila-Cohen’s integration to the project (since 2013) goes along with this purpose. Much of her research work has to do with the quantitative treatment of data in ancient history. A third phase in the history of the BDEG begins then. Sylvain Lebreton and Karine Karila-Cohen consider modifying the philosophy of the BDEG, and then testing a partially automatic migration of a part of the data (the one related to Attica) from the old database structure to the new one. Sylvain Lebreton will continue improving the entries on the website currently available online.    


The BDEG is always evolving. We are aware of the many mistakes that can be made in the process of constructing such an extensive tool. Consequently, we listen carefully to (we could even say we expect) your criticisms and suggestions. If you have any remarks, please do not hesitate to let us know through the Contact section. Your comments will allow us to improve the database and to work on its continuous development, always aiming to foster the study and interpretation of Greek polytheism. 

So as to wish you a “Bon voyage”, and since the use of “navigation” is from now on dedicated to all the manoeuvres to surf the web, let us bless you under the protection of Αφροδίτη Εὐπλοία!


Pierre Brulé, Karine Karila-Cohen, Sylvain Lebreton,

November 2014.



.2. Functionalities




A meta-source like this, given its own nature, implies certain rules of use that are different from those for the constitution of a corpus. The editors of the BDEG did not go to primary documents; there was not a thorough analysis of stones for epigraphic texts, nor critical philology concerning publications. The editors started from publications that worked as a reference (including the SEG,  for epigraphy). The reader has access to these publications for all discussions about the edition of the text. 


The aim of the BDEG is to allow requests according to the different needs of every user. It offers two search possibilities.


2.1. Single-criterion search


The single-criterion search module (or general search), which is accessible directly on the website, is a simple tool which allows a quick search according to the following categories: nom de la divinité (name of the god)nom de l'épiclèse (name of cult-epithet)traduction de l'épiclèse (translation of the cult-epithet)divinités associées ( associated deities)commentaire (commentary)nom du lieu (name of place)nom de la region (name of region)référence de la source (reference of the source). It is not necessary to type the accentuations, or even the complete term. Thus, if the user types “karp”, s/he will find matches for “karpophoros”, but also “karpios”, “kallikarpos”, “epikarpios”, etc. in the name of cult-epithet. If applied to all categories, results like “Karpasia” will also be found on the name of place category. A recap chart with the number of matches in every category is shown and it can be expanded by each one of the matches.    

It can be seen here that accentuation and the order of uppercase and lowercase characters does not affect the search process.





                 By clicking on the number of the entry (at the beginning of each line), the user can see the details (see example below)

This search mode also allows switching between different categories of the BDEG simultaneously based on a single term or root. This helps to measure the quantitative importance of the Epiclèse 1 (Cult-Epithet 1) and Epiclèse 2 (Epithet 2), the Divinités associées (Associated deities) and the Commentaire (Commentary), for example. Nevertheless, the Multi-criteria search module is better adapted to the matches for one epithet (with their locations and references), especially if it is very common ( Soter for example).



2.2. Multi-criteria search

The possibility to specify the variables is what makes the BDEG so interesting. Let us give you a simple example. If you are searching for Athena Polias, in Attica for example, this tool makes it possible to find that one deity very quickly. Type the data in the corresponding fields, namely Divinité (Deity)Épiclèse (Cult-Epithet) and Région (Region), then click on lancer la recherche (Search) and the numbers of the entries that match your criteria will appear.

There are also some criteria to choose how the entries are sorted and shown. These make it possible to choose which criteria you want to see as well as the order in which they appear. This tool becomes indispensable for a user who has to deal quickly with a large number of entries.

            A map tool is also available - click on Carte (Map) -. This tool allows you to localise the results of a multi-criteria search on a map. It is currently available only for Attica as it is still on the test phase


2.3. Search by entry number

This function can become essential, for instance, for a user who works regularly on the database and who needs to go back to an entry previously found. By using the number of the entry, the user can easily find it again. This allows the user to save time and to avoid typing again all the criteria in the fields of a multi-criteria search.




The task of establishing the content of each section, as well as the the way the user interprets it call for comments. The fields of the model entry, just as it was designed by Laurent Piolot and just as it appears when the user clicks on the number on the left in the sorting table, will be examined one by one.  


2.4. Example of an entry: Dionysos Karpios


Registries show two results:


Lieu (Location) Larissa[5]

Région (Region) Thessalie (Pelasgiotide)


It is not always possible to find the location of an entry. Many editors  only mention a location such as « Phrygie » (Phrygia). In this case, the column Lieu (Location) is empty.  The aim is to relate a document to a certain polis (and to its territory); but sometimes only a modern designation  is known. In the case of islands, we have considered that, for the most important ones, their name should be in the Région (Region) column and the name of the polis in the Lieu column. Consequently, it is possible to have an entry with Lieu and Région columns containing the same information; it is the case of Rhodes, for example. For small islands, however, especially those with one only city, it is the name of the island that will appear in the Lieu column and “Egée” (Aegean) (most of the times) in the Région column.




It is found, successively:


Divinité (Deity)




Épiclèse 1 (CultEpithet 1)



« Des fruits » (of fruits)


(Épiclèse 2) (Cult-Epithet 2)



All names are registered in  their nominative singular form (except invariable names, types of epithets such as Tiamou).


Divinité (Deity)

This column contains the name of the deity. It might not be included in the source and we might have only the cult-epithet. We have tried to take it into account in the research process as much as possible, bearing in mind that the absence of such name is a major concern. Brackets are shown (in the fields in French and Greek) if the name has not been retrieved. If it cannot be retrieved due to the incompleteness of the document, the restitution is shown between square brackets (only in the field in Greek)

While all AthanaDamater or Artamis… are registered under the names of Athena, Demeter and Artemis in the French field [6], the slight dialect differences in the Greek written form are respected.   This procedure is usually applied if we learn more from the meta-source about this subject.


Épiclèse 1 and Épiclèse 2 

A reason for having different columns, Épiclèse 1 and Épiclèse 2, is that sometimes, names of deities have not one but many cult-epithets. Therefore, an entry has been built for each cult-epithet so as to find them easily if they are sorted from the Épiclèse 1 column (this does not correspond to the order in which the cult-epithets appear in the document, so we redirect the reader to the source). The Épiclèse 2 column does not appear unless it has been filled. Such is not the case of this example.

The cult-epithet is transliterated according to the most common uses (thus e for eè for hu for ukh for c,ô for w). When cult-epithets include a preposition such asen kèpois”, the preposition is shown in brackets after the noun it refers to. For this example: “kèpois (en)”. In Greek, however, the original order is kept ἐν κήποις).

            Just as in the Divinité column, the restitution square brackets will be used only in the Greek fields.  Certain cult-epithets  indirectly attested are shown in brackets in the Greek and French fields, such as cult-epithets attested through association names deriving from them , for example (cf. the Dios Ataburiastai for Zeus Ataburios in Rhodes ).

            Translation is shown between quotation marks. This is one of the most difficult fields to complete. Certain cult-epithets are still enigmatic, mostly because they do not relate to a Greek  root. Sometimes, a hypothesis is proposed (in this case the hypothesis is followed by a question mark).

Dating and references

In the current version of the BDEG, the total amount of references is not registered, but for us, it was important to indicate the relative volume of each epithet in a given place and within a chronological limit:


- we indicated in the register which were the earliest and latest reference for a given epithet in a given place. Thus, for Dionysos Karpios  in Larissa,  the third quarter of the fifth century  BCE for the  earliest date (Source 1) and the Imperial period for the latest date (Source 2).


-  We have evaluated approximately ( to the date of creation of the entry) the number of references of a cult-epithet found in a single city. This is shown in the column Nombre d’épiclèses (Number of epithets), using the following scale: one (1), between two and five (<5), between five and ten (5<10) and over ten (>10) references.


The entry is shown:


Source 1

D. Theocharis, AD 16 (1960)[1962], B : 185 (BE 1964, 226) = A. Kontogiannis, in Praktika tou a' historikou-arkhaiologikou Symposiou Larisa, Larisa, 1985, p. 111-129 (SEG XXXV 590)

Nature de la source Epigraphique

Date de la source 450-425 a.C.


Source 2

IG IX 2, 573

Nature de la source Epigraphique

Date de la source Impériale                                    


Nombre d’épiclèses <5


Complementary comment.

The  abbreviations used for editions of  inscriptions are those  of the SEG, but also the EBGR or the BE. For technical reasons, the titles of books, newspapers, etc. are displayed in Latin characters only (titles in Greek alphabet have been transliterated) and could not be registered in italic characters.

Nature de la source  (Nature of the source) is chosen from literary, epigraphic, papyrological (rarely) and numismatic (sometimes).

Bearing in mind the uncertainty concerning the dating of epigraphic texts, introducing a chronological segmentation by thirds or halves of century was pointless. The Date de la source (date of the source) is specified by century or group of centuries (ex : IIe-IIIe p.C.), or, by period (Hellenistc, Imperial, Byzantine[8]…). Nevertheless, the dates are specified as precisely as possible in each case in which the source indicates it. In these cases, the dates are shown in  Arabic numerals, like in the following example: 450-425 a.C. for the third quarter of the fifth century BCE.


Four more columns


Culte attesté : oui

(Largement restitué)

Divinités associées 2 : Déméter Phulaka


Commentaire            dédicaces (2 : par une prêtresse) ; cf. P. Chrysostomou, Hyperia 2 (1994) : 113-149 (BE 1997, 285 = EBGR 1996, 40)


Culte attesté 

In most of the cases, the mention of a sacrifice, a priest, a sanctuary or a simple dedication, allowed us to confirm a cult to the deity the cult-epithet referred to. On the other hand, certain divine epithets have been included even if they have been mentioned in certain sources out of the cult context (this might be useful for chronological order purposes). The editor chooses « OUI » or « NON » (“yes” or “no”). Sometimes, it is impossible to make such choice, so the mention « Non renseigné » (Not specified) is used. It is worth mentioning that this undefined choice is possibly due to laconic meta-sources.


Largement restitué

It is applied to epigraphic texts where we consider that a doubt can be accepted according to the risk taken during a  restoration (bearing in mind that we do not  comment on the pertinence of the restoration). A « OUI » (“yes”) is used in this column, which will not be shown otherwise (like  in the case of Dionysos Karpios  of Larisa). As a complement, right square brackets are added sometimes (only in the Greek fields) when the source or meta-source allows it.


The Divinités associées ( associtaed deities) are those which appear next to the deity in the text (before or after). It might be on a coherent list ( an oath or sacrifce) or on some  relief.


Commentaire (Commentary) – longer or shorter according to the meta source and the editor of the entry – aims to provide complementary information (types of support: altar, tablet ; topographical context : temple ; nature of the text : hymn, dedication ; bibliographical additions…).


When Commentaire or Divinités associées deal with only one of the sources, or if they are different for each sources, a number (1 or 2) indicates the sources they refer to. Thus, the Dionysos Karpios  of Larisa is  associated with Déméter Phulaka only in the  inscription that appears in the Source 2 column.


2.5. Contacts


Do not hesitate to let us know about any malfunction or mistake in an entry through the Contact section. We will correct it after verifying it. This interaction with the user is totally positive since it allows improving the information proposed by the BDEG. Thank you for your participation.

.3. User ID and account management.


-It is not necessary to log in to have access to the database.


-The application is developing a tool for creating a user account, changing the password, etc…


-A utilisateur (user) can be designed as auteur (autor) or administrateur (administrator) by an administrator of the application.


-The auteur has the possibility to create, delete or modify entries, apart from the search functionalities.


-The administrateur, besides having the same rights as the auteur, must manage the accounts of the whole group of users..

.4. .   Display of text in ancient Greek:


Texts in ancient Greek can be displayed in different ways:


        Specific Greek fonts (greekgrammataalpha...)

        Unicode fonts, supporting multiple language uses


In the current version, the user can use specific fonts or Unicode fonts. S/he can switch between them by clicking on the following icon:

The site uses the greek  or grammata fonts.


The user can indicate another font type s/he wants to use to display the fields in Greek (because s/he does not have the greek or grammata font, or because a new one is more convenient for her/him).


S/he only has to specify the name of the font in the form that opens by clicking the icon above (it can be found in all the pages that display text in Greek).


This setting is saved by a cookie mechanism (if authorised by the user) which allows using it in different sessions.


Many websites deal with displaying text in ancient Greek, we recommend visiting:


.5. Input in ancient Greek :


The technique used here consists of typing the diacritc mark before the character.


The keys to insert a diacritical mark are the following:


Diacritical mark



Name of the key

Esprit rude (Dasia)



Opening bracket

Esprit doux (Psili)



Closing bracket

Accent grave (Varia)




Accent aigu (Oxia)




Tréma (Dialytika)




Tilde (Perispomeni)


Tilde (Alt Gr + é)

Cédille (Iota souscrit)





The diacritical marks can be combined in any order, for example:


(/,a      /(,a       ,(/a   ... having the same result : ¯



Currently, the key combinations of the non Unicode font “greek” are the only ones that are recognised.


Diacritical marks do not appear directly on uppercase letters. It is advised to type  an  accented spacebar with the mark, followed by the capital letter (the character will not be as elegant as if it had been combined with the diacritical mark directly, but the correct recoding will be done automatically when the unicode version of this tool is released).

Thus, to obtain  Α, you should type the following sequence :   )/ <espace> A  ou /)<espace>A


Finally, to display the comma, you must only tap the key twice, same as for the brackets.


[1]                Kernos 11 (1998) : 13-34.

[2]              Opuscula Atheniensia 28 (2003) : 173-183.

[3]              Nommer les dieux. Théonymes, épithètes, épiclèses dans l’Antiquité (qui publie le colloque du même titre de Strasbourg), N. Belayche, P. Brulé, G. Freyburger, Y. Lehmann, L. Pernot, Fr. Prost (éds), Brepols.

[4]              Kernos 20 (2007) : 217-228. The  Banque de Données sur les Épiclèses Divines (BDDE)  hs since then become the Banque de Données des Épiclèses Grecques (BDEG),  as well as the Crescam (Centre de Recherches et d'Etudes des Sociétés et Cultures Antiques de la Méditerranée) has been reformed in the LAHM (Laboratoire Archéologie et Histoire Merlat, UMR 6566 CReAAH).

[5]              Titles in red bold are those of the columns of the entry; in italics the information relating to  Dionysos Karpios  in Larissa.

[6]              This avoids to split entries relating to the same deity when searching in  Divinité ; indeed, if we had faithfully taken into account all local spellings,  Despoina would have appeared between Damater and Demeter, for example.

[7]              .

[8]               Sources from the Byzantine era, such as lexicographers, can not be considered as evidence for a cult in their own period, of course.