Bilingual Phenomena and Cultu(r)al Interactions in the Roman Empire

In the 1st c. AD, Albinus, the son of Chaerea, an inhabitant of the northern province of Lusitania, erected an inscribed altar dedicated to the patron gods of his city. The inscription combined the local language with Latin, employing the former for the names of the deities and the latter for the name of the dedicator and the votive formula. It could be argued that this bilingualism is a marginal phenomenon or characteristic of a peripheral province, indicative of an early stage of Latinisation and “Romanisation”, yet conservative in its customs. However, such a phenomenon was widespread across the ancient world. Let us entertain a hypothetical scenario in which Albinus emigrates to Rome. In the bustling streets of the imperial capital, the Lusitanian would encounter individuals speaking a multitude of languages and worshipping a plethora of deities from all the corners of the known world. One such individual was Tiberius Claudius Felix, a Palmyrene Roman citizen, who erected an altar to a deity from his homeland together with his family.¹ Unlinke Unlike the Lusitanian, however, Tiberius began his inscription with a Latin dedication to his deity, “translated” into a Roman god, and subsequently provided a translation in his native language, Late Aramaic (Palmyrene), adding different religious nuances to ensure that the essential message was accessible to all.

These two examples illustrate genuine cases of cultural and linguistic interaction, each characterised by distinct mechanisms of ritual communication but united by a shared socio-religious aim: multilingual expression. Yet it is difficult to fully grasp these bilingual scenarios without considering the cultural subtilities subtleties that shaped the lives of people who did not always resort to different languages. In other words, how can we engage with the underlying realities of such actions, moving beyond a purely mechanical analysis of their expression?

It is not sufficient to examine the mechanisms of bilingualism in religious expression; rather, we must also seek to understand the very nature of that expression – often monolingual in form, yet profoundly shaped by the interaction of multiple cultural systems. One Cult, Multiple Cultures invites reflection on precisely this dynamic. The cult of a single god – or of several – in Aantiquity was frequently the result of both linguistic contact and multiple cultures, not of one operating in isolation from the other.

As the first section of this volume demonstrates, in Lusitania, beyond the bilingual inscriptions that preserve traces of the local language, we encounter a rich indigenous theonymy whose religious expression is mediated through Roman epigraphic conventions and cultural frameworks.² What this book seeks to demonstrate is that the material evidence – whether monolingual or bilingual – reveals more than linguistic hybridity; it points to processes of identity negotiation, religious intent, and cultural belonging. Through this lens, multilingualism emerges not merely as a communicative strategy but as a social phenomenon embedded within broader dynamics of acculturation, continuity, and religious innovation.

Throughout its history, the religious landscape of the Roman world was characterised by remarkable adaptability and the coexistence of diverse belief systems and cultures. Roman expansion – initially republican and later imperial – enabled the incorporation of vast territories across the Mediterranean and beyond into the romanitas, while also integrating deities, cults, and practices from a wide range of traditions. The establishment of what J. Osgood has termed the Mediterranean’s first “world state”³ facilitated the circulation of people, goods, and ideas on a scale never seen before in that part of the world.

In Aantiquity, as today, public religious expression was not merely a matter of personal devotion but was deeply intertwined with individual and collective identities. People could worship the gods of the Roman state, participate in local cults, or adopt foreign practices, constructing multiple, complementary religious identities without contradiction – and at times in interaction with one another. This frequently involved expressing devotion to deities from diverse cultural and linguistic backgrounds. Both in the provinces and in Italy, different communities maintained their own religious traditions, honouring (adapted) deities that reflected their heritage while also incorporating new ones, whether drawn from the Roman pantheon or introduced from other regions of the Empire.

Although Latin became firmly established as the administrative and prestige language of the Empire (at least in the western regions), and Greek retained its status as a lingua franca and cultural language, linguistic diversity remained the norm throughout much of Roman history. The religious interplay between political unity and cultural diversity was mirrored in the Empire’s linguistic situation. As M. Dubuisson has argued, the Romans were generally indifferent to foreign languages, with the notable exception of Greek. While a Roman citizen was expected to know Latin, there was never an official language policy imposed across the Empire.⁴ This linguistic “laissez-faire”, combined with the nature of religious practice – so distinct from modern notions of religious denomination, linguistic community, and nation-state – enabled a range of multilingual contexts that this book addresses from the perspective of religion. These included stable forms of bilingualism, diglossia, the use of linguae francae, and instances of language interference. Such a scenario gave rise to a “socio-multilingual” dialogue that persisted for centuries, ultimately leading to the gradual Latinisation of much of the western Empire, though notable exceptions, such as Sicily, remained.⁵

The reality of a multicultural world, where religious ideas and languages circulated, was not unknown to the inhabitants of the Empire, just as it is not unfamiliar to us in the 21st c., despite the differences. Perhaps reflecting our increasingly globalised environment, sociolinguistics has emerged as a distinct discipline since the 1960s, growing rapidly in both scope and influence. This scholarly development has contributed significantly to the rise of studies on multilingualism and cultural contact, which have since been applied to the field of Ancient History. While a detailed survey of the discipline lies beyond the scope of this introductory chapter, it is essential to acknowledge the impact of U. Weinreich’s pioneering work on interlinguistic interference and bilingualism, as well as J. Fishman’s contributions to the study of linguistic domains and ethnolinguistic vitality, particularly with regard toregarding linguistic Latinisation and the decline of pre-Roman languages.⁶

Initially, sociolinguistic analyses of the Roman world focused on the interaction between Latin and Greek, the two dominant languages of the Classical era, which monopolised much of the epigraphic and literary production. In this context, canonical works such as J. Kaimio’s influential monograph on the influence of Greek in the Roman world,⁷ along with other studies, laid the groundwork for more sophisticated discussions on diglossia and bilingualism.⁸ These early contributions have shaped the historical study of multilingual phenomena and continue to inform debates surrounding Latinisation and linguistic interference in Aantiquity.⁹ However, as the perspectives offered by these works evolved at the turn of the millennium, scholars increasingly recognised the need for precise technical definitions to describe the linguistic phenomena attested in written sources.

Since the early 21st c., the technicalisationtechnicalities of discourse have advanced considerably, marking a key milestone in this field with the introduction of terminology borrowed from sociolinguistics and adapted to Ancient History and epigraphy. A major turning point came in 2002 with the publication of Bilingualism in Ancient Society: Language Contact and the Written Text, a collaborative volume edited by J. N Adams, M. Janse and S. Swain, based on a conference held in Reading in 1998. In this volume, alongside a deliberate multifocal approach that addressed language contact across various aspects of everyday life (such as the military, religion, and administration), these phenomena were examined not only in relation to Latin and Greek, which nonetheless dominated much of the discussion, but also with reference to other languages. Notably, D. R. Langslow’s chapter, focusing on the study of fragmentary languages (referred to as corpus languages or Trümmersprache in J. Untermann’s terminology), has had a significant influence on subsequent research into Palaeo-European languages.¹⁰

The following year, J. N. Adams published Bilingualism and the Latin Language, which has since become the most widely cited monograph on bilingualism and cultural contact in the Roman world. It offers a comprehensive analysis that extends beyond Greek-Latin bilingualism. One of its merits is the systematic attempt to define the phenomena of historical language contact through the application of sociolinguistic concepts to the realities of Roman society. However, explanations concerning the cultural mechanics of bilingual phenomena – such as the naming of deities – fell outside the scope of this initial undertaking. J. N. Adams emphasised that bilingualism is not confined to the influence or imposition of one language upon another, and, through epigraphic and literary evidence, he identified three types of bilingual phenomena observable in bilingual situations:¹¹

1. Linguistic interference: the unintentional transfer of features from the dominant language (L1) to the secondary language (L2).
2. Linguistic borrowing: the adoption of linguistic elements from one language into another.
3. Code-switching: the switching from one language to another within a written text or spoken communication. This may occur for various reasons, including linguistic interference or specific ritual objectives. Code-switching can be further classified into tag-switching (the insertion of expressions or elements from another language), inter-sentential switching (switching languages between sentences), and intra-sentential-switching (switching languages within the same sentence).

In terms of the texts that would reflect these phenomena, J. N.  Adams distinguished between:¹²

1. Bilingual texts: texts written in two languages, conveying all or part of the same information.
2. Transliterated texts: texts written in one language, but using a different graphic system (e.g. a Latin text written in the Greek alphabet).
3. Mixed-language texts: texts that exhibit code-switching.
4. Texts which implicitly reflect a bilingual situation: texts that indirectly indicate the presence of bilingualism in their production, including, among others, spelling interference, the change of alphabet, or the translation of formulae or clichés.

J. N. Adam’s monograph helped to shape the Anglophone approach to the topic, fostering growing interest in linguistic contact and the incorporation of sociolinguistic concepts and methodologies into historical analysis. Notable works in this context include A. Wallace-Hadrill’s influential 2008 book Rome’s Cultural Revolution, and the collective volume Multilingualism in the Graeco-Roman Worlds, edited by A. Mullen and P. James.¹³ On the Continent, the conferences held in Lyon in 2003 and Madrid in 2009, whose proceedings were published in 2008 and 2011 respectively, reflect the parallel development of the field. While the Lyon conference focused on Greek-Latin bilingualism and epigraphy, the Madrid event explored linguistic contact in the western Mediterranean, extending the scope of study to western Europe and incorporating the increasingly robust schools of research on Palaeo-Hispanic, Italic or Gaulish epigraphy into the debate.¹⁴ This trend is further exemplified by M. J. Estarán Tolosa’s 2016 publication, Epigrafía bilingüe del Occidente romano, a detailed study of bilingual epigraphy between Latin and the western Palaeo-European languages, with a particular emphasis on new cultural aspects that go beyond the strictly linguistic dimension.¹⁵

In recent years the relationship between Latin and local languages in western Europe has received increasing attention from a holistic perspective, though not necessarily with a focus on the religious expressions addressed in this volume. A few key developments in this area include the ERC LatinNow project, which has advanced research on linguistic Latinisation. This project has resulted in the publication of several volumes exploring the spread of Latin, the disappearance of local languages, the social factors driving these changes, and language contact in Late Antiquity.¹⁶ In France, this approach has also been pursued in collected volumes edited by E. Dupraz and W. Sowa, as well as by R. Roure,¹⁷ whose contributions are particularly notable for advancing the study of linguistic contact and paving the way for future research into multilingual approaches within specific domains, such as religious expression.

Beyond individual case studies, one of the most widely debated aspects of this new line of research has been the methodological classification of texts – particularly epigraphic texts – that exhibit linguistic contact. Despite several revisions, J. N. Adams’ taxonomy has remained especially influential over the past two decades. In Multilingualism in the Graeco-Roman Worlds, A. Mullen acknowledges J. N. Adams’ definitions of linguistic phenomena but proposes a revision of his text-classification system, highlighting the challenges of adapting his categorisation to the fluid realities encountered in linguistic studies:

1. Bi-version (or tri-version) texts: equivalent to J. N. Adams’ Bilingual texts, while recognising the wide variety of outputs.
2. Texts displaying bilingual phenomena: a broader category that combines texts which implicitly reflect a bilingual situation with those exhibiting code-switching. This group is further subdivided into: texts composed in language A but displaying interference phenomena; texts exhibiting code-switching; texts borrowing from a language B. The definitional scale also distinguishes among subtypes of interference (orthographic, morphological, phonetic, lexical, and syntactic), code-switching (as in J. N. Adams), or borrowing (calques and loan-shifting).¹⁸
3. Mixed-language texts: texts written in such mixed languages that the dominant language cannot be identified. 
4. Transliterated texts: similar to J. N. Adams’ category, though at times including bi-version texts or texts displaying bilingual phenomena.

Another revision was offered by M. J. Estarán Tolosa in 2016, who identified the following types:

1. Bilingual and trilingual texts: texts containing the same or similar information in two or more languages, further subdivided into: Bilingual type 1: the complete text is translated (these are generally simple texts in which onomastic information predominates); Bilingual type 2 (the most common type): the information provided in each language differs in quality or quantity.
2. Mixed texts for one statement: a single text written in one inscription with two languages, exhibiting code-switching. This type of texts is further subdivided into: Type 1, namely mixed texts with general content; Type 2, namely mixed texts in which code-switching is limited to standardised formulae.
3. Mixed inscriptions of two or more statements: more complex texts in which each language conveys distinct information.
4. Inscriptions of an indefinite type.

The challenge with these technical definitions – for which no consensus has yet been reached – lies in the emergence of different and sometimes overlapping terminologies within epigraphic studies, reflecting a tension between the technical and social dimensions of bilingualism. This distinction partially separates bilingualism as translation from bilingualism as the conveyance of a message across two languages. InN both cases, the underlying complexity – shaped by the negotiation of discourse – stems from the introduction of sociolinguistic debates on linguistic interaction, diglossia, and cultural interference into the field of Classical epigraphy. ##The technical approach facilitates the precise identification of linguistic mechanisms, whose vocabulary is increasingly being incorporated into scholarly discourse, alongside ongoing efforts to standardise terminologies and categorisations applicable across a range of historical contexts beyond the Classical Mediterranean.¹⁹

It must be acknowledged, however, that the increasingly technical nature of this approach can, in practice, hinder the integration of more nuanced historical explanations of cultural phenomena, as illustrated by the invocations of Albinus and Tiberius Felix at the beginning of this introductory chapter. Monolingual messages also reflect multilingual contexts, revealing the use of L1 to articulate ideas, expressions, or deities originating from other linguistic and cultural domains – an issue central to this volume.

In response to similar challenges, G. Tozzi has stressed the need for more flexible and less rigid technical categories to account for the variety of social situations underlying multilingualism. To this end, she proposes a new classification that departs somewhat from earlier terminology, with the aim of considering the cultural context of bilingual inscriptions. Her classification is as follows:²⁰

1. “Bilingual”²¹ inscriptions in the strict sense: identical or analogous texts, broadly corresponding to previous classifications.
2. Inscriptions with juxtaposed but distinct texts.
3. Texts in Latin or Greek that include words or expressions from the other language.
4. Latin texts written entirely in the Greek alphabet, or vice versa.

The outcome of these studies is a detailed yet varied taxonomic framework that facilitates the identification of numerous nuances related to bilingual processes, applicable across different fields and intellectual traditions, including religious studies. Nevertheless, the core social and cultural questions with which this introduction began remains, in some cases, unanswered. What was the socio-religious function of multilingual expression? Beyond the specific mechanisms of bilingual expression that have been identified, what religious implications or uses do they have in each case?

Recent advancements may not provide straightforward answers, but they offer a range of analytical tools that must be adapted to the study of religious expression in ancient sources. Yet their interpretative limitations require a multiple-perspective approach, drawing on a multidisciplinary and international framework that benefits from, rather than being constrained by, these technical debates. In this regard, research has demonstrated the crucial role of religious phenomena in processes of linguistic change, survival, and revival – phenomena also observable in contemporary examples.²² In contexts of intercultural contact, such as the Roman world, the ritual vocabulary of one language frequently intersected with that of another cultural repertoire. Consequently, multilingualism became a pervasive phenomenon, exerting varying degrees of influence on religious texts: from monolingual inscriptions produced in multicultural settings to bilingual texts, which also display a wide spectrum of nuances. In light of the above, our volume advances the state of the art by adopting a multifaceted approach that integrates diverse perspectives and applies the bilingual taxonomy where appropriate, with the aim of elucidating the religious implications behind multilingual contact and expression.

Building on this approach, it is essential to emphasise that the study of bilingualism and religious expression cannot be reduced to the analysis of bilingual mechanics alone. Rather, it must be situated within broader frameworks, such as comparative epigraphy, which offers a complementary perspective on the linguistic and cultural dynamics at play. This methodological angle, though only occasionally applied to religious inscriptions, has been notably developed in the works of M. J. Estarán Tolosa and G. de Tord Basterra.²³ Both scholars have classified the structural elements of religious expression within distinct sociolinguistic domains of the western Mediterranean, identifying diverse repertoires of religious means. This comparative approach serves as an ideal complement to the discussions of bilingual taxonomies and is reflected in several contributions to this book. While a specific taxonomy is insufficient on its own, its combination with a comparative approach offers deeper insight into the sociolinguistic mechanisms by which individuals articulate religious expression. These social techniques include the use of formulae and concepts that can be compared and subsequently integrated into bilingual contexts – whether for the (in)exact translation of a message into a second language, the enhancement of a religious message through the use of two languages, the incorporation of a borrowed expression, or even the adaptation of foreign religious terms into the primary language.

The analysis of multilingualism also helps to understand, and is better explained through, the interaction of cults and their evolution, which often results in texts falling within J. N. Adam’s category of texts which implicitly reflect a bilingual situation. In the Graeco-Roman world, diverse cult traditions coexisted within the same religious system, presenting complex realities that that resist straightforward definitions of a cult. This complexity was already noted by J.-F. Toutain in 1907, who indirectly addressed it through a then-novel categorisation of Roman, local African, and so-called “oriental cults”.²⁴ Although this artificial division has since been abandoned – with the category of “oriental cults” now recognised as a modern construct burdened with preconceptions –²⁵ it remains true that the Roman religious world was shaped by a multicultural and multilingual horizon, where forms of worship shifted and merged across different cultural references. Within this cultural and linguistic hotchpotch, the development of the imperial cult from the 1st c. AD helped the creation of a common identity among citizens, achieving remarkable success in integrating into and interacting with local religious landscapes.

The interplay between religious mechanics, concepts, divine onomastics, and their adaptation is a multifaceted subject that has been studied for decades, albeit rarely in a unified way. A central aspect of this process is the phenomenon (commonly) referred to as interpretatio, the “assimilation” and “reinterpretation” of foreign cults in the Roman world, which has recently been the subject of renewed scholarly reassessment. The term cannot be reduced to a simple act of translation or a top-down imposition, because it encompasses mechanisms that extend well beyond the linguistic substitution of divine names and central agents.²⁶ Indeed, we often encounter cross-cultural expressions of divine naming strategies that draw upon elements from various cultural and linguistic spheres through monolingual texts.²⁷

This volume builds on this idea in its rationale, exploring these processes through the attestation of divine names in multilingual contexts, in order to shed light on how deities were represented and conceptualised across different cultural environments. The focus, therefore, is not simply on one cult and multiple languages, but rather on the intersection – within a single religious expression – of multiple cultural systems, which may or may not be manifested through their respective languages. Drawing on J.-P. Vernant’s thesis of puissance divines (“divine powers”), J. Rüpke has argued that deities do not possess inherent names; rather, their names are conferred by the cultor in the act of religious communication. The naming relies on the invocation and articulation of specific divine attributes, shaped by the occasion, context, and intention of the worshipper.²⁸ This perspective carries significant implications for multilingual expressions, bringing us back to the central question of this volume: the negotiation – whether conscious or unconscious – of religious discourse in bilingual inscriptions or regions of cultural contact. Such an approach to religious expression is further developed throughout this volume, as it constitutes a line of inquiry that speaks directly to the overarching theme of this book, that is: how religious worship did not rely exclusively on a single culture or tradition, and how multilingualism and multicultural exchange were fundamental to the development of religious practices in the Roman world.

This book addresses these issues from a holistic perspective that helps to explain, in different modern languages, the social mechanics underlying the selection of the religious message. The identification of bilingualism and bilingual modes of expression is no longer regarded as an end in itself; rather, it is integrated as an additional tool, applicable to various approaches to the study of religious cult – understood, perhaps in a broad but nonetheless useful sense, as encompassing the full range of religious practices of the Roman Mediterranean, in order to allow for a wide-ranging overview.

The volume originated from a panel held at the 14th Celtic Conference in Classics, at the University of Coimbra in July 2023. Although the initial focus was on the Roman western Mediterranean, the examination of parallel dynamics of cultural exchange and comparative cases has led to an expanded and enriched version, with contributions extending to other areas. While the leitmotiv of the volume remains epigraphic material, this does not preclude the transversality of the themes and approaches explored. Rather, the aim is to present the different dimensions through which epigraphic religious expression was influenced by multilingual and multicultural interactions. Accordingly, the book offers a series of paradigmatic case studies, organised into three complementary sections, examining the development of the phenomenon in the west of the Roman world (I), its central areas (II), and environments in which inscriptions reveal contact with eastern religious expressions (III).

Lorena Pérez Yarza, Uniwersytet Warszawski. lorenapy@yahoo.com. ORCID: https://orcid.org/0000-0001-6492-3419. This work has been written within the framework of two projects. 1) Proyecto de I+D+I de Generación de Conocimiento Religión: el Individuo y la Communitas (RICO), PID2020-117176GB-I00, funded by the Spanish Ministry of Science and Innovation and the State Research Agency. 2) ERC StG 101040152, funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Council Executive Agency (ERCEA). Neither the European Union nor the granting authority can be held responsible for them.

Javier Herrera Rando, Universidade de Lisboa – UNIARQ. jherrerarando@gmail.com. ORCID: http://orcid.org/0000-0001-7524-0656. This work has been written within the framework of the research project The Birth of the Epigraphic Culture in Roman Lusitania, Fundação da Ciencia e Tecnologia (2022.03547.CEECIND / DOI: 10.54499/2022.03547.CEECIND/CP1762/CT0002).

Sofia Bianchi Mancini, Max-Weber-Kolleg – Universität Erfurt. sofia.bianchi_mancini@uni-erfurt.de. ORCID: https://orcid.org/0000-0002-6167-0986. This work has been written within the framework of the sub-project A01 Ambiguous Property: From Late Antiquity to the Middle Ages of the Collaborative Research Centre TRR 294/2–424638267 Structural Change of Property, funded by the Deutsche Forschungsgemeinschaft (DFG).

Notes