Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

date: 17 September 2021

Multilingualism in the Roman World

Abstract and Keywords

Situated at the intersection of various academic disciplines, multilingualisim has attracted the attention of several categories of researchers specializing in studying the ancient Roman world, including historians, philologists, linguists, literary scholars, epigraphists, and papyrologists. While Greek-Latin bilingualism, which constitutes the most remarkable aspect of ancient multilingualism, has been the subject of landmark, comprehensive studies, and although it is now well described and theorized, there still is not a single study on the relationships between Latin and the other languages of the Empire that takes into account the various aspects of the question and synthesizes the data that have emerged from these disciplines. Looking closely at the lexicon used by the Romans to describe the phenomenon of multilingualism and linguistic interference also shows that, contrary to a commonly held opinion, not only did the Romans not demonstrate an “incuriosity” regarding foreign languages, but they even developed their own linguistic reflection on the subject.

Keywords: languages, Latin, multilingualism, bilingualism, identity, communication, linguistic interference

Vox diuersa sonat populorum, tum tamen una est,/cum uerus patriae diceris esse pater (Mart., Spect. 3.11–12). “All these peoples speak different languages, and yet it is in unison that they proclaim you the true father of the homeland.”1 In these two lines, the poet Martial effectively expressed the constitutive essence of multilingualism:2 the dialectic between the one (una) and the many (diuersa), the relationship between the self and the other, between the center (geographic and political) and the periphery, and the ideological constructs that these engender. Multilingualism was an unavoidable and integral fact of the Roman world. The Romans observed and described it, calling it uariae linguae, and they conceptualized it with the abstract noun phrase uarietas (or diuersitas) linguarum, referring to the “diversity of languages” in use simultaneously across the Empire.

Bilingualism and Multilingualism: The State of the Question

Situated at the intersection of various academic disciplines, multilingualisim has attracted the attention of several categories of researchers who specialize in the study of the ancient Roman world, including historians, philologists, linguists, literary scholars, epigraphists, and papyrologists. While Greek-Latin bilingualism, which constitutes the most remarkable aspect of ancient multilingualism, has been the subject of landmark comprehensive studies,3 as well as a vast bibliography4, for forty years, there still is no corresponding corpus on the relationships between Latin and the other languages of the Empire.5 Over the past two decades, this course has been charted by the work of B. Rochette,6 among others, who takes a historical-philological approach, and J. N. Adams (Adams, 2003a), who takes a sociolinguistic perspective in his substantial, 836-page study on bilingualism, which unfortunately lacks an index of the texts examined. The chapter that Adams devotes to the contacts that Latin maintained with languages other than Greek7 brings together for the first time a collection of ancient evidence that can now be included in multilingualism’s historical record. Added to the information transmitted by the texts is a rich and selective8 corpus that presents precise analyses of epigraphic and papyrological documents, allowing us to access new, more informal levels of language.

Now what is needed is a comprehensive study that takes into account all aspects of the question and synthesizes the mass of individual data that continue to be brought to light and scrupulously analyzed in the various disciplines. The task is complex and delicate. The study of ancient multilingualism requires competence in the languages of antiquity, of which our knowledge is sometimes only fragmentary and epigraphic (as is the case with the Italic languages). Each written record constitutes, in and of itself, a specific case that necessitates both a microstructural and macrostructural study. Each document must first be placed in its historical and enunciative context, and then the linguistic analysis must operate at the different levels of the writing’s constituent elements, based on the text, understood globally and according to the usage of the languages and writing systems that it brings together, and that may be subject to linguistic interferences,9 down to the morphemes, graphemes, and phonemes. The study of an epigraphic document demands an awareness of the phases of its production (Purnelle, 1995): the creation of the “original,” personally put down in writing by the competent authority or orally dictated to a secretary, who later transcribes it; a possible translation, by the secretary, if he is competent enough, or by an interpreter who is more or less trustworthy; and finally, the engraving, created by an engraver who is not necessarily familiar with the language or languages that he is reproducing, nor with the writing system. At each step of the process, which relies on speakers who have mastered, to a greater or lesser extent, the oral and, a fortiori, written codes of the languages concerned, all sorts of errors and interferences may be introduced, rendering the study of multilingualism extremely demanding and delicate.

Multilingualism always arises in diverse and unique situations. Attempts to apply to the ancient languages the categories and formalizations created by and for modern languages become perilous when they are not sufficiently grounded in a precise analysis of the documents. The indirect records are limited and partial, sometimes even bordering on the anecdotal. Transmitted across the filter of writing, they give us but a minimal glimpse into what could have been, in everyday life, the simultaneous use of different languages. It is from this perspective that we must revise the postulate, inherited from Michel Lejeune’s pioneering study, of the supposed “incuriosity” that the Romans showed toward foreign languages (Lejeune, 1949), an assumption that continues, insidiously, to influence contemporary research on the Romans’ attitude regarding foreign languages. Remarkable scientific advances in the fields of epigraphy and papyrology, along with the constant publication of the new corpora and critical studies that they produce,10 bring forth new elements every day. These should be analyzed by taking advantage of the contributions of sociolinguistics, without losing sight of the specificity of the Roman world and the incomplete nature of the documents at our disposal. Paying particular attention to the vocabulary used by the Romans to describe the phenomenon of multilingualism also allows a more precise approach. Contrary to what one may be inclined to believe, the terminology used by the Romans shows that not only were they interested in the contact between Latin and foreign languages, but they also developed a linguistic reflection on the subject that was truly their own.

Orbis Romanus: National Identity, Linguistic Identity

Quot gentes, tot linguae (Isid., Etym. 9.1.1)

The plurality of languages used in the Roman world11 is a historical fact attested to by numerous developments in Greco-Roman literature and, in particular, by the rich epigraphic corpus originating from different regions of the Empire. Even before the founding of Rome and the subsequent propagation of the Latin language and civilization around the world, Italy (and Latium in particular) was a meeting place for people speaking different languages: languages of proximity, cognate with Latin, such as Sabine, Praenestine, and Faliscan; languages established on Italian soil, such as Etruscan, Oscan, Umbrian, Venetic, and Messapian; and languages imported by colonizers (Greeks in particular) and through international commerce, such as Mycenaean, Greek, Phoenician, and Punic. The founding of Rome, in this multilingual and pluricultural context, is presented by Virgil in the Aeneid as resulting from an ethnic intermixing, a triple synoecism12 that propelled the language of the Latins (sermonem partrium, 12.834) to the rank of the dominant and “federative” language: omnis uno ore Latinos (12.837), “all, having become Latin, will speak in a single voice.” Rome’s imperialist politics and its territorial expansion were supported by a strong ideology that took on its full dimension in the age of Augustus. Rome, henceforth situated at the center of the universe, with the Roman people having come to dominate the world, was confronted with a diversity of peoples and languages, as we are reminded of by the topoi, which was traditional in Latin literature, of the catalog of troops or the triumph of the conquering generals:

Verg., Aen. 8.722–723: incedunt uictae longo ordine gentes,/quam uariae linguis, habitu tam uestis et armis, “the conquered nations come forward in a long procession, as diverse in their languages as in their clothing and arms.”

Lucan. 3.289: tam uariae cultu gentes, tam dissona uulgi ora, “such a diversity of clothing, such a dissonance of speech.”

International commerce, moreover, in stretching all the way to the borders of the Empire, put the Romans in contact with an incalculable number of local populations, each speaking different languages or dialects. Pliny13 recalls the port town of Dioscurias on the Black Sea, whose location at the mouth of the isthmus separating the Caspian Sea from the Pontus Euxinus attracted people from the inland region: “three hundred peoples, all speaking different languages” (CCC nationes dissimilibus linguis), which necessitated a team of 130 interpreters to conduct commercial transactions (a nostris CXXX interpretibus negotia gesta ibi). Rome, the capital of the Roman world, which went out to meet people the world over and attracted them to her, was also “the eternal City,” Roma aeterna. Even after the division of the Empire into a Latin-speaking West Empire and a Greek-speaking East Empire, its name continued to be the symbol of a political and cultural entity, that of a Roman world (Romani/[Gr.] Rhômaîoi, Romanitas/[Gr.] Rhômaiótês) that had assimilated Greek culture and made a new, distinctly Roman14 entity of it, and that, surpassing the limits of its political existence, endured well after the fall of Rome in Romania, and for a long time served as a model and a point of reference for the Western world.

Lingua Latina, Lingua Romana

The Latin language (lingua Latina), the language of Latium that became the language of Roman power and the Roman world (lingua Romana), was one of the foundations of Roman identity upon which Rome’s imperialist politics and institutions relied.15

Civitas and the Name of Rome

Along with institutions and laws, language represents one of the distinctive and fundamental traits (inter se differunt) that determine the identity of a people. One could lend a universal significance to what Caesar states at the beginning of the Gallic Wars with regard to the different Gallic peoples: hi omnes lingua, institutis, legibus inter se differunt, “they all have different languages, institutions and laws” (Caes. BGall. 1.1.2). The particular connection between language (sermo) and law (ius) as identity markers is clearly articulated by Cicero: ciues … Romani qui et sermonis et iuris et multarum rerum societate iuncti sunt, “Roman citizens are united by a community of language, of law, and many other things” (Cic. Verr. 2.5 = Suppl. 167). Just like legal status, the tria nomina, or the wearing of the toga (togati, “Romans”), the Latin language (patrius sermo, “the national language,” latine loqui, “to speak Latin”) was integral to the notion of citizenship and Roman identity, or ciuitas.16 Roman expansionism and the formation of a vast colonial empire allowed Rome and its institutions to reach as far as the most distant provinces, at the limits of the Empire and the known world, among people of very diverse languages and systems of rule:

Sall., Iug. 19.7: Mauris omnibus rex Bocchus imperitabat, praeter nomen cetera ignarus populi Romani, itemque nobis neque bello neque pace antea cognitus, “All the Mauri were under the command of the King Bocchus, who aside from their name knew nothing of the Romans, and who was likewise unknown to us either in peace or in war.”

In the De Suppliciis, Cicero evokes several times the glory of Rome and the greatness of the name of the Roman people (Vrbis gloriam, tanta nominis Romani dignitas, Verr. 2.5.100–150) which, even in the remotest regions, ensured respect and protection for every Roman citizen. The affirmation of Roman citizenship (“ciuis Romanus sum,” “I am a Roman citizen”), according to Cicero, was supposed to confer protection and status, everywhere in the world, to those able to claim it. While it granted rights, Roman citizenship also imposed a respect for the principles upon which it was based. Thus, there was a stigmatization of those who, in their official functions outside Rome, transgressed these presupposed principles of Romanitas (such as Verres during his praetorship in Sicily), as well as of foreigners who, having received Roman citizenship, proved themselves incapable of speaking the language implied by their new condition (sermonis ignarum). This was the case for a Greek public figure whose name the emperor Claudius crossed off the list of judges, returning him to the status of a peregrinus: peregrinitatem redegit (Suet. Claud. 16.4).

But evidence of this sort remains rare, and while knowledge of Latin was, in theory, a prerequisite of Roman citizenship, it does not seem that learning Latin was ever a real legal obligation for foreigners who became Roman citizens (Dubuisson, 1982a). In any case, in an empire that was largely illiterate and polyglot, interpreters and scribes existed to mitigate the difficulties and potential ambiguities of oral and written communication.

Patrius sermo: The Utopia of Linguistic Unity

As the language of the ciuis Romanus, Latin was also the language of the Roman state, the national language, the patrius sermo, and a symbol of its identity and its unity. Among the criteria that defined the various social circles and groups (gradus societatis) and the ties that bound them, Cicero grants a privileged role to the community of language (lingua) as a factor of social cohesion (coniunguntur): eiusdem gentis, nationis, linguae qua maxime homines coniunguntur (Cic. Off. 1.53), “belonging to the same race, to the same nation, to the same language which, above all, unites men.” The Latin language contributed closely to the Romans’ awareness of their larger national identity. Pliny celebrates the civilizing role (humanitatem) that Rome and Latin played in exalting terms, as uniting all peoples (cunctarum gentium in toto orbe) and all languages of the universe (discordes ferasque linguas) in a single homeland (una patria), and allowing them to express themselves in a common language (sermonis commercio, colloquia),17 a utopic vision that does not correspond to historical reality.

Even within the two institutions that are accepted as having been bastions of the Latin language (namely, the army18 and the law19), Latin was not able to resist the pressure of Greek and of local languages. Generals communicated with their troops through bilingual officers assigned to each ethnic group (Livy 30.33.8), and at the beginning of the third century, when the Edict of Caracalla (212 CE) granted Roman citizenship to all the free inhabitants of the empire, plurilingualism was officially inscribed into Roman law. Not only was it possible, in oral contracts made by stipulatio, to use any language (latina an graeca uel qua alia lingua), but it was not even obligatory for the two parties to use the same language, provided that they understood each other (intellectum), and that there was congruity between the question and the answer:

Just., Inst. 3.15.1: utrum autem latina an graeca uel qua alia lingua stipulatio concipiatur, nihil interest scilicet si uterque stipulantium intellectum huius linguae habeat. Nec necesse est eadem lingua utrumque uti, sed sufficit congruenter ad interrogatum respondere. Quin etiam duo Graeci latina lingua obligationem contrahere possunt, “It does not matter whether the stipulatio is made in Latin, in Greek, or in any other language; that is of no importance provided that the two parties understand the meaning of what is said. And it is also not necessary that they each use the same language; it suffices to provide the answer corresponding to the question asked. Moreover, two Greeks can enter into a contract in Latin.”

Here, we have an exceptional example of enunciative code-switching, based on a shift in language between the interlocutors (Wenskus, 1997). Juridical documents of all kinds, preserved in epigraphic or papyrological form, show that the use of Latin was often reduced to stereotyped formulas, even abbreviated as acronyms. While the question of multilingualism does not arise in literature, where all people conventionally express themselves in Latin (and in a formal Latin), regardless of their origin or status,20 in reality, Latin had to confront the supremacy of Greek as the language of culture and international communication. Latin failed to take root in the East, and even in the West, it did not hold an exclusive position with regard to other local languages, which continued to be used and recorded. The Etrusco-Italic languages, which were the first to encounter Roman colonization and the diffusion of Latin, completely disappeared from the Roman linguistic landscape at the beginning of the Empire, having been absorbed by Latin, which is illustrated by Quintilian’s remarks on vocabulary:

Quint. 1.5.55-56: uerba aut Latina aut peregrina sunt … Taceo de Tuscis et Sabinis et Praenestinis quoque … Licet omnia Italica pro Romanis habeam, “The words are either Latin or foreign … I’m not speaking of Etruscan, Sabine, or Praenestine words…. We can consider all that is Italic as being Roman.”

Latin was never the sole language of the empire, as it was imagined to be in the empire’s early ideology. Not only was the empire bilingual (speaking Greek and Latin), but in fact it also encompassed a multitude of dialects of differing statuses, the echoes of which can be heard even in Latin, in mixed onomastic formulas,21 and above all, in the numerous lexical borrowings that Latin made from the languages with which it was in contact, and for which Greek often served as an intermediary.22 This multilingualism would be confirmed, several centuries after the end of Roman domination, by the linguistic fragmentation of the Romance regions.23

The Perception and Expression of Linguistic Diversity

The Roman Perception of Foreign Languages

The Oral Dimension

The perception of linguistic diversity forms part of a double relationship of confrontation with the other: with individuals other than oneself (alter, homo alienus) and with foreigners (homo externus), who speak in languages other than the national language (sermo patrius). Linguistic diversity is alluded to in Latin texts by the concept of linguarum uarietas (diuersitas), and more concretely, by the production of sounds attached to linguistic signs: diuersos signorum sonos (Isid., Et. 9.1.1), as well as to differences in accent: illa gentium totque linguarum toto orbe diuersitas, hinc tot cantus et moduli flexionesque (Plin., HN 11.271), “such a variety of peoples and of languages in the world, so many accents, intonations, rhythms.” This diversity was perceived first and foremost in its oral dimension (Biville, 2009). Romans experienced it in the very heart of their city, through the crowd of emigrants and slaves that had come from all over the world: ex toto … orbe terrarum confluxerunt (Sen., Helv. 6.2), “they have flocked here from every corner of the world”; peregrinae multitudinis (Sen., Helv. 6.4), “the crowd of foreigners”; and mixtis seruitiis et dissono clamore (Tac., Hist. 1.32.1), “slaves of every origin, with discordant cries.” It is also visible in literary themes such as the depiction, in satire and epigram, of the loud and cosmopolitan audiences of the circus shows: uox diuersa sonat populorum (Mart., Spect. 3.11), “the diverse voices that resound from the people”; or the catalogs of forces involved in wars and conquered nations, traditional in epic: tam dissona uulgi ora (Lucan 3.289), “the accents, so varied, of this multitude.”

This diversity is simultaneously perceived as an undifferentiated whole and as an identifying sign of vocal recognition that makes it possible to sort everyone into his or her own faction, just as one can do with physical appearance, name, clothing, or arms: mentitaque tela agnoscunt atque ora sono discordia signant (Verg., Aen. 2.423), “they identify them by their usurped arms, and the sounds of their foreign dialects betray them.” This is the case even if it is difficult to describe the exact nature of the specific vocal traits that serve as identifying markers of nationality: sunt etiam proprii quidam et inenarrabiles soni quibus nonnumquam nationes deprehendimus (Quint. 1.5.33), “there are also specific accents, difficult to describe, that sometimes enable us to recognize the nations”; sonis homines ut aera tinnitu dignoscimus (Quint. 11.3.31), “individuals are identified by their accent, just as bronze is identified by its ringing.”

This confrontation with foreign voices led the Romans to question their own identity. In relation to others, they were situated within a binary categorization that at first was defined, within the Roman world, by the opposition between the city and the countryside (urbanitas ~ rusticitas), and then later, from an external point of view, between the Roman (latinitas) and the non-Roman (peregrinitas), between Latin and Greek (Latini ~ Graeci), and above all, between Greek and Latin united (utraque lingua) against other languages relegated to the sphere of the barbaric (barbarus). Quintilian (1.5.56) includes Italic words (Etruscan, Sabine, Praenestine) in the Roman camp (cf. the section “Patrius sermo: The Utopia of Linguistic Unity”), while Cicero (Brut. 170–171) still classifies a provincial orator, originally from Latium, among the externi, or foreigners. A dialectic is thus established between the self and the other (suus ~ alienus, nos ~ illi), between the internal and the external (uernaculus, domesticus ~ externus), the national (patrius) and the foreign (peregrinus, hospes), the natural (genuinus, naturalis) and the transplanted (aduenticius).

The boundaries are often blurred and give rise to mixed creations (mixtus, degener, nothus) and ambiguities of identity, in which ethnic identity and linguistic identity do not necessarily equate. Thus, we have the barbarian capable of speaking in Greek and Latin (cuidam barbaro Graece et Latine disserenti), congratulated by Claudius for his mastery of “our two languages,” utroque … sermone nostro (Suet., Claud. 42.2); the Roman Arrius Antoninus (Plin., Ep. 4.3.5), who speaks such a pure Greek (tam graece loqui), “more Greek than Greek,” that he reveals it to be a learned language (sermone … insiticio et inducto) and not a native one (sermone patrio); and finally, Septimius Severus, originally from Africa, whose speech betrays no Punic inflection (non sermo Pœnus), and who is more Latin (Italus, Italus) than many true Romans (Stat., Silv. 4.5.45–48).

The Postulate Inherited from Lejeune (1949)

The Roman perception of linguistic diversity is tarnished by a postulate inherited from Michel Lejeune’s pioneering study (1949), which continues to influence contemporary research: Greeks and Romans were supposedly not at all interested in foreign languages. Limited in their linguistic curiosity by “the pride they took in their civilization” (l’orgueil qu’ils tiraient de leur civilisation) (Lejeune, 1949: 61), they supposedly showed nothing but contempt for foreign languages, as for everything else that was foreign. Their attitude was characterized “at once by an intermittent curiosity for foreign words, and by a nearly complete incuriosity for the languages themselves” (à la fois par une curiosité intermittente pour les mots étrangers, et par une incuriosité à peu près complète pour les langues elles-mêmes) (Lejeune, 1949: 51).

It is true that Greek and Roman grammarians did not describe languages other than their own, but we cannot expect of them something that they were not capable of producing. It is easy to imagine what a challenge doing this would have been for them, when we can already see the difficulties that Latin grammarians had in describing their own language according to the concepts and methods developed for Greek. The language that they describe is the written language of classical Latin authors, not the spoken language, which was constantly evolving. Phonetic study never really extracted itself from a graphical approach, and it wasn’t until Priscian, in the sixth century CE, that a syntactic dimension24 began to appear.

The Romans’ supposed lack of interest in foreign languages is, therefore, entirely relative. In any case, it did not prevent the Romans from becoming aware of the existence of languages other than Latin and Greek, or from realizing that these languages were different from one another. This awareness even led to a system of thought and expression that bears witness to a certain linguistic observation and reflection and that is illustrated by a whole corpus of bilingual or trilingual documents that epigraphy and papyrology have revealed to us.

The Recognition of Alterity

The Naming of “Foreign” Languages

From the beginning of their history, the Latin people were confronted with other people speaking other languages (Greeks, Phoenicians and Carthaginians, and Etruscans and Oscans, among others), and contacts with other dialects did not cease to multiply upon conquests and the extension of the empire. All the Latin world, in the realm of language as in other areas of reflection, was traversed by a binary and Romano-centrist notion that brings into opposition the other, the foreigner, who has come from elsewhere (alienigena, alienus, exoticus, externus, extraneus, forensis, peregrinus) and “imported” (aduena, aduenticius, adsumptus, inductus), “transplanted” (insiticius, insitiuus, insitus), with the native and indigenous (domesticus, indigena, genuinus, patrius, proprius, suus, uernaculus), which meet in an intermediary category of intermixing (nothus, permixtus) and degeneration (degener) (Biville, 2011). The phrase “foreign language” is expressed in Latin as aliena lingua, alienus sermo, externus sermo, or peregrinus sermo:

Plin., Ep. 4.3.5: non coniectura eget, qui sermone patrio exprimere possis, cum hoc insiticio et inducto [= Graeco] tam praeclara perfeceris, “it is not difficult to imagine what you could have written in your national language, when we see the marvels that you have produced in this borrowed, imported language [= Greek].”

Curt., Alex. 7.5.29: iam bilingues erant, paulatim a domestico externo sermone degeneres, “they already spoke a mixed language; their native language had been contaminated by the foreign.”

Apul., Met. 1.1.5: En ecce praefamur ueniam siquid exotici ac forensis sermonis, rudis locutor, offendero, “Please, forgive me if, as an inexperienced speaker of a foreign and external language, I make some errors.”

The Romans lacked a terminological system equivalent to that of the Greeks,25 in which the compounds in [Gr.] -glôttos (-glôssos)/-phônos: alloglôttos and heteróglôssos form autonymic pairs with [Gr.] homóglôttos/-phônos, “one who speaks another language/the same language.” Instead of this specialized terminology, the Romans used a rich and heterogeneous lexicon, borrowed from lexical domains that were as varied as those of citizenship, kinship, commerce, and agriculture. Nonetheless, they did conceptualize the notion of linguistic alterity by contrasting the peregrinitas (the language’s foreign elements) with the latinitas (the true Latin, unaltered by foreign elements) and the urbanitas (the Latin of Rome).

The Integration of Linguistic Diversity into Latin

The Latin vocabulary of alterity, in its social and spatial dimensions, particularly as expressed by the opposition between the prefixes ad-, in- (importation), and ex- (exportation), could support the Romano-centrist concept of the Romans. But in addition to the fact that any reflection on the other, in terms of language just as in any other domain, can be made only by way of contrast, starting from the self and what one knows, it should be noted that, most important, this opposition prefigures that which, in sociolinguistics, is described by the notions of a “second language” (langue d’accueil), or the “target” language (language 2) and the “source language” (language 1). Whatever applies to language as a whole also may apply to its elements—the accent (sonus), the letters (the litterae peregrinae or Graecae, the “foreign, Greek letters,” y and z, added to the end of the Latin alphabet), and especially the words, in which we again find the tripartition observed in ethnolinguistic adjectives (cf. the section “The Naming of ‘Foreign’ Languages”).

The entire Latin grammatical tradition, from Varro onward, classifies Latin words as either inherited words (Latina), words borrowed from other languages (peregrina, externa), or hybrid words (notha):26 genera sunt tria: unum uernaculum ac domi natum, alterum aduenticium, tertium nothum ex peregrino hic natum (Varr., Ling. 10.69). Parallel to this method being centered on Latin vocabulary and its composition, there exists another approach, which shows that Romans also knew how to take a more direct interest in foreign languages. The lexicographic tradition frequently utilizes bilingual, antonymic formulas that place words of foreign origin alongside their Latin equivalents, with the most widely used formula appearing in the form of a pairing: “what is called X in Latin (Latini, Romani, nos) is called X’ in Greek (Graeci, illi),” or in another language (Nicolas, 2005). Thus, the generic character of the word “foreigner” (peregrinum) makes room for the precise mention of the language of origin: aedificia Numidarum agrestium quae “mapalia” illi uocant (Sall., Iug. 18.8 ), “the built structures of Numidian peasants, which they call ‘mapalia’.”

Varietas linguarum: The Assessment of Diversity

Evocations and Interrogations

Confronted in both their private and public lives with individuals who spoke languages other than Latin, Romans did not fail to mention the multilingual character of their empire and to wonder, more generally, about the origins and plurality of languages. All literature of the Roman period that was truly literary or technical is threaded with allusions to the diversity of languages spoken in the world. These are particularly present in the ethnographic elaborations that can be found in the works of geographers and historians, where they constitute a mandatory theme. It is thus that Caesar begins his narrative on the Gallic Wars: Hi omnes lingua, institutis, legibus inter se differunt (Caes., BGall. 1.1), “All these peoples differ from one another in their language, their institutions, and their laws.” For each region that he studies, Strabo takes note of the diversity of the inhabiting peoples and the plurality of languages that were spoken there, as he did for Gaul (4.1.1) and Asia Minor (13.4.7. 14.2.28). In Germania, Tacitus brings up the linguistic proximity of the Aestionum gentes and the Bretons (lingua Brittanicae propior, 45.3), as well as that of the Peucini and the Germanic people (sermoneut Germani agunt, 46.1).

This multilingualism was the object of a double perception—both ideological and practical: a source of national pride for Rome, which managed to incorporate such an incredible diversity into a single empire (Plin., HN 3.39), but also, first and foremost, a source of communication difficulties with the day-to-day use of language: tot gentium sermones, tot linguae, tanta loquendi uarietas, ut externus alieno paene non sit hominis uice (Plin., HN 7.7), “there exist, among peoples, so many dialects, so many languages, so many diverse means of expression, that a foreigner is scarcely seen as a human by others.”

The Origin of Languages and Words

Following the Greeks, and inspired by their different philosophical doctrines, the Romans27 pondered the origins of human societies and the birth of language, brought about by the need to communicate, first within the group, and later outside it. The dispersion and isolation of the earliest humans led, in each community, to the creation of specific forms of language and different appellations intended to designate analogous realities:28 quot hominum linguae tot nomina deorum […] non enim ut tu Velleius quocumque ueneris, sic idem in Italia Volcanus idem in Africa idem in Hispania (Cic., Nat. D. 1.30.84), “there are as many names for the gods as there are human languages […] it is not in fact the same for you, Velleius; wherever you go, you always have the same name, but Vulcan has different names in Italy, in Africa, and in Spain.” Thus, we have the plurality of languages, which finds another form of expression in the biblical myth of “babelization.”29

These theories on the dialectic between the one and the many are at the very heart of Roman ideology: Rome understood, in unifying the empire under a single name (imperium Romanum) and a single language (lingua Romana), how to re-create an ideal world in which everyone could speak with a single voice. This perspective likewise informed the debate on whether words were of natural ([Gr.] physis/natura) or conventional ([Gr.] thésis/impositio) origin, which can be found throughout all the Latin grammatical literature since Varro. A few rare categories of words could have arisen from a natural, universal language, such as primary interjections, phonic gestures at the margins of articulated language, which are by nature (naturaliter) common to all languages (communes … omnium gentium uoces),30 or the sonically imitative, onomatopoeic formulations. All other words were created by a conventional act, specific to the people in question, and thus fall within the purview of multilingualism and the lexical borrowings that languages can effect between one another.

Denominations and Conceptualizations

This linguistic diversity may be rendered, in a quantitative and undifferentiated manner, by a generic plural that lends to multilingualism its full dimension: quot hominum linguae tot nomina deorum (Cic., Nat. D. 1.30.84), “as many names for the gods as there are human languages”; quot gentes, tot linguae (Isid., Etym. 9.1.1 ), “as many languages as there are peoples”; and omnium linguarum histriones (Suet., Iul. 39.1, Aug. 43.2), “actors performing in all languages.” But it is primarily expressed through a vocabulary of difference, via the use of the prefix dis- (e.g., differre, diuersus, dissimilis, dissonus) and the adjective uarius (in opposition to idem): hi omnes lingua … inter se differunt (Caes., BGall. 1.1), uox diuersa sonat populorum (Mart., Spect. 3.11–12), CCC nationes dissimilibus linguis (Plin. HN 7.15); tam uariae cultu gentes, tam dissona uulgi/ ora (Lucan 3.289–290), “so many peoples with languages as diverse as their dress”; and uaria adhortatio erat in exercitu inter tot homines quibus non lingua, non mos … eadem esset (Livy 30.33.8), “the orders were given in diverse languages in this army composed of so many men who did not speak the same language.” Likewise, it was conceptualized through the abstract substantives uarietas and diuersitas: loquendi uarietas (Plin., HN 7.7), linguarum diuersitas (Plin., HN 11.271).

The Stereotype of the so-called Barbaric Languages

Dissonance and the Absence of Articulated Language

As languages that were “other” in their relationship to Latin, and diverse in their plurality, foreign languages also were generally relegated to the generic category of barbaric31 languages. They did not even deserve to be taken into account and differentiated because, unlike Latin and Greek, they were languages neither of international power nor of international learning, and above all, because very few people could understand them: barbara lingua (Sall., Iug. 18.10: Libyan), sermonis barbarici (Amm. 18.2.2: Germanic), and de ore barbaro (Hier., Hil. 13.7: Francam linguam). They are often presented as purely sonic creations; they are only sound, sonus (Ov., Tr. 3.8.37), and this sound is discordant to Latin ears and dissonant in its diversity: tam dissona uulgi ora (Lucan 3.289), has nationes dissonas et multiplices (Amm. 23.6.75); and dissono clamore (Tac., Hist. 1.32.1).

The barbarian, *bar-bar-us (an onomatopoeic word, which is called in French “mot impressif,” with reduplication), is perceived of as someone who can emit only the most simplistic sound of the human language, the bilabial articulation *ba, which occurs in animal sounds (balare, “bleat,” baubari, “bark”), and also has speech issues (balbus, batulus, bambalo, “stammer”), foolish stupefaction (battare, babae), empty speech (babire, babulus), and idiocy (baburrus, babosus). While exiled in Corsica, Seneca (Pol. 18.9) can only hear sordid gibberish around him (barbarorum inconditus … grauis fremitus circumsonat), which is so grating that it bothers even the ears of somewhat civilized barbarians (et barbaris quoque humanioribus grauis).

An evolution, nevertheless, took place in late and Christian Latinity. In place of the term barbarus, which rejects any specificity of identity, the ethnic adjectives genticus, gentilis, and gentilicius (Amm. 24.4.23) were substituted little by little, and Isidore (Etym. 9.1.8) sketches a brief (even caricatural) typology based on the ways in which consonants are articulated. The Eastern peoples (i.e., Hebrews and Syrians) had a guttural articulation (in gutture, linguam et uerba conlidunt); the Mediterranean peoples (i.e., the Greeks and the people of Asia Minor) had a palatal articulation (in palato sermones feriunt); and the Western peoples (i.e., the Italians and Spaniards) had a dental articulation (uerba in dentibus frangunt). This tripartition in fact covers the three sacred languages, Hebrew, Greek and Latin, that dominated the Judeo-Christian world: tres sunt autem linguae sacrae, Hebraea, Graeca, Latina, quae toto orbe maxime excellunt (Isid., Etym. 9.1.3).

In the literary tradition, foreign languages gave rise to typological stereotypes. Greek benefited from an ambivalent reputation; while it enjoyed a sonic prestige (uerba … bene sonantia, splendore Graecorum, Cic., Orat. 163) commensurate with its cultural prestige, being more “resounding” than all other languages, including Latin (sonantior, Isid., Et. 9.1.4), it very quickly lost its prestige when it was used on an everyday basis by uneducated speakers. It was thus placed on the same level as all other languages indifferently qualified as “barbaric” or “savage” (uox fera, Ov., Tr. 5.7), which were the source of disagreeable auditory impressions (offendere): nullus sonus unquam acidior percussit aures meas … me etiam Vergilius offenderit (Petron., 68.4–5), “never before had such a shrill sound assaulted my ears … even Virgil was annoying to me,” or even sources of fear (timor): omnia barbariae loca sunt uocisque ferinae, omnia sunt Getici plena timore soni (Ov., Tr. 5.12.55), “everything is invaded by barbarism and savage voices, everything is filled with the terror that the Getic accent inspires.”

Voice, along with physical aspect and comportment, are components of the traditional representation of the barbarian, and voice is by nature trux (savage and frightening): Seras … excedere hominum magnitudinem, rutilis comis, caeruleis oculis, oris sono truci, nulllo commercio linguae (Pliny, HN 6.88), “the Seres are larger than normal size, they have red hair, blue eyes, a frightening-sounding voice, they do not communicate with anyone”; and gens truci cantu clamoribusque uariis horrendo cuncta compleuerant sono (Livy 37.8), “this people, with their savage song and dissonant clamors, fill the space with a frightful sound.” Empty of meaning for the Romans, who often perceived them in the context of war, these barbaric voices were reduced to the production of sound (sonus) that can be measured only according to universal, acoustic parameters—its intensity and quality, its rhythmic patterns.

The only sounds that could be identified as meaningful were shouts (clamores), howls (ululatus), and songs (cantus), whether combative or festive: ululatus cantusque dissonos, dissonant howls and songs (Livy 39.5), clamore quoque ac sui moris ululatu, shouts ans howls, according to their custom (Curt. 9.4.24), laeto cantu aut truci sonore, “joyous songs and savage sounds” (Tac., Ann. 1.65.1), as was the case with the famous song of the Germanic people, barditus (or barritus), whose dynamic power was described by Tacitus in Germania as being equal to a force of nature.

Barbarian/Nonbarbarian: Evolution of the Concept and the Place of Latin

The opposition between the barbaric and the nonbarbaric, initially devised by the Greeks to describe linguistic diversity and to contrast Greek with all other languages, progressively took on more distinctly cultural connotations, turning barbarians into unsophisticated (rudes) beings who were culturally inferior. Latin’s ascendance to the international stage and its status as a language of power led to a modification of the original notion, which changed from a binary to a ternary structure before reverting to a bipartite opposition. In the meantime, Latin was at first considered to be a barbaric language by the Greeks (and by the Latin people themselves32), just like all other languages:

Cato ap. Plin., HN 29.7.14: nos quoque dictitant [= Graeci] barbaros, et spurcius nos quam alios,‘opicos’ (var. [Gr.] Opikôn) appellatione fœdant, “[the Greeks] regularly treat us too as barbarians, and they demean us even more disgracefully than the others by calling us Opici.”

Later, in an axiological reversal it would side with the Greeks (Greek and Latin, in opposition to all other languages), before including Greek in the notion of Romanitas and contrasting all that was Roman with all that was not: Romani/barbari, [Gr.] Rhômaîoi/or Rhômaîoi.33

Languages Recognized by the Romans

How many languages, and what languages, were spoken in the Roman Empire? It is impossible to say. There is no reason to believe that the Romans could have created such an inventory, but then again, not everything the Romans wrote has reached us. All we can do, therefore, is to collect, from among all of the Latin documents at our disposal, the sparse and disparate mentions that are scattered throughout. Comprehensive, systematic work remains to be done.

The Naming of Languages

Latin employed a varied, denominative system for designating languages34 (thus, for Latin, lingua latina, lingua Romana, sermo latinus, lingua Latinorum) that was founded on ethnic usage, ktetic adjectives, or toponyms. “In Latin” can be expressed as apud Latinos (“among the Latins”) or in Latio (“in Latium”), as well as by the phrase Latini dicunt (“the Latins say”), the adjective Latinus, -a, -um, (“Latin”) or the adverb latine (“in Latin”). Because the names of peoples are also used to designate their languages, it is not always easy to make the distinction between that which is properly linguistic and that which is more broadly ethnic.35

For our purposes, we will consider as referring to language any statement that associates an ethnicity or a ktetic with an explicit mention of language (lingua, sermo) or a speaking verb (loqui), or that figures in a linguistic formulation of the autonymic (dicitur, appellatur) or etymological (ab, ex) variety. We also will consider as referring to linguistic realities any ethnic or ktetic usage that forms part of a derivational chain containing expressions of the following types: Latium (in Latio dicitur, “in Latium they say”) → Latini (Latini dicunt/appellant, “the Latins say/call”) → adj. latinus, -a, -um (“Latin”) → Latinorum lingua or latina lingua (“the Latin language”), latina uerba ou uocabula (“Latin words”), and the adverb latine (“in Latin”).36 The existence of the adverb is particularly significant. It is corroborated in the writings of Varro and Titinius, who list, and place on the same level, a group of seven languages (namely, Oscan, Etruscan, Greek, Gaulish, African, Volscan, and Latin):

Varr., Ling. VIII ap. Gell., NA 2.25.8: ab Osco Tusco Graeco, osce tusce graece, a Gallo tamen et Mauro, gallice et maurice, “from Oscus, Tuscus, Graecus, we get osce (‘in Oscan’), tusce (‘in Etruscan’), ‘graece’ (‘in Greek’), but from Gallus and from Maurus, we get gallice (‘in Gaulish’) and maurice (‘in African’).”

Greek, for its part, names languages using a lexical microsystem that associates with a nominal syntagma a verb ending in [Gr.] –zô, an adverb ending in –istí, and an abstract ending in –ismós. This system is echoed in Latin: [Gr.] hellênikê phônê, “Greek language”/helllênízô, “to speak Greek”/hellênistí, “in Greek”/hellênismós, “Greek identity”; Graeca lingua/graece loqui, graecissare, [Gr.] graik-ízô/graece, graik-istí, graecitas; Latina and Romana lingua/latine loqui/Latinitas and Romanitas; [Gr.] rhôma-ízô/rhoma- istí/rhôma-iótês. Latin borrowed from Greek the suffix category of verbs ending in -isso/-izo, but only four rare and late forms confirm this linguistic sense: barbarizare and opicizare, “to speak like a barbarian”; graecissare, “to speak Greek”; and latinizare, “to translate into Latin”:

Apul., Apol. 98.8: loquitur numquam nisi punice et si quid adhuc a matre graecissat; enim latine loqui neque uult neque potest, “He only speaks Punic, even though he still knows a little Greek, which he learned from his mother; as for Latin, he neither wants to nor is able to speak it.”

The category of abstract feminine nouns ending in –tas allows us to conceptualize some aspects of linguistic diversity (diuersitas, uarietas, and peregrinitas) and specificity: Latinitas, Romanitas, Graecitas, and Patauinitas. The only live category in Latin is that of adverbs ending in -e/-ice,37 which, associated with the verb loqui, constitute the usual means of identifying a spoken language: graece loquantur et latine et gallice (Varr., ap. Isid., Etym. 15.1.63), “they [the people of Marseille] speak Greek, Latin, and Gaulish”; qui obsce et uolsce fabulantur, nam latine nesciunt (Titinius, Quintus 104 ap. Fest. 204, 28–30L), “They speak Oscan and Volscan because they do not know Latin.”

Thus, we find, in addition to graece and latine: celtice (Sulp. Sev.), gallice (Varr.), getice (Ov.), iudaice (Vulg.), maurice (Varr.), o(b)sce (Titin.,Varr.), persice (Quint), punice (Plaut.), sabine (Varr.), sarmatice (Ov.), syriace (Vulg.), tusce (Fest.), and uolsce (Titin.). All these languages consequently enjoy the status of being recognized by the Romans. (We are not claiming that this is an exhaustive list, though.)

Alphabetic Inventory

The classification of languages adopted by J. N. Adams, in his examination of the languages that were in contact with Latin,38 combines ancient designations with modern typologies. He successively studied Oscan, Umbrian, Venetic, and Messapian; Etruscan, Celtic (Gaulish), and Punic; Libyan and Berber; Hebrew; Germanic; Hispanic languages; Egyptian; Getan and Sarmatian; and finally Thracian. To avoid inflecting the ancient data by projecting onto it a posterior form of categorization, it seems preferable to present the languages in alphabetical order based on their Latin names, which has the advantage of putting all the languages on the same level, as the Romans did, without making assumptions regarding their respective ties (aside from Greek, the distinction between “language” and “dialect” is not pertinent in Latin) and their territorial expansion. The proposed list, most likely far from exhaustive, is given here only as a guide:

  • Aestiorum gentes … quibus…lingua Britannicae proprior (Tac.).

  • apud Afros appellatur, Afrorum lingua (Fest.).

  • e lingua Armenia (Varr.).

  • Assyrius sermo (Ulp.).

  • lingua Britannicae proprior (Tac.).

  • Celtice loqui (Sulp. Sev.), Celtarum lingua (Auson.).

  • sermonem Chaldaicum (Hier.), Chaldaeorum lingua (Hier.).

  • [gr.] homóglôttoi d’eisìn oi Dákoi toîs Gétois (Strabo).

  • linguam Etruscam (Livy), cf. “Tuscus”.

  • Francam linguam (Hier.).

  • lingua Gallica (Enn.), loquantur…gallice (Varr.), “tau”Gallicum (Cimber ap. Quint.), Gallico sermone (HA, Alex.), Gallice dicitur (Marcellus), et passim.

  • sermone…ut Germani agunt (Tac.), sermonis Germanici (Sid.).

  • Getice loqui, Getico sermone, sono (Ov.).

  • “eils” Goticum (Anth. Lat.).

  • lingua Hebraea (Vulg), Hebraeis uerbis (Aug.), sermoni Hebraico (Hier.).

  • “gurdos”…ex Hispania duxisse originem (Quint.).

  • Indorum lingua (Plin.).

  • Italica [uerba] (Quint.), Vituli ab Italis “itali’” sunt dicti (Fest.).

  • Iudaice (Vulg.).

  • Maurice (Varr.).

  • Osce (Varr.), gnari Oscae linguae (Livy), Oscorum lingua (PF).

  • Persice loqui (Quint.), Persica lingua (Plin.), litteris sermonique Persarum (Nep.).

  • Praenestina [uerba] (Quint.).

  • Punice (Plaut.), lingua Pœnorum (PF), sermo Pœnus (Stat.), punica lingua (Plin.) et passim.

  • Sabine, Sabina lingua, Sabini (Varr.), et passim.

  • lingua Samnitium (Gell.).

  • Sarmatice loqui (Ov.).

  • Scythico sermone (Trog. Pomp.).

  • “momar” Siculi stultum appellant (PF).

  • Syro sermone (Hier.), sermo Syrus (Vulg.), Syra uerba (Hier.), nomine Syriaco (Varr.), syriace (Vulg.), siriste (Eger.).

  • [gr.] tôn Getôn homoglôttou Thraxìn éthnous (Strabo). Threicio…circumsonor ore (Ov.). Tusce…dicitur (Fest.). dicunt…Tusci (Varr.), Tuscis [uerbis] (Quint.), et passim.

This list, which remains open, is already sufficient to show the attention that the Romans paid to foreign languages. It would be interesting to compare it with the list of foreign peoples that they inventoried. The study of the relative frequency with which each language is mentioned reveals its degree of proximity and familiarity to the Romans. As for the chronology of the described attestations and situations, it bears witness to the evolution of the linguistic landscape and the contacts between languages.

A Comparative Approach

The interest that the Romans showed in foreign languages went beyond these simple mentions. In addition to vocabulary, which they evoked through the filter of Latin and its borrowings, they occasionally described, by way of contrast with Latin and Greek, the specific traits of certain languages, such as the absence of the vowel o in Umbrian and Etruscan39 or the absence of the neuter in Semitic languages.40 Such remarks can seem anecdotal, but they prove that this sort of reflection did exist, and most likely originated in works that are now lost to us. They show that the Romans pondered the relationships between languages.

Festus (226.30–228.1 = PF 227.2) echoes the etymological debates that were stirred by the Latin compound petor(r)itum, a “four-wheeled” chariot (petor-ritum = quattuor et rota), which is attributed sometimes to Gaulish, sometimes to Oscan, or else to Greek. It is remarkable that a formal link was made between “four” in Latin (quattuor), in Gaulish (petor), and in Oscan (pitora), according to a perspective that, many centuries later, would reemerge in the comparative grammar of the Indo-European languages.41

Strabo42 mentions the commonality of language ([Gr.] homóglôttoi) among the Getans, the Thracians, and the Dacians. Tacitus, in his Germania, notes the proximity of the language of the Aestii with Breton: Aestiorum gentes … quibus … lingua Britannicae proprior (45.3), as well as the resemblance that the language of the Peucini shared with that of the Germanic people: Peucini … sermone … ut Germani agunt (46.1). Punic, Chaldean, Hebrew, and Syriac are said to be similar (similis) by Priscian (GL 2.148.1–3), and Jerome takes advantage of this proximity (uicina) between Chaldean and Hebrew in order to translate Chaldean into Greek, using Hebrew as an intermediary language: quia uicina est Chaldaeorum lingua sermoni Hebraico.43 The seeds of comparative and contrastive grammar can already be found in the Roman world, and they are not limited to the comparison between the Greek and Latin languages.

Multilingual Communication in the Roman World

Commercium linguae

Linguistic Diversity and Multilingualism

The Roman Empire was multilingual in the broad sense of the word because it was meant to unite, under its authority, a multitude of peoples and individuals speaking different languages. But it was not truly multilingual, in the more strictly technical sense of the term, until these different languages found themselves in contact with one another in situations of linguistic exchange (commercium linguae/sermonis). We cannot really speak of multilingualism until the diversity of languages becomes reality, “in action,” in a concrete communicative situation, and within a single entity. Individuals, but also communities,44 cities, and regions,45 as well as texts, can be bilingual or trilingual (and as an exception, polyglot).

Ovid’s personal experiences during his exile at the city of Tomi, on the shores of the Black Sea between 9 and 17 CE, which he wrote about in the poems of the Tristia and Pontica, constitute from this point of view an exceptional testimony, a first hand account. He allows us to glimpse, “in direct,” the manner in which the confrontation and cohabitation of languages in this ancient Greek colony might have been lived and perceived on a daily basis: Greek mixed with Getan (mixta, Tr. 5.7.11. getico barbara facta sono, Tr. 5.7.52); dominant local languages, mostly Getan (Getico sono, ore, passim. Getico sermone, Pont. 4.13.19), but also Sarmantian (sarmatice loqui); and total ignorance of Latin, which led to an axiological reversal in the representation of the barbaric: barbarus hic ego sum, quia non intelligor ulli/et rident stolidi uerba Latina Getae (Tr. 5.10.35–38), “here I am barbarous, because no one understands me, and they make fun of my Latin words, these stupid Getans.” Ovid eventually moved beyond this negative and undifferentiated vision of barbarity (nesciaque … uocis … barbara lingua latinae, Tr. 5.2.67) and applied himself to the study of Getan and Sarmatian (didici Getice Sarmaticeque loqui, Tr. 5.12.58).

The Difficulties of Interlinguistic Communication

The diversity of evidence that can be gleaned from the literature of the Roman period, whether in Greek or Latin and whether properly literary or technical, as well as the documentary richness of epigraphic and papyrological texts, unevenly dispersed across the different regions of the empire, show that the Romans grasped the problem of interlinguistic communication, as well as its difficulties. They were ceaselessly confronted by it, concretely and daily, in their political and administrative management, their military command, their commercial exchanges, and more simply in their private lives, as they came into contact with slaves, foreign communities, and local populations. They observed it, they reflected upon it, and they described it using a vocabulary that, while not exclusive to multilingualism, was in any case specific to it.

Well before the apparition of sociolinguistics and the linguistic modeling of languages in contact with one another, the Romans perceived the degrees of linguistic mastery involved in speaking a foreign language, which modern linguistics describes in terms of passive or active competence, oral or written mastery, and the production of simple or complex utterances. Even leaving aside the question of knowing a foreign language, the difficulties of intercomprehension begin with the mere pronunciation of proper names, specific to each language and endowed with an indispensable referential function. The differences between phonological systems, as well as the pronunciation difficulties to which they give rise, can carry heavy consequences when, during a trial, for example, it is necessary for a witness to recognize a named individual. Quintilian thus brings up two complementary cases—one of a Greek man incapable of pronouncing the Latin [f] in Fundanius, and one of a non-Greek-speaking Latin speaker who was rather unfamiliar with Greek’s aspirated stop [ph] or long, tonic vowels in hiatus (Amphío).46

Multilingual Practice

The Degree “Zero” of Multilingualism

Monolingualism: A “deafness”

Aside from cultured elites, who knew Greek, and the dialects that Romans or foreigners could learn individually in the course of their jobs or activities, it seems that foreign-language learning was generally underdeveloped in the Roman world. Monolingualism, described by Cicero as a “deafness” (surdi sumus) and an inability to understand (non intellegimus) that led to a deaf-mute behavior, was in all likelihood the most common situation. He clearly stated that his compatriots did not know Greek, and likewise that the Greeks did not know Latin, and he extended his remarks to other languages as well:

Cic., Tusc. 5.116: omnesque item nos in linguis quas non intellegimus, quae sunt innumerabiles, surdi profecto sumus, “and we are also ‘deaf’ to all the languages that we do not understand, and which are innumerable.”

The absence of verbal communication

The inability to ascribe meaning to perceived sounds (nesciero uirtutem sonis, Paul, 1 Corinthians 14.11) inscribes the perception of foreign voices (peregrinae uoces) in a relationship of ignorance and incomprehension: ignara lingua (Sall., Iug. 18.6), nescia uocis (Ov., Tr. 5.2.67). This ignorance entails a process of exclusion from the social community (alienus, alienare) stemming from the inability to communicate with others (nulla commercia linguae, Ov., Tr. 3.11.9): tanta loquendi uarietas ut externus alieno paene non sit hominis uice (Plin., HN 7.7), “there exist such a diversity of dialects that a foreigner is scarcely seen as a human by others.” And this process of exclusion carries with it a loss of humanity, a regression of man to the level of the barbarian: si … nesciero uirtutem sonis, ero ei cui loquor barbarus, et qui loquitur, mihi barbarus (Paul, 1 Corinthians14.11), “if I am not capable of understanding what I hear, I will be a barbarian to whomever I am speaking, and whoever is speaking to me will be a barbarian to me,” and even to the level of an animal: linguarum diuersitas hominem alienat ab homine [] facilius sibi muta animalia … quam illi … sociantur […] libentius homo sit cum cane suo quam cum homine alieno (Aug., Civ. 19.7), “the diversity of languages makes man a stranger to man … animals, deprived of language, communicate more easily than them … man is more at ease with his dog than with a foreigner.”

The universality of gestural communication

Interlinguistic incomprehension causes humanity to regress toward a primitive state in which the impossibility of verbal communication (commercium linguae) entails resorting to other, more universal semiotic systems. Body language, phonic noises, cries, and facial and gestural (per gestum) mimicry replace words and even any sound production. For Quintilian, gesture is a form of universal language (omnium hominum communis sermo) that serves as a solution to the diversity of languages (linguae diuersitate).47 This is what Ovid is reduced to in his exile: exercent illi sociae commercia linguae, per gestum res est significanda mihi (Ov., Tr. 5.10.35-36), “they communicate with each other in their language, but as for me, I must express myself through gesture.” This is also what the Seres practiced, in a form of so-called mute barter: nulla sermonum uice, solis oculis aestimantur (Amm. 23.6.68), “they do not need to speak, a simple glance suffices.” Humanity finds itself dehumanized and reduced to an animal state (magis bestias quam homines, Aug., Civ. 16.8.1).

Polyglossia and Trilingualism

At the other extreme of the spectrum of multilingualism, polyglossia was seen as an anomaly and as mirabilia (marvels). The Greco-Latin tradition48 took pleasure in evoking the quasi-mythical figures of sovereigns or leaders capable of expressing themselves in all the languages (most likely dialects) spoken by the people under their authority, such as Mithridates (twenty-two or twenty-five languages) or, in the Greco-Roman world, Crassus and Cleopatra.

Quintilian, at the end of the chapter that he dedicates to memory (11.2.50–51), casts doubt upon the idea that such an achievement could be possible in his time. The concept of polyglossia applies more realistically to social entities than to individuals. Given their composition, with their auxiliary troops, their mercenaries, and their slaves, ancient armies formed a nice illustration of ethnic (tot homines) and linguistic (non lingua … eadem) diversity. Generals communicated with their troops in different languages (uaria adhortatio) through the intermediary of officers assigned to each linguistic group according to their bilingual abilities (Livy 30.33.8).49

It is telling that no equivalent of the Greek term polyglôttos is to be found in Latin, and conversely, that no equivalent of the Latin word trilinguis is to be found in Greek. If Romans conceptualized the notion of trilingualism, it is because it was bound to grant a certain status to their own Latin language within the Greeks’ binary schema (Greek/other languages). If it occurs within a community (such as the army, or a cosmopolitan city like Rome), polyglossia is expressed in Latin with the adjective uarius, but it becomes omnis when referring to individuals: omnes linguas scire, “to know multiple languages,” without any indication of precisely how many languages are involved: omnium linguarum histriones (Suet., Caes. 39.1, Aug. 43.2), “actors speaking all languages/several languages.” In ordinary communication, this typically meant three languages: is omnis linguas scit (Plaut., Pœn. 112–113), “he knows all the languages” (meaning the three languages implied by the communicative situation—namely, Carthaginian, Greek, and Latin), or two, when concerning the mastery of Greco-Latin biculturalism: omnis hominem facundiae, a well-spoken man in each language (Martyr., GL 7.175.10), omnis eloquentiae praesul, the prince of eloquence in each language (Prisc., GL 2.2.31), which was the only recognized cultural model.

The notion of multilingualism encompasses bilingualism and trilingualism, but it refers primarily to the mastery of three languages, as was the case for Ennius: Quintus Ennius tria corda habere sese dicebat, quod loqui Graece et Osce et Latine sciret (Gell. 17.17.1), “Q. Ennius used to say that he had three hearts because he knew how to speak Greek, Oscan, and Latin,” and for the city of Marseille, where the languages Greek, Latin, and Gaulish were spoken: hos Varro trilingues esse ait quod et Graece loquantur et Latine et Gallice (Isid., Etym. 15.1.63). This was also clearly the case in the majority of communicative situations in the Roman world, in which local languages were used alongside Latin and Greek.

Several large trilingual inscriptions demonstrate this, such as the one celebrating Gallus, the first prefect of Roman Egypt, in 29 CE, written in Latin, Greek, and Egyptian hieroglyphics; the epitaph of Hairan, at Palmyra, from 52 CE, written in Latin, Greek, and Palmyrene, and which is the first text confirming the use of Latin by a Palmyrene; and, a trilingual inscription in Greek, Latin, and Hebrew from the sixth century, produced by the Jewish community of Tortosa in Spain.50 Private juridical acts found in Egypt and other regions of the East show this even better: The person that the act concerns often speaks only the local language, with the use of Latin limited to a formula that guarantees the validity of the act and Greek serving as the language of communication and as an intermediary.

The archives of Babatha’s transactions51 offer a series of complex, but particularly revealing, documents concerning a concrete situation of Greek, Latin, and Aramaic trilingualism in Judea, around 120 CE, in the context of Roman administration and jurisdiction. They bear witness to communicative strategies established between administrative subjects who spoke Aramaic (namely, Babatha and her husband, Judah, who served as her tutor) and the Roman, Latin-speaking authorities, as well as the mediating role played by the Greek language. Greek serves to translate the declarations, written in Aramaic, that Judah makes in Babatha’s name, as well as the agreement signed in Latin by the prefect. They show the different stages of the procedure, from the writing of the original document in Aramaic, all the way to the posting of the prefectural decree in Latin, in the basilica, and passing through the intervention of intermediaries, scribes, and translators, who multiply the number of different hands, along with the risks of linguistic interference. They highlight the function of the lingua franca, the language of interlinguistic communication (here played by Greek) and the reduced role, formulaic and artificial, of the Latin language, restricted to its official function as the language of the governing power.

Bilingualism: Utraque lingua and bilinguis

Bilingualism constitutes the most banal and common form of multilingualism. Occasionally, behind the apparent use of two languages lies a third that has served to facilitate translation and has left traces in the form of linguistic interference. While bilingualism was primarily present in its Greek-Latin form, which has been well studied and which led to a Greco-Latin symbiosis, it also potentially concerns all other languages of the empire, in multiple combinations that do not necessarily involve Latin.

Greek occupied a privileged position: It was the “other language” of the Roman world, and it largely prevailed over the sermo patrius, particularly in the East. It also often played a mediating role as a vehicular language between Latin and the local languages. It is thus that we have documented, in very diverse forms, in bilingual documents or interferences,52 Latin and/or Greek bilingualism, with languages as diverse as Oscan, Etruscan,53 Venetic, Gaulish,54 Sicel, and Elymian;55 Phoenician and Punic;56 Libyan, Syriac, Palmyrene,57 Hebrew,58 Egyptian, Coptic,59 Getan (Ovid), Thracian, Germanic, and Celtic dialects;60 or even bilingualism involving neither Greek nor Latin, such as the Pyrgi tablets, written in Etruscan and Phoenician.

These local bilingualisms and trilingualisms are becoming increasingly well studied, thanks to a reevaluation of the epigraphic and papyrological corpus and the discovery of new documents, which enable a better understanding of rare languages, for which we have only fragmentary evidence and that have disappeared from usage. The research remains open concerning these types of contacts, and diverse research programs are currently in progress. We await a comprehensive study that would clarify the exact statuses of the languages concerned, as well as a corpus of bilingual and trilingual inscriptions.

To designate the joint mastery of two languages, Latin employs two expressions of duality, bilinguis (CGL 4.292.37: duas linguas sciens), rare:61 bilingues Bruttaces Ennius dixit quod Brutti et Osce et Graece loqui soliti sint (PF 31,25–27L), “Ennius states that the inhabitants of Bruttium are bilingual because they speak both Oscan and Greek fluently,” and especially utraque lingua (uterque sermo),62 which is not limited to Greco-Latin bilingualism, particularly in its cultural dimension, but which, contrary to what is generally said, also applies to other languages: in utraque lingua (Varr., Ling. 5.74), “in both languages,” referring to Latin and Sabine; facundia sermonis utriusque clarus (Amm. 15.13.1), “reputed for his ability to express himself in both languages,” referring to Musionanus, the Praetorian prefect in the East; utriusque linguae peritissimum loquacem (Hier., In Tobit, Praef., PL 29,23–26), “who has perfectly mastered both languages” (Hebrew and Chaldean).

The duality of bilingualism inevitably leads to a third entity, which can be ideally imagined as a harmonious whole, such as with the Greco-Latin cultural ideal of utraque lingua eruditus, “formed in the two languages,” but which in reality often involves the phenomena of interference and language mixing, which is present to a greater or lesser degree according to the individuals’ linguistic competence and which must have been particularly frequent in oral communication. Pompeian epigraphy offers significant examples of this code mixing that resulted in creating a hybrid and unstable form of language, as demonstrated in the graffiti and the tablets of the banker Jucundus.63

Multilingual Competence

The Latin vocabulary of multilingualism shows that the Romans were attentive to the different degrees of competency that one could attain in the acquisition of languages (gnarus linguae, linguam nouit), whether they were acquired naturally, through the family, through immersion in a minority-language environment, or by scholarly learning. Evidence of this can be found, for the study of Greek and Latin (but also Coptic) as second languages, in pedagogical works such as the manuals (colloquia) and bilingual lexicons found in Egypt.64 We could draw up an assessment scale ranging from the total absence of knowledge, whether due to inability or lack of will: Latine loqui neque uult neque potest (Apul., Apol. 98.8), “he neither wants nor is able to speak Latin,” et ego quidem Graecae linguae perparum assecutus sum et prope nihil (Aug., Petil. 2.91), “for my part, I have acquired very little competence in Greek, none, so to speak,”—to perfect mastery: unum intra annum optime locutum esse persice (Quint. 11.2.50), “in one year he succeeded in learning to speak Persian very well,” eloquentissimus homo in Syro sermone (Hier., Epist. 17.2), “a perfect mastery of the Syrian language.”

This perfect mastery can go so far as to call into doubt the speaker’s linguistic identity, as was the case with Cicero’s aptly named friend, Atticus. A strong mastery is expressed by the adjective peritus (peritus, peritissimus linguae) or by the adverb probe: linguam Etruscam probe nouerat (Livy 9.36.2), “he knew the Etruscan language very well.” A limited understanding was expressed by the adverb uix: uix latine loquens (Hist. Aug, Sept. Sev. 15.7), “barely able to speak Latin.” The Romans likewise observed something that in the modern era would be theorized in the concepts of active (sonare, “to speak”) and passive (legere et intellegere, “to read” and “to understand”) competence: magis possum sermonem Chaldaicum legere et intellegere quam sonare (Hier., Daniel. Praef.), “it is easier for me to read and understand Chaldean than to speak it.”

Bilingual competence also can be assessed according to the degree of accuracy and, for Latin, according to one’s conformity to the codes of the standard language. But the only disparities pertinent to our study are those that resulted from interference with other languages. The errors that can be found in any strictly Latin document should not be taken into account.

The specificity of multilingualism, as it occurs in communicative situations, doubtless resides in the space between that is established in bilingual exchanges, in the zone of intersection in which the process of transfer happens. This zone can first be found in the mental space of the bilingual speaker, in which there arises, more or less spontaneously, an inter language or a cognitive “supra language” that encompasses the spoken languages and enables the transfer from one to the other.65 In the case of intermediaries (interpreters or translators), not only is the process duplicated, but it is also more conscious. More concretely, the zone of intersection specific to multilingualism is also that of a spatial-temporal unity that brings speakers of different languages together around a theme of common discourse.

This enunciative situation is always unique, and it leads to discursive productions that are likewise unique each time, which is what creates all the difficulty in interpreting and theorizing multilingualism. The Romans did not attain the same degree of theorization later reached by modern sociolinguistics, but the analyses that can be found throughout Latinity, as well as the terminology that they employ, prove that they had certainly evaluated the phenomenon, and that they knew how to translate it into words. They did not use the word *multilinguis, for which they nonetheless had a model in the Greek polyglôttos. They preferred instead the concept of trilinguis: It better corresponded to reality as they experienced and perceived it, and which often put Latin, Greek, and a local language in contact with one another. We owe thanks to the Romans for this advancement in the history of the plurality of languages.


Adams, J. N. (2003a). Bilingualism and the Latin Language. Cambridge, UK: Cambridge University Press.Find this resource:

Adams, J. N. (2003b). “‘Romanitas’ and the Latin Language.” Classical Quarterly 53: 184–205.Find this resource:

Adams, J. N. (2007). The Regional Diversification of Latin 200 BC—AD 600. Cambridge, UK: Cambridge University Press.Find this resource:

Adams, J. N., M. Janse, and S. Swain. (2002). Bilingualism in Ancient Society: Language Contact and the Written Text. Oxford: Oxford University Press.Find this resource:

Bellandi, F., and R. Ferri. (2008). Aspetti della scuola nel mondo Romano. Amsterdam: Hakkert.Find this resource:

Benelli, E. (1994). Le iscrizioni bilingui etrusco-latine. Firenze: Olschki.Find this resource:

Biville, F. (1989). “Grec et latin: Contacts linguistiques et création lexicale. Pour une typologie des hellénismes lexicaux du latin.” Cahiers de l’Institut de Linguistique de Louvain 15(1–4): 29–40.Find this resource:

Biville, F. (1990–1995). Les emprunts du latin au grec. Approche phonétique. Tome 1, Introduction et consonantisme. Tome 2, Vocalisme et conclusions. Louvain-Paris: Peeters.Find this resource:

Biville, F. (1993). “Grec des Romains, ou latin des Grecs? Ambiguïté de quelques processus néologiques dans la Koinè.” In: C. Brixhe (ed.), La koinè grecque antique, I. Nancy: Presses Universitaires de Nancy: 129–140.Find this resource:

Biville, F. (2002a). “The Graeco-Romans and Graeco-latin: A Terminological Framework for Casus of Bilingualism.” In: J. N. Adams, M. Janse, and S. Swain (eds.), Bilingualism in Ancient Society. Oxford: Oxford University Press: 77–102.Find this resource:

Biville, F. (2002b). “La dimension grecque de la dérivation latine. Interférences et emprunts.” In: C. Kircher-Durand (ed.), Grammaire fondamentale du latin, Tome IX: Création lexicale: la formation des noms par dérivation suffixale. Louvain-Paris-Dudley MA: Peeters: 353–390.Find this resource:

Biville, F. (2003). “Le latin et le grec ‘vulgaires’ des inscriptions pompéiennes.” In: H. Solin, M. Leiwo, and H. Halla-aho (eds.), Latin vulgaire—latin tardif VI: Actes du VIe colloque international sur le latin vulgaire et tardif. Hildesheim, Germany: Olms: 219–235.Find this resource:

Biville, F. (2008). “Les Institutions de Priscien, une grammaire et une culture bilingues.” In: C. Brunet (ed.), Des formes et des mots chez les Anciens. Mélanges offerts à Danièle Conso. Besançon, France: Presses Universitaires de Franche-Comté: 31–50.Find this resource:

Biville, F. (2009). “Voix étrangères dans la littérature latine.” In: M.-F. Marein, P. Voisin, and J. Gallego (eds.), Figures de l’étranger autour de la Méditerranée antique. Paris: L’Harmattan: 581–592.Find this resource:

Biville, F. (2011). “Dire le métissage linguistique et culturel à Rome.” Revista de Estudios Latinos 11: 11–31.Find this resource:

Biville, F. (2013a). “Le latin et les langues italiques dans la tradition grammaticale latine.” In: J. L. García Ramon and P. Poccetti (eds.), Sprachkontakt und Kulturkontakt im Alten Italien: Onomastik und Lexikon. 10 jahre nach Jürgen Untermanns Wörterbuch des Oskisch-Umbrischen. Pisa and Rome: Fabrizio Serra: 23–44.Find this resource:

Biville, F. (2013b). “Textes et procédures bilingues en droit romain.” In: A. C. Masi Doria (ed.), Modelli di un multiculturalismo giuridico: il bilinguismo nel mondo antico, Diretto, prassi, insegnamento. Naples: Satura Editrice.Find this resource:

Biville, F. (2017). “Le bilinguisme gréco-latin.” In: D. Petit (ed.), Lalies 37, Paris: ENS Ulm: 45–105.Find this resource:

Biville, F., J.-C. Decourt, and G. Rougemont. (2008). Bilinguisme gréco-latin et épigraphie. Lyon, France: Collection de la Maison de l’Orient et de la Méditerranée, CMO 37.Find this resource:

Carling, N. (2013). Une bibliographie critique relative au bilinguisme gréco-latin. In: M.-H. Marganne and B. Rochette, Bilinguisme et digraphisme dans le monde gréco-romain. Liège, Belgium: Presses Universitaires de Liège, 31–40.Find this resource:

Cascione, C., C. Masi Doria, and G. Merola. (2013). Modelli di un multiculturalismo giuridico. Il bilinguismo nel mondo antico. Diritto, prassi, insegnamento (2 vols.). Naples: Satura Editrice.Find this resource:

Clackson, J. (2015). “Local Languages in Italy and the West.” In C. Bruun and J. Edmonson (eds.), Oxford Handbook of Roman Epigraphy (Online Publication: 10.1093/oxfordhb/9780195336467.013.032).Find this resource:

Clackson, J., and G. Horrocks. (2011). A Companion to the Latin Language. Oxford, UK: Wiley–Blackwell.Find this resource:

Decourt, J.-C. (2008). “Le bilinguisme des inscriptions de la Gaule.” In: F. Biville, I.-C. Decourt, and G. Rougemont, Bilinguisme gréco-latin et épigraphie. Lyon, France: Collection de la Maison de l’Orient et de la Méditerranée, CMO 37, 305–319.Find this resource:

Dench, E. (2012). “Roman Identity.” In A. Barchiesi and W. Schneidel (eds), Oxford Handbook of Roman Studies (Online Publication: 10.1093/oxfordhb/9780199211524.013.0018).Find this resource:

Desbordes, F. (1991). “Latinitas: Constitution et évolution d’un modèle de l’identité linguistique.” In: S. Saïd (ed.), Hellênismós. Quelques jalons pour une histoire de l’identité grecque. Leiden, Netherlands: Brill: 33–47, included in Desbordes, F. (2007). Idées grecques et romaines sur le langage. Travaux d’histoire et d’épistémologie. G. Clerico, B. Colombat and J. Soubiran (eds), Lyon, ENS Éditions: 91–105.Find this resource:

Dickey, E. (2013–2015). The Colloquia of the Hermeneumata Pseudodositheana (2 vols.). Cambridge: Cambridge University Press.Find this resource:

Dubuisson, M. (1981). “Vtraque lingua.” L’Antiquité Classique 50: 274–286.Find this resource:

Dubuisson, M. (1982a). “Y a-t-il une politique linguistique romaine?” Ktèma 7, 187–210.Find this resource:

Dubuisson, M. (1982b). “Remarques sur le vocabulaire grec de l’acculturation.” Revue Belge de Philologie et d’Histoire 60: 5–32.Find this resource:

Dubuisson, M. (1983). “Recherches sur la terminologie antique du bilinguisme.” Revue Belge de Philologie et d’Histoire 57: 203–225.Find this resource:

Dubuisson, M. (1984). “Le latin est-il une langue barbare?” Ktèma 9: 55–68.Find this resource:

Dubuisson, M. (1985). Le latin de Polybe. Les implications historiques d’un cas de bilinguisme. Paris: Klincksieck.Find this resource:

Ferriss-Hill, J. (2011). “Virgil’s Program of Sabellic Etymologizing and the Construction of Italic Identity.” Transactions of the American Philological Association 141–142: 265–284.Find this resource:

Ferriss-Hill, J. (2014). “Varro’s intuition of Cognate Relationships.” Illinois Classical Studies 39: 81–108.Find this resource:

Flobert, P. (1988). “Lingua Latina et lingua Romana: purisme, administration et invasions barbares.” Ktèma 13: 205–212.Find this resource:

Fögen, T. (2003). Utraque lingua. A Bibliography on Bi- and Multiculturalism in Graeco-Roman Antiquity and in Modern Times. Essen, Germany: LAUD.Find this resource:

Fournet, J.-L. (2012). “The Multilingual Environment of Late Antique Egypt: Greek, Latin, Coptic, and Persian Documentation.” In: R. Bagnall (ed.), Oxford Handbook of Papyrology (Online Publication: 10.1093/oxfordhb/9780199843695.013.0018).Find this resource:

García Ramón, J. L., D. Kolligan, P. Poccetti (eds.). (2013). Sprachkontakt und Kulturkontakt im alten Italien: Onomastik und Lexikon. 10 jahre nach Jürgen Untermanns “Wörterbuch des Oskisch-Umbrischen.” Pisa and Rome: Fabrizio Serra.Find this resource:

Gitner, A. (2015). “Varro Aeolicus: Latin’s Affiliation with Greek.” In: D. J. Butterfield (ed.), Varro Varius: The Polumath of the Roman World. Cambridge Classical Journal Supplement volume 39, Cambridge: Cambridge Philological Society: 33–50.Find this resource:

Groupe Ars Grammatica (2010). Priscien, Grammaire latine. Livre XVII—Syntaxe, 1. Paris: Vrin (Livre XVIIISyntaxe, 2, in press).Find this resource:

Haase, W. (1983). A.N.R.W. II.29.2. Principat. Sprache und Literatur (Sprachen und Schriften). Berlin: De Gruyter.Find this resource:

Hadas-Lebel, J. (2004). Le bilinguisme étrusco-latin. Contribution à l’étude de la romanisation de l’Étrurie. Louvain-Paris-Dudley, MA: Peeters.Find this resource:

Kaimio, J. (1979). The Romans and the Greek Language. Helsinki: Societas Scientiarum Fennica.Find this resource:

Kearsley, R. A., and T. V. Evans. (2001). Greeks and Romans in Imperial Asia. Mixed Language Inscriptions and Linguistic Evidence for Cultural Interaction until the End of AD III. Bonn, Germany: Habelt.Find this resource:

Kramer, J. (1983). Glossaria bilinguia in papyris et membranis reperta. Bonn, Germany: Habelt.Find this resource:

Kramer, J. (1998). Die Sprachbezeichnungen Latinus und Romanus im Lateinischen und Romanischen. Berlin: Erich Schmidt Verlag.Find this resource:

Kramer J. (2001). Glossaria bilinguia altera (C. Gloss. Biling. II). Munich–Leipzig: Saur.Find this resource:

Laguna Mariscal, G. (1995). “Influencia del griego sobre el latin: guía bibliográfica comentada.” Tempus 9, 5–32.Find this resource:

Leiwo, M. (1995). Neapolitana. A Study of Population and Language in Graeco-Roman Naples. Helsinki: Societas Scientiarum Fennica.Find this resource:

Lejeune, M. (1949). “La curiosité linguistique dans l’antiquité classique.” In Conférences de l’Institut de Linguistique de l’Université de Paris, 8 (1940–1948). Paris: Boivin: 45–61.Find this resource:

Marganne, M.-H., and B. Rochette. (2013). Bilinguisme et digraphisme dans le monde gréco-romain. Liège, Belgium: Presses Universitaires de Liège.Find this resource:

Martorelli, L. (2014). Greco antico nell’Occidente carolingio. Frammenti di testi attici nell’Ars di Prisciano. Olms, Hildesheim-Zürich-New York: Olms.Find this resource:

Mullen, A. (2011). “Latin and Other Languages: Societal and Individual Bilingualism.” Clackson, J., Horrocks, G. (2011): 527–548.Find this resource:

Mullen, A. (2013). Southern Gaul and the Mediterranean. Multilingualism and multiculturalism in the Iron Age and Roman period. Cambridge: Cambridge University Press.Find this resource:

Mullen, A. (2015). “Sociolinguistics.” M. Millett, L. Revell, and A. Moore (eds), Oxford Handbook of Roman Britain (Online Publication: 10.1093/oxfordhb/9780199697713.013.032).Find this resource:

Mullen, A., and P. James. (eds.) (2012). Multilingualism in the Graeco-Roman Worlds. Cambridge: Cambridge University Press.Find this resource:

Müller, C. W., K. Sier, and J. Werner. (eds.) (1992). Zum Umgang mit fremden Sprachen in der griechisch-römischen Antike. Stuttgart, Germany: Seiner.Find this resource:

Neumann, G., and J. Untermann. (1980). Die Sprachen im römischen Reich der Kaiserzeit. Beihefte der Bonner Jahrbücher, 40. Cologne-Bonn: Habelt.Find this resource:

Nicolas, C. (2005). Sic enim appello … Essai sur l’autonymie terminologique gréco-latine chez Cicéron. Louvain–Paris– Dudley, MA: Peeters.Find this resource:

Poccetti, P. (1986). Lat. bilinguis. AION (ling) 8: 193–205.Find this resource:

Purnelle, G. (1995). Les usages des graveurs dans la notation d’upsilon et des phonèmes aspirés: le cas des anthroponymes grecs dans les inscriptions latines de Rome. Genève: Droz.Find this resource:

Rochette, B. (1997a). “Grecs, Romains, et Barbares. À la recherche de l’identité ethnique et linguistique des Grecs et des Romains.” Revue Belge de Philologie et d’Histoire 75: 37–57.Find this resource:

Rochette, B. (1997b). Le latin dans le monde grec. Recherches sur la diffusion de la langue et des lettres latines dans les provinces hellénophones de l’Empire romain. Collection Latomus 233, Brussels.Find this resource:

Rochette, B. (1998). “Le bilinguisme gréco-latin et la question des langues dans le monde gréco-romain: chronique bibliographique.” Revue Belge de Philologie et d’Histoire 76: 177–196.Find this resource:

Rochette, B. (2001). “À propos du terme díglôssos.” L’Antiquité Classique 70: 177–184.Find this resource:

Rochette, B. (2007). “Grecs, Romains et barbares: contribution à l’étude de la diversité linguistique dans l’Antiquité classique.” In: R. Oniga and S. Vatteroni (eds.), Plurilinguismo letterario, Rubbettino: 13–40.Find this resource:

Rochette, B. (2009). “Les noms de la langue en latin.” Histoire Épistémologie Langage 31 (II): 29–48.Find this resource:

Rochette, B. (2011). “Language Policies in the Roman Republic and Empire.” Clackson J., Horrocks G. (2011): 549–563.Find this resource:

Ruiz Darasse, C., and E. Luján. (eds.). (2011). Contacts linguistiques dans l’Occident méditerranéen antique. Madrid: Casa de Velázquez.Find this resource:

Smelik, W. (2012). “The Languages of Roman Palestine.” In: C. Hezser (ed.), Oxford Handbook of Jewish Daily Life in Roman Palestine (Online Publication: 10.1093/oxfordhb/9780199216437.013.0008).Find this resource:

Tribulato, O. (2013), Language and Linguistic Contact in Ancient Sicily. Cambridge: Cambridge University Press.Find this resource:

Vogt, J. (1960). Orbis Romanus. Ein Beitrag zum Sprachgebrauch und zur Vortellungswelt des römischen Imperialismus”. In J. Vogt, Orbis. Ausgewählte Schriften zur Geschichte des Altertums, F. Taeger and K. Christ(eds), Freiburg, Basel, Wien: Herder: 151–171.Find this resource:

Wenskus, O. (1997). “Codewechsel bei der stipulatio. Eine Bermerkung zur Sprachwahl im römischen Recht.” Glotta 73: 116–117.Find this resource:

Werner, J. (1992). “Die Fremdsprachenproblematik in der griechisch-römischen Antike.” In C. W. Müller, K. Sier, and J. Werner (eds.), Zum Umgang mit fremden Sprachen in der griechisch-römischen Antike, Stuttgart: Seiner: 1–20 and 233–252.Find this resource:

Wiele, B. (1979). Lateinische Sprache und römische Nationalität. Ein Beitrag zur Entwicklung des Sprachbewusstseins bei den Römern. Diss. University of Berlin.Find this resource:

Wilson, A. (2012). “Neo-Punic and Latin inscriptions in Roman North Africa: Function and display.” Mullen, A., and P. James (eds.). (2012). Multilingualism in the Graeco-Roman Worlds, Cambridge: Cambridge University Press.Find this resource:

Yon, J.-B. (2008). “Bilinguisme et trilinguisme à Palmyre.” F. Biville, J.-C. Decourt, and G. Rougemont (2008): 195–211.Find this resource:


(1) The poem evokes the national delegations that came to Rome to pay homage to the emperor Vespasian, for the inauguration of the Flavian amphitheater.

(2) The word multilingualism is ambiguous. In French, multilinguisme is used for persons (as a synonym of polyglot) and plurilinguisme for societies, but the two words are often confused.

(4) While we wait for the digitized critical bibliography of Greek-Latin bilingualism to be put online, which is being prepared by the Centre Informatique de Philosophe et Lettres (CIPL) of the Université de Liège (N. Carling in Marganne-Rochette, 2013: 37–40), we may in the meantime consult Werner (1992: 233–252), Laguna Mariscal (1995), Rochette (1998), Fögen (2003), Adams (2003a: 767–804), Mullen and James (2012: 335–379).

(7) See Adams (2003a: 111–296) (“Languages in contact with Latin”).

(8) Pompeian epigraphy, for example, is not considered; cf. Biville (2003).

(9) See Leiwo (1995), Biville (1993. 2002a, 2002b, 2004, 2003).

(12) Biville (2011: 11–12).

(13) Plin., HN 6.15; also see Strabo 11.2.16.

(14) This is the ideal of tertium ex utroque compositum (Prisc., GL 2.2.29); see Biville (2008).

(17) Plin., HN 3.39: tot populorum discordes ferasque linguas sermonis commercio contraheret ad colloquia et humanitatem homini daret breuiterque una cunctarum gentium in toto orbe patria fieret.

(18) See Adams (2003a: 760–761).

(20) In the Aeneid, the Trojan Aeneas has no difficulty talking with Queen Dido at Carthage, nor with the Greek king Evander when he arrives in Latium, nor later, with the Latin king Latinus. The Carthaginian Hannon’s Punic discourse in Plautus’s Pœnulus is one of the rare exceptions to this rule.

(21) This is frequent in epigraphy, but also in literature, as with the freed slave of Syrian origin C. Pompeius Trimalchio Maecenatianus (Petr. 71.12), or Thaïm Iulianus, son of Saad, Syrian, in a bilingual Latin-Greek inscription in Lyon: Thaemi Iuliani Sati [fi]l(ii) Syri (CIL 13.2448).

(22) See Biville (1990–1995, tome 2: 473–503), Adams (2003a: 441–459).

(23) On the regional diversification of Latin, see Adams (2007).

(24) Priscian, Ars Grammatica, books XVII and XVIII, De constructione (GL 3, 106–377) inspired, sometimes down to the phraseological tracing, by Apollonios Dyscole’s Greek Syntax. Groupe Ars Grammatica, 2010. See also Biville (2008), Martorelli (2014).

(26) See Biville (1989. 2002b).

(27) Thus, Lucretius (5.925–1104) and Vitruvius (2.1).

(28) Cf. Diodorus Siculus, Bibl. 1.8.1–4.

(29) Isid. Etym. 9.1.1: De linguis gentium. Linguarum diuersitas exorta est in aedificatione turris post diluuium.

(30) Prisc., GL 2.20.6–8: interiectiones autem pleraeque communes sunt naturaliter omnium gentium uoces.

(32) Plaut., Trin. 19, Asin. 11, Mil. 21. PF 32.14–17L. One of the traits that makes Latin inferior to Greek is the assertion of the poverty (egestas, paupertas) of the Latin language.

(33) See, from the middle of the second century CE, Aelius Aristides’s Regarding Rome (63.1).

(34) See Rochette (2009: 41–43) (“other languages”).

(36) For the pertinence of these criteria to the mention and ranking of Italic languages, see Biville (2013: 28–31).

(37) It is interesting to note that in the situation of oral Greek-Syriac bilingualism in Jerusalem, Egeria twice associates the Greek adverb siriste with the Latin graece: pars populi et graece et siriste nouit … aliqua etiam pars tantum siriste (Eger. 47,3). We also find the word syriace (Vulg. 4 Reg. 18.26).

(38) See Adams (2003a: 111–296), “Languages in contact with Latin.”

(39) Pliny, De dubio sermone, ap. Prisc., GL 2.26.16–27.8: O aliquot Italiae ciuitates, teste Plinio, non habebant, sed loco eius ponebant u, et maxime Vmbri et Tusci.

(40) Prisc., GL 2.148.1–3 (regarding the oppidum Suthul, a Numidian fortress, Sall., Iug. 37.3): cum lingua Pœnorum, quae Chaldaeae uel Hebraeae similis est et Syrae, non habeat genus neutrum.

(41) See Biville (2013), Ferriss-Hill (2011, 2014). On the supposed Greek origin (“aeolic”) of Latin, cf. Gitner (2015).

(42) Strabo 7.3.10 and 13.

(43) Hier., In Tobit, Praef., PL 29,23–26 (Adams, 2003a: 269).

(44) Like the Italici at Delos, the Epirotici (Cic.) in Epirus, the Jewish communities of Rome, and the Syrian merchants who established themselves in Lyon.

(45) PF 31.25–27L: Bilingues Bruttaces Ennius (Ann. 496) dixit, quod Brutti et Osce et Graece loqui soliti sunt. Hor., Sat. 1.10.30: Bilingues Canusini.

(46) Quint. 1.4.14. 12.10.57; see Biville (1990: 190).

(47) Quint. 11.3.87: ut in tanta per omnes gentes nationesque linguae diuersitate hic mihi omnium hominum communis sermo uideatur. On the role of gestural signs in the military command, see Onasandre, Strategikos, 26 (Rochette 1997b: 153–154).

(48) Plin., HN 7.88, 25.6. Gell. 17.17.2. Plut., Ant. 27.4–5.

(49) Cf. the section “Civitas and the Name of Rome.” For a meaningful example of the communication difficulties within the ancient armies, see, among others, Polybius’s writings (1, 67, 2–70, 2 and 80, 5–9) on the multilingualism of the Carthaginian army and the behaviors that resulted.

(50) See the references and bibliography in Adams (2003a: 533–534 and 637–641) (Gallus). 33–34 and 260 (Haeranes; cf. also Yon, 2008). 272, n. 371 (Tortosa).

(51) See Adams (2003a: 264–269. 566–567).

(52) See Adams (2003a: 111–296), Biville, Decourt, and Rougemont (2008), Ruiz Darasse and Luján (2011), Mullen and James (2012), Clackson (2015). The notions of “bilingual” text and of “interferences” include very diverse special cases.

(57) See Yon (2008).

(61) On the uses and meaning of the term, see Dubuisson (1983), Poccetti (1986), Rochette (2001).

(63) For example, CIL 4.3340.32=FIRA 3.129d. See also Biville (2003: 229–231).

(64) See Bellandi and Ferri (2008, bibliography), Kramer (1983, 2001), Dickey (2013, 2015).

(65) See Biville (1993, 2002a, 2008).