2 Indigenous Languages, Bilingualism, and Literacy in Hispania Citerior, Third Century bce–First Century ce*

The Iberian Peninsula has a complex geography, in which the mountain ranges and central plateaux act as natural boundaries creating areas that are difficult to access and culturally isolated.¹ In this general context, Hispania Citerior corresponds grosso modo to a relatively accessible area in the north-eastern part of the Iberian Peninsula (Fig. 2.1). To the north, the Pyrenees formed a natural border separating it from Gaul and on its eastern side the province lay open to the Mediterranean. Here several rivers flow into the sea, and, from pre-Roman times, they were important for the communication between the coast, where influences arrived from the wider Mediterranean and beyond, and inland areas.² Eastern Citerior is characterized by the coastal lowlands in which the first phases of the Roman conquest took place: the plains of Catalonia, bordered to the east by the Pyrenees and by the Catalan Coastal Range, which encloses the Ebro Valley, as well as the plains of Valencia and Murcia. In contrast, according to Strabo’s description (3.4.12), the interior of the province is vast and diverse, for the most part rugged and served by different rivers: the Ebro, which connects to the Mediterranean, and others that flow into the ‘western sea’—that is, the Atlantic—with its different connections.³ Another mountain chain, the Iberian System, follows the course of the Ebro on its western side, marking another border with peoples in the interior of the Meseta. Finally, the province also included the Balearic and Pityusic Islands (see Plin. HN 3.18), which in the pre-Roman period were under Phoenicio-Punic influence, and where the ancient cities of Pollentia and Palma were founded in 123 bce. This geography led to the existence of clearly differentiated cultural areas, which in some cases remained markedly isolated, a situation that evolved with the development in Roman times of a road network, promoting a certain cultural and linguistic unification, as well as the diffusion of some innovations, such as the practice of writing.

Strabo (3.4.19) highlights the difficulty of describing the inhabitants of the Iberian Peninsula, whom he considers lived in small political and geographical entities under constant restructuring. However, he roughly identifies areas to which he attributes different levels of development: from the savage peoples of the north and central areas, described with a long series of traits associated with the representation of the topos of the barbarian, to the highly refined inhabitants of the fertile south, as well as the peoples of the eastern coast, marked by the successive presence of non-local peoples. Epigraphic sources and archaeological research have allowed us to see at first hand the contrasts of this Iberia described by the ancient authors, and to build up a picture of the ethnic and cultural configuration of Citerior.

When the conquest of the Iberian Peninsula began, Citerior was inhabited by Iberians in the Mediterranean regions, the Pyrenees and the Ebro Valley, and Celtic-speaking peoples, especially the Celtiberians, in the central regions, bordering the Vascones in the extreme north.⁴ Of these, the Iberians were those who developed complex and hierarchical forms of organization first, which may relate to their comparatively early adoption of writing, as we shall see.⁵ They were followed in this process by the Celtiberians, whose culture can be considered to a certain extent as transitional between Iberian culture in the east and the cultures of the central Meseta or the north coast, where such developments occurred much later. Finally, the Vascones, located in modern Navarre and the neighbouring lands, are scarcely described by the ancient sources,⁶ but epigraphic evidence has recently begun to offer more for the understanding of the development of a Palaeohispanic written culture in this area, in which Celtiberian and Iberian influences are both seen.⁷

Hispania, and in particular Citerior, is one of the western provinces for which the indigenous languages are best known (Fig. 2.2).⁸ In addition to the onomastic data (anthroponyms, theonyms, and toponyms) that have traditionally been used to identify linguistic areas, this region has also provided an abundance of inscriptions in local languages. These texts are, however, largely obscure: Iberian, a non-Indo-European language whose relationship to other ancient or modern languages is still a matter of debate, must be considered mostly undeciphered; our knowledge of Celtiberian is somewhat better, as it belongs to the Celtic group, although the longer texts also remain partly impenetrable; and, lastly, the evidence in Vasconic or Aquitanian, despite its manifest kinship with modern Basque, is too sparse for a full description of the language. These epigraphic documents show, in any case, contexts of language contact, as well as the sustained coexistence of speakers of different origins, which also must have undoubtedly involved situations of bilingualism, whether between indigenous languages, or with the first exogenous or colonial languages—that is, Phoenician and Greek.⁹

However, the most prolonged situation of bilingualism would be produced from the end of the third century bce by contact between the indigenous languages and Latin.¹⁰ Several bilingual or mixed texts provide evidence for this,¹¹ although, as we shall see, these are scarce and limited to the Iberian language, and often difficult to interpret. In any case, the most revealing evidence indicating the widespread existence of bilingual speakers is language change in Citerior, or, more precisely, the process of Latinization. We know that languages may die because of the disappearance of their speakers, in situations of brutal massacres, mass migration, or natural disasters, for example. However, often languages become extinct through a slow process in which individuals gradually adopt another form of expression, often perceived as more prestigious than their own, to the point that the chain of transmission from parents to children is broken.

The epigraphic record demonstrates that, in the space of two centuries, Latin ended up dominating in Hispania, at least in the epigraphic record, eventually replacing all other languages with the exception of Basque. This process of substitution could be the result of a repressive policy of imposition, but it can also be due to a more subtle phenomenon in which the language is marginalized through social pressure. It is precisely this second mechanism that has traditionally been considered to characterize Roman imperialism. One example of this attitude is provided by the poet from Bilbilis, Martial, with reference to Citerior, who, despite defending his local roots and considering himself ‘born of Celts and Iberians’ (4.55.9), calls the names of his land of birth crassiora ‘cacophonous’ (12.18.12). The sources also speak of the ridicule that certain Hispanic accents of Latin attracted, some of which were probably influenced by the local languages. The Emperor Hadrian was a victim of this attitude when, while still a quaestor, he read a speech in the senate agrestius pronuntians ‘with a provincial accent’, which incited the mockery of senators (De vita Hadriani 3.1). These prejudices create the context in which the epigraphic texts, the most direct sources available for evaluating the process of language shift in Hispania, were inscribed. But they are usually elite generated and focused, and it remains very difficult, though not impossible, to draw any conclusions about broader attitudes and language ideologies. Who can say what a Celtiberian, a Vascon, or an Iberian felt deep down about their language(s)?

In order to understand the situation of bilingualism, or in some cases diglossia,¹² between Latin and the indigenous languages in Citerior, the epigraphic record is of capital importance, both that written in local languages and the earliest documents in Latin, some of which were written by indigenous people and therefore show contact phenomena. The richness of the epigraphic data, which is not replicated in other parts of the Roman West, allows the study of the process of Latinization from a particular perspective, focusing attention on the indigenous side, in contrast to the literary sources. However, inscriptions have serious limitations when one wants to study speakers’ language choice and usage: indeed, they cannot easily indicate the end or survival of the oral use of the languages. This can be deduced only from indirect data, which are not always interpreted in the same way as, for example, in some literary episodes.¹³ In particular, regarding Citerior, we have information from Tacitus (4.45.2) about the use of Celtiberian in 25 ce in Termes (Soria), where one of the inhabitants under torture for supposedly murdering the legate, Lucius Calpurnius Piso, exclaimed voce magna sermone patrio ‘in a loud voice, in the language of his country’ that he would not give away his companions.¹⁴ Centuries later, at the end of the fourth century or beginning of the fifth century ce, the ascetic virgin Cerasia indoctrinated the Aquitanians in Christianity not only ‘in a language which was agreeable and appropriate to each one’ (sermone blando et suo unicuique) but also in ‘a barbarian language’ (lingua barbara).¹⁵ Despite the undeniable interest of this information, in reality these examples are extremely rare, and it is not even possible to say with any certainty that they refer specifically to the use of local languages and not to a particular or dialectal form of Latin.¹⁶

In order to dig deeper into this question and in relation to the local forms that Latin could have adopted, research has looked for other indirect data, such as features that are attributable to autochthonous languages in Latin inscriptions and in contemporary writing of Hispanic origin. However, given the limitations of the available evidence, which is often of a formulaic and standardized nature, current research is somewhat sceptical about the possibilities of identifying the specific characteristics of the Latin spoken in Hispania (Chapter 3).¹⁷

It has often been argued that the Romans did not develop a linguistic policy stricto sensu that imposed Latin on its people.¹⁸ What also seems irrefutable is that language change did not occur naturally and spontaneously. Rather it was determined by the socio-political context, which was not free from impositions, especially military control, and social pressure in relation to the prestige, or lack thereof, of languages and the need to know Latin in order fully to participate and progress in society.¹⁹

The example of Hispania, in particular, shows that, early on, vernacular languages were not abandoned; rather they enjoyed a period of expansion, which is seen in the use of inscriptions. In contrast, from the age of Augustus onwards, they went into a decline and rapidly disappeared from the written record. It is, therefore, possible to distinguish two different stages in the evolution of the use of these languages: the Republican period, a time marked by permissiveness and continuity with the previous period, and the Principate, a time to which the latest indigenous documents in Citerior can be dated and when a boom in Latin epigraphy occurred, in line with the rest of the Empire.²⁰

In order to understand the particularities and the pace of the process of Latinization in Hispania, first of all the early date of the conquest (218 bce) must be taken into account, and the fact that this was especially long, did not follow an orderly progression, and was to a certain extent improvised. Indeed, 200 years passed between the arrival of the Romans in Emporion and their victory against the Cantabri in 19 bce, a long process of military control that involved the violent repression of the local population at times, and the search for collaboration with indigenous elites at others.²¹ During the first centuries of Roman presence, the Iberian Peninsula must be seen as being under fluctuating territorial control of the Romans before full integration into the Roman world (Fig. 2.3). In this context it is understandable that local languages continued to be used with regularity.

The turmoil and trauma that the military conquest brought to local communities should not be underestimated, as well as the repeated conflicts that immersed them in war for generations.²² Citerior was the battleground for wars against the Punic advance, which marked the starting point of the conquest. The subsequent indigenous revolts were harshly put down by Cato, which brought about the destruction of many settlements and changes in the fabric of local populations.²³ However, this context of repression combines later on with a strategy that was not excessively interventionist, which brings with it pacts with the local elite and the continuation of the previous power structures. This may also explain the survival of the use of indigenous writing and language. Later, the bloody conflicts of the second century bce against the Lusitanians and Celtiberians took place (154–133 bce), and, in addition to the revolts and repressions linked to the conquest, the territory was also the battleground during the Sertorian Wars in the 70s bce and finally also between the Pompeians and Caesarians in the 40s bce.²⁴ This must not be forgotten when assessing the extent to which Latin was, or was not, an imposed language.

All these conflicts must have brought about a large reduction in the local population, caused by deaths in war, displacement of the enslaved population, hunger, and post-war poverty.²⁵ But, although local languages certainly lost many speakers, this did not mean that they fell into disuse or disrepute, since they continued to be dominant in the epigraphic record. When considering elements that favoured their survival, it should be borne in mind first that the main attitude of Romans towards the languages of those they conquered was one of indifference, as shown by the general silence of ancient authors on this issue.²⁶ Furthermore, in this period the presence of immigrants from Italy, one of the main vectors for the diffusion of Latin, was not widespread, being limited to soldiers,²⁷ and the inhabitants of the few colonies in existence. The territory should still be seen as a place of war and exploitation, rather than an integrated part of the Roman world. However, as we shall see in the following sections through the detailed study of epigraphic evidence, this does not imply that important factors of contact and cultural transfer did not exist, such as trade and administration, which started a gradual but inexorable cultural change.²⁸

During the Principate, once the territory was under Roman control, a series of substantial changes took place, which can be summarized as follows: the foundation of some twenty colonies, this time with the settlement of a great number of Romans; the collective concession of citizenship to more than seventy indigenous communities promoted to municipia;²⁹ and, ultimately, the spread, with the new political regime, of the so-called imperial culture, characterized by a greater capacity for cultural influence.³⁰ Of course, one of the key tools to convey all these changes was the adoption of Latin as the main language of communication, a process that was also driven by other cultural factors, such as greater access to education for the population and the rise of Latin literature. In Augustus’ reign, therefore, Latin would become the dominant language of epigraphic expression, to the point where it completely replaced indigenous languages in writing at the end of the Julio-Claudian period.³¹ During this period, the Romans still did not develop a repressive language policy, but the reality is that knowledge and use of Latin were indispensable for full integration into the Roman provincial system. Although perhaps it was not always strictly necessary for civic life,³² a command of Latin was expected for Roman citizens. However, this does not seem to have led to the prohibition and active proscription of other languages,³³ which instead were increasingly neglected through lack of prestige.

Before the first century ce, Iberia was a markedly plurilingual territory. When the Romans landed on the shores of Emporion in 218 bce, at least three indigenous languages were spoken in Citerior (Fig. 2.2): Iberian on the coast, Celtiberian and perhaps other Celtic languages in the centre, and probably also Vasco-Aquitanian, a relative of modern Basque, in Navarre and north-west Aragón. Some of these languages, such as Iberian,³⁴ already enjoyed an ancient epigraphic tradition thanks to early contacts with Phoenicians and Greeks, and the Mediterranean world in general, while others had a written culture still in an embryonic stage of development, as is the case for Celtiberian, for whose notation the indigenous peoples had recently adopted the script of their Iberian neighbours.³⁵ The other regions probably lacked writing, and it would not be until the Roman conquest that the first forms of epigraphic expression would be produced, occasionally in vernacular script and languages, as is the case with a few texts that can be interpreted as Vasconic,³⁶ but more usually in Latin.

The adoption of writing by the indigenous peoples of Citerior is a Mediterranean phenomenon,³⁷ which should be directly linked to the ancient commercial contacts with the Phoenicians, which were especially sustained in the south of the Iberian Peninsula,³⁸ and with the Greek world,³⁹ transmitted mainly through the colony of Emporion, founded by the Phokaians in the sixth century bce. It is hardly surprising, then, that at first writing was concentrated in coastal areas,⁴⁰ where the beginning of local epigraphy dates to at least the fifth century bce. The direction of the diffusion of writing always occurred from the coast towards the inland areas, where the penetration of written culture was a later phenomenon, essentially in the Roman period. In contrast, the direction of the initial transmission along the coast, whether it went from south to north, as is the traditional thesis, or from north to south, as has recently been proposed, remains an open question.⁴¹

The early inscriptions reveal, as shall be seen, the marked influence of Greek writing practices, with diverse uses but hardly any monumental epigraphy.⁴² Phoenician and Punic influence is, however, more difficult to gauge: it is not evident that it served as a model for the surviving types of epigraphy—this is probably due to the Phoenicians’ preference for perishable epigraphic supports—but it had enormous importance for the creation of the Palaeohispanic writing systems, whose predecessor was ultimately, without doubt, the Phoenician alphabet.⁴³ Indeed, among the main features of the linguistic and epigraphic situation in Hispania and, in particular, Citerior, are that not only do some of the indigenous languages enjoy a well-established written tradition prior to the conquest, but also that they were written in the so-called Palaeohispanic script, an epichoric semi-syllabic system that was adapted with ad hoc innovations for each of these languages. The use of this writing system lasted well into the Roman period, and, in fact, cultures such as the Iberian were never disassociated from it, as it gives these inscriptions a marked indigenous appearance, including in contexts strongly influenced by the Roman epigraphic tradition. The Celtiberians, for their part, mainly wrote their texts in Palaeohispanic script, despite having types of inscriptions that were based on Roman models, and regardless of the fact that the Latin alphabet was better suited to an Indo-European language, with clusters, such as muta cum liquida, whose notation was impossible in syllabic script. This fact has caused some puzzlement, even being interpreted as a kind of defence of identity or resistance to the Roman model.⁴⁴ Based on the current evidence, it is difficult to affirm that such attitudes existed,⁴⁵ but, in any case, as is argued throughout this chapter, the manifestations of indigeneity can be explained as straightforward elements of continuity with the previous period, in which the Iberian epigraphic culture was sufficiently developed to become a model for neighbouring cultures.

In short, Citerior had a rich variety of vernacular languages, whose written culture preceded the arrival of the Romans in the majority of cases, and in the early centuries of Roman influence preserved many of the characteristics that can be detected in the previous period. This observation is important, as it reminds us to be cautious when linking the changes that took place in the Roman period to the influence of the Roman epigraphic habit: we should not underestimate the dynamics that began before the conquest.⁴⁶

The foreign peoples who settled in Iberia more or less permanently, especially the Greeks and Phoenicians, used their own languages, which are preserved in the written record (Fig. 2.4). The study of these documents is essential to understand the birth and early development of the epichoric written cultures of the Iberian Peninsula and then to evaluate their evolution under Roman influence. In fact, local epigraphy has been considered to be an ‘epigraphy of models’, indicating that influences from peoples who came from elsewhere incentivized the adoption of the different writing practices and types of epigraphy. The interweaving of echoes and influences of the so-called epigraphic koine of the Mediterranean is indeed indisputable,⁴⁷ although this does not mean that the Palaeohispanic peoples did not also develop their own epigraphic practices, sometimes with notable originality.

As we have seen, the main sources of influence before the Italian peninsula became more dominant were Phoenician and Greek, the former being especially important for the south of the peninsula and the latter for the north-east. Although neither of the two corpora brings together a great number of documents, only a few hundred in each case, it is evident that both the Phoenicians and the Greeks in Iberia arrived with an established written culture, which acted as stimulus for autochthonous developments.

The calligraphic style of attested Phoenician texts, perhaps reflecting people accustomed to writing regularly on soft materials, has been highlighted, and the recently discovered cretulae in Cádiz are proof of the existence of archives of documents, perhaps on papyrus.⁴⁸ In fact, the preference for perishable media is one of the essential characteristics of Phoenician written culture. Therefore, one of the influences that can be attributed to the Phoenicians is, paradoxically, the lack of known epigraphic texts in areas where we know that the written culture had already developed, as is the case in the south of the peninsula.⁴⁹ In contrast, the impact of the Phoenicians is less marked in Hispania Citerior. Here, Phoenician inscriptions are concentrated, as in the south, at coastal sites, but the finds are very few in number.⁵⁰ The reduced Phoenician presence on the central and northern Mediterranean coast has been related, as noted above, to the fact that the majority of the Palaeohispanic corpus comes from precisely these areas.

The same assessment of the unequal distribution of the colonial texts is applicable to those written in Greek, which, unlike the Phoenician texts, are concentrated in the north-east, the most significant area being the Greek colony of Emporion and its area of influence, followed to a much lesser extent by the gulf of Valencia and the Bay of Alicante. However, in general terms, Greek epigraphy in the Iberian Peninsula is not abundant. This is due in part to its location in colonial enclaves, and also to the particularities of the early annexation of Hispania to the Roman Empire, which resulted in the linguistic domination of Latin over the other languages of the peninsula, including Greek, which ceased to be used in its main domain, Emporion.⁵¹ Despite this, Greek influence in the development of local epigraphic culture is undeniable and crucial for understanding indigenous written culture before the arrival of the Romans.

Greek epigraphy in Citerior, and especially that of the coast of Catalonia, has close links with the south of France,⁵² and the epigraphy of both has some particularities characteristic of the Ionian colonies in the West.⁵³ Before the third century bce, the corpus is comprised almost exclusively of private inscriptions, the majority found in domestic contexts: inscriptions on pottery with marks of ownership or commerce and messages of ludic or sympotic type, and small lead sheets that contain commercial texts or defixiones. Public or official inscriptions and funerary texts are almost completely unknown, and do not begin to appear until the Roman period. In any case, what seems especially important is that these early uses of Greek epigraphy coincide with the first known epigraphic practices among the Iberian peoples.

These observations serve to contextualize the impact of the rise of Rome on the local epigraphy of Citerior, especially Iberian epigraphy, which, because of its longevity, allows a vision of its evolution over time, distinguishing between the uses that can be attributed to Greek influence in the pre-Roman period and those that were introduced for the first time through contact with Rome.

Roman epigraphy in Hispania during the early centuries shows some similar traits to that of the other foreign peoples established in this region (Fig. 2.4).⁵⁴ Indeed, this is a development that can be described as colonial, concentrated in certain settlements, with reduced uses, and with limited impact for the bulk of local population. These characteristics can be explained in large part by the early date of the Roman conquest: besides the different strategies that Rome adopted for the annexation of the territories conquered in the Republican and imperial periods, it should also be taken into account that the first Romans settling in Hispania still had limited literate baggage and only an incipient epigraphic culture, factors that are key in explaining why local written cultures continued to enjoy great vitality in the first two centuries of Roman control.

The earliest documents in Latin predate the conquest and date to the middle of the third century bce. These are mainly graffiti and stamps made ante cocturam on Graeco-Italic amphorae found at ports in the north-east, or graffiti made post cocturam on Campanian ware from Carthago Nova or Emporion, and at least the second group could be evidence of the occasional presence of Latin traders before the conquest. The influence of peoples who arrived from Italy on the development of the earliest local written traditions is difficult to gauge, although it seems increasingly evident that contact with the Etruscans in the north of the Iberian region—that is to say, in Languedoc⁵⁵—should not be underestimated as a vector of cultural transfer.⁵⁶ In the same way, the early commercial relations with the Italian peninsula, whether with Latins or even Oscans,⁵⁷ also determined the early adoption of certain epigraphic practices.

However, the great change in the epigraphic culture of Hispania occurred with the permanent establishment of the Romans and the deployment of the provincial administration. One of the inscriptions from the beginning of this period is the famous wall graffito on the Torre de Minerva (ELRH C58) in Tarragona, which is also illustrative of the social and linguistic context that must have been experienced in the city from its foundation. The text, in archaic Latin, reads M(anios) Vibio(s) Men(e)rua and must be interpreted as a dedication to Minerva, protector of the fortification to judge by the bas-relief of the tower. Thus it is in a defensive building and a military context that we find one of the first documents written in Latin in Hispania, and, furthermore, it is significant that it shared the same space with other wall graffiti in Iberian, which, despite the obscurity of their content, confirm close contact between the two communities in the capital of the province. This example illustrates how the main port cities in Citerior—namely, Tarraco and Carthago Nova—where the majority of the Roman inscriptions from this period are concentrated, but also Emporion and Saguntum, acted as focal points for the diffusion of Latin among the indigenous population.⁵⁸

From the beginning Latin was the language of the provincial administration. Good examples of this are the earliest bronze tablets with legal content found in Hispania, issued by the Roman authorities and written in Latin,⁵⁹ even those directly concerning the local communities, who were probably still unable to unravel the subtleties of the sophisticated language of Roman administration.⁶⁰ One of the Republican bronze tablets found in Citerior, the Tabula Contrebiensis (AE 1979, 377), reveals how Roman administration was always expressed in one language.⁶¹ The document contains the resolution of a dispute in which three different local communities were involved, Alaun, Salduie, and Contrebia Belaisca, but the judgment being ratified by the governor of Citerior appears written exclusively in Latin. However, this did not stop the same indigenous people from adopting the epigraphic practices of the Roman administration into their own languages to deal with their own communities. Indeed, the large bronze plaques written in the Celtiberian language from Botorrita and Novallas cannot be understood in any way other than as a translation into the indigenous world of the legal and administrative epigraphic practices of Rome.

Another context of official epigraphy where local languages permeate is that of coin legends. The first issues of Roman coins in Hispania appear in the context of the Second Punic War, and in the second century bce some cities, such as Valentia in Citerior, also issued coins with legends in Latin. Nevertheless, in contrast to what occurred in Turdetania, where Latin was adopted by many cities from the beginning of the second century bce, in Citerior the majority of the communities issued coins in the local languages, Celtiberian, Vasconic, or Iberian, with some occasional Iberian–Latin bilingual issues, until the middle of the first century bce.⁶²

Apart from being the means of communication of the Roman administration, Latin was also, of course, the language used by the people who arrived from Italy. First, the impact of soldiers should be considered, whose graffiti in their military camps and inscribed weaponry (helmets or slingshots, glandes inscriptae) have been preserved. Despite the limited evidence, the role of the army in the diffusion of written culture and eventually Latinization takes on a new dimension through the study of the writing equipment, whose concentration in military settlements in the north-east is significant (see below 2.5.2).

Secondly, immigrants are especially important—that is, colonists—mainly discharged veterans in Roman foundations and freedmen who arrived in the Iberian Peninsula attracted by the economic opportunities that the new territories offered them: mining in the region of Carthago Nova and the production of wine and oil in the north-east. However, not all these groups left the same written record. An important observation is made by Francisco Beltrán Lloris and Borja Díaz Ariño on the importance of freedmen settled in the main commercial enclaves in Citerior for the development of the Roman epigraphic habit, contrary to what occurred with the inhabitants of the new colonies, who have hardly left any written evidence.⁶³ The oldest colony in Citerior, Valentia, founded by veterans in 138 bce, and one of the first cities in the province to issue coins in Latin, has provided hardly any monumental inscriptions,⁶⁴ while Tarraco, Carthago Nova, and Emporiae have many funerary inscriptions commissioned by freedmen, as well as evidence for their economic and commercial activities (seals on lead ingots, marks on amphorae, graffiti, and so on). Economic activity, crafts, and trade appear in this way as some of the driving forces for the development of the first Latin written culture in the peninsula, and it is hardly surprising that the indigenous peoples also copied and developed the same practices to express themselves in their own languages. A striking example can be found in the Iberian tituli picti on Graeco-Italic amphorae in Vielle-Toulouse, which are dated to the middle of the second century bce.⁶⁵ It is possible to recognize a series of Latin names adapted to Iberian, as well as formulae that seem to be the local adaptation of typically Italic uses in relation to the commercialization of wine.⁶⁶ It is not possible to determine precisely at which port in the Iberian region they were painted—Narbonne or Emporion would surely be the best candidates—before the distribution of goods towards the interior of Gaul. But what they certainly show is the close collaboration between Italic and Iberian traders, which led to the adoption of new writing practices from the beginning of Rome’s presence.

As has been seen, Roman epigraphic practices are limited to specific contexts and to a few focal points in the urban areas mentioned. In fact, in the interior of the province, where there is little presence of writing in pre-Roman times, Latin inscriptions remain extremely scarce until the first century bce. This means that the local written cultures continued to develop without a break with the previous period. It is, in fact, striking that the Palaeohispanic languages still provide a larger corpus for this period than the evidence in Latin. Nevertheless, this does not mean that they were not influenced by Roman culture from early on, with the introduction of new types of epigraphy and the diffusion of writing in areas that were until that time anepigraphic, as we shall see. At the same time, Latin-language epigraphy was not impervious to local influences, as certain funerary uses show, such as the inscribed cinerary urns in Ulterior,⁶⁷ nor to local influences on some specific types of epigraphy, such as the case of the tesserae hospitales, which are characteristic of the Celtiberian area and which reflect the fusion of two traditions, Roman and indigenous, or the inscriptions made on ex-votos or rock shelters, which are typical elements of Iberian religiosity, to mention just a few examples.

In an influential article on the Romanization of Celtiberian epigraphy, Jürgen Untermann concluded that, ‘in the large majority of epigraphic manifestations, Romanization is not a process of small steps, rather the sudden substitution of one culture for another’.⁶⁸ Despite the fact that the process of Latinization must have been long and complex, the available evidence—an undoubtedly small part of the original epigraphic habit, and a tiny amount of the whole written culture—is insufficient to define precisely each of the stages that led to language shift.

One of the most difficult questions to answer is how, after a period of expansion of local epigraphy in the Republican period, the Palaeohispanic languages stopped being written in the first century ce in favour of Latin. To understand the causes of this phenomenon better, it is necessary to examine the (sometimes subtle) transformations that the indigenous epigraphic culture underwent on contact with Roman epigraphic culture. That is, in parallel with the development and intensification of written culture, the adoption of new types of inscriptions, the adaptation of the Latin alphabet to write local languages, and, finally, the appearance of bilingual inscriptions.

During the Republican period indigenous epigraphy continued to develop extensively. It is not difficult to identify a series of social and cultural phenomena to explain the continuity with the previous period, but more difficult to identify is what motivated the intensification of local literacy from the Second Punic War onwards. This is visible in the wider geographical distribution of vernacular inscriptions, the increase in the number of preserved graffiti, and the introduction of writing equipment, not only in Roman military installations but also in indigenous oppida.

At the beginning of the Roman period, Iberian epigraphy was already fully developed, with a great diversity of epigraphic uses, contexts, and functions for writing, and with an epichoric script that was to a certain extent standardized and supra-local. However, in this period (fifth to third centuries bce), indigenous writing was only found in coastal areas⁶⁹ and mainly in urban centres.⁷⁰ In contrast, in the hinterland of the coast of Catalonia, a short distance from the sea but separated from it by the Catalan Coastal Range, it was not until the second half of the third century bce that the epigraphic habit gradually began to be introduced. There are some important finds of lead tablets written in archaic script (before 200 bce) in the Segre Valley, but most of the evidence dates to the middle of the second century and the end of the first century bce. In the Middle Ebro Valley, the situation is similar: there are practically no known texts before the second century bce, and the majority of the corpus dates to the Roman period. One of the most obvious changes that coincide with the establishment of Roman rule is, therefore, the spread of local writing in hitherto illiterate areas, where literacy seems to advance at roughly the same pace as the conquest. That the Roman advance brought with it the spread of literacy is, to some extent, a banal and perhaps expected phenomenon, but what is surprising is that it involved mainly the proliferation of texts in indigenous languages. There are some details that allow us to understand better the scope and basis of this phenomenon.

First, despite the inscriptions of the interior being dated to the Roman period, palaeographic analysis of the Celtiberian corpus, in which the so-called dual variant of the Iberian script was occasionally used⁷¹ and which fell into disuse in the Roman period, allows us to place the beginnings of Celtiberian literacy to at least the end of the third century bce, if not before.⁷² On the one hand, this would explain why the Celtiberians adopted the Iberian syllabary early on and not the Latin alphabet as their system of writing, and, on the other hand, it suggests that the adoption of writing in Celtiberia was not triggered by Roman presence, although this process could have been intensified because of this. Annexation by Rome accentuated the speed and intensity of a phenomenon initiated by the internal dynamics of local societies, for example, in relation to processes of social stratification and urbanization, and which had begun before the Roman conquest.

Also interesting for this argument is Javier de Hoz’s proposal for the periodization of the Iberian epigraphic corpus,⁷³ which qualifies the commonly used bipartite model of pre-Roman and Roman periods. Indeed, since the publications of Untermann, a useful, yet over-simplified, distinction has been made between an archaic period (before the end of the third century bce), with limited use of writing and characterized by local idiosyncrasies, and a classical period that coincides with the Roman period. However, de Hoz proposes the identification of an intermediate period with its own features in the third century bce, before the arrival of the Romans, during which Iberian epigraphy was in its maturity, with the generalized use of writing and a standardized model of the script. This tripartite model of archaic (fifth–fourth centuries bce), classical (third century bce), and Roman (second–first centuries bce) periods allows us to understand better the characteristics of the indigenous corpus in the Republican period as a phenomenon stemming from the previous period, instead of as an epigraphic revolution specifically motivated by Roman influence.

Secondly, the causes of the large increase in Palaeohispanic inscriptions from the second century bce onwards should be considered. It has already been mentioned that practically all the Celtiberian and Vasconic inscriptions date to the Roman period; as for the Iberian corpus, approximately two-thirds can be attributed to this period.⁷⁴ Naturally, the fact that writing was geographically more widespread had an impact. Important social and material changes undoubtedly also contributed to this, such as the improvements in the road network and the intensification of trade under Rome, although, as has been said, it is possible that the first seeds of this expansion predate Roman influence.

Furthermore, in the case of Iberian, it should be borne in mind that some settlements cause a certain distortion of the data: the settlement of Azaila in the Ebro Valley, one of the few near-fully excavated sites, has provided some 450 inscriptions, a very high percentage in a corpus that totals some 2,250 texts. It should also be taken into account that, in areas where writing was a more recent development, epigraphic practices are, in general terms, less developed and the finds are much more dispersed than in coastal areas. And, even in centres of importance, such as Azaila, the corpus is generally made up of graffiti on pottery or inscriptions on instrumentum, while lead tablets are absent and inscriptions on stone are rare.

In the same way, the type of texts that these inscriptions contain is also illustrative, being usually very short and in many cases consisting of simple marks or abbreviations. This is evident if we compare the texts in the two main corpora in the Iberian zone, Ensérune in Gaul, which is characteristic of the earliest epigraphy, and Azaila, which is especially illustrative of its area and timeline. In the first case, the corpus includes a more varied typology of inscriptions: the texts are generally longer, and, more significantly, even the graffiti on pottery are longer, with complete names, explicit grammatical marks, and formulaic expressions. The corpus of Azaila, in contrast, consists of very short graffiti, the large majority of which are abbreviations or marks of barely a few signs. Other sites with a sufficient number of inscriptions to allow diachronic analysis point to a similar evolution. This is the case of the oppidum of Burriac, which has been the subject of a case study by Alejandro Sinner and Joan Ferrer i Jané, which concludes that, as a result of colonization, the use of writing increased quantitatively in the settlement, but that the complexity of the inscriptions decreased, being largely reduced to short property marks on the bases of vessels.⁷⁵ So, despite the quantitative increase in literacy in the Roman period, the local written culture also starts to show in some areas the simplification of linguistic practices and uses, perhaps in relation to the beginnings of Latinization, which may also explain the somewhat abrupt replacement of Iberian by Latin in the epigraphic record at the turn of the era.

To understand the developments in the culture of writing, it is crucial to analyse writing equipment, a line of research that is still in its early stages for Hispania, but on which a number of articles have been written in recent years.⁷⁶ These contributions highlight the importance of finds in Citerior in the Republican period, especially on the coast (Valentia, Tarraco, Baetulo, Emporion), but also in the Ebro Valley, with notable sites such as La Cabañeta, a settlement from the end of the second century bce with a significant Italic presence, where mainly Latin, but also Iberian, graffiti have been found,⁷⁷ and Azaila and Botorrita, which have provided important epigraphic evidence mainly in the Iberian and Celtiberian languages.⁷⁸ Overall, the main question raised by these finds that coexisted with vernacular epigraphy is whether they should be understood as supplementary evidence for the increase in indigenous literacy, or, on the contrary, as an early focal point of contact with Latin.

Oriol Olesti has addressed this question in two articles concerning the north-east of Citerior, which provide a complementary perspective to that of epigraphy.⁷⁹ He studies more than a hundred writing and associated instruments from the Republican period (second–first centuries bce) from twenty-seven sites in modern Catalonia and highlights the importance of the find-spots, which are focused around three types of site. The majority of the examples come from Roman foundations or re-foundations—that is, centres that were open to trade, with the presence of Italic peoples, and in contact with the provincial administration, a context in which Latin inscriptions are also concentrated and which has already been considered one of the main focal points for Latinization. The second most significant group comes from military establishments, which underscores the importance of literacy among Roman soldiers and reopens the question of the importance of the army in the diffusion of writing and Latin from the Republican period onwards. Finally, the smallest, but most surprising, group comes from indigenous oppida. It is interesting that in these settlements hardly any documentation in Latin has been found, despite the likely presence of Italians in many cases. In contrast, they have provided graffiti in the Iberian language, raising the question whether the tablets were written on in Iberian or Latin. The first possibility is plausible, and it takes on more significance in light of the discovery of at least one engraved stylus in Palaeohispanic script in the Celtiberian area.⁸⁰ This hypothesis might encourage us to assume perhaps even higher levels of literacy among the indigenous populations in this period of intensification of written culture. The second possibility would make up for the surprising scarcity of Latin texts in a region that appears already to be under strong Roman influence from the second century bce onwards and would offer a better explanation for the abrupt disappearance of the Iberian language in the first century ce, by being able to assume an earlier Latinization of these populations than was previously supposed.⁸¹ These options are, obviously, not mutually exclusive, and it is likely that both Latin and the vernacular languages were written with, and on, these implements. This phenomenon, in any case, must be understood as complementary evidence for the adaptation of the Roman habit of writing by indigenous peoples, which will be dealt with in the following section.

One of the most tangible transformations in local epigraphy in contact with Latin is the appearance of new epigraphic types,⁸² which would bring with it as the main innovation the development of official and public epigraphy. As a result of the Roman presence, we find, for example, inscribed coins for practically the first time,⁸³ through which local communities publicly expressed themselves, particularly in the vernacular languages (Iberian, Celtiberian, and Vasconic), and also occasionally with bilingual emissions in Latin. The issuing of coins in local languages is a phenomenon that cuts across the different areas of Citerior. In contrast, other forms of epigraphic expression, such as inscriptions on bronze or stone, show territorial particularities (Fig. 2.5).

In Celtiberia, we find some of the most illustrative examples of the translation of Roman legal texts to the indigenous world. Epigraphy on bronze, a usual epigraphic support for Roman administrative texts, is common in this area. The Celtiberians used it for their two most characteristic types: tesserae hospitales, objects aimed at confirming pacts between individuals and cities and which probably certified the granting of local citizenship,⁸⁴ and large tablets with legal rulings.⁸⁵ Four plaques of this type are known in Celtiberian, three found in Botorrita⁸⁶—in addition to the Tabula Contrebiensis in Latin, mentioned previously—and a fourth more recently discovered in Novallas.⁸⁷ The latter has shone new light on this usage, which appears not to be locally restricted, but with a certain diffusion in the region. This piece is also evidence of a more advanced stage of Latinization: not only is the text written in the Latin alphabet with slight adaptations for the local language, but it also contains borrowings from the Roman administrative lexicon, such as the term publicus, which makes the link between the tabulae aenae in Latin and the indigenous ones even clearer.

The official use of bronze as a medium for writing differentiates Celtiberian culture from Iberian culture, where there is no evidence of official epigraphy, apart from the coin legends. In this area, however, the Roman conquest incentivized the development of epigraphy of monumental, funerary, and honorific types, which until then had been very residual.⁸⁸ On many occasions it has been stated that, before the Roman conquest, Iberian writing was restricted, like Greek, to the private sphere.⁸⁹ Although it is true that publicly displayed inscriptions are scarce, it is also the case that writing is present from very early on in civic and communal contexts, such as commensality or in the religious sphere, so it does not seem entirely accurate to say that the Iberians did not use public epigraphy before Roman rule, if the concept is taken in a broad sense, although it was not until this time that the monumentalization of inscriptions occurred, as well as their public display in the main spaces within the cities.

This process can most clearly be seen in funerary epigraphy, with manifestations in both the Iberian and Celtiberian areas. Following the Roman conquest, the stelae in the indigenous tradition, which until then displayed only an iconographic message, began to have inscriptions in the vernacular language, and it is possible to observe how writing progressively gained prominence in the epigraphic field until it became the sole protagonist.⁹⁰ The Roman influence on the stelae is an interesting phenomenon, because it occurs not only in urban centres, in which cultural transfer is to be expected, but also in rural areas more remote from the habitual routes of communication,⁹¹ which shows the penetrating capacity of Roman provincial culture from an early period, beyond what can be assumed from the small epigraphic record in Latin. This phenomenon can be seen in Fig. 2.6, which shows the distribution of lapidary epigraphy in places relatively far from Roman roads and coin-issuing settlements.

In contrast, a specific phenomenon in cities such as Emporiae, Tarraco, and Saguntum is the presence of monumental inscriptions with clear Roman influence in the execution and shape of the characters, interpunctuation, and ordinatio, and which sporadically exhibit bilingual texts with Latin formulae adapted to the Iberian language. In some places it is even probable that Iberian inscriptions were produced in the same workshops as their contemporary Latin or Greek inscriptions,⁹² and that they shared the same locations, in the forum or the necropolis, which indicates the permissiveness of the Roman authorities towards the use, including public, of indigenous languages. The same local elites who drove the development of these practices also adopted another custom inherited from Italic habits: the floors in opus signinum in aristocratic houses, some of which also bear Palaeohispanic inscriptions, as can be seen in examples both in the Ebro Valley, in Iberian and perhaps Vasconic,⁹³ and in Ilici (La Alcudia, Valencia), this one already in Latin script.⁹⁴

In parallel with public and official writing, other types of epigraphy also continued to proliferate. Throughout the second and first centuries bce, for example, pottery stamps on amphorae, dolia, and mortaria spread, as well as inscriptions on instrumentum domesticum, such as spindle whorls and weights. Although these uses are not an innovation belonging to the Roman period, they do show the growing need for managing production and distribution, in accordance with what is indicated by the profusion of writing instruments and associated objects.⁹⁵

The flip side of this tendency to emulate the Roman epigraphic habit is, however, the disappearance or decline of other epigraphic genres characteristic of pre-Roman Iberian epigraphy. This is what occurs, for example, with the inscriptions on objects associated with the elites, such as the silverware of Tivissa and Abengibre, luxury vessels laboriously adorned with inscriptions of a cultic nature, or the tituli picti on large decorated containers from Sant Miquel de Llíria, the loss of which seems to be indicative of the local aristocracies’ decline in power and prestige or the beginning of their transformation. Important changes are also seen in other manifestations of religious epigraphy—for example, in the disappearance of the ancient cultic vases with the baikar formula—at the same time as typically Roman epigraphic supports are introduced, such as arulae, which are sometimes also inscribed in Iberian. The increase in inscribed stelae accompanied the disappearance of traditional sepulchral vessels with tituli loquentes, and it is even possible to see the transition between the two uses. On the stelae, the formula belonging to the ‘speaking objects’ is slowly abandoned until it is replaced by clearly Roman formulae, such as that of filiation,⁹⁶ the age of the deceased, and, on occasions, funerary expressions of the type aŕe take  hic situs est.⁹⁷ These are subtle tendencies, but ones that, when taken together, suggest underlying cultural transformation, with changes in the form of elite self-representation, and cultic, religious, and funerary practices. We would venture to say that the culmination of this process of substitution of certain epigraphic objects and the changes in the writing habit was the abandonment of the indigenous language in the written record. From this moment on, Iberian culture would have a very weak presence in epigraphic texts, with hardly any representation of indigenous theonyms and few appearances of local anthroponyms.

Another significant change is the decline of the Iberian epigraphic type par excellence, lead tablets. In total, 125 lead tablets inscribed in Iberian are known, making them one of the largest collections in the western Mediterranean.⁹⁸ With Roman presence the use of this medium diminishes drastically, becoming exceptional from 150 bce onwards,⁹⁹ despite this being a time of supposed expansion for the practice of writing. For this reason, some commentators have suggested a relationship between the abandonment of this epigraphic medium and the introduction of wax tablets, which would have replaced the use of metal for writing letters and inventories.¹⁰⁰ It is certainly striking that, in the majority of the places where Iberian lead tablets have been found, styli have also appeared, indicating a timeline that follows that of the lead tablets (the latter in the pre-Roman period, and the writing equipment in the second–first centuries bce). However, it is difficult to assume the complete substitution of lead by wooden tablets, given that this would also require full compatibility between the functions of both epigraphic media: not only mercantile and commercial uses, such as letters and contracts, but also the spheres of magic, funerary rituals, and religion.¹⁰¹ Moreover, although it is probable that indigenous peoples used wax tablets, it is also plausible that these items served for writing, and even learning, Latin. Therefore, the drastic decline in lead tablets could be significant for the advance of Latin in two ways. First, because it implies the regression of the Iberian language on an epigraphic medium that is associated with greater writing ability. Indeed, being able to scratch someone’s name on a vessel, often as a simple abbreviation, or a simple commercial mark, does not imply the same level of literacy as improvising a text of several lines. Secondly, because it is another manifestation of the transformation of local written culture, with the abandonment of older practices and the adoption of Roman epigraphic media.

In parallel to the adoption of new types of epigraphy, another very significant change also occurred in indigenous epigraphies: the evolution of the writing systems and eventually the substitution of local scripts by Latin.

First, a clear transformation can be seen in the palaeographic style of the Palaeohispanic semi-syllabary.¹⁰² The shape of the letters becomes less cursive, straight lines are preferred over curved ones, and, in general, the characters become more similar to Roman capitals. This is the palaeographic style typical of the coin legends, lapidary inscriptions, and those on bronze, but which progressively dominated all epigraphic media and writing techniques (graffiti, tituli picti, and stamps), which underlines the importance of public and official epigraphy in this period and seems to become the model for all other spheres.

At this time, the semi-syllabaries also underwent a process of standardization, with the definitive abandonment of the local allographs and the reduction of variants, as well as the loss of the so-called dual system of differentiation between voiceless and voiced plosives,¹⁰³ which assumes a considerable structural simplification of the writing system. In this restructuring, which in fact may have already begun in the third century bce, it is difficult to see a direct influence from the Latin alphabet or to identify clearly the factors that caused it. As I have argued,¹⁰⁴ it is not impossible that the acceleration of this phenomenon may be related to the greater diffusion of the practice of writing in the Roman period, with more people able to use the Palaeohispanic script, albeit at a rudimentary level. This process must have been accompanied by a certain institutionalization of learning how to write, perhaps with the development of a more organized system. In any case, it should also be considered that this simplification of the orthographic conventions also implies a loss of precision in the notation of indigenous languages, in accordance with the decline in the richness of the texts’ content, at least in the lapidary epigraphic record, as described in the previous sections.

Finally, another aspect of note, which this time indicates interference from the Latin alphabet, is the phenomenon of vocalic redundancy: an anomalous use of the epichoric syllabary consisting of the repetition of the vocalic symbol inherently associated with the syllabic symbol, which implies a certain alphabetic use of the syllabic system characteristic of Palaeohispanic writing. This phenomenon, detectable in a small group of Iberian inscriptions,¹⁰⁵ but especially in the Celtiberian corpus,¹⁰⁶ should be considered as evidence of people who knew both local and Latin writing systems, an indispensable condition for the next stage in the process of Latinization: the use of the Latin alphabet to write the indigenous languages.¹⁰⁷

Despite indigenous inscriptions in Latin writing being clearly in the minority compared to those written in local systems, this is an important group from a sociolinguistic perspective, given that it attests contacts between the indigenous languages and Latin, and the existence of bilingual individuals or even communities, since it is difficult to imagine that learning the writing system would be disassociated from the language.¹⁰⁸ The Celtiberian and Iberian areas again show different behaviours with regard to this practice.

In Celtiberia, this is an important phenomenon with considerable implications. It began sometime in the middle of the first century bce, at a time when Latin writing became the dominant system, although the Palaeohispanic script remained in use in the domestic sphere until the first century ce. The use of the Latin script to write Celtiberian was, however, a short-lived phenomenon, given that the most recent examples can be dated to the beginning of the Julio-Claudian period.¹⁰⁹ The repertoire of inscriptions is not very large¹¹⁰—some 40 texts compared to around 200 in the indigenous semi-syllabary—although there is evidence that this use had taken root. First, is it remarkable that practically all the epigraphic types of Celtiberia are represented, from public and official epigraphy to private, religious, and funerary inscriptions. Secondly, and more strikingly, the Celtiberians did not merely adopt the alphabet, but they also applied slight modifications to adapt it to the phonology of their language. Some characters, such as the letter Q, are used in a particular way, as an abbreviation, perhaps maintaining the features of the local syllabic writing in Latin, a practice that is even documented in the Latin epigraphy of this area for the notation of indigenous family names.¹¹¹ However, the principal innovation was the creation of a new character, an S with a diacritic mark on its base used for the second sibilant of this language, which is identifiable in inscriptions separated by a large distance—in the bronze tablet from Novallas and in the rock inscriptions at Peñalba de Villastar—which allows us to see that this was not an occasional use, but one with a certain diffusion.¹¹² Finally, of interest is the argument of Ignacio Simón and Carlos Jordán that this character continued to be used later in a group of Latin inscriptions in Celtiberia with local onomastics,¹¹³ just as in other areas an alternative form of the letter Z was used,¹¹⁴ or, in a similar way, the aforementioned letter Q. These uses denote not only a link between local epigraphy and Latin, but also the existence of schools in the indigenous context in which it was still considered necessary to mark graphically the subtleties of the Celtiberian phonetic system, even at a time when this language no longer enjoyed its own epigraphic tradition, or at least not one that has survived. Again, these standardized uses and conventions allow us to see, as already noted for Iberian writing, a certain degree of supra-local coordination in the transmission of the knowledge of writing among indigenous peoples.

The use of the Latin alphabet was, however, clearly marginal for writing Iberian. Not only are the examples very rare, but none of them constitutes an absolutely certain piece of evidence for this practice.¹¹⁵ Here, what can be deduced is that in Iberian culture there was an inseparable link between the Palaeohispanic semi-syllabary and local written culture, and that the abandonment of the former occurred at the same time as the extinction of the language in the written record. By contrast, certain graphic particularities are noticeable in some Latin inscriptions containing indigenous onomastics:¹¹⁶ for example, in some cases the digraphs -LD- or -NB- in their conventional notation in local writing, such as Adinildir and Neitinbeles, as opposed to their expected Latin notation as *Adinillir and *Neitimeles; or certain syntactic particularities that were maintained, such as the lack of adaptation to the Latin declension in the expression of the patronym: for example, Bella Bastobles f(ilius) and Bella Gaisco f(ilius), without the genitive marker. These particularities occur, in any case, in a much more sporadic way than in the Celtiberian context and do not denote a uniform way of learning and using the Latin script among the indigenous population, although it is significant that there were individuals familiar with the two writing systems at the beginning of the imperial period, when Iberian epigraphy is already a residual phenomenon.

Some areas of indigenous epigraphy, where local languages and Latin were in contact for a long time, show in a special way the interlinking of the stages that have been outlined in the previous section. Some rock sanctuaries in the indigenous tradition stand out in this respect, such as Peñalba de Villastar (Teruel),¹¹⁷ the rocks in Cerdanya in the Eastern Pyrenees,¹¹⁸ and La Roca dels Moros del Cogul (Lleida),¹¹⁹ where Iberian inscriptions appear with prehistoric schematic paintings and votive texts in Latin. These groups of inscriptions also provide invaluable cultural information: the continued use of certain cult spaces from prehistoric times until the end of the Roman period and even beyond.

Although these are sanctuaries and not settlements, their long period of use allows us to trace the diachronic evolution of the language of the pilgrims. At Peñalba de Villastar, the earliest inscriptions appear in the Celtiberian language,¹²⁰ the oldest of which are in the indigenous semi-syllabary,¹²¹ although the majority are in the Latin alphabet, which is occasionally adapted to the local language. The most recent texts, from the beginning of the imperial period, are written in Latin, and it is worth mentioning the presence of an alphabet, a well-known use of writing in cultic contexts in pre-Roman Hispania, and a quotation from Virgil’s Aen. (2.268–9) followed by an ironic nescio qui, which constitutes an invaluable example of the penetration of Latin literary culture in the interior of Celtiberia through channels that must have involved some form of schooling.¹²²

A similar evolution is seen in Cerdanya. The earliest inscriptions are written in the Iberian language, the majority of them in the older version of the syllabary and others attributable to the Roman period, in which theonyms are abundant and also votive alphabets.¹²³ The presence of the Latin script begins in the age of Augustus,¹²⁴ and there is evidence that points to the indigenous population expressing themselves in this language—not only local onomastics, but also some graphic and syntactic traits attributed to vernacular practices. Latin continued to appear at these Pyrenean sanctuaries including for inscriptions with Christian content,¹²⁵ and its use continued until the Middle Ages.

These environments, which document the coexistence of speakers of different languages, undoubtedly going through stages of bilingualism, however, do not provide us with bilingual or mixed texts. This type of text is evidence for yet another phase in the process of Latinization, and, in the case of Hispania, a stage that can be described as advanced, just before the local languages fell into disuse. The practice of writing bilingual texts in Latin and vernacular languages is very rare in Hispania:¹²⁶ there are very few documents, and they are limited to the Iberian linguistic area, and specifically its northern half—that is, Citerior—the only area that, in addition to coin legends, has provided inscriptions in two languages. Despite this, it should also be taken into account, as María José Estarán Tolosa’s study has shown, that this is a limited phenomenon in the entire western Mediterranean, probably because it required an established epigraphic tradition prior to contact with Latin, a condition that the majority of the regions do not fulfil.¹²⁷

Despite their small number, the bilingual Iberian texts have a typological variety that provides important information. The largest group are the coin legends.¹²⁸ In Citerior, Saguntum began to issue bilingual coinage in the middle of the second century bce, and this practice was also adopted in the following century by the neighbouring cities of Gili and Saetabi. Following the Caesarian wars, Celsa and Oscicerda in the Ebro Valley also issued bilingual coins, in a region that has also provided an interesting series of stamps on mortaria in two languages. The only inscriptions on stone, funerary or monumental, however, come from important urban centres: Tarraco, with three instances, and Saguntum. With the exception of the coin legends from Saguntum, the bilingual inscriptions should be dated generally to the first century bce and correspond to the final stage of epigraphy in the Iberian language.¹²⁹

Leaving to one side coin legends, which give a double version of the toponym, these bilingual inscriptions are of disputed interpretation, but it is believed that each version is adapted to the reality and cultural references associated with each of the languages. They are, therefore, of great interest for evaluating questions about the social status of the two languages. Monetary epigraphy stands out in this regard, given that it represents the community,¹³⁰ which projects its identity to the outside world symbolically through the choice of language. The desire to use two different languages can be seen as a manifestation of the incipient dual identity of the local authorities, which continued, however, to express themselves mainly in the vernacular language in the epigraphic record.¹³¹

Furthermore, the few bilingual inscriptions on stone also reveal the same intentions on the part of the indigenous elites, as well as showing how, at a given moment, cities could display bilingual texts in which Iberian occupied a similar position to that of Latin. Indeed, some of the inscriptions are very carefully made, with elegant letters of a considerable size (c.8 cm, for example, in one of the examples from Tarragona),¹³² and, although in the majority of them Latin comes before Iberian, the two languages appear on an equal footing. This finding is somewhat surprising, especially if we take into consideration that these inscriptions appeared at a time when Iberian epigraphy was about to disappear. It is likely, therefore, that this use should be understood as a sign that the local elites were finally integrating Latin as the language of self-representation, instead of evidence that Iberian had reached a kind of status as ‘co-official’ with Latin. Indeed, the next step towards the abandonment of local languages was the adoption of Latin as the only language of epigraphic communication by the indigenous elites.

A small but significant group of Palaeohispanic inscriptions shows the beginning of the process of adoption of Latin onomastics by the locals in the Republican period. Obviously, the abandonment of traditional anthroponyms for Roman ones is not unequivocal evidence for the adoption of Latin,¹³³ especially in the case of the names of peregrini, but it does indicate the admission of a form of Romanness as part of the individual’s identity.

One of the most illustrative examples is the Celtiberian bronze tablet from Botorrita (Hesperia Z.09.03),¹³⁴ with examples such as the cognomina  salute (Saluta), bolora (Flora), balakos (Flaccus), bubilibor (Publipor), and possible praenomina such as titos (Titus) and markos (Marcus). These forms are inserted in the typical structure of the traditional nomenclature of the Celtiberians, with the indication of the supra-familial unit in the genitive plural, and, when the patronym is mentioned, this is also indigenous. Therefore, the most plausible interpretation is that these are Celtiberians of peregrine status who were given a Latin name, perhaps one that sounded similar¹³⁵ (‘cover name’/‘nom d’assonance’, see Chapter 6), or simply chose Roman names in a context in which Latin was becoming a language of prestige.

Another noteworthy example, this time from the Iberian area, is the mention on a monumental plaque from Emporiae from the first century bce of a certain ma]ŕke koŕneli,¹³⁶ an adaption into Iberian of Marcus Cornelius, from the vocative form. It cannot be ruled out that this was an indigenous person who had gained Roman citizenship and, despite that, continued using his native language as the vehicle of representation in his community, although, given the fragmentary nature of the piece and our poor understanding of the Iberian language, we should recognize that this is not the only interpretation possible.¹³⁷

The most representative evidence for onomastic change in the Iberian sphere is in fact provided by a set of graffiti on pottery with possible Latin praenomina and cognomina,¹³⁸  kai (Caius), luki (Lucius), tite (Titus), likine (Licinus), and bilake (Flaccus), and even some feminine names, such as iunia (Iunia) and uatina (Vatin(i)a). These should probably be understood as a reflection of the cultural, and surely also linguistic, change that underlies the Iberian epigraphic phenomenon in the Roman period and which helps to explain why, despite the apparent vitality of local written culture throughout the first two centuries of Roman presence, Latin became dominant at the turn of the era. Indeed, only a few decades after these graffiti were inscribed, indigenous languages disappear almost completely from the epigraphic record.¹³⁹

Alongside this process, another phenomenon gained importance: the mention of people with indigenous idionyms or cognomina in Latin epigraphy, which gives a complementary perspective from which to understand how language change could have occurred and why.¹⁴⁰ Although this phenomenon still has a very limited reach in the Republican period in Citerior, there are at this time some remarkable exceptions, such as the Tabula Contrebiensis (AE 1979, 377) or the Ascoli Bronze (CIL I² 709), in which some individuals, such as the cavalrymen from Ilerda, have the onomastics of citizens. However, there are still no inscriptions with local onomastics that do not emanate from the Roman authorities and which, therefore, are evidence for the adoption of Latin by the local population. This is, however, a trend that differentiates Citerior from Ulterior, where Latin inscriptions, despite being rarer, contain vernacular elements from a much earlier date.¹⁴¹ This leads us to think that in Ulterior Latin was adopted more quickly for epigraphic communication.

From the Augustan period onwards, indigenous onomastics started to appear more abundantly in the Roman epigraphy of Citerior, especially in funerary inscriptions, although with very notable differences between the Celtic and Iberian areas in terms of the intensity of the phenomenon and its duration. In Celtiberia, the survival of local onomastics in the inscriptions of the imperial period was significant,¹⁴² with the appearance of many anthroponyms, family names, and theonyms. It is interesting to note the persistence of certain elements that link to the previous period, in terms not only of the features of the onomastic formula, but also of certain graphic and morphosyntactic archaisms. A Celtiberian element that was maintained with great vitality in Latin inscriptions is the genitive plural (for example, tirtanos abulokum ‘Tirtano of the Abulocos’ (Hesperia IB.01.01)), which expressed the supra-familial group. However, it is also noticeable that, as the Celtiberians gained Roman citizenship, filiation progressively gained prominence as the more indispensable element for individual identity, and the expression of the family group lost ground to the patronym.¹⁴³ Elements of continuity can also be seen though through specific graphic particularities: as mentioned above, the use of a marked S or the letter Z to reproduce the second Celtiberian sibilant, as well as the genitive plurals of family names written with a Q as an abbreviation for -QVM.¹⁴⁴ At a morphological level, it is highly significant that, in the main, the vernacular genitive ending -um is used to designate the family group, instead of the expected ending -orum.¹⁴⁵

These traits cannot be considered simply as fossilized elements, which therefore has two important implications: first, these Latin texts come directly from the indigenous environment, and, secondly, the people who wrote them still had knowledge of written culture and the vernacular language. This close link detectable between the documents in Latin and the previous period, dominated by the vernacular languages, assumes that, although Celtiberian had already disappeared from the written record, local culture and languages continued to imprint certain features and particularities on the implantation of Roman provincial culture. It is difficult to determine precisely how long the use of indigenous onomastics lasted in Celtiberia, but it is possible that some examples could be as late as the fourth century ce,¹⁴⁶ and, as a result, it must be concluded that this is a deep-rooted phenomenon, both in its intensity and its duration.

Conversely, in the Iberian area this practice had a different dimension. The appearance of local onomastics in Latin sources also began to occur mainly from the Augustan period onwards, while in the second century bce its appearance is very peripheral. However, the available examples are comparatively very scarce. The case of theonyms is illustrative: while, in the part of Hispania with an Indo-European substrate, hundreds of indigenous theonyms are documented in Latin inscriptions,¹⁴⁷ in the Iberian area there are just a handful of known cases.¹⁴⁸ Anthroponyms are slightly more common, but, even so, they remain comparatively very scarce, being limited to around sixty inscriptions, most of them funerary.¹⁴⁹

One of the elements that stands out the most in this Iberian group is the large proportion of people with tria or duo nomina, who, presumably, had the status of Roman citizens,¹⁵⁰ at a time when in Celtiberia the majority of people mentioned were peregrini.¹⁵¹ This is significant, because it could explain the speed at which Iberian onomastics disappeared, just a few generations after the extinction of their written culture. This phenomenon reveals the voluntary integration of the Iberian elites in the Roman provincial world and, with it, the rapid adoption of the new language and culture. As a general pattern, it should also be highlighted that the familial relations mentioned in the epigraphic texts allow us to trace a linear and relentless trend towards the abandonment of traditional names. While the older generations have Iberian names or cognomina, the younger ones have fully Latin onomastics (for example, the son of Marcus Porcius Escerior is Marcus Porcius Nigrinus; the son of Calpurina Vrchatetel is Lucius Aemilius Seranus; the granddaughter of Iunia Tannegadinia is Atilia Potita; and so on). This, against the background of the recent abandonment of indigenous writing, can be taken as evidence that the chain of transmission of the language was also breaking down, as the Iberians gained Roman citizenship. It is likely, in this context, that the Iberian language did not take much longer to disappear from the oral medium as well.

Hispania Citerior brings together a series of particularities that make it a unique territory to evaluate the linguistic impact of Rome on western cultures emanating from the Iron Age. The linguistic heterogeneity of the territory is remarkable, with at least three indigenous languages documented through epigraphy, allowing the study of Latinization through the voices of the indigenous peoples, although we should not hide the difficulties in the interpretation of often fragmentary texts in poorly understood languages. Some regions of Citerior, such as the Ebro Valley, constitute a true cultural trifinium, where three different language areas converged, Iberian, Celtiberian, and Vasconic, with inhabitants who must have communicated with each other, as can be deduced from the existence of documents that implicate the three communities, such as the Tabula Contrebiensis (AE 1979, 377). In the Iberian area, evidence points to the existence of people who were able to express themselves in Gaulish, Iberian, or Greek, as shown by borrowing of writing systems, as well as the diversity of the origins and languages of the people mentioned in the inscriptions. Prior to Latinization, then, bilingualism emerges as a practice that is not exceptional. The nature of the acquisition of Latin by the indigenous communities was driven in part by the relative prestige, or lack of it, associated with each language, and, with the exception of Basque, the only pre-Roman language of Hispania that has survived to the present day, all languages of this territory eventually disappeared under pressure from Latin. Nevertheless, the marked idiosyncrasies of each region, which, on the one hand, obliged Rome to adopt different strategies of control and, on the other, caused different linguistic behaviours and attitudes among the local elites, produced variable paces and patterns in this complex process of language shift.

Another factor in understanding the evolution of the local epigraphic cultures is the early date of the Roman conquest and the epigraphic use of Latin in the first two centuries. Documents before the Augustan Principate are rare, concentrated mainly on the coast and in urban areas, where Tarraco and Carthago Nova stand out. The social groups responsible for these inscriptions are also easily identifiable: freedmen are notable in the context of their economic activities, as is the army, as can be deduced from the epigraphic texts and writing equipment found in military installations. Finally, the Roman administration also expressed itself systematically in Latin, including in its diplomatic relations with indigenous cities. However, these uses of Latin were still limited, and its spread among the indigenous populations must have been moderate. In contrast to what occurred in Ulterior, where the local populations quickly adopted Latin for epigraphic communication, in Citerior the indigenous people continued mainly to use their native languages, even in the public and official sphere. This demonstrates that initially local languages were not prohibited, nor was Latin imposed, nor did the elites choose it, at least in the beginning, as a medium for self-representation.

In contrast to the limited use of Latin, in this period the epigraphic record of Citerior continues to be dominated by local languages, which even went through a moment of expansion. This phenomenon cannot be explained simply by the Roman authorities’ permissiveness or indifference to the use of local languages, nor by the registering and regulatory demands derived from the creation of the province. It is to be expected that the intensification of communication networks, trade, and urban culture with Rome’s arrival would also entail the intensification of written culture, but, in order to understand why this phenomenon occurred mainly in the local languages, we must consider a series of factors that coincided at this time, but that do not exclusively follow from Roman influence. It is important to note that, in the third century bce before the Romans’ arrival, Iberian epigraphic culture was already fully established, with a great diversity of types and a supra-local writing system and was in the process of expanding towards the interior. It is precisely at this time that the Celtiberians adopted the Iberian script with slight modifications to write their language. This allows us to conclude that in local cultures there already existed internal dynamics that point to a stage of consolidation and expansion of vernacular writing. Roman interaction probably intensified this process, but it cannot be considered the only trigger.

In the two centuries of coexistence with Latin, local epigraphy underwent a series of influences and substantial changes that allow us to evaluate the progression of the ‘Romanization’ of written culture and Latinization: the incorporation of new epigraphic types, the evolution of the scripts until the adoption of the Latin alphabet, and the erection of bilingual inscriptions. Without this evolution it would be difficult to explain how, after a time of expansion, local languages were abandoned under Augustus.

Celtiberian epigraphy offers a relatively steady model of evolution, with types markedly influenced by the Roman tradition and the use of the Latin alphabet in the final stages of documentation. The recently published bronze plaque from Novallas also offers notable linguistic and epigraphic data, such as the modification of the Latin alphabet to adapt it better to the local language, a usage that also survived in the Latin inscriptions in the imperial period, and the incorporation of Latin borrowings in the indigenous administrative language. These new data allow us to refine even further our understanding of the succession of steps that led to the adoption of Latin for epigraphic expression.

The case of Iberian, however, shows a more abrupt evolution, which, in my opinion, obliges us to reconsider certain aspects of this process. Among the Palaeohispanic epigraphies, Iberian epigraphy has the particularity of being the one that best allows diachronic analysis, from the pre-Roman period (end of the fifth century bce) to the Julio-Claudian period. One of the most striking aspects is the exponential increase of the corpus in the second–first centuries bce, as well as the incorporation of new epigraphic types under Roman influence, mainly in displayed inscriptions and coin legends. However, it is also vital to contrast the quantitative increase in the documentation with other aspects of a linguistic and epigraphic nature. First, the complexity of the texts decreases in the Roman period—for example, graffiti, although more abundant, generally provide much shorter texts, often simply abbreviations of a few characters. Secondly, it is necessary to contrast the adoption of new types of epigraphy with the disappearance of others. For example, specific practices associated with ancient cults, the representation of local aristocracies, and, even more significantly, trade are abandoned, as witnessed by the loss of lead tablets, which were previously preferred materials. All these aspects are indicative of the strong influence of Roman written culture and perhaps also of an earlier process of Latinization than has traditionally been assumed.

The combination of these epigraphic data with the writing equipment allows us to postulate an intense influence of Roman written culture on local traditions from the Republican period onwards, even in areas in which Latin epigraphy is very scarce, such as Celtiberia. It is likely that this penetration also included some kind of formalization of the learning of writing among the indigenous peoples. The information on this is very scarce, but that provided by Plutarch (Sert. 14) concerning the founding of a school by Sertorius in Osca (Huesca) to teach Greek and Latin to the sons of the indigenous aristocracy fits well with the data that we have analysed: the profusion of writing equipment, the standardization of Palaeohispanic writing systems, the adaptation of the Latin alphabet to local phonetics in certain places in Celtiberia, and the penetration of Roman literary culture in the inland areas from the first century ce, as the citation of Virgil at Peñalba de Villastar shows, or, to cite another example, the dedication of a complex carmen epigraphicum to the Iberian Cornelia Sirasteiun in a funerary inscription found in the Ebro Valley (AE 1994, 1059).¹⁵²

Finally, to determine the pace of the abandonment of the vernacular languages once their written culture had disappeared, the study of onomastics is fundamental, since it allows us to see differences in the two main areas of Citerior, the coast and inland areas. In the Iberian world, the survival of local names in Latin is rare, and is already residual by the beginning of the second century ce. This allows us to postulate a rapid disappearance of the language after indigenous writing had fallen into disuse. In contrast, in Celtiberia and, more generally, in the Indo-European areas of the interior of Citerior, the evidence is much more abundant and reaches to the third or even fourth century ce, which shows a greater survival of indigenous culture and surely also, to a certain extent, of the language in these areas. However, the case of Basque, which has left little trace in the epigraphic record in Citerior, forces us to be prudent: a language’s silence in the surviving sources does not necessarily imply its death.