7 Literacy in Gaul*: The Value of instrumentum

The term ‘literacy’ covers not only the technical knowledge of reading and writing, but also the complex effects of these practices on a given culture. The uptake of writing leads to major changes in the relationships between members of a social group, the management of collective memory, and economic activity. The cascading effects of such changes and their consequences on individual and collective identity mean that the uptake of literacy represents a major event in the history of any culture—a change that is difficult for us to grasp and appreciate in all its magnitude.¹ Ancient Gaul went from an almost entirely illiterate local population at the time of the Caesarian conquests to a society with widespread literate practices (if still not a large percentage of literates) within less than a century.² There were some previous influences from Mediterranean communities, resulting in the earliest preserved writing in France in Greek and Etruscan language from the sixth and fifth centuries bce onwards and later Iberian texts found between the Pyrenees and the French department Hérault.³ But a chain reaction spreading this knowledge was really triggered only in the second and particularly in the first centuries bce, when the first indigenous communities started minting coins with legends, imitating those with Greek legends issued by Massalia (Marseille), and epigraphies for writing the local language Gaulish spread.⁴ It was in this context that the use of writing in Latin started to spread in a much more encompassing and concerted fashion just before the end of the first century bce, first within the higher social strata and the Roman colonies, then among those in contact with them and already communicating in Latin. Once set in motion, nothing seems to have been able to stop the spread of Latin literacy.

The value of small finds for researching this development within different provincial contexts has become more apparent over the past decades. It still largely remains to be explored, but an increasing interest in the materiality of literacy has advanced the topic considerably, as have efforts to devise appropriate ways to categorize small finds—for example, by moving away from a focus on materials and instead prioritizing the function of an object, thus accommodating those objects made of more than one material more easily.⁵ The most direct material proxy for literacy, finds of writing equipment, can be particularly problematic, because it can be difficult to tell an iron stylus from a nail, a bone stylus from a hair pin, or an inkwell from a balsamarium.⁶ With the necessary caution, however, and particularly in combination with the study of graffiti and other non-monumental inscriptions, their evidence can help us grasp literate landscapes, as has been shown for a number of settlements.⁷ In Gaul, the joint analysis of texts and writing implements from the first Roman layers of Lugdunum (Lyon), some of which date back to 43 bce and are therefore contemporary with the colony’s foundation by Munatius Plancus, has shown the potential of this approach.⁸ Depending on the chronological phase, finds include graffiti, styli or other writing utensils, or both. The earliest find, a bone stylus from the earliest layer of the praetorium on the top of Fourvière hill and therefore dated precisely to the year 43 bce when the colony was established, confirms that the Roman colonists brought the technology of writing and the relevant tools with them. Regarding the spread of Latin literacy, which is so central to the present volume, ideally we must assess the extent to which writing was present in the earliest colonial settlements but also contemporary indigenous contexts, and also in different social strata and groups within both, and how this develops over time. This would go beyond the scope of this chapter, but we would like to highlight one group of small finds that will be crucial in answering this question: inscribed small finds, or instrumentum inscriptum.

One of the challenges of studying literacy in the Roman Empire is to appreciate the many different ways in which provincial societies could have encountered writing and learned it themselves. For many, the advantages that literacy skills provided for social mobility and professional careers within the new Roman system must have provided an incentive to acquire them. A certain portion of the population will have had access to school or home education, but, even though its general process is relatively well known, it is rather difficult to weigh the impact of formal education in different social groups. Access must have varied greatly across and even within the different provinces, as well as depending on the circumstances of individuals.⁹ Professional and military environments would also have provided an opportunity for certain provincials to learn how to read and write from their peers or as part of their training.¹⁰ For others, being exposed to writing in their daily lives must have familiarized them with the shapes and meaning of letters. Monumental, publicly displayed inscriptions will have formed part of this exposure in the provinces,¹¹ but, in the earlier periods of the Roman conquest in particular, non-monumental inscriptions may have been a more common sight for some, as many objects circulating in all social layers had writing on them. This includes makers’ marks on imported goods such as stamped ceramic and metal vessels, but also coin legends, which are often among the earliest textual evidence in provincial areas without pre-Roman literacy. Given the wide distribution of coins, anyone could find themselves confronted with the legends spelling out names of peoples and kings, later the names and titles of the Emperor. The pictures on the reverse, frequently accompanied by a legend themselves, acted as visual reading aids. Taking non-monumental epigraphy into consideration thus sheds a light on just how present writing could be in everyday life in the Roman provinces, even for those who did not frequently pass by spaces with a high density of stone inscriptions, and those living in a more rural environment.

Yet, even as literacy became more widespread within provincial communities, being able to decipher coin legends or makers’ marks did not necessarily mean being able to write, or even being able to read other texts as well, for that matter. It seems sensible to take not only different levels but a whole range of different kinds of literacy (and numeracy) into account, which were more or less closely related and overlapping.¹² Such different literacies include functional or ‘specialist’ literacy, which can be purpose bound and limited to the requirements for a specific task. Someone may have been able to read and write labels as part of a production or commercial process, but not much else, while someone who was able to compose poetry may not have been able to understand the abbreviations on such labels, or those painted on amphorae.¹³ Since many of the texts on inscribed small finds relate to production and commerce, it has been argued that they were highly specialized and only ever needed to be legible to a small portion of the population.¹⁴ But when studying literacy and language use, to include such material allows us to look beyond the ostentatious and often ambitious sphere of monumental inscriptions, which tend to be related to the higher echelons of society: the nature of non-monumental inscriptions ranges from fleeting to ornamental and from pragmatic to witty. They can relate to the higher social strata too, as is the case with objects made of precious materials with incised texts such as silver styli or highly ornate inkwells,¹⁵ or with scribbles found on the walls of luxurious domestic spaces, where they allow us to see a different layer of the literacy present in such environments, and potentially different linguistic registers—namely, those meant not for public display but for the domestic sphere. Much more often, however, non-monumental inscriptions give an insight into literacy present in other parts of the society, including more modest segments and those groups that do not often feature prominently in lapidary epigraphy, such as those involved in manufacture and production and peregrini, free inhabitants of the Empire who were not Roman citizens. Non-monumental inscriptions therefore allow us to achieve a much more rounded picture that includes different literacy practices within individual social communities, as well as those of a larger number of social communities—even if these communities were relatively small. In this chapter we will attempt to describe what inscribed small finds can contribute to the study of Latinization in Roman Gaul.

Before we can address the evidence from Gaul, the term instrumentum and its relation to literacy require a brief discussion. In French, instrumentum can be used to designate small finds in general, but in other languages such as English and German it is predominantly used as an epigraphic category. The Corpus Inscriptionum Latinarum (CIL) provided one of the earliest systematic approaches to inscribed small finds, by dedicating specific sections to it,¹⁶ and then with CIL XV a whole volume (Rome). The CIL generally included portable small finds in these sections entitled Instrumentum or Instrumentum domesticum, such as inscribed bricks, tiles, vessels made of ceramic and glass (including amphorae), lamps, metal ingots, finger rings, and various objects made of silver or bronze. Portability is one of the ways in which this category can be defined,¹⁷ and the one we adopt in this chapter. However, the term has increasingly been used more flexibly, perhaps under the influence of the vaguer German term ‘Kleininschriften’, and has become a catch-all for all non-monumental epigraphy—despite some efforts to claim it for the epigraphy of economy and production only.¹⁸ The most inclusive approach is perhaps that of RIB II, which, under the banner ‘Instrumentum domesticum (personal belongings and the like)’, even includes non-monumental inscriptions on non-portable supports such as wall graffiti and mosaic inscriptions. For Digital Epigraphy, the muddiness of the term has further created problems, because it has been deployed in several epigraphic data sets to encompass numerous object types (and even as a text type), lumping together such different finds as a Samian bowl with a graffito and a copper-alloy scabbard with open work lettering.¹⁹ This wide scope of the category and varied usage of the term instrumentum also complicate research on, and systematic overviews of, the material, as well as its inclusion in epigraphic studies on specific geographical areas or aspects of Roman History. This neglect stands in stark contrast with the potential of inscribed small finds for numerous research questions—for example, concerning production, trade and use of ceramics, and associated practices of literacy.²⁰

To counter this situation, efforts have been, and are being, made to gather inscribed small finds and to make them more accessible, both in print and online.²¹ For Roman Gaul, the pioneer project taking on this task is the Artefacts database with its epigraphic module, which will serve as a basis for the analysis in the second half of this chapter. It includes all kinds of inscribed manufactured small finds in instrumentum, but excludes coins and ceramic vessels used for storage or the preparation and serving of food.²²

Regardless of whether inscriptions on non-portable support are included or not, the complexity of the term instrumentum and its usage is further exacerbated by the fact that instrumentum inscriptum itself is so varied, in terms of the objects’ function and materials but also the inscriptions’ function. The most common text types found on instrumentum are makers’ and owners’ marks, but they include religious dedications, jokes, content labels, and much more. Instrumentum was also inscribed in a range of different execution techniques, which further complicates its relationship with language and literacy. To control the variegated material subsumed under instrumentum, epigraphers generally distinguish reproducible or serial inscriptions, such as stamps or moulded marks, from ‘unique’ ones, such as graffiti or dipinti, for example.²³ In some cases, the letters form part of the object, such as cut-outs or open work in metal or glass objects, for example, or fibulae in the shape of letters.²⁴ Such different execution techniques are related to the materials used and to the production processes of both object and text. The point at which an object was inscribed is in turn dependent on the function of the inscription, with instrumentum covering texts that can range from highly decorative examples, which have been carefully preconceived and executed, to spontaneous scribbles. It is not surprising, therefore, that it is often difficult to establish whose literacy a stamp or a graffito is evidence for exactly.

For example, the vast majority of inscriptions on instrumentum are related to production, in the form of makers’ marks.²⁵ They include rarer examples of handwritten signatures or, for example, marks related to the assembly of furniture,²⁶ but mostly they are stamps that detail the producer’s often abbreviated name. They form part of the production process of the objects that bear them, some of which were traded over long distances, meaning that their impact on the literacy at their find-spot is indirect and that they are instead evidence for literacy in the region or workshop where they were made. Furthermore in this context of production, we have to consider the size of the workshop and the status of the workers. In large lamp factories, for example, the stampers²⁷ used to mark dozens of moulds, and thus thousands of lamps were made probably separately, possibly by a maker outside the workshop itself.²⁸ In smaller workshops, however, it is quite possible that the stamper was made by the potters themselves, as is suggested by the many Celtic names on Gallo-Roman ‘Firmalampen’ with limited distribution (see Section 7.4.1), or by the use of the term av(v)ot, the Gaulish equivalent to Latin fecit, on local productions with limited spread, such as ceramic firedogs.²⁹

Texts on instrumentum also arguably display a wider range of literacy skills than the formulaic and often ostentatious stone inscriptions, many of which must have been carved by professionals. Where makers’ marks are restricted to a name, they may reflect so-called signature literacy, which can be considered to be a relatively weak indicator of general literacy skills. The same applies to the owners’/users’ marks that were made after production and are thus better indicators of literacy in the environment of the object’s find-spot. Inscriptions made by the users or owners of an object often consist of a name or simple marks such as crosses,³⁰ but less frequently they can also include more elaborate phrases—for example, relating to a specific event such as a banquet—suggesting more experienced writers. Sometimes evidence for more experienced writers can also be found within workshop contexts, as is the case at the large Samian ware production centre at La Graufesenque (see further Section 7.4.1). Professional writers can, for example, be witnessed through some of the rare votive inscriptions on small finds, inscribed on the object itself—for example, a small sculpture or a vessel—or on plaques that are sometimes attached with chains. Many sanctuaries may have provided votives produced on site, with inscriptions probably made by a professional.³¹

Research on inscribed small finds in the north-western provinces of the Roman Empire has gained momentum since the late 1980s, driven in particular by a series of publications, conferences, and projects in Hungary, the UK, and Italy.³² Until recently, this development seems to have passed by France, where some regional epigraphic corpora include graffiti on pottery, but no other inscribed instrumentum, with the exception of the Recueil des Inscriptions Gauloises (RIG), which dedicated a whole volume to it. In other European countries, this epigraphic treasure trove has been tackled systematically—for example, with the second volume in eight fascicles of the Roman Inscriptions in Britain (RIB) or the corpus of inscribed small finds from the limes region and some neighbouring parts of Germany by Stefan Pfahl.³³ While for France fairly complete corpora have been compiled on amphora stamps, tableware, and architectural ceramics, the only way to get an overview of inscriptions on other kinds of instrumentum from this area is currently to go through the indexes of CIL, CAG, and AE.³⁴

The groundwork for such a corpus is now being prepared based on the Artefacts.mom.fr database. Artefacts was developed over the past thirty years to gather small finds from Europe and the Mediterranean, make them accessible, and provide a platform to establish typologies. It provides a ‘universal’ classification for objects in the form of codes, organized by broad functional groups (personal objects, environment and habitat, activities) and ten chronological periods, from prehistoric to early modern. Attestations in the form of individual finds are added to a typological ‘fiche’ that describes the characteristics of an individual type of, for example, brooch. With this vast scope, Artefacts relies on the collaboration with researchers and projects willing to share their data and establish new typologies. To date (May 2024), Artefacts contains 208,239 objects, with 22,718 typological ‘fiches’, of which 12,600 are for the Roman period alone (Fig. 7.1).

In addition to a number of existing features such as a thesaurus and the option of generating distribution maps and statistical analyses, an epigraphic module has been added to the database infrastructure recently.³⁵ It is in the process of being filled with data and allows contributing researchers to categorize and describe in detail the inscribed objects in the database, with the structure of the epigraphic module ensuring a differentiated and systematic collection of texts on small finds together with related data.³⁶ Out of the 208,239 objects in Artefacts, 13,477 inscriptions on 13,324 objects have been registered to date (May 2024), of which 5,180 on 5,141 objects are from France.³⁷ These data can be analysed through a number of statistical tools, notably Metabase, and the records can be exported together with the relevant references in the form of print corpus entries. This will ultimately allow the production of a corpus of Roman inscribed instrumentum from France, but researchers can also tailor the export to their needs through filtering—for example, focusing on one specific object category (Fig. 7.2).³⁸

Owing to the collaborative nature of Artefacts and because the epigraphic module was set up relatively recently, the status of the different categories varies greatly, but its potential is already becoming clear. Eventually, it will be possible to use this tool to answer, for example, questions regarding the frequency of inscriptions on different object types (Fig. 7.3), the geographical distribution of the names of deities or individuals, the relationship of languages and scripts with object types, or the frequency of tria nomina on small finds.

The more complete categories, such as the fibulae and the so-called ‘Firmalampen’, already allow for an initial assessment in terms of language and literacy in Roman Gaul and shall be discussed in more detail in the next section. The data are moreover starting to show other promising series, often related to specific activities or occupations (Fig. 7.3) and thus allowing us to study the literacy of certain groups or in certain environments. One example are the signacula aenea,³⁹ stampers made of copper alloy. They are sturdy and therefore generally well preserved; to date, 679 inscribed signacula have been registered in Artefacts, of which 119 are from France. They feature the name or names of one or more individuals involved in an economic activity, sometimes including their status. We do not know exactly how these objects were used, but we can imagine that they served for stamping goods, or the documents accompanying them, in the context of production or trade control. Signacula are mainly attested in Italy, particularly in Rome and the south of the peninsula, but their use was taken up in Gaul from the first century bce century onwards, including for local production: one example from Carcassonne features a maker’s mark on its top, giving us the rare excitement of tracing its production to the workshop of a L. Cornelius Sabinianus in Narbonne.⁴⁰ Judging from the onomastics of those whose names were imprinted with such signacula, the individuals featured are often freedmen or slaves,⁴¹ but, interestingly, their cognomina are also largely Greek or Latin. Together with the fact that their distribution is mostly limited to the south of Gaul, this may suggest that they were used in environments in which Latin was commonly used, such as trade with Italy or fiscal matters related to the status of the settlements, as is suggested by the distribution of the finds for which we have the best documentation.⁴² Further research will show whether their geographical distribution is mirrored by, or related to, that of other inscribed instrumentum, and how we might be able to explain it.

There are several other promising sets of inscribed small finds. For example, the oculists’ stamps, an already well-known corpus, are a valuable source for our knowledge, not only about ophthalmological ailments and treatments, but also about the social status of physicians.⁴³ Commercial lead labels, which generally mention products that were the objects of a transaction (purchase, repair, cleaning, dyeing, and so on) and sometimes weights or volumes and prices, also feature the names of customers.⁴⁴ For the religious sphere, ex-votos provide evidence for religious practices and the language used in them, but also the onomastics of local deities and the people worshipping them.⁴⁵  Defixiones, which are often difficult to read, even more so because some of them are in Gaulish, provide information about the magical practices of those whose intimate concerns they show.⁴⁶ Brick and tile stamps, on the other hand, offer a huge amount of data on the producers, both with Celtic and Latin names, and with their distribution in relation to the legal status of municipalities.⁴⁷ These very diverse sources have in common that they shed new light on the daily life of the middle and lower social classes, precisely those provincials who had little or no means or reason to express themselves through monumental stone inscriptions. In what follows, the analysis will focus on two main aspects represented by inscribed instrumentum: the sphere of production and the messages on speaking objects.

To put a maker’s mark onto a crafted object was a common practice in the Mediterranean, particularly in the Roman period, when it was taken up widely. A systematic analysis of makers’ marks can provide unique information on crafts and production, in particular on its organization, and the material and physical conditions for production as well as the origin and social status of the craftspeople. The mark is usually a name that can be followed by a verb such as fecit—that is, ‘made by X’. In the context of larger-scale production and workshop hierarchy, the identity of the named person needs to be considered carefully, as the name could be that of the craftsperson, a ‘team leader’, or the owner of the workshop.

Out of the 13,477 inscriptions currently registered in Artefacts, and thus excluding the frequent stamps on Samian tableware, for example, makers’ marks make for more than 80 per cent of the total, with 72 per cent of them on ceramic, mostly lamps at present, and only 20 per cent on copper-alloy objects.⁴⁸ Although provisional, these numbers are not surprising. The production process of ceramic items lends itself particularly well to adding a mark to an object, and it is more efficient to reproduce a moulded stamp, as it was often used for lamps, than one that is applied individually to each item, as is generally the case with copper-alloy brooches.

Makers’ marks on brooches, or fibulae, are in fact a marginal phenomenon. Fibulae are an indigenous object type that saw a strong development during the Iron Age. During the Roman period, the typology changed, but brooches continued to be used by the local population, both male and female. Certain workshops managed to create ‘niche markets’ for specific types, often skilfully produced or decorated. Makers’ marks do not occur on such elaborate objects, however, but mostly on simpler, utilitarian models, and even on these they are rare: to date, there are ninety-seven makers’ marks on fibulae from France in the database, which, if we assume an estimated total of about thirty thousand fibulae from France, corresponds to around three marked fibulae out of a thousand.⁴⁹

What, then, was the exact function of these occasional marks? Generally, such stamps spell out the name of the bronze smith. On Aucissa-type fibulae from France, for example, so named after a prolific Italian producer of this type of brooch, we find Atrecto/Atrixtos, Aucissa, Durnacus, Iovincilos, Intro[?], and P. Valer(ius). Many of these names are indigenous, some with Latinized endings, and while ten stamps feature the Latin fecit, only one stamp, IVLIOS AVOT, uses the Gaulish equivalent.⁵⁰ There is, however, another way in which local, Gaulish productions are different from the Italian ones. Italian types such as Aucissa-fibulae feature a clearly visible stamp at the head of the bow running perpendicular to the length of the brooch. In the case of the Gallo-Roman series ‘pseudo-La Tène II’, however, which was produced from Tiberius to the end of the first century ce or later, the name of the bronze smith was stamped onto a part of the brooch that was then folded and flattened around the bow, often making it difficult to read the stamp on the finished product.⁵¹ One of the first marks of this type to be read in the nineteenth century, the mark of Litugenus on a fibula from Rennes (Ille-et-Vilaine), provides a good example.⁵² The illustration that accompanied its first publication shows the stamped part of the brooch as if it had been unfolded (see Fig. 4.1). On the other known examples by this producer, only the central part of the stamp, VGEN or GENI, is visible, and often misread, for example, as CEMIF or IVPE (for TVGE; Fig. 7.4).⁵³ Accordingly, the often incomplete or abbreviated makers’ names on this and other types of fibulae have been the subject of revised interpretations over recent years, also benefiting from recent progress in Celtic onomastics. Such revised readings have revealed hitherto unknown bronze smiths, almost always with indigenous names such as Bibidos, Boduacos, Comeilos, Conducios, Daribitus, and Oxtaillos.⁵⁴ If these marks were not meant to be read by the users of the brooches, we can assume that their function was somehow related to the production of the objects instead,⁵⁵ perhaps for quality control or to keep track of batches as the objects went through different stages of production. The writers, as well as the audience, of the inscriptions on these Gallo-Roman brooches were therefore craftspeople, and mostly local ones.

Only recently has the idea that literacy was widely spread among craftspeople become more accepted. Traditionally it was assumed that they led a simple life, which suggested to many that they lacked education and were for the most part, if not altogether, illiterate. The comprehensive publication of the findings from La Graufesenque helped change this picture (Chapter 5).⁵⁶ At this site near Millau in the south of France, a large production centre for Samian ware was excavated, where hundreds of pottery workshops fired their wares in large shared kilns from 20/10 bce onwards. The epigraphy from the site consists primarily of stamps and graffiti. Adding the stamps to bowls or plates during their production was thus part of the potter’s tasks, and we have good reason to think that the tool that was used in this step was made on site. The stampers used were usually made of bone, as found at La Graufesenque,⁵⁷ and spell out the name of the potter(s) or workshop owner(s), often preceded, or followed, by an abbreviation for ‘workshop’: OFMASCLI, for example, for of(ficina) Mascli/ ‘(by/from the) workshop of Masclus’, or OFICBILIC for of(f)ic(ina) Bilic(ati)/ ‘(by/from the) workshop of Bilicatus’.⁵⁸ As is the case with pottery stamps from other sites, some of the names are abbreviated to such an extent that the identification of the potter is difficult or even impossible—for example, Ar(dacus), Cen(sorinus), GR(—), or A(—).⁵⁹ In the most extreme of these cases, at least, the marks therefore seem unlikely to have served for the promotion of a workshop.⁶⁰ Instead, their function is more likely to have been exclusively related to the needs of the workshop and its immediate environment, perhaps to keep track of the volume an individual potter produced in relation to parallel production lines in the workshop or as the goods were given to a different workshop for certain steps in the production.⁶¹ To date, several thousand impressions made by the stamps of about five hundred different potters have been found in Millau, which demonstrates a very common use of epigraphic makers’ marks at this site. The limits of this apparently widespread literacy may be illustrated by the so-called griffes, marks that consist of only one or two letters or of unintelligible signs, and by anepigraphic marks, often rosettes, which are well attested at La Graufesenque as well as through numerous other workshops that produced tableware. From a social point of view, experts have pointed out that the majority of the stamps at La Graufesenque feature single names, so the producers involved may have been indigenous peregrines for the most part, even if a handful of duo and tria nomina seem to indicate a ‘not very ostentatious citizenship’, with gentilicia often reduced to a single letter.⁶² The small number of duo and tria nomina suggests that the more frequent single names are not simply due to a lack of space, which led producers to opt for one name, but that they in fact reflect the reality of the craftspeople’s status.⁶³

In southern Gaul, such stamps were placed on the inside of the vessel, but there are also marks by mould-makers that were traced in cursive directly on the moulds used for Samian bowls, for example. They appear inverted on the vessels made using the mould and are sometimes very faint. Again, these marks seem to have been directed not at the users of the finalized vessels but at others within the production process. There are, however, cases in which such marks were traced carefully in reverse capitals and thus appear perfectly legible on the finished product—for example, SABINI/MANVS ‘hand of Sabinus’, that is, ‘handmade by Sabinus’⁶⁴—demonstrating a very conscious and careful use of writing skills. In such mould-makers’ marks, dating from Tiberius to the second century ce, the names are often indigenous, but most of them, when the name appears complete, show the Latin ending -us.⁶⁵ Other graffiti from the excavations at the pottery primarily consist of firing lists, written with skill and particularly carefully onto plates before firing.⁶⁶ Through this method, the potters kept track of the numbers and types of vessels stacked inside the shared kiln, and these lists contain both Latin and Gaulish language material. We know around two hundred graffiti or fragments of graffiti from La Graufesenque, including the firing lists but also some phrases, mottos, and dedications, which are sometimes personalized and clearly addressed to customers.⁶⁷

While makers’ marks mainly represent a kind of signature literacy, their frequency and early appearance in local productions attests a widespread use of this kind of literacy among local craftspeople. Even if the stamper used to make the stamps was made by someone else for those signing with it, as is suggested by the division of labour seen at large production sites such as La Graufesenque, their significance for the production process implies that they were read by others as well. Together with the evidence of the graffiti by more experienced writers, the example of La Graufesenque thus shows a range of literacies and literacy skills at work, employed by the local population from the early first century onwards, for communication among craftspeople in the first place, but occasionally also with customers, and sometimes mixing Gaulish and Latin.

A slightly different case are makers’ marks on lamps, which are very frequent, particularly on moulded types of the so-called Firmalampen, oil lamps that were produced in large quantities and feature the name of the producer prominently on the bottom (Fig. 7.5). It is usually a single name in the genitive, less frequently in the nominative, with some potters using both cases. The corpus of these makers’ names on Firmalampen in Artefacts, which to date counts almost 190 entries,⁶⁸ already offers an opportunity to study the origin of the makers, their onomastics, and their literacy.⁶⁹ A minority of lamp-makers’ marks includes a gentilicium or initials that suggest tria nomina; in the Artefacts data set of stamps on Firmalampen of all provenances, only 56 out of 189, or 29.6 per cent, do, while 130 stamps bear single names.⁷⁰ No tria nomina on Firmalampen are known from France, but examples of duo nomina occur,⁷¹ and the fact that there are also examples of duo and tria nomina on other small finds as well suggests that the single names on lamps reflect the actual status, probably peregrine, of the craftspeople.⁷²

In Gaul, such oil lamps were originally imported from Italy, but, very soon after the earliest imports, local production started through complex processes of cultural and technological transfer. Fortis, who first produced Firmalampen in Modena at the end of the Augustan period, seems gradually to have extended his business to several workshops that then delivered his lamps to the whole north of Italy.⁷³ Export then seems to have expanded further north from Aquileia, towards the Danubian limes, with secondary workshops popping up between Aquileia and the border from around 60/80 ce onwards.⁷⁴ His name can, however, be found on much later types of lamps as well, with some of them found in contexts dating to the fourth century ce or later, showing that his legacy persisted over centuries.

Over such a long period of time, other potters must have been using Fortis’s name for the production of lamps in a number of places, and other producers also started using their own names, with their stamps turning up in workshops outside Italy as well. In Gaul, the import of Italian Firmalampen seems to start around the mid-first century ce. Almost immediately after that, lamps stamped with the names of Italian producers, but that were clearly produced with local clay and slip features, start appearing along the coast and in south-eastern Gaul. In the second century, local productions with stamps of Italian producers such as Communis, Phoetaspus, Strobilus, or Fortis also appear in eastern central Gaul as well as further east and west. This development can make it difficult to establish the origin of the producers involved: Litugenus and Nerius, for example, produced lamps in northern Italy before their goods were made in Gaul, as is the case with Strobilus, Phoetaspus, and of course Fortis. Yet, productions signed with names that are not attested in Italy, such as Attusa, Butrio, Capito, Catullus, Satto, Talutius, and others, are known from the second century ce and clearly the result of provincial initiatives. Butrio is attested in Lezoux, as Butrio or Putriu,⁷⁵ Talutius in Strasbourg, and others may have been based in Trier or in the Rhine Valley. Their products were traded west as far as Bavay, but seem to have had limited success; most workshops continued to produce lamps marked with Italian names, perhaps answering a market demand for ‘Italian’-style products, even if it was known that they were produced locally.

This, and perhaps their relatively late appearance, may also explain why the local producers’ names are Latinized from the beginning and why there are no instances of AV(V)OT on these lamps. How independent the local workshops were from the Italian ones is unclear, but the local producers clearly understood the function of such stamps, and, when they took over the production of Firmalampen-style lamps, they also took over the custom of adding a stamp. They also clearly had the ability to make this text type work for themselves by putting their own name on the lamps. While this may have required a relatively low level of literacy skills, it is worth noting that indigenous names are very common on local productions with limited geographical distribution, which implies that the local craftspeople made the stamps themselves. The prevalence of lamps with names of Italian producers, however, suggests that, unlike the ones on brooches, these marks may have been made not only—or not primarily—for the production process but also for the customers; they were a recognizable part of the model. As opposed to the brooches, which were commonly known and used in Iron Age Gaul, the Firmalampen may have continued to be perceived as imported objects even in the second century, and the stamps with Italian names may have been part of that perception.

Such subtle differences in the purpose of makers’ marks show how tightly related inscriptions on small finds are to the objects that carry them and reveal different approaches to, and uses of, literacy, and possibly language, in different spheres of everyday life. In the following section, yet another, completely different, kind of literacy will be discussed through ‘speaking objects’.

Strictly speaking, ‘speaking objects’ address their reader or user in the first person through their inscription. This is quite rare and often directed at potential thieves; a graffito on the bottom of a Samian ware bowl from Lattes (Gallia Narbonensis), for example, tells anyone that picks it up to ‘put me down, I belong to Domna’ (or ‘to the domina’).⁷⁶ This kind of inscription is even rarer on objects other than ceramic vessels, and there do not seem to be any examples from France.⁷⁷ By extension, this category also includes objects that carry a message through which one person addresses another, often presupposing an existing relation of some sort, and these are more common. They include the frequent utere felix ‘use happily’ inscriptions on brooches, spoons, and other utensils, as well as bibe ‘drink’ or gaudeas ‘be happy’ on cups, for example.

One particular group of such texts can be found predominantly in Gallia Belgica and part of the Lugdunensis, mainly from the mid-first century ce onwards. They are well known from fibulae but can also be found on other objects such as finger rings or writing equipment.⁷⁸ The discreetly inscribed messages seem to revolve around amorous relationships and seduction, often with intentionally ambiguous phrasing and tiny punched letters of 1–2 millimetres. Traditionally, and perhaps a bit rashly, such inscriptions have automatically been interpreted as messages from men to women,⁷⁹ but it is rare to find a text that is explicit about whom it addresses,⁸⁰ even though the small diameters of the finger rings and the dimensions and types of brooches generally suggest their recipients were female. It can also be disputed that all of these messages belong to the context of amorous relationships, as the less explicit examples could also refer to friendship. But, on the whole, the corpus of these inscriptions, defined by similar style and technique in writing as well as by their focused geographical distribution area, largely seems to carry erotic undertones, even if individual texts can appear innocent. To cite examples on fibulae, if the words misce, sitio, ‘mix [sc. the wine], I am thirsty’,⁸¹ can during a banquet be used to order a drink if addressed to the cupbearer, thirst can also express the longing for a lover. Similarly, ab aquis venio, ‘I come from the baths’,⁸² could indicate that the writer is clean. There are also more direct declarations, such as dulcis amo te, ‘sweetheart I love you’,⁸³  irascor et amo, ‘I am angry and in love’,⁸⁴ or even imploring and plaintive ones, such as succurre amanti si (me) amas, ‘help the lover if you love them/me’.⁸⁵ The request for reciprocity (for example, venio si das, ‘I’ll come if you give’),⁸⁶ also seems to be a theme in Gallo-Roman declarations of love more generally; finger rings and spindle whorls can carry similar declarations. A group of second-century spindle whorls found in southern Gaul is particularly well known for their inscriptions, which mix Gaulish and Latin language material, with some of them entirely in Gaulish.⁸⁷ This mix can also be seen in some of the second-century fibulae from France that, for example, bear the greetings ave adianto (‘hello, loved one’) and ave vimpi (‘hello, pretty’).⁸⁸

Such objects were original productions, with each of the inscriptions presenting its own, often unique, message, which could have been part of the design by the manufacturer or commissioned by a customer, but they can still be grouped into series that seem to exhibit local customs. The well-studied fibulae represent the most obvious case: the fibulae of several series with such inscriptions feature palaeographic characteristics that point towards a restricted area of production.⁸⁹ But some other workshops seem to have specialized in the production of inscribed objects as well. The spindle whorls, for example, seem to have been made in or near Autun, using the local bituminous schist. They sometimes address a particular person, which points to an individual commission rather than standardized production.⁹⁰ A workshop in Bonn was specialized in producing finger rings with messages often identical to those found on fibulae,⁹¹ and another example of a locally restricted phenomenon is found in the decorated styli inscribed with witty or amorous messages that seem to have been particularly fashionable in eastern Gaul and the neighbouring parts of Germania during the second century ce.⁹²

The specialized production implies that the craftspeople were technically those who wrote the texts onto these objects. A striking aspect of these inscriptions, as far as the ones with punched inscriptions are concerned, which include several types of fibulae dating to the first century ce, lies in the technical skills that were involved. The messages often run over several lines, and the tiny letters suggest that the craftsperson had a very steady hand and good eyesight. It is, however, interesting to note that different types of brooches have certain palaeographical traits in common—for example, the almost systematic rendering of the letter L in the shape of a reverse S, as can be found in cursive writing, or a T that resembles a Y, on early hinge brooches (Fig. 7.6). Such particularities might call into question the ability of the craftspeople to interpret a model that they were given. In lapidary epigraphy, for example, the text for commissioned inscriptions could be drafted and handed to the stonemason in cursive, which could result in ‘typos’.⁹³ If we assume a similar process for the fibulae, this could suggest that the craftspeople did not recognize an L in the cursive text that was given to them, since they replaced all the other letters with capitals, and this might also imply that the sense of the inscription may have escaped them. However, in that case one might expect more such ‘mistakes’ with letters that look very different in cursive, such as E, F, B, or D, which the writers seem to have transcribed correctly. It may rather be a case of a local variant, and perhaps a solution for rendering the horizontal hastae of L and T, which are important for deciphering these letters in small writing, in a more visible way when using tiny dots to write them.

In any case, ‘speaking’ objects arguably reveal more about the literacy of the people who commissioned and wrote them, and their audience, than makers’ marks. In the case of regional phenomena in particular, we can assume a certain geographical proximity of the workshop that fashioned the inscribed object, the person who commissioned it, and its end recipient or user. The texts are simple but amusing and suggest a relatively high level of literacy, sometimes including literary references or verse. The writer and the envisaged reader shared an educated background, in which the humorous tone of the inscriptions was presumably meant to be understood.

On the whole, this group of erotic or amorous inscriptions, on fibulae but also on finger rings and spindle whorls, for example, seem to bear witness to social interactions that were established in the Gallia Comata in the Roman period. They are the material evidence for a kind of relationship between lovers that must have been expressed in many ways, with the inscriptions representing only one aspect of it. Considering the scarcity of material and epigraphic evidence for the process of seduction in the Roman provinces, this group of inscriptions offers a unique insight into this topic, complementing in particular the rather conventional picture of marital relationships conveyed by stone inscriptions. The geographical focus of these inscriptions suggests that the success of this epigraphic practice on personal objects, and thus tied to interpersonal relations, could be based on local, perhaps even pre-Roman, indigenous practices of gift-giving and/or communicating in the context of friendly and/or erotic relationships. This may also be suggested by the relatively common presence of Gaulish language material in this category of inscriptions. The fact that these texts are longer, more elaborate, and less formulaic, than, say, makers’ marks, may simply make the Gaulish more visible in such inscriptions, but the playful nature and ambiguity of the messages also implies linguistic dexterity and its intentional use. This becomes particularly evident in cases where Latin and Gaulish are mixed.⁹⁴ Speaking objects thus provide evidence for the use of Gaulish for this specific kind of communication, carried by an educated part of society, probably of indigenous origin. It is tempting to think that they may have used writing onto small objects as an expression of a cultural tradition, keeping it alive well into the Roman period, but, in the absence of further evidence, we can only speculate.

A closer look at some of the inscribed small finds from Gaul has shown that instrumentum should be considered more than just marks that help us trace trade and production. They tell us about the people behind these objects, such as the local craftspeople who started making use of Latin literacy and text types from as early as the second half of the first century bce, for the purposes of their trade, communicating among themselves but also with customers. Among the many perhaps less eloquent makers’ marks, individual texts, including the firing lists from La Graufesenque, for example, reveal more extensive literacy skills and the ability to adapt writing to specific linguistic or material circumstances, on the whole painting a picture of a range of different literacies used by indigenous craftspeople. The different uses of makers’ marks on different objects illustrate an awareness of linguistic subtleties, which can also be seen in the speaking objects, where Gaulish is used well into the second century and languages are sometimes mixed in a playful way.

At the same time, the differences in the significance of makers’ marks on brooches and lamps have made clear just how closely the inscriptions on small finds are related to the objects that carry them, to their materiality, the geography and chronology of their use and production, and even the way they were perceived. Even though makers’ marks and speaking objects are only a small part of the vast category of instrumentum, just these two text types alone already allow insights into different uses of literacy, in different contexts, and for different purposes, which all existed contemporaneously and within the same society. They show that small groups of objects and texts, regional phenomena, and limited productions are key for our understanding of the multilayered nature of literacy and language use in ancient communities.

Initiatives such as Artefacts are needed, as thanks to them future research will be able to describe the literate landscape in the Roman provinces in much more detail, and, importantly, with an interdisciplinary and object-sensitive approach. For Roman Gaul, we will in time be able to expand and complement the already well-developed approaches to studying the complex relations of individuals to the indigenous languages—mostly, but not exclusively, Celtic in Gaul—and users’ marks will, for example, show to what extent the practice of writing entered such varied activities as weighing and measuring, commerce, provisioning of food, or healing, but also more personal concerns such as quarrels between neighbours or love stories. Only a systematic approach that makes inscribed small finds digitally accessible, allowing for complex analyses and international collaboration, will make it possible to identify and contextualize important smaller-scale phenomena across time, space, and different parts of ancient society, and enable researchers to take a much-needed look beyond lapidary epigraphy.