1Studies of literacy in pre-modern societies have most frequently discussed the phenomenon in terms of communication between readers and writers. It is viewed as a ‘technology of the intellect’, the primary function of which is to facilitate communication and the storage and retrieval of information. The classic analysis by Jack Goody interprets literacy as a technology which transforms cognitive processes and relationship with information, and is a fundamental enabling factor in the development of complex state societies. It argues that literacy creates the capacity for complex storage, recall and manipulation of information, and also prompts fundamental cognitive changes in the ways in which people categorise and process information.¹ Many critiques since the seminal work of Goody and Watt have attacked this model, or sought to refine it and produce more nuanced understandings of the relationship between literacy and power.² In particular, the extremely limited levels of literacy in the archaic Mediterranean and the extent to which early Mediterranean societies continued to be oral societies to a significant degree, has been at the forefront of research. Given that the evidence for early literacy is exclusively epigraphic, it is difficult to judge how widespread it was, and to what extent it was used for the sort of record-keeping storage of information envisaged by Goody.³ In the case of early Italy, however, the evidence which does survive points to a very restricted literacy which appears primarily in ritual and funerary contexts. In particular, the adoption of literacy appears to be closely linked with the emergence of dominant élites, and to have been used by those élites as a means of asserting their group identities.⁴

2In all these cases, the predominant trend has been to examine writing as a way of codifying language, and to approach it from the standpoint of language and orality. However, one logical flaw in these arguments is that most start from the standpoint that writing is primarily there to be read and to convey information, but this is problematic when applied to societies in which levels of literacy may be very low, or restricted to specific social groups.⁵ In societies where levels of literacy are very low, and writing is a relatively new technology, the visual impact and the symbolic significance of writing may be important as the content of the text, if not more so.⁶ This is particularly important in societies – such as those of ancient Italy - in which the writing most commonly seen by the population at large takes the form of inscriptions. In communities where writing has rarity value, its visual impact and simply its visual presence are as important as what it actually says.⁷ Given that our evidence for literacy in early Italy comes from writing on durable materials such as stone, metal or pottery, it is important to consider the fact that in some early Italian societies, writing may have been used for its symbolic value just as much (or possibly even more so) than as a technology for communicating information.

3This is a particularly complex issue when writing formed part of the pattern of funerary ritual and behaviour. A high proportion of funerary inscriptions in Italy, particularly those of early date were located inside tombs. Many of these inscriptions are on grave goods deposited in the tomb. These relate to either the giver or recipient of the funerary gift, but are not primarily commemorative in purpose⁸. However, a considerable number of funerary inscriptions inside tombs in some regions of Italy are not grave offerings but are commemorations of the dead. This raises some interesting questions about the purpose of epitaphs, their intended audience or readership, and by extension, the function of writing in the funerary sphere.

4Many studies of grave monuments, with or without written epitaphs, start from the standpoint that they are designed for public display and viewing, and therefore to communicate information about the status, kinship, and identity of the deceased, and his or her family.⁹ This interpretation works well for epitaphs which are inscribed in accessible contexts, such as on grave stelae which are placed in the open air, but is problematic for epitaphs which occur in closed contexts. Epitaphs which are incised or painted on the interior walls of tomb chambers, or are incised on stelae or funerary urns which are placed inside chamber tombs, were clearly accessible to, and aimed at, a much smaller and more restricted audience. However, these were at least accessible to specific groups of people or on specific occasions, such as funerals or rituals of the dead. The most acute interpretative problems are posed by those which are located in a completely closed context, such as those found in Puglia. In these instances, the function of the epitaph is clearly not as a straightforward way of commemorating the deceased or conveying information about him/her, but has a more symbolic function. This paper will explore some of these issues in relation to the epigraphy of Puglia, a region in which a high proportion of epitaphs were located inside closed and sealed tombs.

Literacy and Society in ancient Puglia

5Literacy develops later in Puglia than in some other areas of Italy. At the time of the first inscriptions, in the 6^th/5^th centuries BC, it was an area of developing state societies, particular in the southern part of the region. The 6^th and 5^th centuries are marked by a number of significant developments such as a growth in the size and complexity of settlements, construction of fortifications, developments on the sites of important sanctuaries, changes in territorial organisation, and greater economic specialisation.¹⁰ During the 6^th-4^th centuries, settlements acquire organised street plans, larger and more complex buildings, and areas set aside as public space.¹¹ However the style of urbanism was very different from that of the Greek polis. Many sites – although not all - are characterised by nuclei of houses and associated burials rather than by a single nucleated habitation area and spatial separation between burials and habitation.¹² The fortified areas of sites include areas of burials although some, such as Vaste and Manduria, also have extensive cemeteries outside the walls, sometimes aligned with the roads or gates.

6Socially and politically, there are signs that an élite was emerging in the 6^th century, and it remained an élite-dominated society. Greek sources describe the region as one of tribal or ethnos-based societies ruled by kings, variously referred to as basileus, tyrannos or dunastes.¹³ This broadly agrees with the archaeological evidence. An increase in wealthy grave good in the 6^th century suggests the emergence of a dominant and wealthy élite at this period and supports the assertions of the ancient sources that this was a deeply élite -dominated society, although probably one based on emergent state societies rather than the tribal chiefdoms described by the ancient sources.¹⁴ Overall, the 6^th and 5^th centuries are a phase of significant change in settlement patterns and socio-economic structures, characterised by the early phases of urbanisation, an emergent élite, and the development of strong group identities.

7The earliest inscriptions in the local language (conventionally termed Messapic) date to the late 6^th or early 5^th centuries BC and are contemporary with these developments (fig. 1).¹⁵ They are mostly concentrated in the south of the region and are written in a script closely based on archaic Greek alphabets. It was derived from a number of Greek sources, showing Tarentine/Laconian influence but also characteristics of some other Greek scripts.¹⁶ Innovation and local diversity is perceptible in the addition of a number of specifically local characters.

8For instance, the characters +, X, and may appear to be derived from Greek Χ, φ and Ψ, but there is strong reason to believe that they were used to represent different phonetic values in the Messapic alphabet.¹⁷ The use of inscriptions gradually diffuses throughout the region, although still with concentrations in southern Puglia, but the chronological profile and breakdown of types of inscription show some interesting changes (fig. 2). The number of inscriptions in Messapic remains fairly low in the 6^th-5^th centuries, with a modest rise in the 4^th century, but a huge increase in the period between the late 4^th and early 2^nd centuries.

Fig. 1 - Distribution of archaic inscriptions (6^th-5^th centuries BC). Locations underlined are those where early inscriptions have been found.

Fig. 2 - Distribution of archaic inscriptions (6^th-5^th centuries BC). Locations underlined are those where early inscriptions have been found.

9When these statistics are broken down by type of inscription, there is a clear change in the epigraphic habit, from a roughly equal number of votive and funerary inscriptions in the early period of literacy, to a much greater emphasis on funerary inscriptions in the 4^th-3^rd centuries – the period of the largest numerical increase. Early inscriptions include short votive inscriptions on pottery, deposited on ritual sites and mainly consisting of personal names or the names of divinities and a small number of inscriptions on grave goods, and items such as loom weights.¹⁸ Inscriptions on stelae or cippi are found at all dates, but not in large numbers and there is some doubt over their significance. Many are from contexts which could be votive rather than funerary.¹⁹ There are also a number of longer inscriptions, mostly (although not always) of relatively late date, and probably of a public nature. Given Some appear to be building dedications in the Greek manner, while others may be records of public or ritual pronouncements.²⁰ There is, however, an interesting change in epigraphic culture in the late 4^th century. From this point onwards, a large majority of inscriptions are from a funerary context, and almost all of these are epitaphs rather than inscriptions on grave goods.

Messapic epitaphs and their contexts

10The Messapic epitaphs are unusual in their context and relation to the rest of the burial.

11The predominant élite mortuary practice in Puglia from the 4^th century onwards is inhumation in trench graves lined and with massive stone slabs.²¹ There were also chamber tombs with a vestibule reached by an access staircase and a burial chamber, some of which were decorated with frescoes, although these are much less numerous.²² In addition there were some tomb-types which are categorised as semi-chamber tombs – larger than the usual inhumation grave, but with a similar structure.²³ However, chamber or semi-chamber tombs account for only a small number of élite graves, and inhumation in a stone-lined trench was the predominant form of burial from the 6^th century onwards.

12The corpse and grave goods were usually placed on a wooden couch within the grave, which was then covered with equally massive cap-stones and the trench was filled in. This type of burial was a completely closed context, which limited access to the grave and visibility of its contents and interior to the occasion of the funeral. From the 5^th century onwards, the re-use of graves, was not uncommon, and it seems possibly that they were used for several generations of the same family.²⁴ This seems to have involved the addition of another funeral couch after a period of years, once the original couch had rotted and deposited its corpse and grave goods in the bottom of the grave, a process which would have made the interior briefly visible again. However not all were re-opened and many contain only single burials, and even those which were re-opened for additional inhumations would have been closed and inaccessible for all but brief periods of time.

13A small proportion of these burials have inscribed epitaphs.²⁵ However, despite the sealed and inaccessible nature of most of these burials, these are not displayed outside the tomb but are almost invariably inscribed (or sometimes painted) on the inside of the stone lining, thus placing them in a position in which they would be invisible and inaccessible after the funeral rituals and the closing of the grave. Most were along one wall of the tomb, but others are found on the roof-stones, and one early example from Gnathia occupies an even more hidden location, being placed along the inner edge of one of the roofing slabs, in a position where it would be obscured completely by the adjacent roof-stone.²⁶ Epitaphs in semi-chamber tombs are very similar in that they are also placed in contexts which are almost completely closed. The relatively small number of epitaphs found inside chamber tombs may have been easier to reach, but even so, access is likely to have been restricted. The majority of funerary inscriptions were therefore placed in a position in which they would have been invisible, and consequently unreadable, for most of the time.

14The question of whether graves had any form of visible marker is controversial. Stone cippi, some of them inscribed, have been found at many locations in Messapia, but their function has been widely debated. Some occur in contexts which may be votive rather than funerary, but others appear to be associated with cemeteries. A small number of graves had a small stone cippus placed above them, inscribed with a personal name, which acted as a grave marker and epitaph,²⁷ but relatively inscribed cippi have been found in contexts where they are clearly a funerary marker. This may suggest that external grave markers were uncommon and that the internal epitaph was the only commemoration, but it is also possible that grave markers of this type were more common but failed to survive, or were not found in context. Cippi without inscriptions, however, are relatively common in cemeteries. At Vaste, for instance, the burials excavated at Fondo Melliche were marked with cippi placed above them, and a similar configuration was found at Alezio, but none of these grave markers were inscribed.²⁸

15The rarity of inscribed grave markers placed on top of the grave has led to suggestions that they were not an indigenous Messapic custom but should be viewed as an adoption of a Greek custom by some members of the local élite.²⁹ However, the widespread use of anepigraphic cippi suggests that the custom of marking the grave with an external marker became an important part of Messapic funerary custom. Lombardo suggests that cippi became an important part of both cemetery topography and funerary ritual from the 5^th century, acting as both a symbol of ownership of a tomb, and possibly as the focal point of a tomb cult associated with it.³⁰ However, the fact that a relatively small proportion of these were inscribed means that their primary purpose may not have been personal commemoration. We must consider the possibility that even where the location of the grave itself was marked, writing was reserved for the interior of the tomb.

16In terms of content, the epitaphs are very brief and simple. Almost all consist of only a personal name, usually given as a two part name consisting of name and patronymic (either nom. + gen. or as gen. + gen.).³¹ Single names also occur, expresses in either the nominative or genitive case, and tripartite names are sometimes found but are relatively rare.³² The suffix no added to a very small number of examples has been interpreted as a Messapic variant of the 1^st person dedication frequently found in Etruria and northern Italy (‘I am of...’) but this is rare and also debated by philologists.³³

17The gender profile of the epigraphic data is also interesting. The majority of people who were commemorated with a written epitaph are male, as is the case in most areas of Italy. However, the proportion of male to female epitaphs is unusually high. Only c. 20% of Messapic epitaphs commemorate women, and of those which do so, many are not, strictly speaking, personal commemorations. A relatively small number include a securely identified female personal name and several of these are joint commemorations to a man and a woman.³⁴ The largest category of female epitaphs are the tabara inscriptions, which amount to 37 inscriptions. The term tabara is believed to be a title denoting a priestess or cult official and may be derived from the Indo-European root to-bhora.³⁵ Most epitaphs of this type do not include a personal name, but consist only of the term tabara itself, and in most (although not all) cases, the name of a deity. The relatively small number of personal names which occur in tabara inscriptions are mostly, although not exclusively female, and the deities named are also mostly female.³⁶ These inscriptions are sometimes accompanied by a symbol cross-shaped symbol which may be associated with the cult of Demeter/Damatra.³⁷ Unfortunately, few of the tabara inscriptions have associated grave goods or skeletal material, which could be used to confirm whether these epitaphs are indeed female commemorations, but their identification appears to be plausible given the overwhelmingly female nature of the cults with which they are associated. This category of inscriptions, therefore, seems to commemorate the office held, rather than the personal or family identity of the deceased. The commemoration of women appears to focus on those who held a public role in the community, and commemoration which focuses on the personal and family identity of the dead woman is relatively uncommon, certainly in comparison with the commemoration practices for men.

Epitaphs and Literacy

18This practice of placing funerary inscriptions in closed contexts opens up some very interesting questions about the function of an epitaph and about the role of writing in this society. The single most influential approach to literacy in pre-modern societies – that of Jack Goody – places great emphasis on literacy as a ‘technology of the intellect’ – something which is primarily concerned with transmitting information. By definition, in this model, the relationship between writer and reader is crucial. Alternative approaches which stress the ritual nature of inscriptions in the early Mediterranean place less emphasis on this aspect, but the underlying assumption is still that the primary function of an inscription is to be read, and to convey information.³⁸

19However, in societies of very limited or restricted literacy, the presence and visual appearance of writing may actually more important than what it actually says. The very presence of writing signals the importance of an object and the high status of anyone associated with the commissioning or ownership of that object, irrespective of whether the viewer can read what it actually says, or understand the specific information contained in it. However, when the writing is placed in such an inaccessible position, it has – by definition – no visibility, or at best, visibility only for a very limited time and in a very restricted social context (i.e. the funeral). Most studies of ancient epitaphs have started from the standpoint that they are inscribed principally to commemorate something about the individual or family and their status or role in society. However, the placing of an epitaph inside a tomb which is then sealed, rendering it largely inaccessible, opens up the question of what exactly the epitaph is for, and also the question of what happens to our understanding of the role of literacy once the reader is removed from the equation.

20One important point is that these inscriptions are clearly very different from Greek or Roman epitaphs, which are, in most cases, placed in areas which are publicly accessible (or in tombs to which the families of the dead had access), and which convey significant information about the deceased and his or her status, family relationships and achievements.³⁹ The Messapic epitaphs are, in contrast, placed in a location which would only have been visible for a short period of time and would have been completely invisible once the tomb was closed and filled in after the funeral. This must inevitably cast doubt on whether their purpose was primarily that of personal commemoration, in the manner of Greek or Roman funerary inscriptions. Other factors also suggest that this was not their primary purpose. The fact that a small but significant number of tombs contain more than one inhumation, deposited over a period of years, but few have more than one inscription,⁴⁰ is a further indication that they may not have been primarily personal commemorations of each of the deceased, but relate only to one person, most probably the first interment and establisher of the tomb.

21The content of the inscriptions is also suggestive, as the typical male epitaph places strong emphasis on both ownership, by the frequent use of the genitive case, and on family or kinship group, through the presence of a patronymic. I would therefore suggest that what these inscriptions are doing is recording family ownership of a tomb, in the name of the head of the family or the person who established the tomb. Although we lack detailed information about aspects of Messapic society such as organisation of land-holding or citizenship, it is clear from archaeological, historical and onomastic evidence⁴¹ that the societies of Puglia from the archaic to the Hellenistic periods were hierarchical, and were politically and socially dominated by élites. This suggests that it was important to establish membership of an élite kinship group. Ownership of a family tomb, and celebration of the rites which went with this, may have been one way of doing this, and the addition of a written epitaph naming the first owner of the tomb may have been a way of strengthening this claim and making a permanent record of it.

22This interpretation is consistent with research on Messapic funerary ritual and burial practices, which links the increase in élite burials marked with cippi to the wider changes taking place in Messapic society from the 5^th century onwards, and which also suggests that the establishment of a family tomb (possibly as the centre of a tomb cult) was an important means of establishing and sustaining the group identity of these élites.⁴² Although it is dangerous to extrapolate from other contemporary societies, it is also worth noting that studies of Greek (and especially Athenian) grave stelae have suggested that epitaphs in the genitive case are intended to emphasise possession of a family tomb, and thus assert rights to property and status, and to membership of the citizen body, as well as reinforcing social norms and group identity.⁴³ This interpretation might also provide an explanation for the low number of female commemorations. If an inscribed epitaph was a way of asserting a particular status as head of a family, and as a member of a social élite, it is perhaps less surprising that women, who were unlikely to have had such a social role or exercised direct power and influence, were rarely commemorated in their own right, other than as holders of a particular ritual office which may have conferred a level of public influence and recognition.

23The question of why this practice took root so pervasively in the late 4^th century and why such epitaphs were placed in inaccessibly contexts rather than publicly displayed is still puzzling. However, comparison with other Messapic ritual and epigraphy practices may be helpful. There is strong evidence that underground contexts had ritual significance in Puglia. Use of caves (both natural and artificially constructed) as sites for cult activity is known from the Neolithic onwards, and is documented in the same period as the upsurge of funerary inscriptions.⁴⁴ The two biggest concentrations of ritual inscriptions, in the Grotta Porcinara and the Grotta della Poesia, are both from inside cave sanctuaries with very limited accessibility. Many of those from the Grotta della Poesia date to the late 4^th-2^nd centuries BC.⁴⁵ The placing of epitaphs inside tombs may therefore have been part of established ritual practice, and the intended readers have been of a distinctly non-human variety. In addition, the importance of the inscription seems not to lie in whether it was read for its information or not, but the symbolic fact that it was there.

24The other factor which may shed some light is a wider cultural issue. The period in which these epitaphs appear in large numbers coincides with other changes in epigraphic practices, and with wider cultural changes. It also coincides with an intense period of political and military pressure, in which may parts of the region were under threat from Greek condottieri employed by Tarentum, and then by Roman and Carthaginian armies. The region was in the forefront of the Pyrrhic and Hannibalic wars, being formally brought into alliance with Roman in 270 and subjected to more direct control in 200-150 BC. Evidence of a greater degree of Greek influence can be seen in some of the building styles of the 4^th-3^rd centuries, and also in the greater use of inscriptions such as building inscriptions, public inscriptions and coin legends, which may indicate a greater degree of monumentalisation of the urban landscape and also a greater use and adaptation of some aspects of Greek culture in some spheres of public life. However, this increasing adoption of aspects of Greek urbanism take place at the same time as a widespread adoption of élite burial practices using a very local and non-Greek form of commemoration. The chronological correlation between the large increase in numbers of funerary inscriptions and the pressures on the communities of the region from the late 4th century onward suggests that there may be some connection between increasing cultural, political and military pressure from hostile external forces and an increase in writing in the local language and script. It is possible that the increase in funerary inscriptions is at least in part a response to this pressure and an assertion of their own cultural identities and role in the community by members of the local élite.⁴⁶ The sudden rise in the number of epitaphs was undoubtedly mirrors changes taking place elsewhere in Italy. Numbers of inscriptions in all regions increase significantly in the 3^rd century, as literacy becomes more widespread and the practice of inscribing in various contexts becomes more established, but the increase in Messapia is particularly marked. Given that it also takes place against the background of increased pressure from outside forces, it may reflect an emphasis by the local élite s on family identity and the need to record and preserve it.

25 This evidence raises as many questions as it answers, but it may be possible to reach some tentative conclusions about the function of Messapic inscriptions and the significance of their hidden locations. The inscriptions appear to serve a number of purposes and can be interpreted on a number of different levels.

26Firstly, Messapic epitaphs are a clear example of the symbolic function of writing in early Italian societies. In as far as their message can be read, it appears to be directed at the living only in the context of the funeral. Thereafter, it cannot be read other than in a metaphorical sense by the dead or by chthonic deities. The long tradition of locating ritual sites in underground locations, as well as placing other types of inscription in closed contexts, suggests that this was a well-established aspect of Messapic ritual practice. At one level, therefore, the significance of these inscriptions was a symbolic or ritual one, and their function as a means of conveying information is limited.

27On a different level, however, they have an important role as means of storing information. An important function of the epitaph may have been to establish the ownership of the tomb by recording the name of the person (presumably the first person) buried in it. It records the name, and also usually the family affiliation, of the dead person and by doing so in writing – which was a high-status technology – it not only preserves the information about the owner of the tomb but does so in a form which conveys added prestige. The fact that it was placed in a location which was mostly inaccessible suggests that writing, in this case, had a greater role in storing information than in communicating it. It appears to have been more important that the writing was there, than whether it was visible or could be read. It is not clear whether the presence of an epitaph conferred any more specific benefit than added prestige and proof of ownership, or whether there was any legal or social distinction between people buried in uninscribed tombs and those containing epitaphs. However, ownership of a tomb of this type seems to be an important element in élite identity and was an assertion of membership of a social and political élite. An inscribed epitaph giving the name of the first owner of the tomb helped to assert the right to that membership.

28 Finally, changes in the wider epigraphic habits of the region in the late 4^th-2^nd centuries may reflect a period of broader cultural change in response to contact with (and military pressure from) Tarentum and from Rome. Some forms of inscription, especially building inscriptions and public inscriptions, may reflect the growing influence of Greek forms of urbanism become more frequent, but epitaphs retain their traditional Messapic form. Even though they are hidden and inaccessible, and are not designed to be read other than on very limited occasions, they – and the burial practices with which they are associated – have a strong symbolic function in establishing and maintaining a group identity. They indicate high levels of cultural continuity in funerary and epigraphic practices at a time of intense cultural and political change, and they also demonstrate the strong group identity of the local élites during this turbulent period of the region’s history.