BABesch 81 (2006, 91-133. doi: 10.2143/BAB.81.0.2014425)

The Birth of a Roman Southern : a Case Study Ancient written sources and archaeological evidence on the early Roman phase in the district, (3rd-1st century BC)

Douwe G. Yntema

Abstract

This paper focuses on southeast Italy in the 2nd century BC. This was the crucial period in which the founda- tions of were laid. The image of post-Hannibalic southern Italy was mainly constructed on the basis of ancient written sources. Three fundamental questions are discussed. The first question concerns the nature of both the archaeological evidence and the ancient written sources on this area. The second question explores the reliability of the pictures each of these seems to present. The third question concerns the relation- ship between these pictures. Can they be integrated in order to present a new and fairly coherent narrative?

1. INTRODUCTION1 have entailed an integration process of a native- Italic society, the 1995 paper explored the useful- This paper deals with a number of aspects of the ness of the concept that Martin 170-180 year period following the Roman incor- Millett applied to Britain (dialogue between Ro- poration or conquest of the district of Salento, man and ‘native’ culture) in the regional context of Italy, in the second quarter of the 3rd century BC. Salento (cf. Millett 1990). It emerged that although The Salento district is situated in southern Italy, many things changed in the first two centuries where it forms the heel of the Italian boot (fig. 1). after the Roman triumph over the native groups According to the author , in Salento, very few could be directly linked to Brentesion () and Taras () were the only more or less decent towns there in the early Imperial age, while the same district reportedly had quite a number of important settlements (poleis) in pre-Roman times (Strabo, Geography 6.3.5). However, my concern is not with Taranto, a town on the western fringes of Salento inhab- ited by people who called themselves Greeks. Instead, this paper discusses the district inhabited at the time of the Roman conquest by a non- Greek, ‘native’ population often referred to as the ‘’. The present paper complements an article on a similar subject (Romanization) published a few years ago (Yntema 1995a). In that article I used the black-box system, comparing the archaeological evidence for various aspects of the period pre- ceding the Roman occupation with evidence for the same aspects in the period following Roman intervention. Continuity and change between the two periods were highlighted and explanations offered when the patterns observed in the data seemed to suggest them. Since the de facto incor- Fig. 1. Salento and surrounding area: poration of Salento into the Roman state must sites mentioned in the text.

91 the dialogue between the native societies of evidence and the ancient written sources. What Salento and the Roman world of the 3rd and 2nd exactly are they and what kind of picture do they centuries BC. The Salento groups were in fact present? The second question involves interpre- engaged in a much broader ‘cultural conversa- tation. Are the pictures they present legible and tion’ involving native-Italic speakers with various sufficiently reliable for the period under discussion roots (Messapians, Oscans, , , or do they evoke a hazy, somewhat panoramic Romans etc.) and different types of Greeks (e.g. Ita- landscape, fusing a range of clichés and/or cen- liote, Siciliote, Aegean and Alexandrinian Greeks). turies into a single, blurred and badly distorted I concluded that most of the changes affecting image? The third question concerns the relation- Salento in the first two centuries after the Roman ship between the archaeological evidence and the conquest were certainly not the result of Romani- ancient written sources on Salento. Are the pic- zation, irrespective of how we define the term. tures constructed on the strength of each type of The vast majority would probably have occurred source more or less compatible with one another? - though perhaps not with the same intensity or at Can they be integrated to present a fairly coher- the same pace - even if the Romans had decided ent narrative about early Roman Salento, or are not to intervene and had not succeeded in con- they contradictory, and do we need to explain quering the Salento district. Material, economic, why they differ so much? socio-political and religious features of early The period selected covers the later 3rd, the 2nd Roman Salento actually appear to have derived and the early years of the 1st centuries BC. This is from a wide range of Italic and eastern Mediter- the time between the Roman conquest of the ranean cultures. The dialogue model, with its only Salento district (267/266 BC) and its de iure incor- slightly veiled bipolarity (native versus Roman poration into the Roman state as a result of the culture), appears too simple and too antithetic for Social War in 90/89 BC. These two events have Salento and perhaps in general. Becoming Roman not been chosen as the chronological boundaries in Salento implied much more than an intense for this paper because of their undoubted mile- dialogue between two cultures. The concept leaves stone status in Salento history. Instead, the delim- insufficient room for the substantial regional itations are fairly arbitrary. Events like these are diversities of southeastern Italy and the probably unlikely to have such strongly ‘delimiting’ prop- even greater regional diversities of the eastern erties. When the people of Salento technically Mediterranean. Moreover, contacts with a wide became Romans in 90/89 BC, they did not sud- range of other cultural groups confronted the denly start to speak or to think Roman. The inhabitants of Salento with various possibilities: main reason for selecting this particular period is they could accept, ignore or reject features of the the existence of both a substantial body of archae- cultures with which they came into contact (see ological evidence and a range of written sources also Curti et al. 1996; Terrenato 1998). These cul- concerning those years in the Salento peninsula. tures included the Romans, the neighbouring The preceding era (4th/early 3rd century BC) is native groups of the Lucanians and Peucetians, rich in archaeological evidence, but poor in ancient the neighbouring Greeks from Taranto, written sources, while evidence of both types is and Corfu, and the distant Ephesians, Rhodians scarce for the late Republican and early Imperial or Alexandrinians who may have visited Salento periods. However, we should also concede that ports. there is little ancient written evidence from about The basic aim of the present paper is to con- the mid-2nd century BC onward. tribute to our understanding of a southern Italian The main focus, therefore, will be on the first district (Salento) during the first 170 to 180 years hundred years of indirect Roman rule. The sources of Roman rule. Since I have previously discussed on early Roman Salento discussed here may offer (Yntema 1993a, 195-213; Yntema 1995a) several valuable information about this period of crucial aspects emerging from data recovered by archae- change in the region’s history - the period in ological methods such as excavations and field which Salento became integrated into the Medi- surveys, the subject of this paper is to evaluate terranean world of Hellenistic-Roman times. In the archaeological evidence and data from writ- the following section I briefly outline the events ten sources about the post-conquest period in the surrounding the Roman actions that made up the light of one another. Both types of sources are dis- conquest/occupation of Salento in the second cussed. The present paper hopes to suggest quarter of the 3rd century BC. I then discuss the answers to three fundamental questions. The first ancient written sources (section 3), and briefly concerns the nature of both the archaeological examine the archaeological data (section 4).

92 Section 5 presents the narrative that can be con- mans. In 272 BC Taranto fell into Roman hands. structed on the strength of evidence from the pre- Soon after the capture of the Greek town, the sent research. In the concluding section, I suggest native Salento districts were conquered by the answers to the three fundamental questions for- Romans. This happened in the years 267-266 BC, mulated above. This section in effect summarizes making Salento the last district in peninsular Italy sections 3 and 4, which discuss the nature and to be incorporated into the expanding Roman interpretation of both the archaeological evidence state. and ancient written sources, and seeks to make The Capitolini commemorate explicit the relationship between both types of the Roman successes in Salento by mentioning sources as this emerges from the narrative in sec- two triumphs for the years 267/266 BC. They tion 5. record the names of the Roman commanders who were responsible for these glorious deeds and 2. THE ROMAN CONQUEST OF SALENTO who were accordingly permitted to celebrate a tri- umph: consuls M. Attilius Regulus and L. Iulius Salento is Italy’s gateway to the eastern Medi- Libo on 25 January of the Roman year 486 a.U.c. terranean. Separated from the by the ‘de Sallentineis’, and consuls N. Fabius Pictor and merely 70-km wide Strait of , it is an D. Iunius Pera on 7 February of the year 487 a.U.c. almost mandatory stopping place for those going ‘de Sallentineis Messapeisque’. offers no account to or coming from Greece. The district has often of these events: his ten books covering the periods been regarded as the backyard of the colonial- between the and the second Punic Greek settlement of Taranto, which was reportedly war are lost. The conquest is very summarily founded in the late 8th century BC. The indige- mentioned in Florus’ Epitome (1.3.1-2) and in the nous population who - according to Greek authors Periocha of Livy’s 15th book.4 The general tenor of of the 6th, 5th and early 4th centuries BC - inhab- these and other written sources is that the Romans ited this area were usually labelled ‘Messapians’.2 believed the Salento peninsula to be inhabited by The Messapian groups, however, were not exclu- two tribes - the Calabri and the Sallentini (cf. Lom- sively linked to Taranto. They appear to have had bardo 1992; Lomas 1993).5 The former gave its a much wider range of external contacts during name to the district in Roman times (),6 the period under discussion, for instance with while the latter survives in the present-day geo- parts of the Corinthian Gulf, Corfu and present- graphical name of the peninsula - Salento or Peni- day Albania across the Adriatic and the Metapon- sola Salentina. to district and the north-Apulian of It should be noted that while Greek writers of southern Italy.3 the 6th, 5th and 4th centuries called the natives of There is evidence to suggest that the focus of Salento ‘Messapians’, the Greek and Roman au- the Salento groups gradually turned to Taranto thors of the late Hellenistic/early Roman period from about 370/350 BC. There is probably a close often suggested that the district was inhabited by link between the tremendous boom of Taranto two tribal groups - the Calabri and the Sallentini and the great prosperity of the native districts (most authors). This discrepancy can be explained surrounding that Greek polity during the later 4th in several ways. On the one hand, the earlier label and early 3rd centuries (Yntema 1993a, 1995). ‘Messapians’ may have been a general term given Perhaps common interests resulting from this pri- by outsiders such as Greeks to the population of marily economic tie (although there may have the entire area, or it may be the ethnic label for been social and political links as well) prompted just one group in the area. In this latter explana- the Tarantine Greeks and Messapian natives to tion the term ‘Messapians’ is a kind of collective unite when a dangerous conflict with Rome broke noun covering a complex tribal reality (cf. the out in the 80s of the 3rd century BC. The Sallen- similar use of names such as , Lucanians, tinian native populations sided with Taranto in Samnites etc.). An interpretation in this vein is the polis’s conflict with Rome over the supremacy suggested by Strabo’s texts (see note 5). On the of southern Italy. After the Greek condottiere Pyr- other hand, tribal structures are less stable socio- rhus of Epirus, who had assisted the Tarantines political units than states. Fission and fusion are in their struggle against Rome, left Italy in order inherent in tribal groups: a tribe may split up into to try his luck elsewhere, the weak southern coali- new tribes or two or more tribes may fuse into tion consisting of a series of Italiote towns headed one.7 The coexistence of the names Messapii, Ca- by Taranto and various non-Greek polities in labri and Sallentini in the works of Greek and southern Italy was unable to withstand the Ro- Roman authors, all used for people inhabiting the

93 Fig. 2. Brindisi and surrounding area: sites mentioned in the text. same district, may be due to the fact that the orig- this text probably the north and central Apulian inal Messapian tribe split into two subtribes at tribal groups)8 and the Messapians of southern some point of its history. in his long list of Roman and allied forces In this paper, I generally refer to the native in- drawn up for 225 BC. Since the Romans mostly habitants of the Salento peninsula as ‘Messapians’. allowed defeated groups to manage their own in- This label (probably of Greek origin) is used for ternal affairs as long as they met their obligations the sake of convenience only. I occasionally replace towards Rome (cf. Salmon 1982), the Salento native it with the possibly more correct term Calabri/ populations may have been permitted to govern Sallentini. The term ‘Messapic’ is reserved here for their own territories in much the same way as the language (of the Illyrian branch) spoken by they had before. Of course, their contacts with the various native groups who in pre-Roman times other of Rome and with ‘foreign’ states such inhabited the area of present-day Apulia (i.e. the as the Greek polities across the Adriatic had to be people referred to as , Messapians, Cala- strictly in line with Rome’s wishes. brians, Sallentinians, Peucetians and Daunians in This, then, was probably more or less the gen- the Greek and Roman written sources). eral situation immediately after the Roman con- From 267/266 onward Salento was a - or per- quest in the outlying Italian district of Salento, haps more than one - de facto client state of the some 400 miles southeast of Rome. The tribes of Romans (cf. Salmon 1982). No peace treaties or the Sallentini and Calabri licked their wounds, met provisions have been preserved regulating rela- their obligations towards Rome and quietly con- tions between Rome and defeated states or tribes tinued to mind their own internal affairs. Initially, for the 3rd century BC. Mostly, the terms of peace it was in many ways business as usual in post- have been extrapolated in a very general way Roman-conquest Salento. Soon, however, a new from treaties with other Italian or Mediterranean Roman initiative was to change the native world districts in the 4th or 2nd centuries BC (cf. Lomas of the district decisively. For reasons unknown, 1993, 77 ff.). Therefore we do not know which the Senate in Rome decided to found a Latin official status the Romans granted to the Calabri colony in Salento. This actually happened on 5 and Sallentini. The treaty binding the Messapians August 246 or 244 BC, when the colonia latina of to Rome must have been based on the so-called Brundisium came into existence (e.g. Sirago 1993). Foedus Cassianum (cf. Salmon 1982, Lomas 1993). The foundation of the colony doubtlessly resulted Since we do not know its contents, it is uncertain in an influx of central who settled on for- what rights the Calabri and the Sallentini had and mer Messapian lands, which means that these what obligations they had to the Roman state. lands must have been confiscated at some time They were almost certainly obliged to supply between 266 and 244 BC. The actual number of troops in case of war. (2.24,10-11) seems settlers is unknown, but is estimated to be in the to confirm this by including both Iapygians (in order of 2,000 to 2,500 (Clemente 1988).

94 The Latin colonia was not a completely new set- discussion. The ‘literary’ texts have been collected tlement, however. Brindisi had its origins in the in Lombardo’s I Messapi e la Messapia nelle fonti let- Bronze Age and has probably been continuously terarie greche e latine (Lombardo 1992), and consist inhabited to the present day.9 On the eve of the of relatively short, highly dispersed passages in foundation of the colonia latina, the town of Brindisi Greek and Roman authors like Polybius, , may have been an indigenous settlement of some Livy, and Varro. Inscriptions also inform standing. It was probably one of the walled op- us about early Roman Salento and its inhabitants. pida of the Calabri (Marasco 1988; Burgers 1998; Some are in Greek, and a few in Latin (cf. Susini Lippolis/Baldini Lippolis 1997). Its name is said to 1962). There are probably quite a few inscriptions derive from the Messapic word for stag’s antlers in the regional Messapic language which date to (brention, brenta):10 the settlement is situated on a the later 3rd and 2nd centuries BC. The vast major- rise between two inlets that roughly resemble ity have been collected in the Corpus Inscriptionum antlers (fig. 2). Since the settlement has been in- Messapicarum (Ribezzo/Santoro 1978). As a rule habited without interruption to the present day, they can be dated only very approximately on the pre-Roman (but also Roman) remains can only epigraphical grounds (the shape of the letters be found in small ‘urban’ digs and are therefore etc.). They are not always easy to decipher and very poorly known. For this reason we cannot stem almost exclusively from burials and sanctu- establish with certainty the exact status of the aries. Because of their religious and funerary town of Brindisi in the pre-Roman settlement sys- nature, the inscriptions in Messapic appear to tem of the district of the Calabri. contain mainly personal names and standard for- The foundation of a colonia latina at Brindisi had mulas. A large and highly interesting body of serious consequences for regional power structures Messapic inscriptions from the Grotta della Poesia in Salento. While the Calabrian and Sallentinian sanctuary at the site at Rocavecchia (north of decision-making bodies constituted the sole polit- Otranto) remains unpublished.11 ical system during the 4th and early 3rd centuries Regarding Salento’s history, the Roman con- BC, it is plausible that two different political sys- quest and pacification of the district was described tems existed side by side from 246/244 BC on- in Livy’s now lost second decade. From that time ward: the traditional Messapian system still sur- onward Salento was essentially an outlying dis- vived, while the colonia latina’s political system trict of an increasingly Romanized Italy. The dis- would have been a close copy of that of Rome. In trict was some 600-700 km from Rome, the centre addition, various factors caused Brindisi to become of power in Italy. The distance to the major Greek the regional centre of power. The elite families of centres of the eastern Mediterranean was even several shrinking native oppida in the Brindisi dis- greater. Therefore, most Greek and Roman authors trict are likely to have converged upon this boom- were not even marginally interested in Salento. ing town (cf. Yntema 1995a, 164-165). The founda- Only in the event of any kind of contact between tion of the colonia actually marked the beginning Italy and the eastern Mediterranean did Salento of integration between the local Messapic-speak- enter into the picture. The region served as a step- ing groups and an increasingly large group of ping stone for travel, trade, commerce and mili- newcomers consisting of Latin speaking settlers tary activities to and from the southern Balkans and others - including Greeks - who were attracted and other parts of the eastern Mediterranean. to Brindisi in particular and Salento in general. Since travel and trade were everyday activities in The town was rapidly becoming a focal point in the Hellenistic and Roman world, they were hardly the intensified Mediterranean trade networks of worth mentioning. But once the southern Balkans Hellenistic times (cf. Marasco 1988, Yntema 1995a). and Italy belonged to different and - in particular - Since Rome fought a series of wars against Greek conflicting polities, Salento suddenly featured in polities of the eastern Mediterranean in the 2nd the written sources. Just as throughout its century BC, Greek slaves may also have become long history was a gateway for moves from Italy part of the Salento population in the period dis- towards Africa and vice versa, Salento was the cussed in this paper. gateway for ‘Italian’ ventures into the eastern Mediterranean. 3. ANCIENT WRITTEN SOURCES In view of the above, it is clear that references to Salento and the Salentine situation were mostly The literary sources on early Roman Salento are made only in passing in the Greek and Roman not very numerous. The area was far from being written sources of the Greek-Hellenistic and Roman a centre of political power in the period under Republican periods. Moreover, these writings often

95 do not refer to events, but to the district’s econ- hand, ancient writers describe conquered (mainly omy and geography (e.g. Strabo). Invariably, we tribal) societies that originally had dispersed can view the inscriptions as personal documents forms of settlement in precisely the opposite that can hardly be expected to make a significant terms. There, becoming Roman meant a transfor- contribution to our understanding of the much mation from barbarians to new men (Dench 1995): wider world of the native-Roman societies of Sa- poor and warlike ‘savages’ became prosperous lento of the 3rd and 2nd centuries BC. Together these town dwellers and, ultimately, Romans. In Roman written sources are unlikely to offer a clearly com- eyes, these people benefited greatly from the bles- prehensible picture of the region in the period sings of Roman civilization and culture. We should between the Roman conquest and the Social War. therefore be alert to this potential bias in the They may, however, give us interesting glimpses ancient written sources. A fourth and final bias into Salento societies in the period following the can be found in interpretations by recent histori- Roman conquest of 267/266 BC. ans of the ancient sources about southern Italy in The specific, almost casual nature of the sources the Roman-Republican period (cf. especially on early Roman Salento has both advantages and Toynbee 1965). Toynbee’s belief that the Roman drawbacks. With regard to the former, such casual occupation of southern Italy and ’s pro- remarks clearly do not have a strong ideological longed stay there had disastrous consequences charge. For instance, when Livy (42.27.8; on the has had a major impact on subsequent research year 172 BC) states that ‘three commissioners (cf. Gabba/Pasquinucci 1979; Sirago 1993). It has were sent to Apulia and Calabria in order to buy been a widely-held view that the basis was laid corn for the Roman navy and army’, there is a in the 3rd century BC for the poor, sheep-rearing good chance that Apulia and Calabria produced Mezzogiorno, characterized by large ranches and a surplus of cereals. There are reasons to believe absentee landlords, politically dominated and that this was in fact the case and that surplus pro- economically looted by foreign overlords.12 The duction of corn was a structural element in the 20th-century Mezzogiorno clichés also contributed regional economy, since other sources about ap- significantly to the rather bleak picture that was proximately the same period contain the same or currently being constructed for Roman southern very similar information (e.g. Appian 7.35 on the Italy (Gabba et al. 2001, 29). year 211 BC; Varro, De Re Rustica 2.6.5: on the The specific nature of the written sources on later 2nd or early 1st century BC). Salento as shown above means that ancient However, there are obvious and serious draw- authors highlighted only a few aspects of Salento backs with these types of sources. First and fore- societies of the late 3rd, 2nd and early 1st centuries most, of course, is their extreme patchiness: we BC. These related to transport routes, the econ- peer into hazy landscapes, each with one or two omy, uprisings and people of Messapian descent. small, bright spots. Secondly, we should bear in They will be discussed in the following sections. mind that the ancient authors who wrote these passages were almost invariably outsiders: the 3.1. Transport routes vast majority had never visited the Salento district and were recording second- or third-hand infor- The Salento road systems have been discussed at mation derived from the Roman annalistic tradi- considerable length by Giovanni Uggeri (1975, tion or other sources. Thirdly, we need to remem- 1983). By the late 4th century BC, the Messapian ber that Salento was a de facto conquered area. It oppida were connected by roads which, given the is generally acknowledged today that such dis- fairly considerable surplus production of the area tricts may have been described in cliché terms in (cf. Yntema 1982, 1993), must have surpassed the Graeco-Roman literature. On the one hand, con- humble status of mere paths through the wilds of quered urban societies with strongly centralized Salento (fig. 3).13 Some of the existing roads were forms of government such as Salento were upgraded by the Romans for both military and believed to have been negatively affected by the economic purposes. One example is the south- Roman conquest and were said to have suffered west-northeast road that crossed the Salento isth- decline: they are characterized in the written mus from Taranto to Brindisi (becoming the last sources by economic crisis, depopulation and stretch of the via Appia), and the northwest-south- extensive forms of stock raising (cf. Alcock 1993, east road that followed the Adriatic coast, known 1 ff.; Lomas 1993, 115). In Roman eyes, becoming to Horace as the via Minucia (in Epistulae 1.18.20) Roman in these areas meant loss of prosperity, and renamed in Imperial times as the via Traiana loss of culture and loss of identity. On the other Calabra. They enabled travellers to proceed with

96 attention to other parts of the Mediterranean world, Brindisi’s role in the trans-Adriatic link became especially marked when Rome began meddling in Greek affairs from the early 2nd cen- tury onward. Lombardo (1992) presents a sub- stantial series of texts by both Polybius and Livy suggesting that Brindisi became a crucial base for provisioning the Roman armies and a departure point for military activities in Greece and other parts of the eastern Mediterranean in the first half of the 2nd century BC. Brindisi continued as a major port in Cicero’s time.16 A few years after his consulship, Cicero attended celebrations there of the town’s anniversary. The date, 5 August 57 BC, was also the birthday of his daughter Tullia, who happened to be waiting for him at Brindisi on his return from exile in Thessalonica and Dyrrachium (Cicero, Ad Atticum 4.1.4).

Fig. 3. Pre-Roman road system in Salento 3.2. Economy (adapted from Uggeri 1975). The information on the Salento economy in early considerable speed, as demonstrated by M. Porcius Roman times that can be derived from ancient Cato who reportedly covered the distance between written sources is very scarce indeed. I have the Salento port of Otranto and Rome (some 700 referred above to the general view of historians km) within five days; this occurred under rather that the consequences of Hannibal’s stay in south- stressful conditions in 191 BC (Livy 36.2.6). In ern Italy were rather disastrous; some believed early Augustan times, the poet Horace took the that the area never truly recovered (e.g. Toynbee via Minucia/Calabra in the opposite direction at a 1965; Gabba/Pasquinucci 1979; Sirago 1993). If much more leisurely pace (Sermones 1.5.77-104). we focus on Salento, however, it appears that the The economic importance of the Salento road presence of Carthaginian troops was mainly con- system in Republican times is stressed in a cele- fined to Taranto and its immediate surroundings. brated passage in Varro,14 which suggests that Salento probably had a particularly unpleasant mercatores with mule trains transported olive oil, summer in 213 when Hannibal is reported to have wine, corn and other commodities from the north- roamed through the peninsula. Some Sallentinorum ern Salento plain to Brindisi on the coast. The ignobiles urbes defected to him that year (Livy Salento roads were therefore vital to the trans- 25.1.1). The actions of Rome to retake Taranto and portation of agricultural products to Brindisi, the ‘Sallentinian towns not worth mentioning’ where they could be consumed or packed and once again made Salento the scene of war in 209 shipped on to destinations across the seas. BC.17 Thus the war damage in Salento may not The town of Brindisi, ‘belted by a beautiful fair have been as disastrous as often supposed (cf. haven’,15 is generally regarded as one of the major Lomas 1993, 117-119). While several towns may ports of Roman Italy from the 2nd century BC. It have sustained damage and animals may have continued as such until the late Empire when it been taken by passing troops, agriculture - a major was gradually outstripped by the more southerly source of wealth - is unlikely to have sustained Otranto (e.g. Marasco 1988; Sirago 1993). There is long-term damage.18 reason to believe that the town performed a sim- Written sources on the agrarian sector in Salento ilar, though decidedly less important, function in covering the post-Hannibalic period seems to pre-Roman times (Marasco 1988). The first signs of support this. Salento actually exported wine, cere- Brindisi’s newly acquired importance in a Roman als and olive oil at this time (e.g. Varro, Livy, context can be found in Polybius, who states that Appian, Velleius Paterculus; see texts in Lombardo the town was both point of departure and point 1992).19 The sources attesting to the importance of of return for the sizeable Roman forces setting out Brindisi harbour have been briefly touched on in for the Illyrian war in 229 BC (Polybius, Historiae section 3.1 above. It therefore follows that Brindisi 2.11.7). While the drew Roman must have been a thriving town that developed

97 into the region’s primary economic centre (cf. Ma- ond passage, the ‘conspiracies’ of these herdsmen rasco 1988). This evidence - together with Brin- and investigations resulting from the senatus con- disi’s early 2nd-century role as provisioning base sultum de Bacchanalibus of 186 BC are mentioned for the Roman armies in Greece - suggests that in the same breath. In both cases the praetor re- Salento produced a regular surplus of cash crops. sponsible for Taranto (praetor cui Tarentum provin- It would therefore appear that the bad reputa- cia evenerat) went into action.21 The situation was tion southern Italy earned in Roman times does far from stable at that time in the provincia under not in any case apply to early Roman Salento. The discussion. That the Romans believed the situa- passages cited to reinforce this view (see Toynbee tion to be dangerous is apparent from another 1965, Gabba/Pasquinucci 1979, Sirago 1993) refer Livy passage. Shortly afterwards, in 181 BC, the to other parts of southern Italy or to other periods Senate expressly ordered the praetor L. Duronius, of Roman history, or are simply instances of in- who was charged with Apulian matters (praetor vented history. This is illustrated in a passage cui provincia Apulia evenerat), ‘to extirpate the evil from Strabo who, writing explicitly about Salento, in order to prevent it from spreading again.’22 It stated: ‘Once this whole area was densely popu- seems evident from these three passages that lated and had thirteen towns. But now these are there was considerable unrest in the area that fell small settlements, except for Brindisi and Taranto. under the jurisdiction of the praetor assigned to So badly off are they’ (Strabo, Geography 6.3.5). Taranto or Apulia in the decade between 190 and The passage was written during the reign of Au- 180 BC. These persistent mala required the undi- gustus and tells us only that many Salento towns vided attention of at least three Roman praetors had dwindled into insignificance. However, the or propraetors: L. Postumius in 185 and 184, L. preceding sentences in the same passage from Puppius in 183 and 182 and L. Duronius in 181. Strabo tell us that the soils were of excellent qual- It is not clear, however, whether these three ity, that there were good pastures and that it was sources actually relate to the Salento peninsula. rich in trees, including fruit trees. The decline in Although the wording of the first two might sug- the urban population described by Strabo in Au- gest that the unpleasantness occurred somewhere gustan times was used by ancient historians to in the Taranto area, the words ‘in Apulia’ and the portray early Roman Salento in particular, and mention of L. Duronius’ provincia suggests that southern Italy in general, as a world in decline, the problems arose outside Salento, provided with the usual cliché characteristics: (a) there Livy was using these geographical terms in the were no - or few - larger settlements, i.e. centres customary - probably correct - way.23 The harsh of culture, (b) it was sparsely populated (oligoan- measures taken by Postumius and Duronius in dreia) and (c) there was sheep rearing and exten- the quaestiones de Bacchanalibus and the possible sive forms of agriculture. It is of course likely that admixture of ‘slave revolts’ and ‘conspiracies of sheep rearing was practised in southern Apulia. herdsmen’ demonstrate that something was seri- For instance, there may have been substantial ously wrong in southeastern Italy in the 180s BC. flocks in the less fertile limestone hills (the Murge) Although slaves are unlikely to have been present in northern Salento at both ends of which there in substantial numbers in the early 2nd century BC were towns - Taranto and Canosa - widely re- and slave revolts therefore probably did not occur nowned for their excellent wool and fine woollen at the time indicated by Livy, there was obviously clothes (Roman Canusium; cf. Morel 1978). But the considerable cause for discontent in the early 2nd presence of these activities does not indicate a century BC among large groups of the population decline in agriculture. of present-day Apulia.

3.3. Uprisings/resistance 3.4. People with Messapian roots in ancient written sources Ancient historians have used two passages in Livy to underline the increasing importance of exten- People originating from the Italic civitates that sive stock raising at the expense of agriculture were bound to Rome by a variety of foedera rarely (Livy 29.29.8 and 29.41.6). These relate to the feature in ancient written sources. Because they years 185 and 184 BC, in which ‘slave’ revolts are were not Romans, they usually played only a mentioned ‘in Apulia’.20 In the first Livy passage, marginal role in Roman politics and history. When large groups of herdsmen (pastores) are said to they became Roman citizens after the Social War, have conspired, robbed and looted (Livy implies the integration of their native societies into the that some 7,000 people were executed). In the sec- Roman world had progressed to such an extent

98 that we rarely recognize them as people of Cala- Pyrrhus against the Romans. The town of Rudiae brian, Lucanian or Sabellian origin.24 Moreover, was one of the approximately ten 4th-3rd-century the substantial migrations within Italy through walled settlements in the southern, possibly ‘Sal- colonization make it highly unlikely that a par- lentinian’, part of the peninsula. On the strength ticular gens that was prominent in, for instance, of the present evidence, the settlement may have north-Apulian Luceria or south-Apulian Brundi- originated in the 8th century BC (cf. D’Andria sium, actually had its roots in that region. Equally 1991). Ennius was fluent in both Greek and Latin unlikely is the fact that people who actually had and spoke an obviously native-Italic tongue de- those origins continued to see themselves as Cala- scribed as ‘Oscan’.26 Since he came from Rudiae, brians, Lucanians or Sabellians in the context of he must in any case have spoken Messapic. His the Roman state, some 200 years after the Romans excellent knowledge of both Greek and Latin sug- conquered their district. These observations are gests that he belonged to the elite of the Rudiae especially true of the members of the former native settlement. This view is explicitly expressed by elites which, although some were probably not the poet Silius Italicus, who believed that the Roman citizens before the Social War, functioned mythical King Messapus was among Ennius’ in local, regional or supra-regional contexts which forebears.27 Ennius shared this alleged Messapian had to be compatible with the Roman world. The ‘royal’ descent with the emperor Marcus Aurelius ancient literary sources generally give us only the (see also note 30). The poet’s literary work, espe- names of people belonging to the local elites who cially what remains of his tragedies, demonstrates - viewed from a purely Roman perspective - most convincingly that he was intimately acquaint- behaved particularly badly or conspicuously well ed with Greek literature. He may have been taught in pitched battles, severe crises or other emer- by a Greek paedagogus at Rudiae or have lived for gencies relating to the Roman state, as happened some time in one of the Greek centres of culture for instance during the Hannibalic war.25 like Taranto or Corfu near his native district. Since The Messapians were no exception to this rule. his sister is said to have been the mother of the People with obviously Salento backgrounds are famous ‘Roman’ playwright and painter Pacuvius few, and they feature in our written sources only of Brindisi (e.g. Pliny, NH 35.7.19), we may in passages about the first hundred years after the assume that the Ennii from Rudiae participated in Roman conquest of Salento. The earliest case is a regional elite network. Thus the written evi- that of a certain Dasius from Brindisi. He features dence suggests that Ennius belonged to the upper- prominently in both Polybius and Livy as the most social stratum of the regional society of man who delivered the small north-Italian strong- Salento in the late 3rd and early 2nd centuries BC. hold of Clastidium (now Casteggio near Pavia), with its large stock of supplies, into the hands of Hannibal in 218 BC (Polybius 3.69; Livy 21.48). Both texts make clear that Dasius from Brindisi was the commander of the small fortress. He may have been in charge of a contingent of socii from the Brindisi region which was posted in the far north of the Po valley. We may assume that a comparable function was held in Sardinia by the best-known Messapian in a Roman-Republican context - the ‘Roman’ poet Q. Ennius, whom Horace poetically described as Calabrae Pierides. He is probably the only Messapian whose face we can look at, since portrait herms of him were erected in Rome. Hafner (1968) sug- gested that a particular series of Republican por- traits should be identified as the father of Roman literature (fig. 4). Quintus Ennius was born in the settlement of Rudiae in southern Salento in 239 BC and died in Rome in 169 BC. His parents must Fig. 4. Portrait of the ‘Roman’ poet Ennius born in have been born shortly before or after the Roman the Salento settlement of Rudiae. Vatican Museums victories over the Salento tribes in 267 and 266 (Photo Archive Archeological Institute, Amsterdam BC, and his grandfathers may have fought with Free University).

99 Much of Ennius’ literary work has been lost. so. We should bear in mind that he uttered these Only a few passages have survived because they words in Latin in what was probably a Roman were cited by more recent Roman authors. What context. Another quote, found in Aulus Gellius remains provides little information for our pre- (tria corda sese habere dicebat: ‘he said that he had sent purpose. But since Ennius was regarded as three hearts’), suggests that Ennius identified the father of Latin literature and in that capacity himself with other ethnic labels as well (see note was accorded considerable attention, we are rela- 26). Obviously, he lived in Rome as Romans do. tively well informed about his life. He must have He possessed the paideia that was part and parcel been about 20 years old at the outbreak of the of being a well-educated Greek and was able to Hannibalic war. We have no information about function in Taranto, Corcyra or Dyrrhacchium as what he did during the war. Silius Italicus (Punica Greeks were expected to do. And when he 12.393-395) suggests that he fought in the front returned to his ancestral home in Rudiae, he may line, but he is the only author to do so. In 204 BC, once again have become the native chieftain who towards the end of the war, we find 35-year old discussed matters of local importance with the Ennius in Sardinia, where he may have been part tribal farmers of his clan and other Rudini. of the Roman occupation force that controlled this Ennius, the Sallentine nobleman, obviously former Punic, and hence potentially dangerous, succeeded in a Roman context. He was the most area. As a Salento nobleman, he may have com- venerated poet of Republican times, who mixed manded a Sallentinian contingent of socii. with the greatest Roman politicians of his day, His ascendancy as a literary figure began when visited Greece and eventually became a Roman M. Porcius Cato visited Sardinia on his return from citizen. He was, however, by no means the only Africa. For reasons unknown, Cato took Ennius Messapian who was successful in post-conquest to Rome (Nepos, Cato 1.4). Suetonius’ description Salento. Men with native Salento backgrounds of Ennius’ life in a modest house on the Aventine also played a role in contacts with Greeks in the in Rome is too standardized to ring true; it is as southern Balkans. They are present in sources cliché an image as the one both Cicero and Sue- from the first half of the 2nd century BC and were tonius present of the poor poet (Suetonius, De therefore Quintus Ennius’ contemporaries. The Grammaticis 1; Cicero Cato, 14). It should also be most conspicuous man in the Roman sources is a observed that lavish, opulent dwellings did not yet certain L. Rammius of Brindisi who, according to exist in Rome in the early 2nd century BC. In fact, Livy (42.17.2), played an important role at the Ennius appears to have moved in high Roman start of the second Macedonian War. In 172/171 circles and was acquainted among others with the BC he was allegedly asked by King Perseus of Cornelii Scipiones and the Fulvii Nobiliores, who Macedonia to poison Roman commanders and were highly influential in Rome in the early 2nd ambassadors staying at his Brindisi house. He century BC. An amusing anecdote, told by Cicero naturally revealed this wicked Greek plot to his and often believed to originate from Ennius’ own Roman friends and told the Senate in Rome about Satires, gives an impression of his close friendship Perseus’ intentions. Appian (Rhomaika 9.7) tells with Scipio Nasica (Cicero, De Oratore 2.276). Like the same story but calls the prominent man from the Greek Hellenistic poets who accompanied the Brindisi ‘Erennius’. The context of Livy’s and Greek kings on their campaigns, Ennius followed Appian’s stories shows that Lucius Rammius or his powerful Roman friend, M. Fulvius Nobilior, Lucius Erennius definitely belonged to the elite of to Aetolia in 189 BC during the latter’s consulate. his hometown: Livy calls him princeps Brundisii. On this occasion he composed a kind of epos ex- He was not just an important man in local terms; tolling Nobilior’s performance during the siege of he must also have had connections in both Greece Ambracia. At the start of this campaign, his patron and Rome, receiving Greek princes and highly- came within a short distance of Ennius’ Salento placed Romans in his Brindisi house.28 The pas- hometown, since Nobilior’s forces set out for sages in Appian and Livy suggest that he knew Greece from Brindisi (Polybius 22.24.16). personally both King Perseus of Macedonia and Thanks to the activities of M. Fulvius Nobilior’s King Eumenes of Pergamum. Unfortunately, Livy son Quintus, the Messapian Ennius was granted does not inform us about his origins. Since he Roman citizenship in 184 BC. According to Cicero lived in Brindisi, he could have stemmed from (De Oratore 3.42.162), the poet said: Nos sumus Ro- both the originally native-Calabrian group as well mani, qui fuimus ante Rudini. From 184 BC onward as the Latin/Roman settlers who arrived in 244 Ennius belonged to the socio-political community BC, when the colonia latina of Brundisium was that was Rome and was obviously proud to do founded.

100 That Lucius Rammius (or Erennius) may have network. Their fathers also have ‘Daz’-names had Messapian roots is indicated by a man with (Dazos, Dazoupos, Daziskos), which some believe a very similar, perhaps even identical, gentilicium. to be a title or a name characteristic of the native An inscription in the Dodona sanctuary mentions ruling class of pre-Roman Apulia and Calabria a Gaius Rennius of Brindisi who was definitely of (Mazzei/Lippolis 1984, 226).30 Since they are likely Messapian ancestry. He appears to have been to stem from the traditional, land-owning local granted a proxeny by this important Epirote sanc- elite of the Salento peninsula, they may well have tuary at some time in the first, second or third been large landowners (cf. Yntema 1993a, 207-208). decade of the 2nd century BC. His Salento roots are Moreover, the close connections that Rammius, quite evident, since his father’s name - Dazoupos, Rennius, Pulfennius, Statorius and Dazos, son of a very characteristic ‘Messapic’ name - was added Daziskos, had with the Greek world indicate that to the inscription.29 they may also have been involved in trade with There were more Salento men of this type, Greece and other parts of the eastern Mediter- however. Gaius Pulfennius of Brindisi, son of ranean. Since they had high status in both their Dazos, also had obvious Messapian origins: he was native district and the Graeco-Roman world, they granted a proxeny by the koinon of the Epirotes, may have numbered among the great traders probably during the 170s BC (Cabanes 1976). (probably called negotiatores) who supplied the Gaius Statorius of Brindisi, proxenos of the sanc- people living or operating in Greece, including tuary of Delphi, and Dazos, son of Daziskos from the Roman armies. Products from the Salento large the Sallentinian town of Uxentum (now ), holdings owned by these men and other mem- who appears in the inventories of the sanctuary bers of their class (e.g. corn, olive oil and wine) of Apollo at Delos, are also among those with must have been among the standard products of Salento roots who feature in the Greek and Roman their Salento farmsteads.31 contexts of the early 2nd century BC. The latter Thanks to his literary talents and the patronage two Messapians appear in inscriptions dating to of the Fulvii Nobiliores and the Cornelii Scipiones, the first one or two decades of the 2nd century BC Ennius’ career may have been quite exceptional (cf. Zalesskij 1982; Silvestrini 1998). in a Roman context. However, it is revealing that These Messapians who feature in the written other more or less contemporaries with obvious sources were not simple farmers. The poet Ennius Messapian roots appear in the ancient written is explicitly said to have been of noble or even sources on the early 2nd century BC. Some are royal descent. That such high-ranking Messapians such close contemporaries that they must have were not created in later mythologizations but been acquainted. Lucius Rammius (or Erennius), actually existed in pre-Roman and early Roman who informed the Roman Senate about King Per- Salento is confirmed by Strabo (Geography 6.3.6), seus of Macedonia’s treacherous conduct, must who mentions the presence of a basileion in the have known Gaius Pulfennius. Both had their Messapian settlement of Oria on the Salento isth- home base in Brindisi, Rammius features in Livy mus. Since Ennius’ sister married someone from in a passage concerning the year 172, while Pul- Brindisi, her son, the ‘Roman’ tragedian and paint- fennius was granted a proxeny by the Epirotes er Pacuvius, must have belonged to the same during the 170s. The same holds for Gaius Stato- social group in the Salento district. But in the 2nd rius, Gaius Rennius and Dazos from Ugento, who century BC, the Salento elite probably did not seem to have been men of considerable impor- consist solely of people originating from the for- tance during the 190s or 180s. In the century fol- mer Messapian ruling class; it may also have lowing the Roman conquest, there clearly appear included leading families of Brindisi of originally to have been a fair number of such men in the Latial or central-Italic descent. scant written sources concerning that period. The texts that mention Lucius Rammius or They all belong to the two generations who were Erennius, Gaius Rennius, Gaius Pulfennius, Gaius born within about 30 to 50 years after the Roman Statorius (all from Brindisi) and Dazos from U- conquest. And they are probably just the tip of gento suggest that they were important people. the iceberg: they were among the limited number Dazos - admittedly on flimsy grounds - is believed of people of wealth and high rank who featured to have been an oil merchant at Delos (Zalesskij prominently in ‘international’ contexts. There must 1982). Since three of the others appear in proxeny have been successful and prosperous Messapians inscriptions, they must have been very wealthy in the later 3rd and early 2nd centuries BC who do men who occupied prominent positions in their not appear in our sources simply because their hometown and had a substantial international inscriptions were lost or because they were large

101 Salento landowners who did not attain the emi- the noble Rhomaioi did.32 Here they often used nent ‘international’ status of their ‘compatriots’ names - like Gaios Statorios Brentesinos - that dis- discussed above. play suspiciously close similarities to the tria nomi- Those appearing in the written sources were all na system of the Roman elite. In Greek contexts, members of the Messapian elite. Official inscrip- the poet Ennius may have presented himself as tions and Roman historians rarely mention com- Quintus Ennius Rudinus. moners. It is striking, however, that we would These data provide us with important clues con- never have recognized them as being of Messa- cerning the way the great men of Salento wished pian descent if their fathers’ typically Messapian to be viewed in the international context of the names had not been added to the ‘Roman/Latin’ early 2nd century BC. The text closest to the pri- name they chose in the early 2nd-century context vate sphere of the early 2nd century BC furnishes of the Roman state. This ‘Roman/Latin’ name was us with what may well have been traditional Mes- not only used in exclusively Roman contexts, but sapian practice: the votive offering to Apollo at even features in official inscriptions in the Greek Delos is presented by Dazos, son of Daziskos. No world. Another salient characteristic of these Mes- Latin-style gentilicium or prenomen is given. Iden- sapians is that they did not call themselves ‘Mes- tical onomastic formulae, in which the name of sapians’, ‘Calabrians’ or ‘Sallentinians’. Rennius, the father is added to the name of the dedicator, Pulfennius and Statorius identified themselves as are known from many Messapian inscriptions in Brentesinoi, Ennius called himself a Rudinus and sanctuaries and from tombstones, and have close Dazos professed to being an Azentinos. It appears parallels in the Greek world (e.g. De Simone 1964; that they identified themselves with the town that Pagliara 1987). Thus the use of Latin-sounding they came from, and not with the tribal group to gens names such as Ennius, Rammius/Erennius which they might have belonged in the context of and Rennius and typical Roman prenomina such the originally Messapian societies. as Gaius, Lucius and Quintus found in the sur- It is of course hardly surprising that the men viving written sources may be viewed as a signif- from Brindisi called themselves Brundisini or Bren- icant break with the tribal Calabrian or Sallen- tesinoi. The town of Brindisi was rapidly becom- tinian past. It is perhaps too much of a coincidence ing one of the major ports of Roman Italy - sec- that Ennius assumed the Roman prenomen Quin- ond only to Puteoli - and must have been widely tus and owed his Roman citizenship to Quintus renowned, especially in the eastern Mediter- Fulvius Nobilior, son of Ennius’ patronus Marcus. ranean (Marasco 1988). It is therefore plausible to While the origins of the praenomina are per- assume that Greeks were better acquainted with fectly clear, we may ask where the gentilicium of the town’s name than with terms like Messapos, these originally Messapian gentlemen came from. Kalabros, or Sallentinos. However, Q. Ennius and With regard to Ennius, Erennius/Rammius, Ren- Dazos, son of Daziskos, presented themselves as nius, and Pulfennius, these names - all ending men from Rudiae and Uxentum respectively. These with -ennius - sound like a regional conspiracy two Messapian towns must have been rather and are not really transparent. Marcus Aurelius’ obscure places in the Graeco-Roman contexts of mythical forefather Malennius, allegedly founder the 2nd century BC. This can be explained in two of the Messapian settlement of Lupiae (now ), ways. Either the tribal allegiance and structure of obviously belonged to the same tribe. Gaius Sta- the Messapian groups in Salento had already torius of Brindisi, however, may have based the been strongly eroded by the early 2nd century BC, name of his gens on the Messapian name Thaoto- or the regional elite of Messapian descent no long- ridas, which derives from the regional god Thaotor. er wished to associate itself with the old Messa- Much later, in the 2nd century AD, a gens Tutoria pian political régime. Both explanations imply, whose name derived from Tutor, the heavily Lat- however, that the tribal organization of the Mes- inized form of the same god Thaotor, probably sapian groups no longer played a significant role lived in Lupiae and owned a number of very sub- for people of Salento descent who operated in the stantial Roman villas in the area between Brindisi continuously expanding Hellenistic-Roman world. and Lecce (Marangio 1975; Pagliara 1980). The They suggest that these men wished to be associ- name of the originally Messapian Tutorii is also ated with new socio-political realities. Although it said to live on in that of the village of Tuturano, is uncertain whether they all had Roman citizen- some 10 km south of Brindisi, which is believed ship (most may have had Latin citizen rights), to derive from a fundus Tutorianus (Laporta 1988). these men - although basically Italioi - presented No-one of recognizably Messapian descent ap- themselves in Greek contexts in the same way as pears in written sources for more recent periods.

102 One reason may simply be that we are unable to elites, men like Rammius, Pulfennius and Stato- recognize them: the Italian elites of late Republican rius were probably both large landowners and times operated at a supra-regional level and the very important negotiatores. They may also have local population was gradually becoming a mix been diplomats who mediated between Rome of Latin colonists, immigrant Greeks and Italians, and the Greek states, since they were able to func- people of local origin and imported slaves. An- tion well in two different cultural systems - both other reason may be that the former Greek towns Greek and Roman society. and their originally indigenous neighbours in the Such eminent men from Salento may also have coastal areas of southern Italy were not involved existed in the later 2nd century BC. Ennius’ nephew - and fundamentally not interested - in the Social from Brindisi, the poet and painter Pacuvius, is War and its outcome. The inhabitants of Heraclea the only person with partial or complete Messa- in - since they were already citizens of pian ancestry who by pure chance appears in the Heraclea - even discussed whether they would sources discussed in this section. However, other accept the grant of Roman citizenship in 89 BC outstanding Salento men cannot be traced in the (cf. Ciaceri 1924, 32). As a result, the southern written record since their full integration into the elites were poorly represented among the homines Graeco-Roman world renders them unrecogniz- novi who entered the Senate in the later 2nd and able. Ennius, Rammius and Pacuvius feature in 1st centuries BC (Torelli 1971). As for Salento, only the Roman literary sources. The unique fact that two men of senatorial rank (probably brothers) are a handful of Greek inscriptions from the first half recorded who may have stemmed from Brindisi. of the 2nd century BC add the place of origin and However, the grounds for assigning them to this father’s characteristically native name to the Lat- town are rather flimsy (see Wiseman 1971, 230).33 inized names of some important ‘Italioi’ enables The southern elites of the late 2nd and 1st cen- us to identify a few highly prominent Messapians turies BC seem to have been only marginally whose floruit was between 200 and 160 BC (i.e. in interested in Roman politics. Although they tech- the period directly after the Hannibalic War). nically became Romans in 89 BC and could exer- cise the political aspects of their citizenship rights 4. ARCHAEOLOGICAL SOURCES in the 500-700 km distant centre of power, these Greeks, Messapians and other natives from south- In the introduction to section 3 above, I have ern Italy’s coastal strip were inclined to mind sketched the picture of post-Roman-conquest their own business. They also continued to view southern Italy that was presented until the 1990s the Greek world of the eastern Mediterranean as on the basis of interpretations of historical sources. their main point of reference. It should also be The cliché of post-conquest and post-Hannibalic stressed that the southerners certainly did not decline and impoverishment did not of course consider themselves barbarians: they were cul- encourage archaeologists to invest in research tured men who spoke both Latin and Greek. Thus projects targeting this particular period of south- they might not have had the same keen interest in ern Italy’s history. The general picture, it was truly becoming ‘new men’ that some of the lead- believed, was perfectly clear. It was also a highly ers of Emma Dench’s central-Italic mountain peo- plausible construct of southern Italy in Roman ple had (cf. Dench 1995). times. At first glance it seemed highly likely that the widespread destruction resulting from the The above passages show that, generally speak- Roman conquest and the enormous ravages sup- ing, the Messapian elite prospered in the century posedly caused by Hannibal’s troops in turn pro- following the Roman conquest of their districts. duced the rather depressing landscape described All - or nearly all - of the Salento people we - above. This general impression was reinforced by counter in the ancient written sources emerge as the epigraphical evidence of Roman times. The men with solid ‘Messapian’ backgrounds. They many funerary inscriptions that had fortuitously were also men of very high status who befriended come to light in considerable numbers in both leading Roman senators and/or Greek kings urban and rural contexts in southern Italy revealed (Rammius/Erennius, Ennius) and who were held for instance that the population of Salento consist- in high esteem by famous Greek sanctuaries or ed mainly of freedmen and slaves, many of whom Greek states (Statorius, Pulfennius, Rennius). Like had Greek names (e.g. Susini 1962, 60). This sug- Ennius, they may all have spoken both Greek and gested that the composition of the population of Latin in addition to their native Messapic lan- Salento had completely changed. It should be guage. Since they stemmed from the indigenous stressed, however, that the vast majority of these

103 inscriptions appear to date to the 2nd and 3rd cen- present day. This observation is true of Tarentum turies AD and that inscriptions dating between (Taranto), Brundisium (Brindisi), Lupiae (Lecce) the 2nd century BC and the 1st century AD are very and Canusium (Canosa). The remains of these rare in the area under discussion. If these epi- Roman towns have been buried, destroyed and graphical data can indeed be interpreted in this reused in more recent phases of the rich histories way, a decisive shift in the composition of the of these towns. However splendid their Roman population could in principle have occurred at attire might have been, they are almost invisible any time between the Roman conquest of 267/266 in the archaeological record. In the following sec- BC and the mid-2nd century AD. However, the tions I will briefly discuss the archaeological data close reading of the written sources presented in for the first two centuries after the Roman con- the preceding section reveals that the commonly quest of Salento so that they can be compared accepted picture of early Roman southern Italy is with evidence from the written sources presented probably a mixture of Roman clichés for con- in section three. For the preceding presentations quered territories (cf. Alcock 1993) and southern of data on Roman Salento, I refer to earlier pub- realities and clichés of more recent date (e.g. the lications on this subject (D’Andria 1979; Yntema Roman Imperial Age and the modern Mezzogior- 1995). no). A first sign that the Roman period of southern Italy was attracting the attention of archaeologists 4.1. Rural settlement and economy was the Belgian excavations at north-Apulian (ancient Herdoniae) in the 1960s and 1970s Information on rural settlement in Salento comes (e.g. Mertens 1979 and 1988). Here important mainly from field surveys carried out in the north- urban features such as two temples, a macellum ern part of Salento by teams from Siena Univer- and a basilica of an amusingly provincial town of sity and the Amsterdam Free University (fig. 6). the Roman period came to light (fig. 5). A second While the Siena team worked in the immediate encouragement came from the Magna Grecia in età vicinity of Brindisi and revealed mainly traces of romana congress held in Taranto in 1975.34 A first Roman landscapes (e.g. 2001), teams from Amster- synthesis on the Roman period in Apulia was dam inventoried traces of past human occupation published by Francesco D’Andria in 1979. In that in large ‘rural’ areas on the Brindisi plain at Oria same year the Institute of Archaeology of the (66 km2) and Valesio (20 km2) dating between the Amsterdam Free University decided to carry out Neolithic period and late medieval times (e.g. a series of field surveys in Salento in order to Boersma et al. 1991; Yntema 1993a; Burgers 1998). study changes in rural settlement patterns. One of The Amsterdam Free University is at present the aims was to test the almost generally accept- working in the more hilly area to the north ed view of Roman southern Italy as a particularly of Brindisi and in the hills northwest of Taranto.36 depressing, ancient Mediterranean version of the It was one of the basic aims of both the Oria American Midwest. The fieldwork started in 1981 Field Survey (1981-1983) and the Valesio Field in the area surrounding the important pre-Roman Survey (1989-1991) to trace developments in the tribal centre but relatively unimportant Roman rural economy. The present-day town of Oria is municipium of Oria (Uria Calabra, ORRA on its 2nd- situated halfway between Taranto and Brindisi on century BC coins), situated on the Salento isth- the Salento isthmus. The various concentrations mus (for the final report, see Yntema 1993a). of artefacts were mapped in an area of ca 66 km2 Since the early 1980s considerably more atten- around the ancient settlement of Oria. Though cer- tion has been paid to the Roman period in Apu- tainly not every site was detected in this proce- lia.35 Excavations and other types of research con- dure, there is reason to assume that the field-sur- cerning the Roman period have been carried out vey results do in fact provide a good impression at Taranto, Lecce, Canosa, , Valesio, Vaste of general trends in rural habitation in the Hel- and the area in and around Brindisi. The most im- lenistic and Roman periods (final report: Yntema pressive studies concern a large Roman temple at 1993a). They demonstrate that a fairly intensive Canosa (Cassano 1992; Dally 2000) and the mag- rural habitation began in the native districts of nificent late Roman complex of San Giusto near Salento between 330 and 280 BC. There must north-Apulian (Volpe 1998). But all in all have been dozens of small farmsteads, each with the picture of the Roman period in Apulia is rather one or two adjoining necropoleis, throughout the bleak. This can largely be ascribed to the fact that Oria countryside during the 3rd century BC. Some the most important Apulian centres in the Roman 30 of these were discovered during the field survey period have all continued as settlements to the (Yntema 1993a, 181-186). Judging by the reports on

104 larger farms, we may perhaps assume that the farms were inhabited by a group of people of approximately the same size as the casa colonica described by Cato - i.e. some 25 to 30 people.37 A very similar development can be observed in the Valesio countryside, where the number of farm sites dropped from 49 around the mid-3rd century BC to 27 in the later 2nd century BC (Burgers 2001). Thus both the Oria and Valesio field surveys demonstrate that while the number of rural sites declined during the first half of the 2nd century BC, the sites which continued to be inhabited in the later 2nd century BC (in both cases well over 50% of the sites discovered in the research area) were much larger than the 3rd-century farmsteads. The new farms that were constructed on top of (or using parts of) the 3rd-century tribal farmsteads Fig. 5. Ordona were two to four times larger than their prede- (north-Apulia): cessors (Boersma et al. 1991; Yntema 1993a). On town centre in the Roman the strength of this evidence, it is plausible that Imperial period the rural population in northern Salento (i.e. the (adapted from Mertens 1988). part of the peninsula closest to the Taranto scenes of war in 213-209 BC) actually increased in the aftermath of the Hannibalic war. Agriculture therefore seems to have intensified considerably in Salento in the post-Hannibalic era. The new and much larger farms of the mid- to later 2nd century BC may be regarded as a sign of consid- erable rationalization and intensification of agri- cultural production (cf. Yntema 1993a, 195-213). The large farmsteads are by no means the only evidence for an intensification of agriculture in Salento. Transport amphorae - of the so-called Graeco-Italic type (fig. 8a) - were probably made Fig. 6. Salento isthmus: field survey areas in Salento from about the mid-3rd century BC.38 (Amsterdam Free University, 1981-1995). They continued to be made until about the middle to the third quarter of the 2nd century BC. Their the burials, each of these farms is likely to have appearance may be tentatively linked with the been inhabited by a family unit consisting of ap- surplus production of the relatively small farm- proximately 10 to 15 people. The more sophisticat- steads of contemporary date. Before the mid-2nd ed methods used during the field survey in the century BC, however, a new and larger-scale pro- rural area between Valesio and the duction of amphorae began. These so-called in the early 1990s produced an even larger den- ‘Apulian’ or ‘Brindisi’ amphorae (fig. 8b) were sity of rural settlements for the 3rd and early 2nd produced in the large kilns of the sites of Apani, centuries BC: here 49 farm sites were discovered Giancola, Marmorelle and La Rosa, each of them in an area of about 20 km2 (fig. 7). only a few kilometres from Brindisi (e.g. Mana- However, by about the mid-2nd century BC (i.e. corda 1988 and 1990 and Palazzo 1988 and 1989; in the period after the Hannibalic war), 12 of the Désy 1989). There are indications that similar 30 Oria farmsteads had been abandoned (Yntema workshops, although probably of more modest 1993a, 198-202). Intensive auguring (the standard dimensions, existed in southern Salento.39 These procedure during the Oria survey) has demon- workshops were strictly organized, each with at strated that the remaining 18 farms more than least two or three slaves. The workshop organi- doubled in size. Although no burial plots have zation can be deduced from the stamps on the been found to provide information on the size amphorae (cf. Désy 1989; Palazzo/Silvestrini and composition of the population of these much 2001).

105 Fig. 7. Valesio (Brindisi), farmsteads of the late 4th and 3rd centuries BC discovered during field surveys (Amsterdam Free University, 1989-1992): 1 not investigated 2 farmstead 3 probable farmstead 4 concentration of tiles, amphorae and storage jars 5 pre-Roman sanctuary

It goes without saying that the introduction of large amphora production units near the key port of Brindisi was closely related to agricultural pro- duction in the hinterland: the amphorae probably contained products from the north-Salento Brin- disi plain. Since the start of large-scale production of Apulian trade amphorae (ca 170/160 BC) was contemporaneous with the creation of much larger farmsteads (between about 170 and 150 BC), we may suspect a close relationship: these ‘Catonian’ casae in Salento probably produced the contents of the Apulian amphorae. The distribu- tion of this type of amphora over the Mediter- ranean indicates that the agrarian products of the Brindisi area were widely consumed in consider- able quantities (Cipriano/Carre 1989). Products from Salento were marketed in large parts of the Mediterranean with remarkable success. The dis- trict must therefore have generated very substan- tial surpluses indeed. I have stressed above the probable close link between the appearance of larger farmsteads in Fig. 8. Amphorae produced in the Salento district: rural Salento and the rise of large-scale amphora a. Graeco-Italic amphora from Valesio, late 3rd/early production centres at Apani, Giancola, Marmo- 2nd century BC; relle and La Rosa (all near Brindisi), San Cataldo- b. Brindisi amphora from elite burial at Ascoli Ramanno (the Lecce area) and Felline (near U- Satriano (north-Apulia), mid 2nd century BC. gento). This assumption is strengthened by yet another factor. The urban surveys in the Brindisi (Burgers 1998). This could suggest that the former area mentioned above indicate that those settle- Messapian settlements on these Roman highways ments situated on the well-known, major Roman initially retained at least one of their earlier func- arteries of Salento (the via Appia and via Calabra/ tions: they may have been collecting stations for Minucia) declined less rapidly than those on roads agrarian products from the farms and fields sur- that were peripheral to the Roman road system rounding the drastically reduced settlement itself

106 Fig. 9. Artist’s impression of the settlement of Muro Tenente (Brindisi district) in the 3rd century BC (based on excavations, urban survey and auguring).

and from the farms and fields once belonging to itants. In about 300 BC, shortly before the Roman settlements now virtually abandoned because they intervention in Salento, there were some 25 of were not on a major Roman-era artery. If this is these (D’Andria 1991; Burgers 1998). Since they correct, the via Appia and the via Minucia/Calabra had large and impressive town walls,41 each may were not just Roman highways for interregional have enjoyed a limited measure of political inde- traffic. They played an equally important role in pendence within the regional tribal systems and the regional economy, as the main arteries along must have furnished its inhabitants with a decid- which the very substantial surplus production of edly local identity. However, the largest (e.g. Oria the Brindisi district was transported to the region- and Ugento) may have ranked above most of the al centre of Brindisi for consumption, packing or other walled settlements. They were possibly tribal shipping. centres of regional importance - not just the largest walled settlements of ca 300 BC, but already sub- 4.2. Urban settlement and religion stantial fortified settlements in a relatively distant past. While most of the Messapian walled settle- The profound changes in the rural sector discuss- ments were initially only very small strongholds ed in the preceding section went hand in hand or were not walled until the late 4th century BC, with substantial changes in the settlement hierar- these larger centres in Salento were walled as early chy. Burgers (1998) has highlighted these ‘urban’ as the late 6th or early 5th century BC.42 At the time developments for the northern part of Salento, but of the Roman conquest, the settlement hierarchy we can surmise similar changes in the southern in the Calabro-Sallentine districts probably con- part. Information on this subject can be derived sisted of a few major tribal centres, a series of from a series of ‘urban’ surveys carried out at the about 25 somewhat minor walled towns that har- north-Salento sites of Muro Tenente (52 ha), Muro boured subtribal units, and small rural settlements Maurizio (32 ha), San Pancrazio (16-17 ha), Valesio (hamlets and isolated farmsteads) that were situ- (80 ha), none of which was built over in post- ated within the territories of the walled towns. antique times (cf. Yntema 1993b; Burgers 1998). In section 4.1 I have explained what happened We can discern four different types of settlements to the rural settlements of Salento after the Roman in the Calabro-Sallentine area before the Roman conquest. But the walled centres were also subject conquest. As we have seen above, the smallest to substantial change. Initially very little happen- settlement type was the isolated farmstead. There ed. The distribution of ceramics dating to the sec- must have been hundreds of them in Salento, ond and third quarters of the 3rd century over the sometimes clustered in hamlets of three to five walled sites does not differ from that of the late habitation units.40 In addition to these two types 4th/early 3rd century BC. A peripheral quarter of of rural settlements, however, there was an ample the Muro Tenente site, built in the early decades series of walled, more or less urban settlements, of the 3rd century BC, was not abandoned before each presumably with some 2,000 to 4,000 inhab- 230/220 BC (fig. 9). But before the mid-2nd century

107 BC (i.e. within a hundred years of the Roman con- quest), many of the walled towns had become mere villages with a few hundred inhabitants, and most of these in turn became hamlets of a few dozen people in the 1st century BC (Yntema 1993a; Burgers 1998). According to the results of the four urban surveys in the Brindisi area, the population loss sustained by the walled settlements started in the last quarter of the 3rd century BC and became truly dramatic during the first half of the 2nd cen- Fig. 10. Coins minted at Oria in the 2nd century BC. tury BC.43 Not all the walled settlements of Salento were affected in the same way. As I have said above, those settlements situated on the major Roman highways (the via Appia and via Minucia/Calabra) declined less rapidly than those that were not (Burgers 1998). Moreover, the former tribal centre of Oria on the Salento isthmus managed to retain some of its former grandeur. For instance, the town minted its own coins in the 2nd century BC (fig. 10) and survived as an - admittedly modest - municipium with its own ager Uritanus into the Roman Imperial period. The ports of Gnathia and Otranto, both on the via Minucia/Calabra, retained more or less their former status and survived as towns of secondary importance. However, two Salento towns - Brindisi (Brun- disium) and Lecce (Lupiae) - probably experienced completely different demographic changes. Al- though certainly walled towns in the pre-Roman period, there are no indications that they were major tribal centres. Since both Brindisi and Lecce have survived until the present day, the archaeo- Fig. 11. Canosa (north Apulia). Capital of the Italic logical evidence is necessarily patchy. The ancient temple underneath the church of San Leucio, the 2nd written sources discussed in section 3.1 make it century BC (photo Archive Archeological Institute, quite clear that Brindisi was by far the most im- Amsterdam Free University). portant town during the 2nd and 1st century BC in the area that must have been Calabrian and that Lecce was already a flourishing town in Au- Sallentinian territory in the early 3rd century BC. gustan times. The lavishly decorated theatre with The town’s prosperity in early Roman times is an obviously Augustan programme of sculptures reflected in elaborately decorated capitals belong- (fig. 12) must have been the pride of the town ing to a substantial, 2nd-century BC temple. These (D’Andria 1999).46 This means that Lecce is likely have close parallels in the former native settle- to have been a town of considerable substance in ment of Canosa (Canusium), which was emerging the pre-Augustan period, perhaps second only to as by far the most important Roman town of Brindisi in Salento’s settlement hierarchy from the northern Apulia in that same period (fig. 11).44 later 2nd to 1st century BC onward.47 No such evidence exists for Lecce (Lupiae) in the southern part of Salento. The rise of Lupiae as Many pre-Roman sanctuaries of the Salento district a major regional centre was commonly believed were situated outside the physical boundaries of to have occurred in the 2nd century AD under em- the walled settlements. Located in caves and near peror Marcus Aurelius. Since he believed that King springs, they were probably controlled by local Malennius, the mythical Messapian founder of elites. Judging by the finds, they particularly Lecce, was one of his ancestors,45 it is understood thrived during the late 4th and early to mid-3rd that he upgraded the settlement founded by his century BC. The later 3rd century shows a sub- forebears. Recent investigations, however, suggest stantial drop in the number of finds, which de-

108 Fig. 12. Lecce (south Salento): reconstruction of the scaenae frons of the theatre with its Augustan decoration program containing (upper row) Giustiniani, Alexander the Great, Gabii Artemis and Ares, and (middle row) Doryphorus, Primaporta , Heracles of the Copenhagen-Dresden type and Amazon of the Lansdowne- Sciarra type (adapted from D’Andria 1999).

Roman sanctuaries are almost invariably found in the centre of the settlements. The figured capitals of Brindisi have been discussed above. There is no information on Roman temples in the still existing towns of Lecce and Otranto. At Oria, inscriptions referring to sacred areas or sacred buildings of the Roman period were all found in the very centre of the present-day town (Pagliara 1980). More- over, the modest town of Gnathia, whose Roman part has been almost completely excavated, had two substantial buildings that functioned as tem- ples in early Roman times. These were all situated inside the fortifications and can be characterized as urban cult places. The largest and probably most representative of these sacred buildings probably dates to the 2nd century BC (fig. 13). It was situated on the so-called acropolis, a flat- topped hill within the fortified area towering high above the houses and streets of the Roman town. Changes in the religious domain were rather dramatic. While the Messapian towns had extra- urban sanctuaries with no - or just small - build- ings, the towns of Roman Salento had decidedly urban sanctuaries containing substantial sacred buildings. These occupied conspicuous, often pro- Fig. 13. Gnathia (north Salento): remains of the minent positions within the town. Some must have ‘Acropolis’ temple, 2nd century BC (adapted from Il been embellished in a manner that befitted the Parco Archeologico di Egnazia, Valenzano, 2000). town’s status. While fortifications and elite dwel- lings were among the status symbols of the Mes- clined even further during the first half of the 2nd sapian towns, much time and energy was devoted century. Only very few of these sanctuaries sur- to constructing representative sacred buildings vived into the Roman Imperial period.48 Most and possibly other types of architectural monu- were abandoned during the first half to mid-2nd ments in the hearts of the few Salento towns of century BC (Yntema 1995a). However, traces of early Roman times.

109 4.3. Social relations We have observed that there was a substantial quantity of graves at former Greek Taranto in the Burials and burial plots are sometimes eminently late 3rd and 2nd centuries BC. In addition, there are suited to the study of evolving social relations. It graves of the same period from indigenous set- might therefore be rewarding to analyze the bur- tlements of Salento. Most of these stem from the ial plots of Salento in order to gain insight into this relatively large and prosperous 2nd-century settle- matter. While there is a consistent body of funer- ments of Lecce and Brindisi. While the evidence ary evidence from the former Greek town of Ta- for Lecce was often severely damaged and exca- ranto in Salento for both the 3rd and the 2nd century vated many years ago (see Giardino 1994; Giardi- BC (cf. Lippolis 1994; Gräpler 1997), the data no et al. 2000), a well-preserved burial ground has regarding the remaining, previously indigenous, recently been discovered in Brindisi (Andreassi/ areas of Salento are problematical. If we take at Cocchiaro 1988), dating to between the mid-3rd face value the evidence published thus far, there century and 1st century BC (fig. 14). The tombs have appear to be almost no tombs dating to the period a fairly simple ‘architecture’ and contain only a following the Roman conquest. This is due to a few objects each, suggesting that those buried in series of misconceptions that have led to an incor- the necropolis were either unable or unwilling to rect dating of some classes of Tarantine/Apulian indulge in ostentatious display in the funerary ceramics.49 If this bias is removed, it emerges that domain. Compared to earlier Messapian graves there were numerous tombs in native Salento in and contemporary elite graves, these people prob- the first and second quarters of the 3rd century, ably did not number among the happy few of the that a much smaller number of tombs can be town. dated to the third and especially the final quarter A few tombs dating to the later 3rd and earlier of that century, and that the archaeological record 2nd century BC have also been reported from na- is very meagre indeed for graves from the earlier tive Salento centres that were less successful in the 2nd century BC.50 2nd century BC. Foremost among these are four Of course, in very general terms this funerary large and very special tombs originating from the evidence reflects the settlement history of the vast sites of Oria (2) and (2) in the northern majority of Messapian towns. We have seen that isthmus area (fig. 1). Each of these tombs con- these flourished during the first half of the 3rd sisted of extremely large limestone slabs and con- century, lost some of their inhabitants in the later tained an inhumation burial.51 The scant infor- 3rd century, and were almost completely abandon- mation on the two Oria tombs indicates that they ed during the first half of the 2nd century BC. The must have belonged to the same burial plot. Their depopulation of Messapian towns, which started contents appear to be closely comparable (fig. 15 in the last third of the 3rd century BC and reached and fig. 16): both can be dated to about 150/140 BC dramatic proportions between 180 and 150 BC, is and contain drinking vessels (kantharos or mastos), at least partially responsible for this pattern. It a wine jug (lagynos), a lamp, a series of platters should, however, be observed that in general less and one or two ungentaria. In addition, the two care was bestowed on burials dating to the later Oria tombs appear to have contained a terracotta 3rd century BC: there are fewer pots in the graves, vessel of the Gnathia class dating to the 3rd century the pots are of fairly modest quality, and tile-cov- (i.e. between 70 and 100 years old) and one or ered graves increased at the expense of the tradi- more Greek wine amphorae.52 The two Mesagne tional ‘Messapian’ tombs consisting of impressive tombs originate from two different burial plots stone slabs. Thus the likelihood of these much (some 700 m apart) and date to the first half of the ‘poorer’ burials of the later 3rd century BC turning 2nd century BC. One was found in 1988 and ap- up in the archaeological record is considerably less peared to contain a Rhodian and a Cnidian (or than for the conspicuous cist burials of the earlier possibly west-Cretan) wine amphora, a wine jug 3rd century BC. But whatever biases Salento’s (mykos), a lamp, a strigil, unguentaria and a series archaeological record may contain with regard to of late 4th- and 3rd-century vessels (Apulian red- the funerary domain of local societies, it cannot figured, Gnathia wares). A second equally large be denied that there are indeed burials dating to Mesagne grave displaying comparable features the late 3rd and first half of the 2nd century BC. It was discovered in 1999 and appeared to have is therefore useful to discuss the provenance and been robbed. The few fragmentary objects that character of some of them in order to trace some survived are still unpublished. aspects of the Salento social relationships in the These four tombs from Oria and Mesagne in Hannibalic and post-Hannibalic period. the first half to mid-2nd century BC strike us in

110 Fig. 14. Brindisi. Northeastern part of the extramural burial ground of the Via dei Cappuccini (later 3rd to early 1st century BC; adapted from Andreassi/Cocchiaro 1988).

Fig. 15. Oria (Brindisi district). Local elite burial of Fig. 16. Oria (Brindisi district). Local elite burial of approximately the middle of the 2nd century BC with the third quarter of the 2nd century BC with Cretan Rhodian wine amphora, Micro-Asiatic lagynos and wine amphorae, Brindisi oil amphora and 3rd-century 3rd-century BC Gnathia turreen (photo courtesy BC Gnathia kantharos (photo courtesy Graziella Graziella Maruggi). Maruggi). several respects. First and foremost, their archi- tombs. The link with the native past is also tecture (large cist burials) displays strong links stressed by the presence of ceramics that were at with the past of the Messapian elite. Such large least fifty years old - and in the case of the 1988 elite tombs consisting of large slabs were first con- tomb from Mesagne sometimes more than a hun- structed in the late 6th century and continued to dred years old - when deposited in the grave. be made until well into the 3rd century BC.53 The Secondly, they exhibit ties with Greek areas in the Oria and Mesagne tombs of the 2nd century BC, eastern Mediterranean by means of Greek wine however, are much larger than the earlier elite amphorae and in one case (Oria tomb 1) by a

111 Fig. 17. Giancola (Brindisi). Amphora stamps from the kiln debris (later 2nd-early 1st century BC; based on Palazzo/Silverstrini 2001).

Micro-Asiatic lagynos as well. Thus the tombs discussion. The ‘low visibility’ of most of the buri- suggest that both the Greek world of the eastern als may help account for this: less care may have Mediterranean and the Messapian elite past were been taken over burials for a substantial portion in some way important to the persons buried in of the population. But it is also distinctly possible these graves. Moreover, these four elite burials that an ever-decreasing number of people were from northern Salento are among the most recent given a formal, archaeologically traceable burial tombs displaying traditional Messapian burial from about the middle to third quarter of the 3rd customs. century. Both explanations, however, point to It should be noted that the contents of these social change - perhaps greater social distinctions four Salento tombs have close parallels in both within the Messapian societies in the period fol- northern Apulia and the Marche area (Ancona), lowing the Roman conquest of Salento. While the where elite tombs from approximately the same former tribal elite continued to prosper, express- period have been found, displaying very similar ing their wealth and status in sumptuous graves features. The north-Apulian specimens originate like those found at Oria and Mesagne, the former mainly from the sites of Arpi and Canosa (Roman tribal farmers were becoming increasingly impov- Canusium) and the closely allied, if not Canosa- erished during the final years of the 3rd and the dominated, settlements of (ancient first half of the 2nd century BC. We may well ask Ausculum) and Salapia.54 Although these do not whether Roman taxation and levies (cf. Brunt refer in an obvious way to a past tribal world 1971) were external factors that acted as catalysts (tribal society was still very much alive in Canosa to change. in the period under discussion),55 they retain some highly traditional features and appear to More information on social relations comes from contain amphorae and other objects (Alexandri- the amphorae produced in the coastal area around nian, or more probably Rhodian glass, fayence, Brindisi. As these can be dated to the second half wine amphorae) that stress the links between the of the 2nd century BC and much of the 1st century Arpi and Canosa elites and the Greek eastern Med- BC, they largely fall outside the scope of this paper. iterranean. The elite tombs from Ancona (Marche They bear stamps with the names of people in region) display the same ‘eastern’ connections some way involved in producing these vessels (Mercando 1976; Colivicchi 2002). As manifested and stem from a period of Salento history that is in the funerary domain, the elites of the south- decidedly more recent than the graves discussed Apulian Brindisi district appear to have been above. They may therefore reflect elements of a equally interested in elements that referred to an more recent social constellation. eastern Mediterranean present, but they also Excavations carried out by a team under the di- clung to a noble indigenous past. rection of Daniele Manacorda (1988, 1990) and It is perhaps significant that we encounter the Paola Palazzo (1988, 1989) and analyses conduct- rather modest graves of the later 3rd and early 2nd ed by Philippe Désy (1989), Paola Palazzo and centuries almost exclusively in settlements that Marina Silverstrini (2000) have both demonstrated were later to become the major towns of the that the people who feature on the stamps were Roman period.56 They are absent from the remain- both workshop owners and slaves or artisans. It ing walled centres of Salento which, according to is striking that recognizably Messapian names are the results of the urban surveys, were neverthe- almost completely absent from this substantial set less anything but ghost towns in the period under of names. While patrons/owners such as the An-

112 ninii, Vehilii, and Viselii are almost unanimously more slaves. These workshop owners belonged to believed to be of central-Italic origin (albeit on a prosperous, though not necessarily extremely extremely flimsy grounds; in fact the Anninii and wealthy, stratum of Salento society. Vehilii probably had Messapian roots),57 the arti- sans almost invariably have Greek names, some 4.4. External relationships and newcomers of which evidently refer to slaves (e.g. Kerdos, Noumenios, Pylades, Philemon). Some of these The stamps on the Brindisi amphorae discussed were written in both Greek and Latin letters. Only above suggest the presence of a substantial num- two people have obviously Messapian names ber of probably immigrant slaves. In addition to (Dasus/Dasius and Stabuas). Because such old- the epigraphical evidence, imported and exported fashioned local names would hardly appeal to objects can be adduced in order to provide evi- immigrants, these two men may have had Mes- dence of external relationships. These can of course sapian roots (cf. Désy 1989). The information from be both direct and indirect. The evidence for the amphora stamps of the late 2nd and earlier 1st Salento consists mainly of pottery. While the century BC suggests the presence of quite a num- ceramics of 3rd-century Salento were made almost ber of slaves, perhaps some free artisans of lowly exclusively in local or regional workshops, the status and a group of owners who had enough Salento contexts of the 2nd century appear to con- money to invest in three or more slaves (fig. 17). tain many pottery fragments from other Mediter- The Anninii, Vehilii and Viselii belonged to the ranean districts. latter group and must have been prosperous in- Foremost among the imported ceramics are the habitants of Brindisi.58 There is, however, no need relatively rare amphorae from Cnidus, Rhodes, to assign to all these owners of pottery work- Cos and , which ended up in elite graves shops an elite status comparable to that of people and settlement dumps.60 They contained good like Rammius/Erennius or Statorius in the first Greek wines. While these wine amphorae are rel- half of the 2nd century BC.59 atively rare and were restricted to elite circles, ‘Corinthian’ B amphorae possibly originating from On the basis of the archaeological data discussed Corfu have been reported in considerable quan- above, the following picture of the social landscape tities (e.g. Désy and De Paepe 1990). They date in early Roman Salento suggests itself. In the first mostly to the later 3rd and the early 2nd centuries half of the 2nd century BC, the Brindisi area (and BC. In addition to these transport vessels, there perhaps Salento in general) had a highly pros- are limited quantities of fine wares from the east- perous elite who were connected in some way to ern part of the Mediterranean (fig. 18 nos 1-4). the Aegean and/or other parts of the eastern They include West Slope pots and relief bowls Mediterranean. They adhered to traditional Mes- from across the Adriatic, relief bowls with black sapian burial customs and were buried with heir- to brown gloss and pots with a fine white slip looms that probably referred to the traditional from the west coast of Asia Minor, a few pieces of pre-Roman, ‘Messapian’ society of the late 4th and Syrian black gloss wares and early Eastern Sigil- 3rd century BC. It is perhaps significant that these lata A (Yntema 1995b, 1997, 2001). They all bear tombs are encountered in settlements which were witness to Salento’s contacts with the opposite originally walled native centres that survived as shore of the Adriatic and the Hellenistic king- small towns in the Roman period and did not de- doms of the eastern Mediterranean. Since almost cline into hamlets or ghost towns (as many other all the amphorae and fine wares date to the end Messapian towns did). Both Brindisi and Lecce, of the 3rd and the first half to third quarter of the the substantial Salento towns of the Roman period 2nd century BC, these contacts must have been and intensely inhabited until the present day, have particularly intense in that period. The presence produced ample evidence for the existence of a of fayence and Rhodian or Alexandrinian glass in group of moderately prosperous town dwellers. elite graves of northern Apulia (e.g. Canosa) dat- They can be traced to between the mid-3rd and the ing to approximately the same period reinforces mid-1st century BC. By the late 2nd century BC, the this hypothesis. stamps on the Apulian amphorae testify to the The western Mediterranean is also represented presence of slaves, mostly of Greek origin; their among the imported pottery found in Salento. numbers were probably fairly substantial. They There is an unspecified quantity of amphorae of made their pots in workshops owned by people the Punic type, possibly from the western part of whose names referred either to local origins or to Sicily. Our digs at both Valesio (1984-1990) and central Italy and who could afford to buy three or Muro Tenente (1994-2002) have each produced a

113 Fig. 18. Salento district. Imported pottery of the late 3rd and 2nd century BC: 1-2. West Slope wares from Brindisi (kantharos) and Valesio (calotte shaped bowl), production area uncertain; 3-4. Relief decorated bowls (Ephesos fabric) from Valesio; 5-6 rims of amphorae de tradition punique (west-Sicilian fabric) from Valesio and Muro Tenente; 7-8. Campana A wares (Naples fabric) from Valesio.

few specimens of this type of container vessel (fig. fused. This class is fairly common on the Dalma- 18 nos 5-6).61 They are present in contexts dating tian coast, and apparently rare in mainland Greece, to the middle to later 3rd and earlier 2nd century BC. but occurs in substantial quantities at Rhodes, Western fine wares are represented by the black Crete and the coast of eastern Libya and Egypt gloss class commonly called Campana A wares, (e.g. Sidi Khrebish/Benghazi, Alexandria).63 On which stems from the Naples area (e.g. Morel the strength of the seriation of the Taranto tombs 1981; Yntema 2001). These made their appearance (Lippolis 1994, Gräpler 1997), many of these ex- in the district under discussion around the mid- ported pieces can now be dated to the 2nd, 3rd and 2nd century BC (fig. 18, nos 7-8).62 4th quarter of the 3rd century BC and therefore tes- The Salento district not only imported pottery; tify to Taranto and other production centres in it also exported ceramics, both with and without Salento having external contacts in the period contents. Among the Salento fine wares that reach- after the Roman occupation. It should be noted ed areas outside southern Italy, Gnathia pottery that, once again, the eastern Mediterranean seems stands out because of its wide distribution (Green to have played a dominant role in these contacts 1979). Although this class is traditionally dated to (fig. 19). a period prior to the Roman conquest of Salento It is uncertain whether any Salento fine wares (e.g. Green 1968, 1971), it is now clear that consid- or transport vessels were exported in the period erable quantities of Gnathia wares were still pro- from the end of the 3rd and the first half of the 2nd duced during the mid- to later 3rd century BC. century BC. Some 30 Black Gloss pieces from Especially the late phase of Gnathia wares, char- Libyan Sidi Khrebish (Benghazi) may in fact be acterized by elaborate ribbing and only a limited Apulian, which suggests that Apulian Black Gloss quantity of painted ornaments, was widely dif- wares have simply not been recognized in the

114 eastern Mediterranean.64 By about the later 2nd century BC, however, large quantities of Brindisi amphorae can be shown to have reached foreign shores. They were distributed throughout the Mediterranean, although once again the largest quantities are found in the eastern Mediterranean (cf. Cipriano/Carre 1989).

Changes in regionally produced classes of pottery may also be due to foreign influxes, testifying to direct or indirect contacts with other parts of the Mediterranean world. Since this is the perspective from which the various Hellenistic classes of Sa- lento pottery are now being studied,65 it may be useful to provide some general details here. With regard to the regionally produced fine wares of Fig. 19. Distribution of Apulian Gnathia ware Salento, there is a marked eastern influence in the (from Taranto workshops) outside southern Italy, late 3rd and first half of the 2nd century BC. Some 3rd century BC. forms, which were without antecedents in the regional pottery tradition of Apulia and adjacent parts of Lucania, suddenly made their appearance in this period. Most conspicuous among them are the hemispherical cup (mastos) and the flask (lagy- nos) which occur in both imported and regionally made classes of fine wares (fig. 20), while the characteristic eastern Mediterranean technique for the production of mouldmade so-called ‘Megarian’ bowls was also adopted in southeast- ern Italy.66 Influence from Tyrrhenian fine wares can be found in the later 2nd and early 1st century BC, when large plates and bowls current in the Campanian Campana A and Sicilian Campana C Fig. 20. Eastern shapes in Salento fabrics: (a) mastos pottery were also produced in regional work- (Brindisi fabric: Hard Fired Red), first half of 2nd cen- shops in southeastern Italy. Both the eastern- tury BC; (b) calotte-shaped bowl (Brindisi fabric: Hard Mediterranean and Tyrrhenian influence on pot- Fired Yellow), first half of the 2nd century BC; (c) lag- tery production in Salento coincide with imports ynos (Taranto fabric: Apulian Grey ware), middle to of fine wares from these regions. 3rd quarter of 2nd century BC. A remarkable instance of Tyrrhenian influence, however, can be found in the coarse cooking wares in quite a range of sizes (capacity between approx. of the Salento peninsula. This class of pottery ap- 0.20 and 1.0 litres), but even the largest specimens pears to be very common in Messapian settle- contain only about 50% of the contents of the ments in periods both before and after the Roman average stewing pot of the traditional type. conquest. The repertoire of forms consists mainly Although the shape of this new type of cooking of deep to fairly shallow stewing pots used for pot has absolutely no precursors in Salento, it has simmering and casseroles in which food was baked a long history in central-Tyrrhenian Italy, hence (fig. 21, nos 1-3; cf. Yntema 1991, 178). These shapes the name ‘Latial’ cooking pots (Yntema 2001). The have their antecedents in Greek-speaking areas dimensions of the ‘Latial’ cooking pots and the and may be indicative of the massive adoption by traces of soot on the numerous vessels of this the Messapian populations of Greek methods of shape found during the excavations at Valesio cooking during the 4th century BC. In the last between 1987 and 1990 suggest that they were third of the 3rd century BC, however, a completely used for simmering. This use differs markedly new type of cooking pot emerged in Salento along- from the traditional Salento stewing pots, how- side the traditional shapes. It is relatively tall and ever. While the latter type probably contained the has a flat base and slightly out-turned rim (fig. 21, entire family meal, the ‘Latial’ cooking pots appear no 4). Such tall and flat-based cooking pots occur to have been used to cook various ingredients

115 that together made up a meal. The ‘Latial’ cooking pot is by no means the only new form of coarse ware in the ceramic samples of Valesio of the late 3rd century BC. At least four or five other forms made their appearance in the 3rd-/early 2nd-cen- tury contexts, including the characteristically Roman clibanus or baking lid (fig. 21. nos 4-6). Thus the appearance of the ‘Latial’ cooking pot marks the introduction of a new method of cook- ing. Since it happened in the decades immedi- ately after the influx of central-Tyrrhenian people into Salento (the foundation of the colonia latina of Brindisi in 246/244 BC), it is tempting to see a link between the two phenomena and to suggest that colonists from Rome and the surrounding areas were responsible for the introduction into Sa- lento of what was probably in part a Latial method of cooking.

5. INTEGRATION OF ANCIENT WRITTEN SOURCES AND ARCHAEOLOGICAL EVIDENCE

5.1. Pictures produced in the past: the old story

The construction of post-Roman conquest Salen- to’s past in particular and present-day Apulia in general has to be based on the two main sources currently available to scholars: the ancient writ- ten sources and the archaeological evidence. The picture that emerged from these studies until well in the 1990s was a rather dismal one (e.g. Toynbee 1965; Ghinatti 1977; Gabba/Pasquinucci 1979; Sirago 1993). Roman southern Italy was portrayed as a sparsely populated area of large ranches with extensive stock raising. People living in this de- pressing environment were thought to be mainly or exclusively slaves: the villicus (mostly a freed- man) representing an absentee landlord, and the slaves who did the hard work. Substantial towns, which were invariably regarded as centres of cul- ture in antiquity, were absent there, making Roman Salento a cheerless and culture-less land- scape that held no attraction for further research. The scanty ancient written sources impressed, so to speak, this picture upon the reader. Salento’s misery was believed to have been substantially aggravated by the ‘southern’ episode of the Second Punic war (216-209 BC). Toynbee’s Hannibal’s Legacy (Toynbee 1965) has done much to promote an image of early Roman southern Italy that sus- Fig. 21. Traditional and new ways of cooking in piciously resembles the sub-recent Mezzogiorno Salento: 1-3. Traditional Salento cooking pots with stereotypes (Gabba et al. 2001, 29). Since Salento Greek pedigree (4th-early 1st century BC); 4-6. New is a district of southern Italy, modern historians cooking wares with central-Italic background (late 3rd- of the ancient world clothed it in the same rags as early 1st century BC). other parts of the Roman Mezzogiorno. Salento

116 (and, in fact, every southern Italian district) la- areas where most chronologies are based on pot- boured under a highly generalized image con- tery sequences. Whoever truly believes that Sa- structed for the whole of southeastern Italy. We lento was poor and depopulated after the Roman should realize that this general picture of south- conquest is unlikely to find rich tombs from the ern Italian poverty and submissiveness needs to 3rd and 2nd centuries BC. be carefully checked for each period for which it The second bias influencing the image of post- is proposed. Roman conquest Salento can be found in the ar- chaeological data set itself. Although it is certainly The image of a once flourishing Salento reduced not correct to say that all ‘Messapian’ settlements to a state of poverty and misery in the period fol- faded into insignificance after the Roman con- lowing the Roman conquest and the Hannibalic quest, this did happen to quite a number of them war, was - it seemed - completely confirmed by the in the late 3rd and early 2nd centuries BC (Yntema archaeological evidence. Many burials containing 1995a, Burgers 1998). Especially affected were the bronze vessels, Greek pots with figured decora- native walled settlements that were not built over tion and other objects of high value (e.g. personal in more recent times. These became the most like- ornaments of gold) had been excavated since the ly target for urban surveys and stratigraphical 19th century. As greater insights were gained into excavations, completely confirming the cliché the dating of Greek figured wares, many of the image of poverty and backwardness based on Apulian red-figured and Gnathia wares found in ancient written sources and 20th-century percep- the Salento tombs appeared to date to the 4th or tions of the Italian Mezzogiorno. However, what early 3rd century BC. It was thought that very few actually happened after the Roman conquest was tombs could be dated to the 3rd and 2nd centuries a hierarchization of the Salento settlement pattern. BC. These partly incorrect data confirmed both Sa- A series of formerly substantial native settlements lento’s 4th-century BC prosperity and Salento’s ab- was reduced to villages or hamlets. Other settle- solute poverty in the period following the Roman ments retained their former status and size and a conquest. In addition, the data collected during few Messapian towns developed into major re- the settlement excavations carried out in various gional centres (Burgers 1998). While the Salento native walled centres since the 1970s made it isthmus between Taranto and Brindisi harboured patently clear that these centres flourished in the at least ten fortified settlements around 300 BC,68 4th century BC and had dwindled into insignifi- the same area is likely to have contained only three cance by the 2nd century BC.67 towns with territoria in Roman times: Taranto and Various biases were responsible for this archae- the Ager Tarantinus, Oria and the Ager Uritanus and ological picture of a poor early Roman Salento. Brindisi and the Ager Brundisinus. Lupiae (Lecce) First of all, there were the ancient written sources was the principal Roman town in southern Salen- themselves, which were valued highly by those to (e.g. Giardino et al. 2000).69 Since these four studying aspects of the classical world (including towns have survived until the present day, there archaeologists). The landscapes they evoked were is scant evidence for their Roman phase: Roman also entirely convincing. In fact, the few surviv- strata were obliterated by medieval and post- ing texts by ancient Greek or Roman authors on medieval activity and Roman building materials this subject were fostered and treated with the were recycled. In addition, heritage management utmost reverence: ‘the widespread sentiment that of ‘urban’ archaeology in Apulia is still in its anything written in Greek or Latin is somehow infancy. However, a recent review of the ‘urban’ privileged, exempt from the normal canons of evidence has revealed a more positive image of evaluation’ (Finley 1986, 10). No-one seemed to Roman southeastern Italy (Lippolis 1997). As we notice that the image of a conquered Salento con- have seen above, it is especially the evidence structed on the basis of these texts closely resem- regarding the rural sector that belies the tradi- bled that, for instance, of conquered Greece (cf. tional picture. Alcock 1993) and was too cliché to ring true. Nevertheless, this all means that we have to Moreover, the picture presented of a post-Roman compose a new narrative about the early phase of conquest and post-Hannibalic Salento was exactly Roman Salento. In the above sections, both types what we would have expected since it corre- of evidence have been revised and updated, lay- sponded so nicely with 20th-century perceptions ing the foundations for the creation of another of southern Italy. Of course, chronologies can be narrative on the area under discussion in early adapted to such ill-conceived frameworks. This Roman times (i.e. between ca 260/250 and the late observation is especially true of Mediterranean 2nd century BC).

117 5.2. Early Roman Salento: a new narrative a significant drop in the number of graves has been reported. As for craft, there are reasons to 5.2.1. The decades immediately following the Roman believe that local pottery production intensified conquest in the second and third quarters of the 3rd century BC. This is especially true of the black gloss fine The Roman conquest of 267/266 BC must have wares (Yntema 2001). Furthermore, the composi- been a traumatic event for the various local groups tion of the livestock of Valesio, only 14 km south in Salento. They (or at least some of them) had of Brindisi, barely changed in the later 3rd century lost other battles against ‘foreigners’ in the past,70 and no shifts can be observed in culling patterns but they had never before suffered defeat at the in animal husbandry (Veenman 2002). These data hands of an enemy who controlled the rest of indicate that there were no substantial changes in peninsular Italy. The prospects of regaining their the pastoral strategies used by the people of that independent status must have been rather bleak. settlement. In the period immediately following It cannot have escaped the leaders of the Messa- the Roman conquest, the new elements for Calabri pian tribal groups that Rome was the dominant and Sallentini were their bond to Rome, the group power in Italy and that Greek princes from Sicily of Latins in Brindisi, taxation and levies for or from across the Adriatic (now the only alter- Rome’s foreign wars (the and natives to Rome) were unlikely to start military Illyrian War), and possibly partly dismantled for- interventions in southeastern Italy after the Italian tification systems. It seems that in the second and misadventures of, for instance, Archidamos of third quarters of the 3rd century BC most aspects , Alexander the Molossian and Pyrrhus of of everyday life in many Salento settlements were Epirus. The Roman victory in the war against either unaffected, or only marginally affected, by Pyrrhus and the Italian South was doubtlessly the consequences of the Roman conquest. hard fought, but it was highly decisive. The foedus with Rome regulated the relationships of the Sa- 5.2.2. The final third of the 3rd century BC lento societies with Rome, while the foundation of a colonia latina - probably within the Messapian In the final third of the 3rd century BC the first series oppidum of Brindisi - made it absolutely clear to of major changes can be observed in the archaeo- the conquered that things were happening in their logical record. At the site of Muro Tenente, a per- world on which they had no say. This meant that ipheral quarter inhabited by town-based farmers from 246/244 BC onward, two political systems and artisans was abandoned in ca 230/220 (Bur- co-existed in Salento: the traditional Messapian gers/Yntema 1999). At about the same time a pot- system and the fundamentally Roman system of tery workshop began production at Valesio in the the colonia latina of Brindisi (Yntema 1995a). The very centre of the town, blocking a 4 m-wide main events surrounding the arrival of the Romans and road (Yntema 1994). Urban surveys of walled their further implications can be deduced from towns in the Brindisi area show that the first the written sources. These have been discussed in shrinkages of their settlement areas can be dated section 2 above. to the period 230/200 BC. Moreover, burials al- Outside Brindisi and outside the arena of high most completely vanish from the archaeological politics very little happened. The numerous walled record, while the quantities of votive offerings towns of the Calabri and Sallentini continued to be deposited in the traditional sanctuaries of the prosperous centres, probably with agriculture and Messapian groups display a marked decline (cf. stock raising as their key economic activities. The D’Andria 1990). settlement areas of four walled sites in the Of course, the in 216 BC Brindisi district, investigated by means of urban marked the beginnings of a short period of crisis surveys, neither increased nor decreased in size; and turmoil in Salento. While troops from the area population density was also approximately the under discussion are likely to have played a role same as before the Roman conquest (Yntema in the initial years of the Punic war (cf. the role of 1993b, Burgers 1998). The mainly extra-mural Dasius Brindisinus at Clastidium in 218; see sec- sanctuaries of the Messapian groups continued to tion 3.4 above), Hannibal’s stay in southern Italy bind them together, while the rural farmsteads during his attempt to take Taranto was a severe thrived in the countryside surrounding the towns test of the loyalty of Calabrian and Sallentinian (Boersma et al. 1991; Yntema 1993a). A substantial societies to the Roman cause. The years 213-209 change can be observed in the funerary sphere, must have been particularly tiresome. A few Sa- however. From about the mid-3rd century onward, lento towns went over to the Carthaginians; Man-

118 duria (ca 30 km east of Taranto) and Uxentum ences increased very considerably in native Salen- (Ugento in the southern tip of the Salento penin- to during the later 3rd century BC. This process of sula) are explicitly mentioned by Livy. But the fact social differentiation had a long history in the area, that Livy does not mention important settlements starting with the first formal burial in the final on the Salento isthmus (e.g. the colonia latina of decades of the 7th century and the ‘princely’ tombs Brindisi, the large, major tribal centre of Uria of the tribal chiefs of the late 6th, 5th and early 4th Calabra, now Oria, only 35 km east from Taranto) centuries (cf. D’Andria 1988). The 4th century wit- may indicate that they - and along with them the nessed the birth of wealthy, probably landowning, vast majority of Messapian towns - remained loyal tribal elites that ranked well above the large group to Rome.71 This hypothesis is reinforced if we con- of moderately prosperous tribal farmers (Yntema sider that several high-ranking men with native 1993a). While the elite probably lived in the very Salento roots played a role in the final phase of centre of the Messapian walled settlements and the Second Punic war (Ennius on Sardinia) and in were indeed buried in well-defined, elaborate the three decades following that war (e.g. Ram- burial plots (mainly 4th century BC) or sumptuous mius/Erennius’ role in the beginning of the Third hypogaea (mainly 3rd century BC) in the same cen- Macedonian war). It is unlikely that the Salento tral area of the settlement, the tribal farmers lived societies managed to survive this period of war in more peripheral parts within the town walls or unscathed. The towns that defected to Hannibal on the farmsteads in the countryside surrounding were doubtlessly punished by the Romans when the walled centre. In the later 3rd century BC, how- they retook them in 209. The towns that stayed ever, prosperous farmers had become invisible, loyal to Rome - especially those on the Salento although the probably elite-owned rural farm- isthmus close to Taranto - must have suffered steads where some of them lived still survived. under the Carthaginian troops. Farmsteads out- Moreover, the often extra-mural sanctuaries, which side their walls may have been destroyed and were probably controlled by the elite and played olive groves and vineyards may have sustained a role in forging tribal or sub-tribal allegiance damage. Both the Oria and Valesio field surveys, (Yntema 1995a), were in decline. Thus the read- however, have shown that there are no signs of a ing of these data suggests that these substantial significant break in the rural occupation at the changes were the first signs of the decomposition time of Hannibal’s war: the isolated farmsteads of the tribal world of the Messapians. They her- and hamlets continued to exist. ald the birth of the very different Salento society To sum up, the image that can be constructed that emerged in the 2nd century BC. for Salento in the last third of the 3rd century BC presents both positive and negative aspects. It 5.2.3. The first half of the 2nd century BC shows marked signs of change, which can be extrapolated from the written sources (the nega- In the first half of the 2nd century BC Salento was tive effects of Hannibal’s stay in Salento) and again confronted by wars. This time Salento was which are equally clear from the archaeological the starting point for, rather than the victim of, data: there is an incipient reduction in the settle- warlike activity. The ‘Greek’ wars72 were all fought ment areas of several towns, a sharp drop in the in areas for which the Salento peninsula, as Italy’s number of formal burials, a decline in the number bridgehead to the eastern Mediterranean and sep- of votive offerings in sanctuaries, but a strong arated from Epirus and Corfu by the narrow, 70- continuity of rural occupation. It would, however, km wide Strait of Otranto, was superbly placed be going too far to hold Hannibal responsible for to serve as both starting point and provisioning all the negative developments that occurred in base for Roman campaigns. Dozens of passages this period. It should be noted, for instance, that in Livy’s accounts of these wars confirm that this the number of burials was already on the decline was in fact the case. Salento in general and Brin- well before the second Punic War. If Hannibal’s disi in particular played a key role in Rome’s mil- stay in southern Italy had any role in the long-term itary actions against the Greek states of the south- changes in this area, it was as a catalyst in - and ern Balkans. certainly not as the cause of - such changes. According to these same sources, the Roman The archaeological record for the last third of praetors assigned to Taranto or Apulia faced very the 3rd century BC perhaps displays the first signs substantial unrest in their province, mainly in the of far-reaching changes that were to completely decade between 190 and 180 BC. Livy describes change the societies of Salento in the 2nd century these difficulties as a ‘slave revolt’ (motus servilis) BC. There is reason to suppose that social differ- and ‘herdsmen’s conspiracies’ (pastorum coniura-

119 tiones), but we should be careful not to take these However, the more than 50% that survived were words too literally. Although slavery is likely to two to four times larger than their predecessors. have occurred in the tribal societies of Salento, we Thus the farms of the mid- to later 2nd century BC should question whether it was sufficiently wide- probably had a much larger population than those spread in the early 2nd century BC to trigger mas- of the 3rd century BC. While the number of inhab- sive slave revolts. Of course, a large influx of new itants of the larger, walled settlements probably slaves in the area under discussion may indeed fell sharply (perhaps from a few thousand to a few have occurred as a result of the Second Macedoni- hundred), those in the rural area surrounding an War (200-197) or the first clash with the Aeto- these settlements may have risen. lians (192-189), but it is highly debatable whether However, these patterns of severe urban de- the economic infrastructure of southeastern Italy cline and substantial continuity or even growth was in a position to suddenly absorb the thou- in the rural areas are unlikely to hold for every sands of new slaves needed for a decent slave area in Salento. It should first be noted that most revolt.73 of the field surveys have been carried out in the Section 4.2, moreover, has demonstrated that northern part. As a result, the patterns for rural urban surveys suggest a very substantial decrease occupation may reflect a general trend that applies in the settlement areas of many fortified towns only to that area. Brindisi occupied a special posi- during the first half of the 2nd century BC. At tion there, which means that developments in the Valesio (14 km south of Brindisi), for instance, the hinterland may have differed from other parts. We originally 30 to 35 hectare habitation area (of the have seen above that the decline of many former approx. 80 hectares within the walls) declined to fortified settlements was contemporaneous with about 17 to 20 hectares in the late 3rd century and the steep rise of Brindisi (and probably Lecce). did not exceed 5-6 hectares around the mid-2nd However, this phenomenon would probably only century BC (Yntema 1993b). Comparable decreases partly compensate the losses sustained by other can be found in at least three other walled sites of former larger settlements. All in all, the population the Brindisi area, a fate which may have been trend in Salento between the late 3rd and mid-2nd shared by many Messapian settlements (Burgers century BC is likely to have been a negative one. 1998). Thus these towns sustained extremely heavy The funerary evidence seems to reflect this pat- population losses. Judging solely by the shrink- tern to an extreme. In the first half of the 2nd cen- age of their habitation areas, they may have lost tury BC, a fair number of burials in substantial up to 60 to 80% of their inhabitants between ca and coherent burial plots occurred exclusively at 250 and 150 BC.74 Brindisi and Lecce (Andreassi/Cocchiaro 1988; While many Messapian settlements declined, Giardino 1994).75 Burials of the period under dis- Brindisi boomed. It became a regional centre of cussion are absent from the remaining - formerly considerable consequence and must have had a larger - settlements. The same holds for the large very substantial number of inhabitants by the number of small rural settlements. The grave mid-2nd century BC. The colonia ranked first in the groups close to the isolated farmsteads and ham- now pronounced regional settlement hierarchy. It lets can invariably be dated to between the late was also Italy’s second largest port and the launch- 4th and the mid- to later 3rd century BC. Although ing pad for Roman enterprise in the eastern Med- more than 50% of the farmsteads can be shown to iterranean. It harboured craftsmen working for have thrived in the 2nd century BC, no graves of both naval and overseas activities and the regional this period have been reported. The almost com- market. Vibrant Brindisi expressed its newly-won plete absence of burials outside the new centres status in an urban sanctuary whose temple was of Salento surely does not indicate that no people decorated with magnificent capitals (2nd century lived or died outside the walls of Brindisi and BC). Alongside Brindisi there were a few towns Lecce. From these data it must be concluded that of secondary importance like Lupiae (Lecce), Uria the funerary practices in these two fast-growing Calabra (Oria) and Gnathia which also had urban towns differed from those in the dying ‘urban’ cult places. Otranto (Hydruntum) and Ugento settlements and the rural settlements of the Salen- (Uxentum) may have had a comparable status in to district. the southern tip of Salento. Such differences in funerary practices within the We have seen that changes also affected rural same area may have a variety of causes. They can sites in the first half of the 2nd century BC. In gen- be ascribed, for instance, to the co-existence in the eral, the hamlets survived, but the number of iso- same district of different ideas about burial rites: lated rural farmsteads fell by about 30-40 %. the new Latin settlers at Brindisi probably differed

120 in this respect from the native population of Sa- a sudden increase in the rearing of pigs between lento. This observation does not of course explain the end of the 3rd and the mid-2nd century BC the profound changes in funerary rites (i.e. the (Veenman 2002). disappearance of formal burials) in many other The ancient written evidence indicates that settlements. The differences may therefore be due some of these eminent men were probably in- to variations in the social composition of the pop- volved in overseas trading, which would make ulation between settlements within the same dis- them both large landowners and large-scale trad- trict: the social stratification at flourishing Brindisi ers. They may also have organized in some way may have differed from that of settlements like Va- the transport of products from their farms to the lesio and Muro Tenente that were in great decline. coastal area, carried out by mercatores, and the Since the only graves found in the areas away packing of the products on the coast. Much of the from the new booming centres (Brindisi, Lecce) of agrarian production of the Salento district may in the early Roman period are the highly traditional fact have been in their hands. They were not just and wealthy tombs found at Oria and Mesagne the social and economic elite of Salento, but must (see section 4.3), the latter factor may well explain also have played a similarly key role in regional the problem discussed here. By the first half of the politics, some even figuring on the international 2nd century BC, the social landscape outside the stage (cf. Rammius/Erennius’ role in the early 3rd new regional centres of Brindisi and Lecce was Macedonian war). This perhaps also explains why probably composed on the one hand of a very the majority of these men professed to come from small group buried with ostentatious display and Brindisi. They may well have had their roots and with links to a rich Messapian past and an east- possessions elsewhere, but they had moved to ern Mediterranean present, and on the other hand Brindisi or possessed a pied à terre there, simply of a much larger group of poor people involved because it was both the economic focus of the dis- in agriculture. Those descended from former tribal trict and the new regional centre of power. If they elites or posing as such can be traced, but the arti- wished to gain or retain any political influence in sans and tribal farmers (still traceable in the 3rd the region, Brindisi was the place to be. Many, if century BC) cannot be found in the Salento funer- not all, members of this highly prosperous Sa- ary record of the earlier 2nd century BC. lento elite spoke both Latin and Greek. The poet It should be noted that the ‘native’ sanctuaries Ennius from the Salento town of Rudiae was no near the declining settlements also experienced a exception. The noblemen of Brindisi and other steep decline in the period under discussion. They former Messapian towns travelled widely. Some probably played a key role in the tribal societies were important enough to befriend both Greek of Salento. It was in these sacred places that the princes and Roman senators and to be granted a cohesion between tribal or subtribal units was proxenia by a Greek sanctuary or state (section 3.4). emphasized (Yntema 1995a). Several of them, Since many had their roots in the Salento region, however, had ceased to exist before the mid-2nd they also spoke Messapic and demonstrated their century BC. This also happened at the former allegiance to the Messapian tribal past. For in- tribal centre of Oria, although it did survive as a stance, they were buried in the traditional Messa- Roman municipium. pian way with a number of ‘heirlooms’, and with amphorae from Cos or Rhodes containing exclu- In the first half of the 2nd century BC, therefore, sive Greek wines that showed their links to the Salento is likely to have witnessed a series of Greek world in the eastern Mediterranean. Since highly dramatic changes. Most obvious was the the Greek Hellenistic world was perceived as the enormous increase in social differences. We may epitome of refined culture by both Rhomaioi and assume that the regional magnates with a high Italioi,76 these Salento gentlemen presented them- ‘international’ status evident in both the ancient selves in their graves as highly cultured people. written and archaeological sources were the most The Salento elites of the early 2nd century BC eminent representatives of a class - often it seems were not the only Italic group to display such eco- with Messapian roots - that may have owned nomic and mental ties. They were, in fact, highly large plots of land and many farmsteads. These comparable to the contemporary ‘Daunian’ elites farms produced mainly cash crops such as corn, of the north-Apulian Arpi and Canosa areas (Ro- wine and olive oil (cf. Graeco-Italic amphorae, man Canusium and surroundings), who were in Brindisi amphorae, Varro, RR 2.6.5). The Valesio equally close contact with both Romans and Greeks bone sample suggests that pork may have been (cf. Sirago 1993 and Silvestrini 1998) and were another product of these farms: we observe here buried in large chamber tombs or hypogaea in the

121 traditional ‘Daunian’ way.77 These mainly early Roman occupation of the Salento peninsula, but 2nd-century elite tombs of Daunia also contained since Brindisi was one of the starting points for the precious and rare glass vessels from the eastern the Illyrian war in 229,78 contingents of Calabri Mediterranean and mostly Greek wine amphorae. and Sallentini may well have numbered among the auxiliaries. The Hannibalic war, in which Italy At the other end of the social scale were the slaves itself was the battleground for quite a number of and the former tribal farmers. Judging by the years, was an entirely different matter. Although thousands of Messapian tombs dating to the late Livy - understandably - portrays this war as very 4th and earlier 3rd centuries BC, the latter group much a Roman affair, contingents of socii may have must have been relatively prosperous in that pe- played quite a substantial role in the fighting. The riod, but the subsequent changes observed in the numerous emergencies provoked by Hannibal’s settlements, the funerary sphere and the tribal presence are likely to have required a fairly mas- sanctuaries suggest that they became decidedly sive input of Italic troops in general. As we have less prosperous and ‘tribal’ in the final decades of seen above, snippets of information in ancient the 3rd century BC. The farmers of the 2nd century written sources indicate that Messapian troops BC elude us almost completely in the archaeolog- were indeed involved. They were probably present ical record; their presence can only be postulated in the Po valley in 218 BC (Dasius Brundisinus at because substantial farmsteads existed in that Clastidium), they possibly fought at Cannae in period housing some 20 to 30 people. Of course, 216,79 and can probably be assumed to have served social stratification was already on the increase in Sardinia in 204 (Ennius’ ‘assignment’). These before the Roman occupation of Salento. But what two or three instances, very casually mentioned happened to the social group of Salento farmers by Livy, Silius Italicus and Nepos, may be just the in the early 2nd century BC was truly dramatic. If tip of the iceberg. Since the vast bulk of the Salen- we do not accept the possibility that all the Messa- to population was active in farming and stock pian farmers migrated and were replaced almost raising, the vast majority of the men in the con- overnight by slaves (which I find highly unlikely), tingents of Messapian socii must have been farm- then a substantial number of them became peas- ers. We should also bear in mind that the 2nd ants who worked the lands of the elite-owned Punic War was not the only war fought by the farmsteads. Romans in the late 3rd and 2nd centuries BC: Salen- We may well ask how they came to be in that to was often the area controlled by Rome that was position. It is probably too easy to come back to closest to the scenes of fighting (cf. Macedonian, the seemingly evolutionary process of increasing Syrian/Aetolian and Achaean wars). Thus there social stratification in the Messapian societies is reason to believe that substantial groups of (starting in the Iron Age) that has been mentioned Messapians fought for Rome for many years above. Of course, the difficult situation in the between ca 230 and 150 BC, often in Mediterra- Hannibalic war may have accelerated the steadily nean districts that were relatively far from home. increasing social disparities. Salento is likely to An almost structural absence from home of larger have sustained considerable damage during Han- groups of Messapians was an entirely new phe- nibal’s siege of Taranto, but it remains unclear nomenon which presumably had devastating why these acts of war of themselves could bring effects on smallholders. It certainly helped to turn great wealth to the landowning elite and great small tribal farmers into peasants (cf. Brunt 1971; poverty to the tribal farmers. Since the war did not De Neeve 1984). inflict long-term damage on Salento agriculture In addition to the above factors, the macro-eco- (by the early 2nd century BC Salento was actually nomic developments in the Mediterranean of the exporting agricultural products), large landowners late 3rd/early 2nd century BC were not particularly may have had a somewhat better chance of sur- favourable to small farmers. This period of Medi- viving a succession of difficult years than small terranean history is characterized by two closely farmers. But this factor alone hardly accounts for interlinked phenomena: (a) increasing specializa- the enormous rise in social differences in Salento tion on particular products in particular regions, in the late 3rd and early 2nd centuries BC. and (b) strong intensification of interregional trade It should be realized, however, that Rome’s re- and exchange systems. Salento was drawn into peated demands for troops may have had a much larger economic systems in an early variant of the greater impact on Messapian society (cf. Brunt globalization process. In this respect the later Hel- 1971, De Neeve 1984). We have no information on lenistic times represented a decisive step towards this subject for the first 40 years following the a marked regional specialization. This happened

122 in an almost pan-Mediterranean market that e- for craftsmen. The potters, smiths, stonemasons merged in its most marked form in the Roman and carpenters of the dying, former urban centres Imperial period, when Africa Proconsularis stood of Salento who had lost almost their entire clien- for olive oil and Aegyptus for corn. It is perhaps not tele were probably able to exercise their craft in purely coincidental that the monetarization of the the rapidly growing towns of Brindisi and Lecce. Salento societies took place in the same period. New, specialized crafts may have come into being. Bronze coins minted at Brindisi between ca 215 While the archaeological evidence suggests that and 150 BC (Boersma/Prins 1994) and Roman the processing of wool, for instance, was a house- asses displaying Ianus bifrons and the prora minted hold activity in Salento in the 3rd century BC,82 it between ca 165 and 145 BC appear to be common may have become a specialized métier involving in the 2nd-century BC strata of the site of the dying fullers and cloth makers in the 2nd-century BC town of Valesio. Brindisi area. The buzzing town with its natural These economic developments of the early 2nd harbour offered work for people involved in such century BC were highly problematic for small- trades as ship building and ship repairs, and for holders. They required a significant change to ag- shopkeepers, pedlars and those in other forms of ricultural strategy and substantial investments.80 retail trade. Some of the small farmers of Salento And both had to be achieved in a relatively short who lost or sold their lands may have found em- period. If the small Salento farmers were in any ployment in rapidly expanding Brindisi. position to decide for themselves by the late 3rd/ The increasingly close links between the former early 2nd century BC (and were not - or had not Messapian society and the Roman world created become - strongly dependent on the former tribal still more opportunities. It should be noted that the elite), they were unable to invest and participate Salento ‘crisis’ of the early 2nd century BC coin- in large-scale cash crop farming and thus meet cided with the pacification of the Celtic tribes in the new demands of increasingly supra-regional the Po valley. Control of this area was brought markets (Yntema 1993a, 211). The combination of about by the foundation of a series of coloniae in the rapidly changing economic situation and the the 80s of the 2nd century BC, the foremost of far-reaching social and political changes discussed which were Bononia (Bologna), Mutina (Modena), above (increasing social stratification, farmers Parma and Placentia (Piacenza). The colonia of Aqui- fighting for Rome, and peasantization) meant that leia, founded in 181 BC (the predecessor of Venice the small farmers were badly off. They became in the Caput Adriae) was another new, fast-grow- the underdogs in an elite-controlled peasant soci- ing town in the far north. People belonging to the ety, ranking only slightly above slaves in terms of socii of Rome could in principle be included in the status. groups of new settlers (Gabba 1989, 212), thus acquiring both land and Latin citizenship. There This scenario of rural misery and distress may are indeed reasons to believe that substantial apply to quite a number of the early 2nd-century, groups of Salento farmers emigrated from the far former tribal farmers of Salento: they became ten- south to take part in colonizations in northern Italy ants and farm-hands, mainly or exclusively work- (cf. Verzar Bass 1983). Quite a number of inscrip- ing the fields of the large farmsteads owned by the tions from northern Italy of the 1st and 2nd cen- wealthy members of the former tribal elites and turies AD bear the names of gentes who may have prosperous Latin settlers. Livy’s ‘slave revolts’ and had Salento backgrounds (cf. gens Ennia, gens ‘herdsmen’s conspiracies’ in these years may in Egnatia). It would certainly be going too far to see fact have been peasant revolts sparked off by the the Ennii of the Po valley as the descendants of severe decline in the former farmer’s income and the poet Quintus Ennius of Rudiae. But it is at social status; they can perhaps be read as signs of least an attractive hypothesis that they might be widespread discontent among the new peasants southern immigrants who originally belonged to of Salento and other parts of southeastern Italy.81 the clan or clientela of the elite family of the Ennii There is, however, no reason to suspect that the from Rudiae and/or other Messapian towns. We former tribal farmers were all reduced to such a should, for instance, remember that the poet humble status. While the new Salento economy of Ennius was prominent in Rome during the period the early 2nd century BC was probably a disaster in which these northern coloniae were founded.83 for small agricultural units, it created great oppor- When speaking with his patronus M. Fulvius No- tunities in other areas of the regional societies. For bilior, he may have put in a good word for his instance, flourishing Brindisi with its superb har- impoverished tribal farmers who had fought in bour offered excellent opportunities, particularly Nobilior’s army in Greece. It should also be re-

123 membered that Nobilior’s son Quintus was very yards and olive groves produced cash crops for much involved in the foundation of colonies be- far-away markets. Wool and pork may have been tween 190 and 180 BC.84 Other influential Salento produced in unprecedented quantities and the elite families with good contacts in Rome such as flourishing port of Brindisi is likely to have been the Rennii, Pulfennii and Statorii may have done an important centre of artisan activity.86 the same. The Salento district changed enormously within one life span. Moreover, the changes affected al- Early 2nd century BC Salento was a rapidly chang- most every sector of local societies. Almost noth- ing, highly dynamic world. Of those former Mes- ing remained as it was. Even the physical land- sapians who continued to live in Salento, a few scapes in which people lived changed utterly. acquired great wealth and befriended Roman sen- Several pre-Roman towns were reduced to vil- ators and Greek princes, while others were well- lages, while others boomed. A countryside lit- to-do large landowners of regional significance. tered with small farms and divided into small Much larger groups, mainly former small farmers, plots of land producing a variety of products pre- were reduced to poverty and became peasants. dominantly for local or regional use was trans- But substantial groups - made up of both artisans formed into a landscape of much larger farms, and small farmers - probably migrated from their large cornfields, large vineyards and large olive native home towns. They went to Brindisi, Rome groves producing mainly for far-away markets. If and Corinth, but especially, it seems, to northern Ennius, who was born in 239 BC, had left his Italy. They found prosperity in districts that were home district early in the Second Punic war at the sometimes hundreds of kilometres from their age of 20 and returned to Rudiae shortly before his places of origin. It is likely that many of these death in 169 BC without intermediate visits, he migrants participated in the formation and devel- would have been astounded: of the world he opment of new societies on the borders of the remembered from his youth, almost nothing had expanding Roman state. These people with Salen- survived into his old age. to roots contributed to the Romanization of Gallia Cisalpina, not to the Romanization of Salento, and 6. CONCLUSIONS may claim their rightful place in quite a different narrative about the past. As we have seen in section 3, data from the ancient The first half of the 2nd century BC was there- written sources about Salento in the approximately fore a critical period in the history of Salento. It wit- 100 to 150 years following the Roman conquest nessed the genesis of a Salento society that differed can inform us about various topics. First of all, in many respects from the Messapian world of the they give us the dating of some key events that 4th and 3rd centuries BC. Politically, Salento increas- occurred in Salento: they supply information on ingly became a constituent part of the expanding the histoire événementielle by presenting the mo- Roman state, while intensifying its links with the ments in Salento’s history that the authors - who Greek kingdoms of the eastern Mediterranean. In were mainly Roman - believed to be of major social respects, the period is characterized by importance. These were the defeat and bondage complete loss of tribal cohesion (detribalization) of the Salento polities to Rome with an increasing demonstrated, for instance, by the disappearance loss of independence (the Roman victories of of tribal sanctuaries on the one hand, and by dra- 267/266), a blatantly Roman infringement on matically increasing social differences in rural areas local societies (the foundation of the Latin colony on the other. This period also saw the rise of a of Brundisium, 246/244 BC), a major war fought class that included artisans, small merchants and in and near Salento (Hannibal, 216-209 BC) and pedlars, making Brindisi a consumer town,85 the motus servilis and pastorum coniurationes (prob- whereas the earlier Messapian walled settlements ably a peasant revolt) of the years 186-181 BC. of the 4th/early 3rd century BC - with their high Since these events must have had a substantial percentage of farming inhabitants - had definitely impact on the regional societies of Salento, they been producer towns. The religious and political may also have numbered among the évènements rites that bound the members of this new society remembered by the inhabitants with roots in the together were no longer performed in tribal sanc- district as important landmarks in the past as tuaries outside the settlements, but took place in they saw and experienced it. Thus Q. Ennius (239- the very centre of town (cf. Yntema 1995a). In eco- 169 BC), the Salento native and Roman poet, may nomic respects, Salento flourished as never before. have heard stories from his grandfather about Large farmsteads nestled amidst cornfields, vine- Pyrrhus’ elephants, the battle of Ausculum of 279

124 and the Roman troops entering and perhaps Mediterranean in mid- and late Hellenistic times plundering his hometown of Rudiae in 267/266 with its ever-intensifying interregional contacts BC. When in Brindisi, he may have told his sis- resulting from an ever-increasing regional spe- ter’s young son, the later Roman painter and cialization. The intensity of these contacts and the playwright Pacuvius (ca 220-130 BC), about these degree of regional specialization reached great ‘historical’ events, about the major turning points heights in the Roman Imperial period. in the war with Hannibal and about his personal A fourth subject about which the ancient writ- experiences in the battle of Cannae or his stay in ten sources inform us is that of eminent Salento Sardinia. men. Greek inscriptions, where we encounter In addition to these events, the ancient written Rennius, Pulfennius, Statorius and Dazos, pro- sources provide information on a second topic: vide appealing evidence. If we link these inscrip- the topography and the quality of the soils. They tional data to the evidence from Roman literary tell us the locations and names of a series of set- sources about people with Messapian roots (the tlements. The names can usually be linked to sites, unfortunate Dasius Brundisinus, the influential although there are many more sites in the archae- Rammius or Erennius of Brindisi, the venerated ological record than names in the ancient written poet Ennius ‘Rudinus’ and the painter/tragedian sources.87 Information on soils can of course also Pacuvius), we gain useful insights into the func- be obtained from modern geophysical research, tioning of a particular, highly prosperous social but since the soils in Salento have changed con- group living in the period directly after the Han- siderably (e.g. through severe surface erosion, nibalic war. While data concerning traffic, eco- highly destructive deep ploughing), it is often dif- nomic activities and interregional contacts can ficult to obtain the desired results. All these pieces frequently also be gleaned from archaeological of geographical information contained in the sources, this particular type of information about ancient written sources help form the backdrop people can only be derived from a combination against which the events and medium-term de- of ancient literary sources and inscriptions. For velopments were staged. They may also tell us instance, we learn about their links to other Medi- how outsiders perceived the district in the period terranean areas and sometimes about how Greeks under discussion. But to be frank, these comments and Romans perceived them. We are also offered on Salento barely go beyond the usual clichés and tantalizing glimpses of the identities they con- platitudes expressed about many regions in the structed for themselves in particular situations. Mediterranean: the soils were good, but the area was thinly populated and there were no towns to Information on the same subject can also be found speak of. As we have seen above, the latter two in the archaeological sources. These, too, concern features in particular figure among the recurring eminent men of early Roman Salento. Their graves stereotypes used to describe conquered, town- and the burials gifts they contain tell us about how dwelling societies: they are poor in people, poor they presented themselves in the society where in culture and not particularly prosperous (cf. they had their roots. In their graves, they appear Alcock 1993).88 This aspect of the written sources as wealthy noblemen with evident links to a rich is therefore not especially informative and may native past and an opulent Greek presence. Al- even prove deceptive. though posing as Romans in the inscriptions in the Data of a seemingly comparable nature are given great Greek sanctuaries, their burials in Salento for a third aspect of early Roman Salento: the dis- bear no traces of their ‘Roman’ identity: in the trict’s role in traffic and trade. Information in the society of their ‘roots’ they stress their noble native ancient written sources about this field appears origins and their contacts with the elites of highly much more solid, however. The literally dozens of cultured Hellenistic Greece and other parts of the completely casual mentions of Salento in general eastern Mediterranean (cf. Yntema, in print). In this and Brindisi in particular (for instance by Polybi- particular field there is a happy marriage between us and Livy) as the point of departure for cross- archaeological and ancient written sources. ings to Greece and/or a provisioning area for the It is not really surprising that the archaeologi- Roman armies operating in Greece indicate that cal evidence on Salento illustrates mostly medium- these passages were not just clichés: Salento actu- term developments, the Braudelian conjonctures. ally was a crucial area for the conquest of Greece It is widely accepted that archaeology has an enor- and the provisioning of the Roman armies in the mous contribution to make to the construction of southern Balkans. It was also equally vital to ship- the past in this particular field. Data collected by ping and trade between the eastern and western means of rural and urban field surveys inform us

125 about population trends, changes in rural and ternal and external forces. The present narrative urban settlement and changes in settlement hier- shows an early Roman Salento that underwent archy. Moreover, social and religious change can enormous changes, which - while highly beneficial be studied on the basis of results from both field to some - were rather disadvantageous to others. surveys and excavations. Changing settlement By the early 1st century BC the Salento landscape patterns have social implications, but shifts in the consisted of both boomtowns and ghost towns, of location of religious activities have implications vineyards, olive groves, cornfields and substantial for both the religious and the social landscape. farmsteads, and certainly not solely of thinly pop- The excavation of cemeteries and sanctuaries pro- ulated wastelands with flocks of sheep and goats. vides valuable evidence on social change and the The stereotypical absentee landlords (preferably changes in ritual domains. living in Rome) have been replaced by regional Ancient written sources on the Messapian world noblemen often, it seems, with Messapian roots, in the aftermath of the Roman occupation are who lived and were politically active in Brundisium absolutely scarce. Nor, at first glance, are archaeo- (Brindisi) or Lupiae (Lecce). The number of small logical sources particularly abundant. But a small tribal farmers declined and some became peasants, number of excavations and a series of intensive but the expanding Roman state and the booming field surveys appear to have produced useful ideas economy also offered this group new challenges on quite a number of aspects of Salento societies and prospects. Salento people living in these evi- in the later 3rd and earlier 2nd centuries BC. These dently stirring times were thus actively involved two types of sources provide information on dif- in the ‘Mediterranization’ of their district. More- ferent levels. Whereas archaeological sources tell over, they sometimes participated in the forma- us about developments covering time-spans rang- tion of a substantial series of - to all appearances ing from decades to centuries, the ancient written - Roman communities that were hundreds of kilo- sources tell us about events, people and personal metres from their Messapian roots: the coloniae of views. In quite a number of cases these two sources the northern Adriatic and the Po valley. seem to converge (e.g. economic issues, self-rep- Of course, early Roman Salento may appear to resentation of the elite) or to complement each be an exception to the general rule of a highly im- other (e.g. the decline of tribal farmers). Indeed, if poverished Roman South as laid down by Toynbee we integrate both types of sources on the past and others. Other districts of southern Italy and (e.g. the picture of the Salento elites of the early later periods of Salento’s Roman history may bear 2nd century BC), attractive images of the early a much closer resemblance to what I have called Roman phase of Salento start to emerge. Together the ‘cliché image’ of the poor Roman Mezzogiorno. they produce a nuanced approach in which both The present case, however, is a warning against the textual and archaeological interpretations can accepting images of the past that are based on an sometimes be tested to establish possible biases. uncritical selection and ‘easy’ interpretation of a In this way we begin to see the general outlines limited set of partially biased ancient written and some of the details of another narrative on sources. the former Messapian, now Roman, world in the The combined information from both historical region and period under discussion. Some of these and archaeological sources has produced images have been presented in the above sections. In this of a changing Salento that differs in many respects story about the past, the poor and desolate Roman from those of the past. The period discussed here South of Toynbee (1965), Gabba & Pasquinucci (the 3rd and 2nd centuries BC) was actually the first (1979) and Sirago (1993) does not suddenly ap- stage of a Romanization process that turned a pear to be a world of opulence for all tiers of local partly urbanized Salento, inhabited by the funda- societies. But the integrated approach at least lib- mentally tribal entities of Calabri-Sallentini or Mes- erates us from the long shadows of the subrecent sapii, into a definitely urbanized part of the Roman Mezzogiorno that were cast upon Salento and other state. This process was slow and was certainly not parts of Roman southern Italy during the 20th cen- completed before the birth of the Empire under tury. Early Roman Salento was certainly not a the reign of Augustus. piteous and unresisting victim of Roman occupa- The analyses of material culture alone do not tion, Hannibalic devastations and colonialist im- help us much in establishing how various groups perialism of the Roman senators and their friends. and individuals became Romans. In this respect It was a dynamic world that evolved from a tribal there was little difference between Romans, south- society into a constituent part of the Hellenistic- ern Italian Greeks and non-Greek groups in Salen- Roman world as a result of a combination of in- to. Like Romans, Latins and Campanians, many

126 south Italians and especially the elites of this area and Yntema 2001; for contacts with , Epirus, th rd ‘were interested in participating in the cultural Macedonia in the 4 /early 3 centuries, see D’Andria 1988. The direct or indirect contact with 5th/early 4th cen- koinè’ of the Mediterranean of Hellenistic times tury are indicated by a substantial body (Curti et al. 1996, 182). It is in particular the jux- of Metapontine wares (Metapontine banded and black- taposition of the ancient written sources (ancient figured, Lucanian red-figured wares) on Salento sites authors, inscriptions) and the archaeological evi- (e.g. Yntema 2001, chapters C, H and I). 4 Cf. Florus, Epitome 1.15: Sallentini Picentibus additi caput- dence (artifacts, settlement patterns etc.) that helps que regionis Brundisium inclito portu M. Atilio duce; Titi Livi us to understand how the initial stages of Romani- libri XV periocha: et Sallentini victi in deditionem zation were experienced and perceived by the accepti sunt. people who actually took part in the process. The 5 The geographer Strabo (Geography 6.3.1) is fairly clear evidence primarily concerns the elite, but we can in his definitions of the local groups. He states that the district of Iapygia bordered on the Metaponto area; the also make an educated guess about how the first Greeks called this district ‘Messapia’. But the local pop- stage of ‘becoming Roman’ felt for families that ulation living around Cape Iapygia (now Cape Leuca) did not stem from the former tribal elites. Each were the ‘Sallentinoi’, and the other group the ‘Kalabroi’. group and each individual made its choices and 6 It was not until Byzantine times that the name ‘Calabria’ was applied to the western peninsula of southern Italy. devised strategies that oscillated between resis- 7 Cf. the substantial differences between the Gallic tribes tance (widespread, it seems, among the former reported in Caesar’s De Bello Gallico and those found in tribal farmers), acceptance (by many others) and the same area in the 1st century AD. active involvement in Roman political schemes 8 The Iapygians were usually called the ‘’ in early (e.g. Rammius/Rennius of Brindisi). The result Roman times. 9 For Brindisi in the Bronze Age, see Cinquepalmi/ was that every member of the Salento societies Radina 1998, 167-208 (site of Punta Le Terrare). In the who was a contemporary of Ennius of Rudiae - be Iron Age Brindisi may have held a comparable position it farmer or prince, craftsmen or slave - was forced to that of Otranto. As a dispersed settlement that was to make choices, change attitudes and adapt to both a port of call and a trading port in the early Iron Age (8th to 6th century BC), it may have developed into new stimuli from both inside and outside the a more or less nucleated, walled centre in the 5th to 4th local societies of Salento. Over the course of time, centuries BC. There is no overview which systemati- the inhabitants of Salento began to speak Latin cally collects and interprets the pre-Roman and Roman and to live under Roman laws, while settlements finds from the Brindisi town area. For basic informa- like Brundisium, Lupiae and Canusium were trans- tion on the finds, see Quilici/Quilici Gigli 1975, 65-76. 10 Cf. Lombardo 1992, text nos 188, 205, 400, 429, 444, 458, formed into Roman towns. The tribal past was 460, 476. nd not easily buried, however. As late as the 2 cen- 11 The inscriptions in the Grotta della Poesia date from tury AD, the emperor Marcus Aurelius proclaimed archaic times (6th-5th century BC) to the 1st century BC. - and was obviously proud to do so - that For short preliminary reports, see Pagliara 1987 and 1989. 12 The Romans were followed by the Goths, Lombards, Messapian King Malennius, the mythical founder Arabs, Byzantines, the Staufians from Germany, the of Lupiae (Lecce), was one of his forebears. French Angiovins and the Spanish Habsburg/Bourbon monarchy. After the Risorgimento in Italy in 1870, many NOTES southern Italians felt dominated by Rome and northern Italy. 13 1 I wish to thank my collegues and friends of both Lecce On the ancient roads of Salento, see also B. Fedele, Gli University and the Superintendency at Taranto (Giu- insediamenti preclassici lungo la Via Appia antica in seppe Andreassi, Assunta Cocciaro, Francesco D’Andria, Puglia, ArchStorPugl 19 (1966), 29-43; Uggeri 1975; M.D. Liliana Giardino, Mario Lombardo, Graziella Maruggi Miroslav Marin, La viabilità antica tra Taranto e Brindisi. and Grazia Semeraro) for giving help and advice and La Via Appia, ArchStorPugl 39 (1986) 27-68; M. Lombar- confronting me with their views and ideas on the sub- do, La via istmica Taranto-Brindisi in età arcaica e clas- ject of this paper. The drawings were made by Bert : problemi storici, in Salento Porta d’Italia. Convegno internazionale Lecce 1986, 1989, 167-192. Brouwenstijn (Archaeological Institute Free University), 14 the English of the text was corrected by Annette Visser Varro, De Re Rustica 2.6.5: Greges fiunt fere mercatorum, Translations, Wellington (NZ). ut eorum qui e Brundisino aut Apulia asellis dossuariis com- 2 This name was used by Greeks (e.g. Herodotus 7.170). portant ad mare oleum, vinum itemque frumentum aut quid aliquid. It is uncertain whether they saw themselves as a polit- 15 ical unit and used the same name themselves. For the The quote is from the Messapian-Roman poet Ennius sake of convenience, the term ‘Messapians’ will be used (translation: Warmington, Annales, apud Gellius, Noctes here to denote the native group or groups inhabiting Atticae 7.6.2: Brundisium ... pulcro praecinctum praepete the Salento peninsula. portu; see Lombardo 1992, text 63). He must have 3 For Iron-Age contacts with Albania and the Corinthian known the town well since he came from Rudiae (45 km Gulf, see inter alia D’Andria 1985 and 1995; for imported south of Brindisi) and his sister had married someone from Brindisi (see section 3.4 below). Greek pottery of archaic times, see Semeraro 1997; for 16 6th-century Daunian wares in Salento, see D’Andria 1983 Cf. Lombardo 1992, 39 (comment on text 66) and 91 (comment on text 165); for passages on Brindisi’s role

127 in peaceful shipping, see Lombardo 1992, 53 (comment sapic, which belongs to the Illyrian branch of the Indo- on text 88). European languages (De Simone 1964). Either Gellius 17 If we follow Livy, the Messapian settlements of made a mistake or Ennius knew both Messapic and Manduria and possibly Uzentum or Uxentum (Ugento in Oscan, the latter being the lingua franca in large parts southern Salento) were among these ignobiles urbes. of central and southern Italy. Ugento may have defected to Hannibal after the battle 27 Silius Italicus, Punica 12.393: antiqua Messapi ab origine of Cannae (Livy 22.61,11-12), but the text is fairly cor- regis. rupted (see Ph. Désy, Su alcuni problemi relativi al con- 28 Livy 42.17.2: Princeps Brundisii Rammius fuit; hospitioque flitto tra Annibale e Roma, in L’età annibalica e la Puglia, et duces Romanos omnes et legatos, exterrarum quoque gen- Atti del II Convegno di Studi sulla Puglia romana, Mesagne tium insignes, praecipue regios, accipiebat. 24-26 marzo 1988, Mesagne 1988, 207-208); Manduria 29 For Gaius Rennius Brundisinus, son of Dazoupos, see (some 30 km southeast of Taranto) was retaken by the Griechische Dial. Inschrifte II.1, 1339. Roman consul Quintus Fabius in 209 BC (Livy 27.15.4). 30 It should be noted that Marcus Aurelius believed that 18 Cornfields can be burned, but vineyards and especially the Messapian king Malemnius or Malennius (again an olive groves are difficult to destroy. Thus the damage -ennius name), founder of the town Lecce (ancient may have been limited to burnt farmsteads and the loss Lupiae) and son of Dasummius (again a Daz-name), of two or three years of production; cf. Patterson 1987. was one of his ancestors. Cf. Sirago 1993, 241-242 (notes 19 For Salento wines, see A. Tchernia, Le vin de l’Italie 35 and 36): (Marcus Aurelius) ...probatur sanguinem tra- romaine, Rome 1986. here ... a rege Malennio, Dasummi filio, qui Lupias (Lecce) 20 Livy 29.28.9: Magnus motus servilis eo anno in Apulia fuit. condidit. Daz-names also appear on the opposite shore Tarentum provinciam L. Postumius praetor habebat. Is de of the Adriatic and seem to be characteristic of the lan- pastorum coniuratione, qui vias latrociniis pascuaque pub- guages of the Illyrian group. For such Daz- names, see lica infesta habuerant, quaestionem severe exercuit (Lom- especially O. Parangeli, Studi Messapici, Milan 1960, bardo 1992, text 161). 293-300. Since one of these Daz- names (Dasus) appears Livy 29.41.6: Et L. Postumius praetor, qui Tarentum provi- in a stamp of a potter/slave on the Brindisi amphorae cia evenerat, magnas pastorum coniurationes vindicavit, et of the later 2nd and 1st centuries BC (Désy 1989), we reliquias Bacchanalium quaestionis cum cura exsecutus est may well have some doubts about the elite connotation (Lombardo 1992, text 162). of that name. ‘Dasus’ is more likely to be the Messa- 21 The wording of both passages is so suspiciously simi- pian/Illyrian equivalent of the English name ‘John’ and lar that it is tempting to believe that Livy was recount- may well have denoted both Messapian noblemen and ing the same events twice. commoners. 22 Livy 40.19.9: L. Duronio praetori, cui provincia Apulia 31 For probably elite-owned farmsteads in Salento, see evenerat, adiecta de Bacchanalibus quaestio est, cuius residua Yntema 1993a, 190-191. quaedam velut semina ex prioribus malis iam priore anno 32 The Greeks used the terms Rhomaioi and Italioi fairly apparuerunt ... Id persecare novum praetorem, ne serperet indiscriminately (Zaleskij 1982). iterum latius, patres iusserunt (cf. Lombardo 1992, text 33 These were C. and M. Fabius Hadrianus; the former was 164). a praetor in Africa in the 80s BC, the latter a legatus dur- 23 In the Roman context, the term ‘Apulia’ is almost invari- ing Lucullus’ campaigns in Asia between 72 and 68 BC. ably used for central and northern Apulia (approxi- 34 Magna Grecia in età romana, Atti del XV Convegno di Studi mately the present-day Italian provinces of and sulla Magna Grecia, Taranto 1975, Naples 1976. Bari), while Calabria is used for the present-day Salento 35 Important milestones were the excavation of a Roman peninsula (the modern Italian provinces of Taranto, necropolis at Brindisi and Otranto (Andreassi/Cocchiaro Brindisi and Lecce). Livy seems to have been quite con- 1988: necropoli dei Cappuccini; Excavations at Otranto, sistent in the correct geographical use of these terms. vols. I and II (eds. F. D’Andria and D. Whitehouse), 24 For an unusual case in which a gens with recognizably Galatina 1992, Congedo Editore), the excavation of the Messapian origins (gens Tutoria) appears to own a series kiln sites Apani and Giancola north of Brindisi (Mana- of Roman villas in the area north of Lecce as late as the corda 1988 and 1989: Palazzo 1988, 1989, 2001), the pub- 2nd century AD, see Susini 1962, inscription no 129; lication of several Roman features in the towns of Marangio 1975, 129; Pagliara 1980, 211 ff. The name of Taranto (e.g. Lippolis 1981 and 1984), the research on the same gens may survive in the name of the village Roman Canosa (e.g. Cassano 1992), and Venosa (e.g. of Tuturano ([fundus] Tutorianus), just south of Brindisi. Salvatore 1984), the publication of the Roman mutatio 25 Cf. the ‘treason’ of Dasius Altinius from the north- of Valesio (Boersma et al. 1995), research into the urban Apulian settlement of Arpi during the Hannibalic war archaeology of Lecce (D’Andria et al. 1999, and the ex- (Livy 24.45) or the assistance given to the Roman sur- hibition catalogue Lecce. Frammenti di storia urbana. Tesori vivors of the Cannae slaughter of 216 BC by Busa, mulier archeologici sotto la Banca d’Italia, Bari 2000, Edipuglia). apula from the central-Apulian settlement of Canusium 36 For a first report see Burgers et al. 2004. For Cato’s (Livy 22.52.7). descriptions of such farmsteads, see Cato, De Agricultura 26 Gellius, Noctes Atticae 17.17.1: ... tria corda sese habere 11.1. dicebat, quod loqui Graece et Osce et Latine sciret ... Oscan 37 There was a remarkable shift in burial practices in the was, of course, the lingua franca in large parts of Italy former native districts of Salento sometime around the during the 2nd century BC. But if Gellius meant to sug- mid-2nd century BC. The standard Messapian burial of gest that Oscan was Ennius’ native language, he was the earlier to mid-3rd century BC was an inhumation mistaken. While variants of Oscan - closely related to with a large quantity of pottery in a tomb consisting Latin - were spoken in large parts of central and south- mostly of limestone slabs. While there is little evidence ern Italy (e.g. the Umbro-Sabellic districts, Lucania and for late 3rd- early 2nd-century graves in Salento, crema- Bruttium), the pre-Roman, native language of the pre- tions in completely new cemeteries were practised from sent-day region of Apulia (Salento included) was Mes- about 150 BC onward. A special study on this subject has

128 been dedicated to Lecce (ancient Lupiae): see Giardino Gnathia wares. These were believed to have been pro- 1994. duced until about the end of the 4th century and ca 270 38 For Graeco-Italic amphorae, see inter alia Lyding Will BC respectively (cf. Trendall/Cambitoglou 1982 and 1982 and Vandermersch 1994; the Graeco-Italic amphorae Green 1968 and 1976). Gnathia wares can now be shown were probably produced in many settlements in the to have been produced until the final quarter of the 3rd coastal areas of southern Italy and Sicily. There is no century. On this subject, see Yntema 1990, Lippolis overall study of this type of amphora in which the var- 1994, and Gräpler 1997, 58-60. ious fabrics are identified. 50 It should be noted in this context that rural burial grounds 39 At the site of masseria Ramanno near present-day San of the Roman-Imperial age are almost exclusively known Cataldo (Lecce) close to the Adriatic Sea (see Valchera/ by their inscriptions (exceptions: Brindisi and Lecce; cf. Zampolini Faustini 1997, 152-154: amphora stamps Andreassi/Cocchiaro 1988 and Giardino 1994). Almost with the names DEMETRI, STABUA/STABUAS, LY- nothing is known of the burials themselves. The inscrip- KAON, DIOCLEI, LIBO; these same names are all tions stem from both urban contexts and Roman villa found at the site of Giancola north of Brindisi, i.e. some sites. 50 km north of San Cataldo-Ramanno); a second rela- 51 Since these three tombs are basically unpublished, only tively small amphora production site was traced at a superficial analysis can be offered here. The two Oria Felline in the southern tip of Salento (see Pagliara 1968 tombs were found in 1973 in the northwestern outskirts and Désy 1989: amphora stamps with the names ARIS- of Oria (Ciriaco-Maddalena area) and were believed to TIDES, EROS PVLLI and FELIX PVLLI; not encoun- have contained two depositions each (one deposition tered elsewhere). from the first half of the 3rd century BC and one from 40 For pre-Roman hamlets in northern Salento, see Yntema the mid-2nd century BC; cf. Lo Porto 1974, 343-344, pl. 1993a, sites 6.1 (Santa Cecilia) and 7.35 (Campo Adriano) LVI). The latter observation appears to be due to the and Burgers 1998, 161-169 (Cellino San Marco, masseria incorrect dating of some of the objects found in these Mea). burials. The 1988 tomb from Mesagne is exhibited in the 41 These walls mostly date to the late 4th or early 3rd cen- Mesagne museum (now in the local castle); a short report tury BC; their reconstructed height is 5 to 6 m, while on this burial is published by A. Cocchiaro (1988a); a their width is approximately 5 m. booklet produced for the local market offers more 42 Another, fairly short-lived settlement of this early walled detailed information (Cocchiaro 1988b). The spectacular type is , a few kilometres southeast of Lecce Mesagne tomb discovered slightly south of the Mesagne which had a 5 m-thick wall circuit of more than 3 km Castello in 1999 is still unpublished (personal commu- length. The walls were contructed in the late 6th or early nication by Assunta Cocchiaro). 5th centuries; the settlement came to an end or was 52 Tomb 1 (1973) from Oria contained a Rhodian wine severely restricted in the course of the 5th century BC amphora; tomb 3 (1973) contained 4 amphorae includ- (cf. F. D’Andria, Cavallino (Lecce): ceramica ed ele- ing Cnidian or Cretan amphorae. menti architettonici arcaici, MEFRA 89 (1977) 525-562, 53 For earlier specimens, see inter alia F.G. Lo Porto, Due D’Andria 1988, and O. Pancrazzi, Cavallino, Congedo tombe scoperte a Cavallino, Studi di Antichità 7 (1994) Editore, Galatina 1979). 47-84). 43 The evidence comes from the ‘urban’ surveys of the 54 For elite graves (hypogaea) with amphorae and/or glass walled sites of Valesio (Yntema 1993b), Muro Tenente, vessels from Canosa, see Ciancio 1980 and Ori di Muro Maurizio and San Pancrazio Salentino (Burgers Taranto, 446-452; Stern/Schlick-Nolte 1994, 97-115; for 1998) and from the excavations at Muro Tenente (e.g. similar tombs from Ascoli Satriano, see Tinè Bertocchi Burgers/Yntema 1999). 1985 and Volpe 1987 and Mazzei 1991; for scant remains 44 For the Canosa temple of the 2nd century BC (underneath of a tomb with glass vessel from Salapia, see Mazzei the late Roman basilica of San Leucio), see Principi, 1991; for elite burials with amphorae from the site of Imperatori, Vescovi, 620-654 and Dally 2000). Arpi, see Mazzei 1995 and Mazzei et al. 1996. For tombs 45 Cf. Sirago 1993, 242; Eutropius 8.9: Post hunc imperavit with very similar characteristics from Ancona in the Marcus Antoninus Verus, haud dubie nobilissimus, quippe Marche area, see Mercando 1976 and Colivicchi 2002; cum eius origo paterna a Numa Pompilio, materna a Salen- these contain objects made of precious metals, fayence, tino rege et cum eo Lucius Annius Verus. SHA, Vita Marci Micro-Asiatic white-slipped wares, relief decorated Antonini philosophi 1.6: (Marcus) cuius familia in originem ‘Megarian’ bowls from Greece and oil and wine ampho- recurrens a Numa probatur sanguinem trahere, ut Marius rae (e.g. from Brindisi, Cnidos and Rhodos). Maximus docet; item a rege Malemnio, Dasummi filio, qui 55 It should, however, be noted that the Canosan funerary Lupias condidit. culture retained highly traditional features up until the 46 The Lecce theatre contained, for instance, copies of the 2nd century BC (cf. D.G. Yntema, The Listata Wares from Augustus of Primaporta, the Athena Giustiniani, the Canosa, OudhMeded 77 (1997) 121-134). Gabii Artemis and the Doryphorus by Polycletus; see 56 The only exception reported hitherto is a rural necrop- K. Mannino, Il teatro: la decorazione della scena, in olis probably belonging to a hamlet or larger farmstead D’Andria 1999. dating to the later 3rd and perhaps the early 2nd century 47 It should be noted, however, that Strabo (6.3.5) does not BC at the site of San Lorenzo near the present-day town mention Lupiae as one of the ‘real’ towns of Salento. He of (cf. Marinazzo 1980). may have ranked it among the polismata of Salento. 57 As for the Vehilii, it should be noted that a Vehilius was 48 For a Messapian sanctuary that continued to function buried in Lecce as early as the first half to mid-2nd cen- until well into the Roman Imperial period, see tury BC in a traditional Messapian cist grave (cf. Susini D’Andria et al. 1978 (Grotta Porcinara near Santa Maria 1962, 105; inscription in Latin letters). Thus there may di Leuca). be reason to doubt the central-Italian origin of the gens 49 The datings were based on the (style-based) chronolo- Vehilia. The gentilicium of the Anninii is paralleled by gies of Apulian red-figured pottery and especially similar-sounding gentilicia of families that definitely

129 had roots in Salento (the Ennii, Rennii, Pulfennii from 70 See, for instance Pausanias 10.10.6-8 (Lombardo 1992, Brindisi, the Annii from Canosa, etc.). text 279) on a Tarantine anathema at Delphi, and a pas- 58 Manacorda’s view that the owners of the workshops also sage by Aristoxenos cited in Iamblychus’ Life of Pyth- owned large villas (Manacorda 1988, 1990) has been agoras, 31.197 (Lombardo 1992, text 51). convincingly refuted by Désy (1989). 71 Brindisi even started to mint coins according to the 59 The interpretations proposed by Manacorda and Roman standard in this particular period (cf. Boersma/ Palazzo for the amphora production of Brindisi and Prins 1994); the mint at Oria was operational in the 2nd surrounding area have been strongly influenced by century BC (Travaglini 1990). images constructed for the organization of the produc- 72 The second and third Macedonian War (200-197 and tion of Sestius amphorae of the Cosa area (see, e.g. 171-168) and wars against the Aetolians and the Achaean Manacorda 1981, 1988, recently Cambi 2000). The stamps League (respectively 192-189 and 149-146). indicate that the workshops were relatively small and 73 Livy (39.29.8) states that some 7,000 men were con- suggest that Brindisi kilns were often shared by a group demned because of the revolt. of workshops; they were probably not the prerogative 74 The population decline in these settlements may have of a single, large amphora production unit (Désy 1989). been somewhat less disastrous than the reduced size of 60 For Hellenistic Greek amphorae in southern Italy, see the settlement areas suggests. We should comment that Cento Anni, 60-80 (La collezione Viola); Volpe 1986; the greater social differences in many of the Salento set- Lippolis 1997, 23-28. tlements may have meant that part of the population 61 Amphorae of the Punic type are widely diffused in ceased using the highly diagnostic fine wares, thus southern Italy and stem from contexts covering a fairly becoming virtually invisible. But the severe reduction long period. Therefore they cannot be empty containers in the size of settlement areas is probably real. There is left behind by Hannibal’s troops. A series of specimens a highly diagnostic set of coarse wares in Salento, dat- from Basilicata has been collected by G. Greco (Rivista ing to the close of the 3rd and the first half of the 2nd di Studi Liguri 45 (1979) 13-17). Clay analyses of the century BC (Yntema 2001); fragments of these coarse specimens from two Salento sites (Valesio contacts of wares were also found in a small, mostly central area the early 2nd century; Muro Tenente, 2nd to 3rd quarter of the formerly much larger settlements. This does not of the 3rd century context) suggest a Sicilian origin for rule out the fact that the ‘new poor’ of the 2nd century these amphorae de tradition punique. BC may have lived closer together and that a 75-80% 62 For Campana A wares from Salento, see D’Andria et al. reduction in the settlement area was not linked to a 1978, 123-126 (from S. Maria di Leuca) and Yntema comparable percentage decrease in population. 2001, 329-331 (from Valesio). 75 Perhaps a few smaller Salento towns of the Roman 63 For distribution of Gnathia in general, see Green 1977; period should be added to this list: the coastal towns for Sidi Khrebish/Benghazi, see Kenrick 1985; for of Otranto and Callipolis/Anxa and the former Messa- Dalmatia, see, for instance, Kirigin/Marin 1988; Gnathia pian tribal centres of Oria (Brindisi area) and Ugento wares are obviously rare in (Rotroff 1997) and (southern Salento). Mesagne (Brindisi area) is likely to in south-Albanian Bouthroton (only four pieces in the have been yet another modest town-like settlement of large Hellenistic sample that I analyzed in July 2000). the Roman period. 64 See Kenrick 1985, 37-43 (items B33-B50). 76 At a relatively early stage in Salento history, tribal chief- 65 Such studies have been conducted within the context tains appear to have displayed symbols of Greek paideia of the Valesio excavations (Yntema 2001) and are now in their graves (e.g. A. Rouveret, Les oiseaux d’Ugento, placed in an interregional perspective (B.M.J. Mater, in Mélanges offerts à Jaques Heurgon, Rome 1976, 927-945: Patterns in Pottery. A comparative study of pottery produc- 5th/early 4th-century tomb from Ugento). Ennius’ adap- tion in Salento, Sibaritide and Agro Pontino in the context tations of Greek tragedies display a perfect knowledge of urbanization and colonization in the first millennium BC, of both Greek and Latin and are, like other of Ennius’ PhD Amsterdam Free University, 2005). literary works that were highly dependent on Greek 66 For imported ‘eastern’ fine wares of the late 3rd and ear- examples, indicative of a widely perceived dominance lier 2nd centuries BC, see Yntema 1995c. For ‘eastern’ of Greek cultural models in peninsular Italy. shapes in regionally produced fine wares, see Yntema 77 An inscription with a decree of the sanctuary and peo- 1990 (HFR Group) and Yntema 2005. In addition to the ple, dated between 241 and 232, honored Bouzos, son mastos and lagynos, there are amphorae of West Slope of Orteiros, from Canosa with a proxeny calling him an shape and calotte-shaped bowls in Apulian Black Gloss euergetes (IG XI.4, 642); Livy (22.54) also mentions the wares. rich Busa, mulier apula from Canosa, giving generous 67 Large-scale settlement excavations have been carried help to the Roman survivors of the battle of Cannae. out at the ‘Messapian’ oppida of Vaste, , 78 Cf. Polybius 2.11.7 (Lombardo 1992, text 66). Valesio and Muro Tenente. For Vaste, see Archeologia dei 79 Silius Italicus’ Punica (12.393-395) is the only (and not Messapi, 48-189, D’Andria 1996, 427-438; for Muro Leccese, very reliable) source for Messapians participating in the see Lamboley 1999; for Valesio, see, e.g. Boersma/ battle of Cannae. Yntema 1987; Boersma/Yntema 1989; Yntema 2001. 80 New vineyards and olive groves had to be planted (it 68 These settlements are Taranto, Monte Salete, Vicentino, takes 20 years for an olive tree to bear fruit), and stor- Oria, Muro Tenente, Mesagne, Muro Maurizio, San age facilities had to be expanded. The farms had to be Pancrazio, Valesio, Brindisi (fig. 1). enlarged considerably in order to process and store the 69 The settlement system of southern Salento in Roman products that were no longer consumed in the nearby, times is not yet clear. Lecce (Lupiae), with its theatre and former urban settlement. Olive presses, wine presses amphitheatre, was evidently the main regional centre. and millstones found during the Oria survey at both Both Otranto (Hydruntum) and Gallipoli (Callipolis or 3rd-and 2nd-century BC sites indicate that olive oil, grapes Anxa) must have been ports and were probably Roman and corn were processed at the farmstead (Yntema towns of secondary importance. 1993a).

130 81 Thus the harsh measures taken by the Roman (pro)prae- Cambi, F. 2001, Calabria romana. Paesaggi tardo-repub- tores of the Taranto/Apulia area greatly benefited the blicani nel territorio brindisino, in E. Lo Cascio/A. regional landowning elites (see section 3.3: Uprisings/ Storchi Marino eds., Modalità insediative e strutture resistance). agrarie nell’Italia meridionale in età romana, Bari, 363-390. 82 Spindle whorls and loom weights are extremely com- Carito, G. 1988, L’urbanistica di Brindisi in età repubblicana, mon in the settlement contexts of Salento of the 4th and in La Puglia in età repubblicana. Atti del I Convegno di Studi 3rd centuries BC. In a few cases the latter even occur in sulla Puglia romana, Mesagne 1986, Galatina, 173-179. substantial clusters of 40 to 80 specimens, suggesting the Cassano, R. 1992, Il tempio di Giove Toro, in Principi, impe- presence of a loom or other type of weaving implement ratori, vescovi. Duemila anni di storia a Canosa. Exhibition in that room. catalogue Bari, S. Scolastica, Venice, 741-758. 83 Ennius became a Roman citizen in 184 BC and was Cassola, F. 1971, Romani e Italici in Oriente, DialA 4-5, 305- probably granted a piece of land in the colonia of Poten- 329. tia in Picenum, just south of Ancona. Ciaceri, E. 1924-32, Storia della Magna Grecia, vols. I-III. 84 Cf. Cicero, Brutus 20.79 and Livy 39.44.10. Milan/Rome. 85 The same may hold for Lecce (Lupiae) in southern Salento. Ciancio, A. 1980, I vetri alessandrini rinvenuti a Canosa, in 86 Pliny (NH 33.45.130), for instance, tells us that Brindisi A. Riccardi/A.Ciancio/M. Chelotti (eds.), Canosa I, Bari. mirrors made of bronze and tin were the best available Cinquepalmi, A/F. Radina 1998, Documenti dell’età dell until a certain Pasiteles started to make silver mirrors bronzo. Ricerche lungi il versante adriatico pugliese, Fasano. in the time of Pompeius Magnus (Lombardo 1992, text Cipriano, M.T./M.-B. Carre 1989, Production et typologie 238). des amphores sur la côte adriatique de l’Italie, in 87 In the isthmus area between Brindisi and Taranto, for Amphores romaines et histoire économique, dix ans de instance, there are seven or eight ‘native’ sites which recherche, Rome, 67-104. probably all survived into the 2nd century BC (Burgers Clemente, G. 1988, Introduzione alla storia romana della 1998), while only three toponyms are known from the Puglia romana, in La Puglia in età repubblicana. Atti del I ancient written sources. Greek Hyrie (Roman Uria Cala- Convegno di Studi sulla Puglia romana, Mesagne 1986, bra) can be identified with the important site of Oria, Galatina, 11-20. but the identification of the settlements of Mesochorum Cocchiaro, A. 1989, Nuovi documenti dalla necropoli meridio- (often believed to be the site of Masseria Vicentino) and nale di Mesagne, Fasano. Scamnum (often identified with the site of Muro Tenente) Colivicchi, F. 2002, La necropoli di Ancona (IV-I sec. A.C). is doubtful. Una comunità italica fra ellenismo e romanizzazione, Napoli. 88 On Roman stereotypes for conquered tribal societies Curti, E./E. Dench/J.R. Patterson 1996, The Archaeology (Romans as bringers of peace, prosperity and culture), of Central and Southern Roman Italy: Recent Trends see inter alia Dench 1995 and the introductory part of and Approaches, JRS 86, 170-189. section 3 above. Curti, E. 2001, Toynbee’s Legacy: discussing aspects of the Romanization in Italy, in S. Keay/N. Terrenato (eds.), BIBLIOGRAPHY Italy and the West. Comparative Issues in Romanization, Oxford, 17-27. Andreassi, G./A. Cocchiaro 1988, La necropoli di Via Dally, O. 2000, Canosa, Località San Leucio. Untersuchungen Cappuccini a Brindisi, Fasano. zu Akkulturationsprozessen vom 6. bis zum 2. Jh.v.Chr. am Belotti, B. 1997, Un example de prospection systématique Beispiel eines daunischen Heiligtums, Heidelberg. au sol: histoire de la ville de Vaste et de son territoire, D’Andria, F. 1978, Leuca, Galatina. in F. D’Andria (ed.) Metodologie di catalogazione dei beni D’Andria, F. 1979, Puglia in età romana, in Puglia dal archeologici. Beni archeologici - Conoscenza e tecnologie, Paleolitico al Tardoromano, Milan, 273-360. Quaderno I. 1, Lecce/Bari, 135-166. D’Andria, F. 1983, Nuovi dati sulle relazioni tra Daunia e Boersma, J.S./J. Prins, 1994, Valesio and the mint of Messapia, in Studi in onore di Dinu Adamesteanu, Galatina, Brindisi, Studi di Antichità Lecce 7, 303-325. 41-47. Bradley, G. 2002, The Romanization of Italy in the 2nd c. D’Andria, F. 1985, Documenti del commercio arcaico tra BC, JRA 15, 401-406. Ionio e Adriatic, in Atti XXIV Convegno Int. di Studi sulla Brunt, P.A. 1971, Italian Manpower, 225 B.C.-A.D. 14, Oxford. Magna Grecia, Taranto 1984, Taranto, 321-377. Burgers, G.-J. 1998, Constructing Messapian Landscapes. Set- D’Andria, F. 1988, Messapi e Peuceti, in Italia, omnium ter- tlement Dynamics, Social Organization and Culture Contact, rarum alumna: le civiltà dei Veneti, Reti, Liguri, Celti, Amsterdam. Piceni, Umbri, Latini, Campani e Iapigi, Milan, 653-715. Burgers, G.-J. 2001, L’archeologia e l’Italia meridionale D’Andria, F. 1990, Oria - Monte Papalucio, in Archeologia post-annibalica: una prospettiva regionale e diacronica, dei Messapi. Exhibition catalogue Museo Provinciale in E. Lo Cascio/A. Storchi Marino (eds.), Modalità inse- Lecce, Bari, 239-306. diative e strutture agrarie nell’Italia meridionale in età D’Andria, F. 1995, Corinto e l’Occidente: La costa adriat- romana, Bari, 249-266. ica, in Corinto e l’Occidente, Atti del XXXIV Convegno di Burgers, G.-J./D.G. Yntema, 1999, The Settlement of Muro Studi sulla Magna Grecia 1994, Taranto, 457-508. Tenente, Southern Italy. Third Interim Report, BABesch D’Andria, F. 1996, La casa in Messapia, in F. D’Andria/K. 74, 111-132. Mannino (eds.), Ricerche sulla casa in Magna Grecia e Burgers, G.-J./P.A.J. Attema/M. van Leusen 2004, Walking Sicilia, Atti Colloquio Lecce 1992, Galatina, 403-438. the Murge: interim report of the Ostuni Field Survey D’Andria F. (ed. 1999), Lecce romana e il suo teatro, Galatina. (Apulia, southern Italy), Studi di Antichità Lecce 11, 257- De Neeve, P.W. 1984, Peasants in Peril. Location and Economy 282. in Italy in the Second Century BC, Amsterdam. Calderone, S. 1975, La conquista della Magna Grecia, in La De Simone, C. 1964, Die messapischen Inschriften und ihre Magna Grecia nell’età romana, Atti del XV Convegno Int. di Chronologie, in H. Krahe, Die Sprache der Illyrier II, Studi sulla Magna Grecia, Taranto 1975, Napels 1976, 33-81. Wiesbaden, 1-157.

131 Désy, Ph. 1989, Les timbres amphoriques de l’Apulie républicaine. Lippolis. E. 1997, Fra Taranto e Roma: Societa e cultura urbana Documents pour une histoire économique et sociale, Oxford. in Puglia tra Annibale e l’età imperiale, Taranto. Désy, Ph./P. De Paepe 1990, Torre San Giovanni (Ugento): Lippolis, E./I. Baldini Lippolis 1997, La formazione e lo les amphores commerciales hellénistiques et républi- sviluppo del centro urbano di Brundisium: aspetti e caines, Studi di Antichità Lecce 6, 187-234. problemi di ricerca, Taras 17, 305-353. Donvito, A. 1988, Egnazia dalle origini alla riscoperta archeo- Lomas, K. 1993, Rome and the Western Greeks, London/New logica, Fasano. York. Finley, M.I. 1986, Ancient History: Evidence and Models, New Lomas, K. 1995, Urban elites and cultural definition: Ro- York. manization in southern Italy, in T.J. Cornell/K. Lomas Gabba, E./M. Pasquinucci 1979, Strutture agrarie e alleva- (eds.), Urban Society in Roman Italy, London, 107-120. mento transumante nell’Italia romana (III-I a.C.), Pisa. Lomas, K. 1996, Greeks, Romans, and Others. Problems of Gabba, E. 1989, Rome and Italy in the Second Century BC, colonialism and etnicity in southern Italy, in J. Webster/ in The Cambridge Ancient History, 2nd edition vol. VIII. N. Cooper (eds.), Roman Imperialism: post-colonial per- Rome and the Mediterranean to 133 BC, Cambridge, 197- spectives, Leicester, 135-144. 243. Lombardo, M. 1992, I Messapi e la Messapia nelle fonti let- Gabba, E./G. Benetti/F. Grelle 2001, Hannibal’s Legacy. terarie greche e latine, Galatina. Trent’anni dopo, in E. Lo Cascio/A. Storchi Marino Lo Porto, F.G. 1974, L’attività archeologica in Puglia, in Or- (eds.), Modalità insediative e strutture agrarie nell’Italia fismo in Magna Grecia, Atti del XIV Convegno Int. di Studi meridionale in età romana, Bari, 13-32. sulla Magna Grecia, Taranto 1974, Naples 1975, 337-350. Ghinatti, U. 1977, Magna Grecia post-annibalica, Quaderni Lyding Will, E. 1982, Greco-Italic Amphoras, Hesperia 51, di Storia 6, 99-155. 338-351. Giardino, L. 1994, Per una definizione delle trasformazioni Manacorda, D. 1981, Produzione agricola, produzione urbanistiche di un centro antico attaverso lo studio ceramica e proprietari nell’Ager Cosanus nel I sec. a.C., delle necropoli: il caso di Lupiae (Lecce), Studi di Anti- in A. Giardina/A. Schiavone (eds.), Società romana e pro- chità Lecce 7, 137-203. duzione schiavistica II, Merci, mercati e scambi nel Giardino. L/P. Arthur/G.-P. Ciongoli (eds.) 2000, Lecce. Mediterraneo, Bari, 3-54. Frammenti di storia urbana. Tesori archeologici sotto la Manacorda, D. 1988, Per uno studio dei centri produttori Banca d’Italia. Exhibition catalogue Lecce 2000, Bari. delle anfore brindisine, in Puglia in età repubblicana. Atti Green, J.R. 1968, Some painters of Gnathia Vases, BICS 15, del I Convegno di Studi sulla Puglia romana, Mesagne 1986, 34-50. Galatina, 91-108. Green, J.R. 1971, Gnathia Addenda, BICS 18, 30-38. Manacorda, D. 1990, La fornace di Visellio a Brindisi. Primi Green, J.R. 1976, Gnathia Pottery in the Akademisches Kunst- risultati dello scavo, VeteraChr 27, 375-415. museum Bonn, Mainz. Marangio, C. 1975a, La romanizzazione dell’Ager Brundisi- Green, J.R. 1979, Ears of corn and other offerings, in A. nus, Ricerche e Studi Museo Provinciale Brindisi 8, 105-133. Cambitoglou (ed.), Studies in honour of A.D. Trendall, Marangio. C. 1975b: Ulteriore contributo al Supplemento Sydney, 81-90. del CIL IX, Studi Linguistici Salentini 7, 115-138. Gruen, E.S. 1984, The Hellenistic World and the Coming of Marasco, G. 1988, Economia, commerci e politica nel Mediter- Rome, Berkeley. raneo fra il III e il II secolo a.C., Florence. Hafner, G. 1968, Das Bildnis des Q. Ennius. Studien zur römi- Marinazzo, A. 1980, La necropoli messapica di S. Lorenzo schen Porträtkunst des 2. Jahrhunderts v. Chr., Baden-Baden. (Francavilla Fontana), Fasano di Puglia. Hall, J. 1997, Ethnic identity in Greek antiquity, Cambridge. Mazzei, M. 1991, La ‘tomba delle coppe di vetro’ di Ascoli Hatzfeld, J. 1919, Les trafiquants italiens dans l’Orient hel- Satriano. Nuovi elementi per lo studio della società e del lénique, Paris. commerico della Daunia del II sec. a.C., AnnStorAnt 13, Kahrstedt, U. 1960, Die wirtschaftliche Lage Grossgriechen- 189-208. lands in der Kaiserzeit, Wiesbaden. Mazzei, M. 1995, Arpi, l’ipogeo della Medusa e la necropoli, Bari. Kirigin, B./M. Marin, Excavations at Issa, island of Vis, Mazzei, M. et al. 1996, Il caso di Arpi. Ambiente italico e magno- Jugoslavia 1980, Studi di Antichità Lecce 5, 129-148. greco tra primo e medio ellenismo, Atti Tavola Rotonda, La Bua, V. 1988, Il Salento e i Messapi di fronte al conflitto Foggia 1996, Foggia 1997. tra Annibale e Roma, in L’età annibalica e la Puglia, Atti Mertens, J. 1979, Rapport sommaire sur les travaux des del II Convegno di Studi sulla Puglia romana, Mesagne campagnes de 1975, 1976, et 1977, in J. Mertens (ed.), 1988, Galatina, 43-69. Ordona VI, Bruxelles/Rome, 7-39. Lamboley, J.-L. 1999, Muro Leccese. Sondages sur la fortifica- Mertens, J. 1988, Ordona 1978-1986. Rapport sommaire sur tion nord, Paris/Rome. neuf années de fouilles archéologiques, in J. Mertens Laporta, M.T. 1988, Note sui toponimi in -ano della ‘Calabria’ (ed.), Ordona VIII. Rapports et études, Bruxelles/Rome, romana, in La Puglia in età repubblicana. Atti del I Convegno 7-67. di Studi sulla Puglia romana, Mesagne 1986, Galatina, 233- Morel, J.-P. 1978, La laine de Tarente, Ktema 3, 93-110. 247. Morel, J.-P. 1981, La céramique campanienne: les formes, Lippolis, E. 1981, Alcune considerazioni topografiche su Paris/Rome. Taranto romana, Taras 1.1, 77-114. Ori di Taranto: Gli ori di Taranto in età ellenistica. Exhibition Lippolis, E. 1983, La basilica forense di Egnazia, AnnPerugia catalogue Brera Milan, Milan 1984. 20, 281-321. Pagliara, C. 1968, Bolli anforari inediti di Felline, SCO 17, Lippolis, E. 1984, Le Thermae Pentascinenses di Taranto, 227-231. Taras 4, 119-153. Pagliara, C. 1980, Note di epigrafia salentina, IV, Studi di Lippolis, E. 1994, La necropoli ellenistica: problemi di clas- Antichità Lecce 1, 225-235. sificazione e cronologia dei materiali, in E. Lippolis Pagliara, C. 1987, La Grotta Poesia di Roca (, (ed.), Catalogo Museo Nazionale Archeologico di Taranto Lecce). Nota preliminare, AnnPisa 17, 267-328. III.1, Taranto, 239-281. Pagliara, C. 1986, La costa salentina del Canale d’Otranto.

132 Primi risultati, in Salento, Porto d’Italia: Salento, Porto una carta archeologica della Puglia meridionale, in d’Italia. Convegno Internazionale Lecce 1986, Galatina Metodologie di catalogazione dei beni archeologici. Beni 1989, 121-130. archeologici - Conoscenza e tecnologie, Quaderno I.2, Lecce- Palazzo, P. 1988, Aspetti tipologici della produzione di Bari 1997, 103-158. anfore brindisine, in Puglia in età repubblicana, Atti del I Veenman, F.A. 2002, Reconstructing the Pasture. A recon- Convegno di Studi sulla Puglia romana, Mesagne 1986, struction of pastoral landuse in Italy in the first millennium Galatina, 109-117. BC, PhD thesis, Amsterdam, Vrije Universiteit. Palazzo, P. 1989, Le anfore di Apani (Brindisi), in Amphores Verzar Bass, M. 1983, Contributo alla storia sociale di romaines et histoire économique: dix ans de recherche, Rome, Aquileia repubblicana, in Les ‘bourgeoisies’ municipales 548-553. italiennes aux IIe et Ier siècles av. J.-C., Paris/Naples, 205- Pani, M. 1988, I ‘municipia’ romani, in La Puglia in età 215. repubblicana, Atti del I Convegno di Studi sulla Puglia Volpe, G. 1987, Le anfore della tomba 6 di Ascoli Satriano, romana, Mesagne 1986, Galatina, 21-45. Ricerche e Studi Museo Provinciale Brindisi 13, 105-120. Patterson, J. 1987, Crisis: What Crisis? Rural Change and Volpe, G. 1988, Primi dati sulla circolazione delle anfore Urban Development in Imperial Apenine Italy, BSR 55, repubblicane nella Puglia settentrionale, in La Puglia in 115-146. età repubblicana Atti del I Convegno di studi sulla Puglia Quilici, L./S. Quilici Gigli 1975, Repertorio dei beni culturali romana, Mesagne 1986, Galatina, 77-90. archeologici della provincia di Brindisi, Fasano di Puglia. Volpe, G. 1990, La Daunia nell’età della Romanizzazione, Bari. Ribezzo, F./C. Santoro 1978, Corpus Inscriptionum Messa- Volpe, G. 1998, San Giusto. La villa, le ecclesie, Bari. picarum, Bari. Wilson, A.J.N. 1966, Emigration from Italy in the Republican Salmon, E.T. 1982, The Making of Roman Italy, London. Age, Manchester. Salvatore, M. (ed.) 1984, Venosa: un parco archeologico e un Wiseman, T.P. 1971, New Men in the Roman Senate, 139 BC- museo - come e perché. Exhibition catalogue Venosa, 14 AD, Oxford. Taranto. Yntema, D.G. 1982, Notes on Greek Influence on Iron Age Santoro, C. 1988, La latinizzazzione della regio II: il prob- Salento, Studi di Antichità Lecce 3, 83-131. lema linguistico, in La Puglia in età repubblicana. Atti del Yntema, D.G. 1990, A Specific Group of Apulian Black I Convegno di Studi sulla Puglia romana, Mesagne 1986, Gloss from the Valesio Excavations: the HFR-group and Galatina, 127-166. its connections, Studi di Antichità Lecce 6, 167-186. Semeraro, G. 1997, En Nèusi. Ceramica greca e società nel Yntema, D.G.1993a, In Search of an Ancient Countryside. The Salento arcaico, Lecce/Bari. Amsterdam Free University Field Survey at Oria, Province Silvestrini, M. 1998, Banchieri, mercanti, ‘diplomatici’ schia- of Brindisi, South Italy (1981-1983), Amsterdam. vi tra Puglia, Grecia e Oriente, in R. Cassano (ed.), Yntema, D.G. 1993b, The Settlement of Valesio, Southern Andar per mare. Puglia e Mediterraneo tra mito e storia. Italy, BABesch 68, 49-70. Exhibition catalogue Museo Archeologico Bari and Yntema, D.G. 1995a, Romanisation in the Brindisino, South- Museo Provinciale Brindisi, Bari, 225-232. ern Italy: a Preliminary Report, BABesch 70, 99-123. Sirago, V.A. 1993, Puglia romana, Bari. Yntema, D.G. 1995b, The Artefacts of the Roman Period, Stern, E.M./B. Schlick-Nolte 1994, Early Glass of the Ancient in Joh.S. Boersma (ed.), Mutatio Valentia, The Late Roman world, 1600 BC - AD 50. The Ernesto Wolf Collection, Baths at Valesio, Salento, Amsterdam, 1995, 265-293. Ostfildern. Yntema, D.G. 1995c, Salento and the Eastern Mediter- Susini, G. 1962, Fonti per la storia greca e romana del Salento, ranean in the Middle and Late Hellenistic Period: Some Bologna. ‘Eastern’ Ceramic Evidence (Fine Wares) from Valesio, Terrenato, N. 1998, The Romanization of Italy: global accul- , Studi di Antichità 8.2, 387-404. turation or cultural bricolage, Proceedings of the 7th annual Yntema, D.G. 2001, Pre-Roman Valesio. Excavations of the Theoretical Roman Archaeology Conference 1997, Notting- Amsterdam Free University at Valesio, Province of Brindisi, ham, 20-28. southern Italy, Amsterdam. Tinè Bertocchi, F. 1985, Le necropoli daunie di scoli Satriano Yntema, D.G. 2005, Conspectus of Apulian Grey Gloss wares e Arpi, Genova. (2nd enlarged and updated edition), Amsterdam. Torelli, M. 1971, Contributo dell’archeologia alla storia Yntema, D.G. (in press): Material culture and plural iden- sociale: l’Etruria e l’Apulia. DialA 4-5, 431-442. tity in early Roman Southern Italy, in A. Derks/N. Toynbee, A.J. 1965, Hannibal’s Legacy, Oxford. Roymans (eds.), Ethnic Constructs in Antiquity. The Role Travaglini, A. 1990, La monetazione di ORRA, Studi di of Power and Tradition, Amsterdam. Antichità Lecce 6, 235-255. Zalesskij, N.N. 1982, Les Romains à Délos, in F. Coarelli/ Uggeri, G. 1975, La viabilità preromana della Messapia, H. Solin/D. Musti (eds.), Delo e l’Italia, Opuscula Romana Studi e Ricerche Museo Provinciale Brindisi 8, 75-104. Instituti Romani Finlandiae 2, 21-49. Uggeri, G. 1983, La viabilità romana del Salento, Fasano. Uggeri, G. 1988, Il porto di Brindisi in età repubblicana, in La Puglia in età repubblicana. Atti del I Convegno di Studi INSTITUTE OF ARCHAEOLOGY sulla Puglia romana, Mesagne 1986, Galatina, 47-64. VRIJE UNIVERSITEIT AMSTERDAM Vandermersch, Ch. 1994, Vins et amphores de Grande Grèce DE BOELELAAN 1105 et de Sicile, IVe-IIIe s. avant J.-C., Etudes I, Naples. NL-1081 HV AMSTERDAM Valchera, A./S. Zampolini Faustini 1997, Documenti per e-mail: [email protected]

133