1. General remarks

1In Roccagloriosa I both the survey goals and methodology employed for the nearer territory were outlined. That survey covered intensively, within the limits of the topography and modern land use, an area of ca. 25/30 square kilometers all around the fortified site on the Mt. Capitenali ridge. The proportion of the area covered by intensive survey with respect to the total surface area of the nearer territory (defined as within 4-5 km. of the fortified centre) was extremely high because the survey coincided with a change in the modern land use to deep ploughing which increased the visibility of sites. The verification work carried out during the later phases of the surface survey (1987-91) in the nearer territory have allowed us to add very important data for the settlements dating to the Roman period but in effect have not modified substantially the distribution pattern and density of sites pertaining to the Lucanian period. Such verification on the other hand has insured greater accuracy in the distribution maps and site density for the pre-Roman period which have already been presented and thoroughly discussed in previous studies (Roman Landscapes, 197-201; Velia 1990, 47-54; Gualtieri 1996a).

2As is evident in Roccagloriosa I, fig. 148 (find-spots), the intensive survey of the nearer territory has brought to light a considerable level of ‘background noise’ created by the agricultural activity connected with a typical Mediterranean System of polyculture. It is clear that a noteworthy number of the scatters located by the intensive survey and repeatedly checked in the years 1982-86 do not actually refer to the existence of ancient sites but represent rather the result of more or less intensive agricultural activity, e.g., land fill (‘terreno di riporto’), plough drag, etc. It is always useful to draw attention to the definition of a ‘site’ which remains one of the major problems in interpreting survey data or, more specifically to “the divide between sites and non-sites”, as it has been recently called (Yntema 1993). The remarks of Lloyd (1985, 221) remain fundamental with regard to this problem. Nonetheless, the criterion adopted for the Roccagloriosa area is based not only on the quantitative character of the scatters themselves but also on topographic, geological and environmental factors. The site distribution map presented in fig. 72 is founded on these types of cumulative and varied evidence. It is furthermore now evident that the density of the 4th-3rd centuries B.C. sites located in the nearer territory around the oppidum is even higher than we had identified previously and is, in fact, comparable to the settlement density mapped in flatter, more open landscapes, e.g., that of Messapian Oria, recently published by Yntema (1993).

3The more recent survey covered the wider territory of the entire catchment basin of the Mingardo and Bussento rivers (fig. 71, supra) and was carried out in areas chosen on the basis of various considerations of topography and land use. The extremely rough or uneven nature of the higher Mingardo and Bussento valleys, given the choice of survey areas, permits us to hazard the hypothesis that few if any of the larger sites will have escaped the repeated surveys but it is certainly possible that some of the smaller scatters will have been overlooked. Thus, it is necessary to proceed cautiously when using the distribution maps for the smaller sites in the catchment basin but it is also necessary to bear in mind that the picture is more accurate for the major site distributions of both the 4th and 3rd centuries and for the villae/vici of the Roman period. Concomitantly, it would be much more difficult to delineate a settlement picture for the entire catchment basin in the transitional periods such as the 2nd-1st centuries B.C. and the late antique period.

4The methodological diversity of the surveys carried out in the area also necessitates some general observations regarding the accuracy of the surface retrieval used to reconstruct the settlement pattern evolution in the study area.

5There can be no doubt that the exponential increase in surface surveys conducted in the last two decades has contributed substantially to the reconstruction of the settlement development between the Iron Age and the late antique period for vast areas of the peninsula (Lloyd 1991, 233-234). Thus survey results have become more accurate and more widely comparable on interregional scales as both survey methods and analyses of data become more refined. Recent discussions concerning the methodological aspects of surface surveys have certainly contributed to lessen many of the initial doubts expressed by numerous classical archaeologists about the reliability of such surveys and the validity of historical inferences postulated on the basis of surface survey data, although some of the enthusiasm for such interpretations is often exaggerated (e.g., Snodgrass 1987 is rather moderate but Cherry et al. 1991 is certainly excessive).

6It is not the intent of this brief introductory note to draw attention to the most recent discussions regarding the chance element always present in surface survey data (Ammerman 1993; Cambi/Terrenato 1994, 173) or the historical interpretations based on unreliable or non-representative data. In the specific case of the catchment basin of Roman Buxentum we have already put forward some working hypotheses based on both the surface evidence and important points made by a number of historians (Lepore 1983), which attempt to modify the rather global interpretation (also referred to as the “macro-economic model”) of rural settlement evolution and economic development which Kahrstedt (1960) had postulated for the whole of Magna Graecia in the Imperial period. We believe it necessary, however, to underline two aspects of the archaeological evidence already collected, which in our view allow us to outline some developmental trends in the settlement pattern from the Iron Age through the late Roman period in the study area:

1. the systematic excavation of a major site in the territory under study as well as test trenches excavated for verification in several survey sites have provided, on the one hand, an important chronological framework for the development of a regional centre and a specific settlement typology for the 4th and 3rd centuries B.C. On the other hand, this evidence has confirmed several correspondences (already postulated as working hypotheses) between the surface scatters and the settlement types found in the territory;

2. for the late Republican and Imperial periods, even in the absence of systematic excavations on a large scale, the surface survey data can be put in relation to monumental evidence which, although fragmentary, has provided the foundations for an analysis of the spread of the villa rustica in the study area (Gualtieri 1996a). Such an analysis, additionally supported by literary and epigraphical evidence pertaining to the colonial settlement at Buxentum, provides an extremely useful and valid basis for a more articulated discussion regarding the Romanization process of the study area (supra, chapter 3). Furthermore, the excavation data from Buxentum itself (even though both fragmentary and episodic) as well as the recent excavation carried out by the Superin-tendency at a Roman complex at Sapri/S. Croce, have provided scholars with at least some excavation data and evidence from which to deduce specific planimetric elements and building typologies for sites dating to the Roman period: such documentation undoubtedly substantiates the interpretation of the scatters located by the surface survey.

7Clearly the paucity of data pertaining to the archaic period means that the settlement picture between the end of 6th and the late 5th century B.C. cannot be presented in detail. A significant element, nonetheless, is found in the fact that the survey evidence appears to tally very well with the data from the habitations excavated on the Central Plateau which in their totality indicate that the settlement of the 5th century B.C. was in a phase of profound transformation.

2. The archaic period

8As is evident from the title of this chapter, the pre-and proto-historic periods as well as the period immediately preceding the “Lucanization” of the study area have been excluded from the synthetic discussion of the settlement evolution. Two factors have influenced this decision:

91) above all the scarcity of data (especially the evidence from the surface survey) belonging to the period preceding the sudden growth in the rural settlement: this growth occurred in the decades between the end of the 5th century and the first half of the 4th century B.C. Maximum growth at the Roccagloriosa site can be dated to the second half of the 4th century B.C. and the first half of the 3rd.

102) a second factor was the inevitable and necessary division of labor, especially in such a large territory. The sheer size of the territory prevents us from producing hypotheses on the transformation of settlement models for the period of the later Iron Age and the 6th century B.C. It would not be possible to consider the settlement development for those periods without including sites such as Palinuro (Greco E. 1975), Pyxous (Johannowsky 1985; Johannowsky 1992) and Sapri (Fiammenghi/Maffettone 1990) which are currently being excavated and studied by others. Even more valid then are the reasons for which we have excluded the Neolithic, Eneolithic and Bronze Ages for which it would be impossible to outline a territorial development without including the Coastal strip between Palinuro and Sapri (currently being studied by various teams from the University of Siena) as well as the area of Caselle in Pittari.

11The foregoing having been said, and with the scarcity of Iron Age and Archaic period data firmly in mind, it is now possible to delineate a general picture of settlement evolution for the catchment basin during the 5th century B.C. on the basis of the most recent evidence that has come to light in the territorial survey and in light of the documentation of the 6th and 5th centuries B.C. provided by the excavations of the Roccagloriosa settlement presented in Roccagloriosa I.

12Three main aspects of the 5th century B.C. evidence, which have emerged in the most recent exploration of the area (1987-91) can be singled out in this context:

13a) On the Central Plateau we find during the second half of the 5th century B.C. archaeological documentation for contacts with the southern Tyrrhenian/Siceliote area, which may be matched to the literary references for dose links between Pyxous and Rhegion/Zankle in the second quarter of the 5th c. B.C. At the same time, we have now undeniable evidence for the presence of an outpost of OscoSabellian culture in the form of the cult place located at the very edge of the mountainous area bordering the region to the north. The San Leo site (supra, Part II, chapter 1), only 6 kms. in a direct line to the north from the Roccagloriosa/Central Plateau site, is not by chance located on high ground (alt., ca. 600 m. a.s.l), in a commanding position along the south-western slope of Mt. Centaurino, which provides a spectacular vantage point over the whole Mingardo valley. Both its topographic location (fig. 126) and the evidence of the Italic bronze figurine (although so far an isolated piece) provide compelling evidence for dose cultural links between the Mingardo/Bussento region and the Apennine hinterland and, more generally speaking, with the Samnitic areas further north. What is more relevant, this recent piece of evidence pushes back to the early 5th century B.C. the documentation for the assimilation of Italic elements in the region, and thus provides unique evidence for the earlier stages of the ‘Samnitization’ of the Mingardo/Bussento region. That such a process was already under way can also be gleaned by the scarce textual information concerning the area around the Gulf of Policastro and the succinct reports about the colonization of Pyxous, as, in fact, has already been pointed out in a masterful synthesis by E. Lepore (Lepore 1966) who discussed the situation in the hinterland of Velia during the period following the fall of Sybaris (see now Cordiano 1995 and La Torre 1993).

14b) It is against this cultural-historical background that we have to consider the 5th century B.C. habitation structures documented on the Central Plateau (Roccagloriosa I, 48-51, Phase IB). The 10 m. x 4 m. long rectangular structure, probably of the ‘megaron’ type (Roccagloriosa I, pl. IX) with porch opening onto a north-south paved road (Roccagloriosa I, fig. 51) was decorated, by the mid-5th century B.C., with gorgoneion antefixes of the ‘horrid’ type. The building thus probably assumed the function of an ‘anaktoron’ for the community living on the plateau. Although the remaining walls pertaining to this phase are fragmentary (Roccagloriosa I, pl. IX) it is evident that we are dealing with a habitation area made up of rectangular structures built on low stone socles. In the course of the 5th century B.C. the flimsy huts of the 6th century B.C. phase (and of even earlier date) which have been documented both on the Central Plateau and in the area of Carpineto (possibly in the DB Area as well, supra, 28 and fig. 16) must have been replaced. The change from huts to rectangular structures on stone socles represents a ‘quantum leap’ in the organization of the early settlement at Roccagloriosa. The degree of change is particularly clear when we consider that only a few decades later a large courtyard house, Complex A, complete with porticoes and a stone-built shrine, replaced the 5th century B.C. habitations. The change also entailed both an extension and more regular organization of the settlement areas as well as a concomitant growth in population. It is, however, possible that the site was probably only seasonally occupied in the 6th and early 5th century phase as there are no burials which can be dated to that period.

15The La Scala cemetery appears to have been used as an elite cemetery area from the late 5th century B.C. (Tombs 6 and Tomb 10) onward (fig. 127). It is very tempting, and arguably justified, to connect these two ‘princely’ burials and the 5th century B.C. ‘anaktoron’ on the Central Plateau.

16Clearly, the archaeological evidence for the second half of the 5th century B.C. (mostly from the Central Plateau) is too fragmentary to allow a more detailed picture of settlement change in the region. It does, however, reflect changes in community organization and settlement structure, concurrently with the emergence or consolidation of local elite groups. These 5th century ‘aristocracies’, as already pointed out by a number of scholars, share strikingly similar characteristics across a vast area of the southern Italian hinterland from Lucania to Apulia (particularly the Daunian area). Generally speaking, such developments are best documented by the funerary evidence from Melfi/Pisciolo, Roscigno/Monte Pruno and Roccagloriosa In Lucania; Lavello in Daunia; Gioia del Colle/Monte Sannace in Peucetia and a number of sites, among which Cavallino is programmatic, a site in Messapia (Bottini/Guzzo 1986,213-220; Tagliente 1987). The recent evidence from Lavello, Serra di Vaglio and Roccagloriosa, however, points to a common type of 5th century building at those sites, which highlights the role of the ‘aristocracies’ within the community structure. At Lavello, where several examples of a 5th century B.C. ‘anaktoron’ have been found, burials associated with those structures have also been documented (Bottini/Fresa/Tagliente 1990). The Iron Age settlements in Apulia made no topographic separation or distinction between houses and tombs, and, based on the evidence from Lavello, it is thus possible to envisage a type of settlement structure with a number of habitation clusters, each one with its own ‘anaktoron’.

17c) The later 5th and early 4th centuries B.C. developments within the region, as shown by a number of recently analyzed tomb groups from various areas within the Mingardo/Bussento region, leave no doubt as to a marked change in the settlement pattern. The changed settlement pattern results from the infiltration of new population groups of ‘Oscan’ or ‘Samnitic’ derivation and included a concomitant denser occupation of the countryside. Such a phenomenon is well documented not only in the vicinity of the Roccaglortosa area, as is shown by the finds from Calatripeda (supra, Part II, chapter 1, n. 37) but also further substantiated in greater detail (see, for example, the late 5th century burial with an early type of ‘Samnitic’ bronze belt) by the finds from Torraca and Tortorella (supra, Part II, 113). Sadly, any evidence pertaining to the late 5th/early 4th century B.C. is still missing from the important site at Caselle in Pittari but this may be an archaeological accident pertaining to the area delimited by the excavation. Nevertheless, the evidence so far available from the region leaves no doubt that, by the end of the 5th century, a process of capillary occupation of the countryside and possible population growth was underway: certainly these developments mark the beginnings of a major 4th century B.C. “dispersione della popolazione nella campagna” (to use a very effective definition by Greco E. 1979) which matches an analogous pattern of rural settlement (fig. 128) found in the countryside of Lucanian Paestum (Pedley 1990, 97-99 and fig. 60).

3. The 4th and 3rd centuries B.C.

18The archaeological evidence pertaining to this period in the study region derives from both systematic large scale excavations and surface survey. Thus, the cumulative evidence, both abundant and quite rich, provides a detailed picture of settlement developments at the time when the region fell under control of the Lucanians (Roccagloriosa I, fig. 2).

19The general picture of the area surrounding the oppidum of Roccagloriosa has already been delineated in Roccagloriosa I and has been integrated in Part I with the new data on the organization of the oppidum itself. The most important new data concerning the territorial organization of the Lucanian period derives from the wider territory and especially from the site of Caselle in Pittari (site no. 39, supra, 107-110) which has been extensively excavated in recent years by the Department of Antiquities of Salerno, under the direction of Dr. A. Fiammenghi. A few data of some relevance to the general picture have also been retrieved from Policastro/Pyxous, the site of the 5th century B.C. phrourion established by Greek colonists, which must have fallen under Lucanian control by the 4th century B.C. (Johannowsky 1992).

20The most recent excavations at Caselle in Pittari, as well as the data from the regional survey, add new emphasis to both the picture of the important nucleation process in the region and to our understanding of the dense occupation of the countryside. As hypothesized on the basis of earlier surface surveys (Roccagloriosa I, 195-196), the two adjoining areas of Laureili and Lovito in the Comune of Caselle in Pittari were the site of a substantial nucleated settlement, placed at an elevation of ca. 300 m. and naturally defended by river gullies. Although so far there can be no certainty that it was actually enclosed within a fortification wall, the sheer size and regular organization of dwellings on the area of the Lovito plateau so far explored recall in several respects the Central Plateau at Roccagloriosa. Still, no evidence has been so far unearthed for the type of courtyard houses found at the Roccagloriosa site. It also appears as if the site of Caselle in Pittari would have become, by the second half of the 4th century B.C., a territorial centre which controlled the upper Bussento valley. Although so far the nature of the data retrieved from Caselle in Pittari does not provide any direct evidence for land use, the elements of comparison with Roccagloriosa would allow us to hypothesize that the nucleated site might have controlled a well organized agrarian territory in the surrounding area, including perhaps also part of the lower Bussento valley, thus somehow encroaching on what has been so far considered as the ‘catchment area’ of the Roccagloriosa site. Particularly relevant to this point are the considerations of Vandermersch (1994, 37 and 236, n. 40) concerning wine production. A question which remains unanswered concerns a possible hierarchical relationship between the Lucanian site at Roccagloriosa and the Caselle in Pittari site (supra, fig. 90). If the Roccagloriosa site did function as a cantonal centre then Caselle in Pittari might have represented a hierarchically dependent site, as hypothesized at a previous stage of the research (Roccagloriosa I, 199 and n. 27). There is a strong possibility, however, that we are dealing with two centres which constitute equal territorial communities of a larger polity, the Sontini, one of the Lucanian subgroups listed by Pliny (NH 3, 15, 97). On the basis of what we know about the political and military organization of the Lucanians (Strabo 6, 1, 2-3 = C 254; Chiranky 1982, passim; Pontrandolfo 1994, 187) and in view of the fact that no centralized political structure is ever mentioned by the literary sources, the latter hypothesis would appear to be the more likely one. It would not be incompatible with what is known about the territorial organization of the Lucanians to find two micro-regional systems within the Mingardo/Bussento catchment basin, such as Roccagloriosa and Caselle in Pittari. If this hypothesis regarding the Lucanian settlement pattern in the Mingardo/Bussento region is correct (a rather big ‘if’, indeed), we would then postulate a similar role for the large site at Rofrano/Pruno (site no. 38, supra) as a third ‘centre’, of hitherto unknown size (supra, fig. 126). The site at Rofrano/Pruno could have controlled the upper Mingardo valley, and would certainly have also functioned, at least in the course of the 4th century B.C., as an important link with Moio della Civitella (at the border of the chora of Velia) and the Calore valley, and thus to Poseidonia/Paestum, to the north. It is also to be kept in mind that the archaeological record of 4th century B.C. settlements does not indicate any clear examples of hierarchical dependency among the nucleated/fortified sites of Lucania. The fortified sites, ranging from a circuit of a few hundred metres to the ca. 4 kilomters of Serra di Vaglio or the multiple circuits of the Civita di Tricarico, appear rather to be autonomous communities of the Lucanian ethnos, and as such function as an essential element of the vicus-pagus settlement pattern of a large number of dispersed rural sites (generally labelled as ‘farmsteads’), dependent economically and otherwise on nearby larger fortified sites.

21As already argued above (supra, Part I, ch. 3) in spite of their territorial ‘centrality’ and ‘political’ role, the habitation areas which grew inside such fortified enclosures cannot be compared to the centralized model of the Greek-city, to be found in the Coastal areas of Magna Graecia. The occasional emergence of a full-fledged urban settlement occurs only in particular areas (those found on the Tyrrhenian side of Lucania, such as Lucanian Laos), a fact which does not affect the general picture outlined above.

22Such remarks about the political and ‘administrative’ role of the Roccagloriosa site within the Mingardo/Bussento territory, formulated on the basis of the archaeological evidence discussed in the previous chapters may now receive further light by the text of the recently discovered Oscan inscription on bronze from the Central Plateau (see Part III, infra). No doubt, a deeper and more complete analysis of the text is required in order to clarify the institutional functions of the “meddes” (mentioned in the text) within the settlement System shown by the archaeological record. The first impression, however, after a preliminary reading and discussion of the inscription (Tocco/Gualtieri/Poccetti c.s.), is that the settlement System outlined above may well constitute the touta | civitas, upon which the “meddes” referred to in the bronze tabula exercised their administrative/politicai functions. This may be a reasonable hypothesis for the Roccagloriosa settlement System, if we can accept (as does Letta 1994 and 1997 for the central Italic areas, where a much richer epigraphic record is available) to refer to the local communities (rather than to the ethnos or nomen) the terms touta and meddes.

23A few more details have also been added to the generai picture (Roccagloriosa I, 213-216) of dense rural habitation in the territory surrounding the nucleated site, within a 5/6 km. radius. The most important new elements of the picture come from the Mortelle site, where the 1989 excavation has provided enough data to allow a reconstruction of the generai aspects of a 4th-3rd century B.C. rural site of the type loosely referred to as “farmstead” in the distribution map, fig. 72, supra.

24The analysis of the finds from Mortelle, presented above (supra, 101-107), adds some qualifying elements to our understanding of the nature of this category of rural sites which had been so far skeched exclusively on the basis of the farm building partially excavated in 1977 at Pedale (site no. 24) (fig. 101).

25a) As far as the habitation structures are concerned, the extent of the surface scatter and the results of a ‘line’ of electric resistivity survey¹ together with a limited test trench provide the evidence for a building complex of ca. 400 sq. metres, generally speaking comparable (although on the basis of very limited structural evidence) to the one excavated at Pedale (site no. 24, cit.).

26We would thus assume this to be the average size of the standard 4th/3rd century B.C. farmsite in the rural territory under scrutiny.

27b) The excavated part of the cemetery area and the evidence from other disturbed burials collected, indicate a cemetery area of 10 to 15 burials. The number of 10 to 15 burials seems to represent a fairly typical Lucanian oikos, as we can also infer from the evidence of other 4th century B.C. rural cemeteries documented in the Lucanian area². One aspect of the evidence which should be underlined in our case is the similarity in grave goods between the rural site at Mortelle and the cemetery area at La Scala, in the immediate vicinity of the nucleated settlement. Such a similarity would seem to underline once more the view of a dose link to the “campagna” of the Lucanian elites (Guzzo 1984, 214) and would rule out any clear-cut distinction between an ‘urban’ elite and a rural ‘peasantry’ within the Lucanian world.

28c) Another important qualifying element deriving from the analysis of the Mortelle tomb groups is the chronology of Tomb 2, which dates to the first half of the 4th century B.C. Together with other chronological elements from the area east of the mouth of the Bussento river (Velia 1990, 38), the Mortelle tomb warns us against compressing or forcing an alleged ‘explosion of rural habitation into the second half of the 4th century B.C.³. Due to lack of systematic excavations of rural sites, it is very likely that the data from rural sites provided by surface survey, which most often point to a fairly generic late 4th/early 3rd century B.C. chronology for the phenomenon of dense occupation of the countryside, may be slightly biased by the very nature of survey data, which cannot provide too specific chronological distinctions, and do not allow us to detect the early stages of the phenomenon. The situation in the territory surrounding Roccagloriosa points to the presence of an indeterminate number of rural settlements in the hilly terrain to the east and west of the Capitenali ridge, already in the first half of the 4th century B.C. and throughout the 3rd.

29Because of the very limited excavation of rural sites so far carried out within the region, actual data for the type of agricultural exploitation of the land attached to the individual farmsites is almost completely lacking. We have some indirect evidence for agricultural exploitation in the form of remains of large dolia most often associated to those sites (map in Gualtieri/Polignac 1991, fig. 2). The archaeobotanical evidence from the oppidum of Roccagloriosa, however, systematically collected and carefully analyzed elsewhere (Bökönyi/Costantini/Fitt 1993) leaves no doubt about a phenomenon of agricultural intensification in the course of the 4th c. B.C., with the introduction of a type of polyculture in which vine cultivation might have been prevalent (fig. 129). To judge from the pattern of 4th/3rd century B.C. farm distribution, such a System of agricultural exploitation might have been prevalent at least in the areas between the middle courses of the Mingardo and Bussento rivers and the coast. The added evidence from a group of amphorae of prob able local production (Roccagloriosa I, 281 nos. 365-370) singled out by P. Arthur in his analysis of the amphora assemblage from the oppidum of Roccagloriosa (Roccagloriosa I, 278-289), adds a new element to the picture of 4th/3rd c. B.C. agricultural exploitation, especially so in the light of a recent study of wine and amphora production in Southern Italy (Vandermersch 1994). Indeed, Guzzo (1984, 205) over a decade ago (when still little importance was attached to the careful study of amphora finds from excavations of settlement sites) had cautioned archaeologists about using the evidence of amphorae found in indigenous contexts of Magna Graecia as exclusive proof of a one-way commercial traffic, from the coast to the hinterland (see Gualtieri 1993, 339, with respect to the specific evidence from the Roccagloriosa site). It is now becoming increasingly clear that the adoption of Italiote amphora shapes on the part of local/regional workshops of the socalled Graeco-Italic amphorae has important implications for the problem of wine production in Southern Italy (Vandermersch 1994, 125-126 and 156-157). In the case of the Mingardo/Bussento region, the combined evidence from the archaeobotanical data and amphora production makes it hard to deny a spread of vine cultivation and wine production (for export as well?) in the hinterland areas between the 4th and the 3rd c. B.C.

30A final consideration concerning the 4th-3rd century B.C. developments in the region emerges from the wider analysis of the black-glazed pottery which the analytical presentation of the evidence from the extra-mural areas and from the sites in the wider regional landscape has made possible⁴. It is now clear, especially on the basis of the evidence from Pyxous/Buxentum (Johannowsky 1992, 174; Bencivenga Trillmich 1988; Greco E. 1991, 672) that the Coastal area might have fallen within the sphere of influence of Elea between the late 5th and the early 4th century B.C., to judge from the finds of blackglaze pottery imported from Velia. Some blackglaze fragments of excellent manufacture from the habitation areas on the Capitenali (e.g. Roccagloriosa I, 257, n. 205), might belong to the same category of imports from Velia. In this respect, it is to be pointed out that within the coin finds from the Roccagloriosa area, ca. 80% of an assemblage of over 100 coins were struck in Velia (Roccagloriosa I, 310-313). Such evidence leaves no doubt about the impact of the “monetary economy” of Velia in the hinterland of the Mingardo/Bussento region (Stazio 1975). Nonetheless, it is also very clear that by the second half of the 4th century B.C. a large number of local workshops dominated the ceramic production in the Mingardo/Bussento region. Amongst the local productions it is certainly the case that the Roccagloriosa site could have been a major manufacturer, given the abundant evidence for kilns and other tools for ceramic production (fig. 130), found at the site. Significantly, the large pottery assemblage from Caselle in Pittari, which is mostly dated between the second half of the 4th century B.C. and the beginning of the 3rd, does not seem to include, so far, any relevant import of black glaze pottery from Velia.

4. The 2nd/1st century B.C. ‘transition’

31The centralizing role of the Lucanian oppidum at Roccagloriosa In the course of the 4th and 3rd centuries B.C. and its function within a definable ‘agrarian space’ in the lower Mingardo and Bussento valleys, already underlined in a number of previous discussions, is all the more evident when one considers the settlement pattern for the 2nd and early 1st centuries B.C., as revealed by the analysis of the data from the surface survey (supra, Part II, chapter 2: the so-called ‘transitional’ sites). Two major transformations in the distribution of rural sites within the surrounding territories stand out:

32a) A drastic reduction in the number of sites, with large areas where surface evidence for settlements pertaining to this time period is totally lacking. The hillsides east and west of the Capitenali ridge, with the obvious exception of some of the re-used habitation structures of the oppidum (supra, Part II, chapter 2), and the lower terraces along the Mingardo valley (to the exclusion of the continuing site of Mai) show a density of habitation reduced to a bare minimum. Although we have already formulated a note of caution about the assessment of settlement densities for a period such as the one under consideration for which diagnostic ceramic forms are lacking (Fracchia and Gualtieri in Velia 1990, 57) a substantial drop in the density of rural site distribution is highlighted by the sheer contrast between the exiguous ceramic finds datable with certainty to this period⁵ and the extremely rich evidence pertaining to the preceding phase.

33b) There is also an evident shift in the distribution pattern of rural sites in the territory surrounding (within a 5-6 km. radius) the former nucleated settlement: from a type of ‘centripetal’ distribution centered on the oppidum to a sparser distribution visibly gravitating toward the Coastal area and the mouth of the Bussento river, with the (at least partially) revitalized site of Buxentum⁶. The site of Mai (no. 3) in the lower hillsides flanking the Mingardo valley, may represent an “heirloom” of the former oppidum. In spite of the necessary caution already formulated above, it is altogether clear that the sparse rural settlement has shifted in the course of the 2nd century B.C. to the hilly terrain between the foothills of Mt. Bulgheria and the right bank of the Bussento river (supra, fig. 92, sites nos. 24, 25, 27, 30, 31, 33, 34, 35, 43)⁷. There may be a number of explanations for such a shift. It is evident, however, that, apart from the consequences of the devitalization of the nucleated oppidum in the later 3rd century B.C., the observed shift in the distribution of rural sites is the result of the insertion of the Mingardo/Bussento region into a wider western Mediterranean exchange network in which it is not difficult to detect the growing hegemony of Rome and the presence of the new economic interests of the fast growing Roman power⁸.

34While the distribution of the so-called ‘transitional’ sites may help us envisage some of the transformations occurring in the settlement pattern, it would be very difficult to provide more specific details for the interpretation of the nature of the individual sites. Test excavations of sites revealed by surface scatters have been so far limited to two 4th-3rd century B.C. sites (nos. 25, 37, supra) and, on the other hand, the structural evidence from the Roccagloriosa site pertaining to this period is extremely fragmentary, represented exclusively by minor habitation structures and in most cases limited to various degrees of re-use of the pre-existing buildings. It may be helpful, however, to formulate some hypotheses on some of the site types so far located by surface scatters.

35We may tentatively identify three different types of rural sites for the period under consideration:

1. There are a few larger scatters (3,500 m² or more) which have already been presented as ‘village’-like clusters of farmsteads (Gualtieri/Polignac 1991,199): this label can certainly be attached to sites such as Mai (no. 3), Scudiere (no. 30) and possibly Morigialdo/Celle di Bulgheria (no. 43). The first two (Mai especially) include a good deal of ceramic evidence for the later 4th and 3rd centuries B.C. and can certainly be interpreted as hamlets or clusters of farmsteads within the 4th-3rd century B.C. rural settlement pattern, placed at the centre of (lowland) areas most suited to agriculture and animal husbandry. The surface evidence, however, does not allow us to infer whether they became smaller in the course of the later 3rd/2nd century B.C., as the ceramic record would seem to suggest, or whether they continued as a sort of settlement replacement (vici) concurrently with the devitalization and subsequent disappearance of the oppidum⁹. In any case, the significance of the ceramic assemblage found on all three of the ‘village like clusters’ (a fourth possible such site, Calatripeda, no. 37, cannot be evaluated) suggests that they ought to be considered as a continuation into the 2nd and 1st centuries B.C. of hamlets or clusters of rural habitations developed within the 4th/3rd century B.C. settlement pattern¹⁰. The more specific, and abundant, evidence on the vicus at Celle di Bulgheria/Morigialdo, already discussed in greater detail in Chapter 3, reveals a significant quantity of finds datable to the 1st century B.C. (in particular, the thin-walled pottery and early Italian sigillata) against a fairly scarce ceramic evidence pertaining to the 4th-3rd centuries B.C. Here is an evident case for the growth of a sizable vicus, mid way between the Capitenali and the lower terrain sloping toward the Bussento valley and strategically sited along the major route Buxentum-Palinurus-Velia. Furthermore, it is evident that this site emerges at the expense of the smaller sites which disappear in the later 1st century B.C. Such evidence, if our hypothesis is correct, would allow us to intepret in a much more shaded manner the caesura in the settlement pattern signalled by the visible drop in the number of rural sites after the late-3rd c. B.C. and the marked increase in the number of sites found in the hilly areas sloping toward the Bussento valley during the last decades of the 1st century B.C. (supra, Part II, chapter 3; fig. 103). We may infer here the re-shaping of a new agrarian structure rather than a radical break with the preceding phase; at the ame time it is clear that in the Coastal area the transformation of the settlement pattern follows a different logie, dictated by developments taking place in the ager Buxentinus.

2. Unquestionably, a number of the medium sized scatters (2,500-3,500 m²) of this period represent an early form of villa rustica, on the basis of what we already know about the spread of the ‘Catonian’ type villa in Lucania¹¹. Although such a hypothesis has not been verified by actual excavations, a recent analysis of the evidence from the site of Santa Venere, in particular the funerary relief (Gualtieri, 1996a), suggests the existence of a villa at least by the mid-1st century B.C. on the higher terrain along the south-eastern slope of the M. Capitenali. Considering its proximity to the Lucanian oppidum and its topographic location on the eastern slope of the M. Capitenali overlooking the low hills sloping toward the mouth of the Bussento river and the coastland, one is tempted to think of the owner of the villa at Santa Venere, even in the absence of any epigraphic record, as a Romanized descendant of the Lucanian aristocracy which controlled the oppidum. A similar hypothesis has been formulated for the owner of the atrium house excavated on the ‘Civita’ of Tricarico and dated to the early 1st century B.C.¹². In the case of the Santa Venere site, the hypothesis regarding ownership would be very much in keeping with the situation in the nearby Vallo di Diano, where the abundant epigraphic record indicates a key role of the local, ‘acculturated’ Lucanian element in the process of Romanization of the area (Solin 1983, 414).

3. A number of the smaller scatters (600-1,000 m²) may indeed represent the survival of the simplest type of rural site, such as the ‘farmstead’, which has been well documented for the preceding phase as has been documented amply for this period throughout Italy (Lloyd 1991).

36The survival of the type does not always imply topographic continuity. Rather, as already argued above, there seems to have been an evident shift in topographic location. A number of ‘farm’ sites pertaining to this period are relatively numerous in the area south-east of M. Capitenali, on the hills sloping toward the lower Bussento valley. It is quite possible that some of the earlier farm-sites lasted through the late Republican/early Imperiai period, as is shown in a few cases by the finds of Italian terra sigillata or Dressel 2-4 amphoras associated with small scatters of 4th/3rd century B.C. black-glaze pottery. A similar survival of small farmsteads is implied by the archaeological evidence from the area of Buccino (Volcei) further north (Dyson 1983). In the neighbouring region of the Vallo di Diano, on the other hand, the well documented Gracchan land assignations no doubt created the conditions for a sudden and more radical change in the rural settlement pattern (Coarelli 1981a, 218-220). It cannot, however, be excluded a priori that even in that region the survival of “pre-Roman modes of production” may have entailed some degree of survival of the older type of small farm in the marginal areas¹³.

5. The Roman period

37Clearly it is not easy to provide a clear-cut distinction between the earlier phase and the ‘Roman period’ under discussion, either in terms of general socio-economic developments or in terms of settlement patterns. A visible change in the settlement pattern of the Mingardo/Bussento region was under way in the period between the 3rd century B.C. and the 2nd century B.C. and it can be surmised that at least some elements of discontinuity might have been introduced by the military and political events which mark the history of Southern Italy in the course of the 3rd century B.C.: first and foremost, Rome’s military victories in Magna Graecia in the first decades of the 3rd century B.C. and the foundation of the Latin colony at Paestum (273 B.C.), which provide the background to the settlement developments discussed in the previous paragraph. Secondly, one should take into consideration the aftermath of the second Punic war (especially Kahrstedt 1959) and the foundation (still quite uncertain, however, in its overall impact on the regional landscape) of a maritime colony in 194/186 at Buxentum¹⁴.

38The main scope of the discussion of this paragraph, however, will be to analyze in a more specific manner the nature of the archaeological evidence of the later 1st century B.C. in the Mingardo/Bussento region (supra, Chapter 3,) against the background outlined for the previous phase (the so-called ‘transitional’ sites) and in relation to the more varied and quantitatively significant evidence for the first three centuries of the Empire.

39We have already mentioned the fact that the spread of the villa, as a new type of settlement, was already under way in the course of the 1st century B.C. (possibily, one might surmise, in the period immediately following the Social War) and that the site of S. Venere may represent an early occurrence of a villa rustica and, presumably, a fairly sizeable one, to judge from the evidence of the funerary monument associated with the site (supra, 136-138). Such evidence finds an immediate counterpart in the villa maritima at S. Croce (near modern Sapri) (supra, 139-140), at the extreme south-eastern border of the region under consideration, although possibly still within the limits of the Late Republican ager Buxentinus¹⁵. The recently published evidence for the Roman period from Tortora/Palecastro (most probably ancient Blanda: La Torre 1991, 149; Paoletti 1994, 474-475) and additional recently analyzed data regarding the spread of villae in Bruttium (Sangineto 1994) substantiate the hypothesis that the spread of villas and an associated transformation in the settlement pattern along the Tyrrhenian areas of Lucania and Bruttium was well under way in the course of the 1st century B.C. Indeed, in Bruttium the initial transformation is dated to «fra la fine del II e l’inizio del I secolo» (Sangineto 1994, 563)¹⁶.

40In addition to the phenomenon of the spread of the villa in the course of the 1st century B.C. (witness the two sites of Santa Venere and Sapri/Santa Croce both dated, in their initial phase, around the middle of the century), the survey results also indicate a change in the settlement pattern of the Mingardo/Bussento region, in the second half of the 1st century. Although it is very likely that early 1st century B.C. sites may be under-represented in our survey evidence¹⁷, the visible increase in the number of sites, especially in the lower Bussento valley, and the qualitative composition of the scatters which include early forms of Italian terra sigillata (Goudineau’s ‘arétine Visse’: Goudineau 1968) and Dressel 2-4 amphorae provide an element of discontinuity when compared with the extreme scarcity of Dressel 1 amphorae and an almost total absence of fine wares clearly datable to the first half of the century).

41The actual historical or socio-economic significance of this caesura, however, is more difficult to define. A number of explanations may be hypothesized. Kahrstedt (1960, 20-22), who first attempted an overall picture of the settlement developments in the region¹⁸ saw in this growing importance of the rural sites a counterpart to the relative weakness of the urban site of Buxentum. His analysis of the rural settlement pattern, however, obviously based in many cases exclusively on published or archival evidence, is often biased by an evident attempt to depict in bright colours a contrast between an alleged thriving economic situation in the countryside and the conventional view of a marked, widespread decline of civic life in Magna Graecia after the Hannibalic war (Lomas 1993, 121 and Gualtieri 1996a; also Sartori 1962)¹⁹. Although we can only rely so far on extremely scarce evidence for the development of the colony of Buxentum (supra, Part II, chapter 3, appendix 3), the epigraphic record is consistent in showing a ‘ripresa’ of the city in the Augustan/JulioClaudian period. Such evidence is matched by an evident revival of rural settlement which may correspond to the assignations “in cancellationem limitibus maritimis” mentioned by the Liber Coloniarum in the period between the second triumvirate and the beginning of the Augustan principate (Gualtieri 1996a). The archaeological evidence of sites such as the two small/medium size “farms” at Vallone Pantana (nos. 29, 29bis) also provides an element of specificity to such a hypothesis, in addition to the more general trend of a growth in the number of sites starting from the last decades of the 1st century B.C.

42The two sites lie in the border area between the hilly terrain east of M. Capitenali and the alluvial plain of the lower Bussento river, a few kms. away from the site of Buxentum itself. The regular alignment of the two sites along this borderline parallel to the coastline strongly suggested, at the moment of discovery (Polignac 1986), a new form of ‘planned’ land-occupation to be connected with the above mentioned assignations. Any more systematic exploration of the lowlands toward the mouth of the Bussento river has been later interrupted for administrative considerations²⁰, nor do we have any recent air-photo coverage of the area, but it is clear that a continuation of the survey in the territory of modera Policastro Bussentino might provide clearer results for the evidence of a possible scheme of regular land division.

43The majority of the sites which have provided evidence of the later 1st century B.C./Augustan period (supra, fig. 103) survive at least until the 3rd century A.D. (supra, fig. 111, although at the present time it is not possible to define either the structure or the architectural layout of the smaller sites. Lacking any evidence from excavation or other documentation than surface scatters within the region discussed one can only rely on evidence from other areas for small farmsteads of this period located by survey and later systematically excavated (Jones 1963, 147-158, from the ager Capenas; Jones 1980, from the Tavoliere).

44Among the larger sites, it is quite likely that some were vici rather than villae, given the fact that the only urban settlement in the region was Buxentum in a peripheral location and quite far from the nonurbanized districts of the Mingardo and upper Bussento valleys (Gualtieri/Polignac 1991, 201). In this respect, it is to be underlined that, in the 1st century A.D., this region is amongst those with highest average inter-center distance within the Italian peninsula (Bekker Nielsen 1989, tables 5.1 and 5.2). Indeed, in the less urbanized districts of eastern Lucania and western Apulia the vicus and vicus/villa seem to have been dominant (Small 1991,208-217).

45The site at Celle di Bulgheria/Morigialdo, identified as a vicus on the basis of size and the composition of the scatter, dates from the 1st century B.C. until the 5th/6th centuries A.D. The floruit of the site, however, is between ca. 100 A.D. and ca. 350/400 A.D. The function of the vicus within the territory of the ager Buxentinus cannot be determined on the basis of survey or geophysical testing alone but the long life of the site may indicate the importance of the agglomeration, especially in light of the distance of the area from Buxentum, the administrative center of the region. Still, several points can be raised. The site, as far as we can tell (from absence of mortared remains, mosaic tesserae, etc.) was not associated with a villa and therefore would have always had a more or less independent economic function within the larger territory. In this sense, the vicus can be seen as a holdover or remnant of the indigenous tradition within the Roman landscape (Leveau 1987-1989, 91). The vicus was reasonably robust from its inception as a Roman site since the large quantity of Italian terra sigillata found on the surface seems to have originated from several production centers (supra, Part II, 138-139). The continued economic strength of the site can be judged by the presence of the imported finewares and amphorae dated to the period between 100 and ca. 350/400 A.D. The position of the site, on what was probably a main road connecting Buxentum with Palinurus/Velia, would also have contributed to its growth in the 2nd century A.D. The location of the site would make it a natural market centre for the more inland population during the Roman period. The vicus may have produced amphorae or roof tiles or bricks at some time to judge both from the magnetic anomaly (supra, Appendix 4) and the amount of misfired bricks on the surface. It is likely that some of the population was also involved in agricultural activity, probably the traditional polyculture practiced in earlier times, on the terraces in the Mingardo valley which seems, perhaps falsely, to have been sparsely inhabited in both the Late Republican and Imperial periods. The growth of the vicus at Morigialdo reflects very clearly (and is definitely tied to) the developments in the general settlement picture in the study area: there is a record of growth and strength from the late Republican through the early and mid-Imperial periods while a possible decline in the economic importance and in the size of the site can be considered only for the period after the 5th century A.D. The decline in size and importance is postulated only on the quantity of 5th-6th century forms found at the site. The strength of other vici in the study area during the late antique period may indicate that the apparent decline after ca. 350-400 A.D. at Morigialdo is only an archaeological accident.

6. ‘Villae’ and agricultural exploitation in the Mingardo/Bussento region

46The spread of the villa as a new form of rural settlement in the course of the 1st century B.C. (supra, 172-173) must certainly have involved a number of structural transformations in the local economy and in the land tenure System of the Mingardo/Bussento region. There are visible changes in the survey evidence analyzed above (e.g. the caesura detected in site density) although we ought not to conceive such changes necessarily in terms of a sudden ‘crisis’²¹. Clearly the structural transformations in the System of land tenure brought about by the spread of villas must have entailed the disappearance of some smaller sites, at the lowest level of the settlement hierarchy. Given the geography of (he area, it is very likely that the villa rustica just replaced the traditional farming System with a new type of productive unit which did not necessarily reproduce faithfully the ‘classic’ model of the Catonian villa and which did not imply a marked phenomenon of property concentration²². In addition to the absence of any excavated site in this time period from this region, the epigraphic record, which might provide specific statements or at least remarks about the System of land-tenure, as for other districts of Roman Lucania (Simelon 1993), is totally lacking for the Mingardo/Bussento region, but other considerations and some of the comparative evidence just cited might help to formulate some hypotheses regarding the problem.

47In his overview of the agrarian economy of Lucania in the early Imperial period, Ghinatti (1973, 382-383) accepts at face value Kahrstedt’s settlement reconstruction (Kahrstedt 1960, 22-23) for the Bussento region in imperial times, and thus postulates a rural landscape with “sparse villas”, which would have continued a type of specialized agrarian exploitation “without, however, reaching levels of production comparable with nearby Campania and Apulia”. He also underlines the importance of wine production in the region (fig. 131), recalling the fame of a wine from Buxentum mentioned by Athenaeus (Ghinatti 1973, 384 and n. 143; see also Gualtieri 1993, 334). The more recent evidence from the surface survey cannot per se add very much to this general picture for the Imperiai period except for providing a much more specific map (supra, fig. 111) of the distribution of villas than the one which Ghinatti borrows from Kahrstedt, as well as integrating a number of smaller sites and adding the evidence of at least one substantial vicus to that map. Although still partial and fragmentary, the new survey evidence, together with the evidence from the Roccagloriosa excavations, can also help us to place the picture of agricultural exploitation in the Mingardo/Bussento region into a broader diachronic perspective.

48The importance of viticulture in the hinterland of Magna Graecia and in some of the Italic territories is becoming increasingly clear (supra, 168-170). As recently pointed out in a detailed analysis of amphora production and circulation in South Italy during the 4th and 3rd centuries B.C., it is reasonable to think that at least part of the wine (and oil?) production was for export (Vandermersch 1994, 154). If so, and if we accept the growing realization that the structural changes brought about by the Romanization of South Italy do not necessarily mean “crisi” or “declino senza ritorno” («la romanisation n’effacera pas l’empreinte laissée par les viticulteurs grecs et gréco-italiques sur le paysage», Vandermersch 1994, 155), it is quite possible to think that the kind of polyculture reconstructed for the territory surrounding the Lucanian oppidum of Roccagloriosa (Roccagloriosa I, 213-215 and Pl. 1; Bökönyi/Costantini/Fitt 1993, 304-305) was not substantially “effaced” by the transformations of the 2nd and 1st century B.C. Although possibly implying a more rational form of agricultural exploitation, the villa System in this area may have carried on, in the context of medium-sized estates, the System of poly-culture and the production of wine for export at least until the early Imperial period.

7. Late antique settlement

49The late antique settlements of the study area, like the transitional period sites, are of difficult interpretation in that, although the evidence is quantitatively scarce, imported wares, as a sign of some vigor, are found at all the sites dated to the late antique period. The region is strongly marked by the presence of the bishopric, established in the 6th century A.D., at Policastro Bussentino which was the territorial urban centre in the late antique period as it was for the Roman period²³. The town, and its port, were certainly operative in the late antique period to judge from the presence of imported wares in the rest of the study area. Still, this period in the Mingardo/Bussento region is marked by a decline in the type and number of sites, and thus in general mirrors the phenomenon already documented in other areas of southern Italy (Vera 1995, 194; Small 1991; Gualtieri/Polignac 1991). Also in a general conformation to the overall situation in the late antique period in Italy, the sites located by the surface survey which are datable to this period appear to be, with two exceptions, vici. What is surprising is the long and reasonably healthy life enjoyed by these vici until the 6th century A.D. The vicus at Morigialdo (n. 43) continues throughout the 4th century A.D. and it is only in the 5th century and beyond, to judge from the numbers of forms datable to the 5th-6th centuries A.D. that the site may have become smaller. The two sites at Pruno/Rofrano (n. 38) and at Piano Grande (n. 40) continue as “vici” although the Piano Grande site seems to die out (and this may also not be correct as a great deal of undatable coarse wares were found on the plateau overlooking the scatter of African red slip wares) in ca. 450/500 A.D. Although in many distriets of Southern Italy (Vera 1995, 194) 50/60% of the previous villa sites seem to continue, this does not appear to be the case in this region. Instead in this area the economic vitality of the late antique period is reflected in the vici in the Mingardo and Bussento river valleys. The two sites in the Mingardo valley itself and below the long abandoned oppidum provide some reason to think that the 5th century A.D. was a period of stability in the inland portion of the area. The sites, both vici and otherwise, were all in the midst of good agricultural land and along a route System. Two sites, in the Mingardo river valley, and below the medieval town of Roccagloriosa, may be considered in connection with the noted establishment in ca. 550 A.D. of a church at Roccagloriosa Itself. In fact, the surprising continued strength of the various vici in the 4th, 5th and 6th centuries A.D. in the study area may well be connected in general with the late 4th-5th century Palaeochristian basilica and the later 6th century bishopic at Policastro Bussentino²⁴.