THE CULTURAL DEVELOPMENT IN NORTH-WESTERN LUCANJA C. 600-273 BC The Cultural Development in North-Western Lucania C. 600-273 BC

by HELLE W. HORSN}ES

<> di BRETSCHNEIDER

ROMAE MMII ANALECTA ROMANA INSTITUTI DANIcI, SUPPL. XXVTII Accademia di Danimarca - Via Omero, 18 - 00197 Rome - Italy Lay-out by the editors

© 2002 <> di BRETSCHNEIDER, Rome

PUBLISHED WITH FINANCIAL SUPPORT FROM

The Dr. M. Aylwin Cotton Foundation Landsdommer V Gieses Legat The Danish Research Council for Humanities G. E. C. Gads Fond Novo Nordisk Fonden

Horsnaes, Helle W. The cultural development in North-Western Lucania : c. 600-273 B.C. / by Helle W. Horsnaes. - Romae: <> di BRETSCHNEIDER , 2002. - 195 p. : ill.; 29 cm. (Analecta Romana Instituti Danici. Supplementum ; 28) ISBN 88-8265-194-0

CDD 21. 937.7

1. Scavi archeologici - Lucania - 600-273 a.C.

The journal ANALECTA ROMANA INSTITUTI DANICI (ARID) publishes studies within the main range of the Academy's research activities: the arts and humanities, history and archaeology.

Intending contributors should get in touch with the editors, who will supply a set of guide- lines and establish a deadline. A print of the article, accompanied y a disk containing the text should be sent to the editors, Accademia di Danimarca, 18 Via Omero, I - 00197 Roma, tel. +39 06 32 65 931, fax +39 06 32 22 717. E-mail: [email protected] Contents

Acknowledgements . 7

Chapter 1. Introduction ...... 9 1.1 The background for the study ...... 9 1.2 History of archaeological research in Lucania ...... 13 1.3 Culture and ethnicity ...... 17 1.4 The chronological framework ...... 19 1.5 Summary ...... 22

Chapter 2. Topography ...... 25 2.1 Morphological changes since antiquity ...... 25 2.2 Communication routes ...... 30 2.3 Natural resources ...... 34 2.4 Summary ...... 36

Chapter 3. Archaeological evidence from Northwestern Lucania ...... 39 3.1 Settlements ...... 39 3.2 Cemeteries ...... 52 3.3 Sanctuaries ...... 99 3.4. Oscan epigraphy ...... 105 3.5 Interpretations of the archaeological evidence ...... 107

Chapter 4. Literary sources ...... 119 4.1 The most important sources ...... 119 4.2 Oinotrians and Oinotria ...... 119 4.3 The foundation of Poseidonia ...... 122 4.4 Lucanians ...... 123 4.5 Alexander Molossus' campaigns in Italy ...... 129

Chapter 5. Conclusions ...... 131

Catalogue of sites ...... 139

List of illustrations ...... 179 Index ...... 181 Bibliography and abbreviations ...... 187 Acknowledgements

In the mid-eighties the school of Classical Seminars in London 1998, and The Greeks Archaeology at the University of Copen- and the Others in Athens 1998 .4 I wish to hagen initiated a research programme at thank all the participants of these seminars Pontecagnano (SA) in collaboration with the for their lively interest and many valuable Museo Archeologico Nazionale dell'Agro Pi- comments on my papers. Friends and col- centino at Pontecagnano and the Istituto leagues in Great Britain, Holland, Italy and Universitario Orientale, Naples.' During a the US, have read parts of the dissertation or year of studies at the I.U.O. and periods of discussed special issues in a lively e-mail cor- fieldwork in Pontecagnano I became in- respondence. My thanks go to all of them. creasingly intrigued by the rich archaeologi- While working on the dissertation, travels cal remains of the ancient peoples of the to conferences and exhibitions were of in- hinterland. Undertaking research on this valuable importance. These travels were topic was aimed first of all at satisfying my funded by the University of Copenhagen and own curiosity to know more about the peo- the "Ingeniør Svend G. Fiedler og hustru's ples who lived and worked in this part of the Foundation". I also wish to express my grat- Mezzogiorno. itude to my tutor, Annette Rathje, lecturer in The present volume is a revised edition of Classical Archaeology at the University of the main chapters of my Ph.D. dissertation. Copenhagen, who supervised my work on presented in 1999. The dissertation also in- the dissertation in 1996-1999; to John Pind, cluded appendices that have been or will be who saved me many days of work by helping published as separate articles, and a database me organize the funerary material in a data- of c.1400 published tombs from the study area. base; to Pernille Flensted-Jensen, who dis- To avoid a too voluminous publication these cussed matters concerning Pseudo-Scylax parts are not included in the present volume. with me and translated and commented on The research on which the dissertation Pseudo-Scylax 12 for me; to my friends and was based profited greatly from a number of colleagues Anne Marie Carstens, Helle institutions and individuals who have helped Damgaard Andersen and Mette Korshoim me gain access to publications, first of all the for lively discussions, disagreements and cru- Royal Library of Copenhagen and the li- cial support. braries of the National Museum and the Ny I am most grateful to Dr. Jan Zahle and Carlsberg Glyptotek. Many friends and col- Dr. Gunver Skytte, former and present di- leagues have searched, xeroxed, sent or oth- rector of the Accademia di Danimarca in erwise supplied me with literature. Rome, for accepting my work for publication Particular aspects of the dissertation were in the ARID Supplementum series. The pub- discussed during my participation in the sec- lication was made possible through generous tion on Urbanization processes in early his- grants from my present employer The Royal toric Greece and Italy at the FAA conference Collection of Coins and Medals at the Na- in Ravenna 1997,' the seminar on Ceramics tional Museum, from the Dr. M. Aylwin Cot- in Contexts in Stockholm 1997,' The Ritual ton Foundation, from the C.E.G. Gad Foun-

8 H. W. HORSNS dation, from the Landsdommer V. Giese been discussed, and information deriving Foundation, from the Novo Nordisk Foun- from them have only been hinted at in ad- dation, and from the Danish Research Coun- ditional footnotes. cil for the Humanities. My most sincere Quotation marks have been employed to thanks to all of them. distinguish between ethnic names used by It has not been possible to update the bib- ancient authors and those employed to de- liography and make a thorough revision of scribe archaeological types by modern schol- the material since the end of 1998. Refer- ars. Thus Oinotrian land is the area de- ences to later studies or material published scribed as Oinotrian by Herodotus, while after that date will be found only occasion- "Oinotrian" pottery is the modern type ally, and important works, notably Joseph name. Exceptions to this rule are the names Coleman Carter: The Chora of Metaponto used by A.D. Trendall, i.e. Lucanian, Paestan (University of Texas Press, Institute of Clas- etc. pottery. sical Archaeology, 1998), have therefore not Unless otherwise indicated, dates are BC.

1 Strom 1993. 2 Horsns 2002. Horsris 2001. Horsns 1999. Introduction

Section 1.1 The background for the a certain degree being used as an illustration study - and more rarely as a confirmation - of the literary sources, rather than as an independ- The aim of the present volume is to study and ent source. It is consequently interpreted in discuss the indigenous inhabitants of North- light of a history based on almost exclusive- Western Lucania: their cultural characteris- ly literary sources. By giving the archaeolog- tics, their actions and their interactions. ical material priority in the present study, I The study has been undertaken as a de- do not intend to claim the superiority of one construction: I mean by that a study of the kind of evidence over another, but I wish to archaeological remains freed, as far as possi- point out areas where an archaeological ble, from the successive interpretations im- study may shed fresh light, provide new per- posed on them by the literary sources. The spectives and thus balance the traditional ap- aim is to see whether it is possible, by way of proach. old-fashioned archaeological methods (such as the distribution of types), to reconstruct The area an historical development, to gain knowl- The primary area of research has been de- edge of cultural features, and/or postulate fined topographically as comprising both the existence of cultural groups. If it is pos- coastal zones, river valleys and highlands. sible to write a history by means of the ar- The northern limit is defined by the water- chaeological material, will it then differ from courses from the source of the Tito in the a history based on the literary evidence? Is it mountains south of Potenza, following their possible to define cultural groups, and if so, course towards the sea via the Platano, Bian- is the permeability from one area to anoth- co, Tanagro, and Sele rivers. The southern er so extensive that we cannot speak of a bor- limit is formed by the rivers Laos, from its der between two or more cultural groups, source to its mouth in the Tyrrhenian Sea, but rather a zone of transmission? The liter- and Sinni, on the stretch from its spring to ary/historiographical evidence will be re- the Lago di Monte Cotugno. To the east a viewed and the results of my archaeological line is drawn through the practically unin- investigation compared with the historians' habited mountains (+1300 m) between the views. In particular I will ask whether the two springs of the Tito and the Sauro; it then fol- types of evidence tell the same story, and if lows the course of the Sauro to its confluent it is possible to equate cultural groups with with the Agri. As a general rule, the study peoples, ethnoi, known from ancient literary comprises all known archaeological sites sit- sources. uated within this area, and in the river val- I thus regard the two main groups of evi- leys mentioned above, no matter on which dence, archaeological and literary, as of side of the river the site is found (Figs. 1-2).' equal value for the construction of a history. This principle is violated only in the case of It seems, however, that even for modern the lower course of the Sele, where the sites scholars the archaeological material is still to on the right bank have been treated as sec-

10 H. W. HORSNES ondary evidence. The Etruscan sites in this Material from other parts of Lucania is, area, the Agro Picentino, raise problems however, included when needed. In some with regards to cultural definitions and eth- cases recent excavations (for example in Po- nicity that methodologically and theoretical- marico Vecchio and Tricarico) have provid- ly are parallel to the primary area. The in- ed excellent new publications that throw teractions between the Etruscan culture of light on, the finds from North-Western Luca- the Agro Picentino and the cultures of the nia, in other cases the best available evi- surrounding area, as well as the transmission dence for an argument is found outside the from Etruscan to Lucanian culture within area in question.' the Agro Picentino are highly interesting problems that deserve to be more closely ex- The period amined than has been possible in the pres- The study spans the period of the Greek ent work. presence on the Tyrrhenian coast, from the earliest Greek settlers at Poseidonia around Any definition is artificial, created in order 600 BC to the establishment of the Roman to handle the material. By employing a top- colonialatina at the same site in 273 BC. ographical definition of the research area I Within the phase considered here, a sharp am conscious of cutting through previously division has frequently been detected be- defined cultural groups. I hope that this di- tween the Archaic/ Classical period and the vision will have a psychological effect, re- Lucanian period.' This division is in itself minding us that all sites have had relations very unfortunate, all the more so as the with areas outside as well as inside the re- Greek terms "Archaic/ Classical" are not de- search area. fined. It is clear, however, that in this context

-

INS

14!

_

Fig. 1. Archaeological sites in the reasearch area (nos. 1-66), and other sites mentioned in the text. (Drawing by Lars Davidsen)

INTRODUCTION 11

the transition from "Classical" to "Lucanian" temples, were never "imported" into indige- takes place sometime in the fifth century. nous areas, and have therefore not been dis- Therefore the greater part of the period that cussed. Whereas Velia remained Greek until we normally call Classical in a Greek context the arrival of the Romans, Poseidonia at (namely the fifth and the fourth century to some point became an indigenous town, c.330 BC) is included in the Lucanian phase, which is the reason why Poseidonia has been and thereby seen in opposition to the "Ar- treated more fully than Velia in the present chaic/Classical" period. study. Still, even in the "Greek" period of Po- The artificial division between "Archa- seidonia the two apoikiai were not identical ic/Classical" and "Lucanian" is probably due and a study of their internal relations would to the strong belief among many scholars indeed be interesting, but falls outside the that a profound cultural change took place scope of the present study. in the second half of the fifth century. The change is assumed to have been conditioned The natural environment has conditioned by the wave of immigrants of Samnite de- life in a given area, and is therefore of prime scent, subsequently to become the Lucanians importance for the understanding of a re- and Brettians of the ancient authors. gion. The geology, the geomorphological The chronological limits of the study are changes and the natural resources are treat- as artificial as the geographical ones. The in- ed in Chapter 2. land zones of North-Western Lucania were Chapter 3 consists of a survey and a dis- definitely inhabited in the preceding period, cussion of three groups of contexts (settle- and many Greek settlements had already ments, cemeteries, and sanctuaries) .' These been founded elsewhere in Southern Italy be- groups are the basic types of topographical fore 600 BC. Likewise, life continued in many units. Both the individual structures (e.g. - if not most - sites after the establishment of house, grave, altar) and the small finds the Roman colony of Paestum in 273 BC. from within these structures are, of course, included in each group. The numismatic The sources and epigraphical evidence constitutes a There are two main sources for the cultural fourth group of evidence that can be con- development: archaeological sources and sidered semi-archaeological or semi-literary. written sources. Each line of study presents Coins and inscriptions have a number of a number of problems in itself, but the ma- characteristics in common with archaeolog- jor problems arise when trying to piece to- ical finds as objects recovered from a cer- gether the different kinds of evidence. I am tain (primary or secondary) context. But convinced that no evidence can be left out, the fact that both groups also carry texts but any attempt to correlate literary and ar- puts them in the group of written evidence. chaeological evidence must be made with The letters written on the objects may be of great caution and only after a thorough importance both as evidence for the use of analysis of each of the individual sources by a certain alphabet or language, and as a itself. I intend to examine the archaeologi- text. Here, however, the epigraphic and the cal remains as my primary sources. The ma- numismatic materials are primarily treated terial will be considered irrespective of the as archaeological objects. The coins are dis- ethnic boundaries normally drawn in this cussed in relation to their thematic con- area between Italic and Greeks sites, in or- texts, while the Oscan inscriptions from the der to test the validity of these boundaries. pre-Roman period are examined briefly in The present study focuses on the indige- Section 3.4. nous peoples. The two undoubtedly Greek The state of publication has necessitated apoikiai in the area - Poseidonia and Velia - an intensive search to collect information on have therefore been less thoroughly treated. as many sites as possible - often by means of Specific Greek features, most notable the museum catalogues and in the footnotes of

12 H. W. HORSNZES PII!HIIIIIIiEIIIIIIII!IIIIIIIIIIIIIIIIIH

EII!!IIIIHHIII!IIIIIIIll!IIII!IllhIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIPIIIIIIII!IIIIIIIII !IIIIIHIIII!IIIIIIIIIIIIIIIIIIIII!IIII!!IH! !IIIIIIIIIIIIIIIIIIIIIIIIIIIIllhIIIIHhI!IIII HhIIIHh!IIIHI!IIIIIIIIII!IIIII!IIIII!IIIi HhIIIIIIHIIIIIIHhIIIIIIIHhIII!IIIHhIIII !IIIIIIIIIIIIIIII!IIIIIIIIIIIIIII!IIIII!IIII HIIIIIIIIIIIHIIIIIIIIIIIIIIIIIIIIIIIHIIIII III!I!IIIIIIIIILIuIIIIulIII!III!IIIIIIIIIIII • !IIIHhIIII!IIIIIIIIIIIIHhI!IIIIIIIIIIIIIIII !IIIIII!IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII!IIII !IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII!IIII IH!IIIIIIIlIIIIIIIH!I!I!II!LI!!Hh!IllhII • iIlII!IIIIII!IHII!I!IlII!III!IIllIIIIIHIII •lI!!HhlIIIIlIIIIIII!IIIIllhIlII!!IIIII!IIII

HIMI Fig. 2. Schematical representation of the types of finds from 66 archaeological sites in the research area.

INTRODUCTION 13

analytical articles. This information has been collected in the "Catalogue of Sites".5 After the discussion of the archaeological material, the literary sources will be exam- ined (in Chapter 4), with a view to their com- II!IIIIIIIIIIIIII1 III patibility with the results gained from the analysis of the archaeological material.

Section 1.2 History of archaeological research

The "rediscovery" of Poseidonia and the vas- es of the antiquaries IIIIIIIIIIIIIIIIIIIIII The earliest recorded finds from North-West- ern Lucania took place in the age of the an- HllhIIiIUII!IIIIHHI tiquaries. Although the ruins of Poseidonia were always visible in the landscape, it is cus- tomary to speak of the so-called rediscovery II!IIIIIIIIIIIIIIIIEI of the Greek temples at Poseidonia in the mid-eighteenth century, when their exis- IIIIIIIIIIIIIIIIIIII!I tence became known to travellers from northern Europe .6 The growing interest in IllhIIIIIIIIIIHhIII!I all things Greek interacted with an urge to IIIIIIHIIIIIHIIIII!I travel and led to the publication of descrip- tions and images of the ruins, among which IHIIIIIIIIIIIIIIIIIII the description by Winckelmann is regarded as the first scientific work on Poseidonia. 7 It Iu!IIIuIuIuIuIuIuIIIH was quickly followed by the descriptions by for instance Saint-Non (1783), Paoli (1784), IIlI!I!II!!IUIII!IIII Swinburne (1785), and Delagardette (1799). IIIIIIIIIIIHIIIN!I!I The publication of Sir William Hamilton's first collection of vases in four volumes in IIIIIHHIIIIIIII!!I!I 1766 has been seen as a decisive factor for the "vase rush" of the following years," but it IIIIIHHIIIIIIIIIIIII is perhaps more correct to see it as a symbol of the Age of Enlightenment. At the time of IIII!IIIIuIIII!uuIIIII the earliest visits to Poseidonia excavations were also undertaken in the cities destroyed !!IIIIIuIIIuII!IuIIIII by Vesuvius during the famous eruption in !!ILIIIIIIIIHIHIIIII AD 79, and even some inland sites were known to the collectors who included both local aristocrats and foreign travellers.9

From antiquarianism to modern archaeology When the hunt for vases, gems and other ob- jets d'art gradually ceased, inland sites fell in- to oblivion. The early archaeological re- search was to a large degree focused on the Fig. 2. (continued). Greek cities of Magna Graecia, 10 with some

14 H. W. HORSNES notable exceptions: Scholars interested in to- ly unpublished to our day, the German ex- pography and prehistory alike (among oth- cavations were swiftly published and had an ers Di Cicco, Lacava, Patroni and Viola) did enormous impact. The Early Iron Age mate- not confine themselves to the coastal areas, rial from Sala Consilina was classified by K. but travelled through large parts of South- Kilian, who had been trained as a prehisto- ern Italy. For Lacava the Greek settlement at rian under Muller Karpe, and like him used Metaponto was the starting point for the top- the seriation method for the establishment ographical description of the inland valleys of the relative chronology of the site. 14 Con- and the ancient remains found there." temporaneously, J. de La Genière concen- The first excavations at Poseidonia took trated on the Orientalizing and Archaic ma- place in 1907 under the direction of Spinaz- terial from Sala Consilina, and produced a zola. But it was particularly in the 1930s that similar sequence based on an analysis of the intensive activities were focused on the site indigenous geometric pottery.15 (including the unfortunate construction of the SS 18 right through the centre of the site The 1960s and 1970s and the Roman amphitheatre). The excava- In 1961 the first Convegno di Studi sulla tions at Poseidonia have concentrated on the Magna Grecia took place at Taranto with the urban sanctuaries and the administrative title Greci ed Indigeni in Magna Grecia. Since centres of the Greek/Lucanian and Roman then the Taranto congress has become an in- towns respectively. Furthermore more than stitution in the archaeology of Southern 2500 tombs have been excavated in the ur- Italy, and the annual reports on new finds in ban cemeteries of Poseidonia, and fieldwork the congress proceedings are indispensable. has brought to light evidence of the settle- Modern archaeological research in Basili- ment pattern in the chora. The discovery of cata is inseparably bound up with the life the Hera sanctuary at Foce del Sele in 1934 and work of Dinu Adamesteanu, who be- was made during a survey with the explicit came the first Soprintendente when the Ar- intention of finding the sanctuary known chaeological Soprintendenza of Basilicata from literary sources." In the 1950s the was created in 1964. A few years later, Museo Archeologico Nazionale di Paestum Adamesteanu was able to present summaries was erected with a view to housing the fa- of the settlements of Basilicata", and of the mous sculpted metopes from the Heraion. cultural regions.'7 Excavations are still ongoing at both sites. In 1968, another brilliant figure in the ar- Although the site of Velia had been iden- chaeology of Basilicata, Juliette de La tified already in the seventeenth century, few Genière, provided the first attempt to estab- excavations took place there until the lish a comprehensive chronology of the Iron 1950s." Since then the diateichismata fortifi- Age of Southern Italy. The work was based cation has been exhaustively studied, as have on her study of the material from Sala Con- the different parts of the city and its urban silina, which she updated and divided into development. Contrary to Poseidonia, little main phases closely comparable to the phas- fieldwork has been undertaken outside the es of Kilian. The summaries on the indige- urban area: only one(!) tomb has been re- nous peoples of Iron Age Basilicata pub- ported from the Greek period of Velia, and lished during the 1970s mainly draw on the the suburban territory is practically un- works of de La Genière. 18 Lo Porto's publi- known. cation of old finds from the Materano area In the late 1950s K. Kilian, R. Naumann in the Ridola Museum gave access to mate- and B. Neutsch carried out investigations in rial from another region, 19 and new discov- Sala Consilina and Palinuro respectively. At eries made in the 1970s in the Agri Valley both sites excavations had been undertaken (Alianello, Chiaromonte, and particularly previously by the ubiquitous P.C. Sestieri, but Roccanova) enabled Tocco to add important while Sestieri's excavations remain practical- evidence to de La Genière's study.20

INTRODUCTION 15

From the 1950s to the 1970s the archae- small survey has been undertaken by M. ology of inland or indigenous - Lucania was Skele from Washington University (Mis- mainly a prehistoric archaeology with studies souri). Joint projects in the indigenous areas rarely venturing any later in date than the have already been mentioned (Sala Consili- sixth century BC. na, Palinuro, Satriano), and the recent Cana- An important exception was the American dian project at Roccagloriosa, directed by M. investigations at Satriano in 1966 and 1967, Gualtieri and H. Fracchia, is a very good ex- where material from the period from c.600 ample of the varied information that can be to the early third century was found. The ex- obtained from a large, planned fieldwork cavations were directed by R.R. Holloway,2' project. 14 who also undertook some minor fieldwork in Outside the study area proper, much in- the Buccino area. formation is currently being unearthed at The material available up to c.1980 Pomarico Vecchio by the University of Tori- formed the basis for the first comprehensive no directed by M. Barra Bagnasco 25 and at textbook on I Lucani.22 The subtitle of the Civita de Tricarico by a French team led by book - Etnografia e archeologia di una regione 0. de Cazanove.26 These sites are partly con- antica - is indicative of its approach. Yet the temporary; both existed in the late fourth first chapter is devoted to an analysis of the and early third century. Pomarico seems to literary sources. This analysis shows that Pon- be a single phase site, but it must be under- trandolfo considered the Lucanians to be de- lined that while the excavation method scendants of the indigenous peoples settled (open area excavation) has produced very in the area since the Iron Age, rather than good information on the seemingly last set- immigrants from Samnium. 23 By this inter- tlement phase, deep trenches sunk to virgin pretation Pontrandolfo could proceed to tell soil have only been conducted in a few spots. the continuous story of the Lucanians from It is possible, therefore, that an earlier phase the Iron Age to the Roman conquest. may be hidden beneath the uncovered layers of the town. Fieldwork in the 1980s and 1990s Contrary to Pomarico, continuous habita- Since the publication of I Lucani, intensive tion and even noteworthy construction excavation campaigns have led to a multipli- works, undertaken during the third century cation of archaeological sites and an in- at Tricarico, are now falsifying the assump- creased knowledge of the archaeology of tion that all sites in the interior were desert- Basilicata. Much work followed the major ed after the fall of Taranto to the Romans. earthquake in 1980 that necessitated restora- A common characteristic of the planned tion of medieval towncentres, as for example excavations is that they tend to be settlement at Eboli, and the construction of new hous- sites (although in the case of Roccagloriosa es for the "terremotati". some tombs have also been excavated). Planned investigations in the field are Rescue excavations, on the other hand, mainly carried out as university excavations, mostly deal with cemeteries threatened ei- often in the form of international collabora- ther by modern building activity or by tion between the local Soprintendenza and a tombaroli and the consequent illegal export foreign university. Good examples are the of antiquities. These excavations have led to ongoing excavations at Velia, where the uni- an accumulation of excavated, but unpub- versities of Innsbruck and Vienna are the lished, material over the last decades. This is foreign partners, and Poseidonia, where especially the case with the huge cemeteries scholars from the École Francaise de Rome of remote inland sites as Aliano, Chiaro- have taken part in many excavations in the monte, Roccanova, Guardia Perticara, San- centre, while the excavations of the Santa t'Arcangelo - sites where the numbers of Venera sanctuary were entrusted to an Amer- tombs are counted in hundreds or even ican team directed by J.G. Pedley, and a thousands.

16 H. W. HORSNES

While the results of the planned research l'Italia Meridionale e della Sicilia (DANIMS). excavations have often been published with The BTCGI describes sites under a number impressive speed, publication of rescue ex- of headings (A. literary, epigraphical and nu- cavations tends, almost by law of nature, to mismatic evidence; B. history of archaeolog- lag behind. More often than not, only par- ical research; C. bibliography), and the ar- ticular finds from rescue excavations are de- chaeological sites are marked on a section of scribed in museum catalogues (with the ob- the relevant 1GM map (scale 1:25.000). The vious intention of creating an understanding BTCGI also provides annual lists of "opere di for the work in the local community) or in carattere generale". One would expect only analytical articles. Greek sites to be included in the BTCGI, but In some areas surveys have been con- the criteria for the selection of sites are ducted. A distinction should be drawn be- blurred, and in reality many non-Greek sites tween the systematic surveys, as for example have been listed. the Capodifiume survey,27 and less systemat- The DANIMS project intends to give ho- ic collections of material known from old re- mogeneous summaries of the evidence from ports in the archives combined with some cemeteries in Southern Italy by asking exca- reconnaissance in the field. The latter is by vators to fill in a standard questionnaire. The far the most common form of ricognizioni summaries are very useful for comparing dif- undertaken in North-Western Lucania.28 ferent sites, but they cannot substitute full These regional studies often fall somewhere publications. Furthermore, the DANIMS arti- between a publication of a fieldwork and a cles inevitably become outdated when field- collection of old materials. Through these work is continued after the publication of surveys it has become possible to map a the summaries, as has happened for two of large number of sites with pottery scatters the sites relevant for the present study.32 from various epochs. The material is mainly published with only general descriptions, Typological studies of single groups of ma- e.g. "impasto", "domestic pottery", "tiles" terial are still rare. One of the first groups to etc. The black glaze pottery is often the on- be studied was the mainly Hellenistic ly class that can be dated, and sites with unguentarii,33 followed by the Iron Age types finds of black glaze pottery outnumber all discussed by de La Genière. 34 Since the 1970s other sites, but it is questionable whether the Italiote Red Figured vases have been the this can be interpreted a priori as an indica- object of A.D. Trendall's monumental stud- tion of demographic growth or of a change ies in the Beazley tradition, and the con- in settlement patterns, as is commonly as- temporary black glaze forms are to some ex- sumed.29 Although the survey methods are tent included in the study of the Céramique sometimes described,30 there is no indica- Campanienne. 35 In 1984, the indigenous geo- tion as to what areas exactly were surveyed, metric pottery was the object of a mono- or which of the surveyed areas were com- graphic study by D. Yntema.36 pletely without finds. The many pitfalls in Recently, A. Russo Tagliente has collected the interpretation of survey data are rarely the remains of domestic buildings in Luca- discussed. In one case, however, E. Greco nia and Apulia in an impressive catalogue, noted the surprising find of fifth century which forms the basis of a discussion of the tombs when excavating in an area where the development of domestic buildings in South- survey had led to the presumption that a ern Italy from the eighth to the third cen- sixth century settlement would be found .31 turies .37 Apart from these regional studies, sum- Aspects of warfare have been comprehen- maries of the material are being made by the sively discussed in several studies in the last Bibliografia Topografica della Colonizzazione Gre- decade. Not only the weapons , 38 but also the ca in Italia e nelle Isole Tirreniche (BTCGJ) and fortifications39 and the history of the Italic the Document azione Analitica delle Necropoli del- mercenaries, have been examined .40

INTRODUCTION 17

Modern interpretations of the archaeological A danger when working with the archaeolo- record gy of Southern Italy is the internal rivalry be- In the interpretation of the archaeological tween different scholars (or groups of schol- material, lines of demarcation can be drawn ars). It is of course salutary to have a lively - in terms of theory, methodology, and not debate on the interpretation of the materi- least mentality - between archaeologists al, but in some cases it seems this debate is working in a prehistoric tradition (in this deliberately avoided by not citing views op- connection on Iron Age sites), scholars in- posed to one's own "school" .42 This rivalry vestigating the Greek cities and their chorai, has the unfortunate effect of too easily en- and scholars tackling problems pertaining to snaring the archaeologist in a net of refer- the Romanization of Southern Italy. As with ences that never allows him to see the oth- all classifications there is some overlap be- er side of the picture. In other cases, where tween the categories. Most research of the there seems to be little disagreement be- indigenous cultures from 600 to 273 BC is tween different scholars, neglecting the being carried out by scholars fluctuating be- work of one scholar may be due either to ig- tween one research area and another. norance of it or to a wish to promote friends Regarding the material from c.600-273 and play down the importance of others. BC, the "Greek" explanatory models now This may be the reason why the book by the prevail. These models have now penetrated Dutch scholar D. Yntema on South Italian into the inland river valleys, as a reflection matt-painted pottery seems to have had lit- of the way modern scholars claim that Greek tle impact on studies undertaken by Italian culture, or even the Greeks themselves, in- scholars .4-1 fluenced the indigenous areas in the sixth century BC. A good example of this line of research is the enormous importance as- Section 1.3 Culture and ethnicity cribed to the "destruction of Sybaris in 510 BC" that has become pivotal for the under- The present study will attempt to discuss two standing of almost any find from the sixth very complex issues: culture and ethnicity. and fifth centuries. It can be illustrated by This will not be confined to a theoretical dis- the chronological distribution of imported cussion. The issues will be examined from a Greek pottery found in the Agri and Sinni pragmatic point of view: how can we use the Valleys. In the early 1980s very few examples evidence? of Attic pottery from the late sixth and ear- Both ethnos and culture are heavily loaded ly fifth centuries were known from this area. words. Many attempts have been made to Today, this picture has been completely avoid using them, but in effect this only changed, and the largest quantity of import- transfers the problems of understanding the ed Attic pottery in the area dates to the first meaning of one word to another. In the pres- quarter of the fifth century BC: ent study I will use the word "culture" to des- ignate an entity with a number of common "La documentazione di importazione .. con- denominators that can be recognized in the sente .. di registrare .. uria situazione pifl ar- physical evidence. ticolata rispetto a quella iota in passato I maintain that although it is absurd to (quando la quasi totale assenza di materiali talk about the "X-pot people" '44 it does make importati nel territorio oggetto di atten- sense to talk about an archaeological culture zione era stata letta in relazione al crollo del defined by careful analysis of the material sistema "commerciale" facente capo a culture - an analysis that of course must in- Sybaris) e di interpretare l'addensarsi delle clude considerations of the validity of the presenze nel primo quarto del V secolo nel- single elements as well as of the combination l'ambito di dinamiche successive alla caduta between them .45 della colonia".41 A culture can be recognized by its mem-

18 H. W. H0RSNS bers' use of a particular dress code, their whether it is a conscious use of a feature, it way of life as reflected in their house and is also necessary to keep in mind that the furniture, their way of eating as reflected in same feature may have different meanings, cutlery and tableware, their religious beliefs depending on which society it is found in, as reflected in their graves and sanctuaries. and - perhaps even more important - the These are the kind of cultural traits that meaning of a particular feature within the can most easily be likened with archaeolog- same society may change over a relatively ical cultures: a set of common physical fea- short time .41 Similarly, the use of a cultural tures. In some cases the (archaeological) marker may change from conscious to un- culture may be identical to the ethnic group conscious or the other way round during a in Barth's sense: a group defined as a group relatively short time. by themselves or others .46 A look around "Social" markers, indicating the age, sex, the world today clearly reveals that there profession and/or status of an individual will are many types of cultures, and that they be found as sub-strata within a cultural enti- are not necessarily identical with peoples in ty, but at the same time it is likely that social the political sense. Today, it is often easy to markers will have a wider distribution than tell what nation a person come from by the local/regional culture. looking at his/her clothes or accessories. But there are many transnational cultures, It is still customary to start any book on the and even more examples of different cul- Italic peoples or the archaeology of inland tures co-existing (peacefully or not) within Southern Italy with a chapter devoted to the the same nation. extremely few statements about them in an- This should warn us against equating ar- cient literature .48 Furthermore there is today chaeological cultures with political group- no commonly accepted terminology for ar- ings faintly known from literary sources. I chaeological cultures and types, and the fact therefore find it necessary to employ a sharp that some types have been named after eth- distinction between a culture and an ethnos - noi known from written sources has blurred here defined as a people/political group re- the picture: "Oinotrian" pottery is found ferred to in ancient literary sources. When I over an area much more extensive than - and use the Greek word ethnos, I do so to stress not always overlapping with - the area that that this is the ancient usage with all the lack may have been the home of the ethnos known of homogeneity and precise definiti9ns that as the Oinotrians to some Greeks of the Clas- we may expect from a heterogeneous group sical period. Nor is it impossible to find un- of ancient authors. The English translation considered references to the Lucanian war- will be people, a word that may cover all sizes rior equipped with a "Samnite" belt and of groups (clan, tribe, nation etc.) and types buried in a "Paestan" tomb at Poseidonia. of political organization. Initially, it is therefore important to re- It is also necessary to draw a distinction view, and if possible refine, the definitions between conscious and unconscious cultural and the terminology. Modern scholars are markers. The conscious cultural marker can aware of this problem. A good example is Yn- be exemplified by the wearing of a crucifix, tema49 who suggested that names for stylistic a piercing or the like, intended to signal a groups should be taken from the modern membership of a particular group to those names of their distribution area (e.g. his who are able to read the signals. The un- West-Lucanian and Ruvo-Satriano classes). conscious marker can be exemplified by the Yet, it is quite difficult to rid oneself of well- use of one particular type of functional ob- established terms. Instead of inventing new ject rather than another, caused by tradition terms for artefact groups I have therefore de- or the sheer availability of the object. cided to use the traditional terms derived Apart from arguing whether a specific from ancient texts in inverted commas (i.e. trait is a cultural marker or not, and if so "Samnite" belt; "Ionian" cups etc.).

INTRODUCTION 19

Section 1.4 The chronological frame- The local chronologies are often based on work the absolute chronology of Greek pottery. In the first half of the seventh century BC, This study never intended to discuss - or to Corinthian wares predominated among im- attempt a revision of- the already established ported pottery. Their typological develop- chronologies of the archaeological material. ment and the relative chronology based on The amount of work involved in such a dis- it are decisive for Italic chronology in the cussion would have impeded any other Orientalizing period. Yet for the present theme being studied within the relatively study the Corinthian sequence is of little use; short period available for the present study; firstly because the Corinthian pottery is of it would also demand an extremely good less importance in the sixth century than in knowledge of especially the ceramic materi- the preceding one, secondly because al, based on autopsy. Corinthian pottery is extremely rare on the So, unless otherwise stated, the dates giv- inland sites. en below are taken from the original reports In the sixth century there is a greater dif- or publications. It should be noted that the ference between the fine wares, represented dates of structures are almost invariably mainly by the Attic figured vases, the band- based on the chronologies built on pottery ed or black glazed wares in Greek style, and typologies. These typologies are of course the local painted wares with matt-painted subject to criticism, and perhaps even modi- geometric decoration. fication, as more material becomes available, The Attic figured pottery plays a major and the absolute dates based on the typolo- role for dating single contexts in the late gies may seem misleadingly precise. In many sixth and fifth centuries BC. In the fourth instances the dating material - often restrict- century BC, the Italiote red figured pottery ed to a few diagnostic sherds - has not been takes over the role previously played by Attic published and it is therefore impossible to figured pottery, both as luxury ware found in discuss the chronology of a particular struc- many tombs, and as the prime dating evi- ture. Where I have suggested a date other dence for modern archaeologists. Unfortu- than the one given in the original publica- nately, both the Attic and the Italiote figured tions, it is mentioned explicitly. pottery present problems: While Attic paint- Yet, during the review of the archaeologi- ed pottery is not uncommon in the coastal cal material it became obvious that the rela- sites (for example Greek Poseidonia and tive chronology of both the fifth and the third Pontecagnano in the Etruscan Ager Picenti- centuries may be subject to revision. There- nus), only a few of the vases that reached the fore a short discussion of the absolute as well cemeteries of the Vallo di Diano 5° and the as the relative chronology will be given. Agri51 and Sinni52 Valleys have been pub- It should be noted that pottery can rarely lished with their contexts. The Italiote pro- be absolutely dated with any greater preci- duction is better represented in inland sites, sion than approximately half a century. This but there is little evidence on which to base makes correlations between archaeological the absolute chronology of South Italian red finds and historical events extremely diffi- figured pottery.53 The start of South Italian cult. The dating of pottery is normally based red figure has often been linked to the foun- on a typological sequence, and the absolute dation of Thurii in 443 BC, but it seems dates of each typological step are often esti- rather more plausible to assume that the first mated. Most important for the present study workshops were located in or around is to note the extremely large margins that Metaponto, where a kiln containing late fifth must be allowed for when dating is based on century sherds of the Amykos Painter was style or typology. An absolute date can only found in 1973. Simultaneously a workshop in rare instances be provided by the context. existed at Taranto .54

20 H. W. HORSNIES

The banded and black glazed wares are both argued that the efforts to "reconcile archae- clearly derived from Greek types, but the ological evidence with the information sup- bulk of the material found in Southern Italy plied by ancient authors" have created an ar- seems to have been locally produced. Con- tificial hiatus in the archaeological record in trary to the Greek figured pottery, the plain the fifth century. Instead, Yntema argued, wares are found in large numbers and a few the fifth century hiatus is due to a too high very common types are considered to be Leit- date of the Leittypen in indigenous contexts, typen for particular phases. namely the Ionic cups and the Attic kylikes of In the second half of the seventh century "Bloesch type U. The rarity of the imported East Greek pottery types become increasing- and far better dated Attic figured pottery has ly common in Italy, and the Leittypus for the led scholars to maintain the dates proposed relative chronology from the late seventh for the original production of these Leittypen, through the sixth century are the so-called not considering that local production may "Ionian" cups. The basic description and well have continued after the type had gone classification of the "Ionian" cups were made out of use in its area of origin. by Villard & Vallet. The most common form, I fully agree that the "fifth century dark especially in the West, is type B2, dated c.580- age" is a modern construction .62 The archae- 540 BC. 55 The dates proposed by Villard & ology of the period before the "dark age" is Vallet were based on closed contexts con- traditionally treated with the methods of the taining Corinthian and in rare cases Attic prehistorian, while the post "dark age" period pottery, and are commonly accepted. 56 The is heavily influenced by the interpretation of "Ionian" cups were also produced locally, at the archaeological sources, and the dating of Metaponto at least until 500 BC '51 and possi- the grave contexts is normally based on im- bly into the early fifth century.58 ported luxury items. I am convinced that the In the late sixth and particularly the fifth fifth century hiatus will shrink and eventual- century Attic black glazed pottery was ex- ly disappear when it becomes possible to ported to Italy, where Lamboglia classified study sequences of local material. the Attic imports as "pre-campana".59 Since the 1950s no comprehensive study of Attic Ceramiques campaniennes black glazed pottery in Italy has been at- Since its publication in 1981, Morel's tempted. Ceramiques campaniennes has become the stan- Attic kylikes of "Bloesch type C" are the dard reference for Italic black glaze pottery. type most often quoted as dating evidence .60 The study is a morphologically based open The majority of examples of this type in the typology, comprising material from the west- Athenian Agora were found in pre-Persian ern Mediterranean, 63 irrespective of its pro- contexts, and consequently the type is dated duction area.64 Morel provided a good theo- to the generation before 480,61 i.e. c.510-480 retical discussion of the problems of ab- BC. Similarly, tombs in North-Western Luca- solute dating of the black glaze pottery, and nia containing type C kylikes are normally dat- listed the various termini ante and post quems ed to the late sixth or early fifth centuries BC. as well as associations in closed contexts. Morel stressed the problems in dating: "les A "dark age" in the fifth century? repères véritablement sflr sont peu nom- Very few contexts have been dated to the breux ..... 65 The dates of the individual types half century between 475 and 425 BC. This are rarely given with much argument, but: apparent lack of archaeological material dat- "Les datations proposées ici sont en fonction ing to the fifth century is assumed to be the de considerations très diverses, souvent corn- result of the crisis caused by the invading binées entre elles ...", and "Leur degres de "Proto-Lucanians". But in a recent article Yn- precision - nous ne disons pas d'exactitude - tema has questioned the existence of a "dark peut donc varier de quelques années ... a un age" in fifth century Southern Italy. He has siècle ou plus".66

INTRODUCTION 21

Another problem is a possible circular ar- • recent excavations in Roccagloriosa have gument: One of the dating criteria listed is shown that local pottery production did the "abandon d'un grand nombre de sites de continue after the first quarter of the third Lucanie" around 270 BC .67 But it now seems century, e.g. with the thin walled grooved that many sites lived on through the third bowls found both in connection with the century, which invalidates this particular ar- kiln in A5 and above the dumps (including gument (cf. below). the so-called Pozzo) that according to Lucania and Calabria are not well repre- Morel were closed in the first quarter of sented as provenances for the pottery stud- the century. ied by Morel. Almost no examples are quot- • local petite estampilles workshops were pos- ed from inland sites, while the coastal sites tulated by Morel to have been active at in- in modern Southern Campania (Poseidonia ter alia Paestum immediately after the and Pontecagnano) are better represented, foundation of the Roman colony in 273 being the southern fringes of the well stud- BC .77 ied cluster of sites around the Gulf of • Morel has pushed back the initial produc- Naples. Consequently only one local classe tion of the Campana A to c.230-220 BC. was identified in Lucania: the Atelier Paes- turn 68 Apart from Roccagloriosa, several other sites We clearly need single type studies to gain in Lucania, in particular in the Tyrrhenian more knowledge about the development in part, are now reported to have existed until time of the particular type,69 as well as mod- the end of the third century, namely Roscig- ern single site (or regional) studies of the no, Laos, Gallicchio, and Palinuro. Indeed,78 black glaze production of individual pro- the abandonment of Laos in the late third duction areas. Morel asked for both: century happened "certamente in conse- "Presques tous les types que nous définissons quenza degli eventi della guerra annibalica". devraient en réalité faire l'objet, dans une The suggestion that some of the late veritable publication, d'une étude fourth century forms continued to be in use chronologique particulière",71 and: "l'écart during the greater part of the third century numérique entre les productions réper- has recently been confirmed by the excava- toriées et celles qui doivent encore être tion of the "Banquet Hall" at Tricarico. The defines et étudiées est donc considerable" .71 hall was in use in the period between the Some, still unpublished, single type studies Pyrrhic war (280-275 BC) (post quem date) exist and are being used by scholars with ac- and the Second Punic war (218-201 BC), cess to the unpublished material .12 where it was destroyed (both dates provided Since the publication of Ceramiques cam- by coins). Various groups of pottery, clearly paniennes, local black glaze production has fallen from wooden shelves, were found in- been positively identified at for example Pon- side the hall; they included Gnathia and tecagnano,73 Rivello,74 and Roccagloriosa.75 black glaze types that are normally dated to the end of the fourth or the beginning of the Closing the third century gap third century. 79 Consequently we must allow A gap has been postulated between the end for a dating of these types in the third cen- of the local ceramic production - notably tury as well. Structures that are dated to the black glazed (or painted) wares -, consisting period before the Roman conquest of Taran- of mainly Greek-inspired products, and the to in 272 BC because of association with earliest imports of Roman wares (atélier des these types may therefore be later than nor- petites estampilles and Campana A). This gap mally assumed. was created by the assumption that native ar- eas in Southern Italy were deserted after the In light of the possibility that some black fall of Taranto in 272 BC. This view must now glaze types may have had a longer life than be revised :76 normally believed, an attempt to date the