INTRODUCTION

The data-set on which this study focusses, consists of Mycenaean pottery in areas beyond the Aegean. Catalogue I presents a list of sites in the Mediterranean where such pottery has been found. Considering the large differences between sites and areas in the intensity of archaeological research and publication, this list should be understood as much a result of archaeological research (including my own) as an indicator of the distribution of Mycenaean pottery.

Stylistically, the body of ceramic material under study ranges from LH I to LH IIIB2. The absolute chronology of the Aegean stylistic ceramical phases is by no means undisputed. The chronological frameworks in the Levant, Cyprus and Italy depend to a considerable extent on the stylistical classifications of the Mycenaean pottery. As a result, the interdependant absolute chronological frameworks are all hotly debated. For this reason, I will generally refrain from using absolute dates. A provisional chronological chart based on the traditional view is given in figure 2.1.

THE CHARACTER OF MYCENAEAN POTTERY FOUND OUTSIDE G REECE

Mycenaean pottery, as a class of archaeological artefacts, is very well studied indeed. Already in early studies it was realised that the corpus of Mycenaean pottery found outside the Aegean differs in several respects from that found on the Greek mainland. One of the most readily apparent differences is that wheel-made, decorated fine wares constitute a far larger proportion of the material than in the Aegean. In fact, with a few notable exceptions, undecorated coarse ware and plain fine wares imported from Mycenaean Greece are almost non-existent in most areas. Other differences are visible in the range of pot shapes found in the eastern and central Mediterranean. The occurrence in Cyprus, and to a lesser extent in the Levant, of vessel types which are rare in Greece led to the term Levanto- Helladic ware, to indicate a group of pot shapes believed to have been produced in Cyprus. Shallow bowls (FS 295-296), chalices (FS 278), angular jugs (FS 139), amphoroid kraters (FS 53-55) and zoomorphic rhyta were all believed to belong to this class. The Mycenaean pottery with pictorial decoration, likewise found more often in the eastern Mediterranean than in Greece, was believed to belong to this class of Levanto-Helladic ware as well.

INTRODUCTION

Broglio di Trebisacce is situated 1 km from the coast on a high terrace in the foothills of the Pollino massif at the northern end of the plain of Sybaris. The site is one of more than twenty pre- and protohistoric sites in the hills surrounding the, suggesting that the area was substantially populated from the Middle Bronze Age onwards. Excavations at Broglio began in 1979 and are still continuing. Even though some Neolithic finds have been made, settlement at the site appears to have begun in the Middle Bronze Age and continued into the Early Iron Age.

The site of Broglio di Trebisacce coveres the whole terrace. Five restricted areas have been excavated up to 1985 (fig 17.1). The Areas A and C, were heavily damaged by erosion and ploughing, and the virgin soil appeared directly below the top soil. In Area E, five post holes have been interpreted as the remains of a Middle Bronze Age hut. In the south-eastern part of trench B, a ditch has been discovered with eight post holes and field stones. Together with post holes in the vicinity they may have belonged to a Middle Bronze Age structure, but no pavement or surface has been recognised. In the same area, pits and post holes dating to the Late Bronze Age may represent two or three successive structures. A sequence of terraces has been discovered in trench D, the earliest of which can be dated to the Middle Bronze Age. A short section of a wall as well as pavement dating to an advanced stage of the Middle Bronze Age have been discovered in the eastern and southern parts of the trench.These possible structures may have been the predecessors of a Late Bronze Age structure, of which a ditch with six post holes has been attested. This building was superseded by a structure which was cut into the slope and which has been labeled the ‘central habitation building’, or complesso a monte. This building, which is only partly preserved, had a plan in the shape of a horse-shoe (Fig. 17.2). The few, fragmentary remains of other structures identified at the site all belong to later periods.

MYCENAEAN REPERTOIRES

As is clear from the discussions in chapter 2, the Levant has the most sites with Mycenaean pottery, but the density is highest in Cyprus, as is the absolute number of pots. A common characteristic of the distribution pattern in all areas is that everywhere a large number of sites has produced very few Mycenaean finds, while only a few sites have yielded substantial quantities of it. Several large urban centres along the Levantine coast have produced large amounts of this pottery; however, it also occurs in quantity at sites in the interior of the Levant. This was not the case in Cyprus, where extremely large quantities of Mycenaean pottery have been found exclusively at sites near the coast. Just as the coastal cities of the Levant, all Cypriot urban centres with major deposits of Mycenaean pottery are towns which played important roles in regional and supra-regional trade networks. In Italy, too, the places which have produced substantial quantities of Myceanean pottery can be understood as nodes where international and regional exchange systems coincided. In comparison with the Levant and Cyprus, however, clear shifts in time can be noted with regard to the sites in the central Mediterranean with large quantities of Mycenaean pottery. The unequal presence of large quantities of Mycenaean pottery, which is characteristic of all areas, suggests that everywhere this pottery was imported in only a few places, after which it was distributed through regional exchange networks.

In Cyprus, a few LH I-LH IIA pots have been found, while some contemporary Minoan pottery is also present on the island. An increase in the Aegean imports in the island may be noted during the LH IIB and especially during the LH IIIA1 ceramic phases, of which substantially more finds have been made, even in places far away from the coasts. From LH IIIA2 onwards, large quantities of Mycenaean pottery reached all parts of the island. This remained the case until an advanced stage of LC IIC, when pottery in Aegean style began to be produced on the island itself and Aegean imports diminished.

The present monograph is a revised version of a dissertation written for the department of Classical Archaeology at the University of Amsterdam. The research was made possible by a generous scholarship from the UTOPA foundation. I would like to express my gratitude to the director, Mr L. Dijkman, for providing me with this opportunity and for his pleasant cooperation over the years. The Netherlands Organisation for Scientific Research (NWO) made a travel grant available which allowed me to visit many of the places mentioned in this volume. A publication grant, supplied by the same organisation, made the publication of this book possible.

This study could not have been completed without the active support of a number of friends. Over the years, Professor Joost Crouwel has read my work with alacrity and care. His detailed knowledge of a range of archaeological topics, as well as his consistent emphasis on clarity have contributed greatly to the results of my research. His academic rigour and in his enthusiasm have been a particular source of inspiration.

In the initial phases of my research, the late Vronwy Hankey helped me define the directions and the limitations of my research. My discussions with Vronwy provided an insight into the enormous complexity of the field I was about to enter. I consider myself fortunate to have been one of the many who were able to benefit from her kindness and expertise.

Dr Lucia Vagnetti kindly assisted me with the parts of this study that deal with the central Mediterranean. She also pointed me in the right direction when I was in Italy. Similarly, Professor Albert Leonard helped me with the sections on the Levant and Cyprus. I would like to express my gratitude to both scholars for sharing their expertise and for taking the time to read lengthy instalments.

A number of colleagues have discussed their work with me, shown me sites and material, and have given their views on parts of my research. In particular, I would like to thank Dr P. Attema, Dr M. Bettelli, Dr H. Catling, Dr N. Hirschfeld, Dr B. Kling, Dr G. van der Kooij, Dr S. Levi, Professor M. Marazzi, Mr G. Matteo and Professor R. Peroni.

PREVIOUS RESEARCH

The Syro-Palestinian littoral, commonly known as the Levant, is now taken up by six modern nation states: Turkey, Syria, Lebanon, Israel, Jordan and the Palestine Autonomy. This configuration and the political instability in the region over the last fifty years have influenced archaeological research, and any interpretation of distribution patterns in this area is hazardous. This is true also for the distribution of Mycenaean pottery, which has been found at 111 sites, from Charchemish (site no. 133) in Turkey to Tell es-Shari’a (site no. 243) in northern Sinai (Map 6). The concentration of sites visible in the southern Levant is at least partly caused by the intensity of archaeological research since the foundation of Israel in 1948. Moreover, for a long time research has been difficult at many sites in Lebanon, which has influenced the quality of the information available.

The earliest report, to my knowledge, of Mycenaean pottery in the Levant is given by F.J. Bliss in 1894 who had discovered it at El-Hesy (site no. 231) in Palestine. F.B. Welch named Tell es-Safiyeh (site no. 228) as another site with Mycenaean imports and saw an Aegean influence in Levantine vase shapes and decoration. R.C. Bosanquet probably referred to these publications when he asserted the presence of Mycenaean pottery in southern Palestine. None of these scholars distinguished between actual Mycenaean pottery and so-called Philistine ware, a distinction that was first made by H. Thiersch in 1908.

In the early decades of the twentieth century, several sites with Mycenaean pottery were excavated in southern Palestine, among them Garife (site no. 159), Gezer (site no. 224) and `Ain Shems (site no 227). As a result, D. Fimmen in 1924 was able to list six sites. The discovery of a tomb near Minet el-Beida (site no. 142) in Syria in 1928 and the subsequent excavations by C.F.A. Schaeffer at the site and at nearby Ras Shamra (site no. 141) can be considered a landmark in the research of Mycenaean pottery in the Levant. Not only did these discoveries establish the presence of Mycenaean pottery in the northern Levant, the amounts of that pottery on both sites at the time exceeded that on any other.

INTRODUCTION

Spurs of the southern uplands form an abrupt escarpment in the western part of the Mesaoria, leaving a gap of about 2 km only. Just north of this gap lies the site of Enkomi-Ayios Iakovos, situated some 3 km from the coast. The plundering of tombs seems to have been an important activity at the site already during the 19th century; more systematic research began in 1896, when a team from the British Museum investigated some hundred tombs. Since then, the site has been visited frequently by British, Swedish, French and Cypriot expeditions. Unfortunately, Enkomi has been inaccessible since 1974 due to the occupation of northern Cyprus by Turkish military forces.

The site of Enkomi is defined by a fortification wall, which encompasses an area of ca. 2.5 ha (Fig. 10.1). Late Bronze Age remains have been discovered in every excavated part of the site, but the stratigraphy and absolute dates of the various layers have been heavily debated. Apart from scatter's of MC III and LC I sherds, a few buildings dating to the first settlement period in LC IA have been discovered. Apparently, during this early period a number of fairly large buildings were situated in the settlement area, relatively far away from each other and without a discernible street pattern. The succeeding period LC IIA-LC IIB is relatively poorly documented in terms of settlement architecture, but the city was substantially rebuilt during LC IIC. An important new feature was the defensive wall, which surrounded the city on the northern, western and southern sides (Fig. 10.1). The city wall possessed gates that corresponded with the new street layout of the city, the basis of which is a long north-south street. The streets divide the city area into insulae that were occupied by houses of various sizes. Apart from domestic spaces, there are buildings that can clearly be associated with industrial activities, especially metalworking. Buildings clearly meant for religious purposes dating before the LC III period, have, however, not been identified. This indicates that cult was practiced at Enkomi not in official, public structures, but at group or family level in domestic buildings.

The aim of this chapter is to present the research strategy and research design of the study. As will be further explained below, the overall research strategy of this study is comparable case studies. The chapter therefore starts with a discussion of the choice of cases. This includes both the choice of countries, social-security programmes, and time period. The chapter then goes on to discuss the definition and measurement of the dependent variable, namely retrenchment. The overall aim of the empirical parts of the study is to determine whether or not there is empirical support for the theoretical framework developed in chapter 3. Following King, Keohane & Verba (1994), the best way to investigate the explanatory power of a framework is to derive as many empirically observable implications or empirical hypotheses from it as possible and to see whether these match the actual empirical developments. Therefore, this chapter ends by developing 6 hypotheses about variation in retrenchment across the selected cases.

Denmark and the Netherlands – a most similar systems design

Starting with the country level, the study is based on a most similar systems design (Lijphart 1971; Frendreis 1983). This implies that the two ideal cases for this study would be two countries representing the two different modes of party competition discussed in chapter 3, while at the same time being similar in all other relevant respects. The relevant respects with regard to this study are the institutional and economic factors discussed in chapter 2. Two such ideal countries do not exist. However, as will be argued below, Denmark and the Netherlands are fairly close and have been investigated.

As will be further argued in chapters 6 and 7, the two countries exemplify the two different modes of party competition. Denmark represents the bloc type of party competition. It is either governed by social democratic governments, often including one or several small centre parties, or right-wing governments, mostly involving both the two major right-wing parties, the Conservatives and the Liberals (Damgaard 1992). The Netherlands represents the pivotal type of party competition. The pivotal party is the Christian Democratic Party or cda.

This chapter starts with a brief summary of the main findings of the study. The rest of the chapter aims at discussing a number of implications of these findings at two levels. First, the implications for the scholarly debate about welfare-state retrenchment will be discussed. This includes a discussion of the prospects of generalising the insights of this study beyond the cases studied here. Second, the chapter will address the broader question about the political background for the economic ‘miracles’ in the two countries and the implications that this has for the debate about a country's ability to adjust their models of welfare capitalism to a changed economic environment, and the understanding of the dilemmas of European Social Democracy.

The main findings

The main finding of this study is evidently the importance of party politics for the development of modern welfare states. This study claims that a focus on Dutch and Danish party competition is a must if one wants to know why the Netherlands has generally carried through a larger scale of retrenchment than Denmark. The same goes for the variation over time in the two countries. One must consider how the Dutch pivot system of party competition produced a party consensus around retrenchment. This consensus allowed various Dutch coalition governments to justify retrenchment measures in the eyes of the electorate. One must also consider how, except for the years of economic crisis, the Danish bloc-system of party competition implied a political stalemate in relation to retrenchment during the right-wing bloc rule. However, once the Social Democrats regained power, a party consensus around retrenchment emerged.

If an understanding of the reasons for differences in the extent of retrenchment of the various social-security schemes is desired, the schemes’ connection to the labour market must be considered. It affects governments’ capabilities of justifying retrenchment measures in the eyes of the electorate as well as shapes government incentives for retrenchment. This factor explains why the old-age pension schemes in both countries fared the best in this period, and why the Danish disability-pension scheme fared better than both unemployment benefits and early-retirement benefits.

The aim of this chapter is to take a closer look at Dutch politics and welfarestate retrenchment from 1982 to 1998. Yet, to make sense of what has happened from 1982, it is necessary to start out with a short section on Dutch politics and welfare-state retrenchment in the preceding period, 1973 to 1982. Subsequently, there will be sections describing each of the three types of governments that ruled the Netherlands in this period; the CDA-VVD governments from 1982 to 1989, the CDA-PVDA government from 1989 to 1994, and the PVDA-VVD-d66 governments from 1994.

More specifically, the aim is to establish how party strategies around retrenchment developed in the period in relation to other political parties and in relation to the electorate. For each period, there will be an investigation into the government initiatives concerning the welfare state, the societal reactions, how the governments have tried to legitimise their policies, and the reaction of the opposition. The focus will be on social security. Nevertheless, developments in other policy areas will be included to the extent that they are necessary to make sense of what happened in relation to social security. These issues have been studied using both primary and secondary sources.

Depillarisation and fumbling response to the oil crises

From a historical perspective, Dutch politics are one of the clearest examples of the politics of accommodation inherent in Lijphart's (1969) concept of ‘consociational democracy’. Due to a consensus at the elite level about power sharing, autonomy for the sub-cultures and proportionality, the conflicts stemming from the ‘pillarisation’ of Dutch society never materialised into an adversarial style of politics that could have threatened democratic stability (Lijphart 1979; Gladdish 1991: 33-49; Andeweg & Irwin 1993: 23-49). The consensual style of politics was also clearly visible in relation to party competition. In the two decades following the Second World War, the Netherlands was mainly governed by oversized coalitions. Members of these governments belonged to one or more of the three major denominational parties, KVP, CHU, or ARP, and either the Social Democrats (PVDA) or the Liberals (VVD) (Andeweg & Irwin 1993: 119).

This chapter aims at developing a theoretical framework to explain variation in welfare-state retrenchment, both among countries and across socialsecurity schemes. More precisely, the framework seeks to map the factors influencing government decisions on retrenchment. Theoretically, the framework draws on ideas put forward in a number of the studies referred to above and some insights developed by rational-choice theory.

Political parties: their objectives and incentives for retrenchment

As the framework is centered on political parties, it is worthwhile to start with some theoretical deliberations on them. The literature on political parties points to three different goals: office, votes, and policy (Strøm 1990a; Müller & Strøm 1999b). This framework is based on the assumption that the ultimate goal of political parties is to gain office; that is government power. However, this does not imply that they disregard votes. To gain office, any party needs votes. Consequently, in the following political parties are seen as office- and vote-seekers having office as the ultimate goal. Policy per se is not assumed to be important for political parties. Thus, the framework follows Downs (1957: 28) in arguing that, ‘parties formulate policies to win elections, rather than win elections in order to formulate policies’.

In some situations, the office- and vote-seeking strategies coincide in the sense that the best way to gain office is to gain as many votes as possible. However, in other cases, parties face a trade-off between office and votes, which in effect is a trade-off between the short- and long-run (Müller & Strøm 1999b: 9-11). How parties will respond to this trade-off in a concrete situation is difficult to predict, as it will depend on the actual situation and their time preferences (cf. Müller & Strøm 1999a). Still, as gaining office is the ultimate goal, parties are not expected to sacrifice office for votes repeatedly. In situations where vote- and office-seeking possibilities are unrelated, for instance, because a party is sure to continue in power, political parties are expected to seek as many votes as possible. The reason is that electoral success per se has a number of positive side effects such as a positive spirit within the party and an improvement of the party's bargaining position vis-à- vis other political parties.

This chapter first reports the results of the investigation of the extent of retrenchment in each of the selected social-security schemes carried out as described in chapter 4 and appendix 1. The discussion of each scheme is divided into two parts. First, there will be a short description of the scheme as of 1982. Second, there will be an analysis of developments in the scheme from 1982 to 1998, based on the reported calculations of expected and actual savings. These are reported in tables. The chapter then summarises the findings of variation in retrenchment, and compares them with the hypotheses developed in chapter 4 to see to what extent the theoretical framework developed in chapter 3 has been capable of predicting the actual variation in retrenchment.

Retrenchment in Denmark and the Netherlands

The Danish old-age pension

The Danish old-age pension, ‘folkepension’, was a pension for all citizens aged 67 or more with two types of benefits. A basic amount paid to everyone regardless of other income and an income-tested pension supplement. In 1982, 49% of all pensioners received the full supplement, 19% part of it, and 32% no supplement at all (Danmarks Statistik, Statistiske efterretninger afdeling A 1982: 1282-1283). Compared to the basic amount, the supplement was of limited importance, as the maximum supplement in 1982 amounted to only 18% of the basic amount. Benefits could also be awarded to persons under the age of 67 in two cases. First, single females aged 62-67 were entitled to a pension. Second, if special, social or medical conditions were present, a pension could be awarded to people aged 55-67. In both cases, the basic amount and the pension supplement were subject to an income test. Benefits were taxable as any other income, but single pensioners were entitled to an enhanced personal tax-exemption. Thus, single pensioners living from only the national pension paid no income tax. Benefits were linked to the development in prices and tax financed on a pay-asyou- go basis.

Development of the Danish old-age pension from 1982 to 1998

Table 10 provides an overview of changes in the Danish old-age pension from 1982 to 1998 and their budgetary implications, if measurable. The main conclusion from Table 10 is that the Danish old-age pension has not been retrenched.