1. Introduction

1What are “princely tombs”? The question can be answered in two different ways. The first, and most straightforward, is to treat it as a question about archaeologists’ working definitions: that is, what kind of assemblages do we classify as “princely tombs”? Most archaeologists would probably agree with Patrice Bran’s summary of the situation in the West Mediterranean: «Dans les tombes dites “princières”, le défunt était inhumé accompagné d’un char à quatre roues le plus souvent, d’un service à boisson, de riches parures et parfois d’armes. Parmi ce mobilier, les objets grecs et étrusques constituent un critère indispensable à la reconnaissance du caractère “princier”» (Bran 1987, 96).

2In other parts of the Mediterranean, we need to modify this definition. In Iron Age central Greece, which I discuss in this chapter, Greek objects of course occur in most graves, but chariots are virtually unheard of. But we can nonetheless define a class of rich burials-usually cremations of men, with the ashes placed in a bronze urn, sometimes standing on a bronze tripod, accompanied by weapons, jewelry, Near Eastern imports, and Bronze Age heirlooms, and marked by elaborate vases or tumuli-which have certain similarities to the West Mediterranean princely tombs. Indeed, it has been argued that the princely tomb type was exported as part of an elite lifestyle to Italy by the Greeks who settled Pithekoussai and Cumae in the eighth century, and from there spread to Etruria, Campania, and Latium, and ultimately across the Alps (e. g., d’Agostino 1977; Ridgway 1992, 121-144; Cornell 1994, 89-92).

3But looking at the question this way means making a series of assumptions about a second way in which we might pose it. This second approach is historical. Before we can worry about the kind of collection of artifacts which we should call a “princely tomb” we have to decide what we think this category of burial meant to the people who lived in the Iron Age Mediterranean, and even whether it is legitimate to assume that there was such a category as “princely tomb” at all. Fifty years ago, V. Gordon Childe (1945) pointed out that in many parts of the world, the richest tombs - particularly those commemorating dead men as wealthy warriors - belong to periods when new social systems seem to be taking shape, as in Ur in the time of the Royal Cemetery, Mycenae in the time of the Shaft Graves, or Latium in the Orientalizing period. Childe did not use the expression “princely tomb”, but he clearly had precisely this kind of burial in mind. He suggested that such ostentation is understandable in contexts where violent, new warrior elites are seizing power and trying to legitimize their authority. He argued that their descendants, who inherited stable systems of authority validated by tradition, feel no need for such displays.

4Following in this line of thought, archaeologists often see Mediterranean princely tombs as a stage in a larger evolutionary process, coming between the origins of social stratification and the rise of urbanism and the state (e. g., Ampolo et al. 1980; Champion et al. 1984; Brun 1987; Démoulé 1993). In some parts of the Mediterranean world, princely tombs occur singly or in small groups, which can be seen as belonging to individual rulers or royal families; elsewhere, they occur in larger cemeteries, and which can seen as representing warrior aristocracies.

5This has been a productive way to look at the rich tombs of the Mediterranean, and has led to important advances in understanding. But it also has certain disadvantages. It imposes a single abstract model on the data: it says «this is what such a tomb means, regardless of context (and regardless of the fact that the definition of the “princely tomb” itself varies significantly from place to place)». If we want to construct general models of Mediterranean-wide processes, or even to recognize that there were such vast processes in the Iron Age, we have to operate at a certain level of abstraction. But we also have to face the criticisms which “post-processual” or “contextual” archaeologists level against such grand theories (see particularly Hodder 1992). They object that evolutionists take a category of behavior out of its local context, treating its meaning as given, rather than as something which is constantly being constructed and debated by knowledgeable actors. Evolutionists assume that massive expenditure of energy always has one meaning, that of signaling hierarchy. But postprocessual archaeologists insist that the meaning of hierarchy is itself always constructed in ongoing negotiations, which means that we cannot reduce funerary symbolism to an arbitrary cross-cultural principle of “rank”. We can only understand the ostentation or restraint of funerals in context, within a system of local meanings. Contextual approaches of this kind have had a major impact on studies of the Iron Age of temperate Europe (e. g., Hill/Cumberpatch 1995), but less so on research in the Mediterranean.

6In this paper, I examine the evidence from central Greece in the Iron Age (c. 1200-300 B. C.). Central Greece - the area of the main classical city-states (Fig. 1) - makes a particularly informative case study: not only does it have a rich archaeological record, but from c. 750 B. C. onward literary sources also survive. I argue that the combination of the two types of evidence allows us to see the Greek princely tombs not only as an element in the construction of power relations in a time of rapid social change, but also as part of a long-term religious history. This religious history does not just add the kind of details which would be missing in a purely prehistoric context: it radically transforms the ways in which we can interpret Greek princely tombs. The rich Greek evidence does more than just illustrate the obvious point that any generalizing model will inevitably simplify our understandings and gloss over the interesting details of any specific case. I argue that the standard approaches to princely tombs fundamentally distort our understanding of what was happening in Iron Age Greece.

7In making this argument, I am not trying to provide yet another version of the “archaeologists’ nightmare” approach to interpreting prehistoric burials. Rather, I want to suggest that instead of trying to subsume the details of local archaeological sequences within a single general model, ignoring whatever data do not fit, we would do better to focus on how general, Mediterranean-wide processes played out at the local level, and how these processes interacted with particular structures of class, gender, ethnicity, and cosmology to create highly varied outcomes. In tracing the role of rich burials in the emergence of the Greek city-states, we can develop a more nuanced sense of the range of meanings which “princely tombs” could take on.

8The Greek princely tombs were inextricably bound up with the concept of the hero (hêrôs). Beginning with our earliest literary sources, around 700 B. C., there was general agreement that the hêrôes were an ancient race, which had peopled the earth in the age immediately before that of modern mankind. They were suspended between the gods and the men of later days, in that despite their semi-divine lineage, they were mortal. Only through a glorious death in battle, and subsequent immortalization through the songs of bards and cult honors at their tombs, could they escape their fate, living on forever in the minds of men (Nagy 1979; Vernant 1982). Zeus destroyed the hêrôes in two great wars at Thebes and Troy, and replaced them with modern men (see section 3 below). However, it was still possible for a mortal man to prove through his deeds that he in fact belonged to the race of hêrôes, either by his glorious death in battle or through founding a new city. His funeral promoted him to the status of hêrôs. From Homer onward, there was a standard set of funerary rites which the hero received - cremation, with his bones buried in a metal urn, accompanied by his weapons, and marked by a great mound. Achilles was even said to have sacrificed twelve Trojans over Patroclus’pyre, although Homer criticized this as excessive (Iliad 23. 176). The heroic funeral was so integral to the idea of the hêrôs that in fifth-century Athenian tragedy the urn and mound functioned as metaphors for the hero himself (e. g., Aeschylus, Choephoroi 323-25, 351-53, 686-87, 722-24; cf. Segal 1986, 125-129, on how Sophocles manipulates the image of the empty urn in the Electro).

9In Iron Age central Greece, I argue, the rich warrior burial was not so much “princely” as “heroic”. The distinction was important. In Homer, the hero’s descendants could use his great mound as a source of renewed honor for themselves (Iliad 7. 79-86; 23. 245-58; Odyssey 5. 311; 14. 366-71; 24. 93-95), but this was only one of the options available, and was always contested by a powerful trend toward seeing the tomb of the hero as a communal resource (see Farnell 1921; Nock 1944; Visser 1982). The great warrior, and even more so the founder of a new city (see Malkin 1987), was promoted to semi-divine status. People could pray to him for supernatural aid, and while his descendants could strive to use his greatness to establish special privileges for themselves, his intercession was available to everyone. In a sense, his semi-divinity partially neutralized him as a source of social power for any particular individuals. Putting the dead man on a par with the hêrôes removed him from the games played by mere mortals.

10It would be naive to see central Greek “princely” burials as representing merely an exceptional level of energy expenditure, to be correlated with the degree of control the dead persons family had over the labor of others (an approach pioneered by Binford 1971; cf. Wason 1994, 76-86). Nor were these simply the graves of a warrior aristocracy. No doubt some ancient Greeks did sometimes interpret the graves in these ways. Archaeologists would not exactly be wrong to follow in their footsteps, but they would be guilty of privileging one ancient construction of the social world over others. It was also possible - and, judging from our literary sources, more common - to interpret heroic tombs as a point of contact with higher powers, which benefited everyone. The tomb of a hero, usually in the political assembly place (agora) or near a gate in the city wall, was a place to which everyone turned in times of crisis. We know of relatively few “princely tombs” from Iron Age Greece, and I suggest that this is largely because the complexity of their symbolic associations made them difficult to deploy.

11I argue that understanding “princely tombs” in their local context, and developing models tied specifically to that context, must come before general models of their place in a Mediterranean-wide evolutionary process. In the literature of the Archaic city-states, we see two opposed sets of attitudes toward the kind of aristocracies which a long line of archaeologists since Childe has assumed that “princely” tombs represent. One poetic tradition (all the written sources before 500 are in verse form), what I call the «elitist ideology», celebrated the possibility that luxury could bring aristocrats close to the gods, setting them above their fellow citizens; while what I call the «middling tradition» insisted that luxury was decadent, that all men should hold «middling» (metrios) beliefs, and that restraint was the cardinal virtue. They thought of elitists as would-be tyrants (tyrannoi), unconstitutional and immoral men who grabbed power through violence, and held onto it in the same way. The figure of the tyrant was a stereotype of all that was bad. There were real-life tyrants who seized power in some city-states in the seventh and sixth centuries, but we have no evidence of “princely” tombs or residences from the Archaic period. Kings, whom the Greeks called basileis, feature prominently in the earliest poetry of all, from the end of the eighth century, but by the seventh century basileis had virtually disappeared (with the exception of the unique double kingship at Sparta).

12The principle of civic egalitarianism which made male-citizen democracy possible by 500 B. C. developed out of this background of weak aristocracies and alienated tyrants. In Greece, I suggest, the rarity of “princely” tombs and their association with the category of the hêrôs rather than with that of the basileus is vital for understanding Iron Age social structure. I argue that the tradition of “promoting” great men to heroic status began around 1000 B. C., and played an important part in the great upheavals of the eighth century by making it difficult for would-be princes to establish their power. Instead of seeing the formation of a “princely” aristocracy, the eighth century saw a decisive shift of power toward broad male citizen groups, which became the defining characteristic of Archaic and Classical social structure in central Greece.

13In part 2, I review the archaeological evidence from central Greece between 1200 and 300 B. C. I then turn in part 3 to the literary evidence for aristocracy and sole rulers, and in part 4, I synthesize these data and try to account for the unusual developments in Greece. To do this, I set the Greek evidence into a broader Mediterranean context. Although we can see powerful forces affecting the whole Mediterranean world in the eighth century, I emphasize differences between the Aegean and the both the East and West Mediterranean worlds more than similarities. I do not suggest that the Greek case provides a model which can then be imposed on other areas; rather, it shows the need for detailed investigation of local systems of meaning. I conclude in part 5 by comparing the Greek case briefly with Rome, the other area of Iron Age Europe for which we have a literary record (albeit a highly problematic one). Developments at Rome were very different from those in Greece, but they have enough in common to suggest several conclusions. The first is that the archaeology of resistance needs more emphasis than it has received in Iron Age studies: in both Greece and Rome, popular resistance to the aristocracy and aristocratic resistance to would-be princes were crucial to social dynamics. Secondly, our general models do not just simplify local histories in order to make them fit into a broader pattern: they seriously distort what was happening. We cannot construct cross-cultural models of Mediterranean developments by simply noting the existence of postprocessual criticisms and then continuing with programs of abstraction. We can only generalize about the “princes of protohistory” in the Mediterranean if we first understand the archaeological residues in their local contexts.

2. Greece: the archaeological evidence1

14By 2000 B. C., the first palaces were being built on Crete. Little survives of their earlier phases, but by the middle of the second millennium - by which time palaces were also being built on the Greek mainland - they had much in common architecturally with those of the Near East, and seem to have functioned as redistributive centers. The richest tombs come from the mainland, such as Grave Circle A at Mycenae and the tholos tombs of Messenia. Most of these date to the earlier periods of the mainland palaces (seventeenth and sixteenth centuries). Some tholoi were built as late as the fourteenth century, including the most magnificent of them all, the so-called “Treasury of Atreus” at Mycenae, but after 1400 relatively simple chamber tombs almost completely dominate the record (Dickinson 1994).

15Around 1200, most of the centers of this Late Bronze Age civilization were destroyed. There were major migrations, some into previously marginal areas like Achaea and the Ionian and Cycladic islands, and others overseas to Cyprus and perhaps the Levant. Many Mycenaean high arts disappeared, and the palaces were abandoned. Excavations at Tiryns in the 1970s-80s showed that while the site of the thirteenth-century palace (the “Oberburg”) stood vacant after 1200, a new settlement was established below it on the “Unterburg”, which flourished throughout the twelfth century. The old orthodoxy that 1200 B. C. was a decisive break has been challenged; archaeologists now stress continuities, suggesting that the twelfth century was a golden age for the Mycenaean aristocracy, freed from the demands of the centralized palace authority, but still able to preserve most of what they wanted from their old way of life (see Rutter 1992).

16But the late twelfth century saw a further wave of destructions and the abandonment of many of the new sites on the islands. By 1100, the flourishing settlements on the Tiryns Unterburg and at Lefkandi were in rapid decline, as was the chamber-tomb cemetery at Perati. Contacts with the Near East were dying out, and population levels fell still further. Many of the same crops continued to be cultivated as in Mycenaean times, but settlement contracted rapidly. In the eleventh century, large areas of the countryside were apparently left empty. The northern boundary of “Mediterranean” culture retreated, with more and more Balkan elements appearing in Thessaly after 1100.’

17In central Greece, the eleventh century saw the disappearance of most elements of Mycenaean material culture. Rectilinear multi-room houses were replaced by oval or apsidal one-room structures; multiple burial in chamber tombs was replaced by single burial, usually inhumation in cist graves, but in some regions cremation; offerings stopped at Mycenaean shrines; iron replaced bronze for the weapons and even the ornaments buried in graves; and there was a rapid development away from Mycenaean pottery styles, with Protogeometric vase painting appearing in the Aegean by about 1025.

18By the end of the eleventh century, we can see a new ritual order emerging. Our evidence comes almost entirely from graves. The few excavated houses were flimsy apsidal structures, and the worship of the gods is virtually archaeologically invisible. Mazarakis-Ainian (1997) has suggested that it went on mainly inside chiefs’ houses, leaving few distinctive traces. There are, however, hints of eleventh-century cult activity at some sacred sites, and the fact that some Mycenaean religious centers were also important in Archaic times may mean that worship continued, but in forms that left few physical traces (generally, see Hägg et al. 1988). But in the cemeteries, there are great changes. After the virtual symbolic anarchy of the so-called Sub-mycenaean period (now usually dated c. 1075-1025), the Protogeometric burials of central Greece (c. 1025-900 or 875 in most places) are characterized by restraint and uniformity. By about 1000, gold, ivory, and amber, all of which had to be imported, and even bronze, which depended on imported tin, virtually disappeared from the graves. Nearly all central Greek graves which we can certainly date between 1000 and 925 hold just a few pots and one or two iron objects. Snodgrass (1971, 228-268) suggested that trade with the East Mediterranean collapsed, forcing central Greeks, who had learned about ironworking from Cyprus in the mid eleventh century, to turn to iron for all purposes in the tenth century.

19The one great exception to this pattern is a pair of burials at Lefkandi, dating c. 1000-950 (Popham et al. 1982a; 1993). The bones of a cremated man had been placed in a Cypriot bronze urn, already 200 years old at the time of deposition. His iron weapons were buried with him. Next to this was the inhumation of a woman, adorned with gold jewelry, including an Old Babylonian gorget, one-thousand years older than the burial itself. Four horses had been slaughtered and buried in a separate pit. A large vase stood above the graves, and the entire group was found inside a huge apsidal building, some 50 meters long. At some point probably soon after the funerals, this building was converted into a great tumulus (Fig. 2).

20This group of finds is absolutely unparalleled for the early tenth century, but in the later part of the century we see similar (but less impressive) finds from several sites. At Lefkandi itself, a cemetery grew up around the great tumulus, containing a series of rich graves (Popham et al. 1980; 1982b; Popham/Lemos 1996). One of the earliest, Toumba gr. 49 (c. 950?; Popham et al. 1989), included gold rings and gilt bronze coils at a time when these were still unheard of in the rest of central Greece. By 900, gr. 70 contained five bronze bowls (one of them an engraved and gilded import from the Near East), nine gold rings, six bronze fibulas, a bronze situla, a bronze jug, a faience ring, and two iron pins with crystal globes (Popham 1995); and gr. 79, dating around 875-850, the cremation of a man with his ashes in a bronze bowl, also contained iron weapons and a Syrian cylinder seal made around 1800 B. C. (Popham/Lemos 1995). By the early ninth century, Lefkandi is no longer unique in producing such finds. Graves with Bronze Age heirlooms and/ or Near Eastern imports have also been found at Tiryns, Athens, Argos, and on Naxos; and at the latter site, we also see some kind of cult in honor of the dead, with several parallels at Lefkandi (Lambrinoudakis 1988).

21The increase in the use of bronze and gold grave goods, and also of Near Eastern imports, was a general phenomenon in central Greece, albeit one which is first documented at Lefkandi. The whole Greek world seems to have lost contact with the East around 1025-1000, only to recover it about a century later. Political events in the Levant - particularly the rise and fall of Philistine power - were probably the decisive factors. By 925, Phoenicians were once again voyaging to the central and West Mediterranean, sometimes calling into the Aegean along the way. In the early ninth century some of them may have set up a shrine at Kommos on the south coast of Crete, and even have settled at Knossos (Shaw 1989; Hoffman 1997). Greek pottery once again appears in the Levant around 900, and a new wave of settlement spread across the Cyclades and up the coast of Asia Minor. These new sites have also produced evidence for experiments with domestic space, including multi-room rectilinear houses and enclosure walls around the settled area, and for open-air sacrifice to the gods.

22Grave goods escalated at Athens, Argos, Lefkandi, Knossos, and other sites until about 850 (Coldstream 1977, 55-72), and then declined again. Bronze and gold remained in use, but between 825 and 750 there are far fewer graves with rich offerings than had been the case before 850. The cemeteries at Lefkandi went out of use altogether. Some archaeologists suggest that the decline in imports among late ninth-century grave goods means that Greece again lost touch with the wider world, but since the first substantial evidence for Greek contacts at Al Mina on the Syrian coast may date around 800 (contra Kearsley 1995), this seems unlikely. Rather, funerary rituals changed, and lavish consumption of imports and gold went out of fashion.

23Everything changes around 750. The number and size of settlements increases dramatically, and the numbers of graves and sanctuaries associated with them still more so. Between 750 and 650, multi-room rectilinear houses almost entirely replace oval and apsidal single-room structures in central Greece. Courtyard houses appear as early as 700 at Zagora on Andros (Cambitoglou et al. 1971; Cambitoglou et al. 1988). From the seventh century through the fifth, most houses seem to be roughly the same size, and it has proven very difficult to identify princely palaces, even in cities which we know were ruled by tyrants. At Athens, Shear (1994) suggests that Agora Buildings F, C, and D made up the sixth-century palace of Pisistratus, but even here the evidence is not very strong.

24Enclosure walls become common. In the Greek colony of Megara Hyblaea on Sicily, founded in 728, an open area was set aside for public activities (Vallet et al. 1976). The open feel of earlier Iron Age settlements disappeared; during the seventh century, narrow streets with small doorways leading into courtyard houses became the norm. By 700, even the tiniest hamlet would have a substantial temple, often made of stone, and cities like Corinth or Argos would have a hekatompedon, a “hundred-footer”. Offerings to the gods spiraled rapidly - large numbers of pots by 750, and by 700 metal, sometimes in staggering quantities. Sanctuaries like Olympia, Samos, Ephesus, and Delphi drew visitors from all over Greece, and even from the Near East.

25The variety of graves increased everywhere (Morris 1998). In central Greece, we see two phases in most areas. In the first, graves became more common, the whole age structure of the living population can be seen, and some graves had rich offerings, or were marked by striking monuments like the famous giant pots used in some of the cemeteries of Athens. The forms of burial also proliferated; in Attica, virtually every village had its own variants on the normal adult rite of extended inhumation in an earth-cut pit grave. Different communities went through this phase at slightly different points. It began at Corinth and Eretria by 775, at Athens and Argos by 750, and on some of the islands slightly later. In the second phase, grave goods declined and monuments gradually disappeared. Intramural burial for adults was abandoned. By 700, most of the formative city-states had adopted a new spatial order, with large, homogeneous cemeteries stretching along the main roads leading away from the settlement. Warrior burials almost completely disappeared in central Greece (although not on Crete, in western Greece, or in Thessaly, Macedonia, and Epirus). There were some local variations: Athens and perhaps Aegina only went through part of the second phase, while in Boeotia and eastern Greece we see a direct transition to large homogeneous cemeteries, and even these only became prominent around 550. But overall, by 650 a new funerary order of restraint was in place in the city-states, which lasted for the next quarter of a millennium. There was a slight temporary increase in display between 550 and 500, but the fifth century sees the most uniform and simple cemeteries of any period in Greek history. This restraint and homogeneity began to fragment around 400, and by the end of the fourth century, under the impact of the expansion of Macedonian power, rich warrior graves begin to reappear in central Greece (Morris 1992, 103-155).

26The most “princely” burials in central Greece date around 700, although by comparison with the kind of tombs we find in contemporary Italy or in Hallstatt Europe they are not very impressive. At Eretria, a small group of burials dating around 700 was found near the West Gate, including adult cremations with the ashes placed in bronze urns and children inhumed in coffins. The richest tomb, gr. 6 (possibly a double burial) held two bronze cauldrons, two fragments of gold, a silver ring, fragments of a few burned bronze and iron ornaments, a Levantine sealstone, four iron swords, five iron spearheads, and a bronze spearhead, which was probably a Bronze age heirloom. Although the stratigraphy is not entirely clear, it seems that soon after the last of the burials a triangular stone monument was set up over the graves, which then received cult honors for about a century (Bérard 1970). Argos gr. 45, of about the same date, had been robbed in antiquity, but still contained an almost complete suit of bronze armor (Courbin 1957). The most famous Greek “princely burials”, however, come not from the Aegean but from the Royal Cemetery at Salamis on Cyprus around 700 B. C., with metal vessels, bronze cauldrons and tripods, remains of chariots, and furniture decorated with ivory {e. g., Tombs 1, 2, 79. See Dikaios 1963; Karageorghis 1967; 1974). Here we seem to be dealing with a very different historical trajectory (Snodgrass 1988).

27In conclusion: central Greece has produced few “princely” burials or residences. In the fifty-meter-long apsidal building and associated tombs at Lefkandi we may have examples of both from the early tenth century, and Popham and Lemos (Popham/Lemos 1995, 156) see a continuous tradition of warrior cremations in bronze urns on Euboea, running from this burial through Toumba gr. 79 to the Eretria West Gate graves around 700. There are two peaks of funerary expenditure, one around 900-850, and the other 750-700, but neither of these rival the wealth of the “princely” periods of Etruria, Latium, Campania, or central Europe.

3. Greece: the literary evidence

28The Linear Β tablets, administrative records on clay tablets which were accidentally baked hard when the Mycenaean palaces burned down, show that the palatial administration was headed by an official called the wanax, commonly translated as “king”. While there are disputes over the exact nature of the powers of the wanax, recent research suggests that warfare and religion were vital in legitimating royal authority, and much of the kings power probably came from control over trade and redistribution (Rehak 1995; Laffineur/Niemeyer 1995; Palaima forthcoming).

29The wanax probably disappeared after 1200. The word anax still occurs in Homer’s late eighth-century poetry, but for Homer and Hesiod (probably also dating to the late eighth or early seventh century) the normal word for “ruler” was basileus. This word is etymologically related to pa-si-re-u, a commonly occurring word for some kind of local official in the Linear Β tablets (Carlier 1984, 108-116; Lenz 1993, 92-104). The most popular theory is that after the destruction of the Mycenaean palaces, the pa-si-re-u was the only official whose role still made sense in the contracted world of the Early Dark Age (Andreev 1979; Drews 1983, 109-115; criticized in Lenz 1993, 108-121). The dominant model sees the pa-si-re-u emerging as the “chief” of the small communities of the eleventh and tenth centuries, and dominating Greek society by the eighth century.

30Thucydides (1. 13), writing at the end of the fifth century, seems to have thought in much these terms, arguing that there was once (proteron, “formerly”) a time when there were hereditary kingships (patrikai basileiai) with defined privileges, and later there were tyrants, interested only in their own power. But the most systematic account comes from Aristotle (Politics 1284 b 35-1286 b 40), writing around 330 B. C. Aristotle distinguished five types of kingship (the Spartan kind, the Asiatic kind, the «elected tyrant» [aisymnêtês], the heroic kind, and the absolute monarch [the basileus who is pantôn kyrios, «master of everything»], although, he concluded, in his own day, Spartan and absolute kingship were the only relevant types [1285 b 35]). Like Thucydides, Aristotle suggested that kingship existed proteron, but because of the shortage of truly virtuous men who could exercise such power well, it eventually gave way to aristocracy (1286 b 8-14). He had less to say about tyranny (1295 a 1-24), because, as he said, «it is the least constitutional government» (1293 b 30). He was able to admire the achievements of a few individual tyrants (e. g., 1315 b 2-23; cf. Herodotus 1. 64; 3. 39), but overall he concluded that tyranny was an unnatural form of rule (1287 b 40). Similarly, Plato had argued in the early fourth century that the tyrant negated the civic community, indulging his desires without restraint; and for this, Plato suggested (Republic 616d), he would have to face the worst horrors in the lowest depths of hell.

31Putting together these and other passages, historians have argued that petty chiefs emerged early in the Dark Ages. Homer and Hesiod seem to show that sometimes one basileus would rule a community, and sometimes there would be a college of basileis. At the end of the eighth century, though, the great upheavals which we see in the archaeological record led to groups of aristocrats displacing the basileis. In Archaic times, the word basileus was often relegated to being used a state magistrate with prestigious religious duties. Very few city-states preserved kingship; the most important example was Sparta, where a unique double kingship system survived into the third century. By 650, some members of the aristocracy began to present themselves as the champions of the lower classes against the nobility, and to use popular power to overthrow the internal egalitarianism of the ruling elite and to set themselves up as tyrants. Aristotle argued that it was difficult to preserve a tyranny, and usually within two generations tyrants were overthrown as a result of some elite squabble, often over lovers. By 500, tyrants had vanished from mainland Greece.

32There is a certain amount of agreement among our sources on the basic sequence of events (a progression from constitutional kings to unconstitutional tyrants to oligarchies and democracies), but serious difficulties nevertheless remain. Our information comes from writers active long after the events they are describing, and there is little to suggest that they had access to good primary evidence. Thucydides (1. 1) emphasizes that «I have found it impossible, owing to its remoteness in time, to get a really accurate knowledge of the distant past or even of the period preceding our own times», and Connor (1984, 22) concludes that in his treatment of early times, Thucydides «has constructed an argument, an essay of revisionist history that presents a fresh view of how Greece had once appeared». When Thucydides cites sources, he refers vaguely to «tradition» (1. 4, 9), to analogies with the behavior of contemporary non-Greeks (1. 6), and most often to Homer, who «provides the best evidence» (1. 3). Aristotle may have had a sounder empirical base than earlier writers, having commissioned his students to compile local histories of all the main city-states; but we have no idea how good the information they collected was, or how far Aristotle was prepared to distort it to fit into his philosophical frameworks for the development of Greek politics.

33Ultimately any assessment of historians’models of early Greek political leadership must come down to the interpretation of Homer, who was the major source for Thucydides and Aristotle. Here there is little agreement among historians. The Iliad and Odyssey were probably composed in the late eighth century, describing events in a long-vanished heroic age. By the time of the First World War, Schliemann’s discoveries led most scholars to conclude that this heroic age was in fact the world of the Mycenaean palaces, and that Homer described Late Bronze Age society fairly accurately. But since about 1950, new developments in the study of oral poetry, the decipherment of Linear B, and better knowledge of Iron Age archaeology have forced Homerists to abandon this view. Moses Finley (1954) argued that Homeric society is based on faithful memories of a period somewhere between his own day and the Mycenaean world, perhaps around 900 B. C., but many archaeologists prefer to see the poetic society as an artificial conflation of elements from many different periods (e. g., Snodgrass 1974; Sherratt 1990). Since the 1980s, however, there has been a growing consensus among historians that Homer reflects the social conditions of his own day, in the later eighth century (summarized in Raaflaub 1997). Homer’s basileis are shadowy figures. They generally appear not individually but in groups, leading Finley (1954), Drews (1983), and Carlier (1984) to conclude that the basileis are more “chiefs” or “lords” than “kings”; but in a detailed study, Lenz (1993, 186-256) suggests that we do in fact see strong elements of individual rule in Homer.

34I have argued (Morris 1986) that the nature of oral composition makes it very unlikely that Homer preserved “fossilized” institutions from earlier periods. However, it is equally unlikely that he passively mirrored the situation of the eighth century. He gave his audiences a window onto the lost heroic past, and in so doing he created not a model o/ his own times, but a model for the heroic age - that is, in Homer we see how one eighth-century poet thought the heroic age ought to have been. Interpreting the literary evidence is just as difficult as examining the archaeological record. But I suggest that there are three points worth noting:

1. There was semantic continuity from Mycenaean pa-si-re-u to the Homeric basileus. Since the basileis also appear prominently in Hesiod’s Works and Days and Theogony, it seems reasonable to conclude that basileis were indeed significant leaders in the eighth century. Homer and Hesiod both seem to draw on an older tradition of advice-poetry for basileis (Martin 1984), which again suggests that basileis were a normal political institution not only in the imagined heroic age, but also in the contemporary world.
2. Thucydides and Aristotle believed that the major difference between the early basileus and the Archaic tyrant was that the former ruled constitutionally. He normally inherited his office, but his powers were strictly limited by customary relationships and expectations. For Thucydides and Aristotle, the basileus was someone who governed with the consent of other men who were potential basileis.
3. Basileis had largely disappeared by the seventh century. Few are attested outside Sparta, and Archaic and Classical literature generally represents any form of individual rule as unconstitutional. In the sixth century, Solon and Theognis could use monarkhos, «sole ruler», as a synonym for tyrant; the individual ruler was by definition a man who rejected decency. He could not control his appetites, and would never be satisfied. In agreeing that tyranny was the least constitutional form of rule, Plato and Aristotle followed in a tradition of thought going back far into the Archaic period (see Lanza 1977; Farenga 1981).

35The literary evidence, such as it is, points to the survival of some kind of centralized authority across the Dark Age, headed by basileis, which we might translate as “chiefs”, “princes”, or “kings”, or simply leave in Greek. There is then no reason to see the periods of “princely” burial around 900-850 and 750-700 B. C. as evidence for the emergence of elite groups; rather, they are likely to be related to the transformation of already existing structures of hierarchical power.

4. Greece: “princely burials”, princes, and heroes

36The destruction of the Mycenaean palaces around 1200 ended the age of the wanax, but it did not end Mycenaean culture. During the twelfth century the pa-si-re-u may have come forward as leader in towns like Tiryns, either as one of a group of approximate peers, or in his own right. But it seems likely that this structure would have been changed profoundly by the second set of destructions and migrations around 1100. I have argued (Morris 1991, 27-34) that even in the eleventh century, the population of the largest settlements probably never fell below 1,000 people. While much of the countryside was virtually abandoned, some kind of hierarchy survived at the major centers.

37The Submycenaean cemeteries of the mid-eleventh century suggest that this was a time of collapsing meanings, in which connections to the Bronze Age past were being abandoned, without any very clear alternative taking shape (Morris 1987, 172-173; Whitley 1991, 96-97). It is only at the end of the eleventh century that we once again see highly structured funerary behavior. The new rituals drew lines within and around the community, recreating a sense of order out of the chaos of the recent past (Morris 1987; forthcoming, chapter 9). Within the community, we see new boundaries between adults and children, with the young gradually disappearing from the excavated cemeteries, so that by 900 B. C. no more than 10% of the excavated burials belong to children. The only plausible interpretation is that tenth-century and ninth-century children were buried in ways which have low archaeological visibility. I have argued that the new rituals also drew class lines, between an elite adult group, defined as full members of the community by membership in formal cemeteries, and a larger non-elite group, who were buried in ways which are much harder to detect. Within the elite group, burials are strikingly homogeneous, denying distinctions. Even gender boundaries are poorly defined. Such remains as we have from settlements and the worship of the gods seems consistent: it is as if tenth-century central Greeks went out of their way to avoid any kind of competition, or indeed any kind of activity which would permanently transform the material world.

38Burials also drew sharp lines around tenth-century communities. As Snodgrass suggested (1971, 228-268; see above), by 1000 B. C. there had been a major decline in trade between central Greece and the Near East, and this provides the material background for the decline in gold and bronze which we see in central Greek graves in the early tenth century. But more recent excavations in settlements and chemical analyses of the bronzes show that there was at least some bronze in circulation, and that the near-total dominance of iron in graves was in part a result of deliberate choices. I suggest (Morris forthcoming, chapter 9) that the ritual system which took shape at the end of the eleventh century gave central Greeks a way to make sense of the world they found themselves in. They were surrounded by the ruins of a lost age; every hill and every harbor bore the imprint of the Mycenaean past. Their world had shrunk; some of them even built their rude mud huts in the burned-out shells of Bronze Age palaces. In the twelfth century, some Greeks had tried to preserve the best elements of the old way of life, but that had collapsed by 1050. At the end of the eleventh century, a new, simple, and inward-focused social order was created. Its dominant symbol was iron, a material which cut off the present from the chaos of the bronze-using past, as well as cutting off the locality, where iron ore was abundant, from the broader Mediterranean world, where gold, silver, and bronze had formerly come from. Its dominant themes were stability, hierarchy, and order.

39This is the picture evoked by 99% of our evidence from the earlier tenth century. The two extraordinary graves under the apsidal building at Lefkandi make up the other 1%. These burials are not just grander than the other early tenth-century graves: they completely turn all the symbolism of the period on its head. Where other buriers (including those using contemporary cemeteries at Lefkandi) distanced themselves from the past and the East, these buriers filled the graves with ancient heirlooms from Cyprus and Babylon. Where others emphasized simplicity and homogeneity, these buriers sacrificed horses. Where others marked their graves simply or not at all, these erected a vast tumulus, set on top of a hill, and probably visible for miles around.

40Who could do such a thing? I have argued (Morris forthcoming, chapter 9) that we should understand the new ritual order which emerged around 1025-1000 in terms of myths set down by Hesiod some three hundred years later (Works and Days 106-201). Hesiod tells us that the gods originally made a race of golden men, who lived in peace. But eventually «the earth covered this race», and the gods made a second race, of silver. These were less noble than the gold race, and never sacrificed to the gods, so Zeus destroyed them. He replaced them with a third race, of bronze, «in no way equal to the silver men». The bronze men were violent warriors, who destroyed themselves in war. Zeus then created a fourth race, «more just and nobler, a god-like race of hêrôes who are called demigods». Most of the hêrôes were killed in wars at Thebes and Troy, and Zeus sent the remainder away to live in the Isles of the Blessed. He then created the fifth race, in which Hesiod himself lived. «Now is truly a race of iron», Hesiod explains, «and the gods lay sore trouble upon them». The race of iron lives in constant strife and toil, so that «men never rest from labor and sorrow by day, and from perishing by night... would that I had died before or been born after».

41I suggest that this way of understanding the world took shape in the late eleventh century B. C. It gave the central Greeks a usable past. They saw themselves as belonging to a race of iron, and they gave this symbolic expression in the prominence they gave that material in their rituals. It was a sad decline from a vanished race of heroes, which had filled the landscape with its monuments; but they were not to blame for the situation, which was part of Zeus’plan. We should not see the concept of the heroic race as a distorted reflection of the Mycenaean world; rather, it was an invented tradition, a necessary compliment to the idea of a race of iron. Once the world had been filled with the «god-like race of heroes who are called demigods», but Zeus had taken them away, deliberately starting the wars at Thebes and Troy to destroy them. Now, only a very rare man could show through his deeds that in fact the blood of the heroes flowed in his veins. The Lefkandi burials under the apsidal building are the earliest known example of the process of “heroization” of the recently dead, a tradition which was to continue for a millennium. The great man would normally be cremated, his ashes then placed in a bronze urn and perhaps wrapped in purple cloth. Horses or a whole chariot might be buried with him, along with his weapons, jewelry, and conceivably even human sacrifices; and then the whole complex would be buried under a mound and marked with a stele or great pot. This is the kind of funeral described by Homer for the heroes, and variants on it occur all around the Greek world in later times (e. g., Antonaccio 1995a, 221-243)2.

42A man became a hêrôs by dying. Even in Homer’s world, where all men were by definition hêrôes, a man only truly achieved his heroic status through his kalos thanatos, a «beautiful death» on the battlefield (Vernant 1982). There is nothing to suggest that at any time Greeks thought they were in the presence of living hêrôes: since the heroic funeral was a central part of defining a man as a hêrôs, he had to be dead to achieve this honor.

43This is vitally important for understanding the “princely tombs” of Iron Age Greece. The ritual package which we first see at Lefkandi around 1000-950 elevated the dead man above the status of his peers, but in a very special sense-it made him a demi-god, but also a dead demi-god. It did not reflect the honors and station he had held in life; rather, it was itself the construction of a new status for him. Removing the truly great man to the level of the hêrôes in a sense partly neutralized his glory. The founders of Archaic colonies could be honored like hêrôes after their deaths (see Malkin 1987), and although their descendants could use this to demand special honors in their own right, it was in no sense the same thing as seeing his lavish burial as a claim to “princely” status. A hêrôs was not the same as a basileus, let alone a tyrannos.

44By 900 B. C., many buriers were imitating aspects of the Lefkandi heroic rites (or other early heroic burials which we have not found). In part this must have been because of events in the Near East, which made it possible for the Phoenicians to return to the Aegean, bringing gold, bronze, and other imports which some buriers eagerly seized upon; but buriers also began to place heirlooms in graves and to raise up tumuli, and there were further changes in the age boundaries of the observable burying elite. Whatever the ultimate cause, the years around 900 saw a partial fragmentation of the order of the race of iron. The rich burials of the early ninth century, a kind of “semi-princely tombs”, perhaps do represent attempts to turn heroic status into earthly political power (that seems particularly likely for the people who buried in the Toumba cemetery around the apsidal building at Lefkandi); if so, they failed. Between 850 and 825, the upward spiral of consumption of wealth in central Greek burials ended, and something like the tenth-century ritual order was reasserted. In the late ninth and early eighth centuries, we again see an internally homogeneous and restrained system of burial.

45There is as yet no evidence to help us explain these developments (indeed, precious little evidence of any kind other than burials), but I would speculate that the Phoenicians’ incorporation of central Greece into a larger exchange network around 925-900 disrupted the older homogeneous elite, providing those who wanted them with new ways to claim high standing. For two or three generations, some men perhaps did try to assert a “princely” status, going beyond the customary powers of the basileis. By 850-825, those who defended the traditional order had found ways to incorporate the wider world in a non-disruptive way. The great changes around 900 may have led to new powers being concentrated in the hands of the basileis, but until more settlement evidence is available, we can only speculate.

46Whatever was happening in the late ninth century, in the middle of the eighth the whole Dark Age system collapsed dramatically. The boundaries which funerals drew within the community (and which, if Mazarakis-Ainian is right, worship of the gods reinforced) were swept away. The age structure of late eighth-century cemeteries is close to what we would expect in an ancient society, and I have argued (Morris 1987; cf. Whitley 1991; Houby-Nielsen 1992; 1995) from the variety and structure of the adult burials that around 750 we once again begin to see a cross-section of the whole class structure. The new rituals, and particularly the new sense of space described in section 2, brought the community together where the Dark Age rites had divided it (see also Snodgrass 1980, 15-84; Polignac 1984). The rapid escalation of grave goods, I suggest, was an initial strategy on the part of members of the old elite to preserve distinctions between upper and lower classes, but by 700 such displays had generally been abandoned. Competition shifted to the liminal space of the sanctuary. Here the rich continued to spend lavishly in the seventh century, but in the sixth their activity declined, and the major forms of spending were increasingly dominated by the city-state as a whole (Morris 1998).

47From the beginning of our literary record, around 700, we see a powerful “middling” tradition which insisted that the city-state was a community of fundamentally equal male citizens (Morris 1996). This was the dominant ideology of the Archaic and Classical city-states, and I have argued (Morris 1987; 1998; forthcoming) that it most likely took shape in the great changes of the late eighth century. The new rituals of this period united the civic community, rather than dividing it on class lines, as had been the case earlier in the Dark Age. In the total transformation of the spaces of the living, the dead, and the gods between 750 and 700, we see the material manifestations of a new concept of the state, one which had little or no room for princes.

48This was an important development, which made Archaic Greece radically different from societies in other parts of the Mediterranean world. All our evidence comes to us already implicated in this great social revolution, which means that there is no neutral, direct evidence for the forces behind these changes. Any reconstruction must inevitably be hypothetical, but by putting what we know of Dark Age Greece into a broader Mediterranean context, we can at least identify some of the factors which made Greek social development so different.

49The mid-eighth century was a time of major changes all across the Mediterranean. Everywhere there are signs of rapid population growth, expanding trade, greater wealth, and more intense warfare. In the old kingdoms of the East, local officials and priests were the first to capitalize on the new forces. Assyrian royal power fragmented between 782 and 745, a period Assyriologists call “the interval”. Already in the reign of Adad-Nirari III (810-783) provincial governors were gaining in power at the expense of the crown; and by the 770s, Shamshi-ilu, the governor of Syria, and Shamash-resha-usur, governor of Sukhu and Mari, were virtually independent. In the palace itself, Bel-kharran-beli-usur, the herald, set himself up as a rival source of authority to the kings. Royal building projects ground to a halt, and Urartu emerged as a major military threat. Tiglath-Pileser III (745-727) probably came to power as part of this process of fragmentation. The Eponym Chronicle tells us that there was an uprising in Calah in 746, and two months later, Tiglath-Pileser was on the throne. He apparently understood the forces at work only too well, and succeeded in reestablishing royal power in a series of vigorous military campaigns. He raised the old administrative system to new levels of efficiency, and began a policy of large-scale deportations of peoples, to reduce the risk of future rebellions. The 730s marked a new level of political centralization in Mesopotamia, and some Assyriologists hold that it is only in the days of Tiglath-Pileser that we can really begin to speak of an Assyrian “empire” (e. g., Grayson 1991, 161).

50In Egypt, the whole Third Intermediate Period (1098-730 B. C., contemporary with the Greek Dark Age) was characterized by chronic political disunity, but this reached a head in the middle of the eighth century, often called the time of “Libyan anarchy”. Throughout the ninth century, pharaohs had struggled to maintain control over the priests of Amun at Thebes and Herakleopolis, but by the 820s not only was Upper Egypt escaping from royal control, but even the delta area had fragmented into three warring kingdoms. When Sheshonq III died in 773, a series of Libyan chiefs took advantage of the chaos to challenge the competing Egyptian dynasties. This is one of the most obscure periods in Egyptian political history. The chaos went further in Egypt than in Assyria, and by 728, there were eleven separate, virtually independent kingdoms in Egypt, in addition to a number of “chiefs of the Ma” and powerful priests. None of these competing kings was able to assert himself like Tiglath-Pileser. Instead, beginning in 729, king Piankhy of Napata in Nubia conquered the Nile valley. Tefnakht of Sais was able to impose a degree of unity on the delta in the face of this foreign threat, but in 715 Piankhy’s brother Shabako finally overthrew Tefnakht’s Saite dynasty. For the next forty years the Nubian kings exercised control over the whole of Egypt. Between 674 and 664 Nubians and Assyrians battled for control of Egypt, only for the Assyrian victors to be expelled during the next decade by Psammetichus I, one of their client kings, who enlisted the aid of Greek mercenaries (see O’Connor 1983, 232-249; Kitchen 1986, 334-408; Grimal 1992, 327-359).

51The economic and social factors behind these political events remain obscure (see Postgate 1979; O’Connor 1983; Brinkman 1984a, 1984b; Brink-man/Kennedy 1983; 1986), but the general similarities between the Assyrian and Egyptian cases - political fragmentation in the early and mid eighth century, followed by re-centralization in the late eighth century - suggest that we are dealing with a single process in the Near East. Population was generally increasing, and there is some evidence for an overall growth in wealth; but local markets seem to have been more important than integrated palatial economies, and it is perhaps not surprising that Assyrian and Egyptian kings lost control of their supposed vassals in the 770s. The break-up was not so far-reaching in Assyria as in Egypt. Through systematic violence, Tiglath-Pileser was able to reassert Assyrian royal power after 745, while the Egyptian revival only began in the 720s in the face of Piankhy’s invasions. But all across the Near East, from Elam to Anatolia, the quantity of written records increases dramatically after 720, suggesting a general increase in the ability of central authorities to control lower levels of authority at the end of the eighth century.

52In Italy, by contrast, we see “princely” tombs appearing toward the end of the eighth century. The first of these spectacular burials seems to be the Fondo Artiaco tomb from the Greek colony of Cumae, probably dating around 720; but in the seventh century, Etruscan burials like the Regolini-Galassi tomb at Caere and Latin examples like the Bernardini tomb at Praeneste are just as impressive (the evidence is collected in Spivey/Stoddart 1990; Holloway 1994). These developments are not “explained” by calling them a diffusion of Greek customs, any more than such an explanation accounts for the adoption of Greek wine-drinking technology in Gaul (Dietler, this volume). Etruscans, Campanians, and Latins took over Greek symbolic forms, but reused them in radically new ways (see also Arafat/Morgan 1994). After 700, heroic burials like those at Eretria disappear in the Greek world, but flourish in Italy. Alongside them, we see a second development which has no parallel in Greece: the appearance of “princely residences”, in the seventh century at Murlo and in the sixth at Acquarossa and perhaps at Rome. These great private palaces deploy architectural and iconographic programs which in the Greek world were restricted to temples and other communal buildings. While there is room to argue over what exactly these buildings were, and who lived in them, the combination of burials and houses suggests that in central Italy - unlike central Greece - the eighth and seventh century saw the emergence of a dominating warrior aristocracy. Some scholars couple these finds with evidence for changes in the importance of kinship in funerals, such as the clustering of graves at Osteria dell’Osa and the beginnings of burial in chamber tombs, to argue that powerful Iron Age aristocracies were emerging as the leaders of the descent and patronage groups which our later sources call gentes and were taking control of central Italian towns after 700 B. C. (discussions in Richard 1978, 111-1 15; Franciosi 1984; Bietti Sestieri 1992, 241; Cornell 1994, 84-86).

53In the Aegean, we see a third path of development. Instead of a shift in power toward palaces as in the East Mediterranean, or toward a formative aristocracy as in the West, broad male citizen groups became the dominant force in most central Greek poleis by 700. In each region of the Mediterranean, the way this shared process of population growth and more intense contacts turned out was largely shaped by the details of pre-existing local power structures. Thus in Egypt, local Big Men were able to claim the title of king for themselves, until Piankhy’s invasions compelled them to unite; in Assyria, governors could break away, until crushed by Tiglath-Pileser; and in Italy, warlords could become aristocrats, dominating their clients and struggling among themselves for royal power. There were probably similar openings for individuals to enrich themselves and build up followings in eighth-century central Greece, but in this case, men who tried to do so had to operate in a context where the most important class structure was an internally homogeneous Dark Age elite. The conflicts generated by the new possibilities and problems in Greece were very different from those in other parts of the Mediterranean: would-be princes here found themselves faced by an older elite which insisted that no individual should stand out above all the others - a system perhaps not so far removed from the picture of the ancient basileis which we find in Homer, Thucydides, and Aristotle.

54The overall result of the changes in the eighth century was that sole rulers of any kind virtually disappeared. Aristocratic colleges must have defeated those among them who wished to seize complete power. The place of the hêrôs in Archaic Greek thought may have been important in this process, allowing those who held onto the older conception of an internally egalitarian aristocracy to marginalize any man who seemed to be growing too great. In this time of rapid change, the hêrôes again became very prominent. Besides the explicitly heroizing tombs at Eretria and Salamis on Cyprus, there was also an explosion of cult activity at Mycenaean tombs (Snodgrass 1982; Morris 1988; Whitley 1988; Antonaccio 1995a), the creation of pictorial vase painting for what seem to be mainly heroizing scenes (Snodgrass 1987, 132-169; forthcoming), and quite possibly the creation of the Greek alphabet to record heroic poetry (Powell 1991). As around 900 B. C., profound changes were interpreted through debates over an invented past; it seems reasonable to suggest that once again one of the functions of heroizing burial was to neutralize great men so that they became religious resources for the city-state as a whole, rather than being sources of power for individual aristocratic families.

55We know of very few “princely tombs” in central Greece by comparison with what has been found in contemporary central Italy or Cyprus, and those we do have are relatively poor. The reason for this, I suggest, is that in central Greece the grand burial was implicated in a complex religious history, which made the superman a hêrôs rather than a basileus. Anyone trying to claim a privileged position in the late eighth century by descent from a great hêrôs had to struggle against a long tradition of aristocratic equality and also a growing communal spirit, both of which tried to appropriate the glories of the dead man. As de Polignac (1984, 147) suggests, «tous ces héros sont donc placés au chevauchement de deux mondes, de deux idéologies, de deux systèmes de valeurs militaires et politiques. Le héros fait figure de premier et de dernier champion de la cité».

56The most successful warriors and chiefs of the eighth century perhaps struggled to find ways to combine the semi-divine standing of the hêrôs with the secular power of the basileus to transcend the limits of Dark Age individual power, while other members of the elite resisted any such idea. It was probably in the course of these struggles that the Archaic and Classical concept of the tyrannos took shape (although the word itself was probably a seventh-century borrowing from Lydia [Labarbe 1971]). Any man who tried to be a prince revealed himself to be completely beyond the limits of civilized society, and someone whom all members of the elite should oppose. A few men in the seventh and sixth centuries did get the better of all their rivals, setting themselves up as sole rulers; but aristocratic resistance was so strong that they hardly ever held onto power beyond the second generation. The overall result was that would-be princes of the Italian type were defeated, and relegated to the marginal status of tyrannoi.

57According to the sixth-century poets Solon (fragments 4. 7-8, 23; 9. 3-4; 33; 36. 20-25 [West 1992]) and Theognis (39-52), tyranny was the outcome of the combination of desperate poor men with arrogant rich men. The tyrant is often represented in Archaic poetry as divine retribution for the hubris of the aristocracy (see McGlew 1993, 74-86). Combining these early literary sources with the archaeological record, I suggest that the struggle within the elite to preserve equality against the aggression of tyrants caused the boundaries between the old Dark Age ruling class and the lower classes to collapse. In the modern world, endemic peasant unrest only turns into genuine social revolution in situations where the rulers are internally divided (Skocpol 1979). Eighth-century central Greece was just such a situation. The elite succeeded in defeating the princes and preserving the idea of an internally homogeneous ruling class, but only at the cost of seeing that ruling class generalized to include all adult males. Changes in demography, agriculture, trade, and war affected every part of the Mediterranean in these years, but only in the Aegean did these forces combine with local class structures to cause the partial collapse of the older aristocracy and the formation of a powerful male citizen body rather than the eventual victory of sole rulers.

5. A comparative case: Archaic Latium

58Given such poor evidence, this (or any other) reconstruction must remain conjectural. But the “princely” tombs of Iron Age Greece can only be understood in context, in terms of local mythological and political symbolism, rather than as one example of an evolutionary stage in a Mediterranean-wide pattern. But the very peculiarity of later Greek history could be seen as evidence that the Greek case is unique, of little relevance to archaeologists in other parts of the Mediterranean. The comparative case of Iron Age Latium should dispel this idea. Early Rome, like Greece, has a literary tradition to compliment the archaeological data. The Roman case is very different from the Greek, but it suggests that the symbolism of the rich warrior burial was probably as complex everywhere as it was in the Aegean, and that archaeologists need to put more energy into understanding these finds in local terms.

59As noted in section 4, princely burials appear in Campania in the late eighth century, and in Etruria and Latium in the seventh. No princely tombs have been found at Rome itself, but that is probably the result of post-depositional factors. The main seventh-century cemetery, on the Esquiline hill, was largely destroyed in the 1870s and 1880s by building activity. Pinza (1905) published a body of material from the graves, but Holloway (1994, 21-22) points out that «It is important to remember that these were not archaeological excavations. They were the collection of archaeological material from workmen making the cuttings for streets and sewers», and that «what was found on the building sites has disappeared». Pinza recovered one gold object, two silver rings, bronze arms and armor, jewelry, and eight bronze tripods (1905, 206g, 246a, 562), but nothing on the scale of the finds from Praeneste or Castel di Decima. The quality of the pottery from the Roman burials compares favorably with finds in the Latin “princely tombs” (Sommella Mura 1988), and Pinza (1912, 19) suggested that there was in fact rich gold jewelry in the tombs, which found its way to the flourishing antiquities markets. It remains possible that seventh- and sixth-century Roman chiefs or kings did not celebrate their claims to high status with “princely tombs”, but Pinza’s conclusions seem more likely.

60The settlement at Rome changed dramatically around 600 B. C. Our main evidence from the eighth and seventh centuries consists of foundations for simple one-room oval huts, generally covering an area of 10-15 m^2. Similar huts have been found at other sites in Latium. But in the late seventh century, these huts were abandoned. A much more substantial rectangular stone house, about 10 m. long, has been partially preserved in the Forum, and it seems that there was a major rebuilding project. Part of the forum was paved, and Ammerman (1990) has argued that a vast fill of 10,000-20,000^m3 of earth was dumped there to raise the ground level by 2 m. Probably in the early sixth century, two large buildings were set up. One may be the Curia Hostilia, the first senate house; and the other was the Regia, the residence of the king, a structure which compares in scale and quality with the palaces at Murlo and Acquarossa. A late sixth-century cup found in the Regia had been inscribed rex (king), and an early sixth-century stele from the other end of the forum, although damaged, clearly bore the word recei («to the king»). The overall impression is of powerful sixth-century kings in Rome, organizing impressive urban projects: la grande Roma dei Tarquini, to use Pasquali’s famous expression (see Ampolo 1988; Cristofani 1990; Kuhoff 1995). Later in the sixth century-perhaps around 525-a group of four enormous stone houses was built on the northern slopes of the Palatine, using the atrium-style layout which remained normal at Rome for centuries. House A measures 22 χ 25 χ 31 χ 38 m., far bigger than any contemporary houses in the Greek world. Finally, the first substantial votive deposits begin in the early sixth century on the Capitol and Quirinal, and the earliest temples were built in the Sant’Omobono sanctuary and on the site of the temple of Vesta. At the end of the sixth century, the huge temple of Jupiter (61 x 55 m.), probably one of the biggest temples in the Mediterranean world at the time, was laid out on the Capitoline hill (Gjerstad 1953-63; Cristofani 1990, 29-145; Cornell 1994, 92-97; Holloway 1994, 55-90 and 37-50 on chronology).

61But the archaeological record in Latium is no simpler than that in central Greece. As buildings became more elaborate in the sixth century, “princely tombs”- indeed, tombs of just about any kind-disappear from Latin cemeteries. Very few graves can be dated between 575 and 400. Colonna (1977) has shown that many of the marble ash urns found in the early excavations on the Esquiline hill must date to the fifth century, and suggests that the decline in numbers of burials is simply a result of a shift away from easily datable grave goods. Most archaeologists have followed his theory (e. g., Ampolo 1984; Naso 1990; Holloway 1994, 23; Cornell 1994, 106). It is quite likely that the people digging at Praeneste or on the Esquiline hill in the 1850s-1880s, would have ignored sixth- and fifth-century graves if they were poor (for Palestrina, see Roma 1973, 258-304). But this is not the end of the problem. Not only have recent excavations at Castel di Décima, Laurentina, La Rustica, and Ficana failed to produce burials which can be dated to this period; they have also failed even to produce substantial numbers of burials without grave goods which we could then place in these years. At Castel di Decima, the town was definitely occupied across the sixth and fifth centuries, and Holloway is forced to conclude lamely that «Its later tombs [i. e., post- post-580] must be elsewhere» (1994, 120). If this were the only site lacking sixth- and fifth-century graves, this would be a reasonable argument, but given the virtual absence of this period in any Latin cemetery, it seems that we are dealing with a more complex problem.

62Lavish spending on personal status changed context in early-sixth century Latium, from the cemetery to the house. The process went so far that burials became virtually invisible to the archaeologist. It is of course possible that there was a shift from funerary spending on activities which produced material traces which we can excavate toward those which are archaeologically undetectable. Polybius’famous description (6. 53-54) of second-century B. C. Roman aristocratic funerals might seem to indicate this. Polybius stresses the importance of the display of the masks of ancestors and the speeches delivered over the dead, all of which, he asserts, filled the Roman people with admiration for their aristocratic leaders, but none of which can be excavated. However, prominent Romans of the second century also spent substantial amounts on imposing tombs. Painted chamber tombs returned by the last years of the fourth century, and a large fragment from a tomb painting found in the Esquiline cemetery shows the scale that some of these tombs reached in the third and second centuries (Roma 1973, 188-208). The famous Tomb of the Scipiones was probably dug in the early third century, and substantially enlarged just around the time Polybius was in Rome (Roma 1973, 234-241). The most reasonable interpretation is that spending on funerals declined sharply around 575, to the point that burials of any kind become hard to identify. By 500, the decline in spending became more general: we know of few elaborate fifth-century buildings anywhere in central Italy, and indeed little luxurious material culture of any kind (see Colonna 1990; Pallottino 1991, 97-99). Lavish burials only resumed in the fourth century, with the level of elite display spiraling rapidly from about 300 B. C. onward.

63Our main surviving narratives of sixth- and fifth-century history, by Livy and Dionysius of Halicarnassus, were written in the time of the emperor Augustus (31 B. C. -A. D. 14). Despite major differences in goals and intended audiences (see Gabba 1991; Miles 1995), the two historians more or less agreed on the sequence of events. At the end of the Trojan war, Aeneas of Troy fled to Italy and set up a kingdom at Alba Longa (according to Dionysius [1. 41-44], Hercules had earlier established the principle of legitimate monarchy in Italy). In 753 B. C., Romulus and Remus founded a new city at Rome, and after murdering Remus, Romulus established a dynasty of kings. The city grew and defeated its neighbors, and by the sixth century was a major central Italian power. Its last king, Tarquinius Superbus, ruled like a Greek tyrant, and like many tyrants, was overthrown in 509 B. C. by an oligarchic coup triggered by a sexual escapade.

64The new republic was bitterly opposed to monarchy. Livy (2. 1) has Brutus, leader of the coup, make the people swear that they will never again tolerate a king. However, he insists, it was clear in 509 that the Roman people were not fit to rule themselves, so an oligarchy was created, run by a group of senators. Our sources (Livy 1. 8; Dionysius 2. 8-9; Cicero, On the Republic 2. 16) all say that Romulus had divided the Romans into two status groups: the patres, or patricians, who monopolized political power, and a dependent group of plebs, or plebeians, who, in Dionysius’version, were the clients of the patricians. Varro (On Agriculture 1. 10) and Pliny the Elder (Natural History 18. 7; 19. 50) believed that Romulus had divided the land equally among all Romans, but our major sources do not say this. In 509, the patricians set up a Senate from among their number, and decided that each year two senators would be elected as consuls. They would wield many of the powers of the former kings, but only served for a year. In extreme crises, a dictator could be elected. He was free from all constitutional controls, but could only hold office for six months; and the dictators in fact tried to win glory by resolving the crisis and stepping down from the office as quickly as possible (Livy 4. 14; Dionysius 5. 64-77).

65All the Augustan writers seem to agree that the early Roman patricians were simple and pure in their tastes. Military glory, not luxury, was their goal in life. For Livy (1. 1), as for Sallust (Jugurtha 4-5; Catiline 1-13) writing about twenty years earlier, the most urgent task for a historian was to explain the Romans’moral decline since these days. The most admirable Romans were men like Cincinnatus, who, when called to save Rome from the Aequi in 458 B. C., was busy plowing his tiny farm (Livy 3. 26), or Manius Curius Dentatus, who sat boiling turnips outside his tiny cottage as he received a Samnite embassy in the 290s. Plutarch, who preserves this story, adds that the Samnites «offered him large sums of gold, but he sent them away, telling them that a man who could be content with such a meal did not need gold» (Cato the Elder 2). Similarly, when Rome was overrun by the Gauls in 390, the noble women gladly surrendered their jewelry for the common cause (Livy 5. 25). These upright and stern early Romans competed vigorously for honor, but agreed among themselves that no man should overshadow the other members of the oligarchic college.

66There was rather less agreement on the history of relationships between the patricians and the plebeians. Dionysius seems to feel that from the very beginning of Rome, there were rigid barriers and antagonism over control of the land (Gabba 1991, 187), while Livy (2. 21) suggests that the patricians became much more exploitative after 496, when the last of the exiled Tarquins died, and the oligarchy felt more secure. But there was certainly agreement that rural debt and servitude were such major issues that in 494 the plebeians refused to fight in the army and seceded from the state, forcing the patricians to make concessions. This initiated the so-called “struggle of the orders” (494-287 B. C.). The rich, Livy says, treated the poor as being like children in political terms (3. 21). The state was effectively split in two (Livy 2. 44; 3. 65-67), with the rich claiming to monopolize religion, and eventually even refusing to intermarry with the plebeians (4. 1-6). But after two centuries of class tension, the conflict was resolved with the full integration of the patricians and plebeians, the formation of a new integrated elite of wealth, and the creation of a balanced constitution. According to this theory, set out in detail in the mid second century B.C. by the Greek exile Polybius (6. 3-18), Rome’s power depended on its success in balancing monarchical, aristocratic, and democratic elements, represented by the consuls, the senators, and the popular assemblies respectively. This was a popular vision of Roman history in the first century B. C. (e. g., Cicero, On the Republic 2. 1, 18-21).

67By the 20s B. C., the dominant model was that legitimate kingship gave way to tyranny in the sixth century, followed by a just, harmonious, and virtuous (if sometimes arrogant and exploitative) oligarchy. The patricians gradually worked out a compromise with the plebeians, to produce a perfectly balanced order in the second century, but this collapsed after 133 B. C. into greed, corruption, and civil war.

68There are obvious and tempting points of contact between the archaeological record and this Augustan-era model, and some historians have been quick to claim that recent finds validate the literary tradition. But the written evidence has formidable and well known problems of its own. The most obvious is the date of our authors: Livy and Dionysius wrote nearly 500 years after the fall of the monarchy. Livy described Fabius Pictor as «our earliest authority» for these events (1. 43; cf. 1. 56; 2. 40; 8. 30; 10. 37). Dionysius also relied heavily on Fabius (7. 70-73). But Fabius, the first Roman to write a history, was active around 200 B. C. Several Greek writers had mentioned Rome during the fourth and third centuries, but they did not provide much detailed information. Fabius had no narrative predecessors on whom to draw.

69Nor were there good non-narrative sources. The Twelve Tables, a law code created between 451 and 449 B. C., certainly survived into and beyond Fabius’ times, and he might have consulted the Annales Maximi, chronicles kept by the high priests down to the 130s B. C. and then published. But the primary source which Livy mentions most, what he calls the «Linen Rolls», was only discovered in the early first century (Livy 4. 13, 20). Dionysius (11. 82) also referred to «sacred and secret books», which were probably these same Linen Rolls (Gabba 1991, 86 n. 34). Finally, Livy (2. 37) and Dionysius (1. 87; 2. 54; 3. 1; 4. 26; 10. 32) both refer to ancient inscriptions they had read, but Roman historians generally made little use of such sources.

70Even these meager pre-Fabian texts had their problems. The Linen Rolls and Annales probably recorded only very basic facts, and there were disagreements among first-century historians over what even those meant (e. g., Livy 4. 23; 8. 23). Dionysius (2. 75; 4. 6-7, 30; 7. 1) complains repeatedly about the chronological inaccuracies in his predecessors, who used these lists.

71One of the most disturbing features of the historical record is the tendency for later writers to provide far more detail than earlier ones. Fabius’ History is lost, but an inscription describing its contents suggests that it focused chiefly on the foundation of Rome and on Fabius’own day, passing quickly over the intervening five centuries. Dionysius (1. 6) complains that the earlier writers had too little to say about the Roman kings. But by the first century, historians were finding much more to say. Livy needed five books to relate events down to the Gallic sack of Rome in 390 B. C., while Dionysius took fourteen, and Gellius fifteen (Luce 1977, 155 n. 27; cf. p. 176).

72Possibly these late Republican writers did more thorough research than their predecessors. Livy opens his account (1. 1) by suggesting that writers like to boast that they have uncovered new facts, but little in his writings indicates original archival research. But like most Roman historians, he seems to have relied almost entirely on his historical predecessors of the second and first centuries. Luce (1977, 147) suggests that rather than searching out documentary primary sources, «Livy’s aim was to read through his [literary, secondary] sources with an eye to selecting the one whose version would form the basis of his account... His attitude toward his sources in general was one of circumspect acceptance». In fact, Livy goes so far as to tell us that most of Rome’s records were destroyed in 390 when the Gauls sacked the city, leaving historians little to work with (6. 1). Cicero (On the Republic 2. 33) and Livy (1. 1; 3. 6, 70) agreed that the regal period was hopelessly obscure. Livy suggested that in dealing with these years, the historian could not make the truth his goal (5. 21), and that there was no point in trying to separate myth from fact (1. 1). Dionysius relied even more heavily on myth (e. g., 1. 1, 6, 77).

73The major source for all our historians was probably oral tradition. Some modem writers are optimistic that such traditions would have preserved factual information across the centuries, but the comparative evidence does not make this seem likely, and as Wiseman (1989) points out, we must in any case distinguish carefully between different contexts for the transmission of stories. Livy (8. 40), Dionysius (1. 74), and Cicero (Brutus 62) all complained bitterly about how the nobles distorted the facts to serve family honor, and Livy (7. 9) also accused the early first-century B. C. historian Licinius Macer of exaggerating the prominence of his own clan. The potential for systematic distortion and invention in the long period between the fifth century and the second is enormous.

74For the argument of this chapter, the most important thing about the Augustan accounts of early Rome is the apparent similarity between the attitudes they describe toward monarchy and class relations and those we see in Archaic and Classical Greek writers, which, at first sight, seems like strong evidence that the complexity and ambiguity of Greek notions about “princely tombs” was probably not unique in the Mediterranean. But when we look closely at the Greek-Roman parallels, the problems become even more acute.

75Roman elite culture underwent massive Hellenization in the second and first centuries B. C. as the Romans took political control of the old Greek world. The idea of writing history was itself part of this process; historia was a quintessentiually Greek literary genre, and Fabius Pictor, though Roman, wrote in Greek, for a Greek audience. Some of Rome’s important historical writers, such as Polybius, Dionysius, Diodorus, and Plutarch were themselves Greeks, reading the Roman past through Greek theoretical categories (see especially Momigliano 1975). Hellenism became a major issue in Roman cultural politics, as some members of the elite assimilated themselves to Hellenistic culture, and others resisted it in favor of an idealized image of the “good old days” of Roman purity and simplicity (the most important source is Plutarch, Cato the Elder, see Astin 1978; Gruen 1990; 1992). Some Romans badly wanted to make Rome respectable in Greek eyes by seeing its history as being like that of a Greek city-state. Probably as early as the fourth century, some writers were insisting that Rome had been founded not by Aeneas and the Trojans but by Odysseus and the Greeks (see Wiseman 1995, 50-52).

76Livy was rather cautious in his Hellenism. There is evidence that he modeled some of his stories on the way Greek writers had related similar events (see Luce 1977, 235), and he accepted strong Greek-Roman contacts in very early times, such as the traditions that Tarquinius Superbus had sent an embassy to Delphi (1. 56) and that the elder Tarquin had been the son of the Greek Demaratus (1. 34). However, he was also comfortable criticizing such stories, and he rejected on chronological grounds the idea that king Numa had been a student of Pythagoras (1. 17).

77Dionysius of Halicarnassus is perhaps the most extreme example of a Hellenizing account. Dionysius promoted the fusion of Greeks and Roman into a single ruling class (Gabba 1991, 23-59, 190-216). He recognized a massive increase in Greek elements in Roman culture after 200 (7. 71), but insisted that from its first days, Rome really was a Greek state. He argued that the Italian Aborigines had in fact been immigrants from Greece (1. 11), who were followed by further waves of Greeks (1. 20, 23, 31-33, 41-44). Even Aeneas and his followers, fleeing from the Greek destruction of Troy, turned out to be more Greek than Trojan (1. 61-62). He represented early Roman history as «a slow but continuous prevailing of Greek over autochthonous Italian elements» (Gabba 1991, 110). This Greek element was the key to Roman success (1. 90), and «the motif of the essentially Greek character of Rome as a city... becomes monotonous... [and is] constantly reaffirmed» (Gabba 1991, 153). Elements which other writers saw as typically Italian, such as dictatorship, Dionysius claimed as Greek (5. 73-74, comparing the institution to the Greek aisymnêtês). Where Livy was content merely to note that when the Romans were assembling the Twelve Tables of law in 451-449 they had sent a mission to Athens to consult the laws of Solon (3. 32), Dionysius claimed that the Twelve Tables fused Athenian and Roman principles (2. 27; 10. 55-60).

78The most valuable aspect of Livy’s and Dionysius’ histories is what they tell us about how Romans of the first century B. C. imagined their own origins. Roman ideologies of ethnogenesis were not static; P. M. Martin (1982) and Richard (1987) have shown in detail how the stories of the various kings were constantly reinterpreted across the second and first centuries B. C. in terms of contemporary politics. King Numa, for instance, was converted from a champion of the people into a progenitor for Sulla’s reactionary dictatorship. But on the other hand, neither were the stories which Livy and Dionysius tell completely new inventions of the first century. The main outlines of the narrative, and even many of the meanings assigned to it, were certainly in place by the time Fabius Pictor was writing, around 200. But how much older are these ideas? Did the ways in which the Roman elites of the fifth and fourth centuries saw themselves have anything in common with the Augustan-era ideologies?

79We can say little with confidence. The main Republican institutions of Senate, Assemblies, consuls, and tribunes certainly existed in the third century, and very probably in the fourth. There can be little agreement on the balance of powers between them, and even such a fundamental element of the Livian-Dionysian narrative as the First Secession of the Plebs in 494 may well have been largely invented by projecting the events of 287 back into Rome’s earliest history (Raaflaub 1986). But recently, several historians have argued that popular institutions in the fourth and third century really did exercise some of the powers our sources attribute to them (e. g., Millar 1984; 1989; North 1990). The sixth-century kings certainly existed; but once again there is little agreement on who or what they were. Alföldi (1965) imagined firmly institutionalized Etruscan monarchs ruling Rome, while Momigliano (1989, 97) translated rex as condottiere, and saw the sixth century as a time of roving chieftains at the head of warrior bands.

80The problems are insuperable: as Sailer observes (1991, 159), «The historian who really adheres to the burden of proof... will, I think, have only a very limited narrative left». But we should perhaps not be looking for the archaeological and literary records somehow to confirm or disprove one another. Rather, I suggest, we would do better to follow the postprocessual archaeologists in assuming that our material record is the fragmentary residue of how eighth- through fourth-century Romans constructed images of their world in rituals. Romans of the third through first centuries then transformed these older ideologies to fit their own agendas. The remains of Archaic Roman burials or houses and self-consciously literary productions like Livy’s and Dionysius’ Histories are of course very different categories of evidence, but they form a continuum. In the great crises of the late sixth, late fourth, and late third centuries, the older beliefs and structures probably underwent profound changes. But given the homogeneity and simplicity of the material record from Latium, beginning in cemeteries in the sixth century and spreading to other areas of behavior in the fifth, there is good reason to think that the rituals of this period were indeed characterized by self-restraint; that is, that the stories that Romans of Cato the Elder’s day (234-149 B. C.) were already telling about the simplicity and honor of heroes like Cincinnatus and Coriolanus, and which were so extensively elaborated by Livy and Dionysius, were grounded in the ways that fifth-century elite Romans liked to see themselves. As early as the sixth century, the elites of the Latin towns started to define themselves in their funerals as an equal and restrained community; that, I suggest, was the beginning of the end of Latin monarchy. In such a context, intrigues against a king or his defeat in battle were increasingly likely to lead not to his replacement by another king, but to power passing into the hands of a college of oligarchs. There are some indications that there was a general crisis of monarchy all over Latium (but not Etruria) at the end of the sixth century, and a shift toward republican government (Cornell 1994, 230-232).

81The end of princely tombs in Latium, then, is not a sign of a shift in “levels” of complexity but part of a far-reaching ideological transformation, of new ideas about how aristocrats should act and about the appropriate limits on the power any one man could accumulate.

6. Conclusion

82I conclude by returning to my opening question: what are princely tombs? In the absence of literary sources for most parts of the Mediterranean world, an evolutionary model seeing such burials as an intermediate stage between the origins of social differentiation and proto-urbanization has a certain appeal. We can construct a grand narrative about the history of hierarchy and resistance. But when we have more evidence to put these tombs into a richer context, the distortions which that narrative creates become clear. It is not simply that the rich evidence from central Greece reveals that the evolutionary model glosses over important details; rather, it seems that in the Greek case, and to some extent in the Roman case too, the assumptions of neo-evolutionism are completely misleading. Once we understand the concepts of the hêrôs, tyrannos, and rex, the “princely” tombs of the Iron Age appear in a completely new light. For all the uncertainties of the Roman case, the brief remarks in section 4 should show that there is no reason to assume that the Aegean was the only part of the Mediterranean world in which “princely tombs” were entangled in such symbolic and religious complexities. Current evolutionary archaeology proceeds in much the same way as did nineteenth-century evolutionary anthropology, divorcing cultural traits from their historical context. I am not suggesting that the Greek data undermine the possibility of generalizing about Mediterranean-wide processes, but this evidence does show the need for grand narratives to be built from the ground up, beginning with the details of the local context and the construction of meaning by real people, rather than being handed down from the heights of abstract ethnological reasoning.