If it is undeniable that friendship was an important concept in both Greek and Roman social thought, offering one of the highest ideals for human life, the picture seems at first sight very different when we turn to Christian attitudes, for the common view appears to be that the arrival and spread of Christianity in the ancient world pushed this ideal out and made it redundant: Christian agape and Classical amicitia were not compatible. As Kierkegaard expresses it, ‘Christianity has pushed earthly love and friendship from the throne, the impulsive and preferential love, the partiality, in order to set spiritual love in its place, the love to one's neighbour, a love which in earnestness and truth and inwardness is more tender than any earthly love… The praise of earthly love and friendship belongs to paganism’, and this view is echoed by many. It may be argued that such a radical contrast is misleading, for in its finest forms even earthly love and friendship can be virtuous, spiritual and intent on truth, as many Classical writers have shown; however, there is a problem in the apparent difficulty of reconciling the Biblical commandment to love your neighbour, so central to Christian ethics, with what seems to be an essential characteristic of friendship, namely exclusivity and partiality in our love for others. Later in this chapter we shall consider how this problem was handled by some of the early Christian writers.

When I first read Augustine's Confessions I was intrigued by the prominence of the theme of friendship. As I explored the literature of the fourth and early fifth centuries it became clear to me that this was not unusual in the writings of the time. Thus I was led to the present investigation in which I aim to see how consistent is the picture of the nature and duties of friendship in a Christian context emerging from the letters, theological and ascetic writings of some of the leading patristic writers. This investigation also necessarily related the Christian ideas closely to the main Classical – Greek and Roman – theories on friendship developed in antiquity which generally formed the basis for the Christian theories and therefore offers further evidence for the relations of pagan and Christian at an important stage in the development of the early Church as it negotiated the move from pagan antiquity to Christian Middle Ages.

I am grateful to Andrew Louth not only for introducing me to the world of the Church Fathers, but for all his advice and guidance along the road which led to this book. I was also assisted by Steven Collins who read it through at an early stage, and by the comments of O. M. T. O'Donovan and P. G. Walsh. I am greatly indebted for practical support to my parents and parents-in-law, and to my friends Richard Ruther-ford, Emily Kearns and Nicholas Purcell for discussion and learned advice on so many occasions.

This book is about how a particular sort of evidence, burials, can be used to write a particular sort of history, that of the social structures of classical antiquity. Most of what I say is based on a simple proposition: a burial is part of a funeral, and a funeral is part of a set of rituals by which the living deal with death. All very obvious, perhaps, but it has one major consequence for the historian. Whether we look at graves with religious, economic, social or artistic questions in mind, the analysis of burials is the analysis of symbolic action.

My argument is as follows. In rituals people use symbols to make explicit social structure, an interpretation of the meaning of daily life. Such structure should be central to any attempt to write social history, but on the whole ancient historians have neglected it. This is largely due to the nature of the written sources, which encourages other approaches. Burials are difficult to interpret, but they can be used to augment the written record, giving us for the first time a dynamic account of social structure and how it changed in antiquity.

These are big claims to make in a small book. Most works of ancient history take little account of graves; even the excellent collection Sources for Ancient History only gives seven pages to the topic. When historians do look at burials, it is usually in a ‘bits-and-pieces’ manner, picking out the spectacular or the supposedly ‘typical’ to illustrate arguments based on texts. It is widely assumed that the rigour which philological historians bring to their sources is only required by archaeologists for dating and classifying material.

Cynics sometimes suggest that ‘progress’ in historiography is no more than a series of mutual misunderstandings, as historians rush to disprove things that others had never intended to say. In writing a methodological book like this I perhaps run even more of a risk of creating the wrong impression than I would in a more substantive work. I have argued that a kind of evidence which most ancient historians ignore is in fact a vital source; and that when historians have used it, it has generally been in the least appropriate ways. Some readers will find this annoying, incomprehensible or not worth the bother; but others, I hope, may see ways to extend their own research. I close with a summary of the points I have tried to make, but first I will set out a series of propositions which I have not made.

WHAT IS NOT BEING SAID

Burials are the ‘best’ source

Classicists who rely mainly on texts and those who rely mainly on archaeological evidence often act as if they were two sides in a competition, and one day an impartial observer will judge whose evidence is best and who wins the game. We have to take our evidence where we can find it; and my argument here has been that we need to combine as many genres as possible. Historians who write about community without using material evidence for rituals and archaeologists who write their monographs without reference to historical concepts both lose out.

‘[A]lmost an embarrassment of riches’, says Rick Jones (1977: 21), in describing a cemetery of 150 Roman graves at Lattes, where thirty-four inscribed tombstones were found in situ. If these few are embarrassing, what of the other 180,000 epitaphs from the western Roman empire, or the 10,000 from classical Attica alone? Inscribed tombstones combine the second and fourth levels of the hierarchy of sources which I set out on p. 10, and sometimes the third as well. By examining ancient decisions to inscribe or not to inscribe a monument and then what to say on it, we should be able to enlarge substantially our understanding of the symbolic construction of society that took place in funeral rituals. But on the whole, this has proved extremely difficult to do. Tombstones are only rarely found in direct association with the burials for which they were set up; most have turned up reused as building blocks or in the diggings of antiquarians, who did not bother about recording contextual details. Consequently, there are major problems in treating them as part of the rites which separated the living from the dead, or in assessing their visual impact in the landscape of later generations. Their standard presentation, as serried ranks of repetitive entries in yellowing tomes known to initiates by obscure acronyms like ICUR or IG II/III2, makes the imaginative leap still greater.

In this chapter I examine several recent attempts to use funerary inscriptions in social history.

In his endless search for ‘the good’, Plato has Socrates run into Hippias of Elis. Hippias quickly finds himself boxed into a philosophical corner, and Socrates gets him to define the good as ‘to be rich, and healthy, and being honoured by the Greeks to come into old age, and after providing a fine end for his own parents, to be buried well and megaloprepōs by his own offspring’ (Plato, Hipp. Maj. 291D–E).

Like all Socrates' opponents, Hippias soon has to admit that this is muddle-headed, but Plato most likely intended it to be an attitude which many readers would share. Herodotus (1.30) had had Solon express similar sentiments, and it may have been a wellknown literary theme. Megaloprepōs is usually translated as ‘magnificently’, as seems to follow from Aristotle's discussion (EN 4.1122a 18–23a). Hippias' position would then be one which most archaeologists assume was widespread in the past. One of the commonest guesses we make is that lavish display in grave goods or monuments means a big person, surrounded in life as well as in death by throngs of admirers. But what the word really means is ‘what is appropriate for a great man’ (Hippias even says that ‘the appropriate’ (to prepon) and ‘the good’ (to kalon) are the same thing [293E]; another blunder, of course, as he works out by 294E). In the next two chapters I concentrate on the Athens of Plato (c. 427–347), give or take a generation or two, suggesting that lavish display was not always prepon for a megas.

This chieftain of the Draynflete Culture, the Via Hernia and Professors Spiggot and Hackenbacker were all products of Osbert Lancaster's wicked sense of humour, but truth, as they say, is stranger than fiction. On 1 August 1984 the appropriately named Andy Mould pulled from a bog in Cheshire a 2000-year-old foot. The police handed the matter over to the archaeologists. Five days later a complete body, now known as Pete Marsh or Lindow Man, was in their hands. How would Lancaster have reacted to the news that one day in March or April in the first century A.D. this tormented soul was stunned by two blows to the head, stabbed in the chest, garrotted and had his throat cut?

Eight hundred years later, Ulpian ruled that a funerary monument was by law ‘something which exists to preserve memory’ (D 11.7.2.6). The continuity is striking, but so too are the changes in the functions of monuments. I use most of this chapter to continue the story of display in classical Athens, concentrating on cycles of lavishness and restraint. Grave goods and grave markers are very different kinds of display, the first seen mainly by those taking part in the funeral, the second by those who come later. A Thasian inscription of about 500 b.c. makes this explicit, saying ‘Whoever was not present when they carried me out in death, let him now lament me; the memorial of Telephanes’ (CEG 159]. However, both forms of display are (with a few exceptions) created by the buriers in the ritual process of disposing of the dead, and can only be understood in terms of each other. I argue that they offer a framework around which to organise a history of social structure.

GRAVE MARKERS AND DEMOCRACY, AND BEYOND

Sixth-century Athenian grave markers are impressive: massive, blocked-out kouroi, fetching millions on the art market, even when their authenticity is suspect; slender stelai topped with sphinxes; or huge mounds like that for Cimon, probably murdered by the tyrant family in 528/7 b.c. (Hdt. 6.103.3).

Nine years ago, I started a Ph.D. thesis about Early Iron Age Greece. I knew that very little literary evidence survived, but it came as something of a shock to discover that there was in fact almost nothing to work with except for brief descriptions of graves. At the first social function for new graduate students, I tried to explain to one of my neighbours, a specialist in Anthony Eden's German policies in the first six months of 1936, what I was planning to write about. He looked confused, and then asked me what a lot of graves had got to do with history. After a long delay I have found an answer, and this book is it.

I believe that burials allow us to go far beyond the limits of textual and iconographic evidence in the study of ancient ritual, and that by studying all aspects of death rites as integrated parts of ritual statements about the actors' perceptions of the world we can reach a new understanding of ancient social structure. The disconcerting experience of having nothing to study but graves turned out not to be such a bad thing after all; I argue that burials provide information of a kind which no other sources provide, and that even in the best documented periods of classical antiquity historians cannot afford to neglect them. I try to make this point with a series of examples drawn from 1,500 years of Greek and Roman history. There will be much for specialists to disagree with in each specific case, but overall I hope that there will be more which stimulates research and proves useful to social historians.

This is a poem that archaeologists are fond of quoting; my excuse for using it is that it is peculiarly relevant to the problems discussed in this chapter. So far, I have looked at grand themes: how burials, in context, help us understand the rise of the polis, democratic Athens, the fall of Rome. I have tried to show that burials are worth study. Now I will give an example of how study can proceed at a much more detailed level. I end not with a bang but with a whimper: no all-embracing model of antiquity in ten pages, but a blow-by-blow analysis of Vroulia, a small site at the southern tip of Rhodes (fig. 40).

Vroulia may seem an odd choice for a closing example. It was excavated in 1907–8 and the site report, published in 1914, is not easy to find. The demographic data are scanty and the textual sources available from Rhodes in the period the site was occupied, c. 625–575 b.c., are worse. Further, the site report itself is in many ways ‘pre-modern’, lacking the conventions which nowadays define authoritative archaeological discourse. But some of these failings are actually advantages for my purposes. Much of the archaeologist's energy goes into interpreting the silences and contradictions of site reports, a complex genre which has to be read as closely as any ancient text.