The Class Struggle in the Ancient Greek World: From the Archaic Age to the Arab Conquests (G. E. M. de Ste. Croix)

The Class Struggle in the Ancient Greek World: From the Archaic Age to the Arab Conquests
Cover of the first edition
Author	G. E. M. de Ste. Croix
Publisher	Duckworth Books (United Kingdom) Cornell University Press (United States)
First published	1981 London (United Kingdom) Ithaca, New York (United States)
Type	Book
ISBN	0-8014-1442-3 (hardcover) 0-8014-9597-0 (paperback)
Source	https://archive.org/details/classstruggleina0000dest

The Class Struggle in the Ancient Greek World: From the Archaic Age to the Arab Conquests is a book by British ancient historian G. E. M. de Ste. Croix, first published in 1981 by Duckworth Books in Britain and Cornell University Press in the United States. The seminal work attempts a historical materialist analysis of Ancient Greece.

To Margaret

Preface

The main text of this book is intended not only for ancient historians and Classical scholars but also in particular for historians of other periods, sociologists, political theorists, and students of Marx, as well as for 'the general reader'. The use of Greek text and of anything in Latin beyond very brief quotations is reserved for the Notes and Appendices.

As far as I am aware, it is the first book in English, or in any other language I can read, which begins by explaining the central features of Marx's historical method and defining the concepts and categories involved, and then proceeds to demonstrate how these instruments of analysis may be used in practice to explain the main events, processes, institutions and ideas that prevailed at various times over a long period of history – here, the thirteen or fourteen hundred years of my 'ancient Greek world' (for which see I.ii below). This arrangement involves rather frequent cross-referencing. Some of those who are interested primarily in the methodology and the more 'theoretical', synchronic treatment of concepts and institutions (contained mainly in Part One) may wish for specific references to those passages that are of most concern to themselves, occurring either in other sections of Part One or in the more diachronic treatment in Part Two. Similarly, practising historians whose interests are confined to a limited part of the whole period will sometimes need references to a particular 'theoretical' portion in Part One that is specially relevant. (This will, I think, be clear to anyone who compares II.iv with V.ii-iii, for instance, I.iii with IV.ii, or III.iv with Appendix II and IV.iii.)

The book originated in the J.H. Gray Lectures for 1972/73 (three in number), which I delivered at Cambridge University in February 1973 at the invitation of the Board of the Faculty of Classics. I am particularly grateful to J. S. Morrison, President of Wolfson College, then Chairman of the Faculty, and to M. I. (now Sir Moses) Finley, Professor of Ancient History, for their kindness to me and the trouble they took to make the experiences a delightful one for me and to ensure a large audience at all three lectures.

The J. H. Gray lectures were founded by the Rev. Canon Joseph Henry ('Joey') Gray, M.A. (Cantab.), J.P., born on 26 July 1856, Fellow and Classical Lecturer of Queens' College Cambridge for no fewer than 52 years before his death on 23 March 1932, at the age of 75. His devotion to his College (of which he wrote and published a history), to the Anglican Church, and to Freemasonry (he became Provincial Grand Master of Cambridgeshire in 1914) was equalled only by his athletic interests, in rowing, cricket, and above all Rugby football. From 1895 until his death he was President of the Cambridge University Rugby Football Club; and when that club, in appreciation of his presidency, presented him with a sum of no less than £1,000, he used the money to endow a special lectureship in Classics at Cambridge – 'thus making the gladiators of the football field into patrons of the humaner letters', to quote the admiring and affectionate obituary in The Dial (Queens' College Magazine) no.71, Easter Term 1932. The obituary refers to Gray's 'vigorous Conservative politics' and characterises him as 'an almost perfect incarnation of John Bull in cap and gown'. I am afraid he would have disapproved strongly of my lectures, and of this book; but I am comforted by another passage in the same obituary which speaks of his 'hearty goodwill to all men, even to individual socialists and foreigners'.

This book represents of course a very considerable expansion of the lectures, and it incorporates, almost in their entirety, two other papers, given in 1974: a lecture on 'Karl Marx and the history of Classical antiquity', to the Society for the Promotion of Hellenic Studies in London on 21 March 1974, published in an expanded form in Arethusa 8 (1975) 7–41 (here cited as 'KMHCA'); and another lecture, on 'Early Christian attitudes to property and slavery', delivered to the Conference of the Ecclesiastical History Society at York on 25 July 1974, also subsequently expanded and published, in Studies in Church History 12 (1975) 1–38 (here cited as 'ECAPS'). Parts of this book have also been delivered in lecture form at various universities, not only in this country but also in Poland (in June 1977), at Warsaw; and in the Netherlands (in April-May 1978), at Amsterdam, Groningen and Leiden. I have many friends to thank for their kindness to me during my visits to those cities, in particular Professors Iza Bieżuńska-Małowist of the University of Warsaw and Jan-Maarten Bremer of the University of Amsterdam.

I had intended to publish the Gray Lectures almost in their original form, with little more than references added. However, the comments received from most of those to whom I showed drafts convinced me that owing to the extreme ignorance of Marx's thought which prevails throughout most of the West, especially perhaps among ancient historians (in the English-speaking world at least as much as anywhere), I would have to write the book on an altogether different scale. As I did so my opinions developed, and I often changed my mind.

Friends and colleagues have given me some useful criticisms of the many successive drafts of chapters of this book. I have thanked them individually but now refrain from doing so again, because most of them are not Marxists and might not be happy at finding themselves named here, and partly because I do not wish to debar them from being asked to write reviews, as usually happens to those to whom an author makes a general acknowledgment.

I have incorporated very many essential brief references (especially to source material) in the text itself, placing them as far as possible at the ends of sentences. This, I believe, is preferable, in a work not intended primarily for scholars, to the use of footnotes, since the eye travels much more easily over a short passage in brackets than down to the foot of the page and back again. (Longer notes, intended principally for scholars, will be found at the end of the book.) I give this as a reply to those friends who, out of sheer Oxonian conservatism, have objected to the abbreviation of titles by initial letters – e.g. 'Jones, LRE', for A. H. M. Jones, The Later Roman Empire 284-602 – while themselves habitually using such abbreviations for various categories of references, including periodicals, collections of inscriptions and papyri, and so forth, e.g. JRS, CIL, ILS, PSI, BGU. For me, the only alternatives still allowing the use of references in the actual text itself would have been to abbreviate with date or serial number, e.g. 'Jones, 1964' or 'Jones (1)'; but initial letters are as a rule far more likely to convey the necessary information to a reader who already either knows of the existence of the work in question or has looked it up in my bibliography (pp. - below), where all abbreviations are explained. I should perhaps add that titles abbreviated by initials represent books when italicised, articles when not.

My reading for this book, while concentrated above all on the ancient sources and the writings of Marx, has necessarily been very wide; but there are some 'obvious' works which I have refrained from citing – in particular, books which are specifically philosophical in character and which concern themselves primarily with abstract concepts rather than with the actual historical 'events, processes, institutions and ideas' (cf. above) that are the subject-matter of the practising historian. One example is G. A. Cohen's book, Karl Marx's Theory of History, A Defence, based on much greater philosophical expertise than I can command, but which I find congenial; another is the massive work in three volumes by Leszek Kolakowski, Main Currents of Marxism: Its Rise, Growth and Dissolution, which seems to me to have been vastly overpraised, however accurately it may delineate some of the disastrous developments of Marx's thought by many of his followers.

In an interview printed in The Guardian on 22 September 1970 the released Nazi war criminal Albert Speer said that in the Third Reich 'Each Minister was responsible for his own department, and for that only. Your conscience was quiet if you were educated to see things only in your field; this was convenient for everybody.' Our educational system also tends to produce people who 'see things only in their field'. One of the techniques contributing to this is the strict separation of 'ancient history' from the contemporary world. This book, on the contrary, is an attempt to see the ancient Greek world in very close relation with our own and is inspired by the belief that we can learn much about each by careful study of the other.

The dedication of this book expresses the greatest of all my debts: to my wife, in particular for the perfect good-humour and patience with which she accepted my concentration on it for some years, to the neglect of almost everything else. I also wish to record my gratitude to my son Julian for his valuable assistance in correcting the proofs, and to Colin Haycraft for agreeing to publish the book and accomplishing the task with all possible tact and efficiency.

SEPTEMBER 1980

G.E.M.S.C.

PART ONE

I - Introduction

(i) - The plan of this book

My general aim in this book is first (in Part One) to explain, and then (in Part Two) to illustrate, the value of Marx's general analysis of society in relation to the ancient Greek world (as defined in Section ii of this chapter). Marx and Engels made a number of different contributions to historical methodology and supplied a series of tools which can be profitably used by the historian and the sociologist; but I shall concentrate largely on one such tool, which I believe to be much the most important and the most fruitful for actual use in understanding and explaining particular historical events and processes: namely, the concept of class, and of class struggle.

In Section ii of this first chapter, I state how I interpret the expression 'the ancient Greek world', and explain the meaning of the terms I shall be using for the periods (between about 700 B.C. and the mid-seventh century C.E.) into which the history of my 'Greek world' may conveniently be divided. In Section iii I go on to describe the fundamental division between polis and chōra (city and countryside) that plays such a vital role in Greek history after the 'Classical' period (ending at about the close of the fourth century B.C.) which – absurdly enough – is all that many people have in mind when they speak of 'Greek history'. In Section iv I give a brief account of Marx as a Classical scholar and emphasise the almost total lack of interest in Marxist ideas that is unfortunately characteristic of the great majority of scholars in the English-speaking world who concern themselves with Classical antiquity. I also try to dispel some common misconceptions about Marx's attitude to history; and in doing so I compare the attitude of Marx with that of Thucydides.

Chapter II deals with 'class, exploitation, and class struggle'. In Section i I explain the nature and origin of class society, as I understand that term. I also state what I regard as the two fundamental features which most distinguish ancient Greek society from the contemporary world: they can be identified respectively within the field of what Marx called 'the forces of production' and 'the relations of production'. In Section ii I define 'class' (as essentially a relationship, the social embodiment of the fact of exploitation), and I also define 'exploitation' and 'class struggle'. In Section iii I show that the meaning I attach to the expression 'class struggle' represents the fundamental thought of Marx himself: the essence of class struggle is exploitation or resistance to it; there need not necessarily be any class consciousness or any political element. I also explain the criteria which lead me to define Greek (and Roman) society as 'a slave economy': this expression has regard, not so much to the way in which the bulk of production was done (for at most times in most areas in antiquity it was free peasants and artisans who had the largest share in production), but to the fact that the propertied classes derived their surplus above all through the exploitation of unfree labour. (With this section goes Appendix I, dealing with the technical question of the contrast between slave and wage-labourer in Marx's theory of capital.) In Section iv I demonstrate that a Marxist analysis in terms of class, far from being the imposition upon the ancient Greek world of inappropriate and anachronistic categories suited only to the study of the modern capitalist world, is actually in some essentials much the same type of analysis as that employed by Aristotle, the greatest of ancient sociologists and political thinkers. In Section v I consider some types of historical method different from that which I employ, and the alternatives which some sociologists and historians have preferred to the concept of class; and I demonstrate (with reference to Max Weber and M. I. Finley) that 'status' in particular is inferior as an instrument of analysis, since statuses altogether lack the organic relationship which is the hallmark of classes and can rarely if ever provide explanations, especially of social change. In Section vi I consider women as a class in the technical Marxist sense, and I give a brief treatment of the early Christian attitude to women and marriage, compared with its Hellenistic, Roman, and Jewish counterparts.

Chapter III is entitled 'Property and the propertied'. In Section i I begin with the fact that in antiquity by far the most important 'conditions of production' were land and unfree labour: these, then, were what the propertied class needed to control and did control. In Section ii I explain how I use the expression 'the propertied class': for those who were able to live without needing a significant proportion of their time working for their living. (I speak of 'the propertied classes', in the plural, where it is necessary to notice class divisions within the propertied class as a whole.) In Section iii I emphasise that land was always the principal means of production in antiquity. In Section iv I discuss slavery and other forms of unfree labour (debt bondage, and serfdom), accepting definitions of each of these types of unfreedom which now have world-wide official currency. (Appendix II adds some evidence for slave labour, especially in agriculture, in Classical and Hellenistic times.) In Section v I deal with freedmen (an 'order' and not a 'class' in my sense), and in Section iv I discuss hired labour, showing that it played an incomparably smaller part in the pre-capitalist world than it does today and was regarded by members of the propertied class in antiquity (and by many of the poor) as only a little better than slavery.

In Chapter IV I discuss 'Forms of exploitation in the ancient Greek world, and the small independent producer'. In Section i I distinguish between 'direct individual' and 'indirect collective' exploitation, in such a way as to make it possible to regard even many peasant freeholders as members of an exploited class, subject to taxation, conscription and forced services, imposed by the State and its organs. I also explain that those whom I describe as 'small independent producers' (mainly peasants, also artisans and traders) were sometimes not severely exploited themselves and equally did not exploit the labour of others to any substantial degree, but lived by their own efforts on or near the subsistence level. At most periods (before the Later Roman Empire) and in most areas these people were very numerous and must have been responsible for the largest share in production, both in agriculture and in handicrafts. In Section ii I speak specifically of the peasantry and the villages in which they mainly lived. In Section iii ('From slave to colonus') I describe and explain the change in the forms of exploitation in the Greek and Roman world during the early centuries of the Christian era, where the propertied class, which had earlier relied to a great extent on slaves to produce its surplus, came more and more to rely on letting to tenants (coloni), most of whom at about the end of the third century became serfs. Most working freehold peasants were also brought into the same kind of subjection, being tied to the villages of which they were members: I call such people 'quasi-serfs'. (An Appendix, III, gives a large quantity of evidence for the settlement of 'barbarians' within the Roman empire, the significance of which is discussed in Section iii of Chapter IV.) In Section iv ('The military factor') I point out that in the face of external military threat it may be necessary for the ruling class of a society consisting mainly of peasants to allow the peasantry a higher standard of life than it would otherwise have attained, in order to provide a sufficiently strong army; and that the failure of the Later Roman Empire to make this concession induced in the peasantry as a whole an attitude of indifference to the fate of the Empire, which did not begin to be remedied before the seventh century, by which time much of the empire had disintegrated. In Section v I have something to say about the use of the terms 'feudalism' and 'serfdom', insisting that serfdom (as defined in III.iv) can exist quite independently of anything that can properly be called 'feudalism', and ending with a few words on the Marxist concept of the 'feudal mode of production'. In Section vi I recognise briefly the role of small 'independent producers' other than peasants. That completes Part One of this book.

In Part One, then, I am occupied largely with conceptual and methodological problems, in the attempt to establish and clarify the concepts and categories which seem to me to be the most useful in studying the ancient Greek world, above all the process of change which is so obvious when we look at Greek society over the period of thirteen to fourteen hundred years with which this book is concerned.

In Part Two I seek to illustrate the usefulness of the concepts and methodology I have outlined in Part One in explaining not only a series of historical situations and developments but also the ideas – social, economic, political, religious – which grew out of the historical process. In Chapter V ('The class struggle in Greek history on the political plane') I show how the application of a class analysis to Greek history can illuminate the processes of political and social change. In Section i I deal with the Archaic period (before the fifth century B.C.) and demonstrate how the so-called 'tyrants' played an essential role in the transition from hereditary aristocracy, which existed everywhere in the Greek world down to the seventh century, to more 'open' societies ruled either by oligarchies of wealth or by democracies. In Section ii I make a number of observations on the political class struggle (greatly mitigated by democracy, where that form of government existed) in the fifth and fourth centuries, showing how even at Athens, where democracy was strongest, bitter class struggle broke out in the political plane on two occasions, in 411 and 404. In Section iii I explain how Greek democracy was gradually destroyed, between the fourth century B.C. and the third century of the Christian era, by the joint efforts of the Greek propertied class, the Macedonians, and ultimately the Romans. (The details of this process in the Roman period are described in greater detail in Appendix IV.)

Since the whole Greek world came by degrees under Roman rule, I am obliged to say a good deal about 'Rome the suzerain', the title of Chapter VI. After some brief remarks in Section i on Rome as 'The queen and mistress of the world', I give in Section ii a sketch of the so-called 'Conflict of the Orders' in the early Roman Republic, intended mainly to show that although it was indeed technically a conflict between two 'orders' (two juridically distinct groups), namely Patricians and Plebeians, yet strong elements of class struggle were involved in it. In Section iii I notice some aspects of the political situation in the developed Republic (roughly the last three centuries B.C.). In Section iv I briefly describe the Roman conquest of the Mediterranean world and its consequences. In Section v I explain the change of political regime 'From Republic to Principate', and in Section vi I sketch the nature of the Principate as an institution which continued under the 'Later Roman Empire' from the late third century onwards. In my picture of the Later Empire there is much less emphasis than usual upon a supposed change from 'Principate' to 'Dominate'; far more important, for me, is a major intensification of the forms of exploitation: the reduction to serfdom of most of the working agricultural population, a great increase in taxation, and more conscription. I give a characterisation of the position of the emperor in the Principate and the Later Empire and an outline sketch of the Roman upper classes, not forgetting the changes that took place in the fourth century.

Chapter VII is a discussion of 'The class struggle on the ideological plane'. After taking up some general issues in Section i ('Terror, and propaganda'), I proceed in Section ii to discuss the theory of 'natural slavery', and in Section iii the body of thought which largely replaced that theory in the Hellenistic period and continued throughout Roman times, appearing in Christian thought in an almost identical form. Section iv deals with the attitudes to property of the Graeco-Roman world, of Jesus, and of the Christian Church – or rather, churches, for I insist that the term 'the Christian Church' is not a historical but a strictly theological expression. Jesus is seen as a figure belonging entirely to the Jewish chōra, who may never even have entered a Greek polis, and whose thought-world was thoroughly alien to Graeco-Roman civilisation. The chapter concludes with Section v, which attempts a reconstruction of part of the ideology of the victims of the class struggle (and of Roman imperialism), with some attention to 'Resistance literature' (mainly Jewish) and Christian apocalyptic. The best example that has survived is the fable, which is explicitly said by one of its practitioners to have been invented to enable slaves to express their opinions in a disguised form which would not expose them to punishment, although some of the examples turn out to speak not merely for slaves but for the lowly in general, and of course the fable could also be utilised by members of a ruling class to reinforce their position.

The final chapter, VIII, seeks to explain the 'decline and fall' of much of the Roman empire, leading ultimately to the loss of Britain, Gaul, Spain, and north Africa in the fifth century, part of Italy and much of the Balkans in the sixth, and the whole of Egypt and Syria in the seventh – not to mention the Arab conquest of the rest of north Africa and much of Spain in the later seventh and the early eighth century. Section i shows how the ever-increasing exploitation of the vast majority of the population of the Graeco-Roman world by the all-powerful wealthy classes (a tiny minority) first depressed the political and legal status of nearly all those who were not members of my 'propertied class', almost to the slave level. Section ii describes the way in which, from just after the middle of the second century, the fiscal screw was tightened further up the social scale, on the 'curial class', the richer members of the local communities, who were in theory an 'order', consisting of the town councillors and their families, but in practice were virtually a hereditary class, consisting of all those owning property above a certain level who were not members of the imperial aristocracy of senators and equestrians. Section iii is a largely descriptive account of defection to the 'barbarians', assistance given to them, peasant revolts, and indifference to the disintegration of the Roman empire on the part of the vast majority of its subjects. The last section, iv, explains how the merciless exploitation of a great majority for the benefit of a very few finally led to the collapse of much of the empire – a process too often described as if it were something that 'just happened' naturally – whereas in fact it was due to the deliberate actions of a ruling class that monopolised both wealth and political power and governed solely for its own advantage. I show that a Marxist class analysis can provide a satisfactory explanation of this extraordinary process, which proceeded inexorably despite the heroic efforts of a remarkably able series of emperors from the late third century to near the end of the fourth.

The fact that the whole Greek world eventually came under the rule of Rome has often obliged me to look at the Roman empire as a whole, and on occasion at the Latin West alone, or even some part of it. For example, in Chapter VIII 'barbarian' invasions, internal revolts, the defection of peasants and others, and similar manifestations of insecurity and decline have to be noticed whether they happened in the East or in the West, as they all contributed towards the ultimate disintegration of a large part of the empire. Even the settlements of 'barbarians' within the Graeco-Roman world – on a far greater scale than most historians, perhaps, have realised – need to be recorded (for the reasons discussed in IV.iii), although they occurred on a far greater scale in the Latin West than in the Greek East.

(ii) - 'The Ancient Greek World': its extent in space and time

For my purposes 'the Greek world' is, broadly speaking, the vast area (described below) within which Greek was, or became, the principal language of the upper classes. In north Africa, during the Roman Empire, the division between the Greek-speaking and Latin-speaking areas lay just west of Cyrenaica (the eastern part of modern Libya), on about the 19th meridian east of Greenwich: Cyrenaica and everything to the east of it was Greek. In Europe the dividing line began on the east coast of the Adriatic, roughly where the same meridian cuts the coast of modern Albania, a little north of Durazzo (the ancient Dyrrachium, earlier Epidamnus); and from there it went east and slightly north, across Albania, Yugoslavia, and Bulgaria, passing between Sofia (the ancient Serdica) and Plovdiv (Philippopolis) and joining the Danube at about the point where it turns north below Silistra on the edge of the Dobrudja, an area containing several cities on the Black Sea coast that belonged to the 'Greek' portion of the empire, which included everything to the south and east of the line I have traced. My 'Greek world', then, included Greece itself, with Epirus, Macedonia, and Thrace (roughly the southern part of Albania, Yugoslavia, and Bulgaria, and the whole of European Turkey), also Cyrenaica and Egypt, and all that part of Asia which was included in the Roman empire: an area with an eastern boundary that varied from time to time but at its widest included not merely Asia Minor, Syria, and the northern edge of Arabia but even Mesopotamia (Iraq) as far as the Tigris. There were even Greek cities and settlements beyond the Tigris; but in general, it is perhaps convenient to think of the eastern boundary of the Graeco-Roman world as falling on the Euphrates or a little to the east of it. Sicily too was 'Greek' from an early date and became romanised by slow degrees.

The time-span with which I am concerned in this book is not merely (1) the Archaic and Classical periods of Greek history (covering roughly the eighth to the sixth centuries B.C. and the fifth and fourth centuries respectively) and (2) the Hellenistic age (approximately the last three centuries B.C. in the eastern Mediterranean world), but also (3) the long period of Roman domination of the Greek area, which began in the second century and was complete before the end of the last century B.C., when Rome itself was still under a 'republican' form of government. How long one makes the 'Roman Empire' last is a matter of taste: in a sense it continued, as J. B. Bury and others have insisted, until the capture of Constantinople by the Ottoman Turks in A.D. 1453. The Roman 'Principate', as it is universally called in the English-speaking world ('Haut-Empire' is the normal French equivalent), is commonly conceived as beginning with Augustus (Octavian), at or a little after the date of the battle of Actium in 31 B.C., and as passing into the 'Later Empire' ('Bas-Empire') at about the time of the accession of the Emperor Diocletian in 284. In my view the 'Principate' from the first was virtually an absolute monarchy, as it was always openly admitted to be in the Greek East (see VI.vi below); and it is unreal to suppose, with some scholars, that a new 'Dominate' came into being with Diocletian and Constantine, although there is no harm in using, at any rate as a chronological formula, the expression 'Later Roman Empire' or 'Bas-Empire' (see VI.vi ad init.). Many ancient historians like to make a break somewhere between the reign of Justinian in 527-65 and the death of Heraclius in 641, and speak thereafter of the 'Byzantine Empire', a term which expresses the fact that the empire was not centred at the ancient Byzantium, re-founded by the Emperor Constantine in 330 as Constantinople. My choice of a terminal date is dictated, I must admit, by the fact that my own first-hand knowledge of the source material becomes defective after the death of Justinian and largely peters out in the mid-seventh century: for this reason my 'ancient Greek world' ends not much later than the great book of my revered teacher, A. H. M. Jones, The Later Roman Empire 284-602 (1964), which goes down to the death of the Emperor Maurice and the accession of Phocas, in 602. My own terminal point is the Arab conquests of Mesopotamia, Syria, and Egypt in the 630s and 640s. In justification of keeping within the limits I have described I would plead that virtually everything in this book is based upon first-hand acquaintance with original sources. (in one or two places where it is not, I hope I have made this clear.)

I do believe that 'the ancient Greek world' is sufficiently a unity to be worth taking as the subject of this book: if my knowledge of the source material had been more extensive I should have wished to end the story not earlier than the sack of Constantinople by the Fourth Crusade in 1204, and perhaps with the taking of Constantinople by the Ottoman Turks and the end of the Byzantine empire in 1453. The alleged 'orientalisation' of the Byzantine empire was in reality slight. Although the Byzantines no longer commonly referred to themselves as 'Hellenes', a term which from the fourth century onwards acquired the sense of 'pagans', they did call themselves 'Rhōmaioi', the Greek word for 'Romans', a fact which may remind us that the Roman empire survived in its Greek-speaking areas long after it had collapsed in the Latin West – by something like a thousand years in Constantinople itself. By the mid-ninth century we find a Byzantine emperor, Michael III, referring to Latin as 'a barbarous Scythian language', in a letter to Pope Nicholas I. This contemptuous description of the Roman tongue exasperated Nicholas, who repeated the sacrilegious phrase five times over in his reply to Michael (A.D. 865), with indignant comments.

There is a fascinating account of the Greek contribution to the Roman empire and the relationship of the two cultures in A. H. M. Jones's brief article, 'The Greeks under the Roman Empire', in Dumbarton Oaks Papers 17 (1963) 3-19, reprinted in the posthumous volume of Jones's essays edited by P. A. Brunt, The Roman Economy (1974) 90-113.

(iii) - Polis and chōra

In the Archaic and Classical periods, in Greece itself and in some of the early Greek colonies in Italy and Sicily and on the west coast of Asia Minor, the word chōra (χώρα) was often used as a synonym for the agroi (the fields), the rural area of the city-state, the polis (πόλις); and sometimes the word polis itself, in the special limited sense of its urban area, was contrasted with its chōra (see my ECAPS 1, nn.2-3). This usage continued in the Hellenistic period and under Roman rule: every polis had its own chōra in the sense of its own rural area. However, except where a native population had been reduced to a subject condition there was generally, in the areas mentioned, no fundamental difference between those who lived in or near the urban centre of the polis and the peasants who lived in the countryside, even if the latter tended to be noticeably less urbane (less cityfied) than the normal and in the literature produced by the upper classes are often treated patronisingly as 'country bumpkins' (chōritai, for example, in Xen., HG III.ii.31), an attitude which nevertheless allows them to be credited on occasion with superior moral virtues of a simple kind (see Dover, GPM 113-14). Both groups, however, were Greek and participated in a common culture to a greater or lesser degree.

It is hardly possible to give a general definition of a polis that would hold good for all purposes and all periods, and the best we can do is say that a political entity was a polis if it was recognised as such. Pausanias, in a famous passage probably written in the 170s, in the reign of Marcus Aurelius, speaks disparagingly of the tiny Phocian polis of Panopeus, east of Mount Parnassus – 'if indeed you can call it a polis', he says, 'when it has no public buildings [archeia], no gymnasium, no theatre, no market place [agora], and no fountain of water, and where the people live in empty hovels like mountain shanties on the edge of a ravine' (X.iv.1). Yet Pausanias does call it a polis and shows that in his day it was accepted as such.

In those parts of Asia and Egypt into which Greek civilisation penetrated only in the time of Alexander the Great and in the Hellenistic period the situation was very different. In Asia, from at least the time of Alexander (and probably as early as the fifth century B.C., as I have argued in my OPW 154-5, 313-14), the terms of chōra and polis had come to be used on occasion in a recognised technical sense, which continued throughout the Hellenistic period and beyond in Asia and Egypt: in this sense the chōra was the whole vast area not included in the territory administered by any Greek polis, sometimes referred to as the chōra basilikē (royal chōra), it was under the direct, autocratic rule of the kings, the successors of Alexander, and it was bureaucratically administered, while the poleis had republican governments and enjoyed forms of precarious autonomy which differed according to circumstances. (It will be sufficient to refer to Jones, GCAJ, and Rostovtzeff, SEHHW.) Under Roman rule the same basic division between polis and chōra continued, but the bulk of the chōra came by degrees under the administration of particular poleis, each of which had its own chōra (territorium in the Latin West). The cities in the narrow sense were Greek in very varying degrees in language and culture; native languages and culture usually prevailed in the chōra, where the peasants did not normally enjoy the citizenship of the polis that controlled them, and lived mainly in villages, the most common Greek term for which was kōmai (see IV.ii below). Graeco-Roman civilisation was essentially urban, a civilisation of cities; and in the areas in which it was not native, in which it had not grown up from roots in the very soil, it remained largely an upper-class culture: those whom it embraced exploited the natives in the countryside and gave little in return. As Rostovtzeff has said, speaking of the Roman empire as a whole:

The population of the cities alike in Italy and in the provinces formed but a small minority as compared with the population of the country. Civilised life, of course, was concentrated in the cities; every man who had some intellectual interests . . . lived in a city and could not imagine himself living elsewhere: for him the geōrgos or paganus [farmer or villager] was an inferior being, half-civilised or uncivilised. It is no wonder that for us the life of the ancient world is more or less identical with the life of the ancient cities. The cities have told us their story, the country always remained silent and reserved. What we know of the country we know mostly through the men of the cities . . . The voice of the country population is rarely heard . . . Hence it is not surprising that in most modern works on the Roman empire the country and the country population do not appear at all or appear only from time to time in connexion with certain events in the life of the State or the cities' (SEHRE2 I.192-3).

We can therefore agree wholeheartedly with the American mediaevalist Lynn White, when he says:

Because practically all the written records and famous monuments of Antiquity were produced in cities, we generally think of ancient societies as having been essentially urban. They were, in fact, agricultural to a degree which we can scarcely grasp. It is a conservative guess that even in fairly prosperous regions over ten people were needed on the land to enable a single person to live away from the land. Cities were atolls of civilisation (etymologically 'citification') on an ocean of rural primitivism. They were supported by a terrifyingly slender margin of surplus agricultural production which could be destroyed swiftly by drought, flood, plague, social disorder or warfare. Since the peasants were closest to the sources of food, in time of hunger they secreted what they could and prevented supplies from reaching the cities (Fontana Econ. Hist. of Europe, I. The Middle Ages, ed. C. M. Cipolla [1972], at 144-5).

Actually, as we shall see in IV.ii below, the opinion expressed in that last sentence is less true of the Roman empire (including its Greek area) than of other ancient societies, because of the exceptionally effective exploitation and control of the countryside by the imperial government and the municipalities.

A Greek (or Roman) city normally expected to feed itself corn grown in its own chōra (territorium), or at any rate grown nearby: this has been demonstrated recently by Jones, Brunt, and others, and is now beginning to be generally realised. (Classical Athens of course was the great exception to this rule, as to so many others: see my OPW 46-9.) An essential factor here, the relevance of which used often to be overlooked, is the inefficiency and high cost of ancient land transport. In Diocletian's day, 'a wagon-load of wheat, costing 6,000 denarii, would be doubled in price by a journey [by land] of 300 miles'; and, if we ignore the risks of sea transport, 'it was cheaper to ship grain from one end of the Mediterranean to the other than to cart it 75 miles' (Jones, LRE II.841-2; cf. his RE 37). Jones cites evidence from Gregory Nazianzenus and John the Lydian, writing in the fourth and sixth centuries respectively (LRE II.844-5). According to Gregory, coastal cities could endure crop shortages without great difficulty, 'as they can dispose of their own products and receive supplies by sea; for us inland our surpluses are unprofitable and our scarcities irremediable, as we have no means of disposing of what we have or of important what we lack' (Orat. XLIII.34, in MPG XXXVI. 541-4). John complains that when Justinian abolished the public post in certain areas, including Asia Minor, and moreover taxes had to be paid in gold instead of (as hitherto) in kind, 'the unsold crops rotted on the estate . . . , and the taxpayer was ruined . . . , since he could not sell his crops, living far from the sea' (De magistr. III.61). This evidence, as Brunt has rightly observed, 'is perfectly applicable to every preceding epoch of the ancient world and to every region lacking water communications, for there had been no regress in the efficiency of land transport' (IM 704). I would add a reference to an interesting passage in Procopius, Bell. VI (Goth. II) xx.18, describing how, during a widespread famine in northern and central Italy in 538, the inhabitants of inland Aemilia left their homes and went south-east to Picenum (where Procopius himself was), supposing that that area would not be so destitute of food supplies 'because it was on the sea' (cf. IV.ii below and its n.29).

As I shall not have occasion to refer again to transport in the ancient world, I will give here a particularly striking – though rarely noticed – example of the great superiority of water to land transport even in late antiquity. In 359 the Emperor Julian considerably increased the corn supply of the armies on the Rhine and of the inhabitants of the neighbouring areas by having corn which was already customarily shipped from Britain transported up the Rhine by river-boats (Libanius, Orat. XVIII.82-3; Zosimus III.v.2; Amm. Marc. XVIII.ii.3; cf. Julian, Ep. ad Athen. 8, 279d-80a). The fact that transport against the current of the Rhine was, as Libanius and Zosimus realised, much cheaper than carriage on wagons by road is impressive evidence of the inferiority of the latter form of transport. (It is convenient to mention here that the discovery in recent years of further fragments of Diocletian's Price-Edict of A. D. 301 has advanced our knowledge of the relative costs of land and water transport, a subject I cannot discuss here as it deserves.) I will add a reference to the vivid little sketch in Ausonius of the contrast between river-journeys by boat, downstream with oars and upstream with haulage (Mosella 39-44). It is also worth drawing attention to the repeated allusions by Strabo to the importance of river-transport in the countries where rivers were sufficiently navigable – not so much in the Greek lands, of course, as in Spain and Gaul (see esp. Strabo III, pp. 140-3, 151-3; IV, pp. 177-8, 185-6, 189). In 537 the Emperor Justinian recorded with sympathy the fact that litigants involved in appeals, who therefore needed to travel (to Constantinople), had been complaining that they were sometimes prevented from coming by sea owing to the unfavourable winds or by land owing to their poverty – another testimony to the greater cost of land journeys (Nov. J. XLIX. praef. 2). Yet sea voyages could sometimes involve long delays, because of rough weather or unfavourable winds. The official messengers who brought a letter from the Emperor Gaius to the governor of Syria at Antioch at the end of A.D. 41 are said by Josephus (no doubt with some exaggeration) to have been 'weather-bound for three months' on the way (BJ II.203). In 51 B.C., when Cicero was travelling to Asia to take over his province of Cicilia, it took him five days to sail from Peiraeus to Delos and another eleven days to reach Ephesus (Cic., Ad Att. V.xii.1; xiii.1). Writing to his friend Atticus after reaching Delos, he opened his letter with the words, 'A sea journey is a serious matter [negotium magnum est navigare], and in the month of July at that' (Ad Att. V.xii.1). On his way home in November of the following year, Cicero spent three weeks on the journey from Patras to Otranto, including two spells of six days each on land, waiting for a favourable wind; some of his companions, who risked the crossing from Cassiope on Corcyra (Corfu) to Italy in bad weather were shipwrecked (Ad fam. XVI.ix.1-2).

In point of fact, even the availability of water-transport, in the eyes of Greeks and Romans, could hardly compensate for the absence of a fertile chōra. I should like to refer here to an interesting text, seldom or never quoted in this connection, which illustrates particularly well the general realisation in antiquity that a city must normally be able to live off the cereal produce of its own immediate hinterland. Vitruvius (writing under Augustus) has a nice story – which makes my point equally well whether it is true or not – about a conversation between Alexander the Great and Deinocrates of Rhodes, the architect who planned for Alexander the great city in Egypt that bore (and still bears) his name, Alexandria, and became, in Strabo's words, 'the greatest place of exchange in the inhabited world' (megistron emporion tēs oikoumenēs, XVII.i.13, p.798). In this story Deinocrates suggests to Alexander the foundation on Mount Athos of a city, a civitas – the Greek source will of course have used the word polis. Alexander at once inquires 'whether there are fields around, which can provide that city with a food supply'; and when Deinocrates admits that the city could only be supplied by sea transport, Alexander rejects the idea out of hand: just as a child needs milk, he says, so a city without fields and abundant produce from them cannot grow, or maintain a large population. Alexandria, Vitruvius adds, was not only a safe harbour and an excellent place of exchange; it had 'cornfields all over Egypt', irrigated by the Nile (De architect. II, praef. 2-4).

Now the civilisation of old Greece had been a natural growth ('from roots in the very soil', to repeat the phrase I used above); and although the cultured gentleman, living in or near the city, could be a very different kind of person from the boorish peasant, who might not often leave his farm, except to sell his produce in the city market, yet they spoke the same language and felt that they were to some extent akin. In the new foundations in the Greek East the situation was often quite different. The upper classes, living in or very near the towns, mostly spoke Greek, lived the Greek life and shared in Greek culture. Of the urban poor we know very little, but some of them were at least literate, and they mixed with the educated classes and probably shared their outlook and system of values to a very considerable extent, even where they did not enjoy any citizen rights. But the peasantry, the great majority of the population, on whose backs (with those of the slaves) the burden of the whole vast edifice of Greek civilisation rested, generally remained in much the same state of life as their forefathers: in many areas the majority probably either spoke Greek not at all or at best imperfectly, and most of them remained for centuries – right down to the end of Graeco-Roman civilisation and beyond – at little above the subsistence level, illiterate, and almost untouched by the brilliant culture of the cities. As A. H. M. Jones has said:

The cities were . . . economically parasitic on the countryside. Their income consisted in the main of the rents drawn by the urban aristocracy from the peasants . . . The splendours of civic life were to a large extent paid for out of [these] rents, and to this extent the villages were impoverished for the benefits of the towns . . . The city magnates came into contact with the villagers in three capacities only, as tax collectors, as policemen, and as landlords (GCAF 268, 287, 295).

This of course is as true of much of the Roman West as of the Greek East, and it remained true of the greater part of the Greek world right through the Roman period. The fundamental relationship between city and countryside was always the same: it was essentially one of exploitation, with few benefits given in return. This is brought out most forcibly by a very remarkable passage near the beginning of the treatise On wholesome and unwholesome foods by Galen, the greatest physician and medical writer of antiquity, whose life spanned the last seventy years of the second century of the Christian era and who must have written the work in question during the reign of Marcus Aurelius (161-80) or soon afterwards, and therefore during or just after that Antonine Age which has long been held up to us as part of that period in the history of the world during which, in Gibbon's famous phrase, 'the condition of the human race was most happy and prosperous' (DFRE 1.78). Galen, setting out to describe the terrible consequences of an uninterrupted series of years of dearth affecting 'many of the peoples subject to Roman rule', draws a distinction, not expressly between landlords and tenants, or between rich and poor, but between city-dwellers and country folk, although for his purposes all three sets of distinctions must obviously have been much the same, and it would not matter much to him (or to the peasantry) whether the 'city-dwellers' in his picture were carrying out their exactions purely as landlords or partly as tax-collectors.

Immediately summer was over, those who live in the cities, in accordance with their universal practice of collecting a sufficient supply of corn to last a whole year, took from the fields all the wheat, with the barley, beans and lentils, and left to the rustics [the agroikoi] only those annual products which are called pulses and leguminous fruits [ospria te kai chedropa]; they even took away a good part of these to the city. So the people in the countryside [hoi kata tēn chōran anthrōpoi], after consuming during the winter what had been left, were compelled to use unhealthy forms of nourishment. Through the spring they ate twigs and shoots of trees, bulbs and roots of unwholesome plants, and they made unsparing use fo what are called wild vegetables, whatever they could get hold of, until they were surfeited; they ate them after boiling them whole like green grasses, of which they had not tasted before even as an experiment. I myself in person saw some of them at the end of spring and almost all at the beginning of summer afflicted with numerous ulcers covering their skin, not of the same kind in every case, for some suffered from erysipelas, others from inflated tumours, others from spreading boils, other had an eruption resembling lichen and scraps and leprosy.

Galen goes on to say that many of these wretched people died. He is dealing, of course, with a situation which in his experience was evidently exceptional, but, as we shall see, enough other evidence exists to show that its exceptional character was a matter of degree rather than of kind. Famines in the Graeco-Roman world were quite frequent: various modern authors have collected numerous examples.

There is one phenomenon in particular which strongly suggests that in the Roman empire the peasantry was more thoroughly and effectively exploited than in most other societies which rely largely upon peasant populations for their food supply. It has often been noticed (as by Lynn White, quoted above) that peasants have usually been able to survive famines better than their town-dwelling fellow-countrymen, because they can hide away for themselves some of the food they produce and may still have something to eat when there is starvation in the towns. It was not so in the Roman empire. I have just quoted a very remarkable passage in Galen which speaks of 'those who live in the cities' as descending upon their chōra after the harvest, in time of dearth, and appropriating for themselves practically all the wholesome food. There is a good deal of specific evidence from the Middle and Later Roman Empire to confirm this. Philostratus, writing in the first half of the third century a biography of Apollonius of Tyana (a curious figure of the late first century), could describe how at Aspendus in Pamphylia (on the south coast of Asia Minor) Apollonius could find no food on sale in the market except vetches (oroboi): 'the citizens', he says, 'were feeding on this and whatever else they could get, for the leading men [hoi dynatoi, literally 'the powerful'] had shut away all the corn and were keeping it for export' (Philostr., Vita Apollon. I.15; cf. IV.ii and its n.24 below). And again and again, between the mid-fourth century and the mid-sixth, we find peasants crowding into the nearest city in time of famine, because only in the city is there any edible food to be had: I shall give a whole series of examples in IV.ii below.

We must also remember something that is far too often forgotten: the exploitation of the humbler folk was by no means only financial; one of its most burdensome features was the exaction of menial labour services of many kinds. A Jewish rabbi who was active in the second quarter of the third century by our era declared that cities were set up by the State 'in order to impose upon the people angaria' – a term of Persian or Aramaic provenance and originally relating to forced transport services, which had been taken over by the Hellenistic kingdoms (as the Greek word angareia, plural angareiai) and by the Romans (as the Latin angaria, angariae), and had come to be applied to a variety of forms of compulsory labour performed for the State or the municipalities; 'the Middle Ages applied it to services (corvées) owned to the seigneur' (Marc Bloch, in CEHE I2.263-4), and in fifteenth-century Italy we still hear of angararii, and of those bound by fealty in rustic vassalage to their lords, subject to angaria and perangaria (Philip Jones, in id. 406). An example familiar to most people today who have never heard the word angaria is the story of Simon of Cyrene, who was obliged by the Romans to carry the cross of Jesus to the place of execution: Mark and Matthew use the appropriate technical term, a form of the verb angareuein (Mk XV.21; Mt. XXVII.32). Only an understanding of the angareia-system can make fully intelligible one of the sayings of Jesus in the so-called Sermon on the Mount: 'Whosoever shall compel thee to go a mile, go with him twain' (Mt. V.41). Again, the word 'compel' in this text represents the technical term angareuein. (The passage deserves more notice than it usually receives in discussions of the attitude of Jesus to the political authorities of his day.) Readers of the Stoic philosopher Epictetus will remember that he was less positively enthusiastic than Jesus about co-operation with officials exacting angareia: he merely remarks that it is sensible to comply with a soldier's requisition of one's donkey. If one objects, he says, the result will only be a beating, and the donkey will be taken just the same (Diss. IV.i.79).

As it happens, it is in a speech On angareiai (De angariis in Latin, Orat. L) that the great Antiochene orator Libanius makes a particularly emphatic assertion of the absolute dependence of the cities upon the countryside and its inhabitants. (The word angareia does not actually occur in the speech, and Peri tōn angareiōn as its title may be due to a Byzantine scholar; but no one will dispute that angareiai of a particular municipal kind are the subject of the document.) Libanius is complaining to the Emperor Theodosius I in 385 that the peasants of the neighbourhood are being driven to desperation by having themselves and their animals pressed into service for carrying away building rubble from the city. Permits are given by the authorities, he says, which even allow private individuals to take charge of particular gates of the city and to impress everything passing through; with the help of soldiers they drive hapless peasants with the lash (§§ 9, 16, 27 etc.). As Liebeschuetz puts it, the animals of honorati (acting or retired imperial officials and military officers) 'were not requisitioned; other notables managed to get their animals excused even if with some difficulty. All the suffering was that of peasants. There is not a word about losses of landowners' (Ant. 69). Although he has to admit that the practice has been going on for years (§§ 10, 15, 30), Libanius claims that it was illegal (§§ 7, 10, 17-20). He cleverly adduces the fact that a permit was once obtained from an emperor as proof that even the provincial governor has no right to authorise it (§ 22). He also asserts that visitors from other cities are aghast at what they see happening in Antioch (§ 8) – a statement there is no need to take seriously. Towards the end of the speech Libanius explains that the practice he is complaining about has a bad effect on the city's corn supply (§§ 30-1), an argument that might be expected to appeal strongly to the emperor. (We may compare the complaint of the Emperor Domitian, almost exactly three hundred years earlier, that the infliction on working peasants of burdens of the time of angaria is likely to result in failures of cultivation: IGLS V.1998, lines 28-30.) And then Libanius comes to his climax: he begs the philanthrōpotatos basileus,

Show your concern not just for the cities, but for the countryside too, or rather for the countryside in preference to the cities – for the country is the basis on which they rest. One can assert that cities are founded on the country, and that this is their firm footing, providing them with wheat, barley, grapes, wine, oil and the nourishment of man and other living beings. Unless oxen, ploughs, seed, plants and herds of cattle existed, cities would not have come into being at all. And, once in existence, they have depended upon the fortunes of the countryside, and the good and ill that they experience arise therefrom.

Any foe to the well-being of working farmers and even of their animals, he goes on,

is foe to the land, and the foe to the land is foe to the cities also, and indeed to mariners as well, for they too need the produce of the land. They may get from the sea increase of their store of goods, but the very means of life comes from the land. And you too, Sire, obtain tribute from it. In your rescripts you hold converse with the cities about it, and their payment of it comes from the land. So whoever assists the peasantry supports you, and ill-treatment, Sire, by law, punishment and edicts, and in your enthusiasm for the matter under discussion you must encourage all to speak up for the peasants (§§ 33-6, in the translation of A. F. Norman's Loeb edition of Libanius, Vol. II).

I should perhaps add, not only that the practice against which Libanius is protesting is something quite separate from the burdensome angareiai exacted by the imperial authorities, mainly in connection with the 'public post', but also that Libanius himself sometimes takes a very different and much less protective approach towards peasants in his other writings, notably when he is denouncing the behaviour of his own and other tenants, as well as freeholders resisting tax-collectors, in his Orat. XLVII (see IV.ii below).

The linguistic evidence for the separation between polis and chōra is particularly illuminating. Except in some of the western and southern coastal areas of Asia Minor, such as Lydia, Caria, Lycia, Pamphylia and the Cilician plain, where the native tongues seem to have been entirely displaced by Greek during the Hellenistic age, the great majority of the peasants of the Greek East and even some of the townsmen (especially of course the humbler ones) habitually spoke not Greek but the old native tongues. Everyone will remember that when Paul and Barnabas arrived at Lystra, on the edge of a mountain district of southern Asia Minor, and Paul is said to have healed a cripple, the people cried out 'in the speech of Lycaonia' (Act. Apost. XIV.11) – a vernacular tongue which was never written down and which in due course perished entirely. (And this happened inside a city, and moreover one in which Augustus had planted a citizen colony of Roman veterans.) Such stories could be paralleled again and again from widely separated parts of the Roman empire, in both East and West. And those who did not speak Greek or Latin would certainly have little or no part in Graeco-Roman civilisation.

We must exaggerate the strictly ethnic and linguistic factors, which are so noticeable in the more eastern parts of the Greek area, at the expense of economic and social ones. Even in Greece itself, the Aegean islands and the more western coasts of Asia Minor, where Greeks had for centuries been settled and where even the poorest peasant might be as much a Hellene as the city magnate (if at a much lower cultural level), the class division between the exploiters and those from whom they drew their sustenance was very real, and it naturally deepened when the humble entirely lost the protection many of them had been able to obtain from a democratic form of government (see V.iii below). And in the 'Oriental' parts, newly brought within the great Hellenistic kingdoms, the clear-cut difference between 'Hellene' and 'barbaros' (Greek and native) gradually became transformed into a more purely class distinction, between the propertied and non-propertied. This is true even of Egypt, where the gulf between the Greeks and the native Egyptians had originally been as wide as anywhere, extending to language, religion, culture and 'way of life' in general. In Egypt, indeed, there was more interpenetration between two elements than elsewhere, because until A.D. 200 cities were few (there were only Alexandria, Naucratis, Paraetonium and Ptolemais, and in addition Hadrian's foundation of Antinoöpolis in A.D. 130), and because far more Greeks settled outside the cities, in the country districts, often as soldiers or administrators, but with a strong tendency to gravitate toward the 'metropoleis', the capitals of the districts ('nomes') into which Egypt was divided. The exploitation of Egypt under the Ptolemies (323-30 B.C.) was not as intense as under the succeeding Roman administration, and the rents and taxes exacted from the peasantry were at least spent mainly at Alexandria and Naucratis, and at the other centres of population (not yet poleis) where men of property lived, and were not partly diverted (as they were later) to Rome. Nevertheless, the income of the Ptolemies was enormous by ancient standards, and the fellahin must have been pressed hard to provide it. After 200 B.C. 'some natives rose in the scale and took Greek names, and some Greeks sank; Greek and native names occur in the same family. Some Greeks kept themselves aloof; but a new mixed race formed intermediate between Greeks and fellahin, and Hellene came to mean a man with some Greek culture' (Tarn, HC3 206-7). In Egypt, as elsewhere, 'being a Greek' was certainly very much more a matter of culture than of descent; but culture itself was largely dependent upon property-ownership. Before the end of the second century B.C., as Rostovtzeff says, 'From the social and economic standpoint the dividing line between the upper and lower class was no longer between the Greeks forming the upper, and the Egyptians forming the lower, but between the rich and poor in general, many Egyptians being among the first, many Greeks among the second'; but 'the old division into a privileged class of "Greeks" (which comprised now many hellenised Egyptians) and a subordinate class of natives remained as it had been' (SEHHW II.883). This is true, although some of the documents cited by Rostovtzeff might now be differently interpreted in some respects. In the Roman period, with the growth of the metropoleis into something more nearly resembling Greek cities, where the landowners mainly lived, the propertied classes generally regarded themselves as Greeks and the peasants as Egyptians. In a letter surviving on papyrus from the third century of the Christian era, the writer does not want his 'brethren' to think of him as 'a barbarian or an inhuman [ananthrōpus] Egyptian' (P. Oxy. XIV.1681.4-7).

Marriages between city folk and peasants must have been very uncommon in all parts of the Greek world. Occasionally, no doubt, a peasant girl might be beautiful enough to attract a well-to-do city gentleman, but as a rule he would probably be far more likely to make her his mistress or concubine than his wife. There is, however, one delightful story, which I cannot resist telling, of love and marriage between two rich young city men and two lovely Sicilian peasant girls, who became known as the Kallipygoi. This is transmitted to us through Athenaeus (XII.554cde), from the iambic poems of Cercidas of Megalopolis and Archelaus of Chersonesus. (How much truth there is in it we have no means of knowing.) The two beautiful daughters of a peasant (an anēr agroikos), disputing which of them was the more callipygous, went out onto the highway and invited a young man who happened to be passing by to arbitrate between them. Inspecting both, he preferred the elder, with whom he then and there fell in love. His younger brother, when he heard about the girls, went out to see them, and fell in love with the younger. The aged father of the two young men did his best to persuade his sons to make more reputable marriages, but without success, and eventually he accepted the two peasant girls as his daughters-in-law. Having thus risen greatly in the world and become conspicuously rich, the two women built a temple to Aphrodite Kallipygos – a cult title which was not only most appropriate to the goddess of love and beauty but also made a charming allusion to the circumstances of the foundation. (One may feel that this is one of the cases in which paganism had a distinct advantage over Christianity.) Marriages of well-bred girls to peasants must also have been exceedingly rare. In Euripides' Electra the marriage of the princess Electra to a poor rustic who is not even given a name in the play – he is just an autourgos (a man who works his farm with his own hands) – is regarded even by the man himself as a grave and deliberate slight on the girl, and in his opening speech he alludes with pride to the fact that he has never taken her to his bed and she is still a virgin – tense and neurotic, as we presently discover.

The contrast between superior city-dweller and unsophisticated countryman could even be projected into the divine sphere. In a collection of fables by Babrius we hear of a belief that it is the simple-minded (euētheis) among the gods who inhabit the countryside, while those deities who live within the city wall are infallible and have everything under their supervision (Fab. Aesop. 2.6-8).

In III.vi below I shall mention briefly the creation by wealthy benefactors in Greek and Roman cities of 'foundations' to provide distributions of money or food on special occasions, often graded according to the position of the recipients in the social hierarchy – the higher a person's social position, the more he was likely to get. Rustics, who in the Greek East would often not be citizens of their polis, would very rarely benefit from such a distribution. Dio Chrysostom can make one of his Euboean peasants adduce the fact that his father had once participated in a distribution of money in the local town as evidence that he was a citizen there (VII.49). The only inscription I have noticed that mentions countryfolk benefiting from a distribution instituted by a citizen of a Greek polis is one from Prusias ad Hyphium in Bithynia, which speaks of handouts both to all those 'reckoned as citizens' (enkekrimenois) and to those 'inhabiting the country district' (tois tēn agroikian katoikousin/paroikousin, IGRR III.69.18-20, 24-6).

To conclude this section, I cannot do better than quote two summaries by A. H. M. Jones of his researches into a thousand years of Hellenistic and Roman rule in the Greek East. One, from his first major work, Cities of the Eastern Roman Provinces (1937, 2nd edn 1971), deals specifically with Syria, which had previously been only on the fringe of the Greek world but was brought within it by degrees from the time of Alexander's conquests, from 333 B.C. onwards; but Jones's conclusions are equally, or almost equally, true of the other areas in western Asia, north Africa, and south-east Europe which became hellenised only in Alexander's time or later. Summing up 'the results of the millennium during which Syria had been ruled by the Macedonian dynasties and by Rome', Jones says,

On paper the change in the political aspect of the country is considerable. In the Persian period cities existed only on the sea-coast, the desert fringe, and two of the gangways between them through the central mountain barrier. By the Byzantine period practically the whole of Syria was partitioned into city states; only in a few isolated areas, notably the Jordan valley and the Hauran, did village life remain the rule. In reality, however, the change was superficial. It was achieved partly by assigning vast territories to the old cities of the coast and of the desert fringe, partly by the foundation of a small number of new cities, to each of which was assigned a vast territory. The political life of the inhabitants of the agricultural belt was unaffected; their unit remained the village, and they took no part in the life of the city to which they were attached. Economically they lost by the change. The new cities performed no useful economic function, for the larger villages supplied such manufactured goods as the villagers required, and the trade of the countryside was conducted at village markets. The only effect of the foundation of cities was the creation of a wealthy landlord class which gradually stamped out peasant proprietorship. Culturally, the countryside remained utterly unaffected by the Hellenism of the cities; the peasants continued to speak Syriac down to the Arab conquest. The only function which the cities performed was administrative; they policed and collected the taxes of their territories (CERP2 293-4).

and in a note later in the book Jones adds,

The indifference of the villagers to the cities is, I think, well illustrated by the tombstones of syrian emigrants in the West . . . ; they always record their village, but name their city, if at all, merely as a geographical determinant' (CERP2 469 n.92).

The other passage is from p.vi of the Preface to Jones's The Greek City from Alexander to Justinian (1940). Summarising the conclusions in Part V of that book, Jones says that he discusses 'the contribution of the cities to ancient civilisation' and argues that

Great as their achievement was, it was based on too narrow a class foundation to be lasting. On the economic side the life of the cities involved an unhealthy concentration of wealth in the hands of the urban aristocracy at the expense of the proletariat and the peasants. Their political life was gradually narrowed till it was confined to a small clique of well-to-do families, who finally lost interest in it. The culture which the cities fostered, though geographically spread over a wide area, was limited to the urban upper class.

(iv) - The relevance of Marx for the study of ancient history

So complete has been the lack of interest in Marx displayed by nearly all ancient historians in the English-speaking world[1] that many who begin to read this book may wonder what relevance Marx can possibly have to the history of Classical antiquity. I have heard this lack of interest described as 'a conspiracy of silence'; but that would be to dignify it with a conscious element which in practice is absent: the reality is just silence. I know of nothing comparable as yet in the British Isles to the symposium on the programme of the American Philological Association in 1973, entitled 'Marxism and the Classics', or to the issue of the American Classical periodical Arethusa, vol.8.1 (Spring, 1975), with the same title.[2] (The article included in that volume, with the title 'Karl Marx and the history of Classical antiquity', pp.7-41, is virtually a series of extracts from earlier drafts of this book.) One often hears the view expressed that in so far as the ideas of Marx on history have any validity, they have already been absorbed into the Western historiographical tradition. One thinks here of the late George Lichtheim's description of Marxism as 'the caput mortuum of a gigantic intellectual construction whose living essence has been appropriated by the historical consciousness of the modern world' (Marxism[2] [1964 and repr.] 406). This is altogether untrue, above all in regard to the modern historiography of the Classical world.

Now the situation I have described is certainly due in part to a general ignorance of the thought of Marx, and a lack of interest in it, on the part of the vast majority of ancient historians and other Classical scholars in the English-speaking world. But I shall suggest later that this ignorance and lack of interest can be attributed partly to mistaken attempts in modern times, on the part of those who call themselves Marxists (or at least claim to be influenced by Marx), to interpret the essentials of Marx's historical thought both in general terms and in particular in relation to Classical antiquity. I like to remember that Engels, in a letter written to Conrad Schmidt on 5 August 1890, more than seven years after Marx's death, recalled that Marx used to say about the French Marxists of the late 1870s, 'All I know is that I am not a Marxist' (MESC 496). I think he would have felt much the same about soi-disant Marxists – not only the French ones – of the 1980s. As the German poet Hans Magnus Enzensberger says, in his moving short poem, Karl Heinrich Marx –

I see you betrayed

by your disciples:

only your enemies

remained what they were.

(The translation of the poem by Michael Hamburger is reprinted in the Penguin Poems of Hans Magnus Enzensberger 38-9.) Much modern Marxist writings in languages other than English seems recalcitrant to translation into English. I am inclined to apply to much of this writing some forceful remarks made by Graham Hough in a review in the Times Literary Supplement of two books on Roland Barthes. Approving a statement by Stephen Heath, that the language evolved by Barthes and his school 'has no common theoretical context with anything that exists in English', he continues:

To transfer it bodily – simply to anglicise the words, which is not difficult – produces a wall of opacity that blocks all curiosity at the start. To adapt, to paraphrase, which can also be done and often looks inviting, runs the risk of denaturing the original and reducing disconcerting ideas to acceptable commonplace (TLS 3950, 9 December 1977, p. 1443).

So it is, I feel, with much contemporary Marxist work, even in French and Italian, and still more in German and Russian.

More and more people in my adult lifetime have become willing to take some account of Marx's analysis of the capitalist world in the nineteenth and twentieth centuries. As I am a historian and not an economist, I shall do no more than mention the revival of serious interest in Marx's economics in Britain on the part of a number of leading economists of our generation (whether or not they would describe themselves as Marxists): Maurice Dobb, Ronald Meek, Joan Robinson, Piero Sraffa and others.[3] In the Foreword to the first edition of her Essay on Marxian Economics (1942) Joan Robinson remarked that 'until recently Marx used to be treated in academic circles with contemptuous silence, broken only by an occasional mocking footnote'. In the first paragraph of the Preface to the second edition (1966), she mentioned that when she was writing the original edition, a quarter of a century earlier, most of her 'academic colleagues in England thought that to study Marx was a quaint pastime . . . , and in the United States it was disreputable'. Matters are rather different now. Within the last few years sociologists too have suddenly become far more willing than they used to be to adopt a Marxist analysis of the problems of contemporary society. I may perhaps be allowed to refer to one particularly impressive recent example: a book entitled Immigrant Workers and Class Structure in Western Europe, by Stephen Castles and Godula Kosack, published in 1973, the relevance of which for our present study will emerge in II.iii below. Even so, many people would, I think, agree with the opinion of a leading British sociologist, T. B. Bottomore (who is far from hostile to Marx), that 'while the Marxian theory seems highly relevant and useful in analysing social and political conflicts in capitalist societies during a particular period, its utility and relevance elsewhere are much less clear' (Sociology2, [1971] 201). Those who hold such views may be prepared to concede that a very valuable contribution has been made by certain Marxist historians who have dealt mainly with the eighteenth and nineteenth centuries, for example Eric Hobsbawm, George Rudé, and E. P. Thompson; but they may begin to feel that their premise has been somewhat weakened when they take notice of the work of an American Marxist historian, Eugene Genovese, who has produced work of outstanding quality on slavery in the antebellum South; and it is surely strained to breaking-point and beyond when they have to take account of Christopher Hill (formerly the Master of Balliol), who has done so much to illuminate the history of the sixteenth and seventeenth centuries, and Rodney Hilton, who has dealt with English peasants and peasant movements in the fourteenth century and earlier, in various articles and in two recent books, Bond Men Made Free (1973) and The English Peasantry in the Later Middle Ages (1975, the publication of his Ford Lectures at Oxford in 1973). We are already a very long way from nineteenth-century capitalism; and if we go still further back, into the Bronze Age and prehistory, in Europe and Western asia, we can find archaeologists, in particular the late V. Gordon Childe, also acknowledging their debt to Marx. [See now VIII.i.n.33 below.]

Anthropologists too, at least outside of Great Britain, have for some time been prepared to take Marx seriously as a source of inspiration in their own discipline. French economic anthropologists such as Maurice Godelier, Claude Meillassoux, Emmanuel Terray, Georges Dupré and Pierre-Philippe Rey have operated to a high degree within a Marxist tradition, which they have developed in various ways.[4] Even the structuralists have often acknowledged a debt to Marx. Over twenty years ago Claude Lévi-Strauss himself referred to his 'endeavours to reintegrate the anthropological knowledge acquired during the last fifty years into the Marxian tradition'; and spoke of 'the concept of structure which I have borrowed, or so I thought, from Marx and Engels, among others, and to which I attribute a primary role' (SA 343-4).[5] American anthropologists have also become much more attentive to Marx in recent years: Marvin Harris, for example, in his comprehensive work, The Rise of Anthropological Theory (1969 and repr.), devotes some serious attention to Marx and Engels as anthropologists, including a chapter of over 30 pages ('Dialectical materialism', pp.217-49). And then, in 1972, came what I can only describe as a break-through in British anthropology. An anthropologist of the very first rank, Sir Raymond Firth, delivered the inaugural lecture of a new British Academy series in honour of Radcliffe-Brown, gave it a significant title: not merely 'The sceptical anthropologist?' (An allusion, of course, to Robert Boyle's The Sceptical Chymist) but also 'Social anthropology and Marxist views on society'.[6] I should like to quote part of the last paragraph of this lecture, because it urges social anthropologists to interest themselves in particular aspects of human societies which I think historians of Classical antiquity should also be studying, and which – like the social anthropologists to whom Firth is addressing himself – most of them are not studying. Firth says:

What Marx's theories offer to social anthropology is a set of hypotheses about social relations and especially about social change. Marx's insights – about the basic significance of economic factors, especially production relations; their relation to structures of power; the formation of classes and the opposition of their interests; the socially relative character of ideologies; the conditioning force of a system upon individual members of it – [these insights] embody propositions which must be taken for critical scrutiny into the body of our science. The theories of Marx should be put on a par with, say, those of Durkheim or Max Weber. Because they imply radical change they are more threatening.

That last word is particularly significant. (I shall return to the 'threatening' nature of Marxist analysis in II.ii below.) Now Firth, I am sure, would not describe himself as a Marxist. Shortly before the paragraph I have quoted he expresses the opinion that 'much of Marx's theory in its literal form is outmoded': the examples he gives in support of this claim do not seem to me well formulated or cogent. But what I am primarily concerned to do at the moment is to make a plea for the relevance of Marx's general historical methodology to the study of ancient history. If it can make major contributions to history between the early Middle Ages and the twentieth century, and even in archaeology and anthropology, then there is good reason to expect that it may be able to shed light upon Classical antiquity.

Apart from one negligible book which I shall mention later (in II.i below and its n.20), I know of no single work in English which consistently attempts either to analyse Greek history – or, for that matter, Roman history – in terms of Marxist historical concepts, or to expound those concepts themselves and explain why they are relevant for the purpose of such an analysis. In fact both these tasks need to be accomplished together at least once, within one pair of covers (as I am trying to do here), if the new start that I am advocating is to be made successfully. As I have said, most English-speaking ancient historians ignore Marx completely. If they do mention him, or Marxist historical writing, it is usually with ignorant contempt. An exception is a recent well-chosen selection of source material in translation for Greek economic and social history in the Archaic and Classical periods, first published in French by Michel M. Austin and Pierre Vidal-Naquet under the title of Économies et sociétés en Grèce ancienne (Paris, 1972 and 1973) and then, with some improvements, in English, as Economic and Social History of Ancient Greece: An Introduction (London, 1977). The introduction (mainly by Austin) devotes several pages (20 ff. in the English version) to the notion of 'class struggles'. Now, as I shall explain (in II.iii below), I disagree profoundly with the way these scholars have applied the Marxist concept of class conflict to the Greek world; but at least they are operating with categories that have become thoroughly associated with the Marxist tradition in historiography and are very often repudiated altogether or allowed only a very limited role by non-Marxists.

In languages other than English the situation is much better – although, as I indicated near the beginning of this section, many of the Marxist works on ancient history published on the Continent are as foreign to the English reader in their intellectual and library idiom as in their actual language: they tend to take for granted a whole range of concepts to which most people in the English-speaking world are not accustomed and which they find largely unintelligible.[7] The word 'jargon' is often used in this context, if not always by those who have earned the right to use it by refraining from a different jargon of their own.

At this point I must write briefly about Marx himself as a Classical scholar. He received, in school and university, at Trier, Bonn and Berlin, the thorough Classical education which was given to most young middle-class Germans in the 1830s. At the universities of Bonn and Berlin he studied law and philosophy, and between 1839 and 1841, among various other activities, he wrote, as his doctoral thesis, a comparison of the philosophies of Democritus and Epicurus. This work, completed in 1840-41, before Marx was 23, was not published in full even in German until 1927, when it appeared in MEGA I.i.1 (the first fascicule of Part i of Vol. I of the Marx-Engels Gesamtausgabe, published at Frankfurt and edited by D. Rjazanov) 1-144. It has not been republished in MEW I (the first volume of the complete Werke of Marx and Engels now in course of publication in East Berlin). An English translation (replacing an inferior earlier one) has recently been published in MECW I, the first volume of the new English edition of the Marx-Engels Collected Works (Moscow/London/New York, 1975), 25-107. Cyril Bailey, reviewing the original publication in the Classical Quarterly 22 (1928) 205-6, was greatly impressed with its scholarship and its originality: he found it 'of real interest to a modern student of Epicureanism' and ended by saying that such a student would find in it 'some illuminating ideas'. The thesis looks forward to a larger work (never actually written) in which Marx planned to 'present in detail the cycle of Epicurean, Stoic and Sceptic philosophy in their relation to the whole of Greek speculation' (MECW I.29). It is worth noticing that the Foreword to the thesis ends by quoting the defiant reply of Prometheus to Hermes, in Aeschylus' Prometheus Bound (lines 966 ff.), 'Be sure of this: I would not exchange my state of misfortune for your servitude', and adding that Prometheus (the Prometheus of Aeschylus) is 'the most eminent saint and martyr in the philosophical calendar' (MECW I.31). During this period Marx read extensively in Classical authors, in particular Aristotle, of whom throughout his life he always spoke in terms of respect and admiration which he employs for no other thinker, except perhaps Hegel. As early as 1839 we find him describing Aristotle as 'the acme [Gipfel] of ancient philosophy' (MECW I.424); and in Vol. I of Capital he refers to 'the brilliance of Aristotle's genius' and calls him 'a giant thinker' and 'the greatest thinker of antiquity' (60, 82n., 408) – as of course he was. Later, Marx returned again and again to read Classical authors. On 8 March 1855 we find him saying in a letter to Engels, 'A little time ago I went through Roman history again up to the Augustan era' (MEW XXVIII.439); on 27 February 1861 he writes again to Engels, 'As a relaxation in the evenings I have been reading Appian on the Roman civil wars, in the original Greek' (MESC 151); and some weeks later, on 29 May 1861, he tells Lassalle that in order to dispel the serious ill-humour arising from what he describes, in a mixture of German and English, as 'mein in every respected unsettled situation', he is reading Thucydides, and he adds (in German) 'These ancient writers at least remain ever new' (MEW XXX.605-6).

(This is a convenient place at which to mention that I normally cite MESC, an English translation of 244 of the letters of Marx and Engels, published in 1956, when it includes a letter I am quoting. I need not regularly refer to the German texts, since they print the letters in chronological order, and the dates will enable them to be found easily. The letters exchanged between Marx or Engels to other correspondents, in MEW XXVII-XXXIX.)

Scattered through the writings of Marx are a remarkable number of allusions to Greek and Roman history, literature and philosophy. He made a careful study of Roman Republican history in particular, partly from the sources and partly with the aid of the works of Niebuhr, Mommsen, Dureau de la Malle and others. I have not been able to discover any systematic study of Greek history by Marx after his student days, or of the history of the Graeco-Roman world under the Principate or the Later Roman Empire; but he frequently quotes Greek authors (more often in the original than in translation), as well as Latin authors, in all sorts of contexts: Aeschylus, Appian, Aristotle, Athenaeus, Democritus, Diodorus, Dionysius of Halicarnassus, Epicurus, Herodotus, Hesiod, Homer, Isocrates, Lucian, Pindar, Plutarch, Sextus Empiricus, Sophocles, Strabo, Thucydides, Xenophon, and others. He could also make use of that charming little poem by Antipater of Thessalonica, in the Greek Anthology (IX.418), which is one of the earliest pieces of evidence for the existence of the water-mill (see II.i below). After his doctoral dissertation Marx never had occasion to write at length about the ancient world, but again and again he will make some penetrating remark that brings out something of value. For example, in a letter to Engels of 25 September 1857 he makes some interesting and perfectly correct observations: for example, that the first appearance of an extensive system of hired labour in antiquity is in the military sphere, the employment of mercenaries (how often has that been noticed, I wonder!), and that among the Romans the peculium castrense was the first legal form in which the right of property was recognised in members of a family other than the paterfamilias (MESC 118-19). In a footnote in the Grundrisse (not in the section on 'pre-capitalist forms of production'), written at about the same time as the letter from which I have just quoted, Marx has some acute observations on pay in the Roman army, which need to be put beside the remark in the letter:

Among the Romans, the army constituted a mass – but already divorced from the whole people – which was disciplined to labour, whose surplus time also belonged to the State; who sold their entire labour time for pay to the State, exchanged their entire labour capacity for a wage necessary for the maintenance of their life, just as does the worker with the capitalist. This holds for the period when the Roman army was no longer a citizen's army but a mercenary army. This is here likewise a free sale of labour on the part of the soldier. But the State does not buy it with the production of values as aim. And thus, although the wage form may seem to occur originally in armies, this pay system is nevertheless essentially different from wage labour. There is some similarity in the fact that the State uses up the army in order to gain an increase in power and wealth (Grundrisse, E.T. 529n.; cf. 893).

it came naturally to Marx to illustrate what he was saying with some Classical simile, as when he wrote that the trading peoples of antiquity were 'like the gods of Epicurus, in the spaces between the worlds' (Grundrisse, E.T. 858; cf. Cap. III.330, 598), or when he spoke scornfully of Andrew Ure, author of The Philosophy of Manufactures, as 'this Pindar of the manufacturers' (Cap. III.386 n.75). I have heard quoted against Marx his remark that Spartacus (the leader of the great slave revolt in Italy from 73 to 71 B.C.) was 'the most splendid fellow in the whole of ancient history. Great general (no Garibaldi), noble character, real representative of the ancient proletariat'; so let me mention here that the statement was made not in a work intended for publication but in a private letter to Engels, of 27 February 1861 – in which, incidentally, he also described Pompey as 'reiner Scheisskerl' (MEW XXX.159-60=MESC 151-2).

A recent book by the Professor of German at Oxford University, S. S. Prawer, Karl Marx and World Literature (1976), has shown in detail how extraordinarily wide Marx's reading was, not only in German, French, English, Latin and Greek, but also in Italian, Spanish and Russian.

I shall have something to say in II.iii below on Marx's intellectual development in the 1840s.

I may add that Engels too was very well read and received a Classical education. A school-leaving report testifying to his knowledge of Latin and Greek survives, as does a poem he wrote in Greek at the age of sixteen.[8]

However, it is not so much as the student of a particular epoch that I wish to regard Marx now, but rather as a historical sociologist: one who proposed an analysis of the structure of human society, in its successive stages, which sheds some illumination upon each of those stages – the Greek world just as much as the nineteenth and twentieth centuries.

Let me first mention and dismiss two or three common misconceptions. It is easy to discredit Marx's analysis of society by presenting it in a distorted form, as it is so often presented both by those who wrongly suppose themselves to be employing it and by those who are in principle hostile to it. In particular the thought of Marx is said to involve both 'materialism' and 'economic determinism'. Now the historical method employed by Marx was never given a name by him, but from Engels onwards it has been generally known as 'historical materialism'. (It seems to have been Plekhanov who invented the term 'dialectical materialism'.) It is certainly 'materialist', in the technical sense of being methodologically the opposite of Hegel's 'idealism' – we all know Marx's famous remark that Hegel's dialectic was standing on its head and 'needs to be turned right side up again if you would discover the rational kernel within the mystical shell' (Cap. I.20, from the Afterword to the second German edition, of 1873). But 'materialism' does not, and must not, in any way exclude an understanding of the role of ideas, which (as Marx well knew) can often become autonomous and acquire a life of their own, and themselves react vigorously upon the society that produced them – the role of Marxism itself in the twentieth century is a conspicuous example of this. As for the so-called 'economic determinism' of Marx, the label must be altogether rejected. We can begin with his alleged over-emphasis on the economic side of the historical process, which has even led to the application to his historical methodology – quite absurdly – of the terms 'reductionist' and 'monistic'. In fact the dialectical process which Marx envisaged allowed to other factors than the purely economic – whether social, political, legal, philosophic or religious – almost as much weight as very many non-Marxist historians would give to them. The alleged 'economism' of Marx is no more than the belief that out of all the elements which are operative in the historical process, it is 'the relations of production' (as Marx called them), namely the social relations into which men enter in the course of the productive process, which are the most important factors in human life, and which tend, in the long run, to determine the other factors, although of course these other factors, even purely ideological ones, can sometimes exert a powerful influence in their turn upon all social relations. In five of the letters he wrote between 1890 and 1894 Engels, while admitting that he and Marx had been partly to blame for an unavoidable over-emphasis on the economic aspect of history, stressed that they had never intended to belittle the interdependent role of political, religious and other ideological factors, even while considering the economic as primary. (The letters are those of 5 August, 21 September and 27 October 1890, 14 July 1893, and 25 January 1894.)[9] In an obiter dictum in one of his earliest works, the Contribution to a Critique of Hegel's Philosophy of Law, Marx declared that although material force can be overcome only by material force, yet 'Theory also becomes a material force as soon as it has gripped the masses' (MECW III.182). And Mao Tse-tung, in a famous essay 'On Contradiction' (dating from August 1937), insisted that in certain conditions theory and the ideological 'superstructure' of a society (revolutionary theory in particular) can 'manifest themselves in the principal and decisive role'.[10]

It is true that Marx himself occasionally writes as if men were governed by historical necessities beyond their control, as when (in the Preface to the original German edition of Das Kapital) he speaks of 'the natural laws of capitalist production' as 'self-assertive tendencies working with iron necessity' (MEW XXIII.12. I have altered the misleading translation in Cap. I.8). Such expressions are rare: they probably derive from a conception of historical events in which a high degree of probability has been momentarily taken as certainty. In fact there is nothing in the least 'deterministic' in the proper sense in Marx's view of history; and in particular the role of no single individual is 'determined' by his class position, even if one can often make very confident predictions (of a statistical character) about the behaviour of the collective members of a given class. To give just two examples: if you have an income of more than, say, £20,000 a year, the statistical probability that you will normally hold right-wing views, and in Britain vote Conservative, is very high indeed; and if you do not belong to the lowest social class you will have a far better chance of achieving individual sainthood in the Roman Church – a sociological analysis in the early 1950s showed that of 2,489 known Roman Catholic Saints, only 5 per cent came from the lower classes who have constituted over 80 per cent of Western populations.[11] (Recent proclamations of sanctity, I understand, have not departed from his pattern.)

I believe that some light may be shed on the last question we have been considering (the 'determinism' of which Marx is often accused) by a comparison between Marx and the greatest historian of antiquity, Thucydides – probably the writer who, with the single exception of Marx, has done most to advance my own understanding of history. Thucydides often refers most to something he calls 'human nature', by which he really means patterns of behaviour he believed he could identify in human conduct, partly in the behaviour of individual men but much more emphatically in tht of human groups: men acting as organised states, whose behaviour can indeed be predicted far more confidently than that of most individual men. (I have discussed this in my OPW 6, 12 & n.20, 14-16, 29-33, 62, cf. 297.) The better you understand these patterns of behaviour, Thucydides (I am sure) believed, the more effectively you can predict how men are likely to behave in the immediate future – although never with complete confidence, because always (and especially in war) you must allow for the unforeseeable, the incalculable, and for sheer 'chance' (see OPW 25 & n.54, 30-1 & n.57). Thucydides was anything but a determinist, although he often speaks of men as being 'compelled' to act in a particular way when he describes them as choosing the least disagreeable among alternatives none of which they would have adopted had their choice been entirely free (see OPW 60-2). This common feature of the human predicament, I believe, is just what Marx had in mind when he said, in The Eighteenth Brumaire of Louis Bonaparte, 'Men make their own history, but they do not make it just as they please; they do not make it under circumstances chosen by themselves, but under circumstances directly encountered, given and transmitted from the past' (MECW XI.103).

In every situation in which one is making a judgment there are some factors which cannot be changed and others which can only be partly modified, and the better one understands the situation the less forced and unfree one's judgment becomes. In this sense, 'freedom is the understanding of necessity'. Thucydides, by enabling his readers to recognise and understand some of the basic recurring features in the behaviour of human groups in the political and international field, believed – surely with reason – that his History would be for ever 'useful' to mankind (I.22.4). Similarly, what Marx wished to do was to identify the internal, structural features of each individual human society (above all, but not only, capitalist society), and reveal its 'laws of motion'. If his analysis is largely right, as I believe it is, then, by revealing the underlying Necessity, it increases human Freedom to operate within its constraint, and has greatly facilitated what Engels called 'the ascent of man from the kingdom of necessity to the kingdom of freedom' (MESW 426).

In the third volume of Capital there is a point at which Marx suddenly and quite unexpectedly bursts out into one of those emotional passages 'full of hope and splendour' – an apt phrase of Hobsbawm's (KMPCEF 15) – which look beyond the harsh realities of the present towards a future in which mankind is largely set free from the soul-destroying compulsion which still obliges the greater part of humanity to spend most of their time producing the material necessities of life. This passage, one of many in Capital that reveal the essential humanity of Marx's outlook, must seem less purely visionary and utopian, in our age of increasing automation, than it may have appeared to those who first read it in the 1890s. It occurs in Part VII of Capital III (p.820), in a chapter (xlviii) entitled 'The trinity formula', from which I also quote elsewhere. (The German text can be found in MEW XXV.828.)

The realm of freedom actually begins only where labour which is determined by necessity and mundane considerations ceases; thus int he very nature of things it lies beyond the sphere of actual material production. Just as the savage must wrestle with Nature to satisfy his wants, to maintain and reproduce life, so must civilised man, and he must do so in all social formations and under all possible modes of production. With his development this realm of physical necessity expands as a result of his wants; but, at the same time, the forces of production which satisfy these wants also increase. Freedom in this field can only consist in socialised man, the associated producers, rationally regulating their interchange with Nature, bringing it under their common control, instead of being ruled by it as by the blind forces of Nature; and achieving this with the least expenditure of energy and under conditions most favourable to, and worthy of, their human nature. But it none the less still remains a realm of necessity. Beyond it begins that development of human energy which is an end in itself, the true realm of freedom, which, however, can blossom forth only with this realm of necessity as its basis. The shortening of the working day is its basic prerequisite. (Cf. Marx/Engels, MECW V.431-2, from the German Ideology, quoted in II.i below.)

Marx and Engels were certainly not among those who not merely speak loosely (as any of us may) but actually think seriously of History (with a capital 'H') as a kind of independent force. In a splendid passage in his earliest joint work with Marx, The Holy Family (1845), Engels could say,

History does nothing, it 'possesses no immense wealth', it 'wages no battles'. It is man, real, living man who does all that, who possesses and fights; 'history' is not, as it were, a person apart, using man as a means to achieve its own aims; history is nothing but the activity of man pursuing his aims (MECW IV.93=MEGA I.iii.265).

Excerpt in so far as the concepts of class and class struggle are involved, I do not propose in this book to undertake any comprehensive discussion of Marx's general historical methodology,[12] which of course involves much more than class analysis, although that to my mind is central and its rejection entails the dismissal of most of Marx's system of ideas. Nor do I intend to say anything about such controversies as those concerning 'basis and superstructure',[13] or the so-called 'modes of production' referred to by Marx, in particular in the German Ideology (MECW V.32-5), in Wage Labour and Capital (MECW IX.212), in the section on pre-capitalist economic formations in the Grundrisse (E.T. 471-514, esp. 495),[14] and in the Preface to A Contribution to the Critique of Political Economy (MESW 182). Above all I can legitimately avoid any discussion of the desirability (or otherwise) of recognising an 'Asiatic' (or 'Oriental') mode of production, a notion which seems to me best forgotten.[15] When speaking (for example) of various parts of Asia at times before they had been taken over by the Greeks (or the Macedonians), I believe that it is best to employ such expressions as 'pre-Classical modes of production', in a strictly chronological sense.

It is not my purpose in this book to defend Marx's analysis of capitalist society or his prophecy of its approaching end (both of which in the main I accept); but I have so often heard it said that he did not allow for the growth of a managerial and 'white-collar' middle class[16] that I will end this final section of my Introduction with a reference to two passages in his Theories of Surplus Value which rebut this criticism – and are by no means irrelevant to the main subject of this book, because they serve to illustrate a feature of the modern world to which there was no real parallel in antiquity. Criticising Malthus, Marx says that 'his supreme hope, which he himself describes as more or less utopian, is that the mass of the middle class should grow and that the proletariat (those who work) should constitute a constantly declining proportion (even though it increases absolutely) of the total population'; and he adds, 'This in fact is the course taken by bourgeois society' (TSV III.63).

And criticising Ricardo, Marx complains that 'what he forgets to emphasise is the constantly growing number of the middle classes, those who stand between the workman on the one hand and the capitalist and landlord on the other. The middle classes . . . are a burden weighing heavily on the working base and they increase the social security and power of the upper Ten Thousand' (TSV II.573=MEW XXVI.ii.576).

These passages may remind us of the fact that in the Greek and Roman world there was no proper parallel to our own 'white-collar', salaried, managerial class (we shall see why in III.vi below), except in the Roman Principate and Later Empire, when three developments took place. First, a proper standing army was established in the early Principate, with (for the first time) regular benefits on discharge as well as fixed pay, found by the state. Those who became what we should call 'regular officers', especially the senior centurions, might become men of rank and privilege. Secondly, an imperial civil service grew up gradually, consisting partly of the emperor's own slaves and freedmen and partly of free men who, at all levels, served for pay (and for the often considerable dimensions, although many of its members were technically soldiers seconded for this duty. The third group of functionaries consisted of the Christian clergy, whose upkeep was provided partly by the state and partly by the endowments and contributions of the faithful. I shall have more to say about all these three groups later (VI.v-vi and esp. VIII.iv). Exactly like the middle classes referred to by Marx, they were certainly 'a burden weighing heavily on the working base', and as faithful bastions of the established order they too – except in so far as sections of the army were drawn into civil wars in support of rival emperors – 'increased the social security and power of the upper Ten Thousand'.

To conclude this section, I wish to emphasise that I make no claim to be producing the 'Marxist interpretation of Greek history': it is a would-be Marxist interpretation. After reading by far the greater part of Marx's published work (much of it, I must admit, in English translation), I myself believe that there is nothing in this book which Marx himself (after some argument, perhaps!) would not have been willing to accept. But of course there will be other Marxists who will disagree at various points with my basic theoretical position or with the interpretations I have offered of specific events, institutions and ideas; and I hope that any errors or weaknesses in this book will not be taken as directly due to the approach I have adopted, unless that can be shown to be the case.

II - Class, Exploitation, and Class Struggle

(i) - The nature of class society

'The concept of class has never remained a harmless concept for very long. Particularly when applied to human beings and their social conditions it has invariably displayed a peculiar explosiveness.' Those are the first two sentences of a book, Class and Class Conflict in Industrial Society, by Ralf Dahrendorf, a leading German sociologist who in 1974 became Director of the London School of Economics and Political Science. And Dahrendorf goes on to quote with approval the statement by two prominent American sociologists, Lipset and Bendix, that 'discussions of different theories of class are often academic substitutes for a rael conflict over political orientations'. I fully accept that. It seems to me hardly possible for anyone today to discuss problems of class, and above all class struggle (or class conflict), in any society, modern or ancient, in what some people would call an 'impartial' or 'unbiased' manner, I make no claim to 'impartiality' or 'lack of bias', let alone 'Wertfreiheit', freedom from value-judgments. The criteria involved are in reality much more subjective than is commonly admitted: in this field one man's 'impartiality' is another man's 'bias', and it is often impossible to find an objective test to resolve their disagreement. Yet, as Eugene Genovese has put it, 'the inevitability of ideological bias does not free us from the responsibility to struggle for maximum objectivity' (RB 4). The criteria that I hope will be applied to this book are two: first, its objectivity and truthfulness in regard to historical events and processes; and secondly, the fruitfulness of the analysis it produces. For 'historical events and processes' I should almost be willing to substitute 'historical facts'. I do not shrink from that unpopular expression, any more than Arthur Darby Nock did when he wrote, 'A fact is a holy thing, and its life should never be laid down on the altar of a generalisation' (ERAW I.333). Nor do I propose to dispense with what is called – sometimes with a slight sneer, by social and economic historians – 'narrative history'. To quote a recent statement in defence of 'narrative history' by the present Camden Professor of Ancient History at Oxford:

I do not see how we can determine how institutions worked, or what effect beliefs or social structures had on men's conduct, unless we study their actions in concrete situations . . . The most fundamental instinct that leads us to seek historical knowledge is surely the desire to find out what actually happened in the past and especially to discover what we can about events that had the widest effect on the fortunes of mankind; we then naturally go on to inquire why they occurred (P. A. Brunt, 'What is Ancient History about?', in Didaskalos 5 [1976] 236-49, at 244).

Can we actually identify classes in Greek society such as I shall describe? Did the Greeks themselves recognise their existence? And is it profitable to conduct an investigation along these lines? Is our understanding of the historical process, and of our own society, illuminated and strengthened by thinking in terms of classes and of a 'class struggle' in the Greek world? When I find Lévi-Strauss saying, 'I am not a sociologist, and my interest in our own society is only a secondary one' (SA 338), I want to reply, 'I am a historian who tries also to be a sociologist, and my interest in our own society is a primary one.'

I am not going to pretend that class is an entity existing objectively in its own right, like a Platonic 'Form', the nature of which we merely have to discover. The word has been used by historians and sociologists in all sorts of different senses;[1] but I believe that the way in which Marx chose to use it is the most fruitful, for our own society and for all earlier ones above the primitive level, including Greek and Roman society. Now Marx never, unfortunately, gave a definition of the term 'class', and it is true that he uses it rather differently on different occasions, above all when he is speaking of actual historical circumstances, in which the nature of the particular classes involved could differ considerably.[2] Even when, at the very end of the unfinished third volume of Capital, pp.885-6 (cf. 618),[3] he was about to answer his own question, 'What constitutes a class?' he only had time to say that the reply to this question 'follows naturally from the reply to another question, namely: What makes wage-labourers, capitalists and landlords constitute the three great social classes?' – as indeed they did, at the period of which and during which he was writing. He did not live to write down his answer to even that prior question, which would have produced a definition of the classes of nineteenth-century capitalist society rather than of class in general; and whether he would then have gone on to give an explicit general definition of class, we cannot tell. But after collecting scores if not hundreds of passages in which Marx operates with the concept of class (sometimes without actually using that word), I have little doubt what essential form it took in his mind. (I can give only a preliminary sketch here: I shall attempt to provide a proper account in Section ii of this chapter and subsequently.)

Class as a general concept (as distinct from a particular class) is essentially a relationship; and class in Marx's sense must be understood in close connection with his fundamental concept of 'the relations of production': the social relations into which men enter in the process of production, which find legal expression to a large degree either as property relations or as labour relations. When the conditions of production, such as they are at any given time, are controlled by a particular group (when, as in the great majority of such cases,[4] there is private property in the means of production), then we have a 'class society', the classes being defined in terms of their relationship to the means and the labour of production, and to each other. Some of the most important 'means of production' in the modern world – not only factories, but also banks and finance houses, even railways and aircraft – were of course absent in Classical antiquity, and so, to a great extent, was that wage labour which is an essential element, indeed the essential element, in the relations of production characteristic of a capitalist economy. (As we shall see in III.vi below, free wage labour played an infinitely less important part in the Greek and Roman world than it does today.) In the ancient Greek world the principal means of production was land, and the principal form in which labour was directly exploited was unfree labour – that of chattel slaves above all; but debt bondage was far more widespread than many historians have realised, and in the Roman empire agricultural labour came to be exploited more and more through forms of tenancy (at first involving mainly free men), which in the late third century were converted into legal serfdom. (I shall give precise definitions of slavery, serfdom and debt bondage in III.iv below.) In antiquity, therefore, wealth may be said to have consisted above all in the ownership of land, and in the control of unfree labour; and it was these assets above all which enabled the propertied class to exploit the rest of the population: that is to say, to appropriate a surplus out of their labour.

At this point I must introduce an important and difficult subject which needs careful treatment and can easily lead to serious confusion, and which I intend to deal with properly in Chapter IV below. I refer to the fact that a large part of production in antiquity was always carried on, until the Later Roman Empire (and to a certain degree even then), by small free producers, mainly peasants, but also artisans and traders. In so far as these numerous individuals neither exploited the labour of others (outside their own families) to any appreciable extent nor were themselves exploited to any marked degree, but lived not far above subsistence level, producing little surplus beyond what they themselves consumed, they formed a kind of intermediate class, between exploiters and exploited. As I shall explain in Chapter IV, this exploitation could be not only direct and individual (by landlords or moneylenders, for instance) but also indirect and collective, affected by taxation, military conscription or forced services exacted by the state or the municipalities.

It is very hard to assess the condition of these small free producers accurately. The vast majority were what I shall call peasants (see my definition in IV.ii below), a term covering a wide variety of conditions, which nevertheless can be convenient to use, especially where we are in doubt about the precise situation of the people concerned. In Chapter IV I shall try to show the wide variety of institutions involved, and how the fortunes of some groups might fluctuate very considerably according to their political and legal as well as their economic position.

Other categories than those of class, in the sense in which I am using that concept, have of course been proposed for analysis, or at least the description, of Greek society. I shall consider some of them in Section v of this chapter.

Historians, who are usually dealing with a single society, rarely trouble themselves with any reflections about their choice of categories: they are seldom aware of any problem in this respect; often it does not even occur to them that there is any need to go beyond the concepts employed by the members of the society they are studying. Indeed, a practising historian in the British – and American – empirical tradition may well say to us (as the author of a major recent book on the Roman emperor has virtually done: see the opening of Section v of this Chapter): 'Why on earth should we waste time on all this theoretical stuff, about class structure and social relations and historical method? Why can't we just go on doing history in the good old way, without bothering about the concepts and categories we employ? That might even involve us in the philosophy of history, which is something we prefer to abandon with disdain to philosophers and sociologists, as mere ideology.' The reply to this, of course, is that it is a serious error to suppose that unconsciousness of ideology, or even a complete lack of interest in it, is the same thing as absence of ideology. In reality each of us has an ideological approach to history, resulting in a particular historical methodology and set of general concepts, whether conscious or unconscious. To refuse – as so many do – to define or even to think about the basic concepts we employ simply results in our taking over without scrutiny, lock, stock and barrel, the prevailing ideology in which we happen to have been brought up, and making much the same kind of selection from the evidence that our predecessors have been making and for the same reasons.

Nevertheless, there are very great virtues in the traditional approach of the historian, the essence of which – the insistence on recognising the specificity of the historical situation in any given period (and even area) – must not be abandoned, or even compromised, when it is combined with a sociological approach. Indeed, anyone who is not capable (whether from a deficiency of intellect or from lack of time or energy) of the great effort needed to combine the two approaches ought to prefer the strictly historical one, for even mediocre works produced by the purely fact-grubbing historian may at least, if his facts are accurate and fairly presented, be of use to others capable of a higher degree of synthesis, whereas the would-be sociologist having insufficient knowledge of the specific historical evidence for a particular period is unlikely in the extreme to say anything about it that will be of use to anyone else.

The study of ancient history in Britain has long been characterised by an attitude to detailed empirical investigation which in itself is most admirable. In a recent assessment of Rostovtzeff's great Social and Economic History of the Roman Empire, Glen Bowersock of Harvard University (who had himself been through the Oxford Greats School and was a graduate pupil of Sir Ronald Syme) has spoken of a general raising of eyebrows in Oxford when Rostovtzeff, who had come there in 1918 as an exile from his native Russia, 'announced that he would lecture on no less a subject than "The Social and Economic History of Eastern and Western Hellenism, the Roman Republic, and the Roman Empire"'. He adds, "Together with the immodest grandeur of Rostovtzeff's topic went, perhaps inevitably, an occasional cloudiness of thought'; and he records Rostovtzeff's own remark in the Preface to his book, 'Evidently the English mind, in this respect unlike the Slavonic, dislikes a lack of precision in thought or expression.'[5] Now here we come right up against a problem which faces every historian: how to reconcile full and scrupulous attention to all forms of evidence for his chosen subject and a study of the modern literature relating to it with a grasp of general historical methodology and sociological theory sufficient to enable him to make the most of what he learns. Few if any of us strike exactly the right balance between these very different desiderata. It has been said that the sociologist comes to know 'less and less about more and more', the historian 'more and more about less and less'. Most of us fall too decisively into one or other of these categories. We are like Plutarch's truly pious man, who has to negotiate a difficult course between the precipice of godlessness and the marsh of superstition (Mor. 378a), or Bunyan's Christian in the Valley of the Shadow of Death, treading a narrow path between, on the right hand, 'a very deep Ditch . . . into which the blind have led the blind in all Ages, and have both there miserably perished', and on the left, 'a very dangerous Quagg, into which, if even a good Man falls, he can find no bottom for his foot to stand on'.

I feel much happier, in dealing with the history of the ancient Greek world, if I can legitimately make use of categories of social analysis which are not only precise, in the sense that I can define them, but also general, in the sense that they can be applied to the analysis of other human societies. Class, in my sense, is eminently such a category. Nevertheless, I realise that it is a healthy instinct on the part of historians in the empirical tradition to feel the need at least to begin from the categories and even the terminology in use within the society they are studying – provided, of course, they do not remain imprisoned therein. In our case, if the Greeks did not 'have a word for' something we want to talk about, it may be a salutary warning to us that the phenomena we are looking for may not have existed in Greek times, or at any rate not in the same form as today. And so, in Section iv of this chapter, I propose to begin from the categories employed by the ancient Greeks themselves, at the time of their greatest self-awareness (the fifth and fourth centuries B.C.), to describe their own society. It will immediately become obvious that there is a striking similarity between those categories and some of the features of Marx's class analysis: this is particularly clear in Aristotle's Politics.

Let us now get down to fundamentals. I begin with five propositions. First, man is a social animal – and not only that, but, as Marx says in the Grundrisse (E. T. 84), 'an animal which can develop into an individual only in society'. (Although in the same passage Marx contemptuously and rightly dismissed the individual and isolated hunter or fisherman who serves as the starting-point for Adam Smith and Ricardo – or, for that matter, Thomas Hobbes – as an uninspired conceit in the tradition of Robinson Crusoe, it is impossible not to recall at this point Hobbes's famous description of the life of his imaginary pre-societal man, in Leviathan I.13, as 'solitary, poor, nasty, brutish, and short'.) Secondly, the prime task of man in society is to organise production, in the broadest sense, including both the acquisition from outside his society, by trade or forcible appropriation, of such necessary and desirable things as the society needs but cannot produce, or cannot profitably produce, within itself, and the distribution of what is produced. (In an area which is large or, like the Greek world, much split up by mountains or the sea, the nature of the transport system may be an important factor.) I shall use the term 'production' in this convenient, extended sense, as Marx commonly does.[6] It should hardly be necessary to add that production, in the very broad sense in which I am using the word, of course includes reproduction: the bearing and rearing to maturity of offspring (cf. Section vi of this chapter). Thirdly, in the very act of living in society and organising production, man necessarily enters into a particular system of social and economic relations, which Marx referred to as 'the relations of production' or 'the social relations of production'.[7] Fourthly, in a civilised society such as that of the ancient Greeks or ourselves, the producers of actual necessities must (for obvious reasons, to be noticed presently) produce a surplus beyond what they actually consume themselves. And fifthly, the extraction and perpetuation of such a surplus has led in practice to exploitation, in particular of the primary agricultural producers: this exploitation, with which the whole concept of class is associated, is the very kernel of what I refer to as 'the class struggle'. (I shall deal with it in Sections ii and iii of this chapter. As I shall there explain, when I speak about 'the class struggle' in the ancient world I am never thinking of a struggle on the political plane alone, and sometimes my 'class struggle' may have virtually no political aspect at all.

I should perhaps add, for the benefit of those who are accustomed to 'structuralist' terminology, that I have not found it useful or possible to draw the distinction employed by Lévi-Strauss and his school between social relations and social structure (see e.g. Lévi-Strauss, SA 279, 303-4). I shall sometimes speak of a set of social relations as a social structure, or social formation.

I am of course thinking throughout in terms of the civilised societies of the last few thousand years, which, having developed technologically far beyond the level of primitive man, have aimed at providing themselves with a sufficient and stable supply of the necessities and luxuries of civilised life, and consequently have had to devote a very considerable volume of effort to ensuring that supply. Some anthropologists have argued that by reducing their wants to a minimum, primitives existing in a favourable environment may be thought happier than men in at least the earlier stages of civilisation, and may even enjoy a good deal of leisure; but for my purposes primitive society[8] is irrelevant, since its structure is totally different from that of Graeco-Roman antiquity (let alone the modern world), and any exploitation which may exist at the primitive stage takes place in quite different ways. Moreover, primitive society has not proved able to survive contact with developed modern economies – to put it in the crudest possible way, with Hilaire Belloc (The Modern Traveller, vi),

Whatever happens we have got The Maxim gun, and they have not.

Now in a primitive food-gathering and hunting tribe the mere day-to-day provision of food and other immediate necessities and of defence against wild beasts and other tribes and so on may be virtually a whole-time job for all adult members of the tribe, at least in the sense that in practice they do not extend their economic activities much further.[9] In a civilised community, however, it is not possible for everyone to spend all his time on these basic activities: there must be at least some members of the community who have enough leisure – in the technical sense of being released from directly producing the material necessities of life – for governing and organising and administering a complex society; for defending it against outsiders, with whatever weapons may be needed; for educating the next generation and training them in all the necessary skills, over a period of perhaps ten to twenty years; for the arts and sciences (whatever stage of development these may have reached); and for the many other requirements of civilised life. Such people (or some of them) must be at least partly freed from the cruder tasks, so that they may fulfil their specialised functions. And this means that they will have to be maintained by the rest of the community, or some part of it, in return for the services they provide. The producers will now have to produce more than what they themselves consume – in other words, a surplus.[10] And 'the appearance of a surplus makes possible – which does not mean "necessary" – structural transformation in a society' (Godelier, RIE 274).

In view of the controversy which has been going on for years among economic anthropologists about the whole notion of a 'surplus', I feel it is necessary to make two observations on that concept. First, I use the term in a strictly relative sense and with (so to speak) an 'internal' application, to mean that part of the product of an individual man's labour of which he does not directly enjoy the fruit himself, and the immediate benefits of which are reserved for others. I would distinguish an 'external' application of the term surplus, namely the way in which the notion is employed by anthropologists such as Pearson, to mean something set aside by the society as a whole, or by those who make its decisions, as 'surplus to its needs', and made available for some specific purpose – feats, war, exchange with other societies, and so forth.[11] Secondly, I agree with Godelier that there is no necessary connection between the existence of a surplus and the exploitation of man by man: there may at first be exchange considered profitable by both sides, with certain persons taking upon themselves services genuinely performed on behalf of the whole community[12] – its defence against attack from outside, for example.[13] The precise point in history at which exploitation should be conceived as beginning is difficult to decide, and I have not made up my own mind. The question is not important for my present purposes, because exploitation began long before the period with which I am concerned in this book. Perhaps we could say that exploitation begins when the primary producer is obliged to yield up a surplus under the influence of compulsion (whether political, economic or social, and whether perceived as compulsion or not), at any rate at the stage when he no longer receives a real equivalent in exchange – although this may make it very difficult to decide the point at which exploitation begins, since it is hard to quantify, for example, military protection against agricultural produce (cf. IV.iv below). A much more sophisticated definition of exploitation (which may well be preferable) has been offered by Dupré and Rey on the basis of their anthropological fieldwork in west Africa: 'Exploitation exists when the use of the surplus product by a group (or an aggregate) which has not contributed the corresponding surplus of labour reproduces the conditions of a new extortion of surplus labour from the producers' (RPTHC 152, my italics). Although even a good and fully socialist society must arrange for 'surplus labour' by some, to support the very young, the aged and the infirm, and to provide all kinds of services for the community (cf. Marx, Cap. III.847, 876), it would necessarily do so in such a way that no individual or group of individuals had a right to appropriate the fruits of that 'surplus labour' in virtue of any special control over the process of production through property rights, or indeed except at the direction of the community as a whole or its organs of government.

In every civilised society there has been a basic problem of production: how to extract a sufficient surplus ('sufficient' in a relative sense, of course) from the primary producers, who are not likely to relish their position at the base of the social pyramid and will have to be subjected to a judicious mixture of persuasion and coercion – the more so if they have come to see the favoured few as exploiters and oppressors. Now men's capacity to win for themselves the freedom to live the life they want to live has always been severely limited, until very recently, by inadequate development of the productive forces at their disposal.

All emancipation carried through hitherto has been based on restricted productive forces. The production which these productive forces could provide was insufficient for the whole of society and made development possible only if some persons satisfied their needs at the expense of others, and therefore some – the minority – obtained the monopoly of development, while others – the majority – owing to the constant struggle to satisfy their most essential needs, were for the time being (i.e. until the creation of new revolutionary forces) excluded from any development (MECW V.431-2, from the German Ideology; cf. Cap. III.820, quoted in I.iv above).

If I were asked to name the fundamental features of ancient Greek society which most distinguish it from the contemporary world, I would single out two things, closely connected, which I shall describe in succession. The first, within the field of what Marx called 'the forces of production', is a technological distinction. The advanced countries of the modern world have immense productive power. But go back to the ancient world, and you go down and down the technological ladder, so to speak. The Greek world, compared with the modern one, was very undeveloped technologically, and therefore infinitely less productive.[14] Great advances in technology occurred long before the Industrial Revolution, in the Middle Ages and even the Dark Ages. These advances were far more important than most people realise, not only in the most essential sphere of all, that of sources of energy or 'prime movers' (which I shall come to in a moment), but in all sorts of other ways. To take only one example – I wonder how many people who have not only read Greek and Roman monuments have noticed the absence from antiquity of the wheelbarrow, which at least doubles a man's carrying capacity, but only appears in Europe in the thirteenth century (in China it was known a thousand years earlier).[15] As for sources of energy, I will say only that animal power, in the form of the tractive effort of the horse and ox, was nothing like fully realised in Classical antiquity, in particular because of the extreme inefficiency of the ancient horse-harness;[16] and that only in the Middle Ages do we find the widespread utilisation of two important forms of energy which were very little used in antiquity: wind and water (cf. n.14 below). Wind, of course, was used for the propulsion of merchant ships, though not very efficiently and without the stern-post rudder;[17] but the windmill was not known in Europe before (or not much before) the early twelfth century. The water-mill[18] (hydraletēs) was actually invented not later than the last century B.C.: the last earliest known mention is by the Greek geographer Strabo, in a reference to Pontus, on the south shore of the Black Sea, in the 60s B.C. (XII.iii.30, p.556). But the most fascinating piece of evidence is the delightful poem in the Greek Anthology, by Antipater of Thessalonica, to which I referred in I.iv above as being known to Marx: the poet innocently assures the slave mill-girls that now they have the water-nymphs to work for them they can sleep late and take their ease (Anth. Pal. IX.418: see Cap. I.408). There is a little evidence, both literary and archaeological, for the use of the water-mill in the Graeco-Roman world, but it was rare before the fourth and fifth centuries, and its full use comes a good deal later (see n.14 again). Marx realised that 'the Roman Empire had handed down the elementary form of all machinery in the water-wheel' (Cap. I.348).

That is the essential background to my second basic distinction between the ancient and the modern world, which is intimately connected with the first and indeed largely grew out of it. In the ancient world, as we have seen, the producers, as I am calling them (men engaged in essential economic activities), produced a very much smaller surplus than is necessary to sustain a modern advanced society. This remains vitally important, even if we allow for the fact that the average Greek had a far more restricted range of wants and demanded a much lower standard of living than the modern Englishman, so that the volume of production per head could be well below what it has to be today. But even if we make allowance for this the disparity is still very striking. As I have shown, the ancient world was enormously less productive than the modern world. Therefore, unless almost everyone was to have to work practically all the same, and have virtually no leisure, some means had to be found of extracting the largest possible surplus out of any rate a considerable number of those at the lowest levels of society. And this is where we come face to face with the second of my two fundamental distinctions between the ancient and the modern world, one that occurs this time in the field of what Marx called 'the relations of production': the propertied classes in the Greek and Roman world derived their surplus, which freed them from the necessity of taking part in the process of production, nor from wage labour, as in capitalist society, but mainly from unfree labour of various kinds. The ancient world knew other forms of unfree labour than strict 'slavery' ('chattel slavery', if you like), in particular what I shall call 'serfdom' and 'debt bondage' (see III.iv below). But in general slavery was the most important form of unfree labour at the highest periods of Greek and Roman civilisation; and the Greeks and Romans themselves always tended to employ the vocabulary of actual slavery when referring to other forms of unfree labour.

I have indicated that it is above all in relation to its function of extracting the maximum surplus of those primary producers who were at the lowest levels of ancient society that I propose to consider slavery and other forms of unfree labour in this book. In treating slavery in this manner I am looking at it in very much the way that both masters and slaves have commonly regarded it. (Whether the ancient belief in the efficiency of the institution of slavery in this respect is justified or not is irrelevant for my purposes.) Perhaps I may cite here the opening of the third chapter of one of the best-known books on North American slavery, Kenneth Stampp's The Peculiar Institution (p.86):

Slaves apparently thought of the South's peculiar institution chiefly as a system of labour extortion. Of course they felt its impact in other ways – in their social status, their legal status, and their private lives – but they felt it most acutely in the lack of control over their own time and labour. If discontented with bondage, they could be expected to direct their protests principally against the master's claim to their work.

The feature of slavery which made it appropriate and indeed essential and irreplaceable in the economic conditions of Classical antiquity was precisely that the labour it provided was forced. The slave, by definition, is a man without rights (or virtually without effective rights) and therefore unable to protect himself against being compelled to yield up a very large part of what he produces. Dio Chrysostom, in the early second century of the Christian era, reports an imaginary discussion about slavery in which there was general agreement about the basic definition of the slave's condition: that someone else 'owns him as master, like any other item of property or cattle, so as to be able to make use of him at his pleasure' (Orat. XV.24).

I suggest that the most profitable way of approaching the problem of unfree labour is to think of it in precisely the way in which I have introduced it, in terms of the extraction of the largest possible surplus from the primary producers. I think that in antiquity slavery probably did provide the best possible answer, from the purely economic point of view (that is to say, disregarding all social as well as moral factors), having regard to the low level of productivity, and also to the fact that free, hired labour was scarce, largely confined to unskilled or seasonal work, and not at all mobile, whereas slaves were available in large numbers and at prices the lowness of which is astonishing, in comparison with what is known of slave prices in other societies. But given these conditions – the poor supply of free, hired labour, the easy availability of slaves, their cheapness, and so on – I do believe that slavery increased the surplus in the hands of the propertied class to an extent which could not otherwise have been achieved and was therefore an essential precondition of the magnificent achievements of Classical civilisation. I would draw attention to the fact that the distinction I have just drawn is based not on a difference of status, between slaves and free men, but on a difference of class, between slaves and their owners – a very different matter. (I shall return to this difference later: see Sections iii and v of this chapter.)

It may not have been fully obvious that so far I have been preparing the ground for the definition of the terms 'class' and 'class struggle' which I shall offer in Section ii of this chapter. I had to make clear certain fundamental features of ancient Greek society. I had to make clear certain fundamental features of ancient Greek society. I have now explained one of these, the essential part played by what I am calling unfree labour; and I must now briefly mention another, the fact that by far the most important means of production in the ancient world was land. Wealth in Classical antiquity was always essentially landed wealth, and the ruling classes of all the Greek states, as of Rome itself, invariably consisted mainly of landowners. This is something which most ancient historians now realise; but the whole question, like that of slavery and other forms of unfree labour, will require a more extended discussion than I can give it at this point (see III.i-iii below).

In seeking to use the concept of class as a method of historical analysis there are two quite different dangers that we must guard against: one, a matter of definition, is a question strictly for the historian. After stating them together, I shall briefly discuss them separately. First, we must be quite sure what we mean by the term 'class' (and 'class struggle'), and not slide carelessly and unconsciously from one interpretation to another. Secondly, we must be careful to make a correct historical identification of any class we propose to recognise.

1. The first problem, that of definition, is of a sociological nature. Marx himself, as I said earlier, never gave a definition of class in general terms. Some may feel that no such general definition is possible, but I believe the one I shall produce in Section ii below will serve well enough, although there may be some special cases in which a unique set of historical circumstances makes qualification necessary. Even if it could be shown that there are too many exceptions for my definition to be considered a general one, I would a least claim that it holds for the society, or rather series of societies, of the Graeco-Roman world, discussed in this book. I hope that others will improve upon it.
2. The second problem is purely historical: one must thoroughly understand the particular society one is considering, and know the evidence about it at first hand, before one can expect to identify its classes correctly and precisely. Some serious mistakes have been made in defining the actual classes existing in particular societies, and the results of employing unreal conceptions of those classes, not corresponding closely with reality, have sometimes been disastrous. Misconceptions about classes existing in historical societies have not, of course, been confined to Marxists, by any means, but since they make more use of class categories than other historians they are likely to commit even worse blunders if they start out with misconceptions about the classes they recognise. It has been standard practice among ancient historians to refer to the governing classes of several Greek cities in the Archaic and Classical periods, in particular Aegina and Corinth, as 'commercial aristocracies' or 'industrial and merchant classes' (see my OPW 264-7, esp. n.61; cf. 216, 218-20, and Appendix XLI, esp. p.396). This extraordinary notion, for which there is not a shred of ancient evidence, was adopted without examination by Busolt, Eduard Meyer and other leading historians (even Max Weber was not entirely free of it), and it is still being reproduced today in some quarters. Not a few Marxists have started out from similarly mistaken positions. It is not surprising that attempts by George Thomson (essentially a literary scholar and not a historian in the proper sense) to expound the intellectual development of the Classical Greek world in Marxist terms have not succeeded in convincing historians or philosophers; for Thomson presents the development of Greek thought, and even of Greek democracy, in the sixth and fifth centuries as the consequence of the rise to power of a wholly imaginary 'merchant class'. Thomson even describes the Pythagoreans of Croton as 'the new class of rich industrialists and merchants', who 'resembled Solon in being actively involved in development of commodity production'.[19] In my opinion, this is little better than fantasy. The one book I know in English which explicitly seeks to give an account of Greek history (before the Roman period) in Marxist terms is a prime example of the methodological catastrophe involved in giving a would-be Marxist account in terms of classes that are fictions and correspond to no historical reality. The author, Margaret O. Wason, pretends that in the seventh and sixth centuries, in most Greek states, there came to power a 'new bourgeois class', defined as 'the class of merchants and artisans which challenged the power of the aristocracy'. It is no surprise to find Cleon referred to in the same book as 'a tanner' (this of course reproduces Aristophanes' caricature; cf. my OPW 235 n.7, 359-61, 371) and as 'the leader of the Athenian workers'.[20]

I may add that it would similarly be absurd to speak of a 'class struggle' between Senators and Equites in the Late Roman Republic. Here I am in full agreement with a number of non-Marxist ancient historians of very different outlooks. As P. A. Brunt and Claude Nicolet have so conclusively demonstrated in the last few years, the Equites were part of the class of large landowners to which the Senators also belonged. As Badian has put it, for the Senate they were simply 'the non-political members of its own classes'[21] – those who preferred not to take upon themselves the arduous and often dangerous life that a political career would involve. At certain times a purely political contest might develop between these two groups within the propertied class on specific issues, but this must not mislead us into seeing them as two separate classes having irreconcilable interests. I shall in fact speak sometimes of the Roman Senators (though not the Equites) as a class: the 'senatorial class'. It is possible that some other Marxists may prefer not to break down my 'propertied class' (for which see III.ii below) into two or more classes for certain purposes, as I do – for example, in the developed Principate and the Later Empire, primarily into the senatorial and curial classes, with the Equites perhaps as a kind of sub-class closely attached to the Senators, until in the late fourth and early fifth centuries they were entirely absorbed into the senatorial class (see VI.vi below, ad fin.) But in my set of definitions, early in Section ii of this chapter, I allow for Rechtsstellung (legal or constitutional situation) as a factor that can help to determine class in so far as it affects the type and degree of exploitation practised or suffered; and the constitutional privileges enjoyed by Senators surely did materially increase their capacity to exploit – just as the condition of being a slave, with its severe juridical disabilities, greatly increased the slave's liability to exploitation. But I could quite understand if some other Marxists, feeling that it was above all their great wealth which lay at the root of the Senators' privileged position, rather than the office-holding and the consequential legal privileges it brought them, preferred to treat the Senators merely as an 'order' (which they certainly were) rather than a class. Perhaps 'sub-class' would be a convenient term; but I have avoided it.

I have only one more preliminary point to make before proceeding to a definition of my terms: I am deliberately avoiding, at this stage, discussion of the terms 'caste', 'order', 'estate' (état). Caste is a phenomenon which we do not encounter at all in the Greek or Roman world.[22] We do find what can legitimately be described as 'orders' (or 'estates') – that is to say, status-groups (Stände) which are legally recognised as such and have different sets of juridical characteristics (privileges or disadvantages). Such groups will be noticed when we have occasion to discuss them. I shall have something to say of 'status-groups' in general, and (in Section v of this chapter) of 'status' as an alternative concept to 'class'. But although I shall of course refer at times to particular 'orders' (citizens, slaves, freedmen, senators, equestrians, curials), I shall take no special account of 'orders' as such, treating them as a rule merely as a special form of status-group, except in so far as they materially affect the degree of exploitation concerned (cf. the preceding paragraph).

(ii) - 'Class', 'exploitation', and 'the class struggle' defined

We can now attempt to define 'class', 'exploitation', and 'class struggle'. As I said in Section i of this chapter, I am not going to pretend that there is an objective entity, class, the nature of which remains to be discovered. I would also deny that there is any definition of class which is so generally agreed upon that we are all obliged to accept it or run the risk of being accused of perversity. The concept has been discussed ad nauseam by sociologists during the past few decades (cf. n.1 to Section i above). After working through a good deal of the literature, most of which seems to me almost worthless, I feel entitled to insist from the outset that the disagreement about the best way of using the expression 'class' has been so great that anyone who attempts an analysis of any society in terms of class is entitled to establish his own criteria, within very wide limits, and that our verdict on the definition he adopts ought to depend solely on its clarity and consistency, the extent to which it corresponds with the historical realities to which it is applied, and its fruitfulness as a tool of historical and sociological analysis. If in addition we find (as we shall in this case) that the notion of class in the sense in which we define it corresponds closely we are examining (in our case, that of Aristotle especially: see Section iv of this chapter), then we shall be fortunate indeed.

I should like to quote here a statement by a leading British sociologist, T. B. Bottomore, raising questions which are all too unfamiliar to many historians. Speaking of the construction of general concepts by sociologists, he says:

In some recent attempts to improve the 'conceptual framework' of sociology, and notably in that of Talcott Parsons and his collaborators, the whole emphasis is placed upon definition of concepts rather than upon the use of concepts in explanation. This is a retrograde step by comparison with the work of Durkheim and Max Weber, both of whom introduced and defined concepts in the course of working out explanatory theories. Weber's exposition of his 'ideal type' method deals more clearly with this matter than any later writing, and had his ideas been followed up sociology would have been spared much confused and aimless discussion. In essentials his argument is that the value of a definition (i.e. of a concept) is only to be determined by its fruitfulness in research and theorising (Sociology2 [1971] 37, cf. 121).

I should not like it to be thought, however, that I regard Marx's concept of class as a Weberian 'ideal-type construct', in the sense that Weber himself took it to be. For me, as for Marx, classes and class struggles are real elements which can be empirically identified in individual cases, whereas for Weber all such 'Marxian concepts and hypotheses' become 'pernicious, as soon as they are thought of as empirically valid' (Weber, MSS 103, repr. in Eldridge, MWISR 228).

I propose first to state my definition of class and class struggle, and to explain and justify it in subsequent discussion. I believe that this definition represents the central thought of Marx as accurately as possible: this claim too I shall try to justify.

Class (essentially a relationship)[1] is the collective social expression of the fact of exploitation, the way in which exploitation is embodied in a social structure. By exploitation I mean the appropriation of part of the product of the labour of others:[2] in a commodity-producing society this is the appropriation of what Marx called 'surplus value'.

A class (a particular class) is a group of persons in a community identified by their position in the whole system of social production, defined above all according to their relationship (primarily in terms of the degree of ownership or control) to the conditions of production (that is to say, the means and labour of production)[3] and to other classes. Legal position (constitutional rights or, to use the German term, 'Rechtsstellung') is one of the factors that may help to determine class: its share in doing so will depend on how far it affects the type and degree of exploitation practised or suffered – the condition of being a slave in the ancient Greek world, for example, was likely (though far from certain) to result in a more intense degree of exploitation than being a citizen or even a free foreigner.

The individuals constituting a given class may or may not be wholly or partly conscious of their own identity and common interests as a class, and they may or may not feel antagonism towards members of other classes as such.

It is of the essence of a class society that one or more of the smaller classes, in virtue of their control over the conditions of production (most commonly exercised through ownership of the means of production),[4] will be able to exploit – that is, to appropriate a surplus at the expense of – the larger classes, and thus constitute an economically and socially (and therefore probably also politically) superior class or classes. The exploitation may be direct and individual, as for example of wage-labourers, slaves, serfs, 'coloni', tenant-farmers or debtors by particular employers, masters, landlords or moneylenders, or it may be indirect and collective, as when taxation, military conscription, forced labour or other services are exacted solely or disproportionately from a particular class or classes (small peasant freeholders, for instance) by a State dominated by a superior class.

I use the expression class struggle for the fundamental relationship between classes (and their respective individual members), involving essentially exploitation, or resistance to it. It does not necessarily involve collective action by a class as such, and it may or may not include activity on a political plane, although such political activity becomes increasingly probable when the tension of class struggle becomes acute. A class which exploits others is also likely to employ forms of political domination and oppression against them when it is able to do so: democracy will mitigate this process.

Imperialism, involving some kind of economic and/or political subjection to a power outside the community, is a special case, in which the exploitation effected by the imperial power (in the form of tribute, for instance), or by its individual members, need not necessarily involve direct control of the conditions of production. In such a situation, however, the class struggle within the subject community is very likely to be affected, for example through support given by the imperial power or its agents to the exploiting class or classes within that community, if not by the acquisition by the imperial power or its individual members of control over the conditions of production in the subject community.

There is one aspect of my definition of class which, I realise, may need clarification. Not all individuals belong to one specific class alone: some can be regarded as members of one class for some specific class alone: some can be regarded as members of one class for some purposes and of another class for others, although usually membership of one will be much the most significant. A slave who was allowed by his master to accumulate a considerable peculium, and who (like Musicus Scurranus, mentioned in III.iv below, at its n.13) had even acquired under-slaves of his own, vicarii, might have to be regarded pro tanto as a member of what I am calling 'the propertied class'; but of course his membership of that class would necessarily be qualified and precarious and dependent on the goodwill of his master. A slave who was settled by his landowning master as tenant of a small farm, quasi colonus (see IV.iii § 12 below), would in strictly economic terms be in much the same position as a poor free peasant leaseholder, and we might be inclined to put him in the class of peasants (see IV.ii below); but his legal status would remain greatly inferior and his tenancy would be much more at the pleasure of the landowner, who could therefore exploit him more severely if he were so inclined. And a poor peasant who owned or leased a plot of land so small that he regularly needed to betake himself to a neighbouring city for part of the year to earn wages would be a member of two classes: small peasants and wage-labourers. I also maintain in Section vi of this chapter that women, or at any rate married women (and so the great majority of adult women in antiquity), must be regarded for some purposes as a distinct class, although membership of such a class (because of its consequences for property-ownership) would in a city like Classical Athens be far more important to a high-born woman than to a poor peasant, who would have had no opportunity to own much property had she been a man and whose membership of the class of women would therefore be of far less significance.

Of course I have no wish to pretend that class is the only category we need for the analysis of Greek and Roman society. All I am saying is that it is the fundamental one, which over all (at any given moment) and in the long run is the most important, and is by far the most useful to us, in helping us to understand Greek history and explain the process of change within it. In Section v of this chapter I shall briefly consider alternative approaches, particularly those which have the primary aim – as I have not, and as Marx did not (see Section v) – of establishing a scheme of 'social stratification' according to 'status'. Such activities are perfectly legitimate and may even have quite useful results, provided we keep them in their proper place and realise that they will not by themselves disclose the real secrets of history: the springs and causes of human behaviour and social change. I would say that social status, and even in the long run political power, tended to derive from class position in the first place (as indeed political status always did directly in the commonest form of Greek oligarchy in the Classical period, based on a property qualification), and that in the long run distinctions having any other basis than the economic tended to decay in favour of, and ultimately to resolve themselves into, distinctions based upon economic class. (We shall notice some examples of this process later: see V.iii and VIII.i and ii below.)

Let us be quite clear about one thing. Whereas descriptions of ancient society in terms of some category other than class – status, for instance – are perfectly innocuous, in the sense that they need have no direct relevance to the modern world (which will of course need to be described in terms of a completely different set of statuses), an analysis of Greek and Roman society in terms of class, in the specifically Marxist sense, is indeed (to use Firth's adjective: see I.iv above), something threatening, something that speaks directly to every one of us today and insistently demands to be applied to the contemporary world, of the second half of the twentieth century. If Marx's analysis, originally derived above all from the study of nineteenth-century capitalist society, turns out to be equally well adapted not merely to describe ancient society over a long period of many centuries but to explain its transformations and its partial disintegration (as we shall see it is), then its relevance for the contemporary world becomes very hard to ignore. Of course in some quarters it will be ignored. To quote Marx and Engels, addressing themselves sarcastically in 1848 to the ruling classes of their day:

The selfish misconception that induces you to transform into eternal laws of nature and reason the social forms springing from your present mode of production and form of property – historical relations that rise and disappear in the progress of production – this misconception you share with every ruling class that has preceded you. What you see clearly in the case of ancient property, what you admit in the case of feudal property, you are of course forbidden to admit in the case of your own bourgeois form of property (MECW VI.501, from the Communist Manifesto).

I shall now glance briefly at the use of the conception of class (and class struggle) by Marx himself. I shall maintain that for five different reasons in particular there has been a widespread and serious misunderstanding of the part this idea played in Marx's thought. I believe that my definition represents his fundamental thinking more accurately than do the statements of some modern Marxist and non-Marxist writers who have taken different views from mine. My five reasons are as follows.

First, partly perhaps because of a much-quoted definition by Lenin, in his A Great Beginning, which (as Ossowski says, CSSC 72 and n.1) has been 'popularised by Marxist text-books and encyclopaedias', it has been customary to lay particular stress on relationship to the means of production as the decisive factor (sometimes as the one essential factor) in determining a person's class position. Although his formulation contains a profound truth, it will be seen from the definition of class given above that I regard it as a rather too narrow conception. Secondly, as is well known, Marx himself, although he made important use of the concept of class throughout his work, never gave a formal definition of it, and indeed employed it in very different senses at different times. Thirdly, Marx himself was concerned in his writings almost entirely with a capitalist society which had already undergone a considerable process of development: apart from one section of the Grundrisse (E.T. 471-514) which is specifically devoted to 'pre-capitalist economic formations' (see the excellent edition by Hobsbawm, KMPCEF), the statements in his work about pre-capitalist societies in general and the Graeco-Roman world in particular are all brief, and many of them are in the nature of obliter dicta. In these passages, as a rule, he takes no pains to be precise over terminology. Fourthly (and as a consequence of the facts I have just stated), when Marx spoke in particular about 'class struggle' he tended – thinking almost always, as he was, of nineteenth-century capitalism – to have in mind the kind of class struggle which was so noticeable in the mid-nineteenth century in the more developed capitalist countries: namely, open class struggle on the political plane. Thus when, for example, he spoke in The Eighteenth Brumaire of Louis Bonaparte of the French bourgeoisie as 'doing away with the class struggle for the moment by abolishing universal suffrage' (MECW XI.153), he simply meant that the law of 31 May 1850, by reducing the total number of electors from ten to seven million (id. 147), made it far harder for the French working class to carry on effective political struggle. And finally, in the work often wrongly taken to be the definitive statement of Marx's 'materialist conception of history', namely the Preface to A Contribution to the Critique of Political Economy (1858-9), we find only a passing reference to classes and none at all to class struggle. There is, however, a perfectly good explanation of this, well brought out by Arthur M. Prinz in an article in the Journal of the History of Ideas 30 (1969) 437-50, entitled 'Background and ulterior motive of Marx's "Preface" of 1859'. The Preface was to be published (through the good offices of Lassalle) in Berlin, and it was absolutely necessary for Marx to take careful account of the stringent Prussian censorship and abstain from anything that might be suspected of incitement to class hatred, at that time an actual offence punishable with imprisonment under para. 100 of the Prussian Penal Code. Marx, already well known to the Prussian censors, was now living in England and in no danger of prosecution himself; but he had to be circumspect if there was to be any hope of finding a publisher, for the same paragraph of the Penal Code also prescribed the penalty of confiscation for any offending work. Yet Marx had to publish in Germany, in order to make a bid for the intellectual leadership of the German socialist movement. The Preface, then, had to steer clear of class struggle. But when on 17/18 September 1879 Marx and Engels – thinking back to the Communist Manifesto and beyond – wrote to Bebel, Liebknecht, and others, 'For almost forty years we have stressed the class struggle as the immediate driving power of history' (MESC 395), they were making a perfectly correct statement. Even in those considerable parts of Marx's writing which are concerned entirely with economics or philosophy rather than with the historical process he will sometimes show that the class struggle is ever-present in his mind, as when in a letter to Engels on 30 April 1868 he rounds off a long passage on economics with the words, 'Finally . . . we have as conclusion the class struggle, into which the movement of the whole Scheiss is resolved' (see MESC 250).

From reactions I have had to drafts of this chapter, I know that some people will protest against what will seem to them an excessive emphasis on collective entities, classes, at the expense of 'the individual'. To any such objection I would reply that my main aim in this book is to explain 'what happened in history' on a large scale: the history of the Greek world as a whole over more than 1,300 years – dare I use the rather repellent expression, 'macro-history'? But the history of 'macro-units' (of classes, as of states and alliances) needs to be explained in terms very different from those appropriate to the behaviour of human groups in organised States. Elsewhere I have explained at length how Thucydides – rightly, in my opinion – recognised that the canons of interpretation and judgment applicable to the actions of States are fundamentally different from those we apply to the actions of individuals (see my OPW 7 ff., esp. 16-28). I now wish to advance the following propositions: that the factors governing the behaviour of classes (in my sense) are different again from either of the sets I have just mentioned; that the behaviour of a class as such (that of men as members of a class) may well be inexplicable in terms we can legitimately apply to their behaviour as individuals; and even that a given individual or set of individuals may behave as a constituent part of a class in a way that is quite different from the behaviour we are entitled to expect of him or them as individuals.

If in that last sentence we substitute 'a state' for 'a class', there may be little objection, since the moral standards generally accepted as governing the conduct of individuals are clearly quite different from those applied to the behaviour of states: a man who participated in the bombing of Hiroshima or Nagasaki, Berlin or Dresden, Vietnam or Laos, will not be accounted a mass murderer by most people, because he was acting in the interests – or at any rate on the orders – of his own state, against an 'enemy' state; and those who gave the orders suffered no criminal indictment, for in the event they were not defeated. It would be similarly easy to find examples from the ancient world that would be universally considered morally atrocious behaviour on the part of individuals acting in their own personal interests, but were yet regarded as unobjectionable and even praiseworthy when employed in the service of the state. Most of the Roman generals against 'barbarians' or rebels which are noticed, for example, by Ammianus Marcellinus (a Greek historian who wrote in Latin) are recorded without any sign of disapproval;[5] and the same historian could mention without comment the opinion of 'lawyers of old' that sometimes even the innocent may be put to death (XXVII.ix.5), and felt no need to shed any tears over the wholesale extermination of the children of the Maratocupreni, fierce and wily robbers (XXVIII.ii.11-14). I suspect, however, that many people would be far less willing to accept the propositions advanced at the end of the last paragraph in regard to classes, which I will now demonstrate.

That slaves who rebelled, or who could even be held guilty of failing to protect their masters from being assassinated by one of their own number, were treated with pitiless ferocity by the Romans is well known: I have given one or two prominent examples in VII.i below. The relationship of the Spartans to their Helots – very much a class relationship, of exploiter to exploited – was one of quite extraordinary hostility and suspicion. In III.iv below I draw attention to the remarkable fact that each set of Spartan ephors, upon taking office, made an official declaration of war on their work-force, the Helots, so as to be able to kill any of them without trial and yet avoid incurring the religious pollution such acts would otherwise have entailed. The Greeks on the whole showed less savagery than the Romans towards their slaves; but even in Classical Athens, where we hear most about relatively good treatment of slaves, all our literature takes the flogging of slaves for granted.

Literary sources in abundance from all over the Greek world show that this form of punishment for slaves was commonplace. An epitaph on the tomb of a virtuous matron, Myro (who may be an imaginary character), by the Hellenistic poet Antipater of Sidon, describes quite casually, as if it were the most natural thing in the world, the depiction on her tomb of (among other things) a whip, as a sign that Myro was a 'just character of misdeeds' – though not, of course, a 'cruel or arrogant mistress'! (Anth. Pal. VII.425). No one will doubt that refractory slaves were repressed without mercy, at any rate in so far as this could be done without excessive damage to the interests of their masters, whose property they were (cf. III.iv below).

Whom among our main literary sources might we have thought less likely to order a slave to be flogged than Plutarch? – a man conspicuous, surely, for his humanity. But there is a nasty little story which has come down to us from Calvisius Taurus, a friend of Plutarch's, through Aulus Gellius (NA I.xxvi.4-9). An educated slave of Plutarch's who knew his master's treatise On freedom from anger (Peri aorgēsias, usually referred to by its Latin title, De cohibenda ira) protested, while being flogged, that Plutarch was being inconsistent and giving in to the very fault he had reprobated. Plutarch was quite unabashed. Insisting that he was perfectly calm, he invited the slave to continue the argument with him – in the same breath ordering the flogger to continue applying the lash. The incident was quoted by Taurus, in reply to a question by Gellius at the end of one of his philosophical lectures, and with complete approval. But we need not be surprised in the least at Plutarch's action, if we can bring ourselves to see this particular slaveowner and his slave as 'but the personifications of the economic relations that existed between them' (Marx, Cap. I.84-5).

The class struggle between the propertied class and those who were relatively or absolutely propertyless was also accompanied at times by atrocities on both sides: see e.g. V.ii below. When we hear of particularly murderous behaviour by those who had the upper hand in a stasis (a civil commotion), we can be reasonably safe in concluding that the conflict was basically between social classes, even if our information about it is not explicit.[6]

I forbear to cite contemporary examples of the conduct of class warfare in ways which have been widely accepted as 'necessary' but which have involved behaviour that would be condemned by everyone as morally indefensible in actions between individuals.

PART TWO

APPENDICES