`The decline of Rome was the natural and inevitable effect of immoderate greatness. Prosperity repined the principle of decay; the causes of destruction multiplied with the extent of conquest; and as soon as time or accident had removed the artificial supports, the stupendous fabric yielded to the pressure of its own weight. The story of its ruin is simple and obvious; and instead of inquiring why the Roman empire was destroyed, we should rather be surprised that it lasted so long.' - Edward Gibbon.'
In 476 the last Roman emperor to rule from Italy was deposed at Ravenna. Romulus Augustulus was in his early teens, the puppet of his father who commanded the imperial army. It was not much of an army, but then they no longer controlled much of an empire. The east was ruled by another emperor at Constantinople and he did not recognise the pretender in Italy. Most of the western provinces - Gaul, Spain and North Africa - had been carved up into kingdoms by warlords of Germanic origin. Now the same fate would befall Italy as an army officer of barbarian origin called Odoacer killed Romulus' father and deposed the emperor. The lad himself was not important enough to be worth killing and was permitted to live out the rest of his life in comfortable retirement. There was a bitter irony that he should be named Romulus after Rome's mythical founder and nicknamed `little Augustus' after the first emperor Augustus.
It has been common to name 476 as the year when the Roman Empire ended in the west. If so, then five centuries of imperial rule ended with a whimper. The event did not seem to be of massive importance to contemporaries, and probably passed unnoticed by most of the emperor's subjects. Romulus Augustulus was just the last in a succession of puppet emperors manipulated by powerful generals. The empire had split into eastern and western halves each ruled by its own emperor near the end of the fourth century. The east remained strong, but the west had withered, its wealth and power declining under a succession of blows. By 476 the Western Empire did not have far to fall. In the next century the east Romans would attempt to regain the lost territories, occupying Italy, Africa and part of Spain. They lacked the strength and will to hold on to them in the long run.
The eastern part of the empire - known as the Byzantine Empire to modern scholars, but Roman as far as they were concerned - was a powerful state at the end of the sixth century. Yet it was not a superpower, and its wealth and military might were a pale shadow of the united empire in its heyday, when no enemy or rival had been even remotely Rome's equal. The time when the emperors had governed most of the known world was just a distant memory. By the year boo the world was a very different place. No new superpower had emerged to take Rome's place and instead there were many smaller kingdoms and peoples. The medieval world had taken shape.
There have been a huge number of theories to explain why the world changed in this way, and very little agreement. Many dispute the importance of 476, even as a landmark. Some argue that the empire had already fallen before this and a few, somewhat bizarrely, that it survived afterwards. Not only are the causes of Rome's fall disputed, but also how long the process took. Some, like Gibbon, see the roots deep in the earlier history of the empire, which produced a slow decline over several centuries. Others suggest a shorter time span, although virtually no one has argued that it took less than a few generations. Debate continues to rage, each age answering the question according to its own obsessions and prejudices. The fall of the Roman Empire remains one of the great mysteries of history.
More recent empires have risen and fallen much more quickly. Hitler's thousand-year Reich' and its ally imperial Japan enjoyed spectacular success, both reaching the height of their power in 1942. Three years later they fell in blood and ruin, their power utterly broken. The Second World War also hastened the end of much older empires, whose impact on the wider world was deeper, if often more subtle. Exhausted and impoverished by war, Britain most readily acknowledged the `wind of change' and gave up its empire in just a few decades. Wars were fought to defeat groups determined to seize power by force, but the inevitability of independence was never seriously doubted. Other countries resisted the change more stubbornly, but all failed to cling on to their colonies in the long run.
The great powers of the eighteenth and nineteenth centuries were a spent force, but they left a deep legacy. Newly independent countries had frontiers based on the decisions of imperial administrators - dramatically so where partition was employed, but more generally and less deliberately throughout Africa and Asia. Much of the world now had English, Spanish or French as a second language, which was very often the language of government and education. Legal and political systems were also derived from European prototypes. Ironically enough Latin law in this way spread to a far wider area than the Roman Empire had ever covered. Control passed almost invariably to an elite drawn from the indigenous population, but who were educated in the European style, and often actually in the country of the colonial power. Rarely is it possible to say more than that the wider population has been no worse off since independence, but all too often the new rulers have proved considerably more corrupt and exploitative than their predecessors. Former colonies now form the bulk of the poorest countries in the world.
Soviet Russia, which had inherited the empire and many of the ambitions of its Tsarist predecessor, survived longer than the west European powers and for forty years was one of two superpowers that dominated the world. Finally, Russia collapsed under its own weight. This happened very suddenly, surprising even its Cold War adversaries. The fate of many regions on Russia's fringes remains to be decided, but has already involved considerable bloodshed in several areas. Soviet Russia's fall left the United States of America as the sole superpower in the world, a situation that at the moment seems only likely to change if the forecasts of China's growth prove accurate. (The idea that the EC may become an equal is clearly fantasy. The periodic suggestions that it could join with Marxist China to form a counterweight to the USA are disturbing, but scarcely realistic.)
Once a colony itself, America became a country through rebellion from Britain. Apart from the expansion westwards, it has never shown much interest in occupying overseas territories, as distinct from maintaining bases around the world. Even so, the Cold War led to fighting open wars in Korea and Vietnam, as well as covert support for combatants in many other countries. Currently, the USA and its allies have substantial forces in Afghanistan and Iraq. In each case this is intended to be a temporary operation, until the supported governments are capable of maintaining themselves without direct military aid. Opponents often dub America an empire, but this is largely rhetoric. However, it is overwhelmingly the strongest country in the world and in this sense its position mirrors that of Rome. Yet the very different experiences of other modern empires should make us cautious about pressing this too far. First we must understand the Roman experience.
There is some irony in the coincidence that the first volume of Edward Gibbon's The History of the Decline and Fall of the Roman Empire was released early in 1776, just a few months before the Declaration of Independence. Gibbon was an MP and had been present in the previous autumn, silently approving Parliament's decision to support the government's plan to send more troops against the rebellious colonists. By the time he had completed his mammoth work, Britain had lost the war. It was a serious setback, but proved temporary and the heyday of the empire still lay in the future. The new America was tiny in comparison to today, for the great expansion to the west coast had not yet occurred, and no one would have guessed at its future prominence, although some wild claims were made. America was to have a negligible role in the affairs of the wider world for the next century.'
In the nineteenth century it would become more and more common to compare the grandeur of Britain's empire with that of Rome. For Gibbon and contemporaries the parallel was less specific, but there were a number of reasons why he chose to look at Rome rather than any of the other great empires of the ancient world. The first was quite simply the impact of the Romans on the world, and most of all on the Western world. Their empire had been larger and lasted far longer than any of the other great nations of antiquity. As importantly, it had included Gibbon's homeland, as well as most of western Europe. Christianity emerged in the Roman period and eventually became the religion of the empire, hence a Catholic Church and pope in Rome. Gibbon had dabbled with Catholicism in his youth, before being sent away by his father for a properly Protestant re-education in Calvinist Switzerland. Yet the Catholic Church had ensured the survival of Latin - and helped to preserve Greek - as a language and made possible the rediscovery of Greek and Roman literature in the Renaissance. Men like Gibbon were comfortable in both languages, which remained in his day the central pillars of education. The Greek achievement was admired, but Athens' decline was already chronicled by Thucydides and Xenophon. Alexander's empire was vast, but failed to outlive him. The earlier empires of Persia, Assyria, Babylonia and Egypt were known largely through what the Greeks and the Bible said of them. It was still a generation before Champollion would decipher the Rosetta Stone and little was known for certain about the earliest civilisations.'
There was also a particular immediacy about Rome for Europe's age of Enlightenment. Only now was there general confidence that learning and culture had once again reached the standards of the classical world, and was even beginning to surpass them. Yet the Roman Empire had collapsed in the west some thirteen hundred years before Gibbon began to write, and even the remnant of the Eastern Empire had disappeared three centuries ago. Looking back, the Middle Ages seemed to present a bleak prospect of ignorance and superstition, in stark contrast to the sophistication and apparent rationality of the Greco-Roman world. This reaction is not uncommon, even today. One recent book examining the transition from ancient to medieval was subtitled `the Rise of Faith and the Fall of Reason'.4
For a long time the human race - particularly that part of it living in western Europe - had regressed rather than progressed, and understanding how and why this had happened was central to understanding the modern world. Yet, for all the reverence for the classics, little attention was paid to the world of the Late Roman Empire, primarily because all of the great Greek and Latin authors were earlier. In some ways Gibbon was treading fresh ground in looking at the fall rather than the rise and heyday of Rome. His concept was grand, original and sophisticated. Not only did he refer to ancient sources, but he also noted and assessed the theories of contemporary authors. Gibbon's breadth of scholarship remains exceptional, and in most respects the Decline and Fall can be seen as the first `modern' history of the ancient world written in English, although in fact academic styles would develop in a different way in the following years. It was also from the beginning recognised as one of the great works of English literature.'
The world has changed since the eighteenth century, as have attitudes to both the past and present. Yet the fascination with the fall of the Roman Empire remains. The link may now be less intimate and obvious, but the influence of Rome upon the modern world - and especially Western culture - remains profound. There is also simple curiosity as to how a state that was so successful and so massive for such a long time nevertheless crumbled - or was shattered - and was replaced by far less sophisticated cultures. Rome's fate seems to act as a warning that strength and success will always prove transitory in the end, and that civilisation will not automatically triumph. It was no coincidence that one of Winston Churchill's most famous speeches from 1940 foretold that Britain's defeat would result in a `new Dark Age' - particularly apt since many believed that the Roman Empire had been destroyed by German barbarians in the fifth century.
Each successive generation has returned to the mystery of why Rome fell, and a huge number of different theories have been put forward - not too long ago one German scholar catalogued no fewer than 200. Often parallels have been quite explicitly drawn with problems facing the historian's own country and time, but there is at least one striking contrast between the Roman experience and the demise of the great empires of the twentieth century. Powers like Britain and France were already in decline, exhausted by world wars and their economic consequences, but they also faced huge pressure for independence from their colonies. It is doubtful that either would have had the capacity and will to resist this pressure indefinitely, especially since it was encouraged by the two new superpowers. America had not fought the Second World War to preserve the British Empire and its trade system, while Soviet Russia actively supported Marxist revolutionaries seeking independence.'
There is no trace of a comparable desire for freedom from imperial rule in Rome's provinces. The population of the Spanish provinces did not long to become an independent Spanish state, nor were there movements for the liberation of Cappadocia or Greece. Quite simply there were no equivalents in the Roman period of Gandhi or Nehru, Washington or Bolivar, Kenyatta or Mugabe. Even the empire's Jewish population, which had rebelled on several occasions in the first and second centuries, no longer seem to have wanted their own state by the fourth. People wanted to be Roman and associated freedom with belonging to the empire and not independence from it. This is in spite of the fact that the empire's rulers were not elected and enjoyed, effectively, absolute power. In every case power in Rome's former provinces eventually - and sometimes immediately - passed to new foreign invaders. Strikingly, even these usually wanted to become part of the empire and to enjoy its wealth rather than destroy it. The great paradox of the Roman Empire's fall is that it did not end because people inside it - and, indeed, outside it - stopped believing in it or wanting it to exist.
The Romans wanted the empire to exist, and most could not imagine a world without it, but they did realise that it was facing great problems. Most were inclined to see moral decline as the root cause of these: the empire was struggling because people lacked the stern virtue of the earlier generations who had made Rome great. This was a traditional - and particularly Roman - way of thinking. There was also often a religious element. Pagans blamed everything on the Christians for neglecting the old gods who had guided and protected the empire. In turn, Christians blamed pagans for clinging to the old mistaken beliefs, while a few began to link the end of Rome with the end of the world. St Augustine wrote his monumental City of God to explain to Christians that in the end all human states, including Rome, the greatest of them all, would pass. Christians were all members of a new and perpetual state that God would create. This was not an encouragement to them to despair of the empire or to try to speed its demise, but to reassure them that a better world lay ahead. Some secular historians - mainly men writing in Greek in the eastern half of the empire long after the west had gone - criticised individual emperors for specific military or political decisions, which were claimed to have had long-reaching consequences. However, none of the surviving works from the ancient world attempted any coherent analysis of why the empire, which spanned the bulk of the known world in 200, was reduced to a small fraction of its power and territory by 500.
Gibbon, primarily a narrative historian, was too subtle to present one single cause for the empire's fall. An Englishman in a country where civil war still cast a long shadow - Culloden was fought just thirty years before the first volume of Decline and Fall was released - he drew attention to the frequency of internal strife within the empire and the willingness of Roman armies to fight against each other in support of rival candidates for the throne. With an Anglican's suspicion of the papacy, he saw the adoption of Christianity under Constantine and his successors as a bad thing, which sapped the old Roman virtue and eventually caused too many people to withdraw from public life into unproductive monastic seclusion. His attitude was all the more bitter because he had himself converted to Catholicism during his student days at Oxford. Gibbon's father had withdrawn his son from the university and despatched him for a thorough reprogramming in Calvinist Switzerland. On balance, reflecting both the mood of his sources and the culture of his own day, the sense of moral decline is a constant thread running through Gibbon's account. The Romans failed in the end because they no longer deserved to succeed. At one point, after listing the many problems faced by the empire, Gibbon suggested that we probably ought not to wonder why the empire fell, but marvel that it lasted so long.
In due course many other historians considered this question. Some saw the collapse as internal, the result of failures and decline within the empire. Others preferred to emphasise the attacks on the empire by the Huns, especially the Germanic tribes who forced their way through the frontiers and carved out kingdoms for themselves in the western provinces. In the emotive words of one French scholar, `The Roman Empire did not die. It was assassinated.' Stressing the role of the Germans had particular appeal in the climate of German nationalism in the nineteenth century. Roman texts contrasting the primitive virtue of German warriors with the decadence of fashionable life at Rome were taken at face value. For some, the empire deserved to die so that power could pass to the tribes that would make the countries of modern Europe. Others saw things in almost as blatantly racial terms and viewed the basic failure of Rome as a consequence of permitting too many barbarian Germans to enter its frontiers. The preoccupations of each age have usually been reflected in their views on Rome's fall. Social problems and class tension have sometimes become fashionable explanations, often in combination with economic factors. For some the world of the later Roman Empire was extremely bleak, with over-taxed peasantry being squeezed to pay for the spiralling costs of maintaining the army. In time the strain was too great and the whole system collapsed. Alternative theories would point to military failures or dwindling population. Others have reflected different modern concerns and suggested that environmental or climate change - perhaps increased by the impact of Roman farming and industry - were the root cause of declining agricultural yields and ultimately economic collapse.
In the last few decades the very nature of the debate has changed within the academic community, and there are several reasons for this. One, common in the West, has been a changed attitude towards empires in general, now that the modern ones have gone. These are no longer assumed by their nature to have been good things. Instead, the pendulum of popular - or at least middle-class and academic - opinion has swung to the other extreme. Instead of being forces for order and progress, bringing peace, education, science, medicine and Christianity to the wilder parts of the world, empires have become nothing more than brutal exploiters of indigenous populations. If empires are automatically a bad thing, then it is comforting also to think of them as inefficient. There has been much emphasis in recent studies of the Roman Empire in the first and second centuries to the lack of central control or planning, its unsophisticated economy, limited technology and simplistic thinking in such matters as geography and military strategy. Instead of the apparently sophisticated, the primitive has been stressed!
Curiously, attitudes towards the later Roman Empire have tended to the opposite extreme. For a long time it was unfashionable for academics to work on the later, rather than earlier periods. The main reason was the lack of good sources - particularly reliable and detailed narrative histories - for the third century, much of the fourth century and all of the fifth century. There is a considerable body of literature surviving from these periods, but it deals little with political or military events, much of it being religious - mostly, but not exclusively, Christian - philosophical or legal. Of little value for studying the great events of these years, it does provide considerable material for various aspects of social, cultural and intellectual history, which have become far more popular with academics in the last generation or so. This encouraged a massive boom in the study of the later empire. Many very important and revealing studies have been produced, and it is fair to say that we now know a good deal more about many different aspects of the period.'
Yet something odd has happened as well. In the beginning there was clearly a sense that the historians choosing to work on the later period needed to justify their decision. Many became deeply uncomfortable with the idea of an empire in decline, and emphasised the vibrancy and strength of the fourth- and even fifth-century Roman state. This was especially easy for those dealing with culture and religion. In these fields there was no catastrophic break coinciding with the collapse of the Western Empire. Reassessment of the centuries after Rome's fall has also been an especially fertile field for scholars in recent years, and these two trends have encouraged and fed off each other. Scholars, as opposed to the wider public, had long been unhappy with the term `Dark Age', and instead the fifth to tenth centuries are now universally referred to as the `Early Medieval' period. Medieval history is currently flourishing in universities, making this connection both attractive and instructive. At the same time it has been customary to stop talking about the `Late Roman Empire' or the `Late Roman period', and instead refer to `Late Antiquity, stressing the legitimacy, importance and also separateness of study into this period.
Names can be important, shaping the broad mental framework into which specific studies are fitted. In most respects these trends have been positive. Far more imaginative use has been made of the sources we do possess for these periods. Yet there are also inherent problems. Switching the focus to society, culture, religion and even to government and law tends to produce a rather static view, emphasising continuity rather than change. Events such as wars and revolutions, and the behaviour and decisions of specific emperors and ministers, do not necessarily register, but it would be most unwise to see them as unimportant. It seems very hard for many people working on Late Antiquity to consider the possibility that anything was declining. Instead they prefer to see change and transformation. In a gradual - and in no way traumatic - process the world of the Roman Empire morphed into the medieval world. For instance, one scholar who examined government in the Western Empire concluded that: `It should be clear ... that the Roman Empire did not "fall" in the fifth century, but was transformed into something new."°
The main basis for this conclusion was that some aspects of government, including specific titles and ranks, appear to continue under the Germanic kingdoms. The concept of decline firmly out of fashion, it was probably inevitable that the idea of a fall would also come under pressure. Even when it is admitted that this occurred, it is often portrayed as a matter of little importance. The trend has been for those working on Late Antiquity to be almost relentlessly positive in their assessment of every aspect of it. Institutions such as the army and government are portrayed as very efficient - often more effective than those of the early empire - and any problems seen as inevitable in the conditions of the ancient world and not unique to the later period. Similarly, the slightest trace of continuity is imbued with deep and widespread significance. As an example, the survival of a Roman bureaucratic title in the court of a German king does not necessarily mean that the individual was doing the same job at all, let alone that he was doing it well. Similarly, the find of a stylus pen in a late fifth-century site from Britain cannot be taken to prove widespread literacy in the post-Roman period. Extending the same logic to our own day would mean that the survival of imperial institutions and English as one of the languages of government in India really meant that it was still part of the British Empire. This would doubtless come as a great surprise to the country's inhabitants.
There have been some dissenting voices. Recently two popular books were released in which distinguished specialists on Late Antiquity - curiously enough, both from Oxford - cast doubt on what has become the orthodox view. Brian Ward-Perkins' The Fall of Rome (2005) pointed out firstly that the idea of a peaceful transformation between Roman Empire and barbarian kingdom simply goes against the evidence as well as simple logic. Even more importantly, he used the archaeological record to show just how massive a change there was as a result of Rome's fall. Much of this had to do with the everyday life of ordinary people, who, for instance, now lived in houses with thatched rather than tiled roofs, and used simpler, locally produced pottery, rather than a range of finer imported wares. Cultural sophistication declined so sharply that WardPerkins felt justified in calling it `the end of civilisation'. Peter Heather's The Fall of the Roman Empire (2005) was more concerned with how the Western Empire fell than its consequences. He employs an essentially narrative structure, feeling that the peaceful transition theory of the empire's end `has largely established itself ... only because detailed historical narrative has been ignored for half a generation'. Beginning in 376 he charts the century until the deposition of Romulus Augustulus and, like Ward-Perkins, he sees the `end of the Empire as a major event'. The fourth-century empire is presented as a strong, vibrant state, whose demise was not inevitable. Instead, the new threats posed by peoples from outside, such as the Huns and Goths, presented a challenge that through a mixture of human error and chance was not adequately met."
Each of these books is extremely good in its own way, but both are restricted in what it was possible to cover. Neither makes much effort to link the empire of the fourth century with the earlier empire. Yet this connection needs to be made if we are to understand more fully what the Roman Empire was like and discern why it did eventually fall. Studies of Late Antiquity stress the great strength of the fourth-century empire. They are certainly correct to do so, since Rome in this period was overwhelmingly stronger than any other nation or people in the known world. However, it was not as stable as the empire of the second century, nor was it as powerful. How and why this changed is central to understanding why the later empire was as it was. Put simply, the empire was stronger in the year 200 than it was in 300 - although perhaps it had been even weaker in 250. By 400 the empire was weaker again, and by 500 it had vanished in the west and only the rump was left in the lands around the eastern Mediterranean. A longer perspective is necessary to explain these shifts.
With the concept of decline out of fashion, most historians have tended to stress the pressure from outside the empire. Only recently have some questioned the true scale of the threat posed by the tribal peoples who lived outside the Roman Empire's European frontiers. Even so, many continue to assume that the confederations that appeared by the end of the third century were far more formidable enemies than the barbarian tribes faced by the early empire. It certainly remains an article of faith that the Sassanid Persians who supplanted the Parthian dynasty in the early third century were far more efficient, aggressive and dangerous than their predecessors. This has certainly been repeated so often that no one seems to question its essential truth. The belief that the threats faced by the empire had increased is a convenient one for those wishing to see the massive institutional changes within the empire as sensible reactions to a new situation. Convenience and frequent repetition do not amount to truth, and all of this needs to be questioned."
Civil war was a frequent occurrence from the third century onwards. After 217 there were only a handful of decades without a violent struggle for power within the Roman Empire. Some of these were local rebellions, rapidly suppressed and involving little serious fighting. Others lasted for years and were only decided by one or more major battles or sieges. We have no figures for how many Roman soldiers died or were maimed fighting against other Romans, but the total must have been considerable. It is true that people living in provinces distant from the fighting may not have been directly affected by outbreaks of internal conflict, unless they were related to leading figures on the losing side. This does not mean that such things were of minor importance. Civil war was a fact of life, and everyone who reached adulthood would have lived through one, even if it had no direct impact on them.
Strangely, while most historians note the frequency of internal conflict within the Roman Empire from the third century onwards, they rarely spend much time considering this in any detail. A. H. M. Jones produced a colossal study of the later empire that remains an indispensable reference point even now, more than forty years later. It includes the following curious statement: `Diocletian maintained internal peace for twenty years, broken only by two revolts."3 At this point it is worth noting that one of these revolts lasted for the best part of a decade and both required a major military effort to suppress them. Diocletian had anyway fought and won another civil war to secure himself as emperor in the first place. He was certainly successful by the standards of recent decades, but the stability he gave the empire was limited and brief. His reign was followed by a spate of especially large-scale civil wars. It is significant that Jones devoted only a single paragraph to civil war and internal strife in a long chapter discussing the causes of Rome's fall. His attitude was and is typical, civil wars and usurpations simply being accepted as part of the normal landscape of the later Roman period. One of the reasons for this neglect may simply be that most scholars have worked in countries for whom civil wars were things only of the distant past. It was simply natural for them to assume that foreign threats must always be more serious. In addition, the focus on institutions and culture had little room for civil wars, which rarely if ever involved major changes to such things. Rarely does anyone pause to consider the consequences of this reality for the attitudes of emperors and their subordinates at all levels.
The aim of this study is to look more closely at both the internal and external problems faced by the Roman Empire. It will begin, as Gibbon did, in the year i8o when the empire still appeared to be in its heyday, before moving on to trace the descent into the chaos of the middle of the third century. Then we will examine the rebuilt empire of Diocletian and Constantine, the move towards division into an eastern and western half in the fourth century and the collapse of the west in the fifth. It will end with the abortive effort of the Eastern Empire to recapture the lost territories in the sixth century. Gibbon went much further, continuing to the fall of Constantinople to the Turks in the fifteenth century. That is a fascinating story in its own right, but is too great a one to be dealt with adequately here. By the end of the sixth century the world was profoundly and permanently different from our starting point. The Eastern Roman Empire was strong, but no longer possessed the overwhelming might and dominance of the united Roman Empire. This book is about how this came about. Central is the story of the individual men and women, the groups, peoples and tribes who lived through and shaped these centuries. In telling the story, we will try to assess the more likely theories about why things happened as they did.
We have some important advantages over Gibbon when it comes to considering this theme. Antiquarians had made some effort to collect and catalogue inscriptions from the ancient world, and to describe the visible remains of ancient towns and cities. However, archaeology in any systematic form did not begin until the nineteenth century, and techniques of gathering and understanding data have since become far more refined. New sites are continually being discovered and existing ones better understood, adding to the pool of information about each region and period. Modern methods are very sophisticated and able in the right circumstances to extract a good deal of information. This does mean that the modern trend is to excavate increasingly small areas in greater and greater detail. Given the size of many communities from the Roman period, it is now quite rare for settlements to be excavated in their entirety. Similarly, there are normally only the resources for largescale work on a small proportion of located sites. This can mean that a general picture of rural or urban life in a province tends to be based on a tiny sample of existing remains, even ignoring what has been lost or the sites not yet located. It is also vital to appreciate the limited amount of unequivocal fact discovered by archaeology. All finds require interpretation, especially if wider conclusions are to be drawn. Any study of the history of the ancient world is incomplete without considering the archaeological record, but impressions derived from the latter are liable to change as new discoveries are made or old ones reinterpreted.
The vast bulk of the literature surviving from the Greco-Roman world was available to Gibbon. There have since been a few discoveries - for instance, the letters of Fronto from the very beginning of our period. Conversely, the poems of Ossian - purportedly heroic poetry surviving in Scotland from the Caledonian tribes who had fought against Rome and mentioned in The Decline and Fall- have long since been recognised as an eighteenth-century hoax. However, the genuine finds of texts and fragments from other writers have not fundamentally changed the balance and usefulness of the literary sources. The third century is extremely poorly served. For much of it there are only summaries and epitomes of earlier histories, which are generally brief and often unreliable. There is also the collection of imperial biographies known as the Historia Augusta, which purports to be the work of six authors writing in the late third and early fourth centuries. It is now generally considered to have been written by just one man at least a generation after this. An odd mixture of invention and confusion, the author nevertheless seems to include some reliable information. Yet it is an indication of the poverty of our other sources for this period that we are forced to make any use of it at all.14
Two notable narrative historians provide detailed - and generally reliable - accounts. Ammianus Marcellinus covers part of the fourth century and Procopius part of the sixth. Both were actual eyewitnesses to some of the events they describe. The same was true to some extent of Cassius Dio and Herodian, who cover the beginning of the period. Apart from these, we rely mainly on snippets of information and brief summaries. As we have seen, the overwhelming bulk of literature from this period is simply not concerned with the great events of politics or war. Some, such as the many panegyric speeches, do address emperors and refer to contemporary concerns and events, but in such a stylised and rhetorical form that it is difficult to glean very much information from them. The belief that these contain coded messages is possible, but easily taken too far. It is vital to remember that we have only the tiniest fragment of the literature that once existed. A large chunk of Ammianus' history is lost, while only the names survive of many other authors and their works. There were doubtless far more who do not even get mentioned in what survives. Most works were preserved in manuscripts kept in church libraries. Inevitably, this meant that the prospects for Christian manuscripts were far better, and also that literary merit rather than historical interest played a part. Chance played an even bigger role.
This is even more true of the other documents - mostly written on papyrus, but sometimes on writing-tablets or pottery sherds, preservation has also largely been a question of luck. These continue to be found where the conditions are right, and sometimes appear in considerable quantities and can include such things as census returns. Such information is highly useful, but never exists in sufficient quantity to generate reliable statistics for population size, age range and the general levels of prosperity on more than a very local and short-term basis. All studies of the ancient world are forced to proceed without the support of statistics. This does mean that it is impossible to prove or disprove some of the theories put forward to explain the fall of the Roman Empire. We simply cannot say whether a serious decline in population played a role in this. Similarly, we cannot measure the state of the economy at any set period or trace the real impact of the staggering devaluation of the currency in the third century. What sources we have may hint at trends, but not everyone will interpret these in the same way.
There is a good deal that we simply cannot now know about the history of the Roman Empire in the third and later centuries. To a greater or lesser extent this is true of most periods of ancient history. Yet we must be careful to ask the questions we want to ask, rather than shifting towards those that the sources make it easiest to answer. In addition, the simple fact that so much Greek and Roman literature has failed to survive does rather suggest that the change from a Roman to medieval world was in many ways drastic. Far more of this literature was simply lost rather than deliberately suppressed or destroyed by churchmen. The medieval world was a far less literate place than the classical world that preceded it, particularly in western Europe. None of this suggests transformation. The fall of the Roman Empire was a major event, even if it occurred over considerable time and cannot be assigned to a specific date. This becomes all the more clear when we consider the empire when it was still at its height.