Part Three
Now in general the Romans rely upon force in all their undertakings.
—Polybius (trans. Ian Scott-Kilvert)
Chapter Seven
Early Roman Practices of War
Early in the second century B.C., the eastern Greeks felt the full weight of Roman expansion, beginning the long exchange between the two peoples that eventually produced the dual culture of classical antiquity. In warfare as in much else, the Greek element in this amalgam was the more original, but the Roman contribution was the more decisive influence on later Western civilizations.
Roman warfare was an adaptation of Greek hoplite warfare and the hoplite ideology of decisive battle, but with peculiar features, the most striking of which was its sheer success. This was especially striking to the defeated and humbled Greeks of the second century. At the beginning of his histories, Polybius posed the question that was troubling his compatriots and has never ceased to fascinate the world:
There can surely be nobody so petty or so apathetic in his outlook that he has no desire to discover by what means and under what system of government the Romans succeeded in less than fifty-three years [220-167 B.C.] in bringing under their rule almost the whole of the inhabited world, an achievement which is without parallel in human history. Or from the opposite point of view, can there be anyone so completely absorbed in other subjects of contemplation or study that he could find any task more important than to acquire this knowledge? (1.1 [trans. Scott-Kilvert])
Polybius could not produce an entirely satisfactory answer to this question, but he saw one important key to it. He realized that the rise of Rome owed little to Fortune, the fickle goddess of the Greeks. The triumph of Rome was not the triumph of a meteoric individual like Alexander but the triumph of a system; and the Roman system had come to stay. The Roman republic was a society superbly organized for war. Its capacity for sustained, long-distance, aggressive war making had no earlier parallel and was to have none again until the rise of the modern European nation-states.
The capacity was obvious to the Greeks, but the explanation for it was not, and it has eluded many modern scholars as well. The Romans never interpreted themselves as the Greeks did, and such interpretations as we find in Latin authors were written at a time when the old Roman military system scarcely existed any longer and the Roman aristocracy had acquired a veneer of Greek philosophical culture that made it difficult for the members of this elite to honestly confront their ancestors. Hence, the Roman legacy is largely a legacy of myths and mirages.
The military system itself is well known to us and needs only the briefest description here.1 The study of Roman military institutions in the age of Roman expansion, before the “Marian” reorganization of the first century B.C., raises no problems comparable to those encountered in the study of archaic Greek warfare. The Roman legion was essentially an adaptation of the Greek phalanx, which the Romans broke up into several lines, with each line in turn broken up into small units capable of independent maneuver. Most soldiers were armed with swords and javelins, though the rear line retained the Greek thrusting spears. The Romans sacrificed the depth and cohesion of the phalanx for mobility, sending in their units in waves to attack and retreat in turn, in a fashion that in Greek warfare was associated more with cavalry than with infantry. The system could function smoothly in the heat of battle because the legionaries were subjected to intense drilling and were led by a semiprofessional officer corps, the centurions; neither institution had any parallel in the Greek world outside Sparta.
The legion was obviously a more flexible formation than the phalanx, but in some ways it was less tactically effective. The Roman insistence upon uniformity caused them to neglect cavalry and light infantry, and they therefore never developed the combined-arms tactics perfected by Alexander and his successors. They paid for this when they met a general as brilliant at those tactics as Pyrrhus or Hannibal. But again, the most important feature of the Roman military system was simply the fact that it was a system. Rome did not need brilliant generals and rarely produced them. Romans knew that in the long run nothing could defeat their well-drilled military machine (Roman warfare always evokes the metaphor of a machine, and traditional Greek warfare, that of a duel). Romans knew that all wars with Rome would have a long run because Rome never gave up. Only a society that regarded almost constant warfare and its attendant discipline as normal expectations could have operated such a system. The key to it lies not in the military institutions themselves but in the militaristic culture behind them.
Causes of Early Roman Warfare
When the Greeks thought about societies organized for war, they thought first of Sparta. But Sparta was a hoplite oligarchy whose war aims were essentially defensive, even isolationist at times, and the egalitarianism of the hoplite class, which Sparta carried to communistic extremes, kept its members in line. The contrast between the defensive militarism of Sparta and the expansionary militarism of Rome was perceived by Polybius, but he was too enmeshed in the constitutional theories of Greek philosophy to see the reason for it. And modern scholars have often been misled by Roman historiography, which leaves the impression that none of Rome’s wars were offensive.
Until recently, most historians have followed the interpretation of Roman imperialism laid down in the late nineteenth century by the great scholar Theodor Mommsen. Most have denied that Rome had a conscious policy of imperialism. Some have spoken of a “defensive” imperialism. Cicero said that Rome conquered the world merely by defending its allies. Others have emphasized the element of accident in the Roman conquests, often with reference to the example of the British Empire, which, as the saying went, was acquired in a fit of absence of mind. But if “imperialism” means a policy of expansion—the usual meaning of the word since the late nineteenth century—then “defensive imperialism” seems oxymoronic. Those who think the Roman republic did not consistently seek conquest have never provided a coherent explanation of how it managed nevertheless to conquer so much. The parallel with nineteenth-century Britain should remind us that empires are not really acquired absent-mindedly; there is a wide area between grand strategy and absence of mind. This is another case where we tend to confuse the ancient concept of a just war and the modern concept of defensive war. Just wars could be aggressive, and the just wars of Rome were particularly so.
Recent studies have suggested a new approach to the problem.2 In brief, the Roman public ideology can be described as a highly elitist form of civic militarism. Rome was dominated by a fiercely competitive warrior oligarchy, which at the same time shared the benefits of conquest with the masses, perhaps more fully than any other conquest state had ever done; the dynamism of the system sprang from the interaction between elitist and civic elements.
Unlike most ancient Mediterranean oligarchies, membership in the Roman ruling class was not guaranteed by birth or wealth. This was an oligarchy of officeholders. Admission to the Senate, the ruling council of Rome, was achieved only through election to magistracies, and the functions of magistracies were essentially military. In the early republic, a young noble had to spend years in military service before he was even eligible for office. Admission to the inner circle of the Senate came through election to the consulship, the supreme magistracy, whose supreme function was leadership in war. There were only two consuls, elected annually, and since in the early republic the nobility insisted on sharing office, election to the consulate was usually a once-in-a-lifetime opportunity. Therefore, a noble who won the consulate had a single year to demonstrate his virtus and win laus and gloria for himself and his house. The original meaning of “virtue” was military valor, and “praise” and “glory” originally meant military glory. The highest achievement of a consul was to be awarded a “triumph,” a victory procession through Rome displaying his captives and booty, which could not be awarded unless the consul had killed at least five thousand of the enemy. Much of this seems atavistic, reminding us of the shamanistic killing quotas and head-hunting raids of tribal warfare. The roots of it were primitive, indeed, but here they had been adapted to a highly evolved political system whose net effect was to place enormous pressure on the male members of its elite to compete with one another for military success. Greek cities did not train such elites. The supreme Greek symbol of military glory was the communal dedication of captured arms and armor in a temple, not the general’s triumph. In the Athenian funeral orations no individuals were ever mentioned by name. The Greeks suppressed the element of individual glory in the martial-values complex as systematically as the Romans encouraged it.
This was a very elitist form of militarism, but it never lost its civic foundation. Not only the leadership but the whole society was geared for expansion. Romans expected war to be profitable as well as glorious, profitable for the noble above all, but also offering substantial material rewards to the entire citizen body—in the form of plunder, slaves, distributions of land in the new Roman colonies that soon dotted Italy, and opportunities for social advancement through a centurion’s career. Until the late second century B.C., when the great age of expansion was over, there was little sign that any citizens objected to the constant campaigning that distinguished Rome from all other ancient city-states. The Roman republic went to war almost yearly, and even before the war with Hannibal, it normally had four legions under arms each year, constituting about one-fifth of the eligible citizens. It has been said that this was the highest rate of military participation known in any preindustrial society except those of Prussia under Frederick the Great and Napoleonic France, which matched it only for short periods.3
Warfare in Early Roman Religion
Connections between war and religion, strong in all ancient societies, were nowhere stronger than at Rome; and Roman religion, in warfare as in all else, was characterized by an intense legalism. Rome and the other Latin-speaking cities of central Italy had special colleges of priests called fetiales, whose sole function was to preside over interstate relations.4 A Latin city could not go to war until the fetials had proclaimed the cause to be just, calling upon Jupiter and all the gods, and had demanded reparations from the offending city. No other ancient cities ever seem to have had a special priesthood for war and diplomacy. The rites included elements suggestive of Neolithic antiquity—a pig had to be sacrificed with a stone knife to solemnize a treaty, a spear with a fire-hardened point had to be hurled into enemy territory to declare war—but in spite of these archaic trappings, the likeliest explanation for this strange institution is that it was an artificial creation of the Latin peoples intended to prevent or at least regularize warfare among themselves. The full ceremony required both cities to have fetials, and since fetials were unknown outside central Italy, the custom was gradually dropped as Roman expansion got underway The functions of the fetials were taken over by secular ambassadors such as other cities employed, and only vestiges of the fetial law survived to the late first century B.C., when Augustus Caesar attempted to revive it along with other antique religious customs.
Nevertheless, the fetial law exercised a permanent effect upon the Roman mind. In later centuries, it was remembered that only the Latins had had such an institution, and Roman success in war was widely attributed to the fact that the Romans had always taken such care to ensure that all their wars were pious and righteous. The fetial code included jus in hello as well as jus ad bellum, for the fetials presided over all treaties and oaths with foreign cities and protected ambassadors. All these matters constituted the fides Romana, or the good faith of Rome. The basic concept was not new: Rituals to ensure the justice of war and regulate the conduct of war were universal in ancient societies, and Greeks were also proud of the good faith (pistis) of their city in its dealings with outsiders. But the elaborate legalistic form that these rites acquired in early Rome would make it difficult for the later Romans to separate religion from policy.
In addition to the fetial ceremonies at the start of a war, all campaigns and battles were accompanied by the usual sacrifices, prayers, and auguries. There was an unusually large pantheon of war gods. The principal god of war, Mars, was a far more majestic figure than his Greek counterpart, Ares. Many other deities had military functions. Fides, the good faith of Rome, was herself an ancient goddess associated with Jupiter, and both were responsible for oaths and treaties. In the third century B.C., the pivotal century of Roman expansion, a uniquely Roman cult of the goddess Victoria (Victory) was introduced. The Roman religious calendar was studded with military festivals, from the exercising of the cavalry horses at the Equirria in March to the purification of arms at the Armilustrium in October, which marked the end of the campaigning season.
One aspect of the state cult is of special interest here. William Harris has presented considerable evidence that Rome, alone among ancient city-states, made expansion a public and religiously sanctioned aim. In the reign of Tiberius, a rhetorician named Valerius Maximus compiled an anthology of historical anecdotes for the use of orators that contains some interesting bits of historical information, among them the fact that the duties of the important magistrates called censors had in early times included the recital of a prayer calling upon the gods to make the Roman state prosper and grow: “Quo di immortales ut populi Romani res meliores amplioresque facerent rogabantur” (Valerius Maximus 4.1.10). A. N. Sherwin-White has pointed out (in “Rome the Aggressor?55) that the censors were responsible for taking the census and therefore were concerned with population and fertility, so the prayer just cited, “May the immortal gods make the things of the Roman people better and bigger,” does not necessarily refer to territorial expansion. We have observed that Greeks could use similar expressions, with similar ambiguities. But to my knowledge, Greeks did not say such things in such an official religious forum, and when Greeks and Romans prayed for the expansion of the state, I doubt that they meant to exclude the possibility of territorial expansion. In the case of the Romans, there is reason to think that they had that possibility explicitly in mind. The historical traditions relayed by Livy recalled that in the eastern wars of the second century B.C., the haruspices, the official soothsayers who read the future in the flight of birds and the entrails of sheep, repeatedly prophesied that these wars would extend the frontiers of the Roman people (Livy 31.5, 36.1, 42.30), and there seems to be no question here but that territorial expansion was meant.
There was also a tradition that Rome’s victorious wars always ended with a deditio or unconditional surrender. According to Livy, this practice went back to the period of the kings. King Tarquin in the seventh century B.C. was supposed to have addressed the following formula to the spokesmen of a conquered Sabine city: “Do you surrender yourselves and the People of Collatia, city, lands, water, boundary marks, shrines, utensils, all appurtenances, divine and human, into my power and that of the Roman People?” (Livy 1.38 [trans. B. O. Foster]). Some traditions make a distinction between surrender into the power (potestas or dicio) of Rome and surrender into the good faith (fides) of Rome, implying that a submission to the good faith of Rome guaranteed mild treatment, whereas surrender into the power of Rome put one into a more uncertain status (Livy 36.27, 39.54; Valerius Maximus 6.5.1); but we do not know whether this distinction was more than rhetorical. In any case, it was believed that the Romans had always fought for unconditional surrender. And that was rare in the Greek world until a very late date.
It is possible that all these traditions had been strongly colored by the attitudes of the post-Polybian Roman elite, which took it for granted that the gods had conferred world empire upon Rome and read this worldview back into earlier times. But at least this is how the Romans of the late republic interpreted the mos maiorum, the way of their ancestors: They believed the expansion of Rome had been sanctioned by the gods from the beginning, that Roman wars had always been fought with a dedication to total victory highly unusual among ancient city-states, and that Rome had always had a sense of moral and religious responsibility for those who accepted Rome’s leadership.
The Roman Conquests
The Roman system developed in three stages: the conquest of Italy, circa 400-270 B.C.; the conquest of the western Mediterranean, circa 270-200 B.C.; and the conquest of the Greek world, circa 200-146 B.C.
The first phase was formative. It was during the fourth century B.C., in the course of continual warfare with its neighbors, that Rome, originally not very different from other city-states of central Italy, developed its peculiar military culture. Once the pattern was established, it fed on itself. As soon as a large part of the Italian peninsula had been brought into the Roman alliance, the drive to expand became irreversible because the Roman alliance, like everything else about Rome, was geared for war. The allies of Rome rendered to Rome only military service, not tribute. The contrast with the Athenian empire of the fifth century is striking: Athens had preferred to take tribute, not military service, from its allies. The fact that Rome could demand only military service provided an additional incentive to war, for the only way Rome could profit from its alliances was to make use of them in war, and if a year went by without a successful war, the resources of the hegemony were being wasted. By around 300 B.C., the Romans had brought almost the whole of the peninsula into a Roman confederation with the largest manpower reserves in the western Mediterranean, and the habits of more or less constant warfare had become ingrained.
The second stage, much better documented, brought the two great Punic Wars against Carthage (264-241 and 218-202 B.C.) and the Roman conquest of the western Mediterranean coasts and islands. There has been much dispute over the causes of the First Punic War in 264, but in the long view, it does not seem to matter much what made the Romans cross the straits into Sicily. That venture represented the first departure from traditional Roman policy, which had never looked beyond the Italian mainland, and it may have been a simple miscalculation. The important fact, however, was that once the Romans found themselves in Sicily, they stayed there. Rome managed to convert itself into a naval power, withstood appalling losses, and fought for Sicily for more than twenty years until Carthage conceded. The real secret of Rome’s success was the Roman willingness to persist in warfare year after year. When the war was over, Rome absorbed the Carthaginian thalassocracy and became a Mediterranean power. The second round, the Hannibalic War of 218-202 B.C., the most titanic conflict ever seen in the west, did nothing more than confirm this conclusion and provide an even more impressive demonstration of the invincible tenacity of the Roman war machine.
By 202 B.C., Rome may have been ready to stop. The Senate seems to have been genuinely reluctant to enter the alien and complicated Greek-speaking world to the east. It is doubtful that any Romans at that time had any ambition for, or concept of, world empire. All its traditions rooted Rome in Italy; to hold down a fringe of coastland and island in the western seas did not detract from the Italocentric nature of Roman policy, and these territories provided consuls with opportunities for easy triumph hunting among ill-armed barbarians; but the Hellenistic world was another matter. Also, the Roman capacity to expand may have temporarily outrun the capacity to organize and exploit the conquests. For half a century, Roman policy toward the Greeks alternated between sudden destructive intrusions and long periods of withdrawal. But the mechanisms of expansion in Roman society were still running and would not allow Rome to withdraw completely. During these intrusions, the Roman military system, hardened in the war against Hannibal, won decisive victories over the Greek kingdoms at Cynoscephalae (197) and Magnesia (189); a later intrusion destroyed the Macedonian monarchy at Pydna (168) and left Rome with no rivals. Thereafter, Rome was hegemon of the known world, and the Greek states, by expecting Rome to act like a hegemonic power, drew the Romans ever deeper into their affairs. By the middle of the second century, Romans were becoming accustomed to the idea of empire in the east and felt no more inhibitions about annexing territory there. The process was completed by 146 B.C., when both Carthage and Corinth were destroyed. All the Mediterranean Basin was now within the imperium of Rome, some of it organized into provinces governed by Roman magistrates, the rest reduced to client states.5
By this time, Romans were acquiring a sense that they possessed a world empire. (The phrase imperium orbis terrae first occurs circa 85 B.C. in the oratorical treatise called the Rhetoric to Herrenius.) They began to produce a Latin literature based on Greek models that adapted Greek ideas of war and conquest and sought to explain the Roman empire in those terms.
The emergent Latin literary tradition was also decisively influenced by the fact that it took shape at a time when the great period of Roman expansion had ceased. The machinery of expansion was still running, but now its energies were largely directed inward. The last century of the republic brought a series of devastating civil struggles among the great warlords of the Senate. Expansion continued intermittently, but the direction and nature of it had changed: Instead of being led by the Senate as a whole, expansion was directed by the warlords themselves, who acquired new provinces, like Caesar in Gaul, to strengthen their positions in the civil wars; instead of by the citizens as a whole, wars were fought by increasingly professionalized armies. Hence, the Latin tradition became permeated by a sense of decline and nostalgia for an earlier republic of domestic tranquility and glorious foreign conquest.
The Roman Frontiers
The dependence on Greek models and the ideology of nostalgia for the republican past caused the Latin literary tradition to become generally divorced from current political and military realities, in comparison to Greek literature. The principate never produced a realistic interpretation of itself and bequeathed no theories to posterity. In the eyes of posterity, the Latin writers who mattered most were those of the late republic and the Augustan Age, whose values were republican. Nevertheless the historical fact of the principate had an enormous effect on posterity. It was the major historical example of a great continental state faced with the problem of protecting a long land frontier, and we have ample evidence, mostly of a nonliterary nature, for the evolution of its frontier system over a period of several centuries. The subject deserves attention here because many modern scholars have thought that there must have been more systematic planning and thinking behind this system than appears in the literary record.
Three main stages in this evolution can be distinguished. The first stage was the hegemonial empire established in the late republic. The core Roman territories were surrounded with a cordon of client states, the friends and allies (socii et amici) of Rome. This system remained largely intact well into the first century A.D., but it generally ceased to grow after the Augustan Age, for reasons both internal and external. Although the ideology of expansion continued, the social engine that drove it practically stopped running under the principate, which put a halt to competition for office among the elite. In addition, the Roman war machine, which relied on heavy infantry and siege tactics, was best suited for high-intensity warfare against a dense agricultural population with fixed and vulnerable assets; it was less well suited for mobile warfare against cavalry or light infantry. The further the legions marched from the shores of the Mediterranean, the slower and harder conquest became. The Romans of the principate made repeated attempts to take over northern Europe and the Middle East, but the Germans had no cities to sack and the Parthians had few. The Romans of the republic might have persisted and taken both, but the energies had gone out of the system. Eventually the principate gave up.
By the later part of the first century, the second stage was emerging. The Romans gradually assumed direct control over their client states, and when they were all absorbed, a territorial empire took shape. The legions were now settled in permanent camps behind fortified frontiers, and behind these frontiers all the native elites were gradually incorporated into a single ruling class, united by a uniform Greco-Roman literary culture resembling the mandarin elite of China. The transition to “perimeter defense” (Edward Luttwak’s phrase) was virtually completed by the reign of Hadrian (A.D. 117-138).
But perimeter defense is effective only against weak external enemies, and the enemies of the empire grew steadily stronger. After the defenses collapsed during the crisis of the third century, a third stage emerged: The frontier forces were weakened in favor of mobile central armies. The transition to this mixed security system was complete by the reign of Constantine (A.D. 308-337).
In the 1970s, when William Harris was offering a new approach to the military history of the Roman republic, the military analyst Edward Luttwak made a similar impact upon the study of the principate by applying the concepts of contemporary strategic thought.6 Luttwak’s analysis is illuminating and the sketch given above is indebted to it, but the use of modern strategic language implies a coherent system with an inner logic and the existence of conscious long-range planning such as we expect from the general staff of a modern army. In fact, the evidence for the Roman security system is mostly archaeological, and the existence of deliberate planning behind it is generally a matter of inference. We know there were debates among the elite as to whether the empire should expand here or there, and these have left traces in Roman historiography. But it not obvious whether there was anything that should be called a grand strategy. This question will be taken up in Chapter 9, in dealing with raison d’état among the Romans. But first we must deal with Roman traditions about the morality of warfare.
Notes
1. For an introduction to the Roman army, see the chapters by G. R. Watson, A. S. Anderson, and R.S.O. Tomlin in The Roman World, ed. John Wacher (London, 1987), vol. 1, 75-135. For the army of the early and middle republic, see L.J.F. Keppie, The Making of the Roman Army (Totowa, N.J., 1984). F. E. Adcock’s The Roman Art of War Under the Republic (Cambridge, 1960) is still useful.
2. In the late 1970s, a number of important monographs changed the terms of this debate: Keith Hopkins, Conquerors and Slaves (Cambridge, 1978); C. Nicolet, The World of the Citizen in Republican Rome, trans. P. S. Falla (Berkeley, 1980); and above all, W. V. Harris, War and Imperialism in Republican Rome, 327-70 B. C. (Oxford, 1979). I share the view of J. A. North (“The Development of Roman Imperialism,” Journal of Roman Studies 71 (1981), 1-9) that Harris’s reinterpretation has rendered the theory of “defensive imperialism” untenable, at least in its traditional form. In brief, Harris has argued that republican Rome was persistently aggressive because the ethos of the whole culture was geared to war making, particularly the senatorial elite, and that Rome was unusual among ancient city-states in making expansion a publicly declared aim. Whether Rome had a conscious long-range strategy is a question Harris finds meaningless, because ancient states did not have such strategies. But he does think that Rome had a “continuing drive to expand” (Harris, 107). One weakness in Harris’s argument is that he never fully explains what he means by a “continuing drive.” In practice, he seems to have been thinking of a series of conscious decisions by the elite, for much of his book is taken up by an attempt to prove that virtually all the wars of the Roman republic during the period he studied had aggressive aims. One of his critics, A. N. Sherwin-White, has argued convincingly against this view in “Rome the Aggressor?” Journal of Roman Studies 70 (1980), 177-181, and Roman Foreign Policy in the East, 168 B.C.to Â.D. 1 (Norman, Okla., 1983). It seems to me that Harris’s argument as originally formulated suffers from the Clausewitzian bias of modern military history, which assumes all warfare to be a rational political activity. I suggest the Harris thesis will be strengthened if we adopt a more anthropological perspective: Warfare is everywhere a matter of continuing drives, which are expressions of culture and values more than of politics and policy, and this is especially true of a traditional society. To show that Rome had a continuing drive to expand it is not necessary to prove that most of its leaders had a conscious policy of that kind most of the time, nor need one deny that Rome sometimes acted defensively—as in the long and bloody wars fought in the third century B.C. to defend Italy from Greek and Carthaginian invaders.
3. The estimate of Hopkins, Conquerors and Slaves, 25ff., 102ff. He properly does not count primitive societies, which may have higher rates of military participation than any complex society but which are hardly comparable.
4. The fetial law is described by Livy 1.24, 32; Dionysius of Halicarnassus, Roman Antiquities 2.72; Plutarch, Life of Numa Pompilius 12. See Yvon Garlan, War in the Ancient World: A Social History, trans. Janet Lloyd (Ithaca, 1975), and M. D. Goodman and A. J. Holladay, “Religious Scruples in Ancient Warfare,” Classical Quarterly n.s. 36 (1986), 151-171. The explanation for the origins of the fetial cult that I follow here was suggested by Alan Watson, International Law in Ancient Rome: War and Religion (Baltimore, 1993). I have not followed Watson’s suggestion that early Rome was unique in regarding warfare as a trial before the gods; it seems to me that attitude is very general in primitive and ancient religion.
5. The political and cultural interactions between Romans and Greeks at this period are discussed in detail by Peter Gruen, The Hellenistic World and the Coming of Rome (Berkeley, 1984).
6. The Grand Strategy of the Roman Empire from the First Century A.D. to the Third (Baltimore, 1976). Similar approaches have been adopted by G.B.D. Jones, “Concept and Development in Roman Frontiers,” Bulletin of the John Rylands Library 61 (1978), 115-144; Arther Ferrill, The Fall of the Roman Empire: The Military Explanation (London, 1986).