Strategies or fantasies?
Grand but unfulfilled plans assigned to various leaders and peoples by ancient sources offer some of our most interesting information on ancient strategy. Looking at a selection of these stories, which are usually ignored or dismissed out-of-hand by modern scholars, tells us a great deal about how the classical cultures saw the world in military terms.
In 415 bc, during the Peloponnesian War, the Athenians sent an expedition to Sicily. The biographer Plutarch (ad c. 50-c. 120) claimed that for the Athenian politician Alcibiades Sicily was just the start of a campaign of conquest that would encompass Carthage, Libya, Italy, and then the Peloponnese. In Plutarch’s account this plan caught the imagination of Athenians young and old: people sat in the wrestling schools and other public places sketching in the sand the outline of Sicily and the positions of Carthage and Libya (Alcibiades 17). The contemporary historian Thucydides gave a different order to the projected campaign in a speech he put in the mouth of Alcibiades after his defection to the Spartans: first Sicily, then the Greeks of Italy, after them the Carthaginians, and finally the Peloponnese (6.90).
After the death of Alexander the Great, a memorandum was produced which was said to contain his plans for further conquests: first the Carthaginians, then the peoples bordering the coasts of Libya and Spain, and back to Sicily (Diodorus 18.4). As another source elaborated it (Curtius Rufus 10.1.17-9), Alexander aimed to defeat the Carthaginians, then, after crossing the deserts of Numidia (North Africa), go to Spain, then skirting the Alps and the Italian coastline, return to Epirus (Albania). A third source said that some writers gave an even more ambitious plan: to sail around Africa, enter the Mediterranean via the Pillars of Hercules (the mountains flanking the Strait of Gibraltar), and then add Libya and Carthage to his empire (Arrian 7-1.1-3).
When he was assassinated in 44 bc Julius Caesar was about to leave Rome to campaign in the east against the Parthian empire, which was centred in modern Iraq and Iran. Plutarch credits Caesar with a grandiose scheme: after defeating the Parthians he intended to cross the Caucasus Mountains, march around the Black Sea, crush the Scythians (peoples north of the Danube) and the Germans, and thus return to Italy via Gaul (Caesar 58.3).
Barbarians could be thought to have far-flung ambitions. When Mithridates of Pontus (in Asia Minor) had been driven to the Crimea in 63 bc, he was held to have planned to march around the Black Sea, up the Danube, and, with some Gauls, invade Italy by crossing the Alps (Cassius Dio 37.11.1; Appian, Mithridatic Wars 102,109). Ardashir, the first king of the Sassanid Persians, who overthrew the Parthians in the ad 220s, was thought by those within the Roman empire to want all the territory once held by the Achaemenid Persian empire (550-330 bc). The Sassanids thus were considered to be a threat to the Roman territories of Egypt, Syria, and Asia Minor (Cassius Dio 80.4.1; Herodian 6.2.2).
The writers who give us these stories use them to illustrate the ambition of the characters involved. It can be given either a positive interpretation, constituting a great-minded search for glory (so Arrian on Alexander), or a negative one, indicating over-reaching pride (so Curtius Rufus on Alexander). It was almost always the latter for barbarians, who were thought by their nature to be disposed to such irrational desires.
Modern scholars tend to downplay these stories; ‘rationalizing’ them down to a more ‘achievable’ scale, or dismissing them either as contemporary wishful thinking or as a later invention. Possibly modern scholars are too quick to condemn. If they had not happened, Hannibal’s march from Spain to Italy, crossing both the Pyrenees and the Alps, and Alexander’s conquests from Greece to India, might well have been considered mere pipedreams.
Let us look more closely at one example: Athenian ambitions in the west. The evidence of Plutarch can be considered to derive from Thucydides, and thus to have no independent value. From Thucydides passages can be used to argue for and against the reality of the plan. It is most fully explained in a speech of Alcibiades which seeks to persuade the Spartans to renew the war (6.90). Here there is every reason for Alcibiades to exaggerate. It is clear that the ‘grand plan’ was not raised openly in the Athenian assemblies which discussed sending the expedition (e.g. 6.16-18), or in the strategy meeting of the generals when they reached Sicily (6.48). From Sicily the Athenians actually asked the Carthaginians for aid (6.88). On the other hand, it was Thucydides’ own opinion that the Athenians desired all Sicily (6.6; c.f. 6.1), and that Alcibiades aimed at Carthage (6.15). Any attempt on all Sicily automatically would have caused conflict with Carthage, which controlled cities in the island. The Athenians receiving help from the Etruscans of northern Italy (6.103) indicates the wide scale of their involvement in the western Mediterranean. In a speech which Thucydides gives to a politician in the Sicilian city of Syracuse, the Carthaginians are said to be constantly apprehensive that they will one day be attacked by the Athenians (6.34). Already in a comedy produced in 424 bc the possibility of an Athenian attack on Carthage had been mentioned, albeit with comic exaggeration (Aristophanes, The Knights 1302-5). It was not openly acknowledged policy, but the idea was in the air. Had the Athenian expedition met with more success in Sicily, the ‘grand plan’ may well have appeared attractive.
The big plans can seem more ‘rational’ and attainable if we think about how the Greeks and Romans imagined the world. In the schemes outlined above, certain geographic features stand out: coastlines, rivers, and mountain ranges. These point towards the way in which the ancients thought about geography. Lacking accurate topographical maps, they tended not to think, as we do, in terms of blocks of territory, so-called ‘cartographic thinking’, but in linear terms, such as the lines of coasts, rivers, or mountain ranges, so-called ‘odological thinking’. The products of this ‘odological thinking’ were written and illustrated periploi, lists of ports and landmarks for coastal sailing, and itineraries, lists of towns and stopping places along roads and land routes. It is these that seem to have been employed in strategic planning. Periploi and itineraries were equally as practical and as divorced from topographic reality as the map of the London Underground.
There were, of course, ancient geographers who attempted to produce topographically accurate descriptions of the world. Three points about them can be noted here. First, although their works were to be of immense importance in early modern Europe during the ‘Age of Discovery’, they remained specialist literature with little broad impact in the classical world. Second, their estimates of the ‘inhabited world’ (in Greek the oikoumene, in Latin the orbis terrarum), in other words the known part of the world and the more or less mythical places surrounding it, were vastly too small. The Greek geographer Strabo argued that the oikoumene was about 8,046 miles ‘long’ (east to west) and about half as ‘wide’ (north to south). Third, as with Strabo, the oikoumene was thought of as an oval along an east-to-west line. Europe north of the Danube was considered to be much smaller than it is in reality (thus making the plans attributed to Mithridates and Caesar less daunting for contemporaries than for us), and the west of Africa and Europe also were vastly compressed (with attendant effects on ancient attitudes to the ‘plans’ of the Athenians and Alexander).
In a sense, geography was of secondary importance in classical strategic thinking. The Greeks and Romans thought of conquering peoples not places. The peoples they looked to conquer were barbarians, and, as we saw in Chapters 1 and 2, barbarians were naturally inferior. Big plans of conquest, in a small world comprised of inferior peoples, who could be reached by following or crossing the lines of certain rivers, coasts, or mountains, could seem far more achievable to the ancients than they do to our eyes.
A ‘grand strategy’ for the Roman empire?
In 1976 the American strategic analyst Edward Luttwak published The Grand Strategy of the Roman Empire. As we shall see, for various reasons, some more pertinent than others, few works have caused such controversy among ancient historians. Looking at two plans from Luttwak’s book (see Figure 8) provides a good path into his ideas.
In the first model, Luttwak sets out his schematic view of the Roman empire of the late Republic and early principate (down to the mid-1st century ad). It is a ‘hegemonic’ empire, where areas directly ruled by Rome (Italy and the provinces) are surrounded by ‘client states’. The latter take responsibility for their own internal order, deal with low-level external threats, and delay higher-intensity ones. As the Roman legions, and in the principate professional auxiliaries, are not responsible for day-to-day defence of the borders of the empire, they form a mobile strategic reserve, which is available to crush independent-minded ‘clients’, defeat high-intensity threats while they are still in the territory of‘client states’, and pursue further conquests.
The second model gives Luttwak’s vision of the grand strategy which prevailed in the empire from the second half of the 1st to the later 3rd centuries ad. It is a ‘territorial’ empire. The ‘client states’ have gone. The legions and auxiliaries now are stationed in permanent bases along the frontiers, where they are responsible for perimeter defence. Further conquests are dangerous, as a troop build-up on one frontier involves stripping them from others. Luttwak implied that this grand strategy was a plan consciously worked out by the emperors and their advisors. It was a defensive policy that aimed at the use of an ‘economy of force’, and so thought carefully about geography, choosing good, defendable boundaries, preferably natural (rivers, seas, deserts, or mountains) and reinforcing them, or if necessary replacing them, with man-made defences (walls, ditches, and cleared ground such as Hadrian’s Wall). It was defensive, but not inert. Good intelligence was to allow the Roman forces to head off threats before they reached the frontier.
8. Luttwak’s two models of empire
Luttwak’s ideas have found some support, but the majority of scholarship provoked has been hostile. It can be suspected that some of this hostility stems from the fact that Luttwak is not a professional ancient historian, and so was seen as an interloper. Certainly, Luttwak’s book, with its enthusiasm for fighting wars in the territory of‘client states’, made for uncomfortable reading in Western Europe during the Cold War (Luttwak was a security advisor to Ronald Reagan in the 1980s). Whatever the contemporary factors at work (and it is good to be reminded that modern historians do not work in an historical vacuum), and despite the tendency of Luttwak’s critics to oversimplify his arguments (as space has forced me to do here), several telling objections have been raised against the ‘grand strategy of the Roman empire’. We can look at some of them here.
‘Client states’ never disappeared. There was a tendency for the empire to turn existing ‘clients’ into provinces, especially in the east, but the Romans never ceased trying to turn peoples outside their direct control into ‘client states’.
Another line of attack is conceptual. The Romans lacked the necessary mental tools to produce grand strategy akin to Luttwak’s. We saw in the previous section that they lacked accurate large-scale topographic maps. Although some records were kept, treaties with foreign peoples and grants of citizenship to favoured foreigners, we can find no trace of any archive on diplomacy or foreign policy. Decisions were made by the emperors, who were expected to consult with their council (consilium). Yet the consilium, which the emperor could overrule, consisted of whomever the emperor invited to attend, and we never hear of specialists on foreign policy in specific areas or in general. The contemporary historian Cassius Dio (76.9.4) tells us that the emperor Septimius Severus was short of information when campaigning in Mesopotamia in ad 198. This was some two hundred years after the first Roman campaigns in the area.
Defensive ideals were voiced. The historian Appian in the 2nd century ad described the Romans surrounding ‘the empire with a circle of great camps’ (pr. 7). In the next century another historian, Herodian, spoke of Augustus having ‘fortified the empire by hedging it round with major obstacles, rivers, trenches, mountains, and deserted areas’ (2.11.5). But such statements scarcely amount to anything like Luttwak’s geographically sophisticated grand strategy, and they sit uneasily alongside expressions of ideals of further conquest. Herodian also claimed that, had he not been stopped by a rebellion in ad 238, the emperor Maximinus would have conquered the Germans as far as the Ocean, and the text leaves no doubt that this would have been a good thing (7.2.9).
The Romans discussed strategy in terms that were rational for them, but can look odd to us. Cassius Dio, who had served on the consilium of Septimius Severus, wrote of that emperor’s annexation of Mesopotamia that: he used to declare that he had added a vast territory to the empire and had made it a bulwark of Syria. On the contrary - this conquest has been a source of constant wars and great expense to us. For it yields very little and uses up vast sums.
(75.3.2-3, tr. E. Cary)
We need not believe that Cassius Dio actually voiced these views to the emperor, but, with the exception of Severus’ open vaunting of imperialism as a good thing in itself, this all seems quite normal strategic discourse to us. The same cannot be said of the reasons Cassius Dio gives for another of the emperor’s expeditions.
Severus, seeing that his sons were changing their mode of life (i.e. for the worse) and that the legions were becoming enervated by idleness, made a campaign against Britain.
Campaigns and logistics: some general considerations
For a modern state, fighting a war, win or lose, is very expensive. Warfare in the classical world also could involve huge costs, especially siege and naval war. The exactions the emperor Maximinus imposed to pay for his German war led to a revolt which ultimately cost him his life (Herodian 7.3.1-4.6). Yet wars could also make huge profits. It all depended on who you were fighting and how successful you were. It has recently been demonstrated that an inscription on the Colosseum in Rome stated that it had been funded by booty. This massive project was paid for by a part of the wealth gained by the emperor Vespasian and his son Titus crushing a revolt in just one quite small province, Judaea.
One type of expense incurred in modern wars was not always present in the ancient world: paying the troops. As we have seen, in the classical city state citizenship was bound up with military service. For a long time city states thus had no need to pay their citizen-soldiers. In the Greek world the only city for which we have much evidence of military pay is Athens. There, pay appears to have been introduced in the 5th century bc as Athens acquired an empire, and at first it seems to have been a form of living allowance. During the Peloponnesian War the concept of military pay broadened to include remuneration for service, and other Greek cities began to pay their soldiers. The Romans introduced military pay during the siege of Veii, which ended in 396 bc. The emperor Augustus set up a special treasury and introduced two new taxes to pay the professional army of the principate. It is to be doubted if basic military pay was ever a road to riches. In the principate if a soldier lived long enough to collect his retirement bonus, he would be comfortably set up for the rest of his life. Otherwise, throughout the classical world, a soldier would have to look to booty for serious economic advancement.
Even some mercenaries did not need paying. Some, such as Thracian tribesmen in the Peloponnesian War, served for free in the hope of booty. Yet most mercenaries had to be paid. Before the Peloponnesian War, most Greek mercenaries came from the poorest parts of Greece, such as Arcadia, and tended to serve either non-Greek paymasters, such as the kings of Lydia or the Egyptian pharaohs, or Greek tyrants. After the Peloponnesian War, mercenaries appear from all areas of the Greek world, and are employed not only by non-Greeks, like the Persian pretender Cyrus, the service of whose 10,000 Greek mercenaries was immortalized in Xenophon’s Anabasis, but by Greek cities of any type of constitution. After the death of Alexander, the wars of his successors marked the high point of mercenary service in the Greek world.
Very occasionally we hear of Romans serving as mercenaries abroad (for example, one Lucius set up an inscription some time between 217 and 209 bc commemorating his service in Egypt, IC III 4, no.18). Yet most Romans who served foreign rulers would not have seen themselves as mercenaries. They were voluntary or enforced political exiles fighting for their restoration, as were those in the army of Mithridates of Pontus. Under the Republic, Rome, whose main strength was legionary heavy infantry, often needed additional cavalry and light infantry. Some of these were provided by mercenary service, such as the Cretan archers who fought in the Second Punic War. But, although the line between the two was blurred, the majority came from allied contingents. The professional auxiliary units of the principate removed the necessity for mercenaries, although supporting troops could still be provided by allies, and these allies might receive subsidies from Rome. The use of mercenaries revived in the late empire. In the 5th century ad prominent individuals began to hire private troops of barbarian mercenaries. Rufinus, the praetorian prefect of the emperor Arcadius, maintained a personal guard of Huns. Such troops came to be called Bucellarii (‘military biscuit eaters’), and in the 6th century ad often formed a significant part of Roman imperial armies. By the early 7th century ad ‘Bueellari’ had become the title of a regular cavalry unit.
Logistics, the supply of water, food, firewood, fodder, and other material, were of vital importance to ancient armies, but are easily overlooked. Logistics were not often discussed in ancient literature. Normally they only got a mention when things went disastrously wrong, as when Alexander the Great crossed the Gedrosian desert (Arrian 6.22-6). Our poor evidence makes reconstructing ancient logistics especially difficult. Nevertheless, enough material can be assembled for book-length studies; particularly notable are two works on Roman logistics in general, and one on Alexander’s conquest of Persia. These works have to proceed by taking logistical evidence from more recent, better-documented periods and extrapolating it backwards onto the scattered ancient evidence. This has to be done with great care. For example, estimates for the needs of modern soldiers cannot be automatically applied to ancient ones, who tended to be older, smaller, and more inured to hardship. Again, modern estimates for such things as the weights that can be carried by various pack animals vary by huge percentages. The findings of these modem studies are provisional in the extreme.
Ancient armies seldom campaigned in the winter. A winter campaign demanded special reasons. For example, Alexander’s expedition against the Cossaeans of the Zagros Mountains in 324-323 bc was held to have been inspired by either the king’s desire to find in action solace for his grief at the death of his friend Hephaestion (Plutarch, Alexander 72.3) or, the explanation preferred by most modem scholars, the strategic purpose of catching the tribe when weather conditions precluded flight uphill (Arrian 7.15). The reluctance to campaign in the winter was less to shelter the troops from bad weather, than because of the unavailability of fodder, and the difficulties of moving supplies. Water transportation was the preferred method for bulk goods in antiquity, land transportation being slow and inefficient, and sea travel was especially dangerous in winter.
Logistics become more of a problem the larger armies become, the longer they are assembled, especially in one place, and the further from home they operate.
Armies of the ancient Near East clearly had good logistic capabilities. Although Herodotus’ figures must be wildly inflated (7.186), the Persian army that invaded Greece in 480 bc consisted of tens, if not hundreds, of thousands of men. Similarly, the Carthaginians by the mid-4th century bc were able to supply armies of tens of thousands of men operating abroad. Probably logistics were not usually a big problem for classical Greek armies, which tended to be relatively small, operate at no great distance from home, and not stay in existence for long. Seemingly in the normal run of things, the authorities set a date for assembly and instructed the troops to provide themselves with a certain number of days’ provisions. Alexander’s logistical difficulties would have been eased by his customary rapidity of movement and, after capturing the Persian treasuries, his limitless wealth. Alexander’s successors and the Hellenistic kingdoms had to develop sophisticated logistic arrangements because their armies were large, often contained a high proportion of cavalry, and might stay in the field for years at a time. In 306 bc Antigonus I managed to cross the Sinai desert with an army of almost 90,000 men and a siege train (Diodorus Siculus 20.73.3-74.5).
When the Romans first fought overseas, in Sicily during the First Punic War, they had to begin to develop a high logistic capacity, based on a mixture of foraging, requisitioning, and supply lines. Although it is hard to ascertain normal practice, the Romans were capable of extraordinary feats of logistics. In the Third Macedonian War (172-167 bc) they ran a supply route for about 100 miles through mountainous terrain in the Balkans, and in ad 73/4 managed to supply a large army for a siege of several weeks, if not months, at the waterless site of the desert fortress of Masada. After their defeat of the Hellenistic monarchies, the Romans enjoyed a greater logistic capacity than any of their enemies. This advantage was noted. Cassius Dio said that the Parthians’ war-making efforts were hampered because they ‘do not lay in supplies of food or pay’ (40.15.6).
Greeks and Romans liked to think of huge baggage trains as being typical of barbarian armies. Quintus Curtius Rufus gives a splendid description of the huge retinue, including carriages for 360 concubines, which followed the Persian king Darius III as he set out to fight Alexander (3.3.8-25). There was an awareness of the problems of large baggage trains. They slowed the army down, and if they became mixed up with the troops the confusion impaired fighting ability. The latter was held to be partly responsible for the disaster in the Teutoburg Forest in Germany, where Varus’ entire army of three legions was wiped out in ad 9 (Cassius Dio 56.20.5). Several famous generals, including Alexander, destroyed, or sent away, what they considered inessential baggage. The intention in part was to restore discipline. Scipio Aemilianus in Spain in 134 bc ‘expelled all traders and whores, as well as the soothsayers and diviners, whom the soldiers were consulting’ (Appian, Wars in Spain 85). When ancient sources give numbers for camp followers or wagons, it is because they are exceptionally large. In 171 bc in a raid the Antigonid king Perseus captured 1,000 wagons from his Roman opponents (Livy 42.65.1-3). At the battle of Orange in 105 bc the Romans are said to have lost 80,000 soldiers and 40,000 servants and camp followers (Livy, Periochae 67). Modern scholars, however, can be thought to play down the size of baggage trains in the armies they study. For example, the major work on Macedonian logistics accepts as both true and normal campaigning practice a figure for camp servants that is both dubious and refers explicitly to one training exercise (Frontinus, Stratagems 4.1.6).
Ancient writers tend to record the speed of march of armies only when they are exceptional. Usually these are very fast. In 329 bc Alexander, with a specially selected ‘flying column’, covered about 185 miles in three days (Arrian, Anabasis 4.6.4). Sometimes figures are given because the march was so slow. A Roman army campaigning in Asia Minor in 189 bc was so laden with booty that it made a bare six miles in a day’s march (Livy 38.15.15). There are so many potential variables - the state of the roads/paths, the weather, the composition of the force, the size and type of the baggage train, the time taken setting up a camp, the proximity of the enemy, and the perceived need for haste - that attempts to produce average figures are difficult. Archaeology can help with some campaigns. Although it is notoriously hard to date Roman marching camps, two groups in northeast Scotland have been identified as belonging to the campaigns of Agricola, in the late 1st century ad, and Septimius Severus, in the early 3rd century ad. The distances between these camps indicate a slow rate of march, less than 15 miles a day.
Campaigns and logistics: ‘unhorsing the Huns’
The explanatory possibilities and pitfalls of logistical analyses of ancient armies can be illustrated by taking R. P. Lindner’s article ‘Nomadism, horses and Huns’ as a case study. This work’s stated aim is to ‘unhorse most of the Huns’. Its conclusion is that when the Huns settled on the Great Hungarian Plain (ad c. 410/420-c. 465) they ceased to be nomads, and thus ceased to fight as cavalry. Two lines of argument are deployed to support these findings: one textual and one ecological. Lindner points out that some contemporary sources do not explicitly refer to the Huns as cavalry, while others that do are dismissed as unhistorical because they follow an earlier description (that of Ammianus Marcellinus 31.2, written ad c. 395). On the ecological line Lindner estimates that the pasturage of the Great Hungarian Plain could feed only about 150,000 grazing nomad horses. By analogy with the later Mongols, it is considered that each Hun needed ten horses, and thus there could have been only 15,000 Hun cavalry in this period.
Some scholars in passing have agreed or disagreed with Lindner, but, as far as I am aware, to date there has been no extended academic engagement with his views. A few comments can be made here.
Literary arguments from silence are always suspect. It could be that some contemporaries did not explicitly describe the Huns as cavalry precisely because everyone knew that they were. To dismiss those sources that do portray the Huns as cavalry as unhistorical because they are indebted to Ammianus might be to read them in an anachronistic way. In classical literary culture it was good to show one’s wide reading. As such, it was always apposite to allude to earlier distinguished writers. To give an example, after Thucydides had written a famous description of the plague in Athens in 430 bc (2.47-55), subsequent authors who did plague scenes tended to draw on Thucydides. Which is not to say that the plagues they depicted were imaginary. Ammianus’ description of the Huns may have come to occupy a similar position. Furthermore, Lindner makes no mention of some contemporary literary evidence, uninfluenced by Ammianus, which does talk of the Huns as cavalry. Vegetius, whom Lindner takes to be writing in the mid-5th century ad, in his Epitoma rei militaris, twice holds up the Huns as model cavalrymen (1.20; 3.26). Also it can be noted that in the next century the Huns who served as mercenaries in the Byzantine army of Belisarius are ‘all mounted bowmen’ (Procopius 3.11.12).
Even if Lindner’s estimates for the grazing potential of the Great Hungarian Plain are correct, problems remain with his deductions. Firstly, Hun society now was very different from its earlier form on the Steppes to the northeast of the Black Sea. The Huns had acquired an autocratic monarchy, social stratification, a huge empire, and, as literary sources show them living in villages, they had indeed ceased to be ‘pure’ nomads. The Huns were equally adept at extracting agricultural produce from their subjects and tribute from the eastern Roman empire. These could supplement the forage of the Plain, if the Huns were prepared for another change, feeding their horses partly by grain. Secondly, Lindner’s analogy with the Mongols is arbitrary. Other figures can be played with. As Lindner states, three changes of horse a day might wear them out. But if the Huns were prepared to risk this, then on the given figures the Plain could have supported 50,000 cavalrymen.
Finally, there is a problem with the idea that if the Huns stopped being nomads then they stopped being cavalry. All Steppe nomads were cavalry, but not all cavalry were nomadic.
After a defeat in Gaul in ad 451 we are told (Jordanes, Getica 41.215) that Attila contemplated suicide by burning himself to death on a pyre of his followers’ saddles. Lindner comments that this ‘is proof not that he had many horsemen but that he led too few’. This may be true in the sense that by this time a large proportion of Attila’s army was made up of subject peoples, many of whom fought as infantry. Yet it takes quite a leap of faith to see this as evidence for most of the Huns having given up their horses.