Figure 5.1 Reconstructed gate of Roman fort
The strategic infrastructure of the Empire consisted primarily of the road network and fortified structures, which include cities, towns, and the limes along Rome’s frontiers. Of course, other infrastructure, such as canals, irrigation systems, and aqueducts, were all crucial in the maintenance of the Roman economy, which provided the foundation of all strategic actions. But this work assumes that these could be built and sustained only in an environment in which Roman military power maintained a general peace. And the maintenance of peace required, above all else, roads and forts.
The Roman road system was built solely for the purpose of speeding the movement of the legions. Everything else the roads were used for was simply an added benefit. Most everyone has heard the phrase “all roads lead to Rome.” But to better comprehend how the Romans understood the role of their road network, it is best to reverse that statement: “all roads lead away from Rome.” Rome was the center of the Empire, from which, at least through the centuries of the Principate, all power emanated. Rome built its thousands of miles of roads for the singular purpose of pushing power out from the Empire’s core to its frontiers and then, later, from one part of the Empire to another.
By the time of Diocletian (284–305), the Roman road network consisted of 56,000 miles of roads, supplemented by over a thousand bridges, an impressive total that was not exceeded until the early twentieth century.1 One estimate places the cost of building this network at approximately the same as that of maintaining thirty legions for the entire five hundred years of the Empire.2 As this is only the construction cost and does not include yearly maintenance, it is clear that the cost of the Roman road network greatly exceeded the cost of the army during the span of the Empire. As such, it was a major drain on the Empire’s financial resources. Of course, in the same way that the Roman army paid for itself by maintaining the Pax Romana, which underpinned Roman prosperity, the road network, by hugely easing the cost of trade, more than justified its expense.
The primary military purpose of the road network was to speed the movement of the legions from their bases to either a point of crisis or a mobilization point. It has often been theorized that the roads were necessary to ease the movement of supplies, particularly food and fodder. But there is no evidence in the extant records of this ever having taken place.3 There is a very good reason for this: a well-defined limit exists on how far premodern military formations could move if they had to carry their own food and fodder. In a modern example, the German army in 1914 rapidly exhausted its supplies once it got over a hundred miles from its railheads. This more than any other factor explains the failure of Germany to defeat the French army in the first year of World War I. General William Tecumseh Sherman, in 1864, did lead his army 285 miles in his famous “March to the Sea,” but his soldiers mostly lived off the land, and the animals were expected to make only a one-way trip. A Roman army, no matter how liberally supplied with wagons and pack mules, could travel at most about two hundred miles, and then only if the beasts of burden were allowed to perish rather than make the return trip. Any farther and the wagons and mule packs would consist of nothing but grain and fodder for the animals, which would be rapidly consumed. For a Roman army to move over a thousand miles, as it often did, orders would have to be dispatched all along the march route weeks or months ahead of time, requiring local officials to gather supplies at selected points all along the route. Unless the march route could be supplied by sea or river transport, there was no other way to accomplish the task.
The reality of an extended Roman march then was that the beasts of burden carried the soldiers’ requirements—tents, utensils, weapons, siege trains, and artillery—but little or no food. But because food was a necessity, the only time roads were intensively used for the logistical support of the armies was for short-distance affairs, moving food from nearby port facilities to bases or from nearby agricultural centers to encampments. When the army needed to move bulk supplies to a distant location, maritime transport was the method of choice, as it was faster and cheaper. Wheat moved by sea could travel at twice the speed of land transport, cover nearly unlimited distance, and do so at about one-fiftieth the cost.4 The second major strategic purpose of the road network was moving key personnel and information, which was often one and the same. To accomplish this, the Romans maintained the cursus publicus—Rome’s nervous system. This system was first established by Augustus and maintained until the fall of the Western Empire.5At intervals along all of the Empire’s major roads, there were stations where animals were maintained, along with wagons and carts. These were made available only to those on public business and high-ranking soldiers. Rome did not maintain anything like the American west’s Pony Express, with a corps of dedicated messengers. Rather, if a governor wished to send a note to the emperor, he would have to send someone from his own staff. At these stations the messenger would be able to change modes of transport (get a new horse), but he would not find a team of professional messengers waiting to take the note on the next leg of its travels. The crucial advantage of this system is that the messenger came from the scene of events whence the message was dispatched and could be further interrogated to provide context and additional information as required.
To make sure the cursus publicus was not overused, only a few persons within the Empire were permitted to issue permits (diplomata) for the use of the system’s resources. Such permits would often limit the number of resources a traveler could use. For instance, a messenger who had to get to Rome with vital information—such as if a mass of barbarians had crossed the Danube and were making their way south—might be authorized a new horse at every stop, while a messenger delivering another of Pliny the Younger’s incessant requests for guidance might have to make do with a new and rested horse every half dozen posts. One of Pliny’s messages dealt directly with the workings of this system:
I beg you, Sir, to write and tell me whether you wish the [diploma] permits, the terms of which have expired, to be recognized as valid, and for how long, and so free me from my indecision. For I am afraid of blundering either one way or the other, either by confirming what ought to lapse, or by putting obstacles in the way of those which are necessary.
The permits, of which the terms have expired, ought not to be recognized, and consequently I make it my special duty to send out new permits to all the provinces before the day when they are required.6
Most permits were issued by the emperor, who would send fixed quantities of preapproved permits to each governor to make use of them as he saw fit. In the later Empire, a few high-ranking officials—praetorian prefects—were also entitled to do so.
While the speed of information traveling from one end of the Empire to the other appears agonizingly slow to modern readers, who can reach practically anyone in the world instantaneously, for a strategist the speed at which information travels is relative. While Rome ruled, its capacity to move information long distances was unparalleled, giving it a decisive informational advantage over its enemies. Even the Persians, who had made use of a royal communications system since the days of Cyrus and Darius, could not match the speed and efficiency of Rome’s network, as the Persians could not make use of sea or river transport, except for short distances along the Tigris and Euphrates rivers. How the Romans thought about this network is captured in an oration given around the middle of the second century by Aelius Aristides:
And if the governors should have even some slight doubt whether certain claims are valid in connection with either public or private lawsuits and petitions from the governed, they straightway send to him [the emperor] with a request for instructions what to do, and they wait until he prepares a response, like a chorus waiting for its trainer. Therefore, he has no need to wear himself out traveling around the whole empire nor, by appearing personally, now among some, then among others, to make sure of each point when he has the time to tread their soil. It is easy for him to stay where he is, and manage the entire civilized world by letters, which arrive almost as soon as they are written, as if they were carried by winged envoys.7
No doubt Aristides was overstating the capabilities of Rome’s communication system, but the big picture is correct—a Roman emperor could influence events in every corner of the Empire, no matter his location. In fact, the weakening of this capacity likely had a large influence upon Rome’s ability to meet successive crises during the later Empire. In this regard, Richard Duncan Jones, in a remarkable piece of detective research, calculated the differences in communication times between the two eras. By examining data on how long, on average, it took for news of a new Emperor to reach Egypt—fifty-seven days during the Principate—and comparing it to the travel times indicated in the double-dated edicts of the Theodosian Code (438 CE), which took, on average, 134 days, he was able to document the collapse of Rome’s transport network.8 A more than doubling of communications times from the core of power to the frontiers had to have had a huge impact on the conduct of strategic policy and military operations.
The term limes was originally used to describe a boundary, not a frontier, and certainly not a fortified frontier. By the end of the first century, the term was still used to describe the boundaries of the Empire, and only later as a frontier district. It was also often employed in reference to the roads linking Roman military installations along the frontier. Only in the third century does the word become synonymous with a wall or fortified boundary. Although it never carried the same connotations as we give to national borders in the modern world, as the Romans always understood it, there were many reasons to place forts and other structures beyond the boundary.9 This is a key point: The limes might be the boundaries of the Empire’s territory, but they were never meant to mark the limits of Roman power or influence. The Roman Empire might end at the Danube, the Rhine, or the Euphrates, but there was never any question of Rome’s acting far beyond these boundaries when it had an interest in doing so. In the Roman mind, the Empire’s territory might be measurable, but Roman influence was boundless.
Although Rome never lost its expansionist impulse, and while successful foreign wars were typically the surest way to ensure regime longevity, by the time Tiberius ascended to the purple Rome’s boundaries were increasingly fixed. In fact, by the first decades of the Principate, Rome was much more concerned with internal security than with expansion. In this regard, it is worth, once again, examining the placement of the legions as presented by Tacitus in 23 CE:
But chief strength was on the Rhine, as a defense alike against Germans and Gauls, and numbered eight legions. Spain, lately subjugated, was held by three. Mauretania was king . . . Africa was garrisoned by two legions, and Egypt by the same number . . . Syria . . . was kept in restraint by four legions . . . the bank of the Danube by two legions in Pannonia, two in Moesia, and two also were stationed in Dalmatia, which, from the situation of the country, were in the rear of the other four, and, should Italy suddenly require aid, not too distant to be summoned. But the capital was garrisoned by its own special soldiery, three city, nine praetorian cohorts.10
Note that Tacitus tells us that the eight legions on the Rhine were not just there to guard the frontier—they also had an internal stability role in the event the recently conquered Gauls became restive. The same is seen in the east, where the primary role of the four legions in Syria is to keep the eastern provinces “in restraint.” Guarding the frontier against the Parthians is assigned to client kingdoms, which are kept safe from major attack by their ability to call on the legions for aid. Even the Danube is lightly held by four legions—two on the upper Danube (Pannonia) and two on the lower Danube (Moesia). Notably, the two legions held in reserve, deeper in the Balkans, do not have a primary responsibility of aiding hard-pressed forces on the frontier. Rather, they are held in reserve in case Italy required aid.
It is worth taking a moment to note that Tacitus’s recounting of the locations and strategic purpose of each of the Empire’s major concentrations of military power is positive proof that the Romans clearly understood the geography of their Empire and could arrange forces in accordance with strategic considerations. But it also reveals that Rome was not yet overly worried about protecting the frontiers. What concerns they did have were subsumed by the strategic necessity of maintaining internal stability and securing the position of the emperor. Given the weakness of Rome’s external enemies, it would be odd if we saw large expenditures on permanent frontier defenses that were not yet required.
It is not until the reign of Claudius that we first see Rome building rudimentary permanent defenses along the frontiers. As there likely was no increase in the threat level, an explanation is required. Two are readily at hand. The first is that the legions had now been largely static on the Rhine and the Danube for almost a generation. It is only natural that their encampments would transform over time into more comfortable and secure permanent installations and that connecting roads would be built along with smaller outposts to watch the areas between major camps. The second is a consequence of Claudius requiring more than the two new legions he built—XV Primigenia and XXII Primigenia—to conquer Britain. These troops had to be taken from the Rhine, and removing them greatly weakened the forces along that frontier. Interestingly, we find that the legionary base at Xanten was rebuilt using stone at about this time. Stone, of course, provided a much-enhanced defense for a smaller garrison than the earlier wooden palisade. Moreover, there was an expansion of smaller forts, called watchtowers, all connected along the frontier. On the Danube, Thrace, long occupied and still holding its client status, was annexed, legions were moved from Raetia to the Danube, and new roads such as the Via Claudia were built. In short, Claudius’s need for troops to expand the Empire and to give him a military reputation necessitated the installation of Roman limes.11
From these small beginnings, Rome increasingly emplaced a fortification system that stretched along its entire 4,800-mile frontier system, reaching its apogee in the forty years between 120 and 160 CE. This is the period, starting with Hadrian, when Rome constructed the bulk of its fortified lines along the frontiers.12 Crucially, as you move from region to region, there is no set pattern to which the limes’s engineers were required to adhere, as many factors went into their building. The first and possibly the least important was local resources, as it was much easier to build from local material than to have to import stone or wood from distant portions of the Empire. More important was regional geography. Britain could have a solid stone wall along its entire land frontier because it spanned barely sixty miles from end to end. Similarly, having the ability to incorporate a major river into the defensive system allowed Rome to construct defenses relying on infrastructure other than solid walls from the North Sea to the Alps. Similarly, Rome did not need to defend the vast desert tracks along its African frontiers—here a few good roads and a couple of forts could do the job. The most crucial consideration was always the threat presented by Rome’s enemies. Walls and rivers could stop almost any threat the Principate faced in the Western Empire. In the east, however, a more sophisticated enemy capable of massing huge armies was not going to be deterred by a wall or any kind of linear defense (Fig. 5.1). In the east, Rome required strong points arranged in depth. These strong points, placed along key avenues of approach and incorporating the walls of the east’s great cities, were designed to attrit a Persian army in a series of wasting sieges that bought time for the Roman field army to mobilize and march.
Figure 5.1 Reconstructed gate of Roman fort
The purpose of the limes has been much disputed in recent decades. As one historian writes: “Some of the activities of the Army in these regions [Northern Europe] continue to baffle scholars. There is no real consensus as to what such monumental linear boundaries as the walls of northern Britain or between the Rhine and the Danube were for or how they functioned.”13 Another historian has identified twenty-one separate reasons for the construction of the limes system.14 They range from “a piece of rhetoric” to a place to reflect on “Rome’s failure to conquer the world.” Much of this “limes denial” has already been discussed earlier in this work. Unable to accept that a fort or system of fortifications can serve multiple purposes, too many historians have employed alternative reasons ranging from the probable to the absurd to obscure the primary reason Rome would undertake such a vast expense. But if we were to allow just a bit of common sense to enter the argument, the answer is apparent: For Rome, the limes served the same crucial purpose all walls at all times have served—to keep the people on the other side of the wall out.15 The limes were first and foremost a military system designed to aid the legions in the defense of the Empire. One can easily see this in the Roman general Arrian’s thoughts on the limes in a report from one of his inspection tours:
At Apsaros, where the five cohorts are stationed, I gave the army its pay and inspected its weapons, the walls, the trench, the sick and food supplies that were there. [At Phasis] the fort itself, in which 400 select troops were quartered, seemed to me, owing to the nature of the site, to be very secure, and to lie in the most convenient spot for the safety of those who sail this way. In addition, a double ditch has been put round the wall, each ditch as broad as the other. The wall used to be of earth, and wooden towers were set up about it; now both it and the towers are made of baked brick. And its foundations are firm, and machines are installed, and in short, it is fully equipped to prevent any barbarians from even approaching it, let alone to protect the garrison there against the dangers of a siege.16
But just how effective were the limes, and was the system worth the cost? General George S. Patton considered all fortifications a waste of resources and monuments to the stupidity of man, but that was before his Third Army’s rapid advance was halted by the broken-down German border defenses, manned mostly by raw recruits. Of course, the expense of fortifications always appears wasteful once they are pierced. But to strategists no walls or fortified zones are meant to be impassable. Thus, Rome never intended the limes to be an impenetrable barrier; they had too great an understanding of war to cloud their thinking with fanciful ideas. After all, the Romans were the ancient world’s master of siege warfare. They knew only too well that any fortress can be taken by a large enough force of men determined to take it. But walls make everything harder for an enemy. For instance, when the Goths, in the fourth century, abandoned the siege of Adrianople because it was costing them too many men, their leader, Fritigern, declared he would “keep peace with walls.”17 In that regard, Patton was right: if men can overcome oceans and mountains, fortifications can also be overcome. Of course, fortifications that reinforce such natural barriers can be fiendishly difficult to defeat.
For Rome, the limes served several strategic purposes depending on their location, including acting as a shield to mass forces behind for a surge into enemy territory, deterring or defeating raids and other small incursions, establishing points of contact for trade and diplomacy, delaying a major invasion force, and making it difficult for a booty-laden enemy force to escape Roman territory. But their primary strategic purpose was to permit an economy of force.18 A lone legion in a fortress could defend against enemy armies that it would take multiple legions to confront in the open (Fig. 5.2). Rome was able to concentrate forces for a true crisis because it could leave only skeleton forces in the limes and count on them to contain all but the most serious threats. In this role, the limes succeeded spectacularly for over two hundred years, until during the Crisis of the Third Century the defenses were struck at multiple points by large numbers of enemies. Still, in the wake of the crisis, during the reigns of Diocletian and Constantine, the limes were largely rebuilt and strengthened. It was not until they were weakened by civil war and neglect that the western limes were overrun and the integrated defensive systems wrecked beyond repair.
Figure 5.2 Reconstruction of a Roman fortress