Bryn Mawr Classical Review 1999.11.01
Jonathan P. Roth, The Logistics of the Roman Army at War (264 BC - AD 235). Leiden: Brill, 1998. Pp. xxi, 400. ISBN 90-04-11271-5.
Reviewed by Adrian Goldsworthy, Penarth, South Wales
Word count: 1962 words
The supply of ancient armies in the field rarely receives much scholarly attention, despite its widely acknowledged importance. Strategy and tactics are easy to understand and, with hindsight, even easier to criticise, but the science of equipping and feeding armed forces on campaign is extremely technical, its mysteries understood by only a few initiates. Yet more than anything else logistical constraints limit the activities of any army during a war. The Roman Republic's aggressive and highly successful war-making was made possible not just by its massive reserves of manpower, but its ability to supply armies campaigning great distances away from Italy. Similarly, all the discipline, tactical flexibility and superior equipment of the later professional army would have been of little use without the ability to feed men and animals in the field. How the Romans did this is the subject of R.'s book.
Although there has been some work on the army's wartime supply system, nothing as detailed and wide-ranging as R.'s book has appeared until recently.1 On the whole Roman Army Studies have relied heavily on the steadily increasing archaeological and epigraphic record, focusing attention mainly on the army of the Principate and its peacetime routines. By far the most influential work on the supply of ancient armies remains D. Engles, Alexander the Great and the Logistics of the Macedonian Army (Berkeley, 1978), which emphasised the difficulties of transporting food supplies overland, problems that Alexander's genius and the administrative skills of his officers overcame by massing large amounts of food in central locations ahead of his advancing army. Whether or not such a model is indeed valid for Alexander's Eastern Expedition, R. presents a very different picture of the Roman army's operations. Although making full use of other types of evidence, including archaeology and comparative material from nineteenth and twentieth century armies, the bulk of R.'s evidence comes from the literary record. R. demonstrates that whilst the classical authors seldom, if ever, discuss logistics in detail, they contain many snippets of information about supply. This is especially true of Polybius, Livy and Caesar, leading to an emphasis on the practices of the Republican armies, although R. argues that the less full record for the Principate suggests considerable continuity.
R. begins with the basic daily ration of the Roman soldier. Here he argues that modern historians have overestimated the basic nutritional requirement, or in US Army parlance Recommended Daily Allowance (RDA), for the ancient soldier. Engels based his calculations for the needs of Alexander's soldiers on the US Army's RDA for a 19 year old soldier in the 1960's, namely 3,600 calories. R. points out that the average Roman legionary was both smaller and older than the rank and file of modern armies, suggesting an RDA of nearer 3,000 calories. This may seem a minor point, but a one sixth reduction in the size of the basic daily ration becomes a highly significant factor when it is remembered that we are dealing with the problems of supplying armies of tens of thousands. The Roman soldier's ration was divided into two main elements, frumentum (the grain element) and the cibaria (everything else including meat and drink). R. argues for a daily grain ration of 2 sextarii, so that an eight man contubernium received 1 modius, a reconstruction which is made all the more plausible because it results in the allowance of every unit being a whole number, for instance 60 modii for a cohort and 600 for a legion. Precise allowances for the other parts of the ration are harder to judge, but R. argues that most modern scholars have overestimated the amount of meat issued, basing it on nineteenth and twentieth century ration scales for western armies all produced by cultures who have had an untypically high proportion of meat in their diet. There may have been a little more variation in the type and amount of meat issued to different types of soldier that R. allows. Bone finds from military sites in Britain show a higher proportion of pig remains in legionary fortresses than auxiliary forts. Whether this is a reflection of different dietary requirements, preferences, or simply the problems of transportation is unclear.2 This is one area where new finds of military documents on papyrus or wax writing tablets may add to our knowledge, as for instance one of the recently published late first century AD documents from Carlisle deals with the issue of barley to cavalry turmae.3
Food and drink made up the vast bulk of the supplies needed by a Roman army, since in comparison to modern armies their requirement for large amounts of ammunition was minimal. In addition to feeding its soldiers, the army had also to provide fodder for its mounts and baggage animals. Here R. emphasises that the amounts of fodder given as the recommended requirement in recent military manuals tend to be very high, and that much smaller amounts were actually issued by those same armies on campaign. In all but a few areas, the fodder issued was supplemented by extensive grazing. Firewood was another vital requirement for any Roman army, but one which it is easy to overlook from a modern perspective, since the ration food was prepared and cooked at the level of the 8-man contubernium and not issued ready to eat.
How the army's supplies were carried is discussed in detail. R. supports the views of Junkelmann and others who argue that soldiers were accustomed to carry great weights in addition to their equipment so that they might well carry several weeks rations of grain or biscuit (bucellatum). The section on the carrying capacity of the various pack animals draws extensively on comparative material, but once again R. emphasises that recent theoretical manuals tend to recommend ideal loads, intended to keep the animals healthy and prolong their useful life. In the field, many armies have tended to work their pack animals to death, frequently overloading them. We may recall the problems Germanicus had in obtaining replacement horses from the Gallic provinces as his German campaigns dragged on (Tacitus, Ann. 2. 5). R. divides the army's baggage into four categories, the troop train attached to the unit, the army train for the entire force, the officers' train for their personal equipment, and the siege train, although noting that our sources subsume everything under the labels of impedimenta or agmina. A vital role was clearly played by the servants and camp followers accompanying the army and R. discusses in some detail the different groups amongst these. Lixae occasionally include sutlers following the army, but seems most often to refer to the employees of the soldiers, some of whom were free, and may have included sutlers licensed to a particular unit, a phenomenon certainly seen in many more modern armies. The other main group, the calones, were slave servants, possibly state owned, and by the late Republic usually armed and sometimes organised under the command of galearii, leaders whose name seems to have derived from the wearing of a helmet. R. suggests a soldier to servant ratio of 4:1, but this is conjectural and in reality the ratio may have varied considerably, and as R. himself notes, it was considered a sign of bad discipline to have too high a proportion of servants.
R. extends Labisch's model of Caesarian logistical organisation to the rest of the period. In this view the army employed three types of bases, Strategic, Operational, and Tactical. Strategic bases were found in the area producing the food to be used by the army, collecting this at a central point, and were often in a province some distance from the theatre of operations. Operational bases were often ports in the area or province where the war was being fought, locations providing massive storage facilities where supplies could be accumulated and sent forward. Tactical bases were set up fairly close to where the army was fighting to provide for its immediate needs. Tactical and Operational bases were often connected by a string of guarded depots to ease the passage of convoys. There are no real Latin or Greek terms to conform with these categories, but R. marshals strong evidence from the Republic to support the model, which seems to have evolved in the late third century BC. Initial problems in the first overseas campaign during the First Punic War, when the Roman armies in Sicily had repeated crises when their supply lines threatened to collapse, and during the early stages of the Second Punic War, for instance in Spain, prompted the rapid creation of a far more effective system. By the end of the Hannibalic war, Scipio had few supply problems in his invasion of Africa, and the Romans had little difficulty shifting vast amounts of food over great distances in their wars in the east during the next century. Administratively this seems to have been worked by appointing someone to supervise the collection of supplies for an individual army, not by the area producing them. R. admits that the evidence is inconclusive, but believes that this continued under the Principate, doubting the existence of a central logistical organisation. Instead, campaigning armies drew upon the resources of all of the neighbouring provinces for what they required.
R. depicts a very flexible system, with the same army often employing several different methods to supply itself. An army would forage where possible to supplement the food carried with it or being brought along its supply lines. Resources could be transported by land, sea or river depending on which was the most suitable. In extreme conditions, such as in the desert, an army might carry all of its water and the bulk of its other supplies. The picture emerges of a very efficient, adaptable support logistical structure, capable of supplying an operating army almost anywhere, and it cannot be over emphasised just how much of an advantage this gave the Romans over the majority of their opponents from the Later Republic onwards.
The efficiency of the Roman army's supply system and its ability to move huge amounts of material over vast distances have implications for our understanding of the ancient world outside the military sphere, for it is in this context that we must judge the limits of civilian bureaucracy and long distance trade. The forward planning involved in military supply was massive. The neatness of the ration allocations for different types of units suggested by R. allowed the army to prepare effectively for the operations of armies of a set size. R. notes that the variation in the size of Republican legions suggests that at this period planners had to base their calculations on numbers of men rather than numbers of units. However, it is also important to remember that the planning stage was not necessarily reflected in the actual distribution of rations to the men, which occurred at lower level.
This is a very good book on an important subject. In a brief review of this nature it is impossible to discuss all the issues R. raises. Some are minor, others less so, but for very many there is not enough evidence to reach a final conclusion. Yet even if the reader does not agree with all of R.'s interpretations on these points, this is the nature of the evidence and this work will certainly help to clarify the debate on these matters. A few years ago I expressed the opinion that there was not enough information to write a proper study of the logistics of the Roman army. Many gaps remain, although fewer than I believed, and it is unlikely that most of these will ever be filled, but this book is a major contribution to our understanding of this topic and I am delighted to have been proved wrong so quickly.
1. Notable works include A. Labisch, Frumentum Commeatusque: Die Nahrungsmittelversorgung der Heere Caesars (Meisenheim an Glan: Verlag hein, 1975); M. Junkelmann, Panis Militaris: Die Ernährung des römische Soldat im archäologische Experiment, (Mainz, 1997); D. Breeze, 'The Logistics of Agricola's Final Campaign', Talanta 18/19 (1986-7) pp. 264-286; J. Shean, 'Hannibal's Mules: The Logistical Limitations of Hannibal's Army and the Battle of Cannae', Historia 45 (1996), pp. 159-187; J. Peddie, Invasion (London, 1987) & The Roman War Machine (Gloucester, 1994).
2. A. King, 'Military and Civilian Dietary identity in Roman Britain: an update', in A. Goldsworthy & I. Haynes, The Roman army as a community in Peace and War. JRA Supplementary Series (forthcoming, 1999/2000), pp. 139-149.
3. R. S. O. Tomlin, 'Roman manuscripts from Carlisle: the ink-writing tablets', Britannia 29 (1998), pp. 31-84.