Roman Ideology on Warfare: The Background to Recruitment III: Grand Strategy and Defence-in-Depth

The incompleteness of the archaeological record makes strong conclusions about military dispositions and trends imprecise, leading to much discussion on Late Roman frontier organisation. In its simplest terms, the argument revolves around whether Diocletian and Constantine intentionally replaced “reactionary defence inside Roman territory [with] occasional aggressive pre-emptive defence across the frontiers,” (Southern and Dixon (2001), 276; Pan. Lat. VIII[5].3 comments on the Tetrarchs having “carried forward the boundaries of Roman power by their courage”) or did the frontier system of the fourth century arise by default. So varied is the argument and so inconclusive is the source material that it is possible to argue that Diocletian re-established a more linear preclusive strategy. However, the retention of strategic troop concentrations along with a more comprehensive system of roads suggests that Rome had not strayed far from her offensive stance (Many of Rome’s fourth century frontiers retained a permeability that was required for “connivance, commercial interchange or political control” (Arce in Pohl, Wood and Reimitz (2001), 8). This is perhaps best seen in the east (Isaac (1992), 254-55). Even the famous strata Diocletiana which, on the surface, looks like a fortified line was in reality a series of roads and installations connecting legionary bases (Isaac (1992), 164-65). The large concentration of troops in Cappadocia suggests that strategic designs could differ greatly depending on numerous geographical, topographical and social considerations (Isaac (1992), 171). The building of strongholds on the barbarian territory further supports this (Wolfram (1997), 63; Eutropius, Brev. III.10 on the Rhine; CIL XIII.8502 on the building of Deutz for the “subjection and control of the Franks;” Wolfram (1988), 61 on the Danube; John Lydus, de Mag. II.11 on building fortifications from Egypt to the Persian frontier, although John Malalas, Chron. XII claims that this was done exclusively by Diocletian; Johnson (1983), 166; Petrikovits (1971), 182-87, 207-18).

Another problem with defining the reforms of Diocletian and Constantine is the tendency to attribute modern notions of grand strategy and defence-in-depth. These require an opponent who has an overtly offensive and co-ordinated policy and the means to carry it out (Isaac (1992), 374). While the Romans could respect the martial prowess of a German tribe or Persian king, they never feared anyone enough to institute a frontier system based on defence. This seems particularly peculiar in the modern day. That many of the Empire’s fringes were established along geographical features such as rivers, mountains, deserts and seas makes it more difficult to see these frontiers as zones instead of linear constructs. However, to the Romans it was normal and throughout their history they never fully developed the concept of frontiers.

There were de facto boundaries, but the differences between frontiers of provincial administration, military control and political influence are difficult for modern historians to define, while Roman contemporaries were never inclined to do so (Whittaker (1994), 195, 199-200). The numerous connotations of the term limes further demonstrate these uncertainties. Its modern derivative ‘limit’ bears little resemblance to what it meant to the Romans. Originally a surveyor’s term, it was adopted to mean “a military road linking camps on the fringes of empire, often lined with watchtowers and staging posts” (Whittaker (1994), 201). These links to the frontier zones gradually invested limes with some idea of a boundary. By the fourth century, it came to mean a frontier district under the control of a dux. However, “in no single case is a limes described as something made or constructed” (Isaac (1988), 146). The picture is further blurred by the term ripa. It too took on notions of a frontier and was increasingly used in conjunction with the Rhine, Danube and Euphrates at the edges of the Empire. However, it is likely that the term meant a border region rather than the rivers themselves, suggesting that the rivers were not considered a defensive barrier (Whittaker (1994), 201). This may be borne out by the continued use of both the Rhine and Danube as lines of communication (Ammianus XVIII.2.3; Libanius Or. XVIII.83; Zosimus II.10.2) and to use rivers in this way required control of both banks (Ammianus XXVIII.2.1). Perhaps this need for political control beyond the administrative boundary explains why limes came to mean “a region within which military buildings were constructed both in advance of and behind the line of administered frontiers” (Whittaker (1994), 201).

Therefore, the use of terms like ‘grand strategy’ and ‘defence-in-depth’ suggests that the Romans could realise in practice what they could not define verbally. A grand strategy perhaps requires a co-operation between economic and manpower resources that was beyond Roman capabilities. Furthermore, it would have been unwise to impose a strict set of frontier rules on such a vast area of geographical, topographical and demographical varieties. And being an absolute monarchy would not have helped. Whereas a grand strategy can take years to come to fruition, Roman military strategy could change overnight with the accession of a new Augustus. It may be unrealistic to expect any pre-industrial state to achieve a coherent grand strategy (Whittaker (2004), 28-49) and without it frontier policy was left in the hands of individual emperors. Most changes were made in reaction to trouble rather than innovation on a large scale. The “prudence and tactical common sense” (Whittaker (1994), 209) of men like Constantius II and Valentinian I is confused as defensive strategy, but in reality any frontier system remained ad hoc and regional (Ferrill (1986), 45). 

It is also difficult to accept defence-in-depth, allegedly a “combination of self-contained strongholds with mobile forces deployed between or behind them,” (Luttwak (1976), 131) as an overall system. While Diocletian strengthened the frontiers, there is little evidence “for the stationing of units in the hinterland” (Mann (1979), 180; perhaps only in the east is there any evidence of fortified positions removed from the frontiers), which would seem to rule out defence-in-depth. The contention that Constantine employed such a system is also uncertain. Constantine had only one field army, and therefore the formation of comitatenses cannot be seen as part of a system of defence-in-depth. Only the separation of this central field army into three regional groups by his sons can give rise to a suggestion of defence-in-depth (Whittaker (1994), 207 on the splitting of the field army being a political decision, not a strategic one). The Notitia Dignitatum suggests that over 50% of the army was still positioned near the frontiers by the fifth century (Whittaker (1994), 207).

The use of forts, walls and cities may constitute a “calculated policy of nodal defences” (Whittaker (1994), 207), akin to defence-in-depth. During the Principate, forts could not be expected to withstand a concerted attack due the absence of strong defences and poor location (Isaac (1992), 186; Tacitus, Hist. IV.23 claiming that the lack of quality in fort building was explained by the idea that “our vigour and our arms seemed enough”). Under Diocletian forts better designed for defence were introduced. Rectangular constructs were replaced with square forts with wider berms, thick walls with ramparts, projecting corner and interval towers and fortified gates, placed on sites for tactical dominance. Walls were also built but most served as internal tactical controls (CIL III.6151 suggests that the Tetrarchs “established an eternal guardpost” at Transmarisca on the Danube), such as around Constantinople (Harrison (1969)). and the claustra Alpium Juliarum controlling traffic between Illyricum and Italy (Christie (1991)). However, neither of these was built to keep out foreigners. Due to the events of the third and fourth centuries, these and other walls were more likely built as tactical battle defences against rival political pretenders (Whittaker (1994), 197). “It does not appear that the Romans ever seriously contemplated walls as a means of frontier defence” (Whittaker (1994), 197). The use of cities in a frontier system was also not a new development (Isaac (1992), 254-55). By the late third century some cities had acquired more defensive features and a detachment of troops (Johnson (1983), 114-15, 189, 251).

However, despite some regional similarities, Roman defences show “a studied lack of standardisation” (Johnson (1980), 92-98). and “an indifferent understanding of the tactics involved.” (Wacher (1974), 77-78). This suggests that there was “no central directive covering fortification in the north-western empire.” (Petrikovits (1971), 205). The stationing of men in these cities was arguably in order to keep them better supplied (Ammianus XVI.4.1; XXXI.16.8; XVI.11.4). Therefore, these changes emerged out of the necessity of individual incidents rather than an overall strategic upgrade of all fortifications. Consequently, either the Roman government of the late fourth century was incapable or unwilling of mounting a uniform defence system or was happy to react to problems instead of trying to pre-empt them. Therefore, to speak specifically of defence-in-depth during the fourth century is going beyond the evidence (Isaac (1992), 170 highlights that no contemporary source suggests that a strategy of defence-in-depth was used; Mann (1979), 180 suggests that -in-depth could be used by the Romans when circumstances like financial or manpower dictated). The very idea of defence-in-depth suggests that the Romans had changed their traditional psyche by thinking defensively at all. Defence-in-depth, by its very nature is inherently negative and in conflict with the Roman ideology that they ruled the world either directly or through political influence (Whittaker (1994), 196). The Rhine, Danube and Euphrates were only temporary impediments to the Roman war machine as such a limitless empire “almost prevented the Romans from admitting that a frontier ever existed.” (Whittaker (1994), 198; the Romans had numerous terms to describe the edge of their Empire – finis, terminus, modus and meta in Latin and κρασπεδον, όρος and έσχατια in Greek – but all had indefinite meanings (Elton in Mathisen and Sivan (1996), 126); Caesar, BG IV.16; Velleius II.101; Josephus, BJ II.16.4; Tacitus, Ann. I.9, II.58; Strabo XVI.1.28; Appian praef.4; Herodian IV.10.2; Ammianus XV.10.2 all describe rivers, mountains, deserts and even man-made features as the extent of Roman authority).