1 The military transformation of the Roman world Walter Pohl [Accepted manuscript of the chapter published in: In tempore Sueborum. El tiempo de los Suevos en la Gallaecia (411-585): el primer reino medieval de occidente. Exhibition catalogue (Deputaciã³n Provincial De Ourense 2018); ISBN: 978-8416643110]. The Roman army and the fragmentation of the Roman Empire As many ancient empires, Rome rose as a city state that gradually expanded its power over its hinterland and over many other cities. Rome’s success in maintaining its empire for centuries, however, was not least due to the way in which it co-opted warbands and groups of soldiers of widely different origin. The late Roman Notitia dignitatum, a list of early-fifth century units of the Roman army, preserves this exceptional multiplicity. Many names represent the ethnic or territorial origin of the respective units, usually (and for the empire, quite conveniently) serving far from their places of origin: Sarmatians in Britain or Franks in Egypt. Even if they came from recently- or non-conquered regions, these soldiers enjoyed freedom and privileges, and their leaders had splendid career opportunities (HOFFMANN 1969; DEMANDT 1980; ISAAC 1990; SOUTHERN/ DIXON 1996; DEMANDT 2007: 303-24). Late Roman Emperors, who often had been successful generals before rising to power, could come from all parts of the Empire, from Spain to Syria. We may see the endless internal conflicts between emperors and usurpers in Late Antiquity as a sign of decline and weakness. That may not be wrong; but it also meant that power struggles revolved around the imperial throne, and therefore remained within the Roman system – no alternative power structure could emerge. The successful integration of warriors and warbands into an overarching imperial system only reached its limits when several developments roughly coincided (In general, see POHL ²2005; WOLFRAM 2005; HEATHER 2006; HALSALL 2007; WICKHAM 2006; POHL 2008). The tax by which Roman citizens could buy themselves off from being drafted into the army, the aurum tironicum, was reformed and generalized in 375 (LENSKI 2002: 313-14; ROTH 2016: 38-39). This led to a growing demand for ‘barbarian’ soldiers, and increasingly, whole bands of warriors under their own commanders were drawn in (O’FLYNN 1983; MAC GEORGE 2002). The bitter throne conflicts in the empire were increasingly decided by barbarian contingents: for instance, one of the biggest battles of the period was the battle at the