The gradual changes in the nature of the Roman army between the time of Tiberius and 235 certainly affected both Roman society and the empire’s internal power-struggles. Did they also affect Rome’s strength at the periphery? The really big changes were three, though they had all started well before Tiberius’ accession. The first was the regular organization of ‘auxiliary’ troops into quasi-permanent units in which they would normally serve for twenty-five years before being made Roman citizens on discharge, a system set up by Augustus and refined by his successors. As in centuries past, such troops often outnumbered the legionaries, and their effectiveness was of profound importance.
That leads, secondly, to the matter of recruitment. In the era of the civil wars of 49 to 31 bc, and under Augustus, provincials had entered the legions in large numbers. Recruits came from Roman-colonial or Romanized communities, but also from others: thus an inscription of the early Principate (ILS 2483) shows that almost all the soldiers in the two legions stationed in Egypt had been recruited in non-citizen communities in the eastern provinces (their lingua franca was Greek). All over the empire, the more Romanized provinces provided more and more of the legionaries, while Italians – who made up the bulk of the better-paid praetorian guard – provided fewer and fewer. The authorities were now quite willing in practice to recruit non-citizens, giving them citizenship when they were sworn in. This ‘provincialization’ probably reflected some Roman/Italian reluctance to serve (Italy was too prosperous) but also some intention on the emperors’ part to bring provincials into the mainstream. From Hadrian’s reign on, the normal pattern (though not in Britain) was to recruit legionaries in the provinces where they were needed, but from relatively Romanized/Hellenized elements (and legionaries were more likely than ‘auxiliaries’ to be literate). This was by and large a well-organized and disciplined force; and fighting spirit was probably not lacking either, at least down to Trajan’s time – when battle-commanders chose to entrust the initial impact of the fighting to ‘auxiliary’ units and keep the legionaries in reserve, a procedure that is first attested in a major battle at Idastaviso in Germany in AD 16 (Tacitus, Annals 2.16.3), there could be a variety of tactical reasons.
‘Auxiliary’ recruitment was quite different: the government concentrated on fringe areas such as Iberian Galicia and Thrace, simply supplying officers from the core area of the empire; such units were commonly posted away from their home areas, Britons for example in Upper Germany, while the auxilia in Britain itself might, for example, be Batavian or Syrian. Eventually, but unfortunately we do not know when, Rome also began to employ soldiers who are unlikely to have felt themselves to be Roman subjects: Marcus sent 5,500 cavalry of the Transdanubian Iazyges, whom he had just subdued, to serve in Britain (Dio 71.16). There were Goths garrisoning Arabia in 208, and Goths later took part in Valerian’s war against the Persians. This was probably an increasing trend, but it is hard to tell how much the armies of, say, Constantine and Licinius were really dependent on Goths or Arabs, whom they are known to have made use of.
The other military change of potentially great importance in the period prior to 235 was not so much that many units in the Roman army became ‘sedentary’ from generation to generation, becoming deeply involved in essentially administrative duties, but that many Roman soldiers never experienced battle. This army had never been invincible, but its deplorable failure to protect the Danube frontier in 170–1 suggests significant changes for the worse. Enemy forces reached northern Italy for the first time in some 270 years, while others, as already mentioned, raided as far south as Attica. Our sources on all this are poor, but it may be conjectured that a shortage of officers and soldiers seasoned by warfare had a great deal to do with Rome’s failure, and this in turn was the indirect result of conscious policy. In other respects, the Romans were normally at an advantage: throughout this period they were superior to their opponents in important areas such as artillery and engineering (‘the soldiers are always practising bridge-building’, Dio 71.3).
Temporary causes admittedly contributed, and the Danube line still had a long future. Marcus Aurelius, as we have seen, had had to raise two new legions about 165 to replace the three which his co-ruler Verus had taken from the Rhine and Danube to the east in order to fight the Parthians. Shortly thereafter, the Roman military in the north suffered seriously from the Great Pestilence, as recent studies have demonstrated. Marcus himself had had no military or even provincial experience before 168 – and it showed. Imperial coin-types furthermore had often exaggerated the emperors’ military achievements, and there was a risky deception involved when coin-types absurdly declared in 172–4 ‘Germania subacta’ – ‘Germany has been vanquished’.
Few historians have really tried to evaluate the Severan army, and the evidence is slippery. Even republican armies sometimes mutinied, and there were whole rhetorical topoi about undisciplined soldiery. But an army stationed in Mesopotamia that was mutinous enough to assassinate the provincial governor (about 227, Dio 80.4.2) was a very negative symptom (and see below on the year 235).
We have quite a lot of information about how the Roman army changed between Severan times and Constantine, but assessing its ability to do its job is nonetheless difficult. On the one hand it never, unlike the republican army, won battles it might well have lost, on the other it never, unlike the late-antique Roman army, lost battles that it ought to have won. We have little option but to judge it by its results, though these may be mainly attributable not to its own qualities but to those of its generals, or its logistics, or its enemies, or to any combination of these factors. Recent accounts of Rome’s military performance in this 100-year period are unsatisfactory, but our sources are admittedly tenuous to a degree, whether it is for the defeat at Abrittus in 251 or the battle nine years later in which, or after which, the Persians captured the emperor Valerian (some Roman sources naturally preferred to claim that he was captured by trickery).
Tiberius already knew that it was worth keeping two legions in Dalmatia partly in order to back up the legions on the Danube (Tacitus, Annals 4.5). Later Roman emperors eventually concluded that the long-standing dispositions of the Roman army, with the great majority of the soldiers stationed on or near the frontiers, were ill adapted to resisting major invasions that might come from different directions. It had always been necessary to balance the needs of the Danube frontier and the Euphrates frontier, but both became more dangerous in late-Severan times. Once Rome surrendered the initiative, the distances involved presented an almost insoluble problem: it took something over two months, for example, for troops to travel from Rome to Cologne. The best that could be done was to create a reserve army that could be sent wherever it was needed without weakening some vital garrison. It appears to have been Gallienus who created a central cavalry force (cf. Zosimus, New History 1.40, Cedrenus, i, p. 454 Bekker). The development of these comitatenses, as they came to be called, is impossible to follow in any detail, but Constantine apparently expanded their role (Zosimus 2.21.1 may refer to such troops), while also centralizing the command structure of the army by means of an overall infantry commander (the magister peditum) and a parallel cavalry commander (the magister equitum). Nonetheless it remained difficult to counter any large invasion once it had passed the northern or eastern frontiers. An enterprising governor might raise a local militia (populares: AÉ 1993 no. 1231b shows us a governor of Raetia doing this in 260), but they would be largely untrained and untried.
The reliefs on the Arch of Constantine distinguish between his Roman and his ‘barbarian’ troops, which raises again the complex question of whether Rome was now relying too much on troops who were merely mercenaries. According to the emperor Julian (Caesars 329a), Constantine ‘practically paid tribute’ to the barbarians, and modern accounts suppose that he and his rival Licinius made Rome significantly more reliant on German and other non-Roman troops than any previous ruler; but the ill effects do not yet seem to be visible.
The strength of the Roman Empire’s numerous and various neighbours to the north, east, and south can only be judged, once again, by the results, their aims likewise. From Tiberius’ time to Trajan’s, those who kept their freedom from Rome and their territorial integrity were doing well; this applies mainly to the Romans’ failure to advance far beyond the Rhine and to hold on to Mesopotamia. The incursions of the 160s–70s and of the 240s–60s showed a great deal of vigour. The invaders’ goal was often plunder, including human beings, which the Roman Empire offered in abundance. Dio (71.16) asserts that the Iazyges had taken far more than ‘ten myriads’ of prisoners in Roman territory – a five- rather than a six-digit number, one might think. (Some of the third-century booty has been recovered from the bed of the Rhine, rafts having apparently sunk). Not even Sasanian Persia, the most powerful external enemy Rome faced in this period, showed any determination to hold on to any Roman province, and in fact it had no reliable means of protecting its own core area against Roman forces that were always relatively near. But northern peoples had already in the second century extracted territorial concessions of a sort, obtaining lands within the Roman frontier. This practice went back to Julio-Claudian times. Initially the advantages to Rome probably outweighed the disadvantages; whether that continued to hold true in and after Marcus Aurelius’ time we shall consider in a moment. It certainly looks like a major surrender to strong outside pres- sure. Purchasing the docility of outside enemies by means of payments, unless it was a short-lived tactical expedient, was likewise a recognition of real enemy strength: this started with Domitian, but involved Trajan, Hadrian, and many later emperors. Yet from a Roman point of view, this was by no means an irrational policy, within limits.
Fundamental changes had taken place by the time the conglomeration of Germans known as the Alamanni (‘All Men’), who are first attested in a Roman source in 213, inflicted quite serious harm in 232–3. This was nothing less perhaps than the birth of a new national formation. What made a difference here was probably in the end quite simple: such a new grouping, like the Franks from about 260, could put larger forces into the field than any single German people. But the tetrarchs and Constantine could always, it seems, defeat the northern peoples on the battlefield.