As already stated, typically auxiliary troops were recruited from peregrine (free noncitizen) peoples, although some recruitment of men from beyond imperial boundaries was a recurring feature of Roman armies throughout history. Units were raised in all parts of the empire, as unit titles on military diplomas show, but certain traditionally warlike (and less Romanized) provinces such as Gallia Belgica and Lugdunensis and Thrace were of particular importance early on. The degree to which units might continue to recruit from their native province seems to have varied over time and from region to region of the empire, but within a few decades of moving from their place of origin most existing units seem to have raised replacement recruits from the province in which they were based and from other nearby provinces (for a general summary see Haynes 1999: 166, and Holder 1982: 52-3 on this process in Britain).
While this distinction between recruitment of citizens into legions and peregrines into auxilia served to emphasize and perpetuate existing legal and social distinctions within the empire, army recruitment was also a mechanism for social mobility. The diplomas of auxiliary veterans show that they received Roman citizenship on retirement. Moreover, citizenship was also granted to their wives and children. Ordinary serving Roman soldiers could not contract legally valid marriages from at least the reign of Claudius (Dio 60.24.3) to that of Septimius Severus (Herod. 3.8.4; see Campbell 1978 and Phang 2001: 115-33). However, the formula on the diplomas makes it clear that military authorities accepted that informal relationships did exist and that they recognized them on soldiers’ retirement.
The degree to which soldiers formed family relationships with local populations may have varied a great deal from province to province and even unit to unit (see, for example, Alston 1995: 54-6, 58-9; Pollard 2000: 152-9). At Dura-Europos in Syria, for example, one document (P. Dura 32 [254 ce], hence after the Constitutio Antoniniana) shows intermarriage of a legionary and of a local civilian, suggesting a degree of integration between soldier and civilian population. Conversely another (P. Dura 30) shows a woman who is probably the sister and widow of soldiers marrying another auxiliary soldier in his unit’s winter quarters, witnessed by soldiers. Such ‘‘institutional endogamy,’’ within the confines of families already associated with the army, may have been particularly important where castris recruitment was common and probably tended to separate the army from local civilians.
World History
