Skip to main content

A remarkable number of Latin and Greek inscriptions from the first three centuries CE, some 915, mention slaves and liberti alongside Roman soldiers and veterans. About eight percent (80 texts) mention slaves, while the rest relate to liberti (835 texts). This starkly contrasts the literary evidence, which speaks of slaves rather than liberti in the company of soldiers. Close analysis of some especially detailed epitaphs suggests that the disproportionately large number of liberti epigraphically attested is tied to the distorting effect of testamentary manumission among soldiers and veterans. Though testamentary manumission likely also underlies many civilian inscriptions, it is of particular interest in military contexts in light of the special legal privileges granted to Roman soldiers.

Scholarship on slavery in the Roman military is sparse, and relevant inscriptions are particularly unexplored in a focus on the literary evidence (Rouland 1977, Welwei 1988) or in selective presentation of a few exemplary texts (Bennett 2007, Kampen 2013, Speidel 1989). Even Saller and Shaw (1984) do not comment on “military” slaves and liberti in their landmark study of family relations in the epitaphs of civilian, military, and servile populations, despite including 110 such texts (8 slaves, 102 liberti) in a ratio almost identical to my own findings. In contrast, Josephus, Tacitus, and other authors almost always use terms denoting slave status when mentioning servants alongside Rome’s forces (calones, mancipia, δοῦλοι, θεράποντες, and οἰκέται; rarely, if ever, liberti, or ἀπελεύθεροι).

The discrepancy between the two types of sources may be explained by a widespread use of testamentary manumission among soldiers and veterans. Such grants of freedom are explicit in four inscriptions in my database, CIL 6.32881, AE 1961, 17, CIL 13.8293, and ILS 8269. In each the grant of freedom is linked to the death of the soldier or veteran and the terms of his will. For example, in CIL 6.32881 (86 CE), L. Vafrius Epaphroditus is commemorated by Helius, slave of another soldier, as manumissus testamento (l. 3), “manumitted through the will” of L. Vafrius Tiro, centurio of legio XXII Primigenia. The slave Irolis in AE 1961, 17 (first c. CE) was freed in order to take care of the tomb of his master, decurio of an auxiliary unit (ex testamento ... manumiserunt servom Irolem uti praestus sit eo sepulchro, ll. 5–8). In addition to such explicit references to testamentary manumission, a further 9% (84) of my database’s texts include the words (ex) testamento, indicating a will, and a further 17% (156) mention liberti as heredes, heirs, also raising the possibility of a will.

Postulating a widespread use of testamentary manumission among soldiers also meshes well with our legal sources. Soldiers’ wills (testamenta militaria) stood in a privileged category of their own (Ulp., Dig. 29.1.1 pr). While civilian wills had to follow strict formalities to be valid, a much more lenient stance was taken towards testation in military contexts, both in terms of the form and the required qualifications of an heir (Champlin 1991, 56-58). Soldiers and veterans held special rights which allowed them to dispose freely, sui iuris, of assets, including slaves, acquired during military service (peculium castrense) in their wills (Inst. Iust. 2.12 pr). This would have made the manumissio testamento an exceptionally attractive and easy method of freeing slaves for soldiers and veterans.

The popularity of testamentary manumission among soldiers and veterans apparently explains why liberti so heavily outnumber slaves in inscriptions from military populations. This not only reconciles our conflicting literary and epigraphic record, but also is consistent with the special rights for soldiers and veterans described in our legal sources. The investigation thus throws into relief interesting commonalities and distinctions of Roman soldiers and civilian populations.