Soldiers marriages and recruitment in Upper Germany and Numidia

di David Cherry

The Ancient History Bulletin, 3/6 (1989), pp. 128-130



At least from the time of Claudius to that of Septimius Severus, legionaries and auxiliaries were forbidden to marry during service.1 The prohibition evidently did not apply to officers, at least not to those at or above the rank of centurion.2 And it did not prevent some soldiers from forming stable unions with women, whom they treated as their wives. The evidence is mostly epigraphic � epitaphs on which a soldier commemorates or is commemorated by a woman described as uxor, coniunx, etc., or a soldier described as maritus, coniunx, etc. commemorates or is commemorated by a woman.

In a ground-breaking study published in the Journal of Roman Studies in 1984, R. Saller and B.D. Shaw calculated (among other things) the epigraphically attested rate of marriage in the military populations of Rome and of each of the western provinces for which there is a substantial number of Latin inscriptions.3 Their data revealed wide fluctuation in the rate of marriage from one province to the next, which they attributed mainly to variations in local patterns of recruitment and posting, concluding, for example, that the low rate of marriage attested in the legions and auxilia of Upper Germany is ascribable to less local recruitment and posting, and therefore less frequent commemoration by wives and other family members. Conversely, they took the high rate of marriage attested at Lambaesis as evidence of a greater degree of local recruitment and stationing. Origines recorded on military inscriptions reveal that there was less local recruitment in Upper Germany than in North Africa.4 But the lower rate of marriage attested on the epitaphs of Upper Germany is probably the result of more rigorous enforcement of the ban on soldiers� marriages.

The table shows the number of married soldiers and officers attested on tombstones in Upper Germany and in Numidia:5


Upper Germany






























N1 = Number of men attested
N2 = Number of men attested with wives

The ratio of recorded soldiers to officers is roughly 1:1 in Numidia. It is more than 4:1 in Upper Germany. And because proportionately more officers are attested with wives, if the Numidian ratio is applied to the figures for Upper Germany (i.e., if a similar pattern of commemoration, preservation and discovery is imposed on the data), the overall rate of marriage in the military population of Upper Germany rises to 17.1%, lower than that attested for Numidia (26.1%), but very similar to that which Saller and Shaw calculated for the soldiers, officers and veterans of Rome, Britain, Lower Germany, Noricum and the Pannonias.6 The high rate of marriage attested in the military population of Numidia is mainly a result of the disproportionately large number of officers� epitaphs unearthed at Lambaesis.7

The significant difference between Upper Germany and Numidia is in the rate of marriage recorded for soldiers, 3.0% in Upper Germany and 22.6% in Numidia, not in the rate of marriage attested for officers, 35.6% and 30.6% respectively. And it cannot be maintained that the officers of Upper Germany were locally recruited and posted. Of thirty-one whose origo is recorded, only four were from Germany.8

Variation in enforcement of the ban on soldiers� marriages is explained by the nature and conditions of service at Lambaesis and in Upper Germany. From the end of the first century A.D., when legio III Augusta was transferred to Lambaesis, find-spot of 90% of the Numidian military epitaphs, the base was not threatened and the men posted there were not required to be in a constant state of readiness for combat.9 Such was not the case in the legionary bases along the upper Rhine, notably Mogontiacum (Mainz), a legionary base from the time of Augustus and find-spot of 70% of the military tombstones of Upper Germany. It faced the mouth of the river Main, which constituted one of the three principal invasion-routes for the tribes living beyond the limes.10 The almost incessant danger of invasion in Upper Germany, and perhaps especially at the confluence of the Rhine and the Main, made it imperative that the legionaries stationed there not be encumbered with wives and children.



1     The ban is widely attested, expressly in BGU 114.1. Cf. P. Cattaoui 3 and 4; text in L. Mitteis and U. Wilcken, Grundz�ge und Chrestomathie der Papyruskunde (Leipzig, 1912) II, no. 372. It can be inferred from BGU 140 (Hadrian to Rammius, prefect of Egypt, A.D. 119) that the authorities considered marriage and fatherhood to be inimical to military discipline. The introduction of the prohibition can be taken back at least to A.D. 44: Dio 60.24.3. It may go back to Augustus, perhaps to 13 B.C. when he altered the conditions of military service: see B. Campbell, �The Marriage of Soldiers under the Empire,� JRS 68 (1978) 154, and J.E.G. Whitehorne, �Ovid A.A. 1.101-132, and Soldiers� Marriages,� LCM 4 (1979) 157-158. The ban seems not to have been in effect in the late Republic: Caes. Civ. 3.110.2. It was lifted probably by Septimius Severus in A.D. 197: the evidence is examined in Campbell, JRS (1978) 153-166; cf. P. Garnsey, �Septimius Severus and the Marriage of Soldiers,� CSCA 3 (1970) 45-53.

2     Plin. Ep. 6.31.4-6, Tac. Hist. 1.48 and 4.5; cf. Epict. Ench. 3.22.79.

3     �Tombstones and Roman Family Relations in the Principate: Civilians, Soldiers and Slaves,� JRS 74 (1984) 124-156 (on soldiers: 139-145 and 152-155).

4     J.C. Mann, Legionary Recruitment and Veteran Settlement During the Principate, ed. M.M. Roxan (London, 1983) 12-21, 28-29, 73-79, and 100-106. See also G. Forni, Il reclutamento delle legioni da Augusto a Diocleziano (Milan, 1953) 220-221, 225, 227, and 235, K. Kraft, Zur Rekrutierung der Alen und Kohorten an Rhein und Donau (Bern, 1951) 44-45, and B.D. Shaw, �Soldiers and Society: The Army in Numidia,� Opus 2 (1983) 143-148.

5     The figures are mine. Sources: (a) Upper Germany: CIL 13.5001-7775 and 11468-11980, H. Finke, �Neue Inschriften,� Bericht der R�misch-Germanischen Kommission des Deutschen Arch�ologischen Instituts, 17 (1927) 1-107, H. Nesselhauf, �Neue Inschriften aus dem r�mischen Germanien und den angrenzenden Gebieten,� B.R-G.K, 27 (1937) 51-134, H. Nesselhauf and H. Lieb, �Dritter Nachtrag zu CIL XIII Inschriften aus der germanischen Provinzen und dem Treverergebiet,� B.R-G.K, 40 (1959) 120-229, U. Schillinger-H�fele, �Vierter Nachtrag zu CIL XIII und zweiter Nachtrag zu Fr. Vollmer, Inscriptiones Bavariae Romanae. Inschriften aus dem deutschen Anteil der germanischen Provinzen und des Treverergebietes sowie R�tiens und Noricums,� B.R-G.K, 58 (1977) 447-603; (b) Numidia: CIL 8.1837-8366a and 17585-20206, Inscriptions latines de l�Algerie, Vol. 1, ed. S. Gsell (Paris, 1922), vol. 2, pt. 1, ed. S. Gsell, J. Zeiller, E. Albertini, H.-G. Pflaum (Paris, 1957), Vol. 2, pt. 2, ed. H.-G. Pflaum (Alger, 1976). Soldiers or officers appearing on more than one inscription were counted once.

6     JRS (1984) 152-153 (including �no commemorator� and se vivo � sibi): Rome (equites singulares and others) 14.9% (68 of 455), Britain 17.6% (29 of 165), Lower Germany 15.5% (23 of 148), Noricum 17.2% (29 of 169), the Pannonias 19.7% (65 of 330).

7     Lambaesis: 197 soldiers, 164 officers (sources: note 5). The ratio of soldiers to officers in a legion of 6,000 was in the order of 8:1 (a first cohort with five double centuries, nine regular cohorts, 120 horsemen). Other western provinces: Britain, 82 soldiers: 41 officers (The Roman Inscriptions of Britain, ed. R.G. Collingwood and R.P. Wright, 1965); Lower Germany, 40:20 (CIL); Noricum, 56:24 (CIL); the Pannonias, 120:83 (CIL); Spain, 33:42 (CIL).

8     Officers (31): Italy 10 (32.3%), Narbonensis and Spain 1 (3.2%), Gaul 5 (16.1%), Upper and Lower Germany 4 (12.9%), elsewhere 11 (35.5%). Soldiers (174): Italy 97 (55.7%), Narbonensis and Spain 43 (24.7%), Gaul 3 (1.7%), Upper and Lower Germany 13 (7.4%), elsewhere 18 (10.3%). Source: Mann, Legionary Recruitment, Table 13, 100-106, to which add CIL 13.5206 (Julia Dertona), 6212 (Arretium), 6236 (Termes), 6449 (Sicca Veneria), 6620 (Saldae), 6621 (Ostia), 6658 (Berytus), 6952 (Aequum), 11782 (Colonia Aelia Augusta Mercurialis Thaenitanorum), 11801 (Antioch), 11811 (Arethusa Suriae), 11837 (Venafrum). In the period from A.D. 117 to c. 300, which witnessed a sharp increase in local recruitment to the German legions and auxilia, less than one-third of the officers whose origo is recorded (3 of 11) were from Germany (10 of 21 soldiers).

9     See M. Janon, �Paysans et soldats,� L�Afrique romaine: les conf�rences Vanier 1980, ed. C.M. Wells (Ottawa, 1982) 51-67 (especially 56ff.), and Shaw Opus (1983) 133-160.

10     C.M. Wells, The German Policy of Augustus (Oxford, 1972) 96-7: from the Lippe route on the Upper Weser southwards through the Wetterau to the lower Main or westwards along the Main and its tributaries (description and plan of Mogontiacum, 138-146). See also D. Baatz, Neue Untersuchungen am r�mischen Legionslager in Mainz-Mogontiacum (Berlin, 1962) 68-73. 1

Archive index