Skip to main content

Amidst the standard ethnographic and epic topoi employed in his catalogue of Scythian enemies (Argonautica 6), Valerius Flaccus provides some unexpected contemporary details on Sarmatian arms and tactics (6.162, 231-238). This paper considers how Romans processed new information about barbarian peoples within a conceptual system where established stereotypes and traditional topoi held primacy of place.

Isaac (2004) has convincingly demonstrated that the Romans, building on Greek traditions, tended to view foreign peoples through the lens of stable ethnographic stereotypes, frequently connected to geographic location. These tropes had great staying-power because they described foreign customs as distortions or inversions of Greco-Roman cultural norms (Hartog 1980). Ethnographic topoi permeated every genre of literature and were thus deeply inculcated in the minds of Rome's political decision makers, educated as they were in the ancient liberal arts (Mattern 1999). Lee (1993) has shown that military intelligence-gathering did occur, albeit in a haphazard manner, but what room was there in the broader Roman ethnographic system for such new information, once acquired?

I argue that new facts could be included in literary descriptions of barbarian peoples, but that fresh knowledge rarely displaced older stereotypes. Rather, the new was often simply presented alongside the old creating potentially paradoxical descriptions. Understanding that Roman popular perceptions of barbarian peoples operated on an agglutinative principal can help us decipher contradictory descriptions when we encounter them.

Together, Valerius Flaccus and Ammianus Marcellinus provide a good example of this process in action. Most of the tribes included in Valerius Flaccus' Scythian catalogue originate from Herodotus and are described using Herodotean tropes about nomadic behavior and warfare. Not so the Sarmatians, who, unlike their Herodotean selves, are described here as armored lancers (6.231-238). Syme's point was well made: these lancers reflect the real Sarmatians Rome was then dealing with in the plains beyond the Danube (1929). Valerius presumably included this modern information in order to connect his narrative to a topic of contemporary interest. With Vespasian warring against the Roxolani in 69-70 CE, and Domitian facing down the Iazyges during his Dacian campaigns in 89 and 92, Sarmatians would have been on people's minds at Rome, wherever we may date the Argonautica.

Looking ahead almost three centuries to Ammianus Marcellinus, we find a more complex treatment of the Sarmatian peoples. Here we can identify three contradictory characterizations. First, the Alani, a Sarmatian tribe, are described by Ammianus using details lifted directly from Herodotus (31.2.13-25). Elsewhere, the Sarmatians of the Hungarian Plain are introduced as armored lancers famed as plunderers because of their great mobility (17.12.2-3). Were these two passages all we had to account for, interpretation would be easy: like Valerius, Ammianus has used contemporary data in one place and relied on established tropes to describe the distant Alani. There are, however, other Sarmatian scenes. Whenever Ammianus actually describes Sarmatians engaged in battle (17.13.8-15; 19.11.13-16), they appear mainly as infantry, fighting to defend settled villages, not nomad horsemen in the vein of either Herodotus or Valerius Flaccus. Ammianus, a career military staff-officer, has been judged largely reliable in his presentation of battles and tactics (Austin 1979), so we cannot simply dismiss these descriptions as military commonplaces.

Ammianus, then, presents a highly paradoxical picture of Sarmatian warfare and society. The Alani are described as bow-wielding nomads, while the more westerly Sarmatians are first identified as rapacious lancers only to be shown in battle as common barbarian footsoldiers. This three-part characterization is telling. The battle narratives show that by the later 4th century, Valerius' armored, lance-wielding Sarmatians - once the product of contemporary knowledge - no longer reflected reality. The fact that they still make an appearance in Ammianus' Sarmatian introduction shows how ethnographic stereotypes could snowball over time. Valerius' description became canonical, to be repeated with the older tropes by successive generations even as updated information occasionally revealed the lie; careful reading of other Roman ethnographic material will certainly reveal further examples.