Rome's Gothic Wars_ From the Third Century to Alaric - Michael Kulikowski [43]
The same holds true for the terms of the treaty. Fourth-century evidence is limited, while Jordanes imposes on it an anachronistic Byzantine interpretation. He supposes that Ariaric’s Goths became foederati, a word that by the sixth century had a technical legal content implying specific responsibilities on the part of both empire and federate allies. In 332, however, the formal status of foederatus did not exist, and the word for treaty, foedus, is not a technical term. Even though many scholars think that the treaty of 332 invented the type of technical foedus known in the sixth century, nothing in the fourth-century evidence makes that plausible. The peace of 332 marks a significant stage in both Roman and Gothic history not because of any legal innovations, but because it was so very decisive. It imposed more than thirty years of peace on the lower Danube or, as bishop Eusebius of Caesarea put it in the Life of Constantine that he wrote shortly after the emperor’s death in 337, ‘the Goths finally learned to serve the Romans’.[74] Indeed, some of the defeated Goths would continue to claim a special loyalty to the Constantinian dynasty for many years, decades later supporting a usurper named Procopius on the grounds of his dynastic connections.[75] In the interim, they offered tribute to the emperor, and provided a large supply of military recruits for the Roman army. Such military service was not explicitly required by the terms of 332, as Eusebius’ testimony makes clear: he is nowhere able to state that Goths served in the army as a result of the treaty, even though elsewhere in his Life he is consistently very enthusiastic, and very specific, about Constantine’s recruitment of defeated barbarians.[76] Regardless, the peace brought benefits to both sides.
The Peace and the Gothic Economy
The frontier was opened to trade all along its length, a most unusual measure, given that Roman emperors had for centuries regulated the export of Roman technology outside the empire. Yet the fact that trade surged all along the river is demonstrated by the large number of bronze coins found in the band of territory north of the Danube. Bronze issues of the late 330s to the early 360s dominate the archaeological record, which suggests that the Gothic side of the lower Danube came to be quite thoroughly integrated into the Roman monetary economy in those years. In fact, the distribution of bronze coins in the region immediately beyond the frontier is very nearly as intense as in the Roman province of Scythia itself.[77] That such coins were used for commercial exchange is placed beyond serious doubt by the existence of locally produced imitations of Roman coins which must have been struck to eke out insufficient supplies of genuine Roman coinage in commercial circulation. It must be noted that bronze coin finds are dramatically concentrated right beside the frontier, generally within fifteen or twenty miles of it, but less so in the Gothic regions opposite Scythia and Moesia Secunda than those across the river from Moesia Prima. Although this fact has led some scholars to question the level of monetization of the Gothic economy, the sheer quantity of low-value coinage beyond the frontier make these objections hard to sustain.
That Roman diplomatic connections with the Gothic elite also increased rapidly from the 330s onwards is suggested by the distribution of Roman silver coins. Much less common in the immediate vicinity of the Danube, silver is instead found in large quantity further north and east, in modern-day Moldova and Ukraine. Unlike the bronze, silver coinage is uncommon in stray finds at industrial and residential sites. Instead, silver siliquae are concentrated in small hoards, for instance one found at Kholmskoě near Lake Kitaj or another at Taraclia in Moldova. The Kholmskoě hoard is especially significant: its ninety-three silver coins of Constantius Ⅱ were all of the same value and type, struck between 351 and 355, bearing the legend VOTIS.ⅩⅩⅩ – MULTIS.ⅩⅩⅩⅩ, and virtually unused. This fact raises