These recent studies on ethnography in the classical world have sought to analyze the ideological nature and alienating function of ethnological texts, rather than to assess the historicity of their data, as a previous generation might have done (and with value: Momigliano 1971). By contrast, a major sector of research in Late Antiquity over the past two decades has undertaken a different project, seeking to deploy early medieval and Byzantine texts in order to reconstruct the self-identity and social dynamics of ‘‘barbarian’’ groups such as the Goths, Franks, Lombards, and others that bulk large in the history of Late Antiquity. This research continues the trajectory of one of Late Antiquity’s tributary disciplines, the study of Germanic antiquity. Current approaches in this field are associated with the term ‘‘ethnogenesis,’’ a word borrowed from mid-twentieth-century social sciences, referring to processes by which ethnic groups form (Wolfram 1981, 1988, 1994; Geary 1988,1999, 2002; Pohl 1991, 1994). For a period in which ‘‘barbarians’’ take center stage - with ‘‘barbarian invasions’’ and ‘‘barbarian kingdoms’’ in ‘‘the barbarian west’’ - investigation into the self-identity of barbarian groups seems desirable: an opportunity to see nonclassical peoples in their own terms, and to extend to them the sort of examination previously undertaken for the literate cultures of the Mediterranean. This investigation features intellectual tools similar to those that configure the studies of the classical world outlined above. Group identity is analyzed as a construct manipulated by social elites in order to generate social and political cohesion, and identity is seen as oppositional, constantly defined and asserted against other competing identifiers. Texts are understood to serve ideological purposes, as communications of ideologies around which group identity can cohere (Pohl 1998; 1999).
Nevertheless, this project is not the same as that of the classical research discussed above. Greek and Roman constructs of self and Other can be examined through Greek and Roman sources. To identify the dynamics of self-identification among the barbarians of Late Antiquity, we also have Greek and Roman texts, only - not ‘‘barbarian’’ ones. We can look through Greco-Roman eyes to see how they perceived the Other and, therefore, themselves; we cannot do the same for the barbarians of Late Antiquity. Late antique ethnographic discourses, even when describing the barbarians who gained control of sections of the Roman Empire, remain Greco-Roman. There are, of course, differences between the ethnographic genres of earlier antiquity and the major relevant texts of Late Antiquity: primarily Christianity, which provided an additional register of conceptions of ‘‘ethnic’’ identity, with notions of divinely favored (and disfavored) peoples, an ‘‘ethnic’’ history stretching over millennia with periods of exile and wandering alternating with military and political glory, and extensive genealogies of royal houses. The deployment in recent research of late antique, Christian, Greco-Roman sources in order to reconstruct ‘‘barbarian’’ identities is problematic. These approaches do not simply trawl late antique texts in order to gather isolated fragments of data that report on barbarian groups, but see certain Latin works themselves as evidence for communications of identity-forming processes within those groups; Roman writings are appropriated as, essentially, ideological discourses of ‘‘Gothic’’ or other non-Mediterranean provenance, albeit preserved under strata of Judaeo-Christian and classical topoi.
Central to the ‘‘ethnogenesis’’ approach that seeks to reconstruct ‘‘barbarian’’ dynamics of identity is the text of Jordanes’ Getica. The Getica is a short narrative written in Constantinople in the 550 s, during the latter days of Justinian’s war in Italy. It outlines a pseudo-history of the Goths over the previous 2,000 years, a history dominated by epic migration and royal dynasties. In ‘‘ethnogenesis’’ interpretations, the Getica is treated as a recension of genuine Gothic self-understanding, transmitted through one of several sources, known or putative: either the lost Gothic History of the Italian Cassiodorus, former bureaucrat to the crushed Ostrogothic regime, who is cited by Jordanes; or other, postulated lost ‘‘Gothic’’ histories; or via hypothetical Gothic oral sources. It is not claimed to represent ‘‘real history,’’ but rather is seen as a mythic narrative propagated by Gothic elites in order to generate ethnic unity through belief in a shared history and origins (Wolfram 1981; 1988; 1994). The text is not seen as preserving isolated ‘‘factual’’ elements of Gothic cultural provenance, scattered throughout a Greco-Roman literary work (as, e. g., Herodotus preserves isolated data about Scythian burial practices or royal Persian propaganda), but as replicating a coherent and articulated narrative, Gothic not classical in origin. The framework of the Getica, its migrations and dynasties, is understood as determined by the structure and key themes of an otherwise unreachable, oral ur-narrative cultivated by the Gothic elite. Though the Constantinopolitan Jordanes is belittled as a poor transmitter of this tradition, introducing many irrelevant classical cliches, the Getica is deployed as the residue of a barbarian ‘‘ethnic discourse’’ aimed at attracting adherence to Gothic group identity, in the face of competition both from other barbarian identifications and from Roman imperial allegiance. The evidence of Jordanes for the function of this ‘‘ethnic discourse’’ is fundamental for extrapolating the ‘‘ethnogenesis’’ model to other northern European barbarian tribes.
In this model, as in contemporary research on classical ethnography, a text serves to reify group identity. But the role of the text in this model is in fact diametrically opposite to the function of texts as understood in the current approaches to Greek and Roman writings discussed above. Current interpretations of Greco-Roman works understand classical ethnographic discourse as a communication within Greco-Roman society, functioning to alienate its audience from the object of its discussion, ‘‘the barbarian’’: Greek speaks to Greek about what makes the Scythians so different. Recent interpretations of Jordanes, however, see his text (or its postulated underlying ur-narrative) as a communication within the society of the object of discussion, the Goths, serving to reinforce group identity with the object: Goth speaks to Goth about what makes them Gothic. Classical ethnography seeks to repel, but ur-Jordanes seeks to consolidate. The ‘‘ethnogenesis’’ reading of the function of Jordanes’ text (or of his sources) is at odds with the direction of the work of classical studies on the role of ethnographic texts as instruments of exclusion. The shared vocabulary of ‘‘constructing identity'' conceals a more fundamental epistemological divergence.
There are practical stumbling blocks to this understanding of Jordanes as a recension of a ‘‘Gothic'' ethnic ideology. The relationship between the extant text of the Getica and one known but lost source, Cassiodorus’ Gothic History, is extremely problematic; the likelihood that Jordanes reflects Cassiodorus closely and consistently, or intended to, is not strong (Croke 1987; Christensen 2002: 127; Goffart 2005; Merrills 2005: 101-8). Hypotheses constructed on even earlier links in a hypothetical chain of lost sources are correspondingly more fragile. Interpretation of the Getica as a recension of an oral, Gothic tradition runs counter to the explicit statements of both Jordanes and Cassiodorus that they worked up their accounts of the Goths from written Greco-Roman sources, and that oral sources (i. e., Gothic traditions) were either contemptible or nonexistent (Cassiod. Var. 9.4; Jord. Get. 38; Goffart 1988: 38-9, 86-7; Gillett 2000: 484-5). But the deeper concern with this model is epistemological: the nature of Jordanes as a source, a late Roman work appropriated as evidence for ‘‘Gothic’’ self-identification.
In fact, Jordanes’ Getica looks very much like a classical ethnography, put together with more care and wit than is usually credited (see Merrills 2005: 115). Read in the light of current work in classical ethnography, Jordanes’ text not only becomes more familiar, but also more informative of the circumstances of its composition, Justinian’s Constantinople. The author exploits, from the outset, those features of classical ethnography that function to alienate the audience from his barbarous subject. The work commences with a geographic overview before proceeding to historical narration, a marriage of geographia and historia found not only in Herodotus but also in other Greek writings from the fourth century bc onward (Momigliano 1971: 58). The opening description of the northern islands of Britain and Scandza (whence the Goths are said to come) is a pot-boiler of Greco-Roman ethnographic geographia. Particularly prominent are the recurring references to Oceanus, the location of these islands. Impassable, sluggish, and hostile, Jordanes’ Oceanus is the product of the long literary-geographic tradition in which Oceanus constituted the border between order and chaos (Jord. Get. 4-12, 16-18, esp. 5; Romm 1992: 11-26; see, on the literary role of islands, Merrills 2005: 164-5).
Jordanes’ description of the tribes living in Scandza, neighbors to the ‘‘original’’ home of the Goths, is a systematic catalogue of ethnographic ‘‘barbarian’’ typologies, many already appearing in Herodotus: they live in lands where the laws of nature are turned upside down; they are pre-agrarian; they produce primitive primary goods, breeding horses and capturing animals for fur; and they exist, in multitudinous numbers, in a vast series of tribes that pressure one another like a long series of dominoes (Jord. Get. 19-24; see, e. g., Hdt. 4. 13; Gillett 2002b: 16 n. 32). The theme of migration, which provides one of the main narrative dynamics of the work (and a cornerstone to the early modern construct of the Volkerwanderung), derives from classical ethnographic writings, and is exploited in other late antique works (e. g., Hdt. 1. 16, 103; 4. 1. 11-12; and esp. 4. 13; Thuc. 1. 2. 12; Late Antiquity: Amm. Marc. 31. 3. 1; Priscus, fr. Blockley 1981: 40. 1). It evokes a key concept in classical ethnographic thinking: the hierarchy of autochthonic over migratory peoples, the latter inherently inferior (Loraux 2000; Dench 2005: 18-20, 96-101, 244; by contrast, some other late antique explanations of origins of barbarian groups saw them, more charitably, as new names for already existing aboriginal peoples: e. g., Agathias 1. 2. 1). Jordanes’ short text includes frequent geographic digressions, and the author underscores the geographic-literary context within which he sets his depiction by citing an impressive array of geographic authors, including Ptolemy, Dio Chrysostom, Pomponius Mela, and Strabo, and snippets from writers in other genres who are often presented as if they also were geographers (Gillett 2000: 487-8).
Such features, consciously selected by the author and prominently displayed, signal that this sixth-century Byzantine text was intended to be understood in terms of the very long tradition of ethnographic geographia and historia, and that in doing so it exploits the potent, alienating conception of the barbarian Other. Jordanes stands in the company of many other late antique authors who adapted for their own purposes venerable Greco-Roman portraits of ‘‘barbarians’’ and other enemies of order: such as Claudian, who depicts Alaric as a new Hannibal; Synesius, who breezily transports rebellious Gothic soldiers back in time to an ancient Egypt culled from Plutarch; or the anonymous author of the romance-hagiography Euphemia and the Goth, who uses the stereotypically barbarian faithlessness of his villain as the pivot of the plot (Claud. De bello Getico 78-82; Dewar 1994; Synesius, De providentia; Euphemia and the Goth).
More specifically, Jordanes’ ethnographic monograph (like his summary of Roman history, the Romana) was very much a part of contemporary literary activity in Justinian’s Constantinople. In the early to mid sixth century, Constantinople and the eastern empire more generally witnessed the production of an impressive variety of texts with historical, geographical, and ethnographic interests that intersect with Jordanes (Amory 1997: 135-47; Goffart 2005; Merrills 2005: 129-30, 145, 162-7; on historical literature of the Justinianic period more generally, see Scott 1990; Wilson 1996: 55-6; Rapp 2005b). Perhaps most importantly, the historical narratives of Jordanes and Procopius not only overlap, inevitably given their subjects, but speak to each other, sometimes contradictorily, in a historiographic dialogue yet to be fully elucidated (Goffart 1988: 94-6). Other interactions or striking parallels are also evident: with the author Capito, who like Jordanes wrote both a breviary of Roman history and a history of an ‘‘ethnic’’ group within the empire, the Isaurians (Amory 1997: 136, 304-5); with the chronicler Marcellinus comes, who helped ‘‘manufacture’’ the east Roman assertion of ad 476 as the definitive end of the western Roman empire, and was followed in this by Jordanes (Croke 1983); with the ethnographic victory ideology espoused in Justinian’s imperial titulature and legal rhetoric (Amory 1997: 140-1); with the attempt in the so-called Frankish Table of Nations to categorize contemporary barbarian peoples in terms drawn from classical ethnography (Goffart 1983); and more broadly with the geographical and ‘‘ethnological’’ interests of the Cosmographia of Julius Honorius, the Christian Topography of Cosmas Indicopleustes, and the Ethnica of Stephan of Byzantium. The range of points of contact between Jordanes and his contemporaries may, with further study, help triangulate his position within the debates of his time; together with the lively tradition of classical ethnography, they provide a potential matrix of interpretation for this often misconstrued text. Whatever the author’s elusive purpose, the work is clearly a ‘‘mirror’’ not of barbarian but of Byzantine society and concerns.
Emphasizing the classicizing nature of Jordanes’ text need not preclude the possibility that individual filaments of data genuinely of Gothic provenance may be preserved in the Getica, or in other late antique sources. Jordanes embellishes his text with a few items of Gothic vocabulary, just as Herodotus includes a smattering of Scythian, Persian, and Egyptian words (Waterfield and Dewald 1998: 742-4). But Herodotus does not represent a Scythian ideology, or Jordanes a Gothic one.
Approaching Late Antiquity, we need to be aware that large areas of the field have already been mapped in detail by earlier visitors, albeit that they knew the region by different names. Mindful of our debt to their pathfinding, we are nevertheless not obligated to adhere to their routes, or refrain from disturbing their markers. The habit in both past and current Germanicist scholarship of reading Greco-Roman sources as if they were ‘‘barbarian’’ ones obscures a genuine historical phenomenon: the vitality of classical conceptions of the barbarian and of ethnographic discourses in Late Antiquity, and consequently the scale of their impact on our perceptions of the period. Just as in earlier antiquity widely held ethnographic assumptions informed a variety of discourses ranging from drama to philosophy, so too in Late Antiquity the classical concept of the barbarian contributed to current thought and debate. Amid the religious disputes of the fourth century, Julian defended the reasonableness of polytheism on the basis of the varied inherent natures of different peoples; and Epiphanius of Salamis developed a schema of religious evolutionism, commencing after the Fall with ‘‘barbarism,’’ in order to identify the origin of heresy (Julian. Contra Galilaeos 115 D-131 D; Epiph. Adv. haeres. 1.1. 1-9). The moral diatribe of Salvian exploits the ‘‘qualities of peoples’’ literary tradition and the topos of barbarian simplistic virtue in castigating Roman decadence (Salvian, Degub. 4. 14. 67). The classical conception of the barbarian remained in Late Antiquity a ubiquitous and unquestioned intellectual resource, drawn upon reflexively.
The degree to which our vision of the barbarians of Late Antiquity is shaped by the tradition of Hellenistic ethnographic thought cannot be too heavily emphasized. The ‘‘Scandinavian origins’’ of Jordanes’ Goths (and of their medieval imitators, Paul the Deacon’s Lombards and Widukind’s Saxons) derive from the classicizing milieu of Constantinople, not from barbarian cultural beliefs (Goffart 2005). The very names by which we think of barbarian groups, even if ultimately derived from genuine autonyms, have been reconstituted into Greco-Roman frameworks, as the pseudo-ethnic, administrative terminology used by the author of the Life of Orientius of Auch suggests. By contrast, our expectations of meeting ethnically proud Goths, Franks, and Lombards has been conditioned not by the force of late antique evidence but by the weight of five centuries of modern Eurocentric scholarship. The historical counterparts of these figures appear rather differently in our sources: as the pious Christians, avaricious landowners, and wily negotiators of Roman social and governmental structures who feature in the narratives of Gregory of Tours and Paul the Deacon, and in the transactions recorded by charters.
Several desiderata for future research have been mentioned in this discussion, and before concluding two more may be added. First, it would be healthy for study of the postimperial western kingdoms to be disturbed from its customary, Eurocentric context; the construct of Late Antiquity should facilitate this. The western ‘‘barbarian kingdoms’’ were not the only new states formed on or within the borders of the Roman Empire, combining ‘‘barbarian’’ rulership and Mediterranean culture. On its Syrian frontier, at roughly the same time, Rome dealt with the kingdoms of the Ghassanids and Lakhmids, ruled by Christian Arabic courts influenced culturally and politically by Hellenistic (and Iranian) culture (Shahid 1984a, 1984b, 1989, 1995). Comparative studies of these western and eastern ‘‘provincial’’ or ‘‘transliminal’’ states, levering both away from their traditional scholarly contexts of medievalism and orientalism respectively, may well offer new insight not only into these kingdoms themselves, but into Late Antiquity more broadly; cultural dynamics are sometimes most tellingly revealed at peripheries.
Second, the energy of recent research into classical ethnography and its role in the social, political, and cultural dynamics of antiquity should be tapped by research on antiquity’s heirs. In what ways and to what extent classical ethnography affected texts and thought in Late Antiquity is an open question. How widely were Herodotus and other ethnographical writers read in Late Antiquity? Some forty-six papyrus fragments of Herodotus from the Christian period are known, a respectable number indicating currency of his text (Clarysse et al.). References to Herodotus by late antique authors, writing in Greek and even in Latin, are dauntingly numerous (some 900 references by Greek authors between the second and seventh centuries AD are listed in the Thesaurus linguae Graecae). Many, perhaps most, will be mere name-dropping; nonetheless, the frequent citations attest consciousness of Herodotus as a weighty literary model and source of information, however outdated. A study of the late antique afterlife of Herodotus, and of the classical ethnographic tradition in general, has yet to be undertaken. Our understanding now ofthe concept of the barbarian then is yet to be enriched by appreciation of the force of the classical ethnographic tradition in Late Antiquity.
BIBLIOGRAPHICAL NOTE
At the time of writing, the following represented the major treatments of the theme.
For overviews of recent work on the concept of ‘‘the barbarian’’ in classical antiquity: Dench 1995; Cohen 2000; Malkin 2001; Cartledge 2002: 51-77; Harrison 2002; and see now Isaac 2004; Campbell 2006 (not available to me when this chapter was written). A useful index of the growth of this field is OCD 3rd edn.: 233, s. v. ‘‘barbarian’’; there are no entries for this headword in the earlier editions. For similar approaches to other ancient societies: Poo 2005.
For Late Antiquity, essays on the concept of‘‘the barbarian’’ include Dauge 1981; Chauvot 1998; Heather 1999. Important earlier approaches, sidelined by recent debate, include Reydel-let 1981; Teillet 1984; and the English Marxist Thompson 1982. For historiographic discussion of earlier traditions: Goffart 1980: 3-39; 1989b; 1995; Halsall 1999; Nicolet 2003 (on French ‘‘Romanists’’ and ‘‘Germanists’’); Pizarro 2003: 43-7 (on ‘‘Germanic antiquity’’).
Introductions to ‘‘ethnogenesis’’ approaches include: Wolfram 1981; 1988: 1-18; 1994; 1998; Pohl 1991; 1998; Geary 1988; 1999; 2002. There are several parallel or modified forms of this approach, not necessarily fully in agreement with each other, including Heather 1991; Amory 1997. The latter, the most valuable single contribution using this approach, embraces a range of methodologies and evidence. Critiques of‘‘ethnogenesis’’ as an explanatory model are gathered in Gillett 2002b (essays by Bowlus, Gillett, Goffart, Kulikowski, and Murray; the latter is the fullest historiographic analysis of the model and its development from antecedents); see also Bowlus 1995; Goffart 1995; 2005; Gillett 2005.ForJordanes: Croke 1987; Goffart 1988:20-111; 2005; 2006; Gillett 2000; Christensen 2002; Pizarro 2003: 47-51; Merrills 2005: 100-69.
Moving from the ideological construct of‘‘the barbarian’’ to the underlying economic and military realia of Roman-barbarian relations: Whittaker 1994; 2004; Wells 1999; Burns 2003. For important new approaches to material evidence from European ‘‘barbarian’’ societies: Veit 1989; Halsall 1992; Brather 2002; 2004.
World History
