Empires and Barbarians

by Peter Heather

  30 ‘He was a Greek trader . . .’: Priscus fr. 11.2.422–35. Edeco: PLRE 2, 385–6.

  31 You can tell how people were dressed by where they wore their safety pins, which is all that tends to survive of clothing in most graves. Possible reasons for archaeological invisibility can range from the dramatic (where bodies are left exposed to the elements and wild animals) to the prosaic (cremation followed by scattering of ashes), or the generation of customs where bodies are buried without any chronologically identificatory gravegoods (something which often makes medieval cemeteries undatable in northern Europe once populations convert to Christianity). The horizons of the Hunnic Middle Danube are differentiated from one another by slight changes in the manner in which broadly similar sets of gravegoods were decorated. In chronological order (and there are overlaps between them), the sequence starts with the Villafontana horizon, succeeded in turn by those of Untersiebenbrunn and Domolospuszta/Bacsordas. For introductions to this material, see Bierbrauer (1980), (1989); Kazanski (1991); Tejral (1999). There are excellent illustrations in Wolfram (1985).

  32 Many of the Germanic groups of central Europe had practised cremation in the first to the third century, but inhumation was already spreading more widely before the arrival of the Huns.

  33 Historical sources do occasionally supply enough information, however, which can be used in conjunction with the archaeological evidence approximately to identify some particular groups.

  34 See most recently Halsall (2007), 474–5; for a similar view of the Avar Empire, see Pohl (1988), with Pohl (2003).

  35 Priscus fr. 14.

  36 Priscus fr. 11.2, p. 259. See Chapter 4 for the Tervingi and Greuthungi; cf. the junior status and grimmer treatment handed out to the Sciri after Uldin’s defeat: see above, note 2.

  37 For references, see note 2 above; and see p. 248.

  38 Priscus fr. 2, p. 225.

  39 Priscus fr. 2, p. 227.

  40 Priscus fr. 2, p. 227.

  41 Priscus fr. 49.

  42 The Romans provided Attila with a succession of secretaries, and a prisoner called Rusticius wrote the odd letter (Priscus fr. 14, p. 289). This governmental machine could keep lists of renegade princes who had fled to the Romans, and could keep track of the supplies required from subject groups, but little more. Akatziri: Priscus fr. 11.2, p. 259. Goths: Priscus fr. 49.

  43 Jordanes, Getica 48.246–51, with Heather (1989), (1996), 113–17, 125–6.

  44 Gepids: Jordanes, Getica 50.260–1.

  45 Franks: Priscus: fr. 20.3. Akatziri: see note 42 above. Of the subject groups in between, the most dominated were apparently the Goths who appear in Priscus fr. 49, part of which is quoted above; the least dominated were the Gepids, who led the revolt against Attila’s sons (see previous note). In between were the Pannonian Goths of Valamer: see note 43 above.

  46 Miracles of St Demetrius II.5.

  47 See e.g. Agadshanow (1994).

  48 For further discussion, see Heather (2005), 324ff., with references.

  49 As we have seen, modern anthropological evidence indicates that the most you will sometimes find in such circumstances is that a very few particular items have significance for signalling group identity, but that does not mean that the group identity is in any sense unreal: above, p. 26, after, in particular, Hodder (1982).

  50 While Attila could extract annual subsidies measured in thousands of kilos of gold, the most that even a successful Hunnic successor group like the Amal-led Goths could manage was three hundred: Priscus fr. 37.

  51 Jordanes, Getica 50.265–6. Jordanes himself came from this Balkan military milieu, and there is every reason to suppose this catalogue correct. When exactly in the 450s or 460s these settlements occurred is not clear: that of Hernac is firmly dated to the later 460s, however, and it may be that they all belong to the post-465 meltdown of Hunnic power that also saw moves into Roman territory by Bigelis and Hormidac. Hernac’s willingness to have his power base broken up might explain why he was treated more favourably than Dengizich (see note 23 above).

  52 Jordanes, Getica 53.272; cf. Agathias 2.13.1ff.

  53 Paul the Deacon, History of the Lombards 2.26ff.; cf. Jarnut (2003). It is a consistent theme both within Paul’s narrative and some of the other early Lombard texts that victory led to the inclusion of warriors in the group, but not always on terms of equality: see e.g. Origo Gentis Langobardorum 2 (as aldii: ‘half-free’); History of the Lombards 1.20, 1.27, 5.29.

  54 Goths: Heather (1996), Appendix 1. Lombards: ibid., and see previous note. See also Chapter 2 above.

  55 Lombards: see e.g. Jarnut (2003), who argues that kingship among the Lombards may have been a temporary phenomenon restricted to the leading of expeditions. Goths: Heather (1989), (1996), chapters 8–9.

  56 For the Rugi joining Theoderic in 487: John of Antioch fr. 214.7; on their still being identifiable in 541: Procopius, Wars 7.2.1ff. (they had swapped sides twice during the Gothic conquest of Italy). Heruli: Procopius, Wars 6.14–15.

  57 The account of the Heruli is doubted by Goffart (1988), 84ff.; that of the Rugi by Halsall (1999). See Chapter 1 above for general comments on modern understandings of group identity.

  58 The Gundilas papyrus (translated by him as Appendix 1) is central to Amory (1997). But see also Heather (1996), chapter 9, and Appendix 1, (2003).

  59 Malchus fr. 20, p. 446.215ff. (the 6,000 men), p. 440.83ff. (non-combatants and baggage). Cf. Jordanes, Getica 55.281–2 (Theoderic had earlier also used 6,000 men in the expedition that proved his manhood following his return from being a hostage in Constantinople). For further commentary, see Heather (1991), chapter 7.

  60 See Amory (1997); but see also, in addition to Procopius, Wars 5.1.6ff., Ennodius, Panegyric on Theoderic 26–7 and Life of Epiphanius 118–19 (cf. 111–-12).

  61 The east Romans captured 2,000 wagons in a surprise attack (Malchus fr. 20), but there is nothing to suggest that this was the total baggage train. The Goths were offered ‘unoccupied’ land, which strongly implies that they were to do their own farming, as do all the negotiations between Theoderic and Constantinopolitan representatives: Malchus fr. 18.3, p. 430.5ff.; fr. 20, p. 438.55ff., p. 446.199ff.; cf. Heather (1991), 244ff.

  62 For fuller discussion and complete references, see Heather (1991), 259–63; for Bigelis, see note 22 above.

  63 For pay and rations for 13,000, and 910 kilos of gold per annum, see Malchus fr. 18.4, p. 434.12ff. and fr. 2, p. 408.22ff. For full discussion and references, see Heather (1991), 253–6.

  64 For Strabo’s death, and Recitach’s assassination, see John of Antioch fr. 211.4 and fr. 214.3. For Theoderic’s forces in Italy, see Hannestad (1960). For full discussion, see Heather (1991), 300–3.

  65 For references, see notes 22 and 23 above.

  66 On Herule numbers in 549: Procopius, Wars 7.34.42–3. It is generally tempting to think that the Heruli were smaller than the Amal-led Goths because the latter are portrayed as so victorious in the post-Attilan competition on the Middle Danubian plain. Our only source for this, however, is Jordanes, and it may be that Theoderic’s following only acquired superpower status when he added the Thracian Goths to his following.

  67 The migration topos entirely suffuses Paul the Deacon’s History of the Lombards: the brothers Ibor and Agio lead the first move from Scandinavia, Agilmund the second into Bohemia, Godo takes them into Rugiland, Tato fights the Heruli, and Wacho leads the annexation of part of Pannonia. For modern secondary comment, see the works cited in note 27 above.

  68 See especially Jarnut (2003), with references, and for the thought – as note 55 above – that early Lombard kings may fundamentally have been expedition leaders; cf. Christie (1995), 14–20.

  69 See Curta (2001), 190–204, with his figure 18.

  70 On various occasions, groups of Ostrogoths, Heruli, Huns, Rugi, and Lombards all fall into this category of mass migration. Lombard migration may well have taken the form of an initial flow that had to reorganize itself in mass form whe
n it was necessary to fight the Heruli head-on. In this, it resembles the third-century Goths: see Chapter 3.

  71 Vidimer: Jordanes, Getica 56.283–4. Procopius, Wars 1.8.3 explicitly names Bessas and Godigisclus among the Thracian Goths who didn’t follow Theoderic; see Heather (1991), 302 for some other contenders.

  72 The Amal-led Goths were receiving 136 kilos of gold per annum in the 460s (Priscus fr. 37), while the Thracian foederati pulled in 910 (see note 63 above). On Theoderic and the wealth of Italy, see Heather (1995b).

  73 For Hun-generated wealth, see note 31 above. It is possible, however, given their seemingly non-centralized political structures, that the further spread of Lombard groups south of the Danube into old Roman Pannonia may have again taken the form of a variegated flow rather than a single directed movement.

  74 Life of Severinus 6.6.

  75 For references, see note 56 above. Alternatively, it may be , given Theoderic’s subsequent success, that they had no real choice in the matter.

  76 For markets, see Priscus fr. 46. For other references, see note 23 above.

  77 On Theoderic’s spell as a hostage, and its ending, see Heather (1991), 264–5. On his mention of Italy in 479, see Malchus fr. 20.

  78 Rodulf of the Rani: Jordanes, Getica 3.24.

  79 For the route of the 473 trek: Jordanes, Getica 56.285–6.

  6. FRANKS AND ANGLO-SAXONS: ELITE TRANSFER OR VÖLKERWANDERUNG?

  1 Campbell (1982), chapter 2.

  2 The old maximalist tradition runs from scholars such as Freeman (1888) to Stenton (1971). It never went unchallenged, but scholars such as Higham (1992) and Halsall (2007), especially 357–68, are representative of the more substantial minimizing tradition of recent years. Recent scholars thinking in terms of large-scale migration include Campbell (1982), Härke (1992), Welch (1992). Hills (2003) is representative of an ultra-minimalist position adopted by some younger archaeologists. A good introduction to the variety of opinion is Ward Perkins (2000).

  3 See Woolf (2003).

  4 H. R. Loyn, quoted in P. Sawyer (1978). The best introduction to late Roman Britain remains Esmonde-Cleary (1989).

  5 For an overview of Anglo-Saxon settlement and the development of place names, see Hooke (1998).

  6 See Heather (1994).

  7 Esmonde-Cleary (1989) is very balanced on the end of Roman Britain, as is Halsall (2007), 79–81, 357ff. For an introduction to the literature on systems collapse, see amongst others Faulkner (2000); Jones (1996); Higham (1992). Dark (2002) stands against this position.

  8 One recent example is Halsall (2007), 519ff., with references to some of the alternatives.

  9 For an introduction to such materials, see Dumville (1977).

  10 Campbell (1982), chapter 2 provides a clear introduction to the Chronicle.

  11 For useful introductions to this material, see Campbell (1982), chapter 2; Arnold (1997); Welch (1992).

  12 This much is accepted even by such a general anti-migrationist as Halsall (1995b), (2007), 357ff.

  13 See e.g. Arnold (1997), 21ff.

  14 Ine’s Law 24.2 (cf. 23.3); cf. Arnold (1997), 26ff., with discussion of Warperton.

  15 Compare, for example, Weale et al. (2002) with Thomas et al. (2006). The sample was of men whose pre-Industrial Revolution ancestors can be shown to have been living in the same area as the modern descendant.

  16 One other line of thought has therefore taken a more indirect route, attempting to identify and analyse so-called ‘epi-genetic’ features of the skeletal remains unearthed from the inhumation cemeteries of the fifth to seventh centuries. Such factors reflect the impact of inherited genes rather than diet or environmental factors. The work was able to establish that the element of the population buried with weapons was noticeably taller than those buried without. The argument continues as to whether the height differences should be explained genetically – i.e. as a sign that the weapons-bearers were an intrusive population – or by something else, such as differences in diet, and no firm conclusions have yet emerged: see Härke (1989), (1990).

  17 For 446 AD, see Bede, Ecclesiastical History 2.14; 5.23, 24 (after Gildas, Ruin of Britain 20). For 450 AD, see The Greater Chronicle, year of the world 4410; cf. Anglo-Saxon Chronicle (449 AD) for the arrival of the Kentish dynastic founders, Hengist and Horsa.

  18 Everything is reasonably clear up to about 409, when Zosimus 6.5 records a British revolt. Controversy really begins with Zosimus 6.10, which is traditionally interpreted as Honorius telling the British provincials to look after their own defence, although the text is corrupt. For an introduction to these events and the historiography, see Salway (2001).

  19 Gildas, Ruin of Britain 23–6.

  20 On the Saxon attacks in c.410, see Gallic Chronicle of 452 (though this chronicle does not always date events to individual years). For the first datable remains, see Welch (1992), chapter 8. The appendix to Halsall (2007) attempts to extend the generally accepted sequence still further, arguing that Gildas’ unnamed tyrant, who issued the invitation, usually thought of as a post-Roman figure, was in fact the usurping Emperor Maximus (383–7), and that it was Maximus who brought the first Saxon mercenaries to Britain. This is not an impossible reading, but neither is it the most obvious, so the jury is still out. The further arguments which Halsall erects on the back of this first hypothesis are unconvincing: see notes 44 and 46 below.

  21 Gildas’ report that Roman Britain’s final appeal to the central imperial authorities came when Aetius was (or had been) consul for the third time (446 or after) might provide some further confirmation that the 440s were a period of particular disaster. The British leader on the Loire was Riotamus: see PLRE 2, 945.

  22 See e.g. Campbell (1982), chapter 2; Higham (1994); Halsall (2007), Appendix.

  23 See Dumville (1977).

  24 Gregory of Tours, Histories 4.42; cf. Paul the Deacon, History of the Lombards 2.6ff.

  25 Bede, Ecclesiastical History 1.15 (Angles, Saxons and Jutes); 5.9 (the others).

  26 Higham (1992), 180–1.

  27 Gregory of Tours, Histories 5.26, 10.9; Procopius, Wars 8.20.8–10; cf. especially Woolf (2003).

  28 Famously, the supposed Gothic migration from Scandinavia is also said to have taken place in three ships: Jordanes, Getica 1.25, 17.95.

  29 See Chapter 4 above.

  30 For Norse DNA evidence, see Chapter 9. For language change, see p. 296 above.

  31 Gildas, Ruin of Britain 23–6.

  32 On the Saxon attack on Gaul: Ammianus 28.5. For introductions to the ‘Saxon Shore’, see Johnston (1977); Rudkin (1986).

  33 On coastal inundation, see the excellent discussion of Halsall (2007), 383ff. On Frankish pressure: Gregory of Tours, Histories 4.10, 14.

  34 Carausius: PLRE 1, 180. On parallel phenomena in the Viking period, see Chapter 9.

  35 Gildas, Ruin of Britain 20. The archaeological evidence for Pictish and especially Scottish (= Irish) intrusion into western Britain is irrefutable, even if there is little in the way of historical evidence. A good recent account is Charles-Edwards (2003), Introduction and chapter 1.

  36 See Woolf (2003), 345f.

  37 On the nautical evidence, see Jones (1996), though his discussion includes neither a consideration of Roman ships nor the extended nature of the Anglo-Saxon migration flow. For Goths and the Black Sea, and Vandals and North Africa, see above Chapter 4.

  38 For further exploration of these issues, see e.g. Dark (2002); Woolf (2003).

  39 See e.g. Higham (1992); Halsall (1995a), (2007), 357ff.

  40 An excellent general survey is Hooke (1998).

  41 Relevant general surveys include Hooke (1998); Williams (1991). An excellent case study is Baxter (2007), chapter 7. On the demotion of the peasantry, see Faith (1997), chapter 8.

  42 For introductions to this issue, see Hooke (1997); Powlesland (1997).

  43 See Esmonde-Cleary (1989), 144–54; cf. Loseby (2000) and Halsall (2007), 358f., both with references, on attempts to generate a sub
stantial post-Roman urbanism.

  44 On the peasants’ revolt, see e.g. Jones (1996); cf. Halsall (2007), 360ff.

  45 Constantius, Life of St Germanus 13–18, 25–7. For the Romance-speaking elite, see note 3 above. The famous Llandaff charters may provide further confirmation of essential sub-Roman continuity, although this has been disputed: see Davies (1978).

  46 ‘Is simply to dispose . . .’: Halsall (1995), 61. This ‘before’ and ‘after’ approach to migration is quite common. For another example, see the comments of Higham in Hines (1997), 179, where a reinterpretation of a set of remains – by Hines (1984) – is praised as ‘more complex’ because it ejected migration from its usual role in their discussion. See pp. 160 and 192 for two instances where the determination to minimize the importance of migration has led scholars, including Halsall again, to make methodologically problematic choices in their handling of the evidence.

  47 There are many parallel examples, but for a recent overview of the decline of Roman structures in the Balkans, see Heather (2007).

  48 See the review of the literature in Woolf (2007), 123ff., which draws on, amongst others, Denison (1993) and Hall (1983), which have effectively countered the attempts of Preussler (1956) and Proussa (1990) to detect deeper Celtic influences on Old English. On later medieval cases of language change, see Bartlett (1993), 111ff.

  49 See further Chapter 2 above.

  50 This emerges with huge clarity from all the literary sources – everything from critiques of individual kings in historical narratives to the value systems underlying heroic poetry. Introductions to the mix of land and cash expected over the course of an individual’s lifetime are provided by Charles-Edwards (1989); Campbell (2000), chapter 10.

  51 For a general introduction to the pre-Viking great powers, see Campbell (1982), chapters 3–4.

  52 See Hooke (1998), chapter 3; Powlesland (1997); Esmonde-Cleary (1989).