Stilicho: The Vandal Who Saved Rome

by Ian Hughes

  77 Claud., de Bello Get. 624f. and VI Cons. Hon. 297–8: Prudentius, Contra Symmachus, II. 696–744. Modern agreement, e.g. Liebeschuetz, 1998, 63.

  78 Cassiodorus, Chron. a 402: Jordanes, Getica, 154–55.

  79 Prosper, Chron. a. 402; Orosius, 7.37.2; and Jerome Ep. 107.2.3. Modern agreement, e.g. Wolfram, 1990, 152.

  80 For example, Wolfram suggests that Stilicho was able to stand his ground but was forced to let Alaric go due to the Gothic cavalry arm remaining intact (1997, 96). Halsall suggests the battles were indecisive, although Stilicho had the better of them (2007, 201).

  81 Veg., Epit. Rei Mil. 3.9.

  82 Maur., Strat. 8.2.4.

  83 Maur., Strat, 10.2. The passage rewards careful reading by illuminating the profound shift in emphasis that the Roman army had undergone from the times of the Early Empire.

  84 These are the tactics idealized by Vegetius in his Epitoma rei Militaris: 3.9f.

  85 E.g., O’Flynn, 1983, 40.

  86 E.g., O’Flynn, 1983, 40.

  87 Claud., de Bello Get., 580f.

  88 Claud., VI Cons. Hon. 223f.

  89 O’Flynn, 1983, 54.

  90 Burns, 1994, 193.

  91 Christie, 2007, 547f.

  92 Ferrill, 1988, 97.

  Chapter 11

  1 Matthews, 1998, 309 n.3.

  2 Amm. Marc., 17.8: 20.10.2.

  3 Burns, 2003, 325

  4 Burns, 2003, 357

  5 Drinkwater, 1998, 273.

  6 It is probable that, along with recruiting barbarians for their armies, both Maximus (383–8) and Arbogast (388–95) made treaties with the tribes on the frontiers to defend the same frontiers against attack, so allowing them to withdraw the troops usually stationed there to face Theodosius in the ensuing civil wars.

  7 Burns, 2003, 353.

  8 For a more detailed discussion, see Halsall, 2007, 207–10.

  9 For a more in-depth discussion of the date and references to related literature, see Halsall, 2007, 209f. See also Wolfram, 1997, 95.

  10 cf. Drinkwater, 1998, 274 and 277.

  11 Drinkwater, 1998, 277.

  12 Burns, 2003, 361.

  13 Halsall, 2007, 347f.

  14 Matthews, 1998, 256

  15 This section is heavily indebted to the works of E A Thompson, especially ‘Peasant Revolts in Late Roman Gaul and Spain’, Past and Present, No. 2. (Nov., 1952), pp. 11–23.

  16 Thompson sees the revolt of Maternus during the reign of Septimius Severus (193–211) as the earliest known example of an uprising that would later be known as bacaudae (1952, 12).

  17 Thompson, 1952, 1506.

  18 Burns, 2003, 359. Poulter, 2007, 67f. and Heather, 2007, 170–71, describe the process in more detail with reference to the extensive ongoing archaeological investigations.

  19 Matthews, 1998, 264.

  20 Gaudentius, ‘Tract.’ 17.1f. cf. 13.21 (CSEL LXVIII, pp. 141, 120) referenced in Matthews, 1974, 59: Symm. Epp. 4.9: 7.13–14.

  21 Matthews, 1975, 281, n.4.

  22 Collectio Avellana, No.38. p. 85 (ref. Wolfram, 1990, 153): ‘Etiam super excidio pereuntis Illyrici pio apud vos prodiderimus affectu esse nobis dolori, cur ista nos detrimenta rei publicae nolueritis agnoscere et aliis potuis indicibus quam pietatis vestrae litteris fuerint nuntiata.’.

  23 Honorius apologizing, Wolfram, 1990, 153; Honorius accusing Arcadius, Halsall, 2007, 202 n.69; Heather suggests Honorius is complaining that Arcadius has not informed the West of Alaric’s attacks after he was evicted from Italy in 402; Liebeschuetz, 1991, 65, avoids the issue of the meaning of the letter, only commenting upon the nature of the barbarians.

  24 Williams and Friell, 1994, 155: On the dating, see the previous chapter.

  25 This section is based mainly on Liebeschuetz, 2004, 167f. See also Norwich, 1988, 125f.

  26 This is the letter Collectio Avellana, No. 38. p. 85, mentioned above.

  27 For the complexities surrounding this event, Williams and Friell, 1994, 155; on the date, 223, note 52; see also Heather, 1994, 221.

  28 Soz., HE 8.25.3; 9.4.2-4. Kulikowski, 2007, 171; Williams and Friell, 223, note 52. Zos. 5.26.

  29 405, Seeck, Stein, Schmidt and Mazzarino; 406, Demougeot: source Zosimus: New History, Trans. Ridley, p. 215, n. 84.

  30 Heather, 1991, 77–78.

  31 PLRE II, Iovius 3, 623.

  32 It is possible that Stilicho ‘revived or manufactured a claim’ that Theodosius had ordered that the Prefecture of Illyricum be attached to the West: CAH, 2004, 121.

  33 Rutilius Namatianus, de Reditu suo, 41f: http://penelope.uchicago.edu/ Thayer/E/Roman/ Texts/Rutilius_Namatianus/text*.html (July 2009).

  34 ‘seeing that Arcadius’ ministers were alienated from him’: Zosimus, New History, VIII. 26. 1–2:

  35 ‘King of the Goths’, for example: Aug., Civ. Dei., 5.23; Oros. 7.37.4; Prosp. Tiro. s.a. 400; Chron. Gall. 452 no. 50.

  36 E.g., Williams and Friell, 1995, 155: Wolfram, 1997, 97.

  37 Burns, 1994, 198 believes that Radagaisus entered Italy via the Brenner Pass: Kulikowski, 2007, 171 also implies the same when he notes that Radagaisus invaded via Raetia. For the alternative, see e.g. Wolfram, 1997, 97, who believes that Radagaisus entered the empire via Pannonia.

  38 This obviates the need to understand why Alaric did not join the forces passing through ‘his’ territory; cf. Williams and Friell, 1994, 155, who claim that Alaric was still too weak to join in.

  39 This is the traditional date for the exchange of those hostages including Aetius which may have instead taken place in 402; cf., Liebeschuetz, 1998, 63–4; Halsall, 2007, 202.

  40 Oros., 7.37.4.

  41 For a more detailed discussion of the numbers, see below.

  42 Chron. Gall. 452 no. 52: Paulin. Vita Amb. 50.

  43 Oros. 7.37f.

  44 Chron. Gall., 452. No. 52.

  45 Aug., Civ. Dei. 5.23; Oros. 7.37.6.; Zos. 5.26.4.

  46 Zos., 5.26.4.

  47 Williams and Friell, 1995, 155.

  48 Cod. Th, 7.13.16: 7.13.17. Interestingly, Freeman (1886, 54) saw this as a measure to counter the invasion of the Vandals, Alans and Sueves, which didn’t happen until the end of the year.

  49 Zos., 5.26.4; Huns led by Uldin, Wolfram, 1997, 97.

  50 Williams and Friell, 1995, 155.

  51 Zos. 5.26.4 claims the Danube, but see Ridley’s translation, p 215, n.87 for clarification.

  52 Paulin. Vita Amb. 50.

  53 Aug., Civ. Dei. 5.23 (http://www.ccel.org/ccel/schaff/npnf102.iv.V.23.html). Orosius 7.37.14 also notes that the Romans suffered few casualties.

  54 Captured and executed, Oros. 37.13–15; with his sons, Aug. Civ. Dei. 5.23.(http://www.ccel.org/ccel/schaff/npnf102.iv.V.23.html); date, Addit. ad Prosp. Hann. Ad. A. 405 (Radagaisus, PLRE II, 934). Possibly in Rome, rather than at Faesulae, Paul. Nola. Carm. 21. 1–34.

  55 Olymp. fr. 9.

  56 For additional information and clarification on detailed aspects of the Roman army, refer to Chapter 4. See also Burns, 1994, 356, n.53 for an outline of the problems surrounding the size of Stilicho’s forces.

  57 Zos., 5.26.3: Oros. 7.37.4.

  58 For example, Wolfram simply states that the army was ‘many tens of thousands’ (1997, 96–7). Others simply bypass the problem: for example, Liebeschuetz declares it was a ‘vast war band’ (2004, 65), Williams and Friell describe it as ‘a massive invasion’ (1995, 155), and the CAH has Radagaisus leading ‘a large force of barbarians’ (2004, 120–21)

  59 Olymp. fr. 9.

  60 Burns, 1994, 198.

  61 Chron. Gall. 452 no. 52; Paulin. Vita Amb. 50.

  62 Zos., 6.3. Zosimus may have confused events, with the last remnants of Radagaisus’ forces that escaped by crossing the Alps turning into Vandals, Sueves and Alans. However, this may actually represent a historical tradition, as discussed during the events of 406 below.

  63 O’Flynn, 1983, 41.

  64 Oros., 7.37.16; Zos. 5.26.4.

  65 Paul. Nola, Carm. 21.1-34.


  Chapter 12

  1 CAH, 2004, 123.

  2 CAH, 2004, 123.

  3 For the possibility of Stilicho laying claim to the guardianship and Illyricum, see Chapter 11.

  4 Collectio Avellana, No.38. For a full discussion, see Chapter 11

  5 Baynes, 1922, 216.

  6 Soz., 25.1–2.

  7 This section has been based mainly upon Sodini, 2007, who has demonstrated that the large-scale decline of Illyricum may have begun only after the attacks of the Huns in the 440s, since they had siege techniques that could capture Roman cities.

  8 Poulter, 2007, 18.

  9 Sodini, 2007, 315–16.

  10 Sodini, 2007, 328f.

  11 Jones, 1966, 82

  12 Proc. III. 12.6–7.

  13 Heather, 2005, 219–20.

  14 Soz. HE 8.25. Compare this to Kulikowski’s claim (2007, 171) that there is no evidence that Stilicho was going to join Alaric in Epirus.

  15 Heather, 2005, 220.

  16 Kulikowski, 2007, 172.

  17 Cod. Th. 7.16.1.

  18 This should be compared with the appointment of comes rei militaris in Africa (Chapter 8), Britain (Chapter 9), Spain (Chapter 9) and Illyricum (Chapter 10). Mann, 1979, 182.

  19 Cod. Th. 7.13.18.

  20 Burns, 1994, 205.

  21 Cod. Th. 11.7.14 (April 9, 407).

  22 Zos. 5.27.2f. Stilicho was in Ravenna and Honorius in Rome when news of the death of Alaric reached them. Bury, 1923, 60.

  23 Zosimus, Ridley, 1982, 216, n.91.

  24 Soz., HE 8.25: Zos., 5.27.2.

  25 Zos., 5.29.8.

  26 Zos., 5.27.2.

  Chapter 13

  1 Zos., 6.3.1.

  2 Drinkwater, 1998, 271.

  3 Salway, 1993, 300. This can be compared to the often repeated complaint concerning lack of pay to be found throughout Procopius, for example during the reconquest of Africa and Belisarius’ campaigns in Italy.

  4 E.g. Salway, 1993, 300: Bury, 1923, 188.

  5 Olymp., fr. 12, perhaps giving the date as late 406, cf. PLRE II. Marcus 2, 719: Zos. 6.2.1, giving the date as 407; cf. Soz. 9.11.2.

  6 ‘municeps eiusdem insulae’, Oros. 7.40.4.

  7 Kulikowski, 2000, 332 and note 44, as opposed to e.g. Burns, 1994, and Matthews, 1998, 308.

  8 Freeman, 1886, 55, 60.

  9 Drinkwater, 1998, 273; Gibbon, 1861, II, 249.

  10 Possibly adding to the anger of the Gauls and British at the lack of protection from the emperor.

  11 Lands becoming insufficient, Bury, 1923, 186, quoting Proc. III.22–3. Threatened by Huns, Bury, 1923, 186

  12 Godigisel, Greg. Tur. HF 2.9.

  13 Jer. Ep. 123.16: early modern historians appear to have followed Gibbon, who claims that the tribes that crossed the Danube comprised at least one third of the forces that Radagaisus led into Italy. These were numbered at 200,000 warriors and 400,000 overall, so leading to an estimate of around 60,000 warriors and 130,000 people overall (Gibbon, 1861, II, 246): cf. Freeman, 1904, 19f.

  14 Liebeschuetz, 1990, 13–14.

  15 Matthews, 1998, 275f: Kulikowski, 2007, 171f: Williams and Friell, 1994, 84.

  16 Drinkwater, 1998, 273f.

  17 Muhlberger, 2002, 29–30.

  18 Jer., Ep. 123.15.

  19 Jer., Ep. 123.15.

  20 Greg. Tur., HF 2.9.

  21 Kulikowski prefers to date the invasion to 405 rather than 406, based upon a preference for Zosimus’ dating at 6.3.1 rather than Prosper’s chronicle at 1230 (2000, 326f). Unfortunately, he weakens his own argument by later noting that Book 6 of Zosimus, upon which he bases the theory, is ‘the deeply unreliable Book VI of Zosimus’ (2000, 332, n.40, 41).

  22 Gibbon, 1861, II, 250: ‘The victorious confederates pursued their march, and on the last day of the year, in a season when the waters of the Rhine were most probably frozen, they entered, without opposition, the defenceless provinces of Gaul’; based on, for example, Herodian, 6.7.6–7

  23 Elton, 2004, 40, suggests that the invasion was not coordinated. The events surrounding the nature of the invasion, with the tribes arriving separately, to some degree supports this theory. However, the retreat of that section of the Alans under Respendial and their participation in the battle with the Franks (see below) implies that a low level of coordination existed.

  24 Chron. Gall. 452, No. 127 (s.a. 442).

  25 Once inside any further Burgundian movement was only permitted with Roman agreement: Oros. 7.38.3.

  26 Oros., 7.40.3.

  27 Bury, 1923, 186.

  28 Succession, Greg. Tur. HE 2.9; duration of Gunderic’s rule, PLRE II 522.

  29 Kulikowski, 2000, 326 and n.8; Greg. Tur. 2.9; Oros. 7.40.3.

  30 Drinkwater, 1998, 273–4.

  31 Drinkwater, 1998, 273.

  32 Drinkwater, 1998, 274.

  33 Jord., Get. 22 (115); Jer. Ep. 123.15.

  34 Jer., Ep. 123.15.

  35 Greg. Tur., 2.9.

  36 cf. Drinkwater, 1998, 271.

  37 Jord., Get. 22 (115).

  38 Oros., 7.38.

  39 Freeman, 1886, 54 and n.2.

  40 For further details and bibliography on the proposed itineraries, see Kulikowski, 2000, 331.

  41 Salvian, de Gub. Dei. 6.82–4. However, it should be noted that this episode is not clearly dated by Salvian and so could have occurred at any time during the period 406–9.

  42 Jer., Ep. 123.15.

  43 Paul., Euch. 239.

  44 The section covering the advance of Constantine into the heart of Gaul is in very large part based upon the analysis made by Drinkwater (1998), since his study clearly explains many issues which before then remained in doubt.

  45 Paschoud 1989, 20, on Zos. 5.27.1–2, suggests a date in February (Kulikowski, 2000, n.48); Burns, 1994, 210, suggests the date of 1 March.

  46 Dating, Burns, 1994, 210; reason for overthrow, Kulikowski, 2000, 332.

  47 Oros. 7.40.4; Proc. III. 2. 31. Burns argues that he was of a relatively high rank, 1994, 213.

  48 Soz., 9.11.

  49 Drinkwater, 1998, 272.

  50 For a discussion of the changes, see Drinkwater, 1998, 272.

  51 Kulikowski, 2000, 328, following Paschoud, 1989, 20 and Zos. V. 27. 1–2.

  52 Cod. Th. 7.13.18.

  53 Bury, 1923, 188, referencing Evans, A.J Num. Chron. 3rd Series VII, 191.

  54 Justinianus in Zosimus, Justinus in Olympiodorus; PLRE II Justinianus 1, 644; Nebiogastes in Zosimus, Neobigastes in Olympiodorus; PLRE II, Nebiogastes, 773–4.

  55 E.g., Burns, 1994, suggests 5,500 men, and Drinkwater, 1998, 275 6,000 men.

  56 Drinkwater, 1998, 275, n.42 referencing Hoffmann, 1973, 15–17.

  57 Zos., 6.3.

  58 Pollentia, Ridley, ‘Zosimus’, 1982, 227.n. 19; Goths, ibid; Freeman, 1886, 57.

  59 Freeman, 1886, 54 and n.2.

  60 Zos., 6.3.

  61 Drinkwater, 1998, 277.

  62 Drinkwater, 1998, 277–8.

  63 Drinkwater, 1998, 277.

  64 Drinkwater, 1998, 277.

  65 Drinkwater, 1998, 275.

  66 Coins: Kulikowski, 2000, 332 and footnote 51. Kulikowski links the coins to alleged victories over the invaders, whilst suggesting that they may instead proclaim ‘Constantine’s intention of fighting the barbarians and restoring the respublica’.

  67 We are not informed directly that Constantine attempted to ally himself with the Vandals, Sueves and Alans. However, Oros. 7.40.4 and 7. 28. describes ‘unreliable alliances’ with barbarians and how these treaties were ‘not strictly kept’. This can only relate to treaties with the invaders.

  68 Jer., Ep. 123.15.

  69 Cavallera 1922, 2:52, referenced in Kulikowski, 2000, 331, n.38.

  70 Kulikowski, 2000, 331.

  71 On the reliability of the various sources: Kulikowski, 2000, 331, footnote 39; ‘Apart from Jerome’s letter, Sozomen and three more or less contemporary poems (the Ad uxorem, the Carmen de diuinaprovidentia, and the Epigram
ma of Paulinus … ) attest to a barbarian presence in southern Gaul. Soz. 9.12.3 explicitly ties that presence to the rebellion of Gerontius against Constantine III in 409. Oros. 7.40.3 condenses more than three years of events into a single sentence which tells us nothing about chronology’.

  72 Oros., 7.28

  73 Kulikowski, 2000, 332.

  74 Drinkwater, 1998, 276.

  75 Drinkwater, 1998, 278.

  76 Oros., 7.40.5.

  77 Oros., 7.40.5.

  78 Drinkwater, 1998, 275.

  Chapter 14

  1 The exact chronology for this section is extremely confusing, especially taking into account the speed of messengers carrying news and orders. The majority of this section is derived from Drinkwater, 1998, passim. Although this interpretation is viable, others can be constructed that differ in large part from this one yet remain valid. See, for example, Heather, 2005, 220f and Liebeschuetz, 2004, 66f plus accompanying footnotes and bibliographies.

  2 The attempt to date these events by Burns using the Codex Theodosianus are, at the very least, unconvincing. All that is certain is that news arrived very quickly that Alaric was still alive: Zos. 5.27.3.

  3 Zos., 5.27.2.

  4 Zos., 5.27.2.

  5 Elton, 2004, 193.

  6 Halsall, 2007, 213.

  7 Honoré interprets passages in the Scriptor Historiae Augustae as evidence of some of the natural assumptions of the Western Senate: 1987, 156f.

  8 Cod. Th., 7.13.8: 7. 20.13.

  9 Wolfram, 1997, 97.

  10 Bury, 1923, 61.

  11 Possible rift between Honorius and Stilicho; Matthews, 1975, 279.

  12 Williams and Friell, 1994, 157. The two consuls were Bassus and Philippus.

  13 Soz., 9.11.4: Or. 7.40.6.

  14 Soz., 9.12.1.

  15 For example, Salway, 1993, 304.

  16 See Chapter 13.

  17 Liebeschuetz, 1990, 66. His claim is based upon the concept that Stilicho needed Alaric’s forces to make good the losses suffered by Sarus, but the chronology seems to work against him.

  18 Zos 5.29.1; cf. for example Kulikowski, 2007, 172.

  19 Possibility of Alaric acting to defend Noricum, Burns, 1994, 215.

  20 Zos., 5.37.1f.

  21 Drinkwater avoids having to make a decision on Sarus’ appointment, leaving it open as to whether Sarus was comes rei militaris or magister militum vacans. For a discussion and further bibliography, see Drinkwater, 1998, 279. n. 63.