Jump to ContentJump to Main Navigation
Late Roman Warlords$

Penny MacGeorge

Print publication date: 2002

Print ISBN-13: 9780199252442

Published to Oxford Scholarship Online: September 2007

DOI: 10.1093/acprof:oso/9780199252442.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 24 February 2017

(p.306) APPENDIX Naval Power in the Fifth Century

(p.306) APPENDIX Naval Power in the Fifth Century

Late Roman Warlords
Oxford University Press

(p.306) APPENDIX

Naval Power in the Fifth Century

Appendix Appendix

All naval historians of the ancient world agree that Roman sea power had been run down to almost nothing by the late fourth century AD,1 and most pass over the next couple of centuries in a few sentences. The only major sea battles in the late antique period were those between the emperors Constantine and Licinius in AD 324, involving over 500 vessels, mainly levies from the ports of the eastern Mediterranean,2 and the naval battle off Italy between the Ostrogoths and the Emperor Justinian's forces in AD 551.3

This disappearance of sea power applies especially to the western empire, but it is difficult to establish just when permanent fleets would have ceased to be available. In the early fifth century AD the Notitia recorded some naval units, both sea and river, in the West,4 but we do not know how run‐down these were, nor how out of date this information was. Some naval writers have considered that the fourth‐century imperial fleets in the Mediterranean were adequate for policing purposes, but not strong enough to deal with barbarian attacks by sea.5 But Vegetius, writing in the late fourth century, actually referred to the (p.307) Roman navy in the past tense6, as did Sidonius in the mid fifth century;7 and Priscus said explicitly that the western Romans lacked a fleet in the 460s.8 On the other hand, Claudian wrote of Stilicho in c. AD 400 preparing a fleet of corn transports and another of warships9 (Claudian is, though, notorious for poetic licence).

The paucity of evidence has not prevented scholars arguing for a fully active fifth‐century navy. Moss, for instance, argued, ultimately unconvincingly, that the western empire still retained a navy and was able to carry out naval blockades in the early fifth century.10 The extreme opposite view was taken by Courtois: that in the West from the Severan period onwards naval tactics were unknown and that state ships were not warships but transports only, either for corn or men.11 Courtois' theory does seem to be supported for the West by some of the scanty information from the late fifth century AD. Marcellinus used ships to transport men to Sicily, but we do not hear of him using them in naval engagement. Ricimer's so‐called naval victory over the Vandals was probably nothing of the sort (see Pt. III, pp. 184 and 186). If he did have naval resources in AD 456, when he defeated the Vandals in Sicily and Corsica, they were not capable of defending Italy against Gaiseric's raids. The fleet built and marshalled by Majorian in AD 460 against Gaiseric (see Pt. III, pp. 206–7) seems to have been intended only to transport his army from Spain to Africa.

(p.308) Fleets for major campaigns had to be assembled by levying ships and crews from the commercial sector,12 combined, if it was important enough, with shipbuilding, as for Majorian's campaign:

Meanwhile, you construct, on two shores, fleets for the Upper and Lower Sea. Into the water fall all the forests of the Apennines; for many a long day there is felling on both slopes of those mountains so rich in ships timber…Gaul…is now eager to gain approval by a new levy for this purpose13

In the sixteenth century AD it took the arsenal at Venice about two years to build a galley of seasoned wood (which in turn takes decades to be ready). In emergencies ships can be made of unseasoned timber in a few months, as above, but they would be of inferior quality and seaworthiness.14 Perhaps a life of one or two seasons was considered adequate for Majorian's ships, or perhaps they had no option but to use unseasoned timber.

From the descriptions by Procopius and others of the fleet sent against Gaiseric in AD 468 (see Pt. I, pp. 56–7), it is clear that Constantinople was still capable of raising a substantial fleet for an important campaign. However, the fleet sent against Gaiseric in AD 468 was not a standing force; Leo ‘collected a fleet of ships from the whole eastern sea and he showed great generosity to the soldiers and sailors…they say he spent one hundred and thirty thousand pounds’.15 The important factors were adequate financial resources and a pool of available seagoing expertise, neither of which was readily available in the West.

The question of the West's naval resources has direct bearing on that of Vandal naval strength. It is beyond dispute that the Vandals made use of substantial numbers of ships. In AD 440 a novella of Valentinian records that Gaiseric had brought ‘a large fleet out of Carthage’16 (to be used to raid the coasts of Italy). He went on to capture Corsica and Sardinia, and make repeated raids on Italy, Sicily, and other Mediterranean coasts. In AD 456 we hear of a Vandal fleet of sixty ships (see Pt. III, p. 184). This evidence has caused some historians to write of widespread Vandal piracy and of the Vandals gaining supremacy at (p.309) sea and control of the Mediterranean.17 An important consideration here is what sort of ships the Vandals had. The only way in which the Vandals could have come into possession of a fleet of warships would have been by capturing them from the western imperial navy in their conquest of Africa. But were such ships stationed there?

The Notitia does not list any naval units stationed at Carthage or elsewhere in western North Africa either; most obviously because there were none, or conceivably as a result of an accident of organization, recording, or textual omission. Orosius and Marcellinus comes did record Heraclian using a fleet to transport his troops from Africa to Rome in AD 413, but this seems to have been the corn fleet. Heraclian is said to have had 3,700 ships! Even Orosius is doubtful about this figure.18 The supply of corn from North Africa to Rome still continued in the mid fifth century, but there is no hard evidence that the late empire maintained a naval force to protect it.19 Common sense suggests that cargoes of grain would not be prime targets for pirates, who would prefer cargoes of valuable luxury goods (and slave‐taking). Courtois argued that the vessels that the Vandals came into possession of were the transports of the corn fleet.20 These were, of course, large, round‐bottomed, sailing vessels, broader in the beam and heavier and much less manoeuvrable than war galleys.21 Courtois concluded that ‘no Vandal navy existed except in the minds of some historians’.22

Transport vessels would have enabled the Vandals to carry the men and horses of their raiding parties to Sicily and Italy and other coastal areas. It is known from Sidonius23 that the raiding Vandal ships carried men and horses. Horses would not be easily transported in fighting galleys. Were cargo ships adapted to enable horses to be loaded and offloaded? For raiders to take horses with them is unusual; most, like the Vikings later, would have rounded up horses when they had landed. The immediate possession of mounted men may well have given the Vandals an advantage. There were horse transports in the (p.310) Hellenistic and Republican periods, which were converted old triremes rowed by only sixty oarsmen and carrying thirty horses in the lower rowing decks.24 Such vessels would have been unable to defend themselves. The eastern empire transported cavalry in the fifth and sixth centuries AD;25 we do not know that these ships were ‘specially designed horse transports’.26 However, reliance mainly on this sort of vessel would not have enabled them to engage in large‐scale piracy, naval actions at sea, or widespread disruption of Mediterranean communications. That the Vandals did not in fact do this in the fifth century AD has, of course, been argued by a number of scholars from Pirenne onwards,27 and has some support from the archaeological record.

In AD 468 Gaiseric defeated a (mainly eastern) naval attack. Do the tactics of that victory provide any information on the type of ships under his command? Procopius (using Priscus) describes how the Vandals took advantage of Basiliscus' hesitation28 to set sail towards the imperial fleet, towing boats which they then used as fire‐ships, causing panic and disorder, then moving in to ram and sink ships.29 Although this is open to different interpretations, on balance it sounds as if Gaiseric did possess some warships.30 Of course, the Vandals had acquired shipbuilding facilities, and even if they had not captured a fleet of warships it is possible that they had had some constructed.31

The evidence overall suggests that while the Vandals possessed some warships (perhaps of the newer type described below), they were primarily a threat as raiders rather than as pirates, mainly using (p.311) their ships to transport forces rather than to attack ships at sea. The Vandals do seem to have developed naval skills (or skilfully used those of the conquered provincials), but their naval superiority may have been more a result of Roman weakness and disorganization than their own expertise.32 Rougé has argued that the fleet of ships which Gaiseric acquired would have deteriorated in quantity and quality over his reign and that of his successors,33 and this is quite possible.

The late imperial period was a watershed between two different types of navy, both in strategic and tactical terms and in terms of ship design. Roman naval squadrons had traditionally been made up of a combination of large triremes (often with platforms for artillery) to which the faster and lighter liburnae had been added. But from the second century AD there was a trend towards the use of general‐purpose military vessels, which could be employed as troop carriers.34 From the third century AD ship designs were moving towards the typical Byzantine type, with high fore and aft posts, and fixed masts, rowed from deck level; described by Orna‐Ornstein of the British Museum as ‘similar to a Viking longship’.35

The early Byzantine navies of the sixth century were made up of cataphract galleys, called dromons, with only one bank of rowers, protected by the decking.36 The Ostrogothic ships built in response to the Justinianic conquest of Italy were similar.37 This was a reversion to an older type of vessel, largely due, no doubt, to lack of resources.38 These dromons were very different in size and fighting power to the war‐galleys of previous centuries, and this affected naval tactics and strategy.39 The ships available to Gaiseric, Marcellinus, and Ricimer may well have been closer to this type of vessel than to those of the old imperial navies.


(1) See e.g. Starr (1941: 198); Lewis (1951: 18–19); Haywood (1999: 57); Casson (1991: 213). To keep warships and crews in a state of readiness and efficiency is a costly business. Well‐made ships of the classical period only had a life of twenty to twenty‐five years (Casson (1971: 90, 119)). Ships could be stripped down and kept in storage in dockyards, in which case they might have been reused up to twenty years later. Casson (1971: 120) gives an example of this (though it dates to 48 BC). A recent book on the Roman army supported the continued existence of a standing fleet into the later fifth century, and explained the contradictory evidence by suggesting that it was not in permanent commission (Elton (1996: 98–9)). However, a fleet consisting of hulls in dockyards, without trained crews, would, given western lack of resources, have been much the same as no fleet.

(2) Zosimus, II, 22, 26.

(3) Procopius, Wars VIII, xxiii, 9–13, 29–38.

(4) The Notitia did not list the main eastern fleets, such as the one known to be based at Constantinople (although it did record some minor eastern units, such as river flotillas in Moesia). This is probably a textual omission.

(5) Lewis and Runyan (1985: 4–8).

(6) Vegetius, 4, 31. He is referring to the fleets at Misenum and Ravenna. However, he described the river patrol boats on the Danube as being in increased use (4, 46). He stated with approval that the Roman empire (presumably in contrast to present‐day practice) did not ‘fit out the fleet on the spur of the moment in response to some crisis, but always kept it in readiness’.

(7) Sidonius, Pan. II, 386: ‘Romula desuetas moderentur classica classes’.

(8) Priscus, frag. 39(1).

(9) Claudian, Cons. Stil. I, 308: ‘duplices disponere classes, quae fruges aut bella ferant’.

(10) Moss (1973: 725). However, the sources he cited are hardly evidence of the existence of a proper navy. His example of Fl. Constantius hampering the Visigoths at Narbonne by ‘forbidding and cutting off all passage of ships and the importation of foreign merchandise’ (Orosius, 7, 43) is ambiguous. The cited case of Stilicho actually says that he ‘shut off the coasts and ports with many guards to prevent access’ (Codex Theod. 7, 16, 1); and the cited blockade of the Vandals in Baetica by Castinus (Hydatius, 68 (76) s.a. 422) seems actually to have been a siege.

(11) By the fifth century, fleets came under the command of army generals; no naval command structure existed.

(12) Which goes some way towards explaining their ineffectiveness.

(13) Sidonius, Pan. V, 441–8: ‘Interea duplici texis dum litore classem inferno superoque mari, cadit omnis in aequor silva tibi nimiumque diu per utrumque recisus, Appennine, latus, navali qui arbore dives … Gallia … hoc censu placuisse cupit nec pondera sentit quae prodesse probat’.

(14) Casson (1971: 120 and n. 82).

(15) Procopius, Wars III, vi, 1–2.

(16) Nov. Val. 9.

(17) See e.g. Schmidt (1911: 309); Stein (1959: 324); Barker (1911: 412); Moss (1973: 728). For Moss, this is vital to his central thesis that Aetius was culpable in not using Roman naval resources (which existed according to Moss) to counter the Vandals.

(18) Orosius (7, 42); and Marcellinus comes (s.a. 413).

(19) See e.g. Schmidt (1924: 306, 309); Moss (1973: 723); Lewis and Runyan (1985: 6).

(20) Courtois (1955: 207). His arguments are very detailed and only reported very briefly here.

(21) Casson (1994: 126).

(22) Courtois (1955: 207).

(23) Sidonius, Pan. V, 395–400 (see Pt. III, pp. 203–4).

(24) Morrison and Williams (1968: 93).

(25) Menander, frag. 23 (1); John Lydus, De Mag. III, 43; Procopius, Wars III, xi, 2–16. In the last case, the way in which their capacity is given suggests that they were grain carriers.

(26) As described by Elton (1996: 98).

(27) Pirenne (1922: 223–35); Baynes (1955: 309–16). More recent writers have not taken such a strong line as Pirenne: e.g. Lewis and Runyon (1985: 10); Hodges and Whitehouse (1983: 169). Interestingly, the latter detect a brief revival in Mediterranean economy and communications in the mid to late fifth century; in other words, during the dominance of Aetius, Ricimer, and Odovacer (Hodges and Whitehouse (1983: 52)).

(28) Bass (1972: 134) suggested the Romans were trying to avoid a sea battle.

(29) Procopius, Wars III, vi, 12–24.

(30) As Gautier concluded, there is not enough evidence to be certain of the size or nature of Gaiseric's fleet (1932: 217–21), but it is hard to believe he did not have some warships not reliant on wind power (1932: 219).

(31) Construction did not cease at Carthage under the Vandals (Courtois (1955: 208)). Victor Vitensis (III, 20) recorded prisoners being sent to ‘cut timber for the king's ships’ under Gaiseric, or his son.

(32) Lewis and Runyan (1985: 11).

(33) Rougé (1961: 135–7). This would be a factor in the ease of the Justinianic reconquest.

(34) Orna‐Ornstein (1995: 7). According to Elton (1996: 97–8), late Roman navies consisted of both warships and transports for troops and supplies (with sails only and unable to defend themselves). In the early fifth century AD, up to 5,000 men could be transported (Claudian, De Bell. Goth. 418–23; Zosimus, 6, 8, 2). This capacity probably included requisitioned merchant ships.

(35) Orna‐Ornstein (1995: 7). This type of vessel reached its greatest size and efficiency in the tenth century AD (Lewis and Runyan (1985: 30–1)).

(36) Procopius, Wars III, xi, 15–16.

(37) Procopius, Wars VIII, xxii, 17; III, xxiii, 29–38.

(38) Casson (1971: 148) and (1994: 95–6); Bass (1972: 134). The dromon relied greatly on speed (the ram was the main offensive weapon) and carried only a small contingent (100–200 men) armed with light weapons.

(39) Rougé (1961: 150).