MEDIEVAL AND RENAISSANCE STUDIES
Volume 35 (2004)
PUBLISHED UNDER THE AUSPICES OF
THE CENTER FOR MEDIEVAL AND RENAISSANCE STUDIES
UNIVERSITY OF CALIFORNIA, LOS ANGELES
THE CAROLINGIAN ARMY AND THE STRUGGLE
AGAINST THE VIKINGS
by Simon Coupland
There have been many previous studies of the Carolingian army, but none
examining the ninth-century armies which faced the Viking invasions. Earlier works
have tended to focus on the age of expansion in the eighth century, particularly the
capitularies and campaigns of Charlemagne.' Yet the armies which fought
Charlemagne's offensive campaigns differed significantly from those which
defended the Empire against the Vikings in terms of their composition, size,
armament and purpose. The present study will thus begin by considering the
different types of defensive force which opposed the invasions: the host, the coast
guard and the lantweri. The royal army has naturally been the focus of previous
studies, but against the Viking incursions the initial line of defense was the coastal
guard, both at sea and on land. Another form of resistance was that offered under the
lantweri, the defense of the realm in case of invasion, which was obviously of
primary importance against the Vikings, but which has been accorded scant attention
in the past. The article will then turn to three more general issues: the size of the
Frankish armies which fought the raiders, the leadership and muster of those armies,
and the particular role played by the cavalry. The final three sections will address the
range of military tactics employed against the Northmen: the strategy of
containment, by which the Franks attempted to confine Scandinavian fleets to the
rivers; siege warfare; and finally pitched battle. Throughout the study a more
fundamental question will also be kept in view, namely, why it was that the a11-
conquering war machine of Charlemagne's time apparently found itself unable to
deal with the Scandinavian incursions barely half a century later.
After a long period in which the only interest shown in Carolingian military matters
was in the form of short survey articles, two significant recent books have
turned the spotlight on warfare in the Carolingian period. The first, by Bernard
Bachrach, considers the Frankish military in considerable depth, though again it
concerns itself primarily with the eighth rather than the ninth century.i The second is
Guy Halsall's important survey of warfare and society in the early medieval period,
including discussion of the ninth century and paying particular attention to the
Viking raids.' However, the fact that Halsall discusses a 450-year period across the
whole of Western Europe allows the present study to go into detail where he uses a
'se, for example, an article about the Carolingian anny by F.-L. Ganshof contained such a preponderance
of references to the reign of Charlemagne that the English translation was entitled "Charlemagne's
Anny": "L'armee sous les carolingiens" in Ordinamentl militari in occidente nel/'alto medioevo, 2
vols., Settimane 15 (1968) 1.109-130, trans. in F.·L. Ganshof, Frankish Institutions under
Charlemagne (New York 1970) 57-68.
2Bemard S. Bachrach, Early Carolingian Warfare: Prelude to Empire (Philadelphia 2001).
3G. HaIsaII. Warfare and Society in the Barbarian West, 450--900 (London 2003).
50 SIMON COUPLAND
broader brush; and the conclusions reached here will frequently complement those in
his work. As for earlier studies of the Carolingian army, a valuable detailed survey
can be found in a bibliographical article by France published in 2001, though as was
noted earlier, the great majority of these sources concentrate on the era of
Charlemagne," The occasional references to the struggle against the Vikings in these
previous works tend to be brief and relatively superficial.' One incident which has
received unusually close attention is a battle between Franks and Vikings at the Dyle
in 891, whose significance for the role of cavalry will be discussed below, but this is
a rare exception." The present study is consequently long overdue. It should be
emphasized that this article will focus exclusively on the armies which opposed the
Viking invasions; broader defensive strategies employed by the Franks, such as the
fortification of bridges, the payment of tributes and the hiring of Viking leaders as
mercenaries, have all been discussed elsewhere, as has the role of the church in
resisting the invasions.'
MILITARY SERVICE (1): COASTAL DEFENSE8
The first reference to defensive measures taken specifically against sea-borne
raiders from Scandinavia dates from 800, when Charlemagne ordered a fleet to be
stationed on the North Sea coast against "pirates."? Ten years later further measures
were taken, and in 811 the emperor inspected ships of the newly constructed
fleet at Ghent and Boulogne." Despite this, there is no record of a North Sea fleet
ever having seen action, and in 820 it was the shore-based guards who repelled an
attempted Viking raid on Flanders and the Seine mouth. I I A North Sea fleet was
not mentioned again until 837, when Louis the Pious ordered the construction of
ships in Frisia to counter repeated Viking raids on the region around Dorestad.
FCJ.France, "Recent writing on medieval warfare: from the fall of Rome to c. 1300," The Journal of
Military History 65 (April 200 I) 441-473, esp. at 445-450.
'Bachrach, Early Carolingian Warfare (n. 2 above) mentions them just five times in his index; see
also C. Oman, A History of the Art of War in the Middle Ages, 2nd ed., 2 vols., (London 1924) 1.101-
107; F. Lot, L 'Art militaire et les armees au moyen age en Europe et dans le proehe Orient, 2 vols.,
(Paris 1946) 1.105-111; J. F. Verbruggen, "L'art militaire dans I'empire carolingien (714-1000),"
Revue beIge d'histoire militaire 23.4 (1979) 289-310, at 302-306; 23.5 (1980) 393-412, at 406-407;
P. Contamine, "L'espace carotingien: dilatation, dislocation, invasion" in P. Contamine, ed., Histoire
militaire de la France, 4 vols. (paris 1992-1994) 1.36-42.
6See below, "The role of the cavalry."
's. Coupland, "The fortified bridges of Charles the Bald," Journal of Medieval History 17 (1991)
1-12; "The Frankish tribute payments to the Vikings and their consequences," Francia 26/1 (1999)
57-75; "From poachers to gamekeepers: Scandinavian warlords and Carolingian kings," Early Medieval
Europe 7.1 (1998) 85-114; "The rod of God's wrath or the people of God's wrath? The Carotingians'
theology of the Viking invasions," Journal of Ecclesiastical History 42 (1991) 535-554, esp.
547-553. For the construction of fortifications against the Vikings, see F. Vercauteren, "Comment
s'est-on defendu, au IXe siecle dans I'empire franc contre les invasions normandes?" Annales du AXXe
congres de la Federation archeologlque et historique de Belgique (1935-1936) 117-132; R. M. van
Heeringen, P. A. Henderikx and A. Mars, eds., Vroeg-Middeleeuwse ringwalburgen in Zeeland (Goes
and Amersfoort 1995); and S. Coupland, "Charles the Bald and the Defence of the West Frankish
Kingdom Against the Viking Invasions, 840-877" (Ph.D. diss., Cambridge University 1987) 167-186.
·See also H. Sproemberg, "Die Seepolitik Karls des Grossen" in Beiträge zur belgisch-niederländischen
Geschichte (Berlin 1959) 1-29; and John Haywood, Dark Age Naval Power (London and
New York 1991) 118-130, both of which concentrate on the first quarter of the ninth century.
9Annales Regni Franeorum [ARF] 800 and 800 (revision); ed. F. Kurze, MGH Scriptores rerum
Germanicarum in usum scholarum (hereafter MGH SSrG) (Hanover 1895) 110-111.
IOCapirulare missorum Aquisgranense primum c. 16: MGH Capitularia regum Franeorum (hereafter
MGH Cap.) J.J53;ARF811 (n. 9) 135.
IIARF820 (n. 9 above) 153.
12Annales Bertiniani [AB] 837: ed. F. Grat, 1. Vielliard, and S. Clernencet, Annales de Saint-Bertin
THE CAROLINGIAN ARMY 51
This implies that the existing coastal defenses, which had been reorganized in 835
and again in early 837,13 did not include a fleet at that time. Although the ships
were definitely built," there is no reference to them actually being deployed
against the Viking raids, and the lack of an attack in 838 was said to be due to a
storm which sank part of a Scandinavian fleet on its way south." There is indeed
no indication that any Carolingian ruler ever used a naval force at sea against
Scandinavian attack, in contrast to the explicit references to the deployment of
Carolingian fleets against Arab raiders in the south 16 and also to the naval
victories won by Kings Ethelred and Alfred against the Vikings on the other side
of the Channel."
On land, it was the coast guard which had the dual function of keeping watch
for sea-borne raiders and defending the shore if such raiders attempted to land.
Although no one text describes how the system functioned, a number of details
can be gleaned from various sources. For instance, in 800 all the North Sea ports
and river mouths navigable to ships were guarded by detachments of troops. IS
Where such guards might come from is revealed in a letter of Einhard, abbot of
St.-Bavo in Ghent, in which he reported that the abbey's men had been unable to
attend the autumn assembly of 832 because they were serving in the coast guard.
Einhard complained that it was unfair to fine these men for failure to perform their
military service when they were defending the coast at the emperor's command."
The maintenance of the coastal defense was evidently the responsibility of the
aristocracy, for in 865 Pope Nicholas I wrote that "the majority of the bishops and
other royal vassals are guarding the coast day and night against sea-borne
raiders," in this case undoubtedly from Scandinavia.2o If danger threatened,
everyone living along the coast had to turn out to support the regular guards. Free
Franks who failed to respond faced a fine of twenty solidi, while the unfree would
pay half this sum and receive a flogging.i' To sound the alarm, some sort of
signaling network was employed, though it is unclear whether it involved
beacons, flags, bells, or some other means of communication."
The potential effectiveness of the system was illustrated in 820, when thirteen
Danish ships tried to plunder the coast of Flanders. They were repelled, managing
to steal only a few cattle, and so sailed on to the mouth of the Seine, where they
met even stiffer resistance from the coast guard, losing five of their men in a skirmish.
In the end they were forced to travel as far as Bouin in Aquitaine to find a
(paris 1964) 22; ed. and trans. J. L. Nelson, The Annals of St-Bertin (Manchester 1991) 37.
13AB 835, 837 (n. 12): Grat 18,21; Nelson 33, 37.
"Annales Fuldenses [AFJ 838, ed. F. Kurze, MGH SSrG (Hanover 1891) 28; T. Reuter, ed. and
trans., The Annals of Fulda (Manchester 1992) 15.
"AB 838 (n. 12 above): Grat 24; Nelson 39.
'~aywood, Dark Age Naval Power (n. 8 above) 113-116.
"Anglo-Saxon Chronicle 851, 875, 882, 885, 896: D. Whitelock, D. C. Douglas and S. I. Tucker,
eds., The Anglo-Saxon Chronicle (London 1961) 43, 48, 50, 51, 57-58.
'IARF 800 (n. 9 above) 110; Einhard, Vita Karoli Magni c. 17, ed. F. Kurze, MGH SSrG (Hanover
'9Einharti epistolae no. 23: MGH Epist. Karo!. Aevi 5.121; trans. in P. E. Dutton, Charlemagne's
Courtier (Ontario 1998) 139.
7I:iNicolaiI epistolae no. 38: MGH Epist, Karo!. Aevi 6.309.
2'Capitulare missarum specialia (802) c. 13: MGH Cap. 1.100-101.
~ithard, Historiarum libri Illl 3.3: P. Lauer, ed., Nithard: Histoire des fils de Louis le Pieux
(paris 1926) 94. The word signum can have all these meanings: J. F. Niermeyer, Mediae Latinitatis
lexicon minus (Leiden 1976) 971.
52 S[MON COUPLAND
site which was sufficiently poorly defended for them to pillage.P Yet by the late
830s the increasingly successful Scandinavian raids on Frisia and Aquitaine indicate
that the system was breaking down, and after 840 Viking fleets apparently
penetrated Frankish rivers at will, with repeated raids on the Seine, Loire, and
elsewhere.r' A monk at Noirmoutier, Ermentarius, blamed the civil war: "Their
disagreement gives strength to foreigners ... the defense of the shores of the ocean
is abandoned.?" Another monastic chronicler, this time on the Seine, blamed the
cowardice of the magnates, who "prepared to flee rather than to resist, being absolutely
terror-struck.?" A third factor, not mentioned by contemporaries, was the
increased size of the Viking fleets. Whereas earlier raiders had operated in small
groups-such as the thirteen vessels of 820 or the nine which raided Noirmoutier
in 835-a fleet which attacked Nantes in 843 numbered sixty-seven ships, and
two years later 120 ships entered the Seine." Nevertheless, in 854 Charles the
Bald issued a capitulary decreeing that the coastal defense should be deployed
with the usual vigilance," and the papal letter cited earlier proves that some kind
of defense was still being offered in the 860s. The Flemings also repelled an
attempted Viking landing on their coast in 864,29 but the regularity and apparent
ease with which the Vikings entered Frankish rivers suggests that-as
Ermentarius of Noirmoutier claimed-the coastal guard was largely inoperative or
ineffectual, so that increasingly the fight had to be taken to the Vikings on land.
If a Scandinavian fleet overcame any resistance offered by the coast guard and entered
Frankish territory, the king might muster the royal host to oppose them.
Mobilization of the army could be cumbersome and slow, however, and it has been
claimed that this was one of the reasons for the Vikings' success." What this
argument overlooks is that there was a second, much more rapid type of
mobilization, specifically intended to counter invasion. This was the lantweri, when
"all shall come to the defense of the fatherland, without any excuse.,,32 The
distinction between the usual situation and the lantweri, "that is, the defense of the
2JARF820 (n. 9 above) 153-154.
2"1be raids in the 840s are discussed exhaustively in Coup land, "Charles the Bald" (n. 7 above) 7-
2'Ennentarius, De translationibus et miraculis sancti Filiberti, preface to bk. 2, in R. Poupardin,
ed., Monuments de /'histoire des abbayes de Saint-Philibert (Paris 1905) 60.
26Translatio sancti Germani Parisiensis c. 3: Analeeta Bollandiana 2 ([883) 72.
27ARF 820 (n. 9 above) 153; Ermentarius, preface to bk. 2, 2.11 (n. 25 above) 59, 66-67; AB 845
(n. 12 above): Grat 49, Nelson 60. See further S. Coup land, "The Vikings in Francia and Anglo-Saxon
England to 911" in R. McKinerick, ed., The New Cambridge Medieval History 2 (Cambridge 1995)
190-201, at 194.
28Capitulare missorum Attiniacense c. 2: MGH Cap. 2.277.
29AB 864 (n. 12 above): Grat 113, Nelson 118.
3'Mentioned briefly, without discussion, by Bachrach, Early Carolingian Warfare (n. 2 above) 53-
54; Contarnine, "Espace carolingien" (n. 5 above) 399-400; and T. Reuter, "The End of Carolingian
Military Expansion" in P. Godman and R. Collins, eds., Charlemagne's Heir. New Perspectives on the
Reign of Louis the Pious (Oxford 1990) 391-405; idem, "The recruitment of armies in the Early
Middle Ages: what can we know? in Military Aspects of Scandinavian Society in a European
Perspective, AD 1-/300 (Copenhagen 1997) 32-37, at 34-35.
llA claim made, e.g., by Oman (n. 5 above) 1.101; H. Delbrück, Geschichte der Kriegskunst im
Rahmen der politischen Geschichte, 7 vols. (Berlin 1900-1936) 3.76; A. d'Haenens, Les Invasions
normandes, une catastrophe? (Paris 1970) 55.
12Edictum Pistense (864) c. 27: MGH Cap. 2.322.
THE CAROLINGIAN ARMY 53
fatherland/'" is most clearly expressed in Charles the Bald's proclamation at
Meerssen in 847: "It is our wish that a man serving anyone of us [se, Charles,
Lothar I, or Louis the German] ... shall join the army with his lord or perform his
other duties, unless, God forbid, an invasion of the land should occur, which is
called lantweri, in which case all the people of that land shall turn out together to
repel it.,,34 The lantweri thus involved the conscription of the entire Frankish
population, both those who regularly served in the host and those who were usually
exempt. Even those who were normally forbidden from bearing weapons on
penitential grounds were permitted to take up arms against pagan invaders; at times
like these every able-bodied fighting man was expected to turn OUt.35 Another
example of the way in which the customary criteria for conscription were ignored
under the lantweri can be found in the Capitulary of Quierzy of 877. In the decree,
provision was made for any royal vassal who might wish to relinquish his benefice
to his heir on the king's death, declaring that the only duty required of such a man
would be to turn out in defense of the realm."
The procedures to be followed in the event of invasion were laid down in detail
in a capitulary of 865.37 The missi in charge of the region under attack were
responsible for assembling an army to oppose the invaders. The local bishops,
abbots, and abbesses were instructed to send their full complement of troops, fully
equipped and led by a standard-bearer, while counts and royal vassals were to
command their contingents in person. Although the capitulary referred to the
defense of Burgundy against hostile incursions from neighboring lands, similar steps
were almost certainly taken when any local army was raised to resist the Vikings. It
has been suggested that the aristocracy were less motivated to defend their own
kingdom than when taking part in potentially profitable foreign campaigns.P but the
fact that these men were defending their own territories and ultimately their own
properties seems to me to outweigh such considerations. Likewise, although such
forces were undoubtedly less well equipped than the regular army, since many of the
common people were too poor to afford much more than a spear or bow," the fact
that the men were defending their own region with their local leaders means that
their morale and loyalty may have been as high as or even higher than that of the
royal host. It would also have been possible to mobilize such troops much more
quickly than the regular army.
Many of the recorded instances of opposition to the Vikings almost certainly represented
examples of the lantweri, even if the term was not used in contemporary
accounts. For instance, in 869 the Loire Vikings were defeated by the Transsequanani,
that is, the men from between the Seine and the Loire, under the command of
3~is definition is found in an otherwise unknown capitulary included in Paris, BN lat. 10758 and
4628A, reproduced in MOH Cap. 2.71, note.
)4Hlotharii, Hludowici et Karoll conventus apud Marsnam primus, Adnuntiatio Karoli c. 5: MGH
3lK. J. Leyser, "Warfare in the Western European Middle Ages: The Moral Debate" in K. J. Leyser,
Communications and Power in Medieval Europe: The Gregorian Revolution and Beyond (London
1994) 189-203, at 197. Cf. Coupland, "Rod of God's wrath" (n. 7 above) 547-548.
36Capitulare Carisiacense c. 10: MGH Cap. 2.358.
37Capitulare Tusiacense in Burgundiam directum c. 13: MGH Cap. 2.331.
31HalsaII, Warfare and Sociery(n. 3 above) 134, cf. 90-91.
39S.Coupland, "Carolingian arms and armor in the ninth century," Vialor 21 (1990) 30, 42, 46-47,
54 SIMON COUPLAND
the local counts, Hugh and Gauzfrid." Similarly, in 880 Abbot Gauzlin of St Denis
"sent word to the people on the other side of the Scheldt that they should come on an
appointed day, and with one group on each bank of the river, they would between
them annihilate the Northmen.?" In areas where royal authority was weak, local
resistance might be organized independently. This seems to have been the case in
867, when the Viking chieftain Rorik was driven out of Frisia by the local inhabitants,
as a result of which Lothar called up the host to defend the kingdom against
Rorik's expected return.f Local initiatives were not always so successful, however.
In 880 Gauzlin's army not only failed to achieve any success, but only just managed
to escape by fleeing in disgrace, and a great many of them were captured or killed.
The potential weakness of these local forces was even more forcefully brought out in
882, when the inhabitants of the region around Prilm banded together to try to drive
off a Viking army. As Regino commented, because they were "not so much
unarmed as lacking any military discipline," they were slaughtered by the Northmen
"not like human beings but like brute beasts.?" In 859 it was the common people
between the Seine and the Loire who were massacred, not on this occasion by
Scandinavians, whom they had actually defeated in battle, but by the local Frankish
magnates. The reason was that exasperation with their leaders' failure to offer
resistance had led the local people to form a sworn association (coniuratio) to fight
the Vikings, and such armed groups were seen as too dangerous to the authority of
the establishment to leave unchecked."
MILITARY SERVICE (3): THE HOST4S
It is clear from contemporary sources that the entire army of the kingdom, the royal
host, was called up to fight the Vikings on numerous occasions. To give two early
examples, in 845 Charles the Bald "ordered that the whole army of his kingdom
should, once summoned, mass together to fight," and in 852 Charles and Lothar I
besieged the Vikings at Jeufosse ''with their entire army.?" Arrangements for
conscription were largely unchanged from earlier in the century, as is evident from
the repetition of earlier capitularies." Every man who could afford to go on campaign
was ordered to join the army, and those owning a horse had to bring a mount
as well. The less well-off had to pool their resources, enabling one man to serve with
the help of up to four others, while the poorest free men were exempt from
conscription. Even so, a point which has been insufficiently appreciated in the past is
that even these pauperi played their part in financing the army's campaigns through
the payment of the army tax, the hostilitium. A number of references in ninthcentury
texts suggest that this tax was levied on those who owned property worth
less than one pound.
The first indication is in the Edict of Pitres of 864, in a decree about the heriban-
4°AB 869 (n. 12 above): Grat 166, Nelson 163.
41Annales Vedastini [AV] 880, ed. B. de Simson, MGH SSrG (Hanover and Leipzig 1909) 47, 48.
42AB 867 (n. 12 above): Grat 137, Nelson 139-140.
4lRegino ofPrüm, Chronicle 882, ed. F. Kurze, MGH SSrG (Hanover 1890) 118.
«AB 859 (n. 12 above): Grat 80, Nelson 89. SeeJ. L. Nelson, Charles the Bald (London 1992) 194.
45A fuller discussion of the question of military service can be found, with references to the secondary
literature, in Coupland, "Charles the Bald" (n. 7 above) 85-93; see also Halsall, Warfare and
Society (n. 3 above) 93-101.
46Translatio sancti Germani c. 12 (n. 26 above) 78; AB 852 (n. 12 above): Grat 65, Nelson 75.
47E.g., Praeeeptum pro Hispanis (844) c. 1 (MGH Cap. 2.259) repeats a capitulary of 815: MGH
TIlE CAROLINGIAN ARMY 55
num. Originally a fme on those who failed to perform their military service, this
came to signify a payment in lieu." The full heribannum was a fme of sixty solidi,
but the Edict of Pitres repeated a system of graded penalties which required
individuals owning property worth more than one pound to pay according to their
means/" Those free men who were even poorer than this were apparently exempt,
suggesting that they were not required to serve in the host. Secondly, a tribute
payment to the Vikings in 866 differentiated between those who paid the
heribannum and those who farmed free mansi, suggesting that the latter did not
serve in the army;" and an Italian capitulary of the same year similarly distinguished
between those who could afford to serve in the host, either independently or with
assistance, and ''the poor," who were required to defend the coast and country, or, if
they owned less than ten solidi, exempted from any duties whatsoever/" Ninth century
polyptychs likewise record that differing military obligations were required
of tenants according to their means. The wealthy few served in the host, as for
example those at St, Maur-des-Fosses who owed ''the heribannum for two oxen,
twenty solidi."52 The vast majority of free tenants, however, paid a tax to the army,
the hostilitium, evidently instead of serving in the host themselves. The maximum
appears to have been four solidi, just below the smallest amount levied under the
heribannum, five solidi.s3 Given that the latter sum was exacted from men owning
property worth only one pound, the obvious conclusion is that those who paid the
hostilitium possessed even less than that. The overwhelming majority of free tenants
on the great ecclesiastical estates were consequently too poor to serve in the host,
even with the assistance of others, but paid the army tax instead.
One other aspect of the hostilitium is significant in the present context. Records
from the early ninth century indicate that the tax was originally paid in oxen and
carts," and its primary function was evidently to provide the army supply train.ss
Even at the beginning of the century, however, the payment could be in cash rather
than in kind,56and this had apparently become the norm by the time of the Viking
invasions. Even so, the carts and oxen which Charles the Bald sent to the
fortification work at Pont-de-I'Arche in the 860s were presumably provided through
the collection of the hostilitium.S7
According to the Edict of Pitres, the hostilitium was not the only duty imposed
upon those free Franks ''who cannot join the army." They were also required to
work on the construction of new fortifications, bridges and swamp crossings, and to
Cap. 1.261; Edictum Pistense (864) c. 27 (MGH Cap. 2.321) reproduces one of829: MGH Cap. 2.7.
"See, e.g., Capitulare Bononiense (811) cc. 1-2: MGH Cap. 1.166; Nienneyer (n. 22 above) 481.
.9Edictum Pistense c. 27: MGH Cap. 2.322.
50AB 866 (n. 12 above): Grat 126, Nelson 130. See Coupland, "Tribute payments" (n. 7 above) 62-
51Constilutio de expeditione Beneventana e. 1: MGH Cap. 2.94.
52C.6: B. Guerard, ed., Polyptyque de l'abbe Irminon, 2 vols. (Paris 1845) 2.284.
53Coupland, "Tribute payments" (n. 7 above) 64.
54F._L. Ganshof, ed., "Le polyptyque de I'abbaye de Saint-Bertin (844-859)," Memoires de
l'institut national de France, Academie des inscriptions et belles-lettres 45 (1975) 57-200, at 80, 81,
82 (twice) 83, 84, 85, and St. Maur-des-Fosses: Guerard (n. 52 above) 2.285 (c. 9).
55A. Longnon, ed., Polyptyque de l'abbaye de Saint-Germain des Pres, 2 vols. (Paris 1886-1895)
1.121-123; 1. Durliat, "Le polyptyque d'Inninon et l'impöt pour I'armee," Bibliotheque de l'ecole des
chartes 141 (1983) 185, 197; see also Nienneyer (n. 22 above) 504.
56t.ongnon (n. 55) 1.122.
57AB 869 (n. 12 above): Grat 153, Nelson 153-154; cf. AB 862, 866 (carts alone): Grat 91, 127,
Nelson 100, 131.
56 SIMON COUPLAND
perform guard duty in forts and border regions." Charles the Bald emphasized that
these duties were required "following ancient practice and that of other nations"
(iuxta antiquam et aliarum gentium consuetudinem). The reference to "other
nations" probably refers to the Anglo-Saxons, while "ancient practice" may refer to
established Frankish custom-a charter for Metz cathedral of 775 described the
duties of the church's free tenants as military service, guard duty and work on
bridges-or possibly to late Roman precedents.i" At this time, 864, the king was
undoubtedly keen to assemble sufficient manpower to continue the important
defensive works at Pont-de-l'Arche." while the bridges and swamp crossings served
to increase the army's mobility against the Viking raiders.
THE SIZE OF ARMIEs
The size of Carolingian armies has long been a subject for debate, with the
discussion typically focusing on the size of the expansionist forces of Charlemagne
in the late eighth century. On the assumption that the Carolingian administration was
too primitive to raise large forces, DelbrUck offered the guesstimate of 5000 to 6000
troops in anyone army, though this figure appears to have been plucked from thin
air.61 It has been roundly rejected by Bachrach, whose own estimate of "a total
mobilization of armies for expeditionary operations on all fronts in the 100,000
range" would nevertheless have few supporters.f Verbruggen proposed that
Charlemagne's armies might have contained 2500 to 3000 cavalry and between
6000 and 10,000 infantry, a figure followed by Contamine, though again neither
offered any real justification for these figures." A minimalist view has been put
forward by Reuter, proposing armies "numbered in two or at most three figures,"
though he did emphasize the difficulty if not impossibility of assessing the size of
anyone particular army."
These variations are due in large measure to the paucity of figures in contemporary
sources, and the likely unreliability even of those that are reported. From the
ninth century I am aware of only two authors who specified the number of troops in
a Frankish army, Regino of Prüm and Abbo of St. Germain. Regino stated that
Charles the Bald had over 50,000 men at Andernach in 876, but the qualifying
phrase "so people say" (utfenmt) and Regino's unreliability about events earlier in
the century raise considerable doubts about the accuracy of his report." Figures
5BEdictum Pistense C. 27: MGH Cap. 2.321-322. For other instances of civitas meaning "fortification"
see, e.g., ARF 809,817 (n. 9 above) 129, 147.
59C. GiIlmor, "Charles the Bald and the small free farmers, 862-869" in Anne Nergärd Jergensen
and Birthe L. Clausen, eds., Military Aspects of Scandinavian Society in a European Perspective, AD
1-/300 (Copenhagen 1997) 38-47, at 39-40; J. L. Nelson, 'Translating Images of Authority: the
Christian Roman Emperors in the Carolingian World" in M. M. MacKenzie and C. Roueche, eds.,
Images of Authority; Papers Presented to Joyce Reynolds on the Occasion of her 70th Birthday
(Cambridge 1989) 194-205, at 197. •
6OCoupland, "Fortified bridges" (n. 7 above) 4-6; C. Gillmor, "The logistics of fortified bridge
building on the Seine under Charles the Bald," Anglo-Norman Studies 11 (1988) 87-106, at 99-106.
6JDelbrUck, Geschichte der Kriegskunst (n. 31 above) 3.16.
62B. S. Bachrach, "Early Medieval Military Demography: Some Observations on the Methods of
Hans DelbrUck" in D. J. Kagay and L. J. A. Villaion, The Circle of War in the Middle Ages (Woodbridge
1999) 3-20; Early Carolingian Warfare (n. 2 above) 58.
63J. F. Verbruggen, 'L'armee et la strategie de Charlemagne" in W. Braunfels, ed., Karl der Grosse,
Lebenswerk und Nachleben, 4 vols. (DUsseldorf 1965) J.42a-436; Contamine, "Espace carolingien"
(n. 5 above) 30.
64Reuter, "Recruitment of armies" (n. 30 above) 36.
·'Regino 874, 876 (n. 43 above) 107, 112. On Regino's unreliability see at nn. 79, 135, and 136 be
THE CAROLINGIAN ARMY 57
given by Abbo in his poem about the siege of Paris are scarcely more credible,
including the claim that a Frankish army of just 1000 men defeated 19,000 Vikings
at Montfaucon in 888.66 Although no concrete evidence exists to contradict Abbo's
figure, throughout his work the Franks are said to have overcome massively superior
odds: in the most extreme case a force of at most 200 Franks is supposed to have
resisted a force of 40,000 Vikings."
This highlights the hyperbole in many contemporary accounts. For instance,
Hildegarius of Meaux described the army which Charles the Bald mobilized against
a Viking incursion in 845 in the following terms: "We can but marvel at how large
an army the kingdom fruitlessly brought there with its king, Charles: the earth could
scarcely sustain it, and overshadowed the sky under its covering, though we do not
know how many there were.'.68 Given that the Viking fleet in question numbered
120 ships, it is likely that the Frankish force was a sizeable one, but hardly of the
magnitude suggested by Hildegarius. As Halsall has recently underlined, reflection
shows that medieval clerics are unlikely to be reliable guides as to the true size of
contemporary armies, for even if they had the testimony of eyewitnesses, the
numbers involved were too large, their ability to estimate too poor, and their concern
for precis.e accuracy too Ii'mi.te d.69
Given this dearth of trustworthy material, some authors have tried a more deductive
approach. Ferdinand Lot drew on nineteenth-century census figures to assess the
potential number of West Frankish troops from the area of Francia he regarded as
loyal to the king, but unfortunately his unsupported assumptions about the criteria
for conscription and the equivalence of the rural population in the ninth and
nineteenth centuries render his conclusions worthless.I" Karl Ferdinand Werner
consequently put forward two other methods for estimating the size of Carolingian
armies." The first was to multiply the number of districts (pagl) by an estimated
minimum of fifty horsemen per district; the second to assume that each royal vassal
could have mustered at least twenty mounted followers, a figure amply supported by
contemporary sources. These calculations were not intended to determine the size of
any particular army, but rather to give a ball-park estimate of the kind of numbers
involved. If we apply Werner's calculations to the supposedly vast army assembled
in 845, they suggest that Charles the Bald had a potential pool of 7500 to 13,000
horsemen, not including footsoldiers." Given that not all the available troops would
low; also S. Coupland, "The Vikings on the Continent in Myth and History," History 88 (2003) 187-
"Abbo, Bella Parisiacae urbis 2, lines 492-493, 496-497: H. Waquet, ed., Abbon: Le siege de
Paris par les Normands (Paris 1964) 102, 104.
67Abbo I, lines 114-115 (n. 66) 24. This figure is also quoted in N. P. Brooks, "England in the
Ninth Century: the Crucible of Defeat," Transactions of the Royal Historical Society, 5th ser., vol. 29
(1979) 1-20, at 4, who judged Abbo "in a class of his own as an exaggerator": 6. Wallace-Hadrill was
unwilling to believe that Abbo was "simply employing a literary device to highlight their heroism" but
it seems to me that in this instance at least that is precisely what Abbo was doing: J. M. Wallace-
Hadrill, The Vikings in Francia (Reading 1975) 5.
61Vita Faronis c. 122: MGH SS rer. Mer. 5.200.
6'11alsall, Warfare and Society (n. 3 above) 12Q...124.
wr.ot, Art militaire (n. S above) 1.94-98.
71K. F. Werner, 'Heeresorganisation und KriegfUhrung im deutschen Königreich des 10. und 11.
Jahrhunderts" in Ordinamenti mi/itari (n. Iabove) 2.791-843, at 813-822.
72Based on ca.ISO West Frankish pagi or some 650 royal vassals, i.e., ISO counts, 300 vassi
dominici (following Wemer, who doubled the number of counts), 70 bishops, and an estimated 130
58 SIMON COUPLAND
have mustered on any given occasion, the West Frankish army would have been
smaller than this.
A further pointer to the scale of Frankish armies is the size of the Viking forces
they opposed. Again contemporary figures must be handled with caution: for example,
when the Viking leader Rodulfwas killed on Oostergo in 873, were 500 Danes
killed with him, as the Annals of St. Bertin and Annals of Xanten reported, or 800, as
the Annals ofFulda claimed?73The correct answer is undoubtedly "a large number,"
probably several hundred; we cannot be more specific than that, nor should we try to
be. Although Sawyer proposed in 1971 that even the Scandinavian "Great Army"
consisted of no more than a thousand men, this has since been demonstrated to
have been an underestimate, and a more likely figure is that the armies contained
thousands, rather then hundreds, of men." This is consistent with most contemporary
reports of Viking fleet sizes, casualty figures, and references to raiding parties
such as those on the Seine in 865 numbering 200 and 500 men, or at St. Omer in
891 consisting of 550 men.7S The figures on the Frankish side are likely to have
been similar: armies numbering in their thousands, but not tens of thousands, a conclusion
also reached by Halsalllooking at a broader range of evidence." At the same
time it is important to bear in mind that the Frankish armies which opposed the Vikings
were not always the entire host, but individual squadrons (scarae) which presumably
numbered only in the hundreds. For instance, in 868 Charles the Bald sent
his son Carloman into Neustria with a single squadron, and in 871 it was only the
Neustrians whom Gauzfrid and Hugh the Abbot led against a Viking camp on the
Finally, one ninth-century text which has been overlooked in previous
discussions provides valuable support to these conclusions. In a detailed account of
the battle of Saucourt in 881, the Annals of SI. Vaast described how a few Northmen
attacked the Franks, "and killed many of them, namely up to a hundred men.,,78
Whether plures is understood as relative to the size of the army or as an absolute, the
implication is clearly that the army was sufficiently small for the loss of a hundred
men to have made a noticeable difference. It therefore seems highly unlikely that
such an army could have killed eight or nine thousand Vikings, as claimed by
Regino and the Fulda annalist respectively"
LEADERSIllP AND COMMAND
The overall commander of the Carolingian army was the king; he gave the order to
mobilize the host, and on major campaigns led the army in person. For example,
Charles the Bald defeated a Viking army besieging Bordeaux in 848, capturing nine
enemy ships and killing their crews, and eight years later the same king won a re-
73AB 873 (n. 12 above): Grat 193, Nelson 184; Annales Xantenses [AX] 873, ed. B. de Simson,
MGH SSrG (Hanover and Leipzig 1909) 32-33; AF873 (n. 14 above): Kurze 80, Reuter 72.
"Coupland, "Vikings in Francia and Anglo-Saxon England" (n. 27 above) 194-195 engages with
earlier discussions by P. H. Sawyer, The Age of the Vikings, 2nd ed. (London 1971) 128; and Brooks
(n. 67 above) 4-11.
"AB 865 (n. 12 above): Grat 123, Nelson 127; Miracula sancti Bertini c. 6: MGH Scriptores
76Halsall, Warfare and Society (n. 3 above) 123-133; see also T. Reuter, "Carolingian and Ottonian
warfare" in M. Keen, ed., Medieval Warfare: A History (Oxford 1999) 13-35, at 28-30.
77AB868, 871 (n. 12 above): Grat 151, 181, Nelson 152, 174.
71AV881 (n. 41 above) 50.
79AF 881 (n. 14 above): Kurze 96, Reuter 90; Regino 883 [sic] (n. 43 above) 120.
THE CAROLINGIAN ARMY 59
sounding victory over the Northmen in the forest of Perche." There was evidently
an expectation that in times of crisis the ruler should take on this significant role, so
that when Paris was besieged by the Great Army in 886, Count Odo left the tOMIto
plead that Charles the Fat should bring the host to their rescue." Although Charles
did assemble an army and eventually assume personal command, it was his inability
to defeat the Vikings which reportedly led to his deposition shortly afterwards/" By
the same token, it was undoubtedly Odo's heroic resistance at Paris which made him
the popular favorite to succeed to the throne of the western kingdom.
Although the king was the supreme army commander, conscription, mobilization,
and command were organized on a regional basis by the magnates. The first step
was the creation of lists recording how many men in each district were liable for
conscription. The Edict of Pitres repeated the demand for such rolls to be
maintained, and Hincmar of Rheims is known to have supplied just such a list for
Count Theoderic's campaign against the Vikings in 882.83 The magnates were
subsequently responsible for mustering their followers and leading them to the place
of assembly. Thus Abbot Odo of Ferrieres wrote in 840 that illness had forced him
to dispatch his men with the local count.84 They also commanded squadrons in the
field: a capitulary of Lothar I from 846 reveals that each squadron was led by
between two and four miss;,85 though it appears that many squadrons which opposed
the Vikings were commanded by just one missus, accompanied by one or more other
magnates. For instance, in 854 the Vikings on the Loire were prevented from
attacking Orleans by Bishops Burchard of Chartres, a royal missus'" and Agius of
Orleans." To give another example, a squadron sent against the Vikings on the
Seine in 866 was commanded by Rohert, missus in charge of western Neustria, and
Odo, count of'Troyes/"
The same capitulary of Lothar I from 846 reveals that the second important group
in the military chain of command consisted of the standard-bearers. It listed between
three and six per scara, all of them royal vassals, several of them counts." Standard-
bearers were clearly officers of high rank, although subordinate to the missi.
One of the functions of standards was to act as symbols of authority, so that to capture
the enemy's banners was proof of victory. Robert the Strong proudly presented
standards captured from the Vikings to Charles the Bald in 865, and Adrevald wrote
that in 843 Lambert had failed to carry off Rainald's victory standards
10AB 848 (n. 12 above): Grat 55, Nelson 65; Annales Fontanellenses [AFont] 848, 855: J. Laporte,
ed., "Les premieres annales de Fontanelle," Melanges de la Societe de I'histoire de Normandie, XVe
serie (Rouen and Paris 1951) 63-91, at 81, 91; Tertium missaticum ad Francos et Aquitanos directum
c. 3: MGH Cap. 2.285.
"AV 886 (n. 41 above) 60. On military command as an essential attribute of the monarchy, see Halsall,
Warfare and Society (n. 3 above) 25-30; K. J. Leyser, "Early Medieval Canon Law and the
Beginnings of Knighthood" in 1. Fenske, W. Rösener and T. Zotz, eds., Institutionen; Kultur und
Gesellschaft im Mittelalter (Sigmaringen 1984) 549-566, esp. 560.
12AV887 (n. 41 above) 64.
11Edictum Pissense c. 27: MGH Cap. 2.321; Flodoard, Historia Remensls ecc1esiae 3.26: MGH
Scriptores 13.545; for the dating, see AB 882 (n. 12 above): Grat 246, Nelson 223.
"'Lupus of Ferrieres, letter 15: L. Levillain, ed., Loup de Ferrieres: Correspondance,2 vols. (paris
ISHlotherii capitulare de expeditione contra Sarracenos facienda: MGH Cap. 2.67-68.
II>Capitularemisserum Silvacense, list of misst, c. 9: MGH Cap. 2.276.
17AB 854 (n. 12 above): Grat 69, Nelson 79.
uAB 866 (n. 12 above): Grat 125, Nelson 129.
19As n. 85 above.
60 SIMON COUPLAND
signa).'XJ That they also served a tactical purpose, by giving signals to the army, is
evident not only from the term signum itself, but also from the recurrent use of
phrases such as levatis vexillis or erectis vexillis to indicate the army's departure,"
and from the comments of Hrabanus Maurus that soldiers should use their position
in relation to the standards to keep in formation.92 No contemporary text describes
the appearance of a standard," though manuscript illuminations depict simple
three-tailed banners and a dragon." The latter probably consisted of a metal head
attached to a cloth windsock, to judge from a comparable Roman standard found in
Germany." Horns were also used to give commands, such as setting the army in
motion or summoning aid in battle," and are likewise pictured in contemporary
Although the king was commander-in-chief of the army, he was utterly reliant on
the magnates who controlled the military organization beneath him. Therein lay a
major weakness, for if the magnates failed to fulfill their obligations, not only was
the king's authority undermined, but also the kingdom left defenseless. This was
precisely the charge leveled against Bishop Hincmar of Laon in 871 when he
retained men who, it was argued, should have been fighting the Northrnen."
Furthermore, the loyalty and determination of the common soldiers frequently
depended on their lord, and if he were killed, they would often give up the fight.
Several instances of this are recorded, such as in 885, when "as soon as they came to
join battle, it so happened that Duke Rainald of Maine perished with a few others,
and as a result they all turned back in deep despondency, having achieved nothing of
any use.,,99Similarly, if the magnates withdrew their support, the king would be left
powerless. This occurred in 858, when the blockade of a Viking camp on the Seine
had to be abandoned because many of Charles the Bald's leading men defected to
his brother Louis, who had invaded the West Frankish realm.i'"
These factors must be taken into account in the ongoing debate about the degree
to which the Carolingian army (and indeed the Carolingian state) was centralized
90AB 865 (n. 12 above): Grat 122, Nelson 127; Miracula sancti Benedicti c. 33: MGH Scriptores
15.1.493. See also Halsall, Warfare and Society (n. 3 above) 199-200.
9IE.g., AB 876 (n. 12 above): Grat 208, Nelson 196; Regino 891 (n. 43 above) 137; or Hincmar, De
fide Carolo regi servanda: Patrologia Latina 125, col. 963; compare also Ludwigslied I. 27: Huob her
gundfanon ul W. Braune and E.A. Ebbinghaus, Althochdeutsches Lesebuch, 16th ed. (TObingen 1979)
92Bachrach, Early Carolingian Warfare (n. 2 above) 97, 189, citing De procinctu Romanae
miliciae, c. XIII.
9lIt has been suggested that the amictum / auribus inmodicis croceum carried by two signiferi was
an oriflamme, but the meaning of the phrase is unclear: Abbo I, lines 153-155 (n. 66 above) 26-28,
esp. 28, n. I.
94A. Goldschmidt, Die Elfenbeinskulpturen aus der Zeit der karolingischen und sächsischen Kaiser,
4 vols. (Berlin 1914-1926) I, pI. XIX and XX; or Psalterium aureum 140-141: F. Mütherich and
J. E. Gaehde, Carolingian Painting (London 1977) pI. 46-47.
9sR. Bruce-Mitford, The Sutton Hoo Ship-Burial, 3 vols. (London 1975-1983) 2, fig. 312.
~rmold the Black, Carmen in honorem Hludowici line 1589: E. Faral, ed., Ermold le Noir: Poeme
sur Louis le Pieux et epitres au roi Pepin (Paris 1964) 122; Abbo I, lines 91-92; 2, lines 234-235,
511-518 (n. 66 above) 22, 82-84, 104.
97DerStuttgarter Bilderpsalter, 2 vols. (Stuttgart 1965-1968) I, fol. 88r; San Paolo Bible, fol. 59v:
C.R. Dodwell, Painting in Europe. 800--1200 (Harrnondsworth 1971) pI. 43.
91Concilium TuJ/ense 1apud Saponarias, Epistola synodalis ad Wenilonem; Concilium Duziacense
I: I.-D. Mansi, ed., Sacrorum conciliorum col/ectio IS, col. 530; 16, col. 662.
99AV 885 (n. 41 above) 57; see also AV 886, 62; Regino 867 (n. 43 above) 93; AF 882 (n. 14
above): Kurze 97, Reuter 91.
If"LibeJlus procJamationis adversus Wenilonem c. 5: MGH Cap. 2.451.
THE CAROLINGlAN ARMY 61
and controlled.!" On the one hand Carolingian monarchs such as Charlemagne,
Louis the Pious, and Charles the Bald clearly pursued an imitatio imperii, not least in
the appearance of the royal guards;I02 and the apparatus of state, including the
arrangements for equipping, mobilizing and directing the host, represented
significantly more than the leadership of a loose confederation of magnates'
warbands. On the other hand, as Charles the Bald found to his cost in 858, the
Frankish rulers' leadership of the army was completely dependent on the magnates.
They could and did let their royal masters down, sometimes through negligence,
sometimes through weakness and sometimes through their own deliberate fault.
When royal concerns and aristocratic interest coincided, the Franks could offer a
potent defense against the Scandinavian raiders; when they differed, as often as not
it was the Vikings who benefited. As a result, the Franks' internal conflicts
repeatedly gave the Northmen a window of opportunity. This was not just the view
of the monk Ermentarius; the St. Vaast annalist explained the devastating arrival of
the Great Army from England in 879 as follows: "While they were quarrelling
amongst themselves, the Northmen who were situated across the sea heard about
their discord and crossed the sea,,103Finally, another harmful consequence of the
Frankish internecine struggles noted by contemporaries was the fact that the deaths
of the flower of the Frankish nobility in the civil war and its aftermath deprived the
land of some of its ablest defenders, men who would otherwise have offered stout
resistance to the Vikings.l"
THE ROLE OF 1HE CAVALRY
There is widespread agreement among historians that cavalry formed an important
element in Carolingian armies, though disagreement over whether horsemen outnumbered
infantry, or vice-versa 105 A passage relating to a battle against the
Vikings has been central to the debate and much misunderstood, and so deserves
close consideration.l'" At Louvain in 891, a Scandinavian camp was protected on
one side by the river Dyle and on the other by a swamp. King Arnulf therefore
hesitated to advance, because the situation gave the cavalry no chance to attack, and
because Francis pedetemptim certare inusitatum est.107 This was long held to
indicate that the Franks no longer used infantry,108but that is clearly incorrect: for
IOIThe Problemstellung is usefully summarized in Halsali, Warfare and Society (n. 3 above) 89-
I02Coupland, "Carolingian arms and annor" (n. 39 above) 41; Bachrach, Early Carolingian Warfare
(n. 2 above) 70-71, 86.
1.3N.25 above; A V 879 (n. 41 above) 44.
I04Miracula sancti Benedicti c. 33: MGH Scriptores 15.1.494; Andreae Bergomatis Historia, MGH,
Scriptores rerum Langobardicarum et ltaIicarum 221; Regino 841 (n. 43 above) 75.
losGanshof, "Charlemagne's Army" (n. I above) 66; Lot, Art militaire (n. 5 above) 1.92-93; Oman
(n.5 above) 1.l03-105; B. S. Bachrach, "Charlemagne's Cavalry: Myth and Reality," Military Affairs
47 (1983) 181-187; J. France, "The Military History of the Carolingian Period," Revue beige d'histoire
militaire 26 (1985) 81-99; Verbruggen, "Art militaire" (n. 5 above) 295-297, 304-310.
I06B.S. Bachrach, "Charles Martel, Mounted Shock Combat, the Stirrup, and Feudalism," Studies
in Medieval and Renaissance History 7 (1970) 47-75, at SO-53; J. France, "La guerre dans la France
feodale a la fln du IXe et au Xe siede," Revue beige d'histoire militaire 23 (1979) 177-198, at 191-
192; Verbruggen,"Art militaire" (n. 5 above) 305; Reuter, "Carolingian and Ottonian warfare" (n. 76
above) 30; Halsall, Warfare and Society (n. 3 above) 186-188.
107AF891 (n. 14 above): Kurze 119-120; Reuter 122.
100See,e.g., L. White, Medieval Technology and Social Change (Oxford 1962) 3; Lot, Art militaire
(n. 5 above) 94; J. F. Verbruggen, The Art of Warfare in Western Europe during the Middle Ages
(Amsterdam, New York and Oxford 1977) 96.
62 SIMON COUPLAND
instance, foot-soldiers defended the monastery ofSt. Bertin in 891 and infantry were
deployed in Conrad's army in 906.109 Reuter, believing he was following Bachrach,
translated as ''the Franks are not used to fighting while advancing step by step,"!"
Bachrach has since explicitly contradicted this interpretation, however, setting out
his own understanding of the passage as rather that "in this context ... the Franci
were unaccustomed to advancing slowly over very rough terrain under a barrage
of enemy missiles.t''!' This may have been true, but it is hardly a plausible
translation of the Latin text, which makes no mention of missiles or terrain. The
more straightforward and likely interpretation is that these particular Franks
(compare primores Francorum-"Ieaders of the Franks"-in the same sentence)
were unused to fighting on foot, a fact which was amply demonstrated by
subsequent events.112 In other words, this text indicates that there were Carolingian
troops in the late ninth century who fought only on horseback. but not that
Carolingian armies consisted exclusively of such men.
The composition of ninth-century Frankish forces is seldom described; a rare exception
is a reference to the army at Paris in 845 as consisting of equitum et peditum.
113 The general impression is nonetheless that the host was predominantly
mounted, as at Louvain in 891.114 For instance, when in 835 a Viking attack on
Noirmoutier was repulsed by the local count, Rainald of Herbauge, it was reported
that the Frankish casualties included ''very many horses killed and several horsemen
wounded."m Or when the Vikings launched a counter attack on King Louis III at
Saucourt in 881, it was said that the Franks would have fled "had not the king
swiftly dismounted from his horse to give his men courage and a point of
resistance,"!" Evidently the king, and presumably also at least some of his men,
were fighting on horseback. This is consistent with reports of engagements which
did not involve the Vikings: Nithard stressed the importance of horses in the armies
of the civil war, and the battle of Andernach in 876 evidently included cavalry, since
Charles's troops were described as trying to spur their horses into the battle.'!" It is
also consistent with the contemporary pictorial evidence, since several Carolingian
manuscripts depict pitched battles on horseback. including the Stuttgart Psalter and
San Paolo Bible.lls These illustrations can be taken as reliable indicators of
'I»Miraeula sancti Bertini c. 7: MGH Scriptores 15.1.512; Regino 906 (n. 43 above) 151.
II°Reuter (n. 14 above) 122, citing Bachrach, "Charles Martel" (n. 106 above) 51-53.
IIIBachrach, Early Carolingian Warfare (n. 2 above) 320 n. 108, in which he states clearly: "There
is no reason to suggest on the basis of this text that the Franci were unaccustomed to advancing in a
'step by step' fashion."
1I2Coupland, "Charles the Bald" (n. 7 above) 105. See also Halsall, Warfare and Society (n. 3
113Audradus Modicus, Liber revelationum: L. Traube, "0 Roma nobilis: Philologische
Untersuchungen aus dem Mittelalter," Abhandlungen der bayrischen Akademie der Wissenschaften,
philosophisch-philologische Klasse 19 (1892) 297-395, at 380.
'''References to the presence of horses are not necessarily significant, since troops may have dismounted
to fight, and the army's baggage was usually transported on horseback: AF 896 (n. 14 above):
Kurze 127, Reuter 132.
'''Ermentarius 2.11 (n. 25 above) 66-67.
116AV881 (n. 41 above) 50.
117Nithard2.6, 2.8, 2.9, 2.10, 3.6 (n. 22 above) 56, 60, 66, 72,112; AF876 (n. 14 above): Kurze 89,
IIIDer Stuttgarter Bilderpsalter, 2 vols. (Stuttgart 1965-1968) I, fol. 71v; 1. E. Gaehde, "The
Pictorial Sources of the Illustrations to the Books of Kings, Proverbs, Judith and Maccabees in the
Carolingian Bible of San Paolo Fuori Le Mura in Rome," Frühmittelalterliche Studien 9 (1975) 359-
389, pI. 75, 100 (fol. 83r, 243v); A. Boinet, La Miniature carolingienne (Paris 1913) pI. 124 (fol. 59v).
THE CAROLINGIAN ARMY 63
contemporary conditions because they include ninth-century Frankish military
equipment, even if their composition may have been influenced by antique
iconography.i" The significance of horses in resistance to the Vikings is underlined
by the provisions in the Edict of Pitres for all Franks with the means to own horses
to bring them on campaign, and the warning to state officials not to hinder such men,
although these could of course have been for mobility as much as for combat.120
The proportion of cavalry would undoubtedly have been higher in the royal host
than in local forces raised under the lantweri given the high cost of horses: up to
twenty or even forty solidi in the mid-ninth century.'!' For example, when a Viking
army landed between Bordeaux and Saintes in 845 and Count Siguin of Bordeaux
raised a local army to resist them, the ensuing battle was fought on foot (pedestri ...
proelio). It was a disaster for the Aquitanians: Siguin was captured and killed, and
after suffering heavy casualties the remainder of the army fled.122 Similarly, when
the common people at Prüm turned out to fight the Northmen in 882, there were no
horsemen, but "a countless mass of foot-soldiers, gathered together from the fields
and villages into a single unit.,,123Here, too, the result was a massacre, and even
though on this occasion the lack of military discipline and armor were doubtless also
contributory factors, such defeats as these help to explain why Carolingian rulers
concentrated on raising a well-armed, well-trained fighting force containing a
significant proportion of cavalry.
STRATEGY AND TACTICS (1): CONTAINMENT
When Viking fleets began to penetrate ever deeper into the West Frankish river
system in the 840s, Charles the Bald had to develop tactics to stop them. He
adopted a policy of containment, lining the banks with troops to restrict the invaders
to the rivers. This was first employed on the Seine in 845, but failed when the
Northmen unexpectedly landed on the south bank and put the Frankish troops to
flight. Realizing that the Vikings could not be expelled by force, the king agreed
to pay a tribute.!" This highlights the fundamental weaknesses of the strategy: the
soldiers on the banks were unable to reach the enemy but were themselves vulnerable
to attack. and the tactic was dependent on the discipline, courage and loyalty
of the Frankish forces, who were not always equal to the task.
Nevertheless, Charles adopted the same strategy against a Danish fleet which
camped on an island on the Seine in 852.125 Once again his plans were frustrated,
because the Franks evidently lacked the right sort of ships to launch an assault on
the island. Although river craft were employed in military operations, they invariably
acted as ferries, not as springboards for attack, since they had a deeper
draught than the Viking longships and so could not land on the flat, sloping shores
119Coupland,"Carolingian anns and armor" (n. 39 above) 31-50.
,21>Edictum Pistense c. 26: MGH Cap. 2.321.
'1'20 solidi: Gesta sanetarum Rotonensium 1.7: C. Brett, Irans. and ed., The Monks of Redon: Gesta
Sanetarum Rotonensium and Vita Conuuoionis (Woodbridge 1989) 128-129; L. A. J. W. Sloet, ed.,
Oorkondenboelc der grofschappen Gelre en ZutJen (The Hague 1872-1876) 43; 40 solidi: Miracula
sancti Vedasti 1.5: MGH Scriptores 15.1.398.
'Z2Lupus letter 44: Levillain (n. 84 above) 1.186; cf. Annates Engolismenses 845: MGH Scriptores
'13Regino 882 (n. 43 above) 118.
""Vita Faronis c. 122: MGH SS rer. Mer. 5.200; Translatio sancti Germani c. 12 (n. 26 above) 78;
Coupland, "Tribute payments" (n. 7 above) 59--60.
123AB 852, 853 (n. 12 above): Grat 65, 66, Nelson 7S.
64 SIMON COUPLAND
of islands.i" Faced with the inability of his troops to mount an assault, Charles
subsequently abandoned the siege and paid a tribute to one of the Viking leaders.
In 854 Bishops Burchard and Agius nonetheless proved that the strategy of
containment could be made to work with the help of river vessels. They prevented
the Vikings from advancing up the Loire "by making ready against them ships
and soldiers": presumably the ships blocked the river while the troops lined the
banks.128 It was perhaps this episode which inspired Charles the Bald and Lothar
11to deploy a river fleet "the like of which had never before been seen in our land"
when another Viking band camped on an island at Oisse! in 858.129 The ships
were presumably moored in midstream to prevent the Vikings' passage up- or
downriver, while the army blockaded the island from the shore. Yet as was
mentioned earlier, the siege had to be abandoned when Louis the German invaded
the West Frankish kingdom, and in the subsequent debacle, the entire fleet fell
into the Vikings' hands.l3O Only two other attempts to attack Viking island bases
are recorded, both similarly ending in failure. In 864 Lothar 11fitted out ships to
launch an assault on an island on the Rhine, but his men refused to attack, and
seven years later Hugh the Abbot and Count Gauzfrid of Le Mans tried to capture
an island camp on the Loire, but made little headway and suffered heavy losses.'!'
The Vikings' island camps appear to have been virtually impregnable.
Despite its repeated failures, Charles the Bald nonetheless still believed in the
strategy of containment, and in 862 he at last found a defensive tactic which could
enforce it effectively. He managed to trap a Viking fleet by blocking a bridge
across the Marne at Isles-les-Villenoy and deploying troops along the banks as
usual. The Vikings were forced to come to terms, and the king subsequently
pressed ahead with the fortification of a bridge at Pont-de-I' Arche to try to
achieve the permanent exclusion of the Vikings from the upper Seine.l32 The
inherent weaknesses of containment were again exposed in 865, however, when a
fresh fleet entered the Seine. First the guards failed to turn up on time and then the
raiders fell on one of the Frankish detachments, which turned and fled.m This is
the last occasion on which the strategy is known to have been pursued; Charles's
successors appear to have abandoned it. It had seemed inherently sound, and had
indeed succeeded on one notable occasion, but consistently failed because of the
inability of the Frankish troops to land on the Vikings' island bases and their consequent
vulnerability to surprise attack.
STRATEGY AND TACTICS (2): INvEsTMENT
The Franks were past masters in the art of siege warfare, and Bachrach in particular
has underlined the central role played by the engineers in the Carolingian army
under Pippin the Short and Charlemagne.l" Yet the Vikings' mobility, coupled
126B.Almgren, "Vikingatäg och vikingaskepp," Tor 8 (1962) 192.
I21Coupland, "Poachers" (n. 7 above) 94-95, and "Tribute payments" (n. 7 above) 60.
mAB 854 (n. 12 above): Grat 69, Nelson 79.
129Vita Faronis c. 125: MGH SS rer. Mer. 5.201.
I30Vita Faronis c. 125: MGH SS rer. Mer. S.201; AB 858 (n. 12 above): Grat 78, Nelson 88.
IJIAX864 (n. 73 above) 21;AB 871 (n. 12 above): Grat 181, Nelson 174.
Il2Coupland, "Fortified bridges" (n. 7 above) 2~.
133AB86S, 866 (n. 12 above): Grat 123, 125; Nelson 127, 129-130.
134Bachrach, "Charlemagne's Cavalry" (n. 105 above) 181-187; "Charles MarteI" (n. 106 above)
lHE CAROLINGIAN ARMY 65
with their habit of keeping to their ships or camping on islands, meant that they
rarely allowed themselves to be trapped by the Franks. In fact, during the whole
period of the Scandinavian invasions there are only three occasions when Viking
armies are known to have been besieged: at Angers in 873, Asselt in 880, and a
small fortress in the Hesbaye in 885.135
In any siege, the attackers' first priority was to surround the enemy, sometimes
throwing up a rampart around the site as well, as is attested at Angers in 873.
They would then seek to stop all food and water reaching the beleaguered
garrison, a tactic which is said to have proved decisive in the Hesbaye, where the
Franks had reportedly captured the Scandinavians' supplies before the blockade
began. A protracted investment could, however, also create logistical problems for
the besieging army, and Regino claimed that the Franks at Angers began to run
out of food, as well as being stricken by sickness. Even though the historicity of
Regino's account is dubious, as we shall see, clearly such situations did occur in
Apart from undertaking a direct assault, the assailants might try to undermine
the walls, scale the ramparts on ladders, or burn down the stronghold. Siege engines
could also be employed, though not apparently at any of the places where the
Vikings were besieged; in all three cases it appears that the Franks settled in to
starve their opponents into submission. At Angers this was successful: the
Scandinavian leaders commended themselves to Charles the Bald and promised to
leave the kingdom as soon as the winter was over. This is the version of events
reported in the contemporary Annals of St. Bertin, and it is not only quite different
from the account in the later Chronicle of Regino of Prüm, but also significantly
more reliable. As well as claiming that the Franks used siege engines, Regino
alleged that Charles the Bald took a tribute from the Vikings and asserted that the
deciding factor in the Frankish victory was a Breton plan to divert the Maine by
the digging of a channel "of amazing depth and width," leaving the longships high
and dry. If the second detail is implausible, the third is impossible. The river was
nearly 100 meters wide at Angers, rendering the task beyond the capability of the
army in the time available. By way of comparison, the fossa Carolina dug in
Bavaria in 793 was only some 30m wide, 1500m long and 6m deep, yet it has
been calculated that it would have taken a work-force of 4700 men ten weeks to
excavate.l36 Less dramatic but much more plausible is Hincmar's report that the
Northmen were simply worn down by the length of the siege, a claim which is given
added weight by the likelihood that Hincmar was himself present at the siege."?
Once the Vikings had agreed to leave the town, the king ordered the fortification of
47-75; and most recently Early Carolingian Warfare (n. 2 above) 102-118.
IJSAngers:AB 873 (n. 12 above): Grat 193-195, Nelson 183-185; Regino 873 (n. 43 above) 105-
107; Asselt: AF 882 (n. 14 above): Kurze 98-99, 107-109; Reuter 92-93, 104-106; AB 882 (n. 12
above): Grat 247-248, Nelson 224-225; Hesbaye: AF 885 (n. 14 above): Kurze 102, Reuter 97. Although
Regino described a purported siege at Brissarthe in 866, his account cannot be trusted: Regino
867 [sic] (n. 43 above) 92-93. (1) The encounter is wrongly dated to 867; (2) Regino makes no mention
of details in other, more reliable sources, such as the presence of Bretons, the context of a raid on
Le Mans, or the fact that the raiders were mounted; (3) he states that the Franks intended to use siege
engines, even though they had intercepted the Vikings with no expectation of a siege; (4) he names the
Viking commander as Hasting, a notorious leader by the time Regino was writing, but not otherwise
attested on the Continent before 882.
I*H. H. Hofmann, 'Fossa Carolina" in Braunfels, ed., Karl der Grosse (n. 63 above) 1.446.
Il7Nelson,Annals of St-Bertin (n. 12 above) 185 n.17.
66 SIMON COUPLAND
the bridge over the Loire at Les Ponts-de-Ce to prevent them from returning
As for the siege at Asselt, our understanding of events is complicated by the fact
that conflicting accounts are given in the two continuations of the Annals of Fulda.
They at least agree on the outcome of the siege-one of the Viking leaders, Godfrid,
commended himself to Charles the Fat, while the rest left on payment ofa tributebut
differ significantly over why it was lifted. One version claims that the stronghold
was about to surrender, and attributes the deal to bribery, treachery and weakness,
but the author is deeply hostile to those counselors who brokered the agreement. The
other, which is more likely to be reliable,139reports that the presence of large numbers
of putrefying corpses (it was July) spread not only disgust but also sickness
among the besieging army. This is not only highly plausible but also perhaps paralleled
at Angers, as we have seen. Even so, it did nothing for the emperor's
In sum, the Franks suffered the frustration of having an army which was skilled
in siegecraft, but fighting an enemy which hardly ever allowed itself to be
trapped.l'" When it did come to a siege, the Franks apparently held the upper
hand, in that two of the three occasions on which Viking forces were besieged
were clear Frankish victories. This was the case at Angers, where the
Scandinavian leaders were forced to come to terms, and in the Hesbaye, where the
Northmen fled by night. The third, the siege of Asselt, was seen at the time as a
capitulation by Charles the Fat, but did result in Godfrid's commendation, which
was a proven and effective method of turning an enemy into an ally."!
Furthermore, we have seen that the emperor's failure to press home the siege was
probably due to the spread of sickness among his troops, a common hazard of
siege warfare, especially in the summer.
STRATEGY AND TACTICS (3): ENGAGEMENT142
Nemine resistente: "resisted by no-one"; the phrase echoes as a plaintive litany
through an array of ninth-century clerical sources describing the Viking raids.!"
Yet it should not be thought that the Franks fought shy of engaging the Vikings in
battle. The description of Charles the Bald in the Annals of Xanten as "suffering
frequent onslaughts from the pagans, continually offering them tribute, and never
emerging victorious in battle" owes more to political rhetoric than historical reality.
l44Time and again Frankish armies and Carolingian rulers (including Charles the
Bald) came against the Vikings in battle, sometimes suffering defeat, but sometimes
emerging victorious. Any roll of honor would include Noirmoutier (835), the Dordogne
(848), Poitiers (855 and 868), le Perehe (856), the Charente (865), the Loire
(862, 865 and 869),.Oostergo (873), West Frisia (876), Saucourt (881), Avaux
13BCoupland,"Fortified Bridges" (n. 7 above) 9-10.
139Coupland,"Myth and History" (n. 65 above) 198-199.
J~e present discussion has been kept short for this reason.
J4JCoupland, "Poachers" (n. 7 above) 108-114.
J42Seealso Halsall, Warfare and Society (n. 3 above) chap. 9, "Battle."
J4JThe phrase or an equivalent is found in the Annals of St. Bertin, Annals of Fulda, Annals of St.
Vaast, Chronicle of Nantes, Life of St. Faro, Miracles of St. Benedict, Miracles of St. Germanus,
Miracles of St. Remaclius, Translation of St. Philibert and works by Abbo and Audradus Modicus.
J44AX869(n. 73 above) 27; Coup land, "Tribute payments" (n. 7 above) 71-72.
THE CAROLINGlAN ARMY 67
(882), Montfaucon and the river Aisne (888), Louvain (891), and le Vimeu (898).145
These victories were, moreover, won by a variety of Frankish forces. In 848, 856,
881, and 882 it was the king leading the host who was victorious; in 835,862,865,
and 869 a local count, possibly under the auspices of the lantweri; while at Poitiers
in 855 and 868, on the Charente in 865 and in Frisia in 873 and 876 it was the local
populace who were defending their territory. These military successes serve as a
reminder that the weaknesses of the Frankish military machine and the failings of
locally raised forces should not be overestimated.
This list of triumphs over Viking armies would also undoubtedly be longer were
it not for the raiders' frequent avoidance of battle. For example, on the Seine in
841 "Wulfard, a royal vassal, opposed them with an army, but the pagans were not
at all prepared to fight"l46 The Vikings' aim was to acquire plunder, not to win
battles, so why risk their lives and loot in combat when a tactical retreat would
preserve both? As a result, they generally sought to pillage poorly defended
targets and withdraw without risking casualties, particularly in the earlier
incursions. Thus when a warband was intercepted at Vardes in 852, they fled into
the woods to escape the Frankish horsemen, who could not pursue them among
the trees.147The same tactic was used by the Great Army in 891 when they were
attacked by Odo at Wallers, and the armalist adds that the Northmen had been
deliberately traveling through inaccessible areas (invia loca).148What was more, if
Viking forces were defeated they often simply withdrew and regrouped. The
lament uttered by the St. Vaast annalist after Carloman's victory at Avaux applied to
many similar situations: "When battle commenced, the Franks emerged victorious,
but this engagement did nothing to subdue the Northmen.,,149
Contemporary sources reveal regrettably little about the tactics employed by
the Franks against the Vikings, though one vital factor was intelligence. Scouts
were seen as extremely important in keeping abreast of the enemy's movements
and trying to predict their intentions. ISO Thus when Louis III defeated a Viking
army at Saucourt in 881, it was after "he sent out scouts, who reported that they
were coming back laden with plunder."lsl Twenty-five years earlier, Charles the
Bald's successful attack on a Viking warband in the forest of Perehe followed his
muster of the Franks at Neaufles, where, Hincmar wrote, ''we were keeping watch
against the Northmen's incursion."m Equally useful was misinformation, to mislead
or trick the enemy. For example, in 873 Charles the Bald was planning to
attack the Vikings at Angers, but announced a campaign into Brittany in order to
lull them into a false sense of security .IS3
Virtually nothing is recorded about Carolingian battle tactics against the Vi-
I"See at on. 40, 66, 73, 80, and 115 above, n. 168 below; also AB 848, 855, 862, 868 (n. 12 above):
Grat SS, 71, 89, 166, Nelson 65, 81, 99, 152; AF 876,891 (n. 14 above): Kurze 86, 119-121, Reuter
79, 121-123; AV881, 882, 888, 891, 898 (n. 41 above) SO-51, 53, 65, 70, 80.
If6AFont 841 (n. 80 above) 75.
147AFont851 (n. 80 above) 89; see also Regino 892 (n. 43 above) 138.
I"AV89I, cf. 898 (n. 41 above) 69, 80. See also at n. 163 below.
149AV882 (n. 41 above) 53.
uOCf.Bachrach, Early Carolingian Warfare (n. 2 above) 188-189; Halsall, Warfare and Society (n.
3 above) 147-148, 157.
IIIAV881 (n. 41 above) 50.
112AFont855 [sic] (n. 80 above) 91; Flodoard, Historia Remensis ecclesiae 3.16: MGH Scriptores
13.506; Hincmari Epistolae no. 131: MGH Episto1ae 8.1.72.
mAß 873 (n. 12 above): Grat 192, Nelson 183.
68 SIMON COUPLAND
kings, though presumably they were similar to those employed against other
opponents in other campaigns. These could include dawn attacks, ambush,
feigned retreat, and attacking from two or three directions in order to stretch the
enemy forces and probe for weak points.P' The latter maneuver seems to have
been used at Fontenoy during the Frankish civil war, and had also been favored
by Charlemagne and Louis the Pious in the earlier expansionist campaigns.F" The
Frankish cavalry evidently charged at speed in close order, as is described in
Regino's account of a battle fought between Franks and Bretons in 851.156
Although Regino is not always reliable on earlier events, in this instance his
account is corroborated by Ermold the Black, who personally campaigned in
Brittany, and by Nithard's description of Frankish military games fought in
842.157 Regino records that the Franks were drawn up en masse (in unum
conglobatii and unlike the Bretons, who were armed with light throwing spears,
held onto their lances, using them as thrusting weapons, as well as fighting hand
to hand with drawn swords. The Bretons advanced towards the Frankish line
hurling their javelins, then fell back in an attempt to lure their opponents to break
ranks and pursue them. This underlines the Franks' preference for fighting in
close order, in contrast to the wheeling, darting Breton cavalry, and King Odo's
command to his men to keep close together as they charged at Montfaucon in 888
suggests the same tactic was still being employed against the Vikings.I" By
contrast, though it is possible that ninth-century Carolingian armies sometimes
fought in a wedge formation or phalanx, the use of the term cuneus by
contemporary authors should not necessarily be interpreted in that sense. By this
time the term could simply denote a body of troops, including units which were
not in battle formation, or even in a wholly non-military context to describe a
crowd of church-goers.i"
Although the Carolingian armies were well armed, highly mobile and very experienced
in mounted warfare, the Vikings quickly developed effective tactics
against them. As we have seen, warbands tended to return to their ships or camp
on islands where the Franks could not reach them. The rare defeat at Noirmoutier
in 835 was only possible because the tide had gone out, allowing Rainald and his
horsemen to cross a causeway from the mainland. On land, the Vikings learned to
use the local terrain to their advantage. To avoid encountering Frankish troops
they kept to invia loca, inaccessible regions.!" including forest and marshland:
Il4Halsall, Warfare and Society (n. 3 above) 188-191, 195-196, 204-206; Bachrach, Early
Carolingian Warfare (n. 2 above) chap. 5, "Battlefield Tactics."
mNithard 2.10 (n. 22 above) 76-78; Verbruggen, Art of Warfare (n. 108 above) 280-282; Halsall,
Warfare and Society(n. 3 above) 147.
156Regino 860 (n. 43 above) 78-79 (recte 851: see AFont 851 [no 80 above] 87; Annales
Engolismenses 851: MGH Scriptores 16.486). See also Halsall, Warfare and Society (n. 3 above) 194-
J57Ermold lines 1494, 1680-1683 (n. 96above) 114, 128; Nithard 3.6 (n. 22 above) 110-112. Cf.
l-C. Cassard, "La guerre des Bretons armoricains au haut moyen äge," Revue historique 557 (1986) 3-
InARF(revision) 782 (n. 9 above) 63; Abbo 2,line 507 (n. 66 above) 104.
I'"AB 864, 869 (n. 12 above): Grat 116, 165, Nelson 121 ("companies"), 162 ("formations");
Ermentarius 2.2 (n. 25 above) 63. A passage in Richer (1.9) sometimes cited as evidence that the wedge
formation was still employed in the late ninth century is unreliable: R. Latouche, ed., Richer: Histoire
de France (888-995), 2 vols. (Paris 1930-1937) 1.24 and n. I; see ix-xi of the same work and Lot, Art
militaire (n. 5 above) 1.102.
160As n. 148 above.
THE CAROLINGIAN ARMY 69
the Vikings at Louvain camped with a swamp on one side and the Dyle on the
other precisely to preclude a mounted attack. On land, Viking armies regularly
surrounded their camps with a rampart topped by a palisade, and perhaps ditches
as well."! The effectiveness of the latter was proved at Paris in 886, where the
commander of the army sent to break the siege, Duke Henry, rode into just such a
ditch. His horse fell and threw him, whereupon a group of Danes hiding nearby
jumped out and killed him.162Here again we see how the resourceful Scandinavians
repeatedly found ways of blunting the edge of the Carolingian war machine.
This article set out to consider why the Frankish armies which had proved so effective
against other opponents under Pippin III and Charlemagne were apparently
ineffectual against the Vikings in the ninth century. The last section in particular has
highlighted the fact that this is largely a misconception. Carolingian armies won numerous
victories against the Vikings, but to their frustration found that military success
alone was insufficient to deal with the Scandinavian menace. The Northmen
would simply retreat, regroup and later return. What was more, they were masters of
the avoidance of battle, not only through their choice of territory through which to
travel and the location of their camps, but also through their willingness to beat a
strategic retreat. This was equally true in 841, when Wulfard found his opponents on
the Seine unwilling to fight, and fifty years later, when "King Odo caught up with
[the Northrnen] at Wailers, but not in the way that he wanted, for, having abandoned
their loot, they escaped by scattering through the woods, and so regained their
camp.,,163The reason for this behavior is clear, namely that the Vikings were not an
invading army seeking territory to conquer, but a series of fast-moving warbands,
each hungry for loot. Normandy was not conquered by Rollo as Saxony was by
Charlemagne, but granted by Charles the Simple much as tracts of Frisia had been
given in benefice to other Scandinavian chiefs by Frankish rulers before him,
perhaps even including Charlemagne.!" This made it particularly difficult for the
Carolingians to crush the raiders, for defeat in battle might remove them in the short
term, but only some kind of agreement offered the hope of lasting exclusion. That is
why Frankish rulers ended up offering benefices in exchange for conversion and
commendation, or, like their Anglo-Saxon counterparts, paying tribute in exchange
for a promise of departure.l'"
We have also considered the possibility that weaknesses in the Carolingian military
organization might have undermined the defense of Francia and benefited the
Viking attackers. Previous authors have suggested that the mobilization of the army
was too slow to deal with these fast-moving raiders and that the lack of booty might
have lessened the resolve of the Frankish nobility.l'" We have concluded that both
these explanations lack substance, since under the provisions of the lantweri local
armies could have been raised relatively quickly, motivated to defend not simply the
161AF891 (n. 14 above): Kurze 119-120 (sepibus more eorum munlcione septa securi consederunt]
Reuter 121; Regino 891 (n. 43 above) 138 (ligno et terrae congerie more solito se communiunl).
162 AV886 (n. 41 above) 61; Regino 887 (n. 43 above) 126.
163As nn. 146 and 148 above.
164Coupland,"Poachers" (n. 7 above) esp. 87-88.
16sCoupland,"Tribute payments" (n. 7 above) 68-69, 72.
166Seeat nn. 31 and 38 above.
70 SIMON COUPLAND
kingdom, but also their own homes and families. Some of these armies were admittedly
poorly trained, armed and equipped, and paid the penalty for their shortcomings,
167but others were more than equal to the task. Their zeal and potential
effectiveness was exemplified in 865, when Robert the Strong led one local force
which "killed more than 500 of the Northmen who were based on the Loire, without
loss to his own side," and an army of Aquitanians "battled with the Northmen who
were camped on the Charente under Sigfrid, and killed around 400 of them, while
the rest fled back to their ships.,,168The principal weakness of the Frankish forces
which has been identified here was the lack of assault ships which could attack the
Vikings' island bases, leaving the initiative always with the Scandinavians and the
Franks vulnerable to surprise attack.
Ninth-century authors offered other reasons for the Franks' inability to overcome
the Vikings. One blamed the cowardice of the army's commanders.l'" but this
almost certainly reflected the strongly held view of the clerical writers of the day that
it was the duty of the laity to defend the church and the kingdom by defeating and
expelling the pagan enemy.'?" There is ample evidence of resistance to the Vikings
by the aristocracy, several of whom gave their lives in the defense of the realm.i"
Other contemporaries, such as Ermentarius and the St. Vaast annalist, saw the
problem as the Franks' internal divisions.l72 This was clearly a factor which
contributed to the Vikings' success, and here there is a significant difference
between the reign of Charlemagne and the later ninth century. Charlemagne of
course faced internal dissent and had to win over his magnates with gifts of wealth
or prestige. Yet he never risked a rival Frankish ruler invading his kingdom while on
campaign, as happened to Charles the Bald and Lothar IT in 858,173nor suffered
incursions from pagan enemies while he was seeking to consolidate his grip on the
throne, as happened to Charles the Bald in the early 840s. Worse, contemporaries
reported that those enemies had been encouraged to attack the West Frankish
kingdom by Lothar I, initially as a way of weakening the position of his father,
Louis the Pious, and later in support of his dream of empire during the civil war.'?'
In the memorable phrase of Gwyn Jones, "It must have appeared to Charles [the
Bald] as though he was a man with a wolf at his throat and a wasp in his hair, and in
this menagerie of menace the Danes were the wasp.,,175
In short, it appears that it was Frankish politics and Scandinavian tactics, including
their choice of island bases, which represented the most significant obstacles
to the Carolingian armies in their struggles against the Vikings, rather than
any innate weakness in their military structures or strategies.
St. Paul's Vicarage
33 Queen's Road, Kingston upon Thames
Surrey KT2 7SF, UK
167Seeat nn. 39, 43, and 121 above.
I6IAB 865 (n. 12 above): Grat 122, 124; Nelson 127,128.
169Seeat n. 26 above.
17°Coupland,"Rod of God's wrath" (n. 7 above) 547-549; "Tribute payments" (n. 7 above) 71.
I7IE.g.,at nn. 40, 17, 99,115, and 122 above.
JnSee at nn. 25 and 103 above.
I7lSee at n. 100 above.
I74See Coupland, "Poachers" (n. 7 above) 90--91.
mG. Iones, A History of the Vikings (London 1968) 213.