Charles the Simple, Robert of Neustria,

Early Medieval Europe





(



)



–







Garsington Road, Oxford OX





DQ, UK and



Main Street, Malden, MA



, USA

Blackwell Publishing Ltd

Oxford, UK

EMED

Early Medieval Europe

0963-9462

XXX

Original Article

Charles the Simple, Robert of Neustria

Charles the Simple, Robert of Neustria,

and the

vexilla

of Saint-Denis





A recent consensus holds that the deposition of Charles the Simple by Robert
of Neustria in 922 was the result of a sudden deterioration in their rela-
tionship that had nothing to do with any earlier hostilities. A close examina-
tion of contemporary diplomas suggests a very different story. In 898, Charles
did not faithfully implement the agreement he had made with Robert’s
brother Odo that gave him the succession and he resisted recognizing
Robert’s lay abbacies, in particuar that of Saint-Denis. These grievances
shaped the entirety of the reign, and their memory remained quite alive at
the time of the deposition.

If Shakespeare had been French, he would have written

Charles III

. It’s

the ﬁve-act arc of the king’s life – the injustice committed against him
when he was a boy and denied the throne; his long struggle to right the
injustice and gain the throne; the ﬂawed greatness he displayed once he
achieved his goal; the insurgency that deposed him; the pathos of his
imprisonment and death. It’s the national history in which his tragedy
played out – a long conﬂict between two great houses that ended with
the overthrow of a proud dynasty in favour of its enemies, whose
descendants would rule in their place for over 800 years. It’s the impor-
tance of the ideas at stake – hieratic kingship that stands apart from and
above the powerful with its roots in a heroic past; consortial kingship
that stands with the powerful, pragmatically doing whatever is necessary
to make it through a turbulent present. It’s the leading players – two
powerful, powerfully minded protagonists seemingly destined to a ﬁnal
confrontation by values and forces they cannot choose to ignore with-
out denying what they stand for. It’s the cast of secondary charac-
ters: the loyal counsellor whose misguided advice pushes his king
towards the ﬁnal disaster; the evil count who feigns loyalty as cover for
betrayal; the wise prelate who dies on the eve of the battle he had spent
his life trying to prevent. It’s the mystery, the crucial facts about events
and motivations being unknown, and perhaps unknowable.

Early Medieval Europe





(



)



356

Geoffrey Koziol

Most of all it’s the mystery. Mysteries are hard to resist.
Ruler of the West Frankish kingdom in the early tenth century,

Charles III – more commonly known as Charles the Simple or Charles
the Straightforward – was the grandson of Charles the Bald and the
posthumously born son of Louis the Stammerer.

After the deaths of his

two older half-brothers and the cousin who succeeded them, it should
have been Charles’s turn to come to the throne, but he was passed over
because, at eight-years old, he was judged too young to be able to provide
protection against the Vikings, who had recently spent two years besieg-
ing the city of Paris.

Instead, in 888, a group of West Frankish mag-

nates elected as king Odo, the count of Paris and scion of a powerful
aristocratic family whose leadership during the siege of his city had
gained him pre-eminence within the kingdom, including advancement
to the position of margrave. Even so, Odo had been elected by only one
grouping of magnates. Others resisted him; and though over the course
of his ten-year reign Odo outfaced his enemies and ruled with surpris-
ing success, his reign remained insecure and subject to insurgency at the
slightest provocation, not least from supporters of Charles the Simple,
the most loyal and powerful of whom was the archbishop of Reims,
Fulk. In 893 Fulk had gone so far as to crown Charles king. Over the
next several years, Odo did not so much defeat Charles as outlast and
isolate him, until in late 897 Charles sued for peace. Knowing he was
nearing the end of his own life, Odo came to terms with Charles.

The arguments presented here anticipate those I develop at greater length in a forthcoming
book, tentatively titled

The Footsteps of Kings: West Frankish Royal Diplomas and Their Stories

(840–987)

. In the meantime, my reasons for preferring the traditional translation of

Karolus

Simplex

as ‘Charles the Simple’ rather than ‘Charles the Straightforward’ are set out in ‘Is

Robert I in Hell? The Diploma for Saint-Denis and the Mind of a Rebel King (Jan. 25, 923)’,

Early Medieval Europe

14 (2006), pp. 233–67, at n. 14. Brieﬂy,

simplicitas

meant ‘innocence’

or ‘naivety’, for contemporaries an excellent monastic virtue but a tragic characteristic of
kings, as it was with Charles, who allowed himself to be duped and imprisoned by Herbert
of Vermandois. ‘Straightforward’ does not capture these meanings and is not what later
chroniclers meant by calling Charles

simplex

It goes without saying that while the justiﬁcation is plausible, it was also self-serving, not only
for Odo’s supporters but also for Fulk of Reims, when he later had to explain his failure to
have supported Charles in 888: Gerhard Schneider,

Erzbischof Fulco von Reims (883–900) und

das Frankenreich

, Münchener Beiträge zur Mediävistik und Renaissance Forschung 14

(Munich, 1973), pp. 39–41, 47–8, 92–3, 115–19. There is also good evidence that Odo was
more cunning about his elevation than used to be thought: Olivier Guillot, ‘Les étapes de
l’acession d’Eudes au pouvoir royal’, in

Media in Francia: Recueil de mélanges offert à Karl

Ferdinand Werner

(Maulévrier, 1989), pp. 199–223. Though Simon MacLean has criticized it,

Guillot’s interpretation still seems fully justiﬁed by the evidence. In addition, MacLean’s
optimistic account of relations between Odo and Charles the Fat overlooks the heavy irony
of his primary source (Abbo of Saint-Germain’s

Song of the Siege of Paris

), as well as the overt

criticism of other sources, such as the

Annals of Saint-Vaast

(

s.a.

886, p. 62). See MacLean,

Kingship and Politics in the Late Ninth Century: Charles the Fat and the End of the Carolingian
Empire

(Cambridge, 2003), Ch. 3.



Early Medieval Europe





(



)

Charles the Simple, Robert of Neustria

357

According to historians’ current understanding of the settlement, Odo
assigned Charles a small part of the kingdom for his maintenance dur-
ing the remainder of his own reign; in return, he agreed that Charles
would succeed him. When Odo died just a few months later in January
898, the West Frankish magnates accepted Charles as king.

The mystery in understanding Charles’s reign is due to the fact that

our only contemporary West Frankish narrative source, the

Annals of

Saint-Vaast

, gives out in June 900, soon after Charles had been recog-

nized as king. There is then almost twenty years of silence before the
next contemporary West Frankish narrative begins. This narrative – the

Annals

of Flodoard – opens with a frustratingly brief account of seem-

ingly inconsequential events in 919, immediately followed by a long and
complex entry for 920 that shows all hell already breaking loose, as a
group of magnates seized Charles and held him in captivity. Although
Flodoard does not tell us who the leader of the opposition was, its
beneﬁciary is obvious: Odo’s brother Robert. On this occasion, Fulk’s
successor as archbishop of Reims, Heriveus (our Shakespeareanly selﬂess
prelate) negotiated a reconciliation. Even so, it took him seven months
of effort and the peace proved temporary.

In June 922, deﬁance turned

to open rebellion as Robert was elected king. Still, for a year Charles
remained on the loose, as he tried to mobilize support for a counter-
attack. Finally, in June 923, his supporters dwindling, he staked every-
thing on a surprise attack during a truce when Robert had disbanded
the larger part of his forces. Though Robert was killed in the ﬁghting,
Charles still lost the battle and with it the war, for instead of returning
to Charles, the magnates elected Robert’s son-in-law king. Soon after,
lured by the false promise of an alliance that would help him regain the
throne, Charles was captured and imprisoned by Herbert II of Vermandois

The essential works on Odo’s career and reign are Yves Sassier,

Hugues Capet: Naissance d’une

dynastie

(Paris, 1987), pp. 27–71; R.-H. Bautier, ‘Le règne d’Eudes à la lumière des diplômes

expédiés par sa chancellerie’,

Comptes-rendus des séances de l’Académie des Inscriptions et Belles-

Lettres

(1961), pp. 140–57; Guillot, ‘Les étapes de l’accession’. Also MacLean,

Kingship and

Politics

(with the provisos above, n. 2); Reinhard Schneider, ‘Odo (888–898)’, in Joachim

Ehlers, Heribert Müller and Bernd Schneidmüller (eds),

Die französischen Könige des Mitte-

lalters von Odo bis Karl VIII. (888–1498)

(Munich, 1996), pp. 12–21; Auguste Eckel,

Charles le

Simple

(Paris, 1899), Ch. 1; Edouard Favre,

Eudes, comte de Paris et roi de France (882–898)

(Paris, 1893); Schneider,

Erzbischof Fulco von Reims

. The edition of Odo’s acts is

Recueil des

actes d’Eudes, roi de France (888–898)

, ed. R.-H. Bautier (Paris, 1967).

On Heriveus, see Gerhard Schmitz, ‘Heriveus von Reims (900–922): Zur Geschichte des
Erzbistums Reims am Beginn des 10. Jahrhunderts’,

Francia

6 (1978), pp. 59–106, especially

pp. 78–87;

idem

, ‘Das Konzil von Trosly (909): Überlieferung und Quellen’,

Deutsches Archiv

33 (1977), pp. 341–434; and R.-H. Bautier, ‘Un recueil de textes pour servir à la biographie
de l’archevêque de Reims, Hervé (X

siècle)’, in Charles-Edmond Perrin (ed.),

Mélanges

d’histoire du Moyen Age dédiés à la mémoire de Louis Halphen

(Paris, 1951), pp. 1–6.

Early Medieval Europe





(



)



358

Geoffrey Koziol

(our Shakespeareanly evil count). He died in honourable but hopeless
captivity in one of Herbert’s castles six years later, in 929.

According to Flodoard, a single complaint set Robert against Charles: the

favour the king showed to a Lotharingian noble named Hagano (our Shake-
speareanly misguided counsellor).

Yet calling this a matter of ‘favour’ does

not do justice to the issue. It was not simply that Hagano’s advancement
piqued Robert and hindered his ambitions. As understood by historians,
the stakes were much more complicated, much more principled. First, Hagano
was not a magnate, and therefore not even close to the standing of Robert
and other territorial princes of the kingdom. He had no right to advise
the king on matters affecting the West Frankish realm at all, since as a
Lotharingian he was not even a member of the kingdom, he and his
family having held no signiﬁcant ofﬁces there. Second, the competition
between Robert and Hagano was zero-sum: Hagano appears not simply
to have favoured policies Robert opposed; his promotion was explicitly
designed to counteract Robert’s authority in the kingdom. Thus, the decision
that ﬁnally set off the deposition was aimed directly against Robert:
Charles gave Hagano the convent of Chelles, whose abbess had recently
married her daughter to Robert’s son and putative heir. In other words,
by giving Chelles to Hagano, Charles was trying to undermine an important
political alliance Robert had made for his son and successor. At the same
time, Chelles lay right on the edge of Robert’s and Herbert’s own territories.
Its possession gave Hagano (and Charles) the ability to harrass their lands
while interdicting Robert’s easy passage towards Reims and Lotharingia,
key areas of Charles’s support but also areas where Robert had recently
gained allies. Nor was this the only insult against Robert. According to a
later source, often wrong but on this score plausible, Charles admitted
Hagano to private counsels from which he excluded the other magnates,
including Robert. In Robert’s very presence, he allowed Hagano to perform

On Charles’s reign and the rebellion, see especially Karl Ferdinand Werner,

Les origines (avant

l’an mil)

, Histoire de France, gen. ed. Jean Favier (Paris, 1984), pp. 487–515, and Sassier,

Hugues Capet

, pp. 73–87; Bernd Schneidmüller, ‘Karl III. (“Der Einfältige”), 893/898–923/

929’, and Franz J. Felten, ‘Robert I. (922/923) und Rudolf I. (923–936)’, in Ehlers,

et al

. (eds),

Die französischen Könige

, pp. 22–35, 36–45; Philippe Lauer,

Robert I

et Raoul de Bourgogne,

rois de France (923–936)

(Paris, 1910), pp. 1–11; Helmut Schwager,

Graf Heribert II. von

Soissons, Omois, Meaux, Madrie sowie Vermandois (900/06–943) und die Francia (Nord-
Frankreich) in der 1. Hälfte des 10. Jahrhunderts

, Münchener historische Studien, Abteilung

mittelalterliche Geschichte 6 (Kallmünz, 1994);

Recueil des actes de Robert I

et de Raoul, rois

de France (922–936)

, ed. Jean Dufour (Paris, 1978), especially Introduction, Appendix I,

pp. xci–xcvi. The edition of Charles’s acts is

Recueil des actes de Charles III le Simple, roi

de France (893–923)

, ed. Ferdinand Lot and Philippe Lauer (Paris, 1949).

Flodoard,

Annales

s.a.

920, ed. Philippe Lauer (Paris, 1906), p. 2; also, though his account cannot

be fully trusted, since he telescopes different events of the reign in order to bring it dramatic and
narrative unity, Richer,

Historiae

, I.8, 15–16, 24, ed. Hartmut Hoffmann,

MGH SS

38 (Hanover,

2000), pp. 51–3, 58–9. On the relation between the two histories, see Jason Glenn,

Politics and

History in the Tenth Century: The Work and World of Richer of Reims

(Cambridge, 2004).



Early Medieval Europe





(



)

Charles the Simple, Robert of Neustria

359

the personal acts of service and grooming that were the public signs of private
favour. And in public assemblies, he sat Hagano on his left, promoting
a foreigner and relative non-entity to a position of near equality with Robert,
who for over ﬁfteen years had had the privilege of sitting on his right. Galling
enough in itself, the change was not simply a matter of public status.
It meant that at every public assembly, though Robert would speak ﬁrst,
Hagano would speak immediately after, contradicting, undercutting, per-
haps even ridiculing Robert, his words of equal weight, in fact greater
weight, since everyone would have known for whom Hagano spoke.

Calculated insults to Robert, these small performatives also pointed

to matters of principle, different ways of seeing the governance of the
West Frankish kingdom. Beginning in the last quarter of the ninth
century, that governance had increasingly privileged the position and
prerogatives of a super-elite of great magnates. Charles the Bald himself
had promoted powerful members of well-connected families to posi-
tions approaching vice-regal power and inﬂuence – men like Boso of
Vienne, Bernard Plantevelue, Hugh the Abbot, and Gauzlin of Paris.
Under Charles’s sons and grandsons, the power of these men and their
heirs had become entrenched and territorialized, to the point that they
could play king-makers, Hugh the Abbot wielding unchecked power in
the lower Loire valley, Gauzlin nearly equal power in the middle Seine,
Boso even establishing himself as king with limited but still real success
in the valley of the Rhône. Not long before his election as king, Odo
himself had succeeded to both Hugh’s power on the Loire and Gauz-
lin’s on the Seine, becoming a proto-territorial magnate himself. Both
as the tradeoff for being recognized as king by his peers and in order to
regularize a dependable structure of regional authority, Odo subse-
quently accepted the vice-regal power of certain magnates within their
territories by allowing ducal and margraval titles and prerogatives for
Boso’s brother Richard in Burgundy and Bernard’s son William in the
Auvergne. Such quasi-regalian powers enabled these dukes and mar-
graves to stand as a ‘screen’ between the king and their territories. They
appointed their clients to rule counties within their regions as viscounts
on their behalf. They controlled appointments to bishoprics within
them (deposing and even mutilating prelates who resisted their will).
They distributed beneﬁces from ﬁscal lands to their vassals. They
assumed the lay abbacies of the leading regional monasteries, giving

On the importance of such acts of precedence, see Gerd Althoff, Spielregeln der Politik im
Mittelalter: Kommunikation in Frieden und Fehde (Darmstadt, 1997). On performative analy-
ses of ritual, Geoffrey Koziol, ‘A Father, his Son, Memory, and Hope: The Joint Diploma of
Lothar and Louis V (Pentecost Monday, 979) and the Limits of Performativity’, in Jürgen
Martschukat and Steffen Patzold (eds), Ritual, Inszenierung und Performanz vom Mittelalter
bis zur Neuzeit (Cologne, 2003), pp. 83–103.

Early Medieval Europe

  ()



360

Geoffrey Koziol

them access to abbatial lands with which to raise troops and further
reward their followers while also affording them a theatre within which
they could sacralize their authority in liturgical settings. Within these
territories, in other words, the king did not intervene directly; so far as
he ruled at all, he ruled through the great territorial magnates.

This, then, was the kind of kingship Charles the Simple had to

accept: Richard the Justiciar as duke in Burgundy, William as duke in
Auvergnean Aquitaine, and above all Robert as margrave on the Loire
and Seine. For soon after becoming king, Odo had given his younger
brother the counties and abbacies he had controlled before his election,
while the reorganization that regularized the margraval authority of
Richard and William allowed Odo to establish Robert as margrave in
Neustria as well.

During Odo’s reign, Robert was therefore margrave

In general, see Jan Dhondt, Etudes sur la naissance des principautés territoriales en France
(IX

–X

siècles) (Bruges, 1948), especially pp. 81–119; Werner, Les origines, pp. 466–76; idem,

‘Westfranken-Frankreich unter den Spätkarolingern und frühen Kapetingern (888–1060)’,
in Theodor Schieder (ed.), Handbuch der europäischen Geschichte, vol. 1 (Stuttgart, 1976),
pp. 731–83. However, only detailed studies reveal the logic of these transformations, particu-
larly vis-à-vis the formation of the Robertian ‘principality’: K.F. Werner, ‘Gauzlin von Saint
Denis und die west-fränkische Reichsteilung von Amiens (880)’, Deutsches Archiv 35 (1979),
pp. 395–462, reprinted in Von Frankenreich zur Entfaltung Deutschlands und Frankreichs
(Sigmaringen, 1984), pp. 157–224; idem, ‘Les premiers Robertiens et les premiers Anjou (IX

siècle–début X

siècle)’, in Olivier Guillot and Robert Favreau (eds), Pays de Loire et Aquitaine

de Robert le Fort aux premiers Capétiens, Actes du colloque international tenu à Angers en
septembre 1997, Mémoires de la Société des antiquaires de l’Ouest et des musées de Poitiers,
5th ser. 4 (Poitiers, 1997), pp. 9–67, especially pp. 28–30; Hubert Guillotel, ‘Une autre
marche de Neustrie’, in K.S.B. Keats-Rohan and C. Settipani (eds), Onomastique et parenté
dans l’Occident médiévale (Oxford, 2000), pp. 7–13. Also K.F. Werner, ‘Faire revire le souvenir
d’un pays oublié: La Neustrie’, and J.-P. Brunterc’h, ‘Le duché du Maine et la marche de
Bretagne’, in Hartmut Atsma (ed.), La Neustrie: les pays au nord de la Loire de 650 à 850, 2
vols, Beiheft der Francia 16:1–2 (Sigmaringen, 1989), I, pp. xiii–xxxi, 29–127; Walther Kienast,
Der Herzogstitel in Frankreich und Deutschland (9. bis 12. Jahrhundert) (Munich, 1968),
pp. 55–7 with nn. 2–3; Otto G. Oexle, ‘Bischof Ebroin von Poitiers und seine Verwandten’,
Frühmittelalterliche Studien 3 (1969), pp. 138–210; Bernd Schneidmüller, Die Welfen:
Herrschaft und Erinnerung (819–1252) (Stuttgart, 2000), pp. 66–70; K.F. Werner, ‘Missus-
marchio-comes: entre l’administration centrale et l’administration locale de l’empire caro-
lingien’, in W. Paravicini and K.F. Werner (eds), Histoire comparée de l’administration
(IV

–XVIII

siècle), Beihefte der Francia 9 (Munich, 1980), pp. 191–239, reprinted in Von

Frankenreich zur Entfaltung Deutschlands und Frankreichs, pp. 108–56.

Olivier Guillot, Albert Rigaudière and Yves Sassier, Pouvoirs et institutions dans la France

médiévale: des origines à l’époque féodale, 2 vols (Paris, 1994), I, pp. 156–61; J.-P. Brunterc’h,
‘Naissance et afﬁrmation des principautés au temps du roi Eudes: l’exemple d’Aquitaine’, in Guillot
and Favreau (eds), Pays de Loire et Aquitaine, pp. 69–119. Note in particular that although Robert
had been abbot of Saint-Martin since 890 at the latest (Recueil Eudes, ed. Bautier, no. 19
(29 February 888–22 March 890); also Recueil Robert I

, ed. Dufour, Appendix I, no. 37 (13 June

892)), he is not actually attested as margrave until 893, and then only in acts associated with
his Loire holdings, i.e., the command attached to the lay abbacy of Saint-Martin (Recueil Eudes,
ed. Bautier, nos. 33 (28 May 893), 34 (15 October 893)). Signiﬁcantly, then, the regularization
of the vice-regal status of William, Richard and Robert came in immediate response to
the threat raised by Charles the Simple’s anointing by Fulk of Reims (28 January 893) and
was closely associated with the joint stand by William, Richard and Odo in April or May 893
(accepting, provisionally, the reading of Brunterc’h, ‘Naissance et afﬁrmation’, pp. 81 and
107 n. 121, which has William and Richard aiding Odo against Charles, not opposing him).



Early Medieval Europe

  ()

Charles the Simple, Robert of Neustria

361

in the lower Loire and middle Seine. He personally held a large num-
ber of the most important counties throughout the region. With
his brother, he would have controlled the appointment to bishoprics
throughout his march. And he was lay abbot of at least Saint-Martin of
Tours and Marmoutier,

and probably also of Saint-Aignan of Orléans,

Saint-Germain-des-Prés, Morienval, and Saint-Amand.

As historians have reconstructed events, the agreement between Odo

and Charles stipulated that in return for being allowed to succeed Odo
without challenge from Robert, Charles agreed to recognize Robert’s
position as margrave. Accordingly, it is generally believed that on Odo’s
death Robert recognized Charles as king, while Charles recognized Robert’s
margravate. But in 900, Manasses, the leading client of Richard the
Justiciar, made disparaging remarks about Robert to Charles. We do not
know what Manasses said, but whatever it was, the insult was so grave
that Robert immediately left court and stayed away for three years.

When he ﬁnally returned to court in 903, Charles immediately gave him
a position of utter pre-eminence that never weakened, as manifested by
the honoriﬁcs attached to Robert in the diplomas Charles issued for
him. In 903, he is ‘the count much beloved by us’ and ‘our dearest Robert’.

In 914, he is ‘the margrave and abbot most beloved by us’.

Between

914 and 918, the honoriﬁcs become still more extravagant, unprece-
dented formulas assigning Robert not just pre-eminence but an almost
vice-regal power under the king within the West Frankish kingdom:
‘Robert, the most faithful executor of our serenity’; ‘Robert, venerable
margrave, the counsel of our kingdom and, alongside us, its aid’.

These are not insigniﬁcant formulas, cheap because inﬂated. As we will
see, the earliest ones both signiﬁed and were occasioned by a major

At this time the abbacies of Marmoutier and Saint-Martin of Tours were joined: sacked by

the Northmen in 853, the monks of Marmoutier had fled to Saint-Martin, which installed
canons at Marmoutier and ruled it until its reform c.982: Charles Lelong, ‘Etudes sur l’abbaye
de Marmoutier’, Bulletin de la Société archéologique de Touraine 39 (1980), pp. 279–320, at
pp. 282–4. The lay abbacies of Marmoutier and Saint-Martin had been joined to the com-
mand of the Loire march ever since the time of Robert the Strong, father of Odo and Robert,
and his successor Hugh the Abbot: Schneidmüller, Die Welfen, pp. 66–7.

On Robert’s counties and abbeys see (broadly) Sassier, Hugues Capet, p. 73; Werner, Les

origines, p. 495; Dhondt, Etudes, p. 101; Eckel, Charles le Simple, pp. 34–8 (though very dated).
For more nuanced and precise information, Recueil Robert I

, ed. Dufour, Appendix I,

especially pp. xci–xcii; Werner, ‘Gauzlin von Saint Denis’; idem, ‘Les premiers Robertiens’.

Annales Vedastini, s.a. 900, ed. B. de Simson, MGH SRG 12 (Hanover, 1909), p. 82.

Recueil Charles III, ed. Lot/Lauer, no. 45, pp. 94–7 at p. 96 (25 April 903): ‘comes nobis
admodum dilectus, nomine Rotbertus’; no. 46, pp. 97–102 at p. 101 (30 April 903): ‘sepe
nominati abbatis Rotberti carissimi quidem nostri’.

Recueil Charles III, ed. Lot/Lauer, no. 77, pp. 172–4, and no. 78, pp. 174–6, at pp. 173, 175
(both 19 June 914): ‘Robertus, dilectissimus nobis marchio atque abbas’.

Recueil Charles III, ed. Lot/Lauer, no. 81, pp. 180–2 at p. 181 (25 August 915): ‘Rotbertus,
nostre serentiatis exequtor ﬁdelissimus’; no. 92, pp. 209–12 at p. 211 (14 March 918): ‘Rotber-
tus, venerabilis marchio, nostri quidem regni et consilium et juvamen nobiscum simulque’.

Early Medieval Europe

  ()



362

Geoffrey Koziol

upheaval that ousted Robert’s competitors from court and sealed his
dominance there. The later ones suggest, tantalizingly but indetermin-
ately, a new institutional reality that placed Robert in a signiﬁcantly
different, far more important position. For in 911, on the death of Louis
the Child, the magnates of Lotharingia refused to recognize the claim
of Louis’ East Frankish, non-Carolingian successor to rule them. Instead,
they turned to the sole remaining representative of the Frankish Caro-
lingian dynasty – Charles the Simple – and elected him as their king.
This made Charles’s authority almost imperial, in the early medieval
sense of the term: that is, Charles now ruled more than one kingdom.
Robert’s position as his ‘executor’ and ‘aid alongside us’ may therefore
have signiﬁed that he was no longer just a margrave exercising vice-regal
authority within a discrete province of the West Frankish kingdom. He
was, in a sense, the secundus a rege, pre-eminent in the kingdom as a
whole under Charles, anticipating the position of ‘duke of the Franks’
that his son and grandson would later hold.

In decoding these titles and epithets and in trying to understand

relations between Charles and Robert before the ﬁrst attested signs of
conﬂict in 920, historians have most recently argued that Robert’s pre-
eminence was fully supported by Charles and that between 903 and 918
relations between the two were not merely good but excellent. One
of the most critical pieces of evidence for this interpretation is the
honoriﬁcs themselves, which speak of the ‘love’ and ‘dearness’ in which
Charles holds Robert. A second is the absence of any hint of conﬂict
between them in sources covering these years. Other arguments for their
cooperation and good relations hinge less on explicit evidence than on
induction. Of particular importance here is Charles’s settlement with
the Normans of the Seine in 911, by which he granted them lands on
the lower Seine in return for their conversion. It has been increasingly
recognized that this was Charles’s own initiative, worked out with
Heriveus, his archchancellor and archbishop of Reims. Yet the settlement
involved territories that were at least technically within Robert’s pur-
view and concerned a hostile, pagan people whom Robert’s margravate
had been set up to combat. Not only must Robert have been closely
involved for the settlement to have worked, he must have accepted it,
since he stood as godfather to the newly baptized Norman leader. It is
also argued that far from creating a source of conﬂict between them,
Charles’s takeover of Lotharingia in 911 and Robert’s assumption of
effective vice-regal authority in the west actually beneﬁtted both: Robert
gained by seeing his authority in West Francia increased, which left

Thus Werner, Les origines, p. 496; Sassier, Hugues Capet, p. 79. Others do not agree: Guillot,
et al., Pouvoirs et institutions, I, pp. 159–61. As discussed below, another interpretation may
be more accurate.



Early Medieval Europe

  ()

Charles the Simple, Robert of Neustria

363

Charles free to concentrate on extending his control over Lotharingia.
Thus, although Charles’s control of Lotharingia increased his power, it
did not threaten Robert. On the contrary, it actually removed sources
of contention and competition between them in West Francia and
therefore gave each a stake in preserving the status quo between them.

One of the strengths of these arguments for the two leaders’ coopera-

tion before 920 is that they help explain the timing and causes of its
breakdown. Charles had hoped Lotharingia would provide him free-
dom of action by giving him access to new lands and clients. Instead,
he ended up facing exactly the same problems there that he had in West
Francia. In order to rule effectively in Lotharingia Charles had to
diminish the independence of those magnates whose families had come
to power under Arnulf and Louis the Child. This meant, in particular,
Gislebert, son of the margrave Reginar. Reginar had been Charles’s most
important supporter in Lotharingia. When he died in 915, his son
expected to step into his father’s place. Wanting to keep the upper
hand, Charles refused. But denying Gislebert’s claims led Charles to
rely on rival Lotharingian families for support, which only further alien-
ated Gislebert. The conﬂict came to a head in 919, when Charles began
to dismantle Gislebert’s lay abbacies and Gislebert went into outright
rebellion. In order to support his rebellion, Gislebert sought alliances
with the new East Frankish king, Henry I, and with Robert.

At the

same time, Charles’s troubles in Lotharingia forced him into more active
engagement in West Frankish policies, as he now tried to use West
Frankish resources to buttress his power in Lotharingia. Signiﬁcantly,
the most important of Charles’s new Lotharingian clients was Hagano
himself, for whom Charles tried to create a material basis of power in
West Francia, thereby weakening Robert and openly infringing on his
public role as the king’s ‘executor’ and ‘counsellor of our kingdom’.

This is the interpretation especially of Werner and Sassier. For the details, particularly the
proposition that the Norman settlement was devised by Heriveus and Charles rather than
Robert, see Olivier Guillot, ‘La conversion des Normands à partir de 911’, in Histoire religieuse
de la Normandie (Chambray, 1981), pp. 23–53; idem, ‘La conversion des Normands peu après
911: Des reﬂets contemporains à l’historiographie ultérieure (X

–XI

s.)’, Cahiers de civilisation

médiévale 24 (1981), pp. 101–16, 181–219, especially Part I.

Sassier, Hugues Capet, pp. 79–80; Werner, Les origines, pp. 506–9; idem, ‘Westfranken-Frankreich’,
pp. 740–1; Eckel, Charles le Simple, Ch. 4; Rosamond McKitterick, The Frankish Kingdoms
under the Carolingians, 751–987 (London, 1983), pp. 307–9; R. Parisot, Le royaume de Lorraine
sous les Carolingiens (843–923) (Paris, 1899), pp. 574–659; Joachim Ehlers, ‘Carolingiens, Robertiens,
Ottoniens: Politique familiale ou relations franco-allemandes’, in Michel Parisse and Xavier
Barral i Altet (eds), Le roi de France et son royaume autour de l’an mil (Paris, 1992), pp. 39–45;
Rüdiger Barth, Der Herzog in Lotharingien im 10. Jahrhundert (Sigmaringen, 1990), pp. 39–45,
54–8; Eduard Hlawitschka, Lotharingien und das Reich an der Schwelle der deutschen Ges-
chichte, MGH Schriften 21 (Stuttgart, 1968), pp. 194–205; Walter Mohr, Geschichte des Her-
zogtums Gross-Lothringen (900–1048) (Saarbrücken, 1974), pp. 12–21; idem, ‘Die Rolle Lothringens
im zerfallenden Karolingerreich’, Revue belge de philologie et d’histoire 47 (1969), pp. 361–98.

Early Medieval Europe

  ()



364

Geoffrey Koziol

The historiography that has championed these interpretations is

learned, precise and subtle – medieval scholarship at its ﬁnest. Yet one
can sometimes be too learned, as when the tenacity required to master
and interpret a galaxy of details causes one to lose sight of the intuitively
obvious. To begin with, the absence of evidence of conﬂict between the
two men during this period cannot be taken as evidence for the absence
of conﬂict, not when there is no contemporary narrative evidence at all
for the years between 900 and 919, not when contemporary narrative
sources that do exist for the years immediately before 900 and after 919
show the two in ﬁerce, open conﬂict. Related to this problem is a
second issue: in the absence of contextualizing narrative sources, the few
isolated fragments of information we have are capable of diametrically
opposing interpretations, meaning that in and of themselves they prove
nothing. For example, in 911, as part of the settlement between the
Normans and Charles, Robert stood as godfather to the Norman leader
Rollo.

This indicates his participation in the rituals surrounding the

peace, a necessity since the Normans were being settled on the frontier
of his march. Why must it indicate his enthusiastic agreement? From
everything we know, the Norman settlement was principally Charles’s
policy, the Normans principally his allies. They had been so in 893
(against Odo), and as we will see in 898 (against Robert), and yet again
in 923 (against Robert’s ally and successor Raoul).

In marked contrast,

Robert’s father had been killed in battle against Normans. His brother
had been besieged by them in Paris for nearly two years, while his
recognition as king had been sealed by his victory over them at Mont-
faucon. The treaty of 911 itself had been made possible by Robert’s own
celebrated victory over Rollo at Chartres.

Robert and his family there-

fore had a long history of deep antagonism to Normans. Charles had a
history of using Normans against Robert and his family. Why, then,
conclude that Robert was happy to see Charles agree to settle the Nor-
mans on his own Neustrian frontier, directly threatening one of his
own abbeys (Saint-Amand) and enveloping the lands of another (Saint-
Germain)?

Such doubts are reinforced by another consideration: Robert’s

See the references above, n. 17.

Eleanor Searle, ‘Frankish Rivalries and Norse Warriors’, Anglo-Norman Studies 8 (1985),
pp. 198–213; also Guillot, ‘Conversion des Normands peu après 911’, p. 112; idem, ‘Conversion
des Normands à partir de 911’, pp. 28–9; Schneider, Erzbischof Fulco von Reims, pp. 163–4;
Flodoard, Historia Remensis ecclesiae, ed. Martina Stratmann, MGH SS 36 (Hanover, 1998),
IV.5, pp. 384–5; Flodoard, Annales, s.a. 923, p. 14; also below, on Charles’s actions in 898.

Annales de Saint-Bertin, s.a. 866, ed. Félix Grat, Jeanne Vielliard and Suzanne Clémencet
(Paris, 1964), pp. 130–1; Abbo of Saint-Germain, Le Siège de Paris par les Normands, ed. Henri
Waquet, 2nd edn (Paris, 1964), especially II, ll. 491–531, pp. 102–7, for Montfaucon, with
Annales Vedastini, s.a. 888, p. 65; Flodoard, Historia Remensis ecclesiae, IV.14, p. 407.

The problems the settlement created for Saint-Germain are attested and were dealt with in
an act of Charles the Simple, Recueil Charles III, ed. Lot/Lauer, no. 92 (14 March 918).



Early Medieval Europe

  ()

Charles the Simple, Robert of Neustria

365

victory at Chartres was made possible by the fact that he had been
joined there by another magnate whose ascent had been built on his
hostility to the Normans: Richard the Justiciar, duke of Burgundy, hero
of Argenteuil.

Up to that moment Robert and Richard had been

committed enemies, their hostility going back all the way to the siege
of Paris. From that moment, Robert and Richard began to reconcile. It
is quite visible in acts of 914 and 915.

In 918, when Richard’s capacities

started to fail, he began to turn over government to his son Raoul, who
immediately established a committed alliance with Robert directed
squarely against Charles, spelling his doom.

But the reconciliation had

begun with Richard’s decision to aid Robert against the Normans at
Chartres in 911. Against this backdrop, is the fact that Robert acted as

Guillot, ‘La conversion des Normands à partir de 911’, pp. 30–1, with multiple sources quoted
by Maurice Chaume, Les origines du duché de Bourgogne, Part I (Dijon, 1925), pp. 353–8;
though the value of Dudo is minimal, the validity of other accounts is indisputable.

Recueil Charles III, ed. Lot/Lauer, no. 79 (21 June 914), for Richard’s son Hugh, is Charles’s
ﬁrst intervention for the family since 908; ibid., no. 82 (10 October 915) has Robert and
Richard acting jointly. Given Robert’s domination of formal expressions of precedence
around Charles since 903, a domination accompanied by Richard’s almost complete exclusion
(see below), these two acts must be seen as signs of their continuing steps towards reconciliation.

Recueil Robert I

, ed. Dufour, no. 51 (918, after September 1) is the linchpin of a major alliance

between Robert and Raoul, insofar as Raoul and Richard approve Robert’s and Odo’s
association with Charles the Bald in commemoration and prayer while conspicuously exclud-
ing Charles the Simple; and the author of the act is Walo, bishop of Autun and brother of
Manasses, whose insult had triggered Robert’s withdrawal from Charles’s court in 900 and
the severing of all ties between Robert and Richard. In this act, Walo is making amends, not
least for his probable involvement in the assassination of his predecessor, Odo’s chancellor,
Adalgarius (whose donation to Saint-Nazaire of Autun, formerly conﬁrmed by Odo, is here
reconﬁrmed by Walo). Given this charter, 918 must also be the approximate date of Raoul’s
marriage to Robert’s daughter Emma (most likely, the charter’s terms prepared the way for
the marriage alliance by clearing away old grievances). Richard had also failed to support Odo
at the siege of Paris; in return, the peace made by Charles the Fat under Odo’s inﬂuence
ended the siege by giving the Vikings free passage to advance on Richard’s Burgundy. This
is the likely origin of the hostility between the two houses, quite visible in Abbo of Saint-
Germain’s praise-poem of Odo: Siège de Paris, II, ll. 340–6, 340–4, and especially 471–2
(versus MacLean, Kingship and Politics, p. 61: in fact, it was not Richard’s Burgundians who
came to Paris, but Charles the Fat’s loyalists, such as Geilo of Langres, and clients of the late
Hugh the Abbot, whose Burgundian monasteries Richard did not yet control). The hostility
would have deepened with the assassination of Adalgarius; with Richard’s seizure of Sens in
895/6 and his imprisonment of Odo’s kinsman and crucial supporter Archbishop Walter; and
with his takeover of Saint-Germain of Auxerre, which Hugh the Abbot had ruled and which
Odo wanted along with the rest of Hugh’s abbacies. Edouard de Saint-Phalle, ‘Comtes de
Troyes et de Poitiers au IX

siècle: Histoire d’un double échec’, in Keats-Rohan and Settipani

(eds), Onomastique et parenté, pp. 154–70, at p. 155; Guillot, ‘Les étapes de l’accession
d’Eudes’, pp. 202–6; Werner, ‘Westfranken-Frankreich’, p. 737. For Saint-Germain of
Auxerre, see, for example, Die Urkunden Karls III., ed. P. Kehr, MGH Urkunden der deutschen
Karolinger 2 (Berlin, 1937), nos. 143, 145 (886; Hugh still being abbot at his death); Recueil
Eudes, ed. Bautier, no. 12 (889; Askeric, bishop of Paris and Odo’s archchancellor succeeding
Hugh as abbot); Recueil Charles III, ed. Lot/Lauer, no. 38 (901; Charles recognizing Richard’s
abbacy during the period of Robert’s disfavour). This history is one of the reasons, inciden-
tally, that when in 936 Robert’s son Hugh the Great used his power over Charles’s son Louis
IV to take control of Saint-Germain, it was such an insult to Richard’s son Hugh the Black:
Recueil des actes de Louis IV, roi de France (936–954), ed. Philippe Lauer (Paris, 1914), no. 3.

Early Medieval Europe

  ()



366

Geoffrey Koziol

godfather to Rollo in 911 really evidence of his commitment to Charles
and Charles’s Norman policy? Since baptismal sponsorship between
leaders of equal status was one of the conventional formalities governing
peace ceremonies with converting pagans, Robert’s sponsorship of Rollo
can easily be interpreted, with equal if not greater justiﬁcation, as bowing
to political necessity, since Charles and Heriveus had created a treaty
he could not oppose without losing all claim to a moral high ground.

The honoriﬁcs attributed to Robert pose a similar problem. They

simply cannot be taken at face value. 914 and 915 are the years that
Robert’s epithets in Charles’s acts approach a level of exaltation not
seen since Boso had dominated Charles the Bald’s imperial court.

But

as just noted, they also happen to be the very years Robert and Richard
began to form a lasting alliance. Already there is a disjunction between
political image and political reality, and it only gets wider. In 916, we
ﬁnd the ﬁrst trace of Charles’s future favourite Hagano in his retinue.
By 917 Hagano had gained a public position of unusual intimacy with
Charles, since he then appears in one of the king’s most personally
important diplomas involving Compiègne and his late queen Frederuna.
The very next year, 918, Hagano was granted the extreme honour
of receiving commemoration at Compiègne alongside Frederuna and
Charles himself. Not even Robert was ever permitted such favour.

Yet

just two months earlier a diploma of Charles had referred to Robert as
‘counsel of our kingdom and, alongside us, its aid’, while a few months
later, Robert, Richard and Raoul have sealed their alliance against
Charles. So much for the evidentiary value of laudatory epithets. The
honoriﬁcs given Robert in Charles’s diplomas during these years are no
evidence at all of Charles’s real feelings about Robert, nor of Robert’s
commitment to Charles. They are not even evidence of Robert’s stand-
ing at Charles’s court, since he was losing real inﬂuence to Hagano at the
very moment his epithets are the most extravagant. They are the language
of politics, no more a guide to Charles’s true attitude than a conserva-
tive senator’s public reference to his liberal opponent as ‘my esteemed
colleague’. Indeed, because Robert’s grandest epithets come precisely at
the time of Hagano’s rising inﬂuence, they were most likely meant as a
sop to ease his justiﬁable sense of a growing threat, or demanded by him
as public signs of continued commitment in the face of that threat.

Arnold Angenendt, Kaiserherrschaft und Königstaufe: Kaiser, Könige und Papste als geistliche
Patrone in des abendländischen Missionsgeschichte, Arbeiten zur Frühmittelalterforschung 15
(Berlin, 1984). The locus classicus, of course, is the baptism of Harald in 826.

Above, nn. 14–15. For Boso, see, for example, Recueil des actes de Charles II le Chauve, roi de
France, ed. Georges Tessier, 3 vols (Paris, 1943/55), II, nos. 444, 458, 460 (from 875–77).

Recueil Charles III, ed. Lot/Lauer, no. 84 (19 January 916), at p. 189; no. 90 (26 July 917);
no. 95 (26 May 918).



Early Medieval Europe

  ()

Charles the Simple, Robert of Neustria

367

One ﬁnal consideration must be taken into account in trying to

understand the events of these years: Charles and Robert could not
possibly have liked or respected each other. Everything in Charles’s
reign from its beginning to its end speaks of a passionate attachment to
an exalted idea of his kingship aimed at restoring his grandfather’s
dynasty to its rightful greatness, never more so than in the years after
911.

Everything in Robert’s life speaks to a deep conviction that great

magnates like himself possessed a special wisdom (sagacitas) that kings
needed to heed.

Temporary accommodations occasioned by passing

political convergences aside, how could there be abiding sympathy between
them? Quite apart from the principles that opposed them, Robert’s
brother had displaced Charles from the throne in 888, Charles had
fought Odo bitterly but hopelessly for ﬁve years before Odo’s death,
and Robert had been Odo’s chief military executive in the struggle
against Charles. Was all that forgotten in 898 when Charles became
king? For the past forty years, historians have worked very hard to
demonstrate that in the ninth and tenth centuries, ‘memory’ (memoria)
really meant ‘commemoration’ and ‘memorialization’ and that this kind
of memory was essentially a political contrivance, invented, erased and
resurrected according to the transient needs of shifting alliances and
changing circumstances.

This is all very true; unfortunately, the very

success of this new understanding has taken away our word for ordinary
human memories and left us without models for understanding them
and appreciating their power. So what of Charles’s boyhood memories?
How when he was eight he had been passed over for the throne by
Robert’s brother? How when he was thirteen he had been crowned king
by the archbishop of Reims, only to see his legitimate claims ignored
by magnates who sought nothing but their own political advantage?
How between the ages of eight and eighteen he had been shunted back

See especially Herwig Wolfram, ‘Lateinische Herrschertitel im neunten und zehnten Jahrhundert’,
in Herwig Wolfram (ed.), Intitulatio II: Lateinische Herrscher- und Fürstentitel im neunten und
zehnten Jahrhundert, Mitteilungen des Instituts für österreichische Geschichtsforschung, Ergän-
zungsband 24 (Vienna, 1973), pp. 19–178, at pp. 71–2, 114–22; idem, ‘Political Theory and
Narrative in Charters’, Viator 26 (1995), pp. 39–51, at p. 43; Ehlers, ‘Carolingiens, Robertiens,
Ottoniens’, p. 39. Equally signiﬁcant are the strongly, consistently ideological programmes
of Charles’s staunchest supporters among the clergy, Fulk and even more Heriveus, both
archbishops of Reims and archchancellors: Schneider, Erzbischof Fulco von Reims; Schmitz,
‘Heriveus von Reims’; idem, ‘Das Konzil von Trosly’; Bautier, ‘Recueil de textes’.

Koziol, ‘Is Robert I in Hell?’, pp. 250–1.

Patrick J. Geary, Phantoms of Remembrance: Memory and Oblivion at the End of the First
Millenium (Princeton, 1994); Karl Schmid and Joachim Wollasch (eds), Memoria: Der ges-
chichtliche Zeugniswert des liturgischen Denkens im Mittelalters, Münstersche Mittelalter-Schriften
48 (Munich, 1984); Karl Schmid and Joachim Wollasch, ‘Die Gemeinschaft der Lebenden
und Verstorbenden in Zeugnissen des Mittelalters’, Frühmittelalterliche Studien 1 (1967),
pp. 265–405. Similarly the description of acutely partisan, polemical histories as ‘social
memory’ or ‘collective memory’: e.g., Rosamond McKitterick, History and Memory in the
Carolingian World (Cambridge, 2004).

Early Medieval Europe

  ()



368

Geoffrey Koziol

and forth between Poitiers, Reims, Burgundy and Lotharingia, always
under attack, never ﬁnding a safe refuge, every magnate who promised
him support casting him aside, sometimes even turning on him, the
moment it became convenient? Were these memories so easily erased?
And throughout these ten years of his early life, the only two constants
in his life, the only two people who had stood by him no matter what,
were Archbishop Fulk of Reims and his mother Adelaide. But Fulk was
a committed enemy of Odo, Adelaide the widow of an embattled king
and stepmother to two others scarcely less embattled, who had seen
their authority and lineage not merely disputed but openly ridiculed by
Boso, who had had the temerity to have himself elected king against
them.

What stories had Adelaide told her son during the years when

they were little more than refugees? What stories had Fulk told to brace
him during yet another trek in search of yet another promised alliance?

Of course, there is no actual evidence that they told the young

Charles any stories at all. But apart from the fact that it is simply
inconceivable that they did not, something must have happened during
these years that turned a boy-refugee passed over for the throne into the
proudest of all Carolingians (with the possible exception of his grand-
father and namesake Charles the Bald, himself a king with a strong
sense of birthright hardened by an embattled boyhood). And evidence
for Charles’s pride does exist. It runs throughout his diplomas, the most
self-consciously regalian of all Carolingian acta (again with the possible

Regino of Prüm, Chronicon, s.a. 879, ed. F. Kurze, MGH SRG 509 (1890; repr. Hanover, 1978),
p. 114: ‘pro nihilo ducens adulescentes ﬁlios Ludowici et velut degeneres despiciens, eo quod
iussu Caroli eorum genitrix spreta atque repudiata fuerit’ (granting that one does not know
how seriously to take this slur). On Boso’s election and the Carolingians’ anxiety over it:
Simon MacLean, ‘The Carolingian Response to the Revolt of Boso, 879–887’, Early Medieval
Europe 10 (2001), pp. 21–48, and Laetitia Boehm, ‘Rechtsformen und Rechtstitel der Bur-
gundischen Königserhebungen im 9. Jahrhundert: Zur Krise der karolingischen Dynastie’,
Historisches Jahrbuch 80 (1961), pp. 1–57; also Stuart Airlie, ‘The Political Behaviour of Secular
Magnates in Francia, 829–879’, Ph.D. thesis, University of Oxford (1985), Ch. 5; idem, ‘The
Nearly Men: Boso of Vienne and Arnulf of Bavaria’, in Anne J. Duggan (ed.), Nobles and
Nobility in Medieval Europe: Concepts, Origins, Transformations (Woodbridge, 2000), pp. 25–
41; Johannes Fried, ‘Boso von Vienne oder Ludwig der Stammler? Der Kaiserkandidat Johan-
nes VIII.’, Deutsches Archiv 32 (1976), pp. 193–208; R.-H. Bautier, ‘Aux origines du royaume
de Provence: de la sédition avortée de Boson à la royauté légitime de Louis’, Provence His-
torique 23 (1973), pp. 41–68; Recueil des actes de Louis II le Bègue, Louis III et Carloman II,
rois de France (877–884), ed. R.-H. Bautier and Félix Grat (Paris, 1978), pp. xviii–lvi; René
Poupardin, Le royaume de Provence sous les Carolingiens (Paris, 1901), Chs 2–3; Constance
Brittain Bouchard, ‘The Bosonids or Rising to Power in the Late Carolingian Age’, ‘Those
of My Blood’: Constructing Noble Families in Medieval Francia (Philadelphia, 2001), Ch. 5;
François Bougard, ‘En marge du divorce de Lothaire II: Boson de Vienne, le cocu qui fut fait
roi?’, Francia 27 (2000), pp. 33–51. On Adelaide: K.F. Werner, ‘Die Nachkommen Karls des
Grossen bis um das Jahr 1000 (1.

−8. Generation)’, in Karl der Große: Lebenswerk und Nach-

leben, ed. Wolfgang Braunfels, 5 vols (Düsseldorf, 1966–68), IV, pp. 403–82, at pp. 429–41;
Sylvia Konecny, Die Frauen des karolingischen Königshauses: Die politische Bedeutung der Ehe
und die Stellung der Frau in der Fränkischen Herrscherfamilie vom 7. bis zum 10. Jahrhundert,
Dissertationen der Universität Wien 132 (Vienna, 1976), pp. 142–3.



Early Medieval Europe

  ()

Charles the Simple, Robert of Neustria

369

exception of his grandfather’s).

It lies in his commitment to elevating

the palace complex at Compiègne built around Saint-Corneille (founded
by the same grandfather) into the ceremonial capital of the kingdom,
which he turned into a memorial of his dynasty and his birthright.

shows itself in his belief that he, as king, could exalt a low-level noble
like Hagano to a position of equality with a margrave who was the
brother of a king. Such policies speak of an abiding, focused desire to
restore his dynasty to greatness, a greatness that had been forgotten and
discarded by men who were upstarts, no matter what their own claims
to wisdom and authority. Only molten boyhoods harden into adult
character so adamantine. Shaped and hardened by his mentor and his
mother, the only two people he could trust, would not his memories
have been the one certain touchstone of his life? Is it conceivable that
the boy whose rights were cast aside by Odo and his brother could have
become a man who happily exalted that same brother? Is it not more
plausible to believe that Charles’s treatment of Robert was nothing but
a distasteful but necessary accommodation?

And then there is the uprising itself. Seeing it as the sudden collapse

of two decades of harmonious relations between Robert and Charles
leaves too much unaccounted for. The seriousness of the step of actu-
ally deposing a king was extremely unusual in early medieval history –
particularly in late ninth and tenth century West Francia, where kings
were opposed, cornered and badgered, but not deposed.

And the ﬁnal

battle that Charles launched against Robert to decide the contest, by
surprise on a Sunday during a truce, when Robert and his men were at
their midday meal after Mass – this is the desperate gambit of passion-
ate conviction, a hatred so righteous it makes wrong right. To suggest
otherwise, to claim that this was simply a rough age’s business as usual,
is itself an assumption requiring proof and not at all borne out by the
few Frankish battles approaching it in consequence.

It also requires a

See the works cited above, n. 29.

Recueil Charles III, ed. Lot/Lauer, no. 75, pp. 167–70 (913–22); no. 80, pp. 178–9 (915); no. 90,
pp. 202–6 (917); no. 91, pp. 206–9 (917); no. 93, pp. 212–4 (918); no. 95, pp. 217–21 (918);
no. 96, pp. 221–3 (918–23); no. 109, pp. 261–4 (921); no. 122, pp. 288–90 (923). See May
Vieillard-Troïekouroff, ‘La chapelle du palais de Charles le Chauve à Compiègne’, Cahiers
archéologiques 21 (1971), pp. 89–108; Stuart Airlie, ‘The Palace of Memory: The Carolingian
Court as Political Centre’, in Courts and Regions in Medieval Europe, ed. Sarah Rees Jones,
Richard Marks and A.J. Minnis (York, 2000), pp. 1–20, at pp. 14–17.

Koziol, ‘Is Robert I in Hell?’, pp. 242–3, n. 27.

Of the two other great battles that were fought by Franks against Franks (both also thought
of as judgements of God), Fontenoy (25 June 841) was fought on a Saturday, Andernach
(8 October 876) on a Monday. In both, Charles the Bald tried for surprise, and in both
negotiations continued right to the eve of battle (those at Andernach in apparent bad faith),
but neither occurred during truces; quite the contrary, in both cases the two sides had been
publicly escalating military preparations and an imminent battle was expected. (This is well
brought out by Adelheid Krah, Die Entstehung der ‘potestas regia’ im Westfrankenreich während

Early Medieval Europe

  ()



370

Geoffrey Koziol

jaded discounting of what our sources actually tell us. The one closest
to Charles says that after Robert fell in battle, Charles personally plunged
his lance into his mouth all the way through his skull, because that
mouth had sworn false ﬁdelity.

And the shock of the deposition runs

through charters issued after the deposition (and even after Charles’s
ﬁnal defeat and imprisonment), not simply in that they continue to be
dated by Charles’s regnal years, but more tellingly in the judgements
they pronounce: ‘14 Kalends September, the year that Robert reigned
falsely’; ‘the Lord’s day, 3 Kalends October, the 24

year of King

Charles, the ﬁrst year that Robert rose against him’; ‘14

Kalends

March, the third year after King Charles had been dishonoured by
faithless Franks’.

When there is absolutely no narrative evidence at all

for the nearly twenty preceding years, when the argument for peace
must be based on such subtle readings of so little evidence, it simply
seems implausible to argue that an event that spun completely out of
control so quickly and tapped such deep convictions had no precedents,
especially when the reappearance of a narrative source reveals the two
protagonists already at loggerheads. Again, it seems especially implaus-
ible given the fact that Robert’s brother Odo had been crowned king at
Charles’s expense, that Charles and Odo had fought nearly throughout
Odo’s reign, that Robert himself had been Odo’s chief lieutenant in his
battle against Charles, that Charles’s childhood protector had been
Odo’s committed enemy. Could all this simply have been set aside
during the ﬁrst twenty years of Charles’s reign? Is it likely that the
insurrection that began in 919 had nothing to do with the hostilities of
893–7? Does it make sense?

It does not make much sense at all. More to the point of this article,

there is good reason to believe that there were additional grounds for
hostility between Robert and Charles, grounds that historians have
never suspected but which continued, latent, throughout the reign,
from Charles’s ﬁrst public act as king in 898 right to the ﬁnal battle in
923. As common for the period, the most important proof of this
revisionist interpretation is a piece of evidence so slight one could read

der ersten Regierungsjahre Kaiser Karls II. (840–877) (Berlin, 2000), pp. 49–86.) In 844, Pippin
II ambushed Charles the Bald’s relief forces in the Angoumois, gaining a great victory
through surprise, but the attack came on a Saturday (14 June), and far from there being any
truce, at the time Charles was engaged in full-scale military operations against Pippin at
Toulouse. Judging by the frequency with which battles were remembered in later chronicles
and annals and the ways they were described, of ninth- and tenth-century battles among
Franks, only Fontenoy exceeded Soissons in infamy. In any event, their stakes and unpredict-
ability made all-or-nothing pitched battles quite rare.

Regino, Chronicon, Continuatio, s.a. 922 (for 923), pp. 156–7: ‘Karolus tamen ori sacrilego
Ruodberti ita lancea inﬁxit, ut difﬁssa lingua cervicis posteriora penetraret.’

Recueil Robert I

, ed. Dufour, Appendix II, pp. cv–cxxii, from Agde, Cluny and Brioude.



Early Medieval Europe

  ()

Charles the Simple, Robert of Neustria

371

it a hundred times and not notice its signiﬁcance. But once noticed, it
cannot be explained away.

On 25 January 923, seven months after his

election as king, Robert issued a diploma for the monastery of Saint-
Denis, of which he was lay abbot. The diploma foresees a coming battle
against Charles, in which Robert hopes the three patrons of Saint-Denis
will give him victory, ‘so that he [Robert] might return bearing the
unconquered battle standards (vexilla) from their subjection’. What
battle standards? What did Robert mean by saying they were held in
‘subjection’? The only way to explain the statement is to admit that in
923 Charles had possession of the battle standards of Saint-Denis’ three
patron saints, Dionysius, Rusticus, and Eleutherius. Acknowledging
this fact raises another question: how and when had Charles acquired
them? And trying to answer this question calls into doubt many of the
assumptions made by historians about relations between Charles and
Robert in the opening years of Charles’s reign.

We must therefore return to the events of those years.

After ﬁve years of hostility between Charles and Odo beginning with
Charles’s coronation in 893, changing circumstances at last made a
ﬁnal settlement possible in 897. By that point, Charles had lost all hope
of ever displacing Odo. At the same time, Odo realized that he was
nearing the end of his life, and probably also understood that his
brother would never be able to succeed him (if that had ever been in
his mind – we simply do not know). In any case, historians generally
agree about the essential terms of the settlement: Odo accepted that
Charles would succeed him, in return for which Charles promised to
recognize Robert’s position as margrave of Neustria. However, since
‘margrave of Neustria’ is a broad and in some ways anachronistic title, it
is better to break the ofﬁce down into its component elements. Robert
would be recognized as holding the same position after his brother’s
death that he had held during his life: count of Paris, Etampes, Meaux,
Orléans, Tours, Blois, Chartres and Angers; lay abbot of Saint-Martin
of Tours, Marmoutier, and probably Saint-Amand, Morienval, and
Saint-Aignan of Orléans, with a military command (margravate) that
attached to the lay abbacy of Saint-Martin (and perhaps also Saint-
Amand) and covered the valleys of the lower Loire and lower and mid-
dle Seine.

Historians also agree that on Odo’s death, Robert did in

fact recognize Charles from the beginning, while Charles did in fact
recognize Robert’s margravate, including its component abbacies and

My thanks, then, to the unknown questioner who refused to allow me to sidestep the phrase
when I ﬁrst presented a reading of Robert’s diploma in Prof. William Jordan’s Medieval
Studies seminar at Princeton University in December 2001.

See the references above, nn. 10–11, and the discussion below.

Early Medieval Europe

  ()



372

Geoffrey Koziol

counties.

All this is what historians believe happened. However, the

precise sequence of events as given by our only contemporary narrative
source, the Annals of Saint-Vaast, does not fully support this reconstruc-
tion. Since the details of its account are extremely important, they should
be revisited.

According to the Annals, at some point in the middle of 897, Charles’s

few remaining supporters (including probably Fulk of Reims) approached
Odo, asking him to give Charles a portion of the kingdom of Charles’s
father (Louis the Stammerer).

Following further negotiations, Odo

received Charles personally ‘and gave him as much of the kingdom as
seemed appropriate to him, and promised him greater, and sent him
back to his place, after reconciling Herbert [of Vermandois] with him’.
In the ﬁnal days of this same year, as Odo lay ill, ‘he began to ask
everyone to show faith to Charles’.

When he died (in the ﬁrst days of

January 898), he was buried at Saint-Denis. Afterwards, ‘the Franks’
convened at Reims ‘and restored Charles to his paternal seat’. Though
Baldwin of Flanders did not appear (because of a feud with Herbert I
of Vermandois, now one of Charles’s supporters), he sent legates to
Charles saying he would be ‘faithful to him, as was ﬁtting’. That spring
(verno tempore), the Normans began to attack Neustria and part of
Aquitaine. Only now do the Annals mention Robert:

After this Count Robert, brother of King Odo, came to the king; the
king received him honourably, and having been made his ﬁdelis, he

Above, nn. 3, 5.

We do not know the date of either the negotiations or the agreement between Odo and
Charles. Charles’s earlier projected alliance with the Normans culminated in the baptism of
the Norman chieftain, with Charles standing as godfather, on Easter (27 March) 897, so the
negotiations with Odo would have occurred signiﬁcantly after that date. After Charles’s
agreement with Odo (viz., after he had effectively sold them down the river), the Normans
began to rampage; Odo negotiated a truce with them, part of whose stipulations pertained
to their winter quarters. A rare act of Odo for Richard at Nanteuil (near La Fère, where Odo
died) on 21 October 898 (Recueil Eudes, ed. Bautier, no. 42) only makes sense if it was given
during a meeting between the two at which Odo solicited Richard’s agreement to the settle-
ment, though whether that was after the settlement had been negotiated or in order to
proceed with negotiations cannot be guessed. The agreement between Odo and Charles was
therefore negotiated considerably after Easter 897, at a time when winter plans for 897/98
were becoming an issue for the Normans, and either shortly before or shortly after the
diploma for Richard – i.e., in October.

Annales Vedastini, s.a. 897, pp. 78–9: ‘Verum post haec [i.e., the baptism of the Norman
Hundeus] hi qui cum Karolo erant, videntes suam paucitatem et nullum tutum habere locum
refugii, iterum ad Odonem regem dirigunt, quatinus ad memoriam reduceret, quod senior
eorum ﬁlius esset sui quondam senioris, et partem aliquam ei ex paterno regno concederet.
At vero rex cum consilio suorum respondit se illi velle misereri, si sibi liceret; et intercurren-
tibus nuntiis venit Karolus ad eum; quem ille benigne suscepit, deditque ei tantum e regno,
quantum sibi visum fuit, promisitque maiora et remisit eum ad locum suum, paciﬁcato
Heriberto cum eo . . .’ Late that year, on his deathbed, Odo, ‘qui dum languor per dies
singulos incresceret, omnibus rogare coepit, ut Karolo servarent ﬁdem’.



Early Medieval Europe

  ()

Charles the Simple, Robert of Neustria

373

returned to his [territory]. Richard [of Burgundy] and William [of
Aquitaine] did likewise. After this King Charles with a small army
pursued the Normans as they were returning from their plundering
of the pagus of Vimeu . . . and after some of them were killed and
many wounded, the Normans, judging the location poor, returned
to their ships in their usual manner.

Essentially, this is all our sole surviving narrative source says about the
regnal transition. Everything else is reconstruction, though there has
been a fair amount of agreement among historians in their suppositions.

First, historians now accept that the 897 agreement between Odo

and Charles was not a partition of the kingdom, as was once thought.
Negotiating from a position of strength, Odo simply assigned Charles
‘as much of the kingdom as seemed appropriate to him’, that is, ‘a
place’ to which he ‘sent Charles back’. Following Edouard Favre, most
(but not all) historians argue that Charles was probably given control
of Laon; I tend to agree with them, for the reasons Favre gives while
(like Karl Ferdinand Werner) putting more emphasis than Favre on the
importance of Laon to Charles’s son and eventual successor Louis IV.

Second, while giving Charles some part of the kingdom, Odo also

promised him ‘greater things’ (maiora). Most historians assume that
this refers to a promise by Odo to allow Charles to succeed him at the
expense of Odo’s brother Robert, with the assumption that everyone
knew the succession was not far off because of Odo’s ill health. Again
I tend to agree.

Third, historians assume that the agreement between Odo and

Charles had a further stipulation: that in return for Robert’s giving up
any claim to succeed his brother as king, Charles agreed to allow him

Annales Vedastini, s.a. 898, pp. 79–80: ‘Obiit ipse [Odo] in eodem loco Kalendis Ianuarii,
corpusque eius apud Sanctum Dionisium delatum ibique honoriﬁce humatum. Franci vero
rege mortuo die . . . [sic] Remis conveniunt Karolumque in sede paterna restituunt. Balduinus
vero propter Heribertum venire distulit, attamen missos dirigit, qui regi innotescerent se illi
ﬁdelem esse, sicut dignum erat. Nortmanni vero verno tempore rediere ad naves, vastatam
Aquitaniae partem atque Neustriam, insuper plurima eversa castra, interfectis habitatoribus.
Post haec Rothbertus comes, frater regis Odoni, venit ad regem; quem rex honoriﬁce suscepit,
eiusque ﬁdelis effectus rediit ad sua. Similiter fecit et Richardus insuper et Willelmus. Post
haec rex Karolus cum exercitu parvo Nortmannis a praeda revertentibus in pago Vitmau iuxta
quandam . . . [sic] insecutus, aliquibus suorum interfectis plurimisque vulneratis, Nortmanni
more solito loca inoportuna tenentes rediere ad naves.’ The ellipses here are lacunae in the
manuscript. One cannot draw any conclusions from the fact that the annalist refers to Robert
as comes, since he is conservative in his usage of titles, usually assigning none, and at best
using only rex and comes.

Favre, Eudes, p. 190; Werner, ‘Westfranken-Frankreich’, p. 738. Schneider, Erzbischof Fulco
von Reims, pp. 167–8, espouses a maximalist position, which is unlikely given Odo’s strength
and the wording of the Annals of Saint-Vaast, which makes the territory ceded to Charles
seem small and places all the initiative in Odo’s hands (‘se illi velle misereri, si sibi liceret’,
‘quem ille benigne suscepit, deditque ei tantum e regno, quantum sibi visum fuit’).

Early Medieval Europe

  ()



374

Geoffrey Koziol

to retain all the honours he had held under Odo. They also assume that
this alleged stipulation was immediately executed: that upon Odo’s
death Robert recognized Charles as king, while Charles accepted Robert’s
position as margrave with all the lands and ofﬁces he had held, including
his abbacies, and that the transition between the reigns was therefore
amicable.

In fact, the Annals of Saint-Vaast make no such statements. They

must be inferred ﬁrst from the fact that Robert did recognize Charles,
and second from the fact that he is known to have held many of the same
lands and ofﬁces both before and after Charles’s accession. These lands and
ofﬁces included the large concentration of counties already mentioned
and, equally if not even more important, a number of leading lay
abbacies: again, Saint-Germain-des-Prés, Saint-Aignan of Orléans, Saint-
Amand, Morienval, Marmoutier and Saint-Martin of Tours.

Fourth, historians include Saint-Denis among the abbacies they assume

Robert received immediately after Odo’s death. However, it should be noted
that Saint-Denis was not one of the abbacies Robert had held under Odo.
On the contrary, after the death of Abbot Ebalus in 892, Odo kept the
abbacy of Saint-Denis for himself, holding it to his death and being
buried in the church, near Charles the Bald.

It is surely true that Odo

intended his brother to succeed him as lay abbot, and it would have
been hard to prevent him from doing so, since Robert controlled Paris.
Still, it was not an ofﬁce he had held before Charles’s accession in 898.

Considering what the Annals of Saint-Vaast actually say about the

settlement between Charles and Odo, it may or may not be that their
agreement stipulated that Robert would continue in his ofﬁces. Personally,
I believe it likely, for in surrendering the kingship to Charles Odo would
have insisted on protecting his family’s power and in late 897, given his
strength and Charles’s weakness, been in a position to do so. But that
does not mean that Charles did not try to weaken Robert or assert himself
once Odo was gone. Nor is it provable, strictly speaking, that Robert had
continuous physical control of all or some of these monasteries imme-
diately upon his brother’s death. In fact, close reading of non-narrative
evidence suggests that Charles did not immediately accept Robert’s abbacies
(or, to be more precise, that he did not give Robert ofﬁcial recognition

Werner, Les origines, pp. 493–5; Sassier, Hugues Capet, pp. 73–6; Schneidmüller, ‘Karl III.’,
pp. 28–9; Eckel, Charles le Simple, pp. 30–1.

On lay abbacies see Franz J. Felten, Äbte und Laienäbte im Frankenreich: Studie zum Verhält-
nis Staat und Kirche im früheren Mittelalter (Stuttgart, 1980); on their importance in the
development of the lay principalities, Geoffrey Koziol, Begging Pardon and Favor: Ritual and
Political Order in Early Medieval France (Ithaca, 1992), pp. 38–9, 46; Guillot, et al., Pouvoirs
et institutions, I, pp. 155–8.

Giles Patrick Allen Brown, ‘Politics and Patronage at the Abbey of St Denis (814–98): The
Rise of a Royal Patron Saint’, Ph.D. thesis, University of Oxford (1989), pp. 164–5; Alain
Erlande-Brandenbourg, Le roi est mort: Etude sur les funérailles, les sépultures et les tombeaux
des rois de France jusqu’à la ﬁn du 13

siècle (Geneva, 1975), p. 73.



Early Medieval Europe

  ()

Charles the Simple, Robert of Neustria

375

of them). It also indicates that he directly challenged Robert’s control
of Saint-Denis. Most of this evidence is found in the royal diplomas
issued by Odo or Charles, charters issued by Robert, and notices involv-
ing him. However, before discussing these sources, we should make two
statements of principle. In evaluating the signiﬁcance of diplomas men-
tioning Robert as abbot of any given monastery, it must be kept in mind
that we have very few charters and diplomas from this period. In some ways,
we are lucky to have any mentions of Robert as abbot of, say, Morienval.
Given the few charters and diplomas we have and the apparent random-
ness of the survival of those few, it is dangerous to draw any conclusions
from the absence of any that might prove Robert was abbot of this or that
house during any given year or range of years. On the other hand, caution
can be and has been carried too far, because of a misunderstanding of the
way diplomas were actually used and the reasons they were issued. Diplomas
in this period were not administrative documents as we understand them.
They were in some ways props in political ceremonies. Thus, the
absence of diplomas may or may not be signiﬁcant, but the presence of
diplomas is almost always signiﬁcant, because issuing a diploma to a
petitioner was almost invariably tied to some major ceremony by which
a change in alliance or loyalties was publicized. In other words, if we have
a diploma in 903 that mentions Robert’s abbacy of Saint-Denis, there
is a reason we have this diploma giving this recognition in that year.

Unfortunately, a re-evaluation of these sources does not lend itself to

a clean, elegant exposition. The evidence is too scattered, fragmentary
and elusive. Setting it within a narrative before critiquing the fragments
has only allowed historians to slip into presumed continuities and unex-
amined assumptions. It is wiser to isolate the component elements, then
reassemble them into a whole, like the puzzle these years are.

A. Odo died on or about 1 January 898 and was buried at Saint-Denis.

We do not know the exact day on which Charles was enthroned as sole
king at Reims, but it must have been soon after (since Charles dated
his ‘reintegration’ from Odo’s death).

B. Robert did not submit to Charles immediately. He waited. Historians
have not read anything into this: noting that the Annals of Saint-Vaast
speak of a serious Norman incursion in Robert’s march – that is, in Neustria
and part of Aquitaine (probably the northern part along the Loire, abutting
Neustria) – they have assumed that Robert was simply preoccupied with
the defence of his territory. On the other hand, Charles followed a

Equally good sources place Odo’s death on 1 January and 2 January; a single less creditable
source (Regino) places it on 3 January. Recueil Eudes, ed. Bautier, p. clviii.

Early Medieval Europe

  ()



376

Geoffrey Koziol

consistent policy of using Normans against his enemies, and had done
so just the year before against Odo and Robert.

It is more than con-

ceivable that the Normans attacked Robert’s territories on behalf of
Charles, as they had in 897 and would again in 923. In any case, Charles
was crowned in January; the Normans attacked in the spring. There is
no reason Robert could not have attended the enthronement. Indeed,
far from the Normans preventing Robert’s attending, he actually came
only after they had attacked. The entire sequence of events makes it
appear as if Charles had connived (or at least acquiesced) in their attack
on Robert’s territories as a means of bringing him to heel.

C. Most of Robert’s lay abbacies are not actually attested until after 903.
Thus he is not attested as abbot of Saint-Amand until September 906;

Saint-Aignan until June 914;

of Saint-Germain until 918;

of Morienval

until 920.

Of course, it may well be that all or some of these abbacies

had been held by Robert since his brother’s death or before and that
we are simply missing the charters to prove it. The fact remains that there
is no actual evidence documenting his possession of any of these particular
abbacies until well into Charles’s reign, and more particularly until
after his return to court in 903 following his self-imposed exile in 900.

D. Robert is attested as abbot of Saint-Martin of Tours in an act issued
between 888–90, and otherwise in acts of 891, 892, 895, 896, 897, 899, 900,
903 and 904. There is then a gap of some years before the next series of acts
naming him as abbot, which begins in 910–11.

With respect to Saint-

Martin, therefore, we see striking evidence of continuity in Robert’s abbacy
over the course of the two reigns, with three highly signiﬁcant exceptions.

See the references above, n. 20.

Recueil Robert I

, ed. Dufour, no. 46; also no. 110 (921). See below, n. 59.

Recueil Charles III, ed. Lot/Lauer, nos. 77 and 78 (both 10 June 914).

Recueil Charles III, ed. Lot/Lauer, no. 92 (14 March 918); also no. 94 (14 May 918). As
mentioned below, in 903 he petitions a diploma for Saint-Germain, but as count of Paris,
not abbot (no. 45). The difference is slight and technical but important.

Recueil Charles III, ed. Lot/Lauer, no. 105 (20 January 920). Immediately relevant antecedent
diplomas are Recueil Eudes, ed. Bautier, no. 45, pp. 183–4 (888–98); Recueil Louis II le Bègue,
no. 90, pp. 228–9 (882–84).

Recueil Eudes, ed. Bautier, nos. 19 (888–890), 41 (896), Appendix III (891); Recueil Charles III,
ed. Lot/Lauer, nos. 46 (903), 49 (904), 63 (910–11), 101 (919); Recueil Robert I

, ed. Dufour,

nos. 37 (892), 39 (895), 40 (897), 41 (899), 42 (900), 43 (900), 45 (904). Odo had been abbot
of both Saint-Martin and Marmoutier at the time of his accession (Recueil Eudes, Appendix II).
Recueil Robert I

, no. 38 (894) attests Robert as abbot of Marmoutier (joined to Saint-Martin

since c.853 (see above, n. 10)); nos. 47 and 48 (912) attest him as abbot of both Marmoutier
and Saint-Martin. In Recueil Charles III, no. 98 (1 December 918), Charles names Robert as
count and margrave, not abbot, but this may be because the subject of the diploma –
extending Saint-Martin’s immunity to the wall newly built around the monastery – pertained
to comital jurisdiction. On the other hand, the diplomas of Saint-Martin from this period
are exceedingly problematic; one cannot press their language too hard.



Early Medieval Europe

  ()

Charles the Simple, Robert of Neustria

377

In an otherwise continuous sequence of charters and diplomas, there
are gaps in 893–4, 898 and 901–2. 901–2 are the years when Robert had
broken with Charles and completely removed himself from court. 898 is
the ﬁrst year of Charles’s reign after Odo’s death, with Robert’s late display
of acquiescence in his rule. As for 893–4, these are the years of Charles’s
initial uprising against Odo, when he was crowned king by Archbishop
Fulk of Reims. And in those very two years – two years when there is a
gap in attestations of Robert’s abbacy, during the ﬁrst months of Charles’s
ﬁrst challenge to Odo – Charles issued a diploma granting Saint-Martin
to his primary supporter, Fulk of Reims.

Of course the grant was a

dead letter. Robertian control of the Touraine was too strong. Yet Charles
did award Fulk the abbacy, and Fulk did, in fact, claim to be abbot.

E. This alone indicates Charles’s desire early in Odo’s reign, when he
was under the inﬂuence of Fulk and Adelaide, to break Robert’s control
of his most prestigious abbatia, the one most closely associated with the
margravate of Neustria, and his lack of respect for ofﬁces awarded by
his rival. It is not the only example. Robert had a strong claim to hold
the abbatia of Saint-Germain-des-Prés and was certainly recognized as
its abbot by Charles by 918; yet in 903, an act of Charles has Robert
intervening on behalf of the monastery simply as count of Paris, the
monks being referred to collectively, without any mention of Robert’s
being abbot.

Still more striking is the case of Saint-Amand. This was

another monastery to which Robert had a claim to rule, but which no
diploma of Charles acknowledged before 906.

Yet Charles did issue an

Recueil Charles III, ed. Lot/Lauer, no. 1, p. 1 (n.p., n.d.) (28 January 893–December 894), a
lost diploma mentioned in a letter of Fulk to Gui of Spoleto summarized by Flodoard,
Historia Remensis ecclesiae, ed. Stratmann, IV.5, pp. 383–4.

Schneider, Erzbischof Fulco von Reims, p. 132.

Recueil Charles III, ed. Lot/Lauer, no. 45 (25 April 903), at Compiègne: ‘quidam comes nobis admo-
dum dilectus, nomine Rotbertus, necnon et grex sanctus monachorum sancti Germani Parisiacensis,
ad nostram accedentes mansuetudinem . . .’ Nor does this act concern rights pertaining to the
county rather than the abbatia, the distinction which might conceivably explain why a late
act for Saint-Martin of Tours refers to Robert as count and margrave rather than abbot (ibid.,
no. 98 (918); above, n. 55).

Recueil Charles III, ed. Lot/Lauer, no. 54. Bishop Gauzlin of Paris, Odo’s ally, mentor and
patron, had previously held Saint-Amand along with two other abbeys that were key compon-
ents of early Robertian rule and which Robert came to hold – Saint-Germain-des-Prés and
Saint-Denis. After Gauzlin’s death in 886, Saint-Amand appears to have been held by Radulf,
son of Eberhard of Friuli and brother-in-law of Hucbald, Odo’s ally and father of his notary
Heriveus (later Charles’s notary, Fulk’s successor as archbishop of Reims, and Charles’s
archchancellor). Once Odo became king, Saint-Amand must have passed to Robert after
Radulf’s death in 892. Note also that Charles’s 906 act for Saint-Amand was issued in
conjunction with a convenientia that he seems to have required Robert to make with the
monks over their mensa in return for acknowledging his abbacy (Recueil Robert I

, no. 46,

Appendix I, pp. 180–1 and Appendix III, no. 1, pp. 208–10). Thus, the purpose of the 906
diploma was to recognize Robert’s abbacy while also asserting the principle of Charles’s superior
royal authority over it. This principle of royal authority over the magnates’ lay abbacies

Early Medieval Europe

  ()



378

Geoffrey Koziol

earlier diploma for the monks – in 899, i.e., the year after Odo’s death,
at a time when he is supposed to have recognized Robert’s ofﬁces. Far
from recognizing Robert, as at Saint-Martin, Charles set Fulk against
him; for it was Fulk who petitioned the diploma, acting in the place
where one would have expected Robert and presenting Charles with
Saint-Amand’s past diplomas as one would expect an abbot to do.

There is no doubt that this was an intentional, coded repudiation of
Robert’s abbacy, since the diploma has the monks complain about ‘the
questionable actions of a succession of rectors and other evil men’. This
can only mean Robert.

Just as Charles refused to recognize Robert’s abbacy of Saint-

Martin in 893–4 and perhaps those of Saint-Amand and Saint-Germain,
he did the same with Saint-Denis. In 896, while in Lotharingia trying to
regain Zwentibold’s support for his challenge to Odo, Charles made a
play for patronage of Saint-Denis by issuing an act restoring signiﬁcant
lands to Salonnes, a priory of Saint-Denis located in Lotharingia, near
Metz, that Zwentibold himself had recently restored to the abbey.

was an important aspect of Charles’s rule generally. On Saint-Amand’s abbots and these

relations: Henri Platelle, Le temporel de l’abbaye de Saint-Amand des origines à 1340 (Paris, 1962),
pp. 59–61; Werner, ‘Gauzlin von Saint Denis’; Grierson, ‘La maison d’Evrard de Frioul et les
origines du comté de Flandre’, Revue du Nord 24 (1938), pp. 241–66; Christina La Rocca and
Luigi Provero, ‘The Dead and their Gifts: The Will of Eberhard, Count of Friuli, and his
Wife Gisela, Daughter of Louis the Pious (863–864)’, in Frans Theuws and Janet Nelson (eds),
Rituals of Power from Late Antiquity to the Early Middle Ages (Leiden, 2000), pp. 225–80.

Recueil Charles III, ed. Lot/Lauer, no. 18, pp. 29–33 (17 March 899), at Reims: ‘quia, ad
petitionem monachorum ex coenobio almi presulis Amandi, adiens serenitatem culminis
nostri vir venerabilis Fulco, Remorum archiepiscopus, nobis carissimus, presentavit obtutibus
nostris quedam imperialia et regalia . . . scripta . . .’

‘Unde, propter suspectas succedentium rectorum vel aliorum malorum hominum voluntates,
prefati monachi per eundem venerabilem archiepiscopum devote nostram postulaverunt . . .’
The diploma itself excludes Robert’s predecessor Radulf from consideration as one of these
irresponsible rectors by mentioning him favourably as a donor to the monastery; in any case,
Radulf was the maternal uncle of Heriveus, Charles’s notary, in whose name the act was
subscribed.

Recueil Charles III, ed. Lot/Lauer, no. 7, pp. 9–12 (25 July 896, at Gondreville), the act being
petitioned by Charles’s mother, Adelaide. In an act dated 22 January 896, Zwentibold
restored the priory to Saint-Denis (Die Urkunden Zwentibolds und Ludwigs des Kindes, ed.
Theodor Schieffer, MGH Urkunden der deutschen Karolinger 4 (Berlin, 1963), no. 7, pp. 29–30).
The two acts must be related in some way, though how is puzzling, since after Zwentibold’s
abandonment of Charles the previous year, the two should not have been cooperating. It is
therefore hard to know whether their acts for Salonnes (given six months apart) are expres-
sions of a rivalry or an alliance. Signiﬁcantly, Zwentibold’s act was petitioned by the monks
of Saint-Denis. Since at the time Odo was clearly abbot and in control of the monastery, just
who these monks could have been is also something of a mystery. On Salonnes as a priory
of Saint-Denis, see Michel Parisse, ‘In Media Francia: Saint-Mihiel, Salonnes et Saint-Denis
(VII

–XII

siècles)’, in Media in Francia, pp. 319–43; idem, ‘Saint-Denis et ses biens en Alsace

et en Lorraine’, Bulletin philologique et historique (jusqu’à 1610) du Comité des travaux his-
toriques et scientiﬁques (1969), pp. 233–56; Barbara Rosenwein, Negotiating Space: Power,
Restraint, and Privileges of Immunity in Early Medieval Europe (Ithaca, 1999), pp. 115–34.



Early Medieval Europe

  ()

Charles the Simple, Robert of Neustria

379

More surprisingly and not easily explained away, on 8 February 898,
Charles was actually at Saint-Denis, issuing a diploma for the monas-
tery at the request of the monks, with the consent of the bishops of
Beauvais and Laon and also that of his mother, Adelaide. There is,
pointedly, no mention of Robert in any guise. Indeed, in the clause
where the lay abbot would normally be mentioned as the petitioner of
a diploma, Robert is not mentioned but only the monks of Saint-Denis,
without any abbot (or even count) being named at all.

Since Odo had

died holding the abbacy of Saint-Denis, it is clear from this single
instance that a month after Odo’s death, before Robert had submitted
to him, Charles was able to hold an important ceremony at the abbey
and that Robert had not been recognized by Charles as abbot. Remem-
bering that Odo had just been buried in the monastery, Charles’s action
was more than a challenge to Robert; it was an affront.

G. Charles came into sole possession of his kingship soon after Odo’s
death in early January 898. The argument with Manasses that drove
Robert to remove himself from court occurred in the summer of 900.

Between these two dates, there are no diplomas issued by Charles for
or involving Robert in any way.

Indeed, far from their being any

Recueil Charles III, ed. Lot/Lauer, no. 10, pp. 15–17 (8 February 898), at Saint-Denis: ‘quia
accedentes ad nostram presentiam venerabiles fratres peculiaris patroni nostri domni Dionysii
sociorumque eius, deprecati sunt nostram clementiam ut . . .’ This act was entirely written
out by a monk of Saint-Denis itself, as is clear from the ruche, a classic Saint-Denis type not
at all like that of Heriveus (in whose name it was subscribed). Diplomata Karolinorum, ed.
Ferdinand Lot and Philippe Lauer, 9 vols (Paris, 1936–49), VI, plate 10; Peter Worm, Karo-
lingische Rekognitionszeichen: Die Kanzlerzeile und ihre graphische Ausgestaltung auf den
Herrscherurkunden des achten und neunten Jahrhunderts, elementa diplomatica 10:1–2, gen. ed.
Peter Rück, 2 vols (Marburg, 2004), I, p. 134; cf. Bautier in Recueil Eudes, pp. lii–liv.

The argument is narrated only by the Annales Vedastini, s.a. 900, p. 82. It is mentioned after
the assassination of Fulk of Reims by Winemar on behalf of Baldwin of Flanders (16/17 June
900) and the ordination of Fulk’s successor Heriveus and Heriveus’s excommunication of the
assassins (6 July). Assuming that the annalist was keeping roughly to chronological order (but
appending Heriveus’s later consecration and anathema to the assassination as a single block
of events), then the argument with Manasses occurred around the time of Fulk’s assassination
in mid-June, immediately before the sudden profusion of diplomas given for Richard, the
ﬁrst datable acts of the series being issued on June 26 and 30 (nos. 32–33). Note also that
though its day is not given, Recueil Charles III, no. 31 may well have been the very ﬁrst act
of the series. It was issued at Compiègne, after Fulk’s assassination (since his successor Askeric
already appears as archchancellor). Compiègne happens to have been the location of the court
where Fulk and Baldwin had their ﬁnal falling out, and Fulk was assassinated as he was
returning from there (Regino, Chronicon, s.a. 903 (for 900), consistent with the Annales
Vedastini ’s implication that the court had been on the Oise, s.a. 899 (for 900)). All this would
mean that the argument between Robert and Manasses occurred at the very same court that
witnessed the break between Fulk and Baldwin, and that the diplomas marking Richard’s
supremacy were issued immediately after.

I assume that the ‘Rotbertus ﬁdelis noster’ who petitioned no. 13 (pp. 21–3, issued on 23 June
898 at Vienne-la-Ville) is not Robert of Neustria. Its beneﬁciary was a Septimanian ﬁdelis
named Theodosius, a typically southern name impossible to associate with Robert of Neustria
or his circle, and it concerns properties in Roussillon and Besalù that seem to descend from

Early Medieval Europe

  ()



380

Geoffrey Koziol

diplomas for Robert, Charles’s diploma of 8 February 898 for Saint-
Denis was a threat and an insult to him – and from Robert’s point of
view, must have raised memories of Charles’s attempt to give Fulk his
abbacy of Saint-Martin in 893–4 and his attempt to assert his patronage
of Saint-Denis in January 896.

H. During the same opening two-and-a-half years of Charles’s sole
reign, Charles issued no diplomas for or involving Robert; but neither
did he issue any for or involving the two other great magnates of the
kingdom, Richard the Justiciar, duke of Burgundy, and William the
Pious, duke of Aquitaine. Indeed, the only individual clearly favoured
early in Charles’s reign was his mother Adelaide, central to seven of the
twenty-one diplomas Charles issued from his accession to 17 June 900.

I. The break in 900 between Robert and Charles, occasioned by dis-
paraging remarks of Manasses, principal ﬁdelis of Richard of Burgundy,
can be more speciﬁcally dated to June of that year – probably in the
middle of the month

– because in late June, over the course of a few

days, having issued no diplomas at all for the greatest magnates of his
kingdom over the preceding two-and-a-half years, Charles suddenly
issued three diplomas for churches within Richard’s territory, at Rich-
ard’s request, in which Richard is described as ‘dilecti ac carissimi nostri
Richardi, venerabilis comitis’, the kind of honoriﬁc that throughout his
reign marked the men and women Charles favoured. Judging from
Charles’s diplomas, from this point on for at least the next two years,
it was Richard who dominated the court: he and his clients are central

the old Septimanian regime of aprisio. The diploma was also part of a dense series of acts
issued in 898 and 899 exclusively for southern recipients that underscored the adhesion of
southern loyalists to Charles. There is therefore little that would support this Robert being
Robert of Neustria.

Recueil Charles III, ed. Lot/Lauer, nos. 10 (898), 11 (898), 14 (898), 15 (898), 22 (899), 23 (899),
27 (899). It should be noted that the intervention of Adelaide represented more than a merely
formal mention of the queen mother; quite apart from the fact that any mother’s direct
intercessory intervention is extremely rare in West Frankish Carolingian diplomas (that of
wives is more common), Adelaide and Fulk of Reims were the young Charles’s mainstays in
the darkest days of his struggle with Odo, Adelaide, for example, accompanying her son
during his ﬂight to Lotharingia in 896 (ibid., nos. 5 and 7; Parisot, Royaume de Lorraine,
p. 531). The relationship between Charles and Adelaide was therefore unusually close. I would
go so far as to argue that it was Adelaide, even more than Fulk, who held out to her son the
injustice of his position and pushed him to reclaim his birthright. On the place of wives and
mothers as intercessors and petitioners in Carolingian diplomas, see Franz-Reiner Erkens,
‘“Sicut Esther Regina”: Die westfränkische Königin als consors regni’, Francia 20 (1993),
pp. 15–38, especially pp. 20–1 on the unusual number of Adelaide’s interventions; also
Knut Görich, ‘Mathilde-Edgith-Adelheid: Ottonische Königinnen als Fürsprecherinnen’,
in Bernd Schneidmüller and Stefan Weinfurter (eds), Ottonische Neuanfänge: Symposion zur
Ausstellung ‘Otto der Grosse, Magdeburg und Europe’ (Mainz, 2001), pp. 251–91.

Above, n. 64.



Early Medieval Europe

  ()

Charles the Simple, Robert of Neustria

381

to seven of the thirteen diplomas Charles issued during this period.

was also Richard, not Robert, who attended the translation of Remigius
at Saint-Remi after Christmas 900.

Thus, Robert’s eclipse after the

previous June corresponded to Richard’s ascension at court.

J. The last diploma in this spate for Richard and his cohort was issued
on 25 July 902. There is then a gap of nine months until Charles’s next
surviving diploma for any recipient of any sort, and it was issued for
Robert, on 25 April 903. We do not know what had happened during
those nine months; we can only be sure that for whatever reasons,
Richard was losing favour and Robert gaining at his expense. In any
event, from 25 April 903 until 907, Robert’s dominance of Charles’s
court was complete: three of all three extant diplomas issued in the
seven remaining months of 903 are issued with him and for him or his
clients. In 904, 905 and 906, he is central to four of Charles’s seven
extant diplomas.

Conversely, during the same period Richard was

completely eclipsed at court, being named in not a single diploma.
Richard returns in two diplomas of 907 and 908, but Robert remains
prominent.

Moreover, in the important years 910–11, immediately

before and during the ﬁrst stages of Charles’s settlement with the Nor-
mans and his takeover of Lotharingia, Robert remains the major ﬁgure
in diplomas, while Richard is, again, completely absent.

K. Robert’s diploma for Saint-Denis of 923 speaks of needing to
return from battle with Charles ‘with the triumph of victory, bearing
the unconquered battle standards from their subjection’ (‘cum trium-
pho victorie invicta . . . , exinde de eorum subjectione vexilla referre’).

Recueil Charles III, ed. Lot/Lauer, no. 31 (after 17 June 900) for Saint-Etienne of Auxerre;
no. 32 (26 June 900) for Saint-Martin of Autun; no. 33 (30 June 900) for Saint-Nazaire of Autun;
no. 38 (22 April 901) for Saint-Germain of Auxerre; no. 43 (902) for Richard’s client and ally
Manasses (here speciﬁcally called Charles’s ﬁdelis). Also demonstrating the strength of the
Burgundian circle at court in this period are no. 37 (901) (supposedly for Adalgarius of Autun;
since Adalgarius had died in 893/4, the act was probably reworked and attributed to Adalgar-
ius as part of his rehabilitation when Richard and Robert established their alliance; above,
n. 25); and no. 42 (902), for the cathedral of Auxerre and Bishop Herifred. Manasses is usually
thought to have been the husband of the daughter of Richard’s brother Boso; Bouchard has
recently argued that he was in fact the son of Richard and Boso’s sister: ‘Those of My Blood’,
Appendix B, pp. 192–3.

Flodoard, Historia Remensis ecclesiae, IV.12, p. 405. This was an important ceremony, in some
sense a continuation of the puriﬁcations that followed the assassination of Fulk and marked
the inauguration of Heriveus’s reform programme.

Recueil Charles III, ed. Lot/Lauer, no. 45 (25 April 903) for Saint-Germain-des-Prés; no. 46
(30 April 903) for Saint-Martin of Tours; no. 47 (5 June 903) for Saint-Denis; no. 48 (904)
for the bishop of Châlons; no. 49 (904) for Saint-Martin of Tours; no. 50 (905) for Saint-
Denis; no. 54 (906) for Saint-Amand.

Richard: Recueil Charles III, ed. Lot/Lauer, nos. 55 (907), 59 (908). Robert: ibid., no. 57 (907).

Recueil Charles III, ed. Lot/Lauer, no. 63 (910–11), no. 65 (911–15, probably earlier rather than
later within this range), no. 66 (911).

Early Medieval Europe

  ()



382

Geoffrey Koziol

The phase is completely mystifying, unless one takes it at face value and
understands that the abbey’s vexilla, the battle-standards of its three
saints, were actually in Charles’s control.

L. This understanding requires renewed consideration of an under-
studied problem involving the relics of Saint-Denis’ patron saint. The
discussion will turn out to be something of a red herring, yet it is still
important, since the details encountered along the way remain highly
relevant to the overall story. As the evidence is again fragmentary and
recalcitrant, it is best simply to present the relevant issues.

1. In the middle of the eleventh century, the monks of Saint
Emmeram at Regensberg made the startling announcement that
they possessed the body of St Dionysius. According to their story,
in the time of King Odo and Emperor Arnulf, when Ebalus was
abbot of Saint-Denis, one Gislebert had, at Arnulf’s request, stolen
Dionysius from the monastery, and Arnulf had later given the relics
to Saint Emmeram, his burial place. Shot through with inconsist-
encies and improbabilities, the claim and the documentation sup-
porting it are regarded as utterly spurious.

This judgement, however,

may be too severe. Though certainly not in possession of the entire
body, Saint Emmeram might still have had partial relics of Dionysius,
received not by theft but through a series of exchanges between
Ebalus, Odo and Arnulf.

2. At the time of his ﬁrst coronation (29 February 888), Odo, with-
out access to Reims (controlled by Archbishop Fulk, who refused

For relations between Arnulf and Saint Emmeram: Karl Babl, Emmeram von Regensburg:
Legende und Kult, Thurn und Taxis-Studien 8 (Kallmünz, 1973), pp. 143–4. On the claims of
Saint Emmeram: Franz Fuchs, ‘Die Regensburger Dionysiussteine vom Jahre 1049’, in Renate
Neumüllers-Klauser (ed.), Vom Quellenwert der Inschriften: Vorträge und Berichte der Fachta-
gung Esslingen 1990, Supplemente zu den Sitzungsberichten der Heidelberger Akademie der
Wissenschaften, Philosophisch-historische Klasse 7 (Heidelberg, 1992), pp. 139–59; Roman
Hankeln, Historiae Sancti Dionysii Areopagitae. St Emmeram, Regensburg, ca. 1050/16. Jh.,
Musicological Studies 65–3 (Ottawa, 1998), pp. ix, xviii–xxiv; Andreas Kraus, Die Translatio
S. Dionysii Areopagitae von St. Emmeram in Regensburg, Bayerische Akademie der Wissen-
schaften, Philosophisch-Historische Klasse, Sitzungsberichte 4 (Munich, 1972); idem, ‘Saint-
Denis und Regensburg: Zu den Motiven und zur Wirkung hochmittelalterlicher Fälschungen’,
in Fälschungen im Mittelalter, Internationaler Kongreß der Monumenta Germaniae Historica,
München, 16.

−19. September 1986, 6 vols, MGH Schriften 33.1–6 (Hanover, 1988–90), III,

pp. 535–49; Wilhelm Störmer, ‘Beobachtungen zu Aussagen und Intentionen der bayerischen
Stammes-‘Sage’ des 11./12. Jahrhunderts’, in ibid., I, pp. 451–70, at pp. 467–9. Philippe Lauer
suggested that the relics Charles gave to Henry came from Saint Emmeram, relying on an
authentic from the Lateran altar of the Sancta Sanctorum, not recognizing its ﬁctitious
origins, and not knowing of Salonnes’ possession of relics of Dionysius: Robert I

, p. 20; idem,

Le trésor du Sancta Sanctorum, Fondation Eugène Piot: Monuments et mémoires de l’Académie
des Inscriptions et Belles-Lettres 15 (Paris, 1906), pp. 129–31.



Early Medieval Europe

  ()

Charles the Simple, Robert of Neustria

383

to recognize him) and therefore lacking regalia, took regalia and
treasures from Saint-Denis for the ceremony at Compiègne. The
transaction was doubtless facilitated (if not forced on the monks)
by Odo’s close ally (and soon-to-be archchancellor) Ebalus, abbot
of Saint-Denis and Saint-Germain-des-Prés. If nothing else, the
transaction shows the ﬁrst Robertian king’s keen sense of depend-
ence on Saint-Denis even before he had assumed the abbacy.

3. In their quest for legitimation and support, Odo, Charles and
Robert were constantly making gifts to Arnulf and Henry, rulers
of the eastern kingdom. Among the most important were those
Odo gave Arnulf when he visited Worms in 895, at a critical junc-
ture in his reign, when Arnulf had just recognized Charles’s com-
peting coronation as king.

We do not know what gifts Odo gave.

However, in a chain of unprovable but convincing associations, it
is often argued that they included Charles the Bald’s Codex Aureus,
which is likely to have come from Saint-Denis and which Arnulf
did, in fact, give to Saint Emmeram.

Though not mentioned by

any sources, Odo would also have given gifts in August 888, when
he visited Arnulf in order to receive his recognition prior to a second
coronation at Reims in November with a crown given by Arnulf
himself. The Codex might equally have been given then.

4. If there were any core of fact to Saint Emmeram’s later story
of a theft of relics of Dionysius, one would expect it to be in these
early transactions involving Saint-Denis, Ebalus, Odo, Arnulf and
Saint Emmeram, since three of the four principals in the story
really did interact intensively as contemporaries at that time
(Odo, Arnulf, and also Ebalus, under whose abbacy the theft is

Werner, ‘Westfranken-Frankreich’, p. 744 n. 1; Schneider, ‘Odo’, pp. 14–5; Guillot, ‘Les
étapes de l’accession d’Eudes’, p. 210.

Annales Fuldenses, s.a. 895, ed. G.H. Pertz and F. Kurze, MGH SRG 5 (Hanover, 1891),
p. 126: ‘Ibi Odo rex Galliae ad ﬁdelitatem regis cum muneribus veniens ab eo honoriﬁce
susceptus est.’ Regino, Chronicon, s.a. 895, p. 143: ‘In eodem placito Odo rex cum magnis
muneribus ad Arnufum venit . . .’

Paul Gichtel, Der Codex Aureus von Sankt Emmeram: Die Restaurierung des Cod. lat. 14000
der Bayerischen Staatsbibliothek München (Munich, 1971); Le trésor de Saint-Denis (Paris, 1991),
pp. 51, 54 n. 68. The inventory drawn up by Saint-Denis listing items given to Odo in early
888 does, in fact, describe a gospel book in terms that ﬁt the Codex Aureus: P.E. Schramm
and Florentine Mütherich, Die deutschen Kaiser und Könige in Bildern ihrer Zeit: 751–1190
(Munich, 1983), p. 95: ‘Evangelium auro et gemmis optime paratum, et intus auro scriptum I.’

Annales Vedastini, s.a. 888, p. 88: ‘Statuto itaque die Odo rex fretus auxilio suorum Worma-
ciam venit, honoriﬁce ab Arnulfo rege susceptus, et facti amici, remisit eum cum honore in
regnum suum . . .’ Annales Fuldenses, s.a. 888, p. 116: ‘veniensque humiliter ad regem et
gratanter ibi recipitur’. On the importance of this second coronation and its relation to the
ﬁrst, see Guillot, ‘Les étapes de l’accession d’Eudes’.

Early Medieval Europe

  ()



384

Geoffrey Koziol

said to have occurred; Gislebert is unknown). Among the various
possibilities:

a) Odo may have received relics of Dionysius from Saint-
Denis through Ebalus at the time of his ﬁrst coronation, along
with regalia and other treasure, to be used for swearing his
oath. He might later have given such relics to Arnulf, either
in August 888 or in 895.
b) Odo may have received relics from Ebalus speciﬁcally to
give to Arnulf in August 888.
c) As abbot of Saint-Denis after Ebalus’s death in 892, Odo
himself could have taken relics from the monastery to give to
Arnulf in 895.

The ﬁrst possiblity (a) is unlikely. It happens that when the monks
of Saint-Denis loaned Odo regalia and treasure for the coronation,
they drew up an inventory of the items he received. The list is
quite detailed. It includes a gospel book that sounds much like the
Codex Aureus, but nothing that can be construed as relics. Yet it is
hard to draw a ﬁrm conclusion from this. On the one hand, there
is an inventory, and it mentions no relics. On the other hand, the
likely presence of the Codex Aureus in the inventory does support
the possibility of a sequence of transmissions of luxury items from
Saint-Denis under Ebalus to Odo, from Odo to Arnulf, and from
Arnulf to Saint Emmeram; while if Saint-Denis had given Odo
relics of Dionysius, it might not have listed them in the inventory,
either because Ebalus’s transaction was entirely separate or simply
to avoid the kind of problem the monks of Saint Emmeram later
created.

The third scenario (c) is possible, save that Saint Emmeram’s

story has Ebalus playing a major role in the transfer, and Ebalus
had died in 892. The middle scenario (b) is likeliest of all. It alone
involves all the identiﬁable principals of Saint Emmeram’s later
account. And relics of the ‘special patron’ of the Frankish kings
would have been a highly appropriate gift for Odo to give Arnulf
in return for the crown Arnulf gave him. Of course, Saint
Emmeram may have invented the entire story, as most historians
have believed; but instinctive scepticism towards the furtum sacrum
of such an important saint may also have led them to unwarranted
rejection of the claim that the monastery had any relics of Diony-
sius at all. In any case, one of the sceptics’ weakest points is that they
can provide no particularly convincing reason for Saint Emmeram’s

Schramm and Mütherich, Deutschen Kaiser und Könige, p. 95.



Early Medieval Europe

  ()

Charles the Simple, Robert of Neustria

385

claims to have Dionysius’ relics in particular, transmitted by these
particular leaders.

It is therefore barely possible that the monks

really did have at least some of his relics.

Even so, all of the above possibilities assume that Saint

Emmeram received the relics from Arnulf (d. 899). None explains
how Charles got control of either the standards of Saint-Denis or
its relics. And the fact is, in 923, Charles had both.

5. In an account that has never been questioned and probably
should not be in its essentials, Widukind of Corvey writes that
Charles the Simple sent a legate to King Henry I of Germany. On
Charles’s behalf, the legate gave Henry the hand of Dionysius,
sheathed in gem-studded gold, ‘as a sign of faith and truth’ and ‘a
pledge of perpetual union and mutual love’.

Though Widukind

does not date the gift, small details suggest that it was made in 923
between February and June – in other words, after Robert’s
diploma for Saint-Denis but before the battle of Soissons. To be
precise, Widukind associates the gift with an amicitia between Henry
and Charles. This might refer to the Treaty of Bonn between the
two rulers (7 November 921). However, Widukind makes the gift
entirely one-sided, whereas the Bonn treaty, even more than most
such amicitiae, was decidedly bilateral, reciprocal and equal. Still,
one might expect Widukind to diminish signs of bilaterality for the
sake of exalting Henry above Charles, so another consideration is
more important in dating the gift. In Widukind’s account, Charles
states that he is making the gift at a time when he is beset on all
sides by enemies – a description that does not easily ﬁt 921 but
obviously ﬁts 922 and 923. Signiﬁcantly, in 923 (that is, right
around the time of Robert’s diploma for Saint-Denis), Henry met
Robert and swore friendship with him – thereby abrogating the
terms of the treaty he had made with Charles at Bonn. Since
Widukind emphasizes that Charles gave the relics of Dionysius as
a pledge and sign of union, love, faith and truth, the most likely
explanation (following Karl Schmid) is that Charles made the gift

The reasons alleged are that the entire affair was a great literary in-joke (Krause, ‘Saint-Denis
und Regensburg’), that it reﬂects heightened Bavarian self-consciousness in the face of Henry
III’s attacks on the duke of Bavaria (Störmer, ‘Beobachtungen’), that Saint Emmeram
suddenly wanted to imitate Saint-Denis’ growing status as a regnal centre (Kraus, Translatio
S. Dionysii ), that monks of Saint Emmeram brought back knowledge of the Pseudo-
Areopagitean writings from Chartres (Hankeln, Historiae Sancti Dionysii Areopagitae, citing
Bischoff). Such reasons are variously implausible, anachronistic, or inadequate.

Res gestae Saxonicae, I.33, ed. and trans. Ekkehart Rotter and Bernd Schneidmüller (Stuttgart,
1981), p. 76: ‘“Et hoc tibi signum ﬁdei et veritatis transmisit”; protulitque de sinu manum
preciosi martyris Dionisii auro gemmisque inclusam. “Hoc,” inquit, “habeto pignus foederis
perpetui et amori vicarii.”’

Early Medieval Europe

  ()



386

Geoffrey Koziol

to recall Henry to the terms of the treaty which Henry had sworn
on relics and then violated – viz., after Robert’s treaty with Henry
(January or early February 923) and before the battle of Soissons
(June 923).

6. As a priory of Saint-Denis, Salonnes would have had relics of
Dionysius. Charles’s diploma of 896 for Salonnes states as much.

M. Conceivably, Saint Emmeram actually did have relics of Diony-
sius, which ultimately came not from a theft but from a gift from
Ebalus to Odo and Odo to Arnulf in 888. Still, even if such a series of
transactions did occur, it cannot explain how Charles ended up with
both relics of Saint-Denis and its vexilla. For Charles almost surely gave
relics of Dionysius (probably part of a ﬁnger encased in a gold-gilded
arm) to Henry I in 923, between February and June – meaning that
Charles had such relics in his possession before then. These relics might
have come from Salonnes. Yet while Salonnes, as a priory of Saint-Denis,
could well have had a ﬁnger of Dionysius, it is hard to believe that it
had a standard. And even if, for the sake of argument, Salonnes did have
a standard of the saint, then this standard would not have been Saint-
Denis’ and therefore would not have been spoken of as being in ‘sub-
jection’. However, Charles had been at Saint-Denis itself in February 898,
in circumstances that suggest an attempt to claim the saint’s ‘special’
patronage of the Frankish kings for himself in opposition to Robert.

Everything being taken together, it was probably at this time that he
had taken both the monastery’s battle-standard and some of its relics.

From all this we can make the following deductions:

I. The break between Charles and Robert in 900 was a major event
that shook and shaped the ﬁrst years of Charles’s reign. This is shown
not only by the fact that the Annals of Saint-Vaast mention it so prom-
inently, but even more by the fact that it is the very last event men-
tioned by the annalist: with the quarrel between Robert and Manasses
and Robert’s withdrawal from court, the annals abruptly end, as if their
author is giving up on a hopeless situation. The quarrel meant Robert’s

Karl Schmid, ‘Unerforschte Quellen aus quellenarmer Zeit: Zur amicitia zwischen Heinrich I.
und dem westfränkischen König Robert im Jahre 923’, Francia 12 (1984), pp. 119–47, at
pp. 141–2; followed by Felten, ‘Robert I.’, pp. 36–8.

Recueil Charles III, ed. Lot/Lauer, no. 7, p. 10: ‘. . . fratribus monasterii Salonae, ob vene-
rationem et amorem sanctissimorum martyrum inibi quiescentium, almi videlicet Privati,
Frodoldi atque Iddonis ac beatissimi Dyonisii martyris Domini et patroni nostri’.

Recueil Charles III, ed. Lot/Lauer, no. 10. On Dionysius as the king’s peculiaris patronus, see
Brown, ‘Politics and Patronage’, pp. 1, 283–5, 314, 326; Koziol, ‘Is Robert I in Hell?’.



Early Medieval Europe

  ()

Charles the Simple, Robert of Neustria

387

complete withdrawal from court, coupled with Richard’s complete takeover
of precedence for himself and his clients. It was such a major event, the
rivalry between Robert and Richard so agonistic, that when Robert
returned to court in April 903, the effect of his return was the complete
removal of Richard and his clients from court for a number of years.

II. Before the quarrel that removed Robert and advanced Richard,
none of the territorial princes had any particularly strong position at
court. This fact, coupled with the suddenness and completeness of
Robert’s withdrawal and Richard’s prominence, is, as Yves Sassier has
remarked, indicative of ‘une sourde lutte d’inﬂuence’ around Charles
early in his reign, but it is ‘muted’ only to us historians because of the
lack of diplomas. But the absence of diplomas is itself the proof that no
one (besides Adelaide and Fulk) had managed to assert their position
around Charles, their return both the proof of Richard’s success and the
ceremony by which the regularization of precedence at court was pub-
licized and manifested. In any case, contemporaries like the annalist
of Saint-Vaast and the recipients (or non-recipients) of diplomas at
Auxerre, Autun, Tours and Saint-Denis knew exactly what the stakes
were, and the stakes were tectonic in their effects.

III. These conclusions give new perspective to several aspects of
Charles’s reign. First, they make more sense of Richard’s position. With
Robert’s success in 903, and even after his own limited return to court
in 907, he does not appear to have been a great supporter of Charles,
but he was no great enemy either. He seems to have kept himself on
the sidelines, having presumably removed himself from political
intrigues that had isolated him and left him without options: he did
not have the standing to oppose Robert openly but could not oppose
Charles without further weakening his position and incidentally
strengthening Robert’s even more. So he watched, and waited to see
what would happen. His opportunity came in 911 when the Normans
attacked Robert’s territory in the Chartrain and Dunois. Immediately
he came west to ﬁght alongside Robert; together they broke up the siege
of Chartres, inﬂicting such a defeat on the Normans that they were
forced to accept a settlement from Charles. It is tempting to believe that
in coming to Robert’s aid, Richard was intentionally undoing his failure
to have aided Odo during the Viking siege of Paris twenty-ﬁve years
earlier; certainly his success at Chartres was trumpeted as loudly as his
failure at Paris had been. In any case, Richard’s solution to his margin-
alization was to slowly, cautiously mend fences with Robert. In 918 his
son Raoul took the last step and completed the alliance. The decision
may speak of Raoul’s pragmatism and Robert’s power (the latter so

Early Medieval Europe

  ()



388

Geoffrey Koziol

entrenched that the former found the better part of valour in embracing
it). With equal plausibility, the alliance may also indicate that Charles’s
actions, in particular his favour of Hagano, perhaps also his Norman
policies, had become so distasteful to such a broad spectrum of the
aristocracy that he had alienated even the son of Robert’s enemy. In any
case, the effects of the alliance were momentous, because it removed a
persistent rivalry that had, from the beginning of Charles’s reign,
helped him stay in power by dividing the interests of the two major
contenders for power.

IV. Most of Robert’s lay abbacies are not attested until after 903 –
that is, after his reconciliation with Charles. The reasons are varied. In
the case of a small and relatively unimportant house like Morienval, it
may be that Robert had long been abbot, but there had been no reason
for him to seek a diploma for the monks or issue a charter as abbot. In
the case of Saint-Aignan of Orléans – a monastery located deep in the
heart of Robert’s home territories without great symbolic importance
either in the mythology of the Franks or in the clash of princes – it is
unlikely that there was any advantage for either Charles or Robert in
disputing or proclaiming the abbacy, and unlikely that Robert had ever
lost it. It may be that we have simply lost diplomas, though more likely
none were ever issued. Neither of these conditions, however, holds true
for four of Robert’s abbacies, including the two most important and
symbolically freighted ones: Saint-Amand, Saint-Germain-des-Prés,
Saint-Martin of Tours, and above all, Saint-Denis. At Saint-Germain
the sequence and language of Charles’s diplomas suggest that he had
not recognized Robert’s abbacy after Odo’s death. At Saint-Amand, he
not only did not recognize it, he may have gone so far as to appoint
Fulk of Reims rector against Robert’s rights, as he had certainly tried
to do at Saint-Martin in 893. Conversely, the ﬁrst sign that Robert had
taken the leadership at Charles’s court is a diploma of 25 April 903 in
which Robert intercedes for Saint-Germain and is described (though as
count not abbot) with the same kind of honoriﬁcs only recently applied
to Richard: ‘comes nobis admodum dilectus’.

The next is a diploma

issued just a few days later, on 30 April, which – not coincidentally –
afﬁrms Robert’s position as abbot of Saint-Martin. This time the hon-
oriﬁcs are even more extravagant: he is ‘abbas reverendus et admodum
amabilis Rotbertus abbas’. Still more signiﬁcant, these honoriﬁcs are
coupled to a statement that he is the brother of ‘the Lord King Odo
our predecessor’. In other words, through this diploma, Charles not
only publicly recognized Robert’s legitimacy as abbot but also gave Robert

Recueil Charles III, ed. Lot/Lauer, no. 45.



Early Medieval Europe

  ()

Charles the Simple, Robert of Neustria

389

a personal recognition of his brother’s legitimacy as king.

Finally, and

most importantly, Robert is attested as abbot of Saint-Denis only in the
third of Charles’s acts following Robert’s reappearance at court, on
5 June 903.

But why were these three acts spread out the way they were rather

than being issued at once? The liturgical year provides the answer. The
ﬁrst two diplomas were given during the week of the octave of Easter.
The last act, the culmination of the series, the one that ﬁnally gave
Robert what he wanted and solemnized his formal reconciliation with
Charles, was saved for Pentecost Sunday.

One of the most solemn

feast days of the church, it was also frequently – as here – a day of high
political liturgy; and no feast day better symbolized unity.

Quite apart

from underscoring the overriding importance of this single act, the
timing demonstrates just how little diplomas had to do with the utilit-
arian recording of legal rights and privileges, how much their issuance
had become integrated into the ceremonial of changing political rela-
tions. In any case, before that Pentecost of 903 there is absolutely no
mention of Robert’s being abbot of Saint-Denis; to the contrary, there
is evidence that Charles had tried to keep the abbacy vacant after Odo’s
death. He may not have been able to take the step of naming himself
lay abbot; after six years of Odo’s domination (and another ﬁfteen by
Odo’s patrons and allies), he may not have had the support within the
community or within the Parisis (a Robertian bastion) to risk such an
open rejection of Robert; but he was trying to keep Robert away from
the abbacy and trying to gain the patronage of the abbey and its three
saints for himself.

We can only guess at the pressure Robert brought

to bear on Charles to return to court and expel Richard, but the price
– and the central issue in the negotiations – was surely Charles’s recog-
nition of his position as abbot of Saint-Denis.

But not before Charles had absconded with some relics and the

standards. Clearly, Charles had the standards in his possession in Janu-
ary 923, when Robert spoke of wanting to ‘return them from their
subjection’ in his anticipated battle with Charles. Charles also had relics
of Dionysius, since he gave some to Henry I at about the same time.
He is unlikely to have received the standards from Salonnes, since it is

Recueil Charles III, ed. Lot/Lauer, no. 46.

Recueil Charles III, ed. Lot/Lauer, no. 47.

Arthur Giry, Manuel de diplomatique (Paris, 1894), pp. 153–4, 192, 248–9.

The most important court of Charles’s grandson Lothar was also reserved for Pentecost, and
was also the occasion for the issuance of the most important diploma of the reign: Koziol, ‘A
Father, His Son, Memory, and Hope’.

Odo’s patron was Gauzlin, abbot of Saint-Denis from 877–86, who was succeeded by his
nephew Ebalus, on whose death in 892 Odo himself took over the abbacy: Brown, ‘Politics
and Patronage’, pp. 164–5; Werner, ‘Gauzlin von Saint Denis’.

Early Medieval Europe

  ()



390

Geoffrey Koziol

unlikely Salonnes ever had them. Neither standards nor relics are likely
to have been part of any treasure Saint-Denis loaned Odo, for whatever
parts of that treasure Odo gave out he gave to Arnulf, while Arnulf
appears to have given them to Saint Emmeram, not Charles. The only
time Charles was in a position to take possession of the standards and
the relics was before he had recognized Robert’s abbacy, and the most
likely occasion was the one time he is recorded as having been at Saint-
Denis, in February 898, after his coronation and before Robert’s sub-
mission to him.

V. The ﬁrst ﬁve years of Charles’s reign were far more tumultuous
than historians have suspected. In many ways, they were an uninter-
rupted continuation of the hostilities that had opposed Charles to
Robert and Odo since 893. To be sure, Odo had wanted there to be
a peaceful succession after his death and to that end had asked his
followers to ‘keep faith’ with Charles; but it did not happen. Robert did
not immediately recognize Charles. Charles did not give Robert pri-
macy in his counsels, did not formally and explicitly recognize the lay
abbacies that were the most important constituents of his margravate,
probably turned the Vikings loose on him, and worst of all, immedi-
ately tried to take over Saint-Denis, in all these ways violating the spirit
if not the letter of his agreement with Odo. It therefore took ﬁve entire
years before the terms of the succession settlement were ﬁnally imple-
mented as Odo had intended – ﬁve years in which Robert saw Charles
resist his patrimonial rights, saw himself suffer public insult at Charles’s
court, saw his rival gain the complete supremacy that should have been
his as the brother of a king. However we interpret the events between
903 and 920 – events largely hidden from us because of the document-
ary silence – this much is clear: in January 923, in the diploma he
issued for Saint-Denis, when Robert stated his cause in a voice as close
to his own as we can reasonably expect, he returned to the grievances
with which Charles’s reign had begun twenty-ﬁve years before. He went
back to his identity as abbot of Saint-Denis and to Charles’s theft of its
vexilla. If the battle of Soissons was a judgement of God on the right-
ness of Robert’s rebellion, then God’s judgement would be declared by
which of the two kings held those battle standards at its end.

University of California, Berkeley

Wyszukiwarka

Podobne podstrony:
K J Charles The Secret Casebook of Simon Feximal Butterflies
The simple joy of living
Parametric Analysis of the Simplest Model of the Theory of Thermal Explosion the Zel dovich Semenov
Dynamic gadolinium enhanced subtraction MR imaging – a simple technique for the early diagnosis of L
Angielski - Gramatyka - opracowania, Simple Past of the verb to be, Simple Past of the verb " t
10 Past Simple or Present Perfect Provide the correct form of the verb
17 Present Perfect or Past Simple Provide the correct form of the verb
Heinlein, Robert A The Good News of High Frontier
Howard, Robert E The Fearsome Touch of Death
Heinlein, Robert A The Green Hills of Earth (SS Coll)
Howard, Robert E Breckenridge Elkins The Conquerin Hero of the Humbolts
Robert Stevenson The Strange Case of Dr Jekyll and Mr Hyde
The Face of the Waters Robert Silverberg
Heinlein, Robert A The Last Days of the United States
Robert P Smith, Peter Zheutlin Riches Among the Ruins, Adventures in the Dark Corners of the Global
The Last Call of Mourning Charles L Grant
Tigers of the Sea Robert E Howard
The Untold Story of State Medicine Robert Henry Williams
Gates of the Universe Robert Coulson

więcej podobnych podstron