The secret of my succession: dynasty

and crisis in Vandal North Africa

emed_293

135..159

A.H. M

errills

The system of royal succession in the Vandal kingdom of North Africa has long
been regarded as idiosyncratic within the early medieval west, but its fullest
implications have rarely been investigated closely. The present article examines
the origins of succession by agnatic seniority under the strong rule of King
Geiseric, and argues that it was one of several innovations intended to
establish the emergent Hasding royal house against other aristocratic chal-
lenges. The article goes on to explore the consequences of this law in the two
major dynastic crises of the Vandal kingdom: under Huneric in c.

481 and

under Hilderic in

530. In both cases, the standard narratives of events are

challenged, and with them assumptions about the ‘constitutional’ status of
Geiseric’s law of succession.

Few individuals dominate the history of the fifth century like the Vandal
king Geiseric. Under this physically unprepossessing ruler, the Vandals
underwent an unlikely transformation from being a small and unimpor-
tant group of mercenaries in the southern plains of Spain to being the
masters of Carthage and the Mediterranean shipping lanes.

Geiseric

oversaw the settlement of the Vandals and their allies in the rich lands of
Africa Proconsularis and their emergence as a major player on the inter-
national scene. His diplomatic brinksmanship infuriated the Visigoths,

I am grateful to Neil Christie, Roland Steinacher, Jen Baird, Christina Pössel, Danuta Shanzer,
Richard Miles and the referees of this journal for suggestions and comments. Mistakes which
remain are, of course, my own.

C. Courtois, Les Vandales et L’Afrique (Paris,

1955) remains seminal on the group. Important

recent studies include those of M.E. Gil Egea, África en Tiempos de los Vándalos: Continuidad
y Mutaciones de las Estructuras Socio-Políticas Romanas, Memorias del Seminario de Historia
Antigua

7 (Alacala, 1998), and now G.M. Berndt, Konflikt und Anpassung. Studien zu Migration

und Ethnogenese der Vandalen, Historische Studien

489 (Husum, 2007). For further bibliogra-

phy and discussion see P. von Rummel, ‘Zum Stand der afrikanischen Vandalenforschung nach
den Kolloquien in Tunis und Paris’, Antiquité Tardive

11 (2003), pp. 13–19; and A.H. Merrills,

‘Vandals, Romans and Berbers: Understanding Late Antique North Africa’, in idem (ed.),
Vandals, Romans and Berbers: New Perspectives on Late Antique North Africa (Aldershot,

2004),

pp.

3–28. Geiseric’s own physical appearance is described by Jordanes, Getica 168, ed. T.

Mommsen, MGH AA

5.1 (Munich, 1882), p. 102.

Early Medieval Europe

2010 18 (2) 135–159

2010 Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350

Main Street, Malden, MA

02148, USA

inflamed the Huns, led to the cataclysmic sack of Rome in

455, and

ultimately provoked a succession of increasingly ambitious imperial cam-
paigns against his kingdom.

But it also led to the betrothal and eventual

marriage of his son Huneric to the Theodosian princess Eudocia, and to
the strong Vandal patronage of the late imperial pretender Olybrius.
Naturally, this vertiginous political rise resulted in massive social and
political changes among the Vandals themselves, but by a combination of
pragmatic land distribution, military reorganization and the brutal sup-
pression of revolt, Geiseric presided over a relatively peaceful kingdom
down to his death in

477.

Yet a kingdom founded upon the strong personality of a king was not a

stable state, as Geiseric well recognized. After his death, the Hasding line
was hardly marked by a distinguished lineage of kings, but the founder of
the African state did his best to smooth the path for his successors. Among
the best known of Geiseric’s political innovations was his establishment of
a clear law of succession for the rulers who came after him.

In place of the

existing system of inheritance (the workings of which are unknown, but
are generally assumed to have been based on primogeniture), Geiseric
instituted a succession by agnatic seniority – rule by the eldest surviving
male member of a ruling house.

The principle is explained clearly by

Procopius and Jordanes, two sixth-century Byzantine commentators:

Geiseric, after living on a short time, died at an advanced age, having
made a will in which he enjoined many things upon the Vandals and
in particular that royal power among them should always fall to that
one who should be the first in years among all the male offspring
descended from Geiseric himself.

Before his death he summoned the band of his sons and ordained
that there should be no strife among them because of desire for the
kingdom, but that each should reign in his own rank and order as he

For a survey of Geiseric’s foreign policy, see M.E. Gil Egea, ‘Piratas o estadistas: La política
exterior del reino vándalo durante el reinado de Genserico’, Polis

9 (1997), pp. 107–29 and the

excellent study of F.M. Clover, ‘Geiseric and Statesman: A Study of Vandal Foreign Policy’,
Ph.D. thesis, University of Chicago (

1967). The subject is addressed further in A.H. Merrills

and R.T. Miles, The Vandals (Oxford,

2010).

H. Schultze, De Testamento Genserici. Seu de antiquissima lege successoria in Germanorum regnis
(Jena,

1859) and D. Claude, ‘Problem der vandalischen Herrschaftsnachfolge’, Deutsches Archiv

für Erforschung des Mittelalters

30.2 (1974), pp. 329–55 provide general overviews. Courtois,

Vandales, pp.

238–44 gives a useful survey.

For speculation on earlier royal practice among the Vandals, compare Courtois, Vandales, pp.

239–40; Schultze, De Testamento Genserici, pp. 14–15; F.M. Clover ‘Timekeeping and Dyarchy

in Vandal Africa’, Antiquité Tardive

11 (2003), pp. 46–7 and 59–61.

Procopius, De Bello Vandalico I.

7.29, ed. and trans. H.B. Dewing, Procopius. History of the Wars

Book III: The Vandalic War, Loeb (Cambridge, MA,

1916), p. 72.

136

A.H. Merrills

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

survived the others; that is, the next younger should succeed his elder
brother, and he in turn should be followed by his junior. By giving
heed to this command they ruled their kingdom in happiness for the
space of many years and were not disgraced by civil war, as is usual
among other nations; one after the other receiving the kingdom and
ruling the people in peace.

Procopius refers to the law as Geiseric’s diathekai or ‘will’, Jordanes to

his praecepta.

The law clearly had a resonance beyond Geiseric’s direct

heirs and was intended to stand in perpetuity. This was not merely a royal
will, but an attempt to promulgate a lasting law. In this sense, the term
constitutio Genserici, employed by the North African historian Victor of
Vita, better illustrates the import of the ruling.

On the face of it, the constitutio was a resounding success. Geiseric’s

eldest son, Huneric, succeeded him, followed in

484 by Gunthamund, the

son of Theoderic and the oldest living Hasding. Gunthamund was suc-
ceeded in

496 by his younger brother, Thrasamund, who was then

followed in turn by Hilderic, Huneric’s eldest son and the eldest male
member of the family. According to laws of primogeniture, Hilderic would
have succeeded Huneric in

484, and would have come to the throne

relatively young. Instead, he succeeded as an old man upon the death of his
cousin in

523. The system only failed in 530 when Hilderic’s own heir

designate – Gelimer, his first cousin once-removed – lost patience with the
elderly king and rose in revolt. Gelimer’s own rule was then cut short by the
Byzantine invasion under Belisarius – an imperialist programme of ‘regime
change’ justified in part by the unconstitutional usurpation (Fig.

1).

Nineteenth- and early twentieth-century scholarship on the Vandal

state frequently placed Geiseric’s law of succession among the king’s most
important innovations.

Very little is known of Vandal legislation beyond

Jordanes, Getica

169, ed. T. Mommsen, MGH AA 5.1 (Munich, 1882), p. 102. The translation

used is that of C.C. Mierow, The Gothic History of Jordanes (Princeton, NJ,

1915).

Cf. Proc. BV I.

9.10, ed. Dewing, p. 86 and Jordanes, Getica 170, ed. Mommsen, p. 102. On the

significance of diatheke, see Schultze, De Testamento Genserici, pp.

13–14.

Victor Vitensis, Historia persecutionis Africanae prouinciae II.

13, ed. S. Lancel, Victor de Vita,

Budé (Paris,

2002), p. 127 and cf. Courtois, Vandales, p. 238.

Compare Schultze, De Testamento Genserici; idem, ‘Geschichtliche Entwickelung der fürstlichen
Hausverfassung im deutschen Mittelalter’, Zeitschrift für Rechtsgeschichte

7 (1868), pp. 323–405;

J. von Pflugk-Harttung, ‘Zur Thronfolge in den germanischen Stammesstaaten’, Zeitschrift der
Savigny-Stiftung für Rechtsgeschichte Germanische Abteilung

11 (1890), pp. 177–205; and D.

Martroye, ‘Le testament de Genséric’, Bulletin de la Société nationale des Antiquaires de France
(

1911), pp. 232–6. E.-F. Gautier, Genséric. Roi des Vandales (Paris, 1929), pp. 227–9; L. Schmidt,

‘The Sueves, Alans and Vandals in Spain

408–429. The Vandal Domination in Africa 429–533’,

in H.M. Gwatkin and J.P. Whitney (eds), The Cambridge Medieval History, I (Cambridge,

1924), p. 318 argues that the constitutio is ‘rightly . . . reckoned in history among the most

remarkable facts relating to public law’. Courtois, Vandales, p.

238 terms it ‘presque comme une

constitution au sens actuel du terme’ – at least in the eyes of later Byzantine commentators.

Dynasty and crisis in Vandal North Africa

137

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

ig.

The

asding

yal

line

138

A.H. Merrills

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

the constitutio, and almost nothing at all of secular law in the kingdom.

More significant still is the fact that Geiseric’s was the earliest known
attempt to establish a legal basis for royal (or imperial) succession, and
provides a rare glimpse into the legal institutions of ‘Germanic’ kingship
before the sixth century. For many commentators, the constitutio repre-
sented a valuable survival of prehistoric patterns of ‘Germanic’ kingship
into the late antique period, and a missing link between the primitive
systems of tribal government described in Tacitus’ Germania and the
more familiar monarchic apparatus of the Middle Ages.

The fact that the

eventual overthrow of the law under Gelimer was commonly ascribed to
the political dissatisfactions of a still-powerful Vandal aristocracy (a view-
point which the account of Procopius certainly encourages), further
suggested that this was a moment of profound political transition from
the ancient to the medieval worlds.

Few scholars in the early twenty-first century would have confidence in

such bold teleologies, of course. Tacitus is no longer read as a reliable
source on the barbarian societies of the first century, let alone the fourth,
fifth or sixth, and attempts to view in his text the embryonic outline of
later political institutions have largely passed out of favour. So, too, has
the notion of a single ‘Germanic’ identity shared by a variety of different
polities across the late antique and early medieval world. As the state
structures of Ostrogothic Italy, Merovingian Francia and Visigothic Spain
are increasingly examined and appreciated in their own provincial con-
texts, the uncritical importation of fragments of information from else-
where is (correctly) viewed with distrust. Most scholars would now agree
that Tacitus’ Germania can provide little illumination on Geiseric’s law of
succession, and that it in turn can tell us little of direct value concerning
the later development of Gothic, Lombard or Anglo-Saxon kingship. But
as a means to explore the political reality of the Vandal kingdom, the

G. Vismara, ‘Gli editti dei re Vandali’, in Studi in Onore di Gaetana Scherillo,

3 vols (Milan,

1972), II, pp. 849–78 provides an overview of Vandal law-making. The paucity of material is

illustrated by the cursory treatment in both H. Conrad, Deutsche Rechtsgeschichte. Band I.
Frühzeit und Mittelalter (Karlsruhe,

1962), pp. 61–3 and P. Classen, Kaiserreskript und König-

surkunde. Diplomatische Studien zum Problem der Kontinuität zwischen Altertum und Mittelalter,
Byzantine Texts and Studies

15 (Thessaloniki, 1977), pp. 107–10.

Discussed in Claude, ‘Problem der vandalischen Herrschaftsnachfolge’, pp.

330–3. The impli-

cation that Geiseric’s constitutio was a reflection of an ancient ‘Germanic’ institution is refuted
by the absence of comparable institutions in the other successor kingdoms – a point noted by
Vismara, ‘Gli editti dei re Vandali’, p.

853. The closest parallel is found in the Visigothic royal

succession of AD

451–84, in which Theoderic I was succeeded in turn by his sons Thorismun-

dus, Theoderic II and Euric. The fact that Thorismundus and Theoderic II were both violently
deposed by their younger brothers suggests that the succession was not formally ordered. Euric
was eventually succeeded by his own son Alaric (II) although it is not clear whether any older
members of the family were alive at that time. See J.R. Martindale, Prosopography of the Later
Roman Empire, vol II. AD

395–527 (Cambridge, 1980), stemma 40.

See for example, Claude, ‘Problem der vandalischen Herrschaftsnachfolge’, p.

352.

Dynasty and crisis in Vandal North Africa

139

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

operation of royal power and the role of tradition in its articulation, the
constitutio remains of tremendous significance.

Crucial to the appreciation of the constitutio is the recognition that it

was not an abstracted political ideal, floating above the brutalities of
Vandal Realpolitik, but a law that was fundamental to these struggles.
This was true of the origin of the legislation, and also of its application
over the generations that followed. To illustrate the point, one need
merely reflect that our knowledge of the law is dependent upon sources
ostensibly hostile to the Vandal state itself. Victor of Vita was an outspo-
ken critic of the Hasding court, and never accepted its legitimacy, yet
happily condemned the actions of its kings for their supposed rejection of
the constitutio; Procopius and Jordanes were less directly critical of the
Hasdings (writing, as they were, almost two decades after the fall of
Carthage to Belisarius), yet presented the emperor Justinian as the only
true defender of Geiseric’s testamentum. These authors framed their
discussions around two periods of intense political crisis in the Vandal
kingdom: first in AD

480/1 when Huneric sought to consolidate his own

position on the throne through the removal of political rivals, and then in
AD

530 when Gelimer successfully usurped Hilderic’s rule and claimed

authority for himself. Critics of these kings found in the constitutio a
useful rhetorical device to frame their condemnation, but there is every
reason to suggest that both Huneric and Gelimer felt themselves to be
acting in the spirit of Geiseric’s law. It is the intention of the current
article to explore the contexts of these two major constitutional crises,
and to examine the role of the succession legislation within it.

It is necessary first to examine briefly the political context in which the

law was formulated. The motivations behind Geiseric’s law have been
variously explained. The king had himself succeeded to the throne in
rather murky circumstances following the death of his half-brother Gun-
deric. Popular rumour suggested that this was followed in short order by
the murder of Gunderic’s widow and children – vicious acts which may
reflect the dubious legality of Geiseric’s claim to the throne.

While some

have suggested that the institution of agnatic seniority was intended to
circumvent the succession of a minor, and hence to avoid the sort of
dynastic crisis that had hit Ravenna in the early fifth century, others have
noted that Geiseric might have been motivated by a desire to legitimate
retrospectively his own ascent to the throne.

These themes are explored briefly by Courtois, Vandales, p.

240.

A rumour happily reported by Vict. Vit. HP II.

14, ed. Lancel, p. 129. On the dating of these

murders to early in Geiseric’s reign, see Claude, ‘Problem der vandalischen Herrschaftsnach-
folge’, pp.

330–6.

Courtois, Vandales, pp.

239–40; H. Wolfram, The Roman Empire and its Germanic Peoples,

trans. Thomas Dunlap (Berkeley,

1997), p. 165.

140

A.H. Merrills

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

For Geiseric himself, however, the firm statement of dynastic principle

that the constitutio represented also had immediate political implications.
In many ways, the whole of Geiseric’s reign may be viewed as a conscious
and deliberate construction of a new form of Vandal kingship for a group
redefining itself within the rich African provinces. Jordanes implies that
the constitutio was promulgated on the king’s deathbed, and hence was
among the last of Geiseric’s actions, but important political innovations
can be traced back throughout his reign, and it is certainly possible that
the formal establishment of the Hasding succession was made long before
the king’s death. As early as AD

442, Geiseric had laid the foundations for

a strong monarchy within the Vandal state.

Prosper reports the suppres-

sion of a major aristocratic rebellion against the king’s rule in that year;

Procopius implies that there was a fundamental reorganization of the
Vandal army at around the same time, and various sources describe the
diplomatic manoeuvrings which followed the formal peace treaty with
Valentinian III in the same year.

It was in this period, too, that the

Vandals were formally settled within North Africa – a process which was
to have fundamental implications for royal power within the group.

The practical operations of the settlement in AD

442 have been the

subject of considerable recent scholarship, and need not detain us exces-
sively here. Broadly, disagreement among modern scholars has centred
upon the nature of the land confiscations themselves, and whether the
invaders were physically settled on estates within their new homeland, or
simply supplied by a proportion of the tax revenues taken from lands
nominally made over to them.

While the popular consensus seems to be

On the year AD

442 as a turning point, see E. Stein, Histoire du Bas-Empire. Tome Premier. De

l’État Romain à l’État Byzantine (

284 – 476), trans. Jean-Remy Palanque (Paris, 1959) p. 326; Gil

Egea, África en Tiempos de los Vándalos, pp.

314–15; Berndt, Konflikt und Anpassung, pp. 149–50;

as well as Merrills and Miles, Vandals, pp.

61–70.

Prosper, Chronicon, a.

442, ed. T. Mommsen, MGH AA 9 (Berlin, 1892), p. 479.

The reorganization of the army is implied at Proc. BV I.

5.18–20, ed. Dewing, p. 52. For

discussion, see Gil Egea, África en Tiempos de los Vándalos, pp.

333–4.

The crucial sources on the settlement are Proc. BV I.

5.11–17, ed. Dewing, p. 50 and Vict. Vit.

HP I.

13, ed. Lancel, p. 102.

The former thesis broadly follows the brilliant argument laid out by W. Goffart, Barbarians and
Romans A.D.

418–584. The Techniques of Accommodation (Princeton, 1980), which is primarily

concerned with hospitalitas in Italy, Spain and Burgundy. For summary of this work and the
scholarship since, see now G. Halsall, Barbarian Migrations and the Roman West.

376–568

(Cambridge,

2007), pp. 422–7. The same model was applied to North Africa by J. Durliat, ‘Les

grands propriétaires africains et l’Etat byzantin (

533–709)’, Cahiers de Tunisie 29 (1981), pp.

517–31; idem, ‘Le salaire de la paix sociale dans les royaumes barbares V

–VI

siècles’, in Herwig

Wolfram and Andreas Schwarcz (eds), Anerkennung und Integration. Zu den wirtschaftlichen
Grundlagen der Völkerwanderungszeit

400–600, Österreichische Akademie der Wissenschaften,

phil.-hist. Kl., Denkschriften,

193 (Vienna, 1988), pp. 21–72; and has since been restated by A.

Schwarcz, ‘The Settlement of the Vandals in North Africa’, in A.H. Merrills (ed.), Vandals,
Romans and Berbers. New Perspectives on Late Antique North Africa (Aldershot

2004), pp. 49–57.

Prior to Goffart, it was always assumed that Vandal settlement had depended upon extensive

Dynasty and crisis in Vandal North Africa

141

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

leaning towards the former view – not least because of the extensive
lamentations regarding lost lands in a variety of textual sources from the
period – much of this discussion has missed a particularly significant
feature of the settlement: both Victor and Procopius note a clear distinc-
tion between the sortes Vandalorum assigned to the ‘Vandals’ as a whole,
and the royal estates divided up between Geiseric and his sons.

Both

stress further that the sortes Vandalorum were limited to the province of
Africa Proconsularis, while the Hasding lands included all of the confis-
cated territories in Numidia and Byzacena, as well as some holdings in the
Proconsular province. These royal holdings evidently included vineyards,
mines, lakes and urban palaces as well as agricultural estates and villas.

The treaty of

442 established the Vandals within North Africa, but it also

provided a social and economic foundation for the power of the Hasding
family among them. Just as important as the provision of sustenance for
the Vandals was the formal institution of a distinction between the royal
family and their followers. This was accentuated still further by Geiseric’s
surprising innovation of establishing a court for each of his sons.

Victor

of Vita alludes to bureaucratic personnel at the court of the crown prince
Huneric, and notes further that the entourage of the younger prince
Theuderic was itself an important political focus during the reign of his
father.

Throughout his legislation, Geiseric was evidently anxious to

secure the position of his sons. The constitutio both established the status
of the princes during the founding king’s lifetime and provided for their
continued prosperity after his death.

All of this had several lasting results. Geiseric’s concern to establish all

of his sons successfully underscored his ambitions for the Hasding house:
following the suppression of the aristocratic revolt in

442 we hear of no

challenge to the Vandal throne coming from outside the family. The
generous distribution of lands and influence created a series of alternative
political foci within the Vandal kingdom, however. While Geiseric
himself was still alive, this seems to have posed few problems, but it
proved to be a lasting threat for his successors. Under Huneric, it was the

land appropriations. The case for this has now been eloquently re-articulated by Y. Moderan,
‘L’établissement territorial des Vandales en Afrique’, Antiquité Tardive

10 (2002), pp. 87–122 and

passim. M. Innes, ‘Land, Freedom and the Making of the Medieval West’, TRHS

16 (2006), pp.

39–76, at p. 48 persuasively argues that the distribution of lands within Vandal Africa must be

considered separately from processes in operation elsewhere in the western empire. His prom-
ised analysis of this is eagerly awaited.

Vict. Vit. HP I.

13, ed. Lancel, p. 102; Proc. BV I.5.11–14, ed. Dewing, p. 50.

Mines – Vict. Vit. HP III.

68, ed. Lancel, p. 210 and Notitia provinciarum et civitatem Africae

II.

76, ed. S. Lancel, Victor de Vita, p. 257. Vineyards – Vict. Vit. HP I.44, ed. Lancel, p. 117. On

Vandal palaces, cf. n.

72 below. Cf. L. Schmidt, Histoire des Vandales, trans. H.E. Del Medico

(Paris,

1953), p. 220, n. 5.

Proc. BV I.

5.11, ed. Dewing, p. 50.

Vict. Vit. HP I.

45, ed. Lancel, p. 117; I.48, ed. Lancel, p. 119.

142

A.H. Merrills

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

presence of powerful rivals which precipitated the first crisis; under his
son Hilderic, it was the creation of new power blocs within an expanding
Hasding family which was the greatest danger.

Huneric: blood and irony

Huneric’s own succession cannot have been in doubt upon the death of
his father in

477. As the eldest son of Geiseric, and the oldest living

member of the family, his claim to the throne could scarcely have been
challenged. His marriage to the princess Eudocia, daughter of the late
Valentinian III, would have lent him legitimacy in the eyes of observers
in North Africa and beyond, and his period as a hostage in the imperial
court at Ravenna doubtless further strengthened his credentials.

This

imperial cachet had been purchased at a high cost – Geiseric’s rejection
and mutilation of Huneric’s first wife, an unnamed Visigothic princess,
probably in around

442 – but on the international stage of the mid-fifth

century, this was evidently felt to be a price worth paying.

Shortly after

coming to the throne, Huneric sent envoys to Constantinople, renounc-
ing all Vandal ambitions in the Mediterranean in return for recognition
in the east, and made similar concessions to Odoacar in Italy shortly
before his death. But for all of these advantages, Huneric’s proved to be
an unstable reign. The period is known to us almost exclusively through
Victor of Vita’s alarmist Historia Persecutionis. As the title of his work
suggests, Victor’s principal concern was with the persecution of the
African Nicene church, first under Geiseric and then under his successor,
particularly during the bloody repression of

484. As has long been rec-

ognized, Victor’s polemical intentions fundamentally shape the form of
the Historia as a whole, and the fragments of secular history which appear
within the text need to be viewed as contributions to this rhetorical
programme.

The account of the first half of Huneric’s reign provides an obvious

case in point. According to Victor, Huneric was initially welcomed
within the kingdom. His persecution of the Manichaeans was regarded as
commendably regal behaviour by even this most virulent Nicene, and

Proc. BV I.

4.13–14, ed. Dewing, pp. 36–8; Merobaudes, Carmina I.7–8, ed. F. Vollmer, MGH

14 (Berlin, 1905), p. 3 which may refer to Huneric while in Ravenna. See F.M. Clover, Flavius

Merobaudes. A Translation and Historical Commentary, Transactions of the American Philo-
sophical Society, ns

61 (Philadelphia, 1971), pp. 20–1 for translation and discussion.

Jordanes, Getica

184, ed. Mommsen, p. 106. On this marriage cf. Clover, ‘Geiseric and States-

man’, pp.

107–8.

C. Courtois, Victor de Vita son oeuvre. Étude critique (Algiers,

1954) remains central on Victor.

And see now T. Howe, Vandalen, Barbaren und Arianer bei Victor von Vita (Berlin,

2007). The

recent edition and commentary of Lancel, Victor de Vita has made criticism of the text
immeasurably easier.

Dynasty and crisis in Vandal North Africa

143

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

under pressure from Constantinople the king allowed the ordination of
Eugenius as bishop of Carthage – the first to hold the post since the death
of Deogratias twenty-four years earlier.

Thereafter, however, things

apparently went downhill. In

480 or 481, Huneric engaged in a substan-

tial dynastic purge, first killing or expelling various members of his
family, and then alienating the church through his clumsy attempts to
redefine the legal succession of the Vandal realm in favour of his son,
Hilderic. In effect, Victor describes three stages to this political extirpa-
tion, although the relationship between them is not completely clear.
First is the active purge of the Hasding house – the murder of some
members of the family, the exile of others, and the torture of an unspeci-
fied number of Vandal aristocrats for their supposed involvement in a
plot. Second was the removal of Geiseric’s comitatus, a small group of
close advisers who had been loyal to the earlier king and had (Victor tells
us) been specifically recommended to Huneric by his father. Third came
Huneric’s rather peculiar scheme to gain Catholic episcopal support for
his redefinition of Geiseric’s constitutio. This provides a precursor to the
religious persecutions proper.

Victor’s account is fullest on the actions taken by Huneric against the

members of his immediate family.

According to the Historia, Huneric’s

earliest political actions were directed against the family of his brother,
Theoderic. The first of these victims was Theoderic’s wife, who was
murdered on suspicion of stirring her husband and her son into revolt.
This son (who is not named in Victor’s account) was the next of Huner-
ic’s victims and was also put to death.

From here, the historian recounts

a catalogue of horrors with relish. First, the Arian ‘patriarch’ Jucundus,
was publicly executed on the strength of his association with Theoderic’s
household – a political connection which stretched back at least as far as
the reign of Geiseric.

Theoderic himself died in exile, and his younger

son and two adult daughters were also expelled from the kingdom. The
final phase of this political pogrom witnessed Huneric attack the numer-
ous counts and nobles associated with his brother, apparently burning
some and executing others.

At around the same time, or shortly there-

after, he exiled Godagis, the son of his youngest brother Genton, and
harassed the surviving members of Geiseric’s court circle who had been
recommended to the new king by his father. This included Heldica, the

Vict. Vit. HP II.

2–6, ed. Lancel, pp. 122–5.

Vict. Vit. HP II.

12–16, ed. Lancel, pp. 127–9.

Vict. Vit. HP II.

12–13, ed. Lancel, p. 127.

Vict. Vit. HP II.

13, ed. Lancel, p. 127. The length of Jucundus’ association with the court of

Theoderic is suggested by HP I.

44, an episode which took place when Geiseric was still king

and Jucundus still a presbyter.

Vict. Vit. HP II.

14, ed. Lancel, p. 128.

144

A.H. Merrills

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

praepositus regni of Geiseric’s kingdom, who was publicly executed along
with his wife, and his brother Gamuth, who was repeatedly humiliated,
condemned to menial labour and flogged each month for a period of five
years.

From the start of his account, Victor is at pains to stress that this whole

brutal process was motivated by Huneric’s desire to secure the succession
of his own son, in preference to the agnatic seniority instituted by his
father. The point is established at the outset of the episode:

This man, who had until then shown himself mild to everyone, wished
to assign the kingdom to his sons after his death; as it happened this
did not come to pass.

The composition of the bulk of the Historia Persecutionis is normally

dated to

484. Consequently, the allusion here to Huneric’s death and the

succession of Gunthamund must have been a later interpolation, and is
generally thought to have been the historian’s own, rather than the work
of a later redactor.

At any event, Victor himself was evidently keen to

emphasize the underlying motivation behind Huneric’s actions, and its
unconstitutional nature. The point is made again in the description of
Huneric’s second murder:

Afterwards, that son, who had received a higher education, was killed
as well; to him, in particular, according to the constitutio of Geiseric,
the chief rule was due among his nephews, because he was senior of
them all.

Compelling as this narrative may seem, the principal motive for

Huneric’s actions cannot have been so straightforward.

A concern to

secure the succession of Hilderic in opposition to Geiseric’s legislation

Vict. Vit. HP II.

15–16, ed. Lancel, pp. 128–9.

Vict. Vit. HP II.

12, ed. Lancel, p. 127.

Lancel, Victor de Vita, pp.

297–8; cf. Courtois, Victor de Vita, pp. 16–17; S. Costanza,

‘ “Uuandali-Arriani” e “Romani-Catholici” nella Historia Persecutionis Africanae Provinciae di
Vittore di Vita. Uno controversia per l’uso del latino nel concilio cartaginese del

484’, in

Oikoumene. Studi Palaeocristiani pubblicati in onore del concilio ecumenico Vaticano II (Catania,

1964), pp. 223–41, at p. 241.

Vict. Vit. HP II.

13, ed. Lancel, p. 127. Moorhead’s rendering of inter nepotes potissimum as

‘eldest’ among the nephews probably reflects Victor’s intended meaning, but the superlative
may hide a subtlety behind the succession convention which is now lost.

Most secondary accounts have followed Victor uncritically. H. Wolfram, The Roman Empire
and its Germanic Peoples (Berkeley,

1997), p. 174; Lancel, Victor de Vita, p. 297, n. 137; Claude,

‘Problem der vandalischen Herrschaftsnachfolge’, pp.

338–42. Courtois, Vandales, pp. 241–2

acknowledges that there is more to the episode than meets the eye, but still assumes that the
whole episode reflected Huneric’s dynastic ambitions, rather than a pragmatic response to an
immediate threat.

Dynasty and crisis in Vandal North Africa

145

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

explains neither Huneric’s strategy of piecemeal execution and exile
towards other Hasdings, nor his brutal purge of the secular and episcopal
aristocracy. Even if we accept Victor’s statement that Theoderic’s son was
older than Hilderic, and hence stood to inherit the Vandal throne before
his cousin, the politic execution of the prince would scarcely have
improved Hilderic’s position. According to Geiseric’s law, Huneric’s
immediate successor would have been Theoderic himself as the oldest
living Hasding. Genton, Huneric’s youngest brother, died before

477,

but he left behind four sons, at least three of whom were older than
Hilderic, and thus had priority over him in the succession.

Victor states

that Godagis was the eldest of these sons, and that Huneric simply sent
him into exile.

Victor says nothing of Gunthamund (who was eventu-

ally to succeed Huneric), Thrasamund (the next king after that), or
Geilarith (the youngest of the four sons, but who still may have been
Hilderic’s senior).

The implications of this deserve some discussion. Huneric’s actions

against his own family cannot have been motivated by the simple desire
to ensure the seamless succession of his son to the throne, as Victor
explicitly states. Of the five Hasdings who certainly stood to come to the
throne before Hilderic (Theoderic, Theoderic’s son, Godagis, Guntha-
mund and Thrasamund), Huneric killed one and exiled two more but
seems to have taken no action against the others. Furthermore, we know
two further Hasdings who may also have been older than Hilderic (and
hence senior to him in the eyes of the law) – Theoderic’s younger son,
and Genton’s youngest son Geilarith. The former was sent into exile with
his father, the latter apparently left unharmed.

Huneric’s actions are far more easily explained as a straightforward

response to a specific political threat, focused around Theoderic’s elder
son.

By executing the prince and his mother, as well as Jucundus and

other long-standing allies of Theoderic’s court, Huneric acted against a
direct threat to his own power, not a perceived danger to that of his son.
The fact that Theoderic and Godagis were merely exiled, and other
Hasding princes left unharmed, further suggests that this was a purge
with an immediate political function. Quite how this related to the
subsequent persecution of Geiseric’s former followers is unclear. They
may have been implicated in the plot surrounding the anonymous
prince, or might have fomented rebellion of their own. They may even
have fallen from favour for entirely unconnected reasons. But these were
pragmatic responses to an immediate political threat, not putative state-

Genton’s death is reported in BV I.

8.1, ed. Dewing, p. 72.

Vict. Vit. HP II.

14, ed. Lancel, p. 128.

Courtois, Vandales, pp.

240–1.

146

A.H. Merrills

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

ments of dynastic ambition. Indeed, Victor rather gives the game away in
his concluding statement on the episode:

Having therefore in short time disposed of all those he feared and
made his reign secure, as he thought . . . he turned all the missiles of
his reign towards a persecution of the Christian Church.

Huneric’s concern was to halt a rising against his own rule, not to

secure the position of his son. That initiative would come later.

Victor’s insistence that Huneric’s actions were determined by dynastic

ambition, rather than political pragmatism is explained easily enough. As
the historian himself acknowledged, secular events rarely fit comfortably
within the Historia Persecutionis.

A work intended for the edification

of Catholics had little use for accounts of the political persecution of
Arians. Victor nevertheless included these passages, with the ostensible
purpose of illustrating the insanity and irrationality of the persecutors. The
purge of prominent Vandals fits well within this schema. Victor’s enu-
meration of the public execution of Theoderic’s widow and son, the
sufferings in exile of Godagis, and the desperation of the Vandal noble
Gamuth in taking refuge in a church cess-pit highlight the cruelty of the
king, in vivid anticipation of the religious persecutions that were to follow.
The reference to the classical education of Theoderic’s son, and the
political prominence of many of his fellow victims, similarly underscores
Huneric’s perceived barbarity in carrying out these purges – a theme to
which Victor repeatedly returned. Finally, by presenting all of these actions
as an expression of Huneric’s dynastic pretensions – as a direct contradic-
tion in other words to the constitutio of his father – Victor could present the
new king as an illegitimate ruler in the eyes of his own laws. Huneric’s
actions could be demonized easily enough without this dissimulation, but
the conflation of his later political pretensions with his earlier pragmatism
proved to be an easy enough process for the Catholic historian.

Victor was helped in this misrepresentation by the fact that Huneric

did eventually attempt to redefine the Vandal succession in favour of
Hilderic, but this was probably not connected directly with the political
purges of

480/1. In a peculiar passage midway through the third book of

the Historia Persecutionis, Victor describes Huneric’s assembly of a
number of Catholic bishops in Carthage, at the site of the former Temple
of Memoria.

Here, Huneric apparently blackmailed the episcopal

Vict. Vit. HP II.

17, ed. Lancel, p. 129.

Vict. Vit. HP II.

13, 16, ed. Lancel, pp. 127, 129. For discussion, Costanza, ‘Uuandali-Arriani’,

pp.

233–4.

Vict. Vit. HP III.

17–20, ed. Lancel, pp. 182–4.

Dynasty and crisis in Vandal North Africa

147

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

aristocracy, effectively intimating that religious freedom might be theirs
in return for support in his new dynastic scheme:

That piece of chicanery read as follows: ‘Swear that, after the death of
our lord the king, you wish his son Hilderit [Hilderic] to be king, and
that none of you will send letters to lands across the sea, for if you give
your oath concerning this, he will restore you to your churches.

This oath vividly illustrates the continued political authority that the

Nicene episcopate enjoyed, even towards the latter stages of Huneric’s
reign. Apparently reluctant to present the church and state in such close
proximity, Victor’s interpretation of Huneric’s actions is typically eccen-
tric. Incredibly, he suggests that the main motive for Huneric’s overture
was not to establish the position of his own son on the throne, but to trick
the more receptive Catholic bishops into breaking Christ’s command-
ment at Matthew

5.13 against the taking of oaths. Victor’s Catholic

solipsism is such that a genuine attempt at political reform is presented
simply as an esoteric attack on the coherence of the church.

Huneric’s gambit, however, was not a success. After the king’s death,

his nephew Gunthamund acceded to the throne, and very probably
instituted a political purge of his own. There is certainly some suggestion
that supporters of the deceased king, perhaps including the poet Dra-
contius, found themselves out of favour.

Whether this is best regarded

as a triumph for Geiseric’s constitutio, or a reflection of the realities of
political power in the Vandal kingdom of

484, is unclear. We simply

know too little about the reign of Gunthamund to state with certainty the
circumstances of his succession.

Gelimer

530, the political landscape of Vandal North Africa had changed

greatly. The vigorous pollarding of the Hasding family tree under Gei-
seric and Huneric doubtless did much to secure the peaceful succession of
Gunthamund, Thrasamund and Hilderic, and hence lent the constitutio
the patina of institutional antiquity.

By the second quarter of the sixth

century, however, the family branches were again starting to spread and
new dynastic tensions began to emerge.

Vict. Vit. HP III.

19, ed. Lancel, pp. 183–4.

For discussion see A.H. Merrills, ‘The Perils of Panegyric: The Lost Poem of Dracontius and its
Consequences’, in idem (ed.), Vandals, Romans and Berbers. New Perspectives on Late Antique
North Africa (Aldershot

2004), pp. 145–62.

Courtois, Vandales, p.

242.

148

A.H. Merrills

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

The ruling monarch at the time of the second crisis was Hilderic,

Huneric’s ill-starred son, who had finally gained the throne in

523.

Hilderic had probably been born around

456, shortly after the arrival of

his mother Eudocia from Rome, and certainly no later than

472 when she

left Carthage for the Holy Land.

Consequently, the new king was at

least fifty-one, and perhaps as old as sixty-seven at the time of his
succession. He was caricatured after his death as a frail and somewhat
indecisive king, and seems to have entrusted to others the thankless
struggle with the neighbouring Moorish polities of the south. But despite
this, his reign was marked by significant innovation. His conversion to
Catholicism was the most dramatic cultural change in the history of the
kingdom, and was apparently made against the dying wishes of his
predecessor Thrasamund.

Scarcely less momentous was the axial shift in

foreign policy which he instituted, imprisoning and later executing
Amalafrida, the Ostrogothic widow of Thrasamund, in favour of increas-
ingly close ties to Justinianic Byzantium.

The challenge to Hilderic’s rule came in the person of the Hasding

prince Gelimer. The son of Genton’s youngest son Geilarith, and hence
the nephew of Gunthamund and Thrasamund, Gelimer was the eldest
living Vandal after Hilderic and stood to inherit the kingdom after his
death. All of our sources for the episode present the crisis from an
essentially Byzantine perspective, and all agree that the usurpation was
the action of a restless and violent subordinate against an ineffective, but
ultimately inoffensive, elderly ruler.

Procopius’ Vandal War provides the

fullest account of these events: in it, Gelimer’s usurpation is explicitly
cited as a central justification for the Byzantine intervention in Africa.
The appeal of this narrative is illustrated by the alacrity with which later
writers adopted Procopius’ account, both in Constantinople and
beyond.

In his epic Iohannis, written in celebration of the Byzantine

Cf. Martindale, Prosopography of the Later Roman Empire, vol II., pp.

564–5 (Hildericus) and

407–8 (Eudocia 1).

Victor Tunnunensis, Chronicon, a.

523.2, ed. A. Placanica, Vittore da Tunnuna Chronica: Chiesa

e Impero nell’età di Giustiniano (Florence,

1997), p. 36 followed by Isidore of Seville, Historia

Gothorum, Wandalorum, Sueborum,

82, ed. C. Rodríguez Alonso, Las Historias de los Godos,

Vandalos, y Suevos de Isidoro de Sevilla (Léon,

1975), p. 302; cf. also Ferrandus, Vita Fulgentii, 28,

ed. and trans. G.-G. Lapeyre, Vie de Saint Fulgence de Ruspe (Paris,

1929); Proc. BV I.9.1, ed.

Dewing, p.

82.

Proc. BV I.

9.4–5, ed. Dewing, p. 84.

Proc. BV I.

9.10, ed. Dewing, p. 86; Corippus, Iohannidos III.262–4, ed. J. Diggle and F.R.D.

Goodyear, Flavii Cresconii Corippi Iohannidos Libri VIII (Cambridge,

1970), p. 57. The text is

usefully translated by G.W. Shea, The Iohannis or De bellis Libycis of Flavius Cresconius Corippus
(Lewiston, NY,

1998).

Sources influenced by Procopius: Theophanes, Chronographia, AM

6026, ed. C. De Boor,

Theophanis Chronographia (Hildesheim,

1963), pp. 186–217 (which provides a compression of

Procopius’ narrative with some minor changes); see also C. Mango, R. Scott and G. Greatrex,
trans. and comm., The Chronicle of Theophanes Confessor: Byzantine and Near Eastern History,

Dynasty and crisis in Vandal North Africa

149

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

occupation, the North African poet Corippus presents the usurpation as
the result of an ambitious tyrant exploiting an exhausted province.

the Chronicle of Victor of Tunnuna – another North African with strong
Byzantine sympathies – Gelimer again is presented as a usurper, but with
no explanatory context.

Yet there is certainly more here than meets the eye.

Procopius may

have damned Gelimer to posterity, but the text of his Vandal War does
not do a particularly good job of hiding the tensions which bubbled
beneath the surface of the crisis. Procopius sets out Gelimer’s justification
for revolt:

. . . allying with himself all the noblest of the Vandals, he persuaded
them to wrest the kingdom from Hilderic, as being an unwarlike king
who had been defeated by the Moors, and as betraying the power of
the Vandals into the hand of the Emperor Justinus, in order that the
kingdom might not come to him because he was of the other branch
of the family; for he asserted slanderously that this was the meaning of
Hilderic’s embassy to Byzantium, and that he was giving over the
empire of the Vandals to Justinus.

The usurpation itself seems to have been carried out with little diffi-

culty. Gelimer seized the throne with substantial support from the Vandal
military, and imprisoned Hilderic and his two nephews Hoamer and
Hoageis.

Procopius goes on to describe the prolonged diplomatic exchange with

Constantinople which followed the usurpation in AD

530.

Justinian, an

ally and patron of Hilderic, lambasted the new tyrant for his unconsti-
tutional actions. In the first of his letters, Justinian demanded that
Gelimer relinquish his power and wait for the throne to pass to him in the

284–813 (Oxford, 1997). Evagrius, Historia ecclesiastica IV.17, ed. J. Bidez and L. Parmentier,

Evagrius Scholasticus: Historia Ecclesiastica (London,

1898), and trans. M. Whitby, The Ecclesi-

astical History of Evagrius Scholasticus (Liverpool,

2000), pp. 217–18. (Evagrius is silent on the

usurpation, but dwells at some length on the subsequent triumph).

Corippus, Iohannidos III.

262–4, ed. J. Diggle and F.D.R. Goodyear, Flavii Cresconii Corippi

Iohannidos seu de Bellis Libycis Libri VIII (Cambridge,

1970), p. 57; trans G.W. Shea, The

Iohannis or De bellis Libycis of Flavius Cresconius Corippus (Lewiston,

1998).

Vict. Tun., a.

531, ed. Placanica, p. 38, and a. 533, ed. Placanica, p. 38; followed by Isidorus

Hispalensis, Historia Gothorum, Wandalorum, Sueborum

82–3, ed. Cristóbal Rodríguez Alonso-

Núñez, Las Historias de los Godos, Vandalos, y Suevos de Isidoro de Sevilla (León,

1975), pp. 304–6.

For the standard narrative, see J.R. Martindale, Prosopography of the Later Roman Empire, vol
IIIA. AD

527–641 (Cambridge, 1992), pp. 506–8 (Gelimer). Again, Courtois, Vandales, p. 242

recognizes suspicious features in the received account, but does not explore the full implications
of the division.

Proc. BV I.

9.8, ed. Dewing, pp. 84–6.

Proc. BV I.

9.9, 14, ed. Dewing, pp. 86, 88.

Proc. BV I.

9.10–26, ed. Dewing, pp. 86–90.

150

A.H. Merrills

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

natural course of things, according to the will of Geiseric. In response,
Gelimer dismissed Justinian’s envoys, kept Hilderic and Hoageis incar-
cerated and blinded Hoamer. In a second letter, an unspecified time later,
Justinian berated the usurper for his actions, threatening imperial inter-
vention in the name of Geiseric if the deposed Hasdings were not sent to
Constantinople.

Gelimer’s response to this approach is not certain. Procopius states that

the usurper insisted upon the justice of his revolt against Hilderic who
was himself ‘planning a revolution against the house of Geiseric’, and
effectively provoked Justinian into action.

John Malalas claims that

Gelimer took a more conciliatory tone, but was himself rejected by the
emperor, who then set about appealing to the Ostrogoths in Italy as part
of a grand coalition against the Vandal kingdom.

Having established his

position, Justinian then used the Hasding succession crisis as one of
several pretexts for his intervention within North Africa. Procopius,
Lydus, Malalas, the Latin chronicler Victor of Tunnuna and the short
king-list conventionally known as the Laterculus regum Vandalorum all
agree that Belisarius’ landing in Africa led Gelimer to initiate a purge of
Hilderic’s family, including the murder of the former king and his two
nephews, although we are told that his children and grandchildren
escaped to Constantinople.

At first blush, this episode reads as a fairly straightforward opposition

between Vandal barbarism and Byzantine legality – which is surely the
impression that Procopius and the other sources wished to give. Gelimer
is cast as a militaristic usurper, encouraged in his actions by a coterie of
aristocratic traditionalists who founded his rebellion upon a deep-rooted
suspicion of the Romanitas of the ruling monarch. The whole episode,
indeed, is a close anticipation of the Gothic aristocratic agitation against
the Athalaric in Procopius’ Gothic War.

Gelimer’s barbarism provides

the only causal explanation for his usurpation – an action determined by
his irrational impatience at taking up his position on the throne and an
unreasonable suspicion at the cultured pretensions of Hilderic and his
courtiers.

The usurpation has almost always been interpreted in these terms, but

it is perhaps appropriate to pay Gelimer the compliment of taking his

Proc. BV I.

9.20–26, ed. Dewing, pp. 88–90.

John Malalas, Chronographia,

459–60, ed. L. Dindorf (Bonn, 1981), trans. E. Jeffreys, M. Jeffreys

and R. Scott, The Chronicle of John Malalas, Byzantina Australiensia

4 (Melbourne, 1986), p.

459. Theophanes AM 6026 provides a further account of the diplomatic exchange, in which

Gelimer made the first contact with Justinian – despite the likelihood that Procopius was the
chronicler’s only source on the episode.

Proc. BV II.

9.13, ed. Dewing, p. 282; Vict. Tun., a. 534.1, ed. Placanica, pp. 38–40.

Procopius, De Bello Gothico I.

2.1–22. ed. and trans. H.B. Dewing, Procopius. History of the Wars

Books V and VI, Loeb (Cambridge, MA,

1919), pp. 14–20.

Dynasty and crisis in Vandal North Africa

151

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

political grievances seriously. Procopius notes that Gelimer’s fear that the
rule of the kingdom ‘might not come to him’ was a major motivation
behind the rising.

More specifically, Hilderic’s own machinations

against ‘the house of Geiseric’ are cited by Gelimer as his central griev-
ances.

Procopius implies, of course, that this fear was based upon the

unfounded suspicion that Hilderic wished to surrender Vandal indepen-
dence to the eastern empire, and thus reflected Gelimer’s barbarism. The
wholesale betrayal of the Vandal kingdom to Constantinople seems
inherently unlikely, but there is every reason to suspect that Hilderic’s
burgeoning Romanitas did, in fact, represent a direct threat to Gelimer’s
inheritance. Circumstantial evidence from Hilderic’s reign, and the pecu-
liar behaviour of Gelimer upon seizing the throne, encourage the view
that the usurpation arose from a genuine split within the Hasding house
– between the aspirations to ‘Roman’ rule on Hilderic’s side of the family,
and the necessarily more traditional leanings of Gelimer and his support-
ers. Hilderic may not have planned to hand his kingdom over to Justin-
ian, but he may well have intended the kingship to pass to his immediate
family after his death, rather than see the rule pass to a distant side of the
family. This is entirely conjectural, but does allow the climactic crisis of
the Vandal kingdom to be viewed from a different perspective. What may
well have been at stake, then, was not the constitutio Geiserici, but pre-
cisely the form of dynastic rivalry that Geiseric’s law was intended to
avoid.

Hilderic’s Romanitas

The most striking evidence for Hilderic’s dynastic pretensions comes from
the Latin Anthology, an eclectic compilation of Latin poetry which was
probably compiled in North Africa during the early years of the Byzantine
occupation. Among the texts therein are the two verses on Hilderic’s
palace. The better known of the two was composed by Luxorius:

The remarkable edifice of King Hilderic gleams, erected with skill,
toil, talent, riches, wealth. From it the sun itself takes its rays to pass
on here. Another day can be believed

<to dwell> in the marble. Here

the firm pavement seems like thick snow spread about. When your feet
stand upon it, you would think they could sink into it.

Proc. BV I.

9.8, ed. Dewing, pp. 86–8.

Proc. BV I.

9.21, ed. Dewing, p. 88.

Anthologia Latina

194, ed. D.R. Shackleton Bailey, Teubner (Stuttgart, 1982), I, p. 145: ‘Hildirici

regis fulget mirabile factum | arte, opere, ingenio, divitiis pretio. | hinc radios sol ipse capit, quos
huc dare possit; | altera marmoribus creditur esse dies. | hic sine labe solum nix † iuncta et †
sparsa putatur; | dum steterint, credas mergere posse pedes.’

152

A.H. Merrills

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

The second is an anonymous ekphrasis dedicated to the king and

written in celebration of the grand wall decorations within the same
palace:

Powerful Vandal-king, inheritor of a twinned diadem,
You adorned your own through your great deeds!
Theodosius the avenger suppressed the aggression of armies,
Capturing the barbarians, in an easy contest.
Honorius subjected the enemy with arms of peace
his better prosperity conquered.
The courage of Valentinian, known to the world in its greatness
by the enemies reduced to his mercy, is displayed through the art of his
grandson.

Both of these poems belong to a tradition of self-consciously classiciz-

ing verse common to the collection as a whole.

Among the better

known examples from the Anthology are Florentinus’ poems in honour of
Thrasamund’s Carthage, Luxorius’ description of the tower of the Vandal
noble Fridamal, and a consolatio on the death of Hoageis’ young daugh-
ter.

The celebration of Hasding building projects is also a common

enough theme: five short poems were composed in honour of Thrasam-
und’s construction of an elaborate bath complex at Alianae.

Little is known of Hilderic’s palace foundation at Anclae beyond what

appears in these poems. According to Procopius, ‘Aclas’ was a suburb of
Carthage, which Belisarius occupied at the time of Gelimer’s surrender,
presumably taking residence in the royal buildings.

The neighbourhood

– and hence the palace – may well have taken its name from a large
waterwheel, known as the ‘ancla’ and celebrated in another poem of the
Latin Anthology.

Similar devices are known to have been erected by

206, ed. Shackleton Bailey, I, p. 154: ‘Vandalirice potens, gemini diadematis heres, | ornasti

<ni>ani virtus cognita mundo | hostibus addictis ostenditur arte

nepotis.’ On this poem, see Michel Chalon, Georges Devallet, Paul Force, Michel Griffe,
Jean-Marie Lassere and Jean-Noël Michaud, ‘Memorabile factum: Une célébration de
l’evergétisme des rois vandales dans l’Anthologie Latine’, Antiquités africaines

21 (1985), pp.

207–62, at pp. 242–7.

On the political and cultural significance of the collection, see Chalon et al., ‘Memorabile
factum’; R.T. Miles, ‘The Anthologia Latina and the Creation of Secular Space in Vandal
Carthage’, Antiquité Tardive

13 (2005), pp. 305–20 and the references therein.

371, ed. Shackleton Bailey, I, pp. 286–8 (Carthage); AL 299, ed. Shackleton Bailey, I, pp.

246–7 (Fridamal); AL 340, ed. Shackleton Bailey, I, pp. 269–70 (Hoageis).

201–5, ed. Shackleton Bailey, I, pp. 150–3. On these poems, see Chalon et al., ‘Memorabile

factum’, pp.

226–41.

Proc. BV II.

7.13, ed. Dewing, p. 268.

278, ed. Shackleton Bailey, I, p. 201. Cf. Chalon et al., ‘Memorabile factum’, pp. 243–4.

Dynasty and crisis in Vandal North Africa

153

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

royal initiative elsewhere in the Vandal kingdom, but the precise circum-
stances of this particular project are unknown.

What is certain is that the

Anclae foundation was distinct from the royal palace complex on Byrsa
Hill, which was the principal seat of Hasding power.

Hilderic’s development of the complex at Anclae need not be read as

a conscious rejection of the Byrsa Hill complex – many of his predeces-
sors are known to have spent time on their own estates, rather than in the
capital.

What is notable, however, is the precise form of wall decoration

described within the anonymous poem. Here, despite the peculiar for-
mulation of the opening apostrophe to the Vandaliricus potens – ‘powerful
Vandal-King’ – the imagery is resolutely imperial. The poet describes
images which celebrate first Theodosius’ victories over the gentes, then
Honorius’ diplomatic success, and finally Valentinian’s virtus. The link
between the Vandal subject of the poem and the imperial images on the
palaces is made explicit in the last couplet: these are the triumphs of
Hilderic’s ancestors, his grandfather (Valentinian III), and two earlier
scions of the Theodosian house, Honorius and Theodosius I. When
Hilderic is addressed as the inheritor of the ‘twinned diadem’, then, it is
for his unification of these two royal lines, not for the hoary claims to rule
over the ‘Vandals and Alans’ made by his father, and later revived under
Gelimer.

This reinvention of royal genealogy was nothing new, of course. When

Dracontius celebrated Gunthamund’s heritage in his long Satisfactio, he
had the tact and good sense to omit Huneric from his encomium on
Gunthamund’s royal line.

Other Vandal kings made little obvious effort

to link their own rule to the prehistoric past – a conceit which may have
suited the tastes and ingenuities of Byzantine antiquarians better than the

382, ed. Shackleton Bailey, I, p. 295, is a verse by Cato on a coclea, or Archimedes screw

erected by Huneric. On this poem, see E. Malaspina, ‘L’idrovora di Unirico. Un epigramma
(A.L.

387 R. = 382 Sh.B). e il suo contesto storico-culturale’, Romanobarbarica 13 (1994–1995),

pp.

43–56.

Vict. Vit. HP I.

17, ed. Lancel, p. 104 (Geiseric at Maxula); AL 371, ed. Shackleton Bailey, I, pp.

286–8, AL 201–5, ed. Shackleton Bailey, I, pp. 150–3 (Thrasamund at Alianae); AL 194, ed.

Shackleton Bailey, I, p.

145 (Hilderic at Anclae); Proc. BV I.14.10, ed. Dewing, p. 128 (Gelimer

at Hermiana). Further foundations (perhaps royal) are alluded to at BV I.

17.8, ed. Dewing, p.

150 (Grasse) and AL 286, ed. Shackleton Bailey, I, pp. 238–9 (an unknown estate with a fish

pond).

Contra Gautier, Genséric, p.

229. The title rex Vandalorum et Alanorum is attested in Vict. Vit.

HP II.

39, ed. Lancel, p. 139 and III.3, ed. Lancel, p. 175 and also appears on a silver salver (as

CIL VIII.

17412). On the title, see Clover, ‘Timekeeping’, p. 47 and esp. A. Gillett, ‘Was

Ethnicity Politicized in the Earliest Medieval Kingdoms?’, in Andrew Gillett (ed.), On Barbar-
ian Identity. Critical Approaches to Ethnicity in the Early Middle Ages (Leiden,

2002), pp. 85–121,

at pp.

109–10.

Dracontius, Satisfactio

49–52, ed. and trans. C. Moussy, Dracontius Oeuvres II, Budé (Paris,

1988), pp. 178–9. On which see Merrills, ‘The Perils of Panegyric’, pp. 157–8. It is surprising how

frequently Dracontius is cited as evidence for the solidity of the Hasding lineage, rather than its
flexibility. See, for example, Gil Egea, África en Tiempos de los Vándalos, p.

313.

154

A.H. Merrills

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

barbarian kings of the period – but this failure may itself be illustrative.

The Vandals betrayed little interest in the history of their campaigns in
Gaul and Spain in the first three decades of the fifth century, and none at
all in the period before that; it seems they were happy to eschew their
claims to distant origins in favour of celebrating their position as the
rightful children of Carthage.

Hilderic’s decision to recast his own

authority in an imperial guise was an entirely understandable action in
the light of his descent from Huneric and Eudocia.

But the golden light of this newly important Roman heritage only fell

on half of the Hasding family. At its centre, of course, was Hilderic
himself, Vandalrice potens, and first-born offspring of a Hasding king and
a Theodosian queen. Alongside him stood his own children and grand-
children, but also his two nephews, Hoamer and Hoageis. Procopius
leaves little doubt that these brothers were two of the major heroes of the
late Vandal state. Hoamer he boldly nominates as the ‘Achilles of the
Vandals’, while Hoageis is celebrated in Luxorius’ poems as a dominus
(‘Lord’: a title reserved for those of the royal house), and the father of the
‘royal child’ Damira.

As the children of Hilderic’s sibling, they too

could claim inheritance from both royal bloodlines and evidently were
celebrities of no small stature within the kingdom. Among those
excluded from this celebration, however, were the four sons of Geilarith:
Tzazo, Gunthimer, Ammatas and of course Gelimer, who could make no
such claim to imperial ancestry. As the Hasding house folded in on itself,
Hilderic’s propaganda may have indicated the new direction he wished
the Vandal monarchy to take.

Hilderic’s celebration of his own Theodosian lineage, and the implied

recasting of Hasding power in the light of this heritage had still wider
implications for the nature of Vandal identity. Crucially, the celebration
of Theodosius, Honorius and Valentinian as Hilderic’s ancestors not only
eclipsed Geiseric’s position at the ideological heart of Vandal kingship, it
also sought to obscure the firmly masculine grounding of Hasding legiti-
macy that the agnatic law of succession had established. Hasding rule was

See W. Goffart, ‘Does the Distant Past Impinge on the Invasion Age Germans?’, in Andrew
Gillett (ed.), On Barbarian Identity. Critical Approaches to Ethnicity in the Early Middle Ages,
Studies in the Early Middle Ages

4 (Turnhout, 2002), pp. 21–37, for a typically provocative

discussion of the relevance (or otherwise) of historical traditions within the successor states.

The Vandals may not have been greatly interested in their history, but they did not ignore it
entirely. Procopius alludes to Vandal sources on events in Gaul and Spain at BV I.

3.33, ed.

Dewing, p.

32 and I.22.1, ed. Dewing, p. 184 but none of these seems to have been written, and

all are predictably ambiguous. BV I.

3.1–2, ed. Dewing, p. 22 and I.22.1–10, ed. Dewing, pp.

184–6 may be regarded as Procopius’ own interpretation of the Vandal past. On the Hasdings

as the children of Carthage, see Florentinus in AL

371, ed. Shackleton Bailey, I, pp. 286–8, and

the discussion in F.M. Clover, ‘Felix Carthago’, Dumbarton Oaks Papers

40 (1986), pp. 1–16.

Proc. BV I.

9.2, ed. Dewing, p. 82; AL 364, ed. Shackleton Bailey, I, p. 283; AL 340, ed.

Shackleton Bailey, I, pp.

269–70.

Dynasty and crisis in Vandal North Africa

155

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

limited to men, of course (itself scarcely an unusual situation in the
successor kingdoms), but the erasure of women within the ruling family
went still deeper. The only royal women we hear of in the entire history
of the Vandal kingdom are unnamed, with the exception of the infant
Damira and the princesses Eudocia and Amalafrida, both of whom came
from outside the kingdom.

Women had no place within Geiseric’s law

of succession but were central to Hilderic’s proposed redefinition of
Hasding authority. The role of gender within the formation of Vandal
identity is a subject that deserves extensive consideration, not least
because of the famous role of the ‘Vandal women’ in the political and
economic disputes that followed the Byzantine conquest.

At the very

least, however, Hilderic’s emphasis upon his matrilineal origins must be
regarded as a major conceptual shift within the ideological politics of the
last years of the kingdom.

Gelimer’s tyranny

Gelimer’s behaviour following his usurpation supports this image of a
divided Hasding house. When he established his junta after the coup, all
of his most important lieutenants came from his own household. Both of
his brothers were given central positions of responsibility within the
military. Ammatas supervised the execution of Hilderic and the family
when news of Belisarius’ approach reached Carthage; he was later killed
resisting the Byzantines at the climactic battle at Ad Decimum.

Gelimer

initially entrusted the government of Sardinia to his domesticus Godas,
who was clearly an important member of the household if not a blood
relation.

When Godas himself rebelled, the king turned to another

brother Tzazo to recover the island and the troops thereon. Tzazo later
returned to Gelimer and joined him in their final stand.

We also hear of

Gelimer’s appointment of two further relatives, Gibamundus and
Gunthimer, to military posts.

These were either his cousins or (more

probably) his nephews by Tzazo or Ammatas.

The most telling detail comes with Gelimer’s treatment of Hilderic’s

circle. The king himself was imprisoned of course, and was later executed
as Belisarius’ troops approached Carthage. Hoageis suffered the same

For Damira, see n.

67 above.

On the centrality of gender to the formation of political identities within this period, see
J.M.H. Smith, Europe after Rome. A New Cultural History

500–1000 (Oxford, 2004), esp. pp.

115–50; and Halsall, Barbarian Migrations, pp. 96–9 and 482–8. The importance of gender to

Vandal identity is discussed in some detail in Merrills and Miles, The Vandals, pp.

106–8.

Proc. BV I.

17.11–12, ed. Dewing, p. 152.

Proc. BV I.

10.25–34, ed. Dewing, pp. 98–100.

Proc. BV I.

11.23, ed. Dewing, p. 106; I.25.19–26, ed. Dewing, p. 204.

Proc. BV I.

18.1, ed. Dewing, p. 154; Vict. Tun., a. 534, ed. Placanica, pp. 38–40.

156

A.H. Merrills

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

punishment, but it is the blinding of Hoamer which deserves particular
comment.

This episode is without parallel in the kingdom, and the

absence of extant Vandal law codes makes it impossible to infer the crime
from its punishment. But blinding occurs frequently enough as a symbol
of irrational tyranny in contemporary western texts to encourage the
suspicion that Procopius included the information simply as an illustra-
tion of Gelimer’s barbaric tyranny.

But if Hoamer actually was blinded

(and the episode was not an invention of the Byzantine historian), it is
worth noting that the vast majority of known instances of this punish-
ment, from early Byzantium down to the fate of Gloucester in King Lear,
are linked to political disputes, and specifically to the punishment of
lèse-majesté. Procopius himself cites an example from Persia, which notes
that blinding rendered an individual ineligible for royal succession in the
kingdom.

According to the mid-seventh-century law of Chindaswinth

in Visigothic Spain, capital punishment for treachery against the crown
might be commuted to blinding and the loss of property at the discretion
of the king.

Further examples abound from the early eighth century, in

both Byzantium and Carolingian Frankia. Through the act of blinding
pretenders to the throne, the king or emperor expressed his ineffable
authority within the state.

These comparanda provide nothing more than circumstantial evi-

dence: the direct identification of blinding as a punishment for lèse-
majesté can only be attested with confidence in the eastern Mediterranean
in the sixth century, and in the west from the middle of the seventh.
Consequently, Gelimer’s punishment of Hoamer in the

530s represents

something of an outlier. But the new king’s insistence that his revolt
against Hilderic was motivated by the latter’s betrayal of Geiseric’s con-

Proc. BV I.

9.14, ed. Dewing, p. 88. On blinding as a punishment, see J. Lascaratos and S.

Marketos, ‘The Penalty of Blinding during Byzantine Times’, Documenta Ophthalmologica

(

1992), pp. 133–44 on Byzantium; and G. Bührer-Thierry, ‘ “Just Anger” or “Vengeful Anger”?

The Punishment of Blinding in the Early Medieval West’, in Barbara H. Rosenwein (ed.),
Anger’s Past. The Social Uses of an Emotion in the Middle Ages (Ithaca,

1998), pp. 75–91 on the

western tradition.

Bührer-Thierry, ‘Just Anger’, esp. pp.

75–8. Typified by the description of Chilperic in Gregory

of Tours, Decem Libri Historiarum VI.

46, ed. B. Krusch, MGH SRM I (Hanover, 1937), pp.

319–21.

Procopius, De Bello Persico I.

11.3–4, ed. and trans. H.B. Dewing, Procopius. History of the Wars

Books I and II, Loeb (Cambridge, MA,

1914), p. 82; BP I.23.4, ed. Dewing, pp. 208–10; BP

II.

9.12, ed. Dewing, p. 340. And cf. other examples in Lascaratos and Marketos, ‘The Penalty

of Blinding’, pp.

133–5.

Lex Wisigothorum II.i.

8, ed. K. Zeumer, MGH Leges nat Germ. 1 (Hanover, 1902), p. 55. See

Bührer-Thierry, ‘Just Anger’, p.

78.

Expressed neatly by Bührer-Thierry, ‘Just Anger’, p.

91: ‘when the emperor blinded someone

who had attacked his ministerium in the hope of usurping or tarnishing it, he did not abuse his
power nor commit an arbitrary act inspired by uncontrollable anger. He acted within a clearly
defined framework: if not a law code, at least a system of references and ideas that recognized
his monopoly on this particular form of violence.’

Dynasty and crisis in Vandal North Africa

157

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

stitutio, may provide a context for his actions. A fear of disinheritance in
favour of Hoamer would explain Gelimer’s otherwise perplexing decision
to rebel, his policies upon seizing the throne and the justification of these
actions to the Byzantine court. Hilderic’s obvious attempts to reposition
the Hasding monarchy along pseudo-imperial lines may be regarded as
an early stage in the same process.

The final piece of evidence in support of this interpretation comes

from a surprising direction. The constitutio Genserici was evidently a
subject of relatively common knowledge in Justinianic Constantinople.
The emperor was able to invoke Geiseric’s ruling in his letters of appeal
to Gelimer, and the law was sufficiently familiar for both Procopius and
Jordanes to provide competent summaries in the

550s. But why was this

law so familiar? And why did Justinian turn to it so readily in his
exchange with Gelimer? The usurper was not, after all, acting in direct
contravention of Geiseric’s law of succession in seizing the throne for
himself. He was behaving illegally, certainly, and might justifiably be
labelled a tyrant (as he was), but if Gelimer was to be condemned,
Geiseric’s law provided poor grounds on which to do so. While Justinian
may simply have wished to appeal to the long-standing Vandal reverence
for their founding (great grand-) father, the specific legal allusion remains
peculiar. It is possible, however, that it was Hilderic, and not Gelimer,
who had first brought the law to the attention of his allies in Constan-
tinople. Any attempt to rework the succession would have necessitated an
engagement of some sort with the hoary traditions of the Hasding house.
When this failed – thanks to Gelimer’s timely usurpation – the ancient
law of Geiseric had little legal relevance, it simply provided further a
cudgel with which Hilderic’s Byzantine allies could lambast the new
tyrant. Gelimer was not specifically acting in breach of his great-
grandfather’s testamentum, it just made sense for his political opponents
to state that he was.

Conclusions

Geiseric’s constitutio had an important role to play within the articulation
of royal power in the early years of the Vandal kingdom. It took its place
alongside a variety of social and economic innovations which both
created a strong political core to the new kingdom and firmly distin-
guished the ruling Hasding family from rival groups among the Vandal
(and Romano-African) aristocracy. The subsequent prominence of the
law in the writing of Victor of Vita, Procopius and Jordanes indicates that
Geiseric’s innovation retained an ideological authority over the remain-
der of the Vandal century, but this should not obscure the complex means
by which royal power was constantly renegotiated over this period. Two

158

A.H. Merrills

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

major succession crises within the Vandal kingdom illustrate this particu-
larly effectively, but Huneric’s attempt to rework the succession in favour
of his own son, and Gelimer’s later usurpation, were not the only
examples of Hasding power straying beyond the careful guidelines laid
down by Geiseric. Close investigation of each case suggests that Huneric
and Gelimer were themselves responding to direct institutional chal-
lenges to established systems of royal power: in one case in the person of
a young royal pretender, in the other through a recrafting of the Hasding
line as descendants of the house of Theodosius.

University of Leicester

Dynasty and crisis in Vandal North Africa

159

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Churches and aristocracies in

seventh-century Spain: some thoughts

on the debate on Visigothic churches

emed_294

160..174

lexandra Chavarri´a Arnau

After a decade of intense debate on the chronology of the so-called Visigothic
churches in Spain the divergent opinions have made little progress owing to
the lack of reliable chronologies and the limits of a very stylistic approach. This
article aims to present an analysis of the social and economic conditions of the
seventh century (particularly the second half ) as a background to understand-
ing and defending the existence of church building in late Visigothic Spain.
Their physical type (monumental but not enormous), the quality of their
construction techniques and decoration, their function (absence of baptisteries
and, often, presence of privileged burials) suggest aristocratic patronage of
these buildings, an identification widely supported by textual and epigraphi-
cal evidence.

Visigothic or Ummayad? Methodological problems

The debate surrounding the chronology and patronage of church build-
ing is one of the most hotly contested topics in current archaeological
research on early medieval Spain. Until now, the study of early medieval
churches in Spain has been approached through a number of fundamen-
tally different methodologies. Probably the most revolutionary new
approach has been via the stylistic analysis of ornamental motifs, a
method that led Sally Garen to re-date the church of S. Maria de Melque
(Toledo) – formerly considered to be of seventh-century Visigothic con-
struction – to the period after the Arab conquest of

711, on the basis of

comparison with the decoration at Khirbet al Mafjar in Palestine.

In this

M. Alba, G.P. Brogiolo, W. Liebeschuetz, A. Vigil Escalera and C. Wickham read, commented
and provided helpful suggestions on the draft of this article. Tamara Lewit and Yuri Marano
helped me with its translation.

S. Garen, ‘Santa Maria de Melque and Church Construction under Muslim Rule’, Journal of the
Society of Architectural Historians

51 (1992), pp. 288–305.

Early Medieval Europe

2010 18 (2) 160–174

2010 Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350

Main Street, Malden, MA

02148, USA

way, Garen triggered a revolution in the field of Visigothic studies,
fostering a profound revision of the chronology of early medieval material
culture in Spain.

Her thesis was widely accepted among archaeologists

and art historians, not least by the excavator of Melque, Luis Caballero
Zoreda, who moved from a defence of the traditional Visigothic period
dating of the church of Melque to a chronology after

711, in the light of

further reassessment of the architecture and decoration of the building.

The new dating of S. Maria de Melque precipitated a revision of the
chronology of other churches which had been traditionally dated to the
second half of the seventh century, but which are now commonly attrib-
uted to the eighth and ninth centuries.

This revision encouraged the development in Spain of the archaeology

of standing buildings, and was applied by Luis Caballero to some of the
contested churches. However, although the stratigraphic analysis of the
masonry of these churches has improved our understanding of the dif-
ferent phases of construction, it has not produced absolute chronologies.
Furthermore, the archaeology of buildings in Spain has not been followed
by the elaboration of interpretative models regarding the construction of
these buildings. A deeper understanding of construction materials
(whether new or reused), the organization and composition of the work-
shops involved in building them, and of the construction cycles would
help us to understand their patrons and the historical context in which
they were built.

Radiocarbon analysis has also proved inconclusive regarding the

dating. Samples taken from the floors at Melque produced a date between

660 and 770. Dendrochronological and radiocarbon analysis of wooden
beams and joists from S. Pedro de la Nave and S. Juan de Baños
place them within a chronological range too broad to assign them to
either the Visigothic or Ummayad period.

The surviving epigraphic

The first scholars to date these churches to the Mozarabic period were M. Gómez Moreno,
Iglesias mozárabes (Madrid,

1919); J. Camón Aznar, ‘Arquitectura prerrománica española’, in XVI

Congrès International d’Histoire de l’Art (Lisbon,

1950), pp. 105–23; J. Camón Aznar, ‘La

arquitectura mozárabe del s. X y de repoblación’, Goya

52 (1963), pp. 206–19; J. Puig i Cadafach,

L’Art wisigothique et ses survivances (Paris,

1961). Other scholars, however, dated them to the

seventh century: H. Schlunk and T. Hauschild, Denkmäler de frühchristlichen und westgotischen
Zeit (Mainz am Rhein,

1978); P. de Palol, Arqueología Cristiana de la España Romana (siglos IV

al VI) (Madrid and Valladolid,

1967).

See, in synthesis (with earlier bibliography): L. Caballero-Zoreda, ‘La arquitectura denominada
de época visigoda ¿es realmente tardo romana o prerrománica?’, in L. Caballero and P. Mateos
(eds), Visigodos y Omeyas. Un debate entre la Antigüedad tardía y la alta Edad Media, Anejos de
AEspA

23 (Madrid, 2000), pp. 207–49. Also L. Caballero (ed.), La iglesia de San Pedro de la

Nave (Zamora) (Zamora,

2004). Many relevant contributions to the discussion were published

in Visigodos y Omeyas. Un debate entre la Antigüedad tardía y la alta Edad Media (Madrid,

2000).

F. Alonso Matthías, E. Rodríguez Trobajo and A. Rubinos Pérez, ‘Datación de madera con-
structiva en San Pedro de la Nave (Zamora) y su interdatación con San Juan de Baños
(Palencia)’, in Caballero, La iglesia de San Pedro de la Nave, pp.

209–37.

Churches and aristocracies in seventh-century Spain

161

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

documentation of some churches, such as the inscription of Mijangos
(Burgos) or that of the church of S. Pedro de la Mata (Toledo), has also
been scrutinized.

Recent studies have even raised doubts about the

authenticity of the famous inscription of Reccesuint in S. Juan de Baños
on the basis of its stratigraphic position in the wall and of written sources
that seem to identify an identical inscription in the church of S. Roman
de Hornija (Valladolid).

Unfortunately, most significant buildings were excavated before the

widespread implementation of modern methodologies and dating tech-
niques (e.g. at S. Juan de Baños, S. Pedro de la Nave, Quintanilla de las
Viñas) and even in the case of churches which have been the subject of
recent research (S. Maria de Melque, S. Lucia del Trampal), their foun-
dation dates have not been precisely determined because of the problems
of distinguishing material culture dating from the seventh century from
that produced in the eighth.

Furthermore, the excavations have generally

concentrated on the inside or immediately surrounding area of the build-
ings, without extensive research into the settlements which were con-
nected to these churches, and which would be of primary importance to
understanding the function and chronology of the whole architectural
complex.

However, even if the precise dating of the artistic culture traditionally

considered typical of the seventh century is a difficult task, we cannot
deny that the church found in the vicinity of Zorita de los Canes (a
settlement identified with the city of Reccopolis) was certainly a work of
the Visigothic period, as the city was founded ex novo around the end of
the sixth century and was destroyed at the end of the eighth century, as
archaeological research has shown.

The same is true of the ecclesiastical

J.A. Lecanda, ‘Mijangos: la aportación de la epigrafía y el análisis arqueológico al conocimiento
de la transición a la alta Edad Media en Castilla’, in Visigodos y Omeyas, pp.

181–206. Caballero

considers it a tenth-century copy, in M. de los A. Utrero, Iglesias tardoantiguas y altomedievales
en la Península Ibérica. Análisis arqueológico y sistemas de abovedamiento, Anejos de AEspA

(Madrid,

2006), pp. 504–5.

I. Velázquez Soriano and R. Hernando, ‘Una notícia desconcertante sobre la inscripción de San
Juan de Baños ofrecida por Álvar Gómez de Castro’, Archivo Español de Arqueología

73 (2000),

pp.

295–307.

And this despite the identification in recent years of archaeological contexts securely dated to
the seventh century, which have permitted the identification of datable pottery typologies. See
on this problem, E. Manzano, ‘La cerámica de los siglos oscuros’, in L. Caballero Zoreda, P.
Mateos and M. Retuerce (eds), Cerámicas tardorromanas y altomedievales en la Península Ibérica.
Ruptura y Continuidad, Anejos de AEspA

28 (Mérida, 2003), pp. 541–57, at pp. 549, 552–5.

L. Olmo, ‘Ciudad y procesos de transformación social entre los siglos VI y X: de Recópolis a
Racupel’, in Visigodos y Omeyas, pp.

385–99; L. Olmo, ‘Transformaciones de un paisaje urbano.

Las últimas aportaciones de Reccópolis’, in Actas del I Simposio de Arqueología de Guadalajara
(Sigüenza,

2000), pp. 545–55; L. Olmo, ‘The Royal Foundation of Recópolis and the Urban

Renewal in Iberia during the Second Half of the Sixth Century’, in J. Henning, Post-Roman
Towns, Trade and Settlement in Europe and Byzantium. Vol.

1. The Heirs of the Roman West (Berlin

and New York,

2007), pp. 181–96.

162

Alexandra Chavarría Arnau

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

complex at El Tolmo de Minateda (Hellín, Albacete), where destruction
and change of function is closely dated to the mid-eighth century.

According to the scholars who support the revised dating of these

buildings (from the seventh to the eighth and ninth centuries), the
political context of the seventh century could not have supported the
vigorous social and cultural climate necessary for the construction of a
sophisticated architecture since it was marked by a deep crisis in the
Visigothic state and by incessant revolts and uprising fomented by the
aristocracy. According to this model, the seventh century is depicted as a
period of regression in building techniques, when simpler structures were
built or pre-existing ones reused.

Only around the end of the seventh

century, or more plausibly, during the eighth, do these scholars see the
signs of a renewal that led ultimately to the architectures of the ninth
century. This renewal is thought to be related to the arrival of Muslim
armies that laid the basis of new social and economic developments
supporting the artistic activity of Christian communities gathered around
monasteries.

Several dissenting voices have been raised against this chronological

shift, notably that of Achim Arbeiter, who has defended the existence of
architectural and sculptural production in the kingdom of Toledo during
the seventh century and who has pointed out how the ornamental
elements considered by the exponents of the ‘Umayyad or Mozarabic
model’ could be seen instead to have originated in the milieu of the
Byzantine culture which dominated the Mediterranean during the
seventh century.

After a decade and more of intense debate, during which important

colloquia have been organized (Visigodos y Omeyas published in

2000,

Escultura decorativa tardorromana y altomedieval en la Peninsula ibérica
published in

2007, and another on the seventh century published in

2009), the divergent opinions have made little progress. This article aims
to present an analysis of the social and economic conditions of the
seventh century (particularly the second half ) as a background to under-
standing and defending the existence of church building in late Visig-
othic Spain.

L. Abad Casal, S. Gutiérrez Lloret and B. Gamo Parras, ‘La basílica y el batisterio del Tolmo de
Minateda (Hellín, Albacete)’, Archivo Español de Arqueología

73 (2000), pp. 193–221; S. Gutiér-

rez Lloret, ‘Algunas consideraciones sobre la cultura material de epoca visigoda y emiral en el
territorio de Tudmir’, in Visigodos y Omeyas, pp.

95–116.

M.L. Real, ‘Portugal: cultura visigoda y cultura moçárabe’, in Visigodos y Omeyas, pp.

21–75, at

34: ‘A sétima centúria parece constituir un período de recessão constructiva, caminhando-se

para soluções relativamente simples, de cunho local.’

A. Arbeiter, ‘Construcciones con sillares. El paulatino resurgimiento de una técnica edilicia en
la Lusitania visigoda’, in IV Reunió d’Arqueologia Cristiana Hispánica (Lisboa,

1992) (Barcelona,

1995), pp. 211–22; idem, ‘Alegato por la riqueza del inventari monumental hispanovisigodo’, in

Visigodos y Omeyas, pp.

249–63.

Churches and aristocracies in seventh-century Spain

163

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

The first rural churches in Hispania

To judge by the available archaeological evidence, the Christianization of
the countryside in Hispania took place at a relatively late date. In com-
parison with other areas such as Italy, where, in the fifth century, several
churches with baptismal structures are documented along the main axes
of communication in villages and in fortified settlements, in the Iberian
Peninsula Christian buildings in rural areas are rare before the sixth
century.

Even though many archaeologists assume that churches built

within late Roman villas point to the early Christianization of the
Spanish countryside, archaeology shows that these churches were built
long after the abandonment of the residential buildings. Connecting
sixth- or seventh-century (or even later) churches to the initiative of a late
Roman proprietor whose villa was abandoned in the fifth century
remains an unproven hypothesis.

Obviously these data could be a simple consequence of the limits of

our knowledge but nonetheless they fit well with the situation docu-
mented by the archaeological research on the Spanish countryside in the
fifth century: a general abandonment of the villas as a consequence of the
crisis and disappearance of traditional elites, a process which coincided
with political instability in some areas and the later arrival of the Suevi,
the Vandals, the Alans, and the first Visigothic incursions.

The first known rural churches date from c.

500 AD onwards.

Among

the most significant examples (although not the only ones), are several
buildings in the vicinity of Mérida (Augusta Emerita) one of the most
important cities in late antique Hispania.

Churches of this date, such as

the sites El Gatillo (Caceres), at Valdecedabar (Badajoz) or Ibahernando

See, as the latest synthesis on this subject, G.P. Brogiolo and A. Chavarría, ‘Chiese, territorio e
dinamiche del popolamento nelle campagne tra Tardoantico e Altomedioevo’, Hortus Artium
Medievalium

14 (2008), pp. 7–29.

A critical analysis of the archaeological evidence in A. Chavarría, El final de las villae en Hispania
(siglos IV–VII), Bibliothèque de l’Antiquité Tardive

7 (Turnhout, 2007), pp. 192–5 and eadem,

‘Aristocracias tardoantiguas y cristianización del territorio (siglos IV–V) ¿otro mito historiográ-
fico?’, Rivista di Archeologia Cristiana

82 (2006), pp. 201–30. Another opinion is provided by K.

Bowes in ‘Un coterie espagnole pieuse: Christian Archaeology and Christian Communities in
Fourth and Fifth Century Hispania’, in K. Bowes and M. Kulikowski (eds), Hispania in Late
Antiquity. Current Perspectives, The Medieval and Early Modern Iberian World (Leiden,

2005),

pp.

189–258; and also in Private Worship, Public Values and Religious Change in Late Antiquity

(Cambridge,

2008).

A defence of more accentuated continuity is expressed by J. Arce, Bárbaros y Romanos en
Hispania:

400–507 d.C. (Madrid, 2005), an exhaustive analysis of the political events of the fifth

century.

For a recent critical catalogue of late antique and early medieval churches in the Iberian
Peninsula, see Utrero, Iglesias tardantiguas y altomedievales.

On these buildings, see P. Mateos and L. Caballero-Zoreda (eds), Repertorio de Arquitectura
Cristiana en Extremadura: época tardoantigua y altomedieval, Anejos de AEspA

29 (Madrid,

2003).

164

Alexandra Chavarría Arnau

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

(Cáceres), are characterized by their small size and the presence of burials,
suggesting a possible function as private churches or oratoria with a
funerary function. At Casa Herrera (Badajoz), by contrast, the larger size
and more complex architecture and liturgical organization of the church
(with two apses and a corridor that connected them), as well as the
presence of a larger cemetery, allow us to hypothesize the existence of a
vicus, or the possibility that this building was frequented by the scattered
population of the territory.

The vitality of the Iberian Peninsula during the sixth century has been

emphasized by several archaeologists, who date to this period the monu-
mentalization of urban centres and, in particular, of the episcopal areas of
the towns along the Mediterranean coasts (Barcelona, Valencia, Carta-
gena) as well as in the interior.

Particularly interesting is the case of

Mérida, the prosperity of which is well known from the Vitae Sanctorum
Patrum Emeritensium – information confirmed by archaeological
research.

It is possible that the erection of the rural churches in the

course of the sixth century reflects the reorganization and/or renewal of
the urban elite and of their properties in the countryside, as well as the
first attempts by the church to provide and organize the pastoral care of
rural communities.

Around the end of the sixth century or the beginning of the seventh,

some of the churches mentioned above (Casa Herrera for example) were
enlarged and provided with baptismal fonts. It is tempting to link the
proliferation of baptismal churches and the monumentality of some
baptisteries to the conversion of the Visigoths to Catholicism decreed in

589 at the Third Council of Toledo. In this context, baptism could
represent an important symbol of the public exhibition of the victory of
the Catholic faith and of the unity of the kingdom.

The Visigothic aristocracies and their churches: textual and

epigraphic evidence

Textual and epigraphic evidence can provide important evidence relating
to building activity during the seventh century.

As early as the sixth

century the ecclesiastical authorities in the Spanish church discussed

For a general synthesis with previous bibliography see the works collected in J.M. Gurt and A.
Ribera (eds), Les ciutats tardoantigues d’Hispania: cristianització i topografia, VI Reunió
d’Arqueologia Cristiana Hispànica (Valencia

2003) (Barcelona, 2005). See also the interesting

article by L. Olmo, ‘The Royal Foundation of Recópolis’.

Vitas Sanctorum Patrum Emeretensium, ed. A. Maya, CCL

116 (Turnhout, 1992).

On the political significance of the Third Council of Toledo, see S. Castellanos, La hagiografía
visigoda. Dominio social y proyección cultural (Logroño,

2004), pp. 205–17.

Contra, Real, ‘Portugal: cultura visigoda y cultura moçárabe’, p.

34: ‘São mínimas as alusões à

actividade construtiva durante este período ao contrário do que sucede para a centúria anterior.’

Churches and aristocracies in seventh-century Spain

165

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

private churches, condemning those proprietors who, having built their
churches, tried to avoid episcopal control.

In the third canon of the First

Council of Ilerda (

546) the founders were accused of conceiving their

churches ‘sub specie monasterii’, surely a strategy to elude economic
submission to the bishop because the monasteries were independent and
immune from the payment of tertia.

Some decades later, the sixth canon

of the Second Council of Braccara (

572) condemned founders who built

‘oratoria non pro devotione fidei sed pro questu cupiditas’ to share with
the clergy the oblations of the faithful.

Moreover, the canon criticized

the bishops for demanding compensation to consecrate private churches,
warning them not to consecrate these buildings without being sure that
there was a sufficient endowment to maintain the clergy and lighting.

The Third Council of Toledo in

589 reorganized the relationship

between the aristocracies and the church, accentuating the need for
submission of the nobility and the private churches to ecclesiastical
authority. Several canons dealt with the organization of the property and
revenues of rural churches, insisting on absolute ecclesiastical jurisdiction
and refusing the lay proprietors any rights over their foundations.

However, in the decades after the Third Council of Toledo, noble

authority over these foundations increased, perhaps as a consequence of
the strengthening of their estates and their power. This development
coincides with the progressive weakening of the Visigothic monarchy, as
revealed by the sequence of usurpations and rebellions that continued
down to the reign of Chindasuint (

642–53).

This process can be also

The bibliography on the private churches (Eigenkirchen), not intended as churches built with
private funds but possessing a private administration, is ample. See now on this subject the
outstanding work by S. Wood, The Proprietary Church in the Medieval West (Oxford,

2006).

Spanish bibliography has a wide tradition: M. Torres, ‘El origen del sistema de iglesias propias’,
Anuario de Historia del Derecho Español

3 (1926), pp. 307–475; R. Bidagor, La iglesia propia en

España. Estudio histórico canónico (Rome,

1933); I. Arenillas, ‘La autobiografía de San Valerio

(siglo VII) como fuente para el conocimiento de la organización eclesiástica visigoda’, Anuario
de Historia del Derecho Español

11 (1934), pp. 468–78; P.C. Díaz, ‘Iglesia propia y gran propiedad

en la autobiografía de Valerio del Bierzo’, I Congreso internacional Astorga romana (Astorga,

1986), pp. 297–303. For arguments against the existence of this type of institution in Spain, see

G. Martínez Díez, El patrimonio eclesiástico de la España visigoda (Madrid,

1957), in particular

pp.

71–8.

The tertia referred to the third part of revenues of the church administered by the bishop and
destined for the restoration of buildings: see Martínez Díez, El patrimonio eclesiástico, pp.

83–94.

Concilios visigóticos e hispano-romanos, ed. J. Vives (Barcelona and Madrid,

1963), p. 83.

In particular the canons

4, 15, 19, 20 concerning the organization of the patrimony and the

revenues of the rural churches; canon

19 insisting on the ecclesiastic property of the possessions

of the rural churches; canon

33 refusing every right of the proprietors on their foundations: La

Colección Canónica Hispánica V, ed. G. Martínez Díez, F. Rodríguez (Madrid,

1992), pp. 103–32.

The weakening of the monarchy can be observed, for example, in the several councils assembled
in the seventh century by the various kings who used the ecclesiastical institution to legitimize
their power and, in some cases, to fight against the aristocracy, as done by King Chindasuint.
See, on the relation between kings and councils, A. Ziegler, Church and State in Visigothic Spain
(Washington DC,

1939), pp. 86–126. On the anti-aristocratic policy of Chindasuint, see D.

166

Alexandra Chavarría Arnau

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

traced in the Spanish councils and, in particular, in the legislation con-
cerning private churches which increasingly came to suit the interests of
the founders, that is the lay aristocracy (‘uiri illustres e omnes senories
gothorum’), whose presence in the councils became compulsory and
authoritative after

653, as demonstrated by the fact that they subscribed

to the acts alongside the bishops and the priests.

In the Fourth Council

of Toledo, assembled by King Sisenand in

633, the noble founders were

safeguarded against the abuses of the bishops (canon

33);

the thirty-

eighth canon stated that the church should sustain the founders and their

Claude, Adel, Kirche und Konigtum im Westgotenrreich (Sigmaringen,

1971), pp. 115–31. For

another opinion, see Wickham, Framing the Early Middle Ages, p.

93, according to whom ‘there

is in fact reasonable evidence of increases both in royal ambition and royal control’.

On the presence and function of nobles at the councils see E.A. Thompson, Los godos en España
(Madrid,

1990), p. 335.

La Colección Canónica Hispánica V, pp.

220–1.

Fig.

1 Locations of the late antique churches quoted in the text: a. Casa Herrera;

b. El Gatillo; c. Ibahernando; d. Valdecedabar and locations of the Visigothic
churches quoted in the text:

1. S. Maria de Melque; 2. S. Pedro de la Nave; 3. S.

Juan de Baños;

4. S. Maria de Mijangos; 5. S. Pedro de la Mata; 6. Quintanilla de

las Vinas;

7. S. Roman de Hornija; 8. Las Tapias at Abelda de Iregua; 9. San

Vicente del Valle;

10. S. Lucia del Trampal

Churches and aristocracies in seventh-century Spain

167

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

successors in times of economic troubles;

the fifty-first canon assured

the economic independence of the monasteries, even though they
remained under ecclesiastical jurisdiction.

This trend was continued by

other canons of the Ninth Council of Toledo, assembled by King Rec-
cesuinth in

655, which allowed the founders of churches and monasteries

suffering from avaricious bishops to manage their church during their
own lifetimes (canon

and to choose their priests (canon

2).

This

legislation supporting the private churches was subsequently reinforced
by King Wamba.

In practice, these decrees recognized the two basic

elements of the institution of private churches: administrative and pas-
toral control by their founders.

Councils reflect the increasing power of

the nobility, whose new status and position was demonstrated through
the foundation of churches and monasteries.

Apart from the documents already analysed, hagiographic texts also

refer to the possession and building of churches by members of the
aristocracy.

One of the most significant is the Life of Valerius of Bierzo

(composed c.

678).

It mentions one Ricimerus, a member of the rural

aristocracy at the time of Wamba, who founded a church in the fundus
Ebronanto and nominated a man named Valerius as a priest of the
building. However, Ricimierus died when the church collapsed before its
completion, and his successors decided to appoint as a presbyter another
man, by the name of Justus.

There are also several inscriptions referring to this type of building

activity in the seventh century.

Important examples are: the famous text

recording the three churches built by Gundiliuva, c.

600;

that of the

inlustis femina Anduires, who supported, along with her husband, the

La Colección Canónica Hispánica V, p.

224.

La Colección Canónica Hispánica V, pp.

230–1.

La Colección Canónica Hispánica V, pp.

492–3.

La Colección Canónica Hispánica V, pp.

494–5.

Leges Visigothorum IV.v.

6 in MGH Legum, Sectio I.1, ed. K. Zeumer (Hanover and Leipzig,

1902; repr. 1973), repeated in the Eleventh Council of Toledo, canon 5: La Colección Canónica

Hispánica VI, ed. G. Martínez Díez and F. Rodríguez (Madrid,

2002), pp. 105–10.

Wood, The Proprietary Church, pp.

11–16.

Castellanos, La hagiografía visigoda, pp.

146–62.

Ordo querimoniae

5–6, Valerio del Bierzo. An Ascetic of the late Visigothic Period, trans. C.M.

Aherne (Washington DC,

1949), pp. 80–93).

Some reflections on these documents in P. Castillo Maldonado, ‘Pro amore dei: donantes y
constructores en la provincia Baetica tardoantigua (testimonias literarios y epigráficos)’, Antiq-
uité Tardive

13 (2005), pp. 335–7 and I. Velázquez, ‘Basilicas multas Miro opere construxit (VSPE

51.1). El valor de las fuentes literarias y epigráficas sobre la edilicia religiosa en la Hispania

visigoda’, Hortus Artium Medievalium

13.2 (2007), pp. 261–9.

For different interpretations of this inscription, see Y. Duval, ‘Nativola-les-Trois-Églises (Éveché
d’Acci,

594–607) d’après Vives ICERV 303’, Mélanges de l’Ecole Française a Rome. Antiquité

103.2 (1991), pp. 807–20; A. Canto, ‘N. 121. Inscripción conmemorativa de tres iglesias’, Arte

Islamico en Granada. Propuesta para un Museo de La Alhambra (Granada,

2005), pp. 343–6;

Velázquez, Basilicas multas Miro opere construxit, p.

262.

168

Alexandra Chavarría Arnau

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

construction of a church;

the inscription of

622 by Belesarius mention-

ing a funerary basilica;

the inscription concerning the Church of Santa

Maria de Cisimbrus (Zambra) founded by three members of a local family
and consecrated by Bishop Bacauda of Egabrum in

660.

In the latter

case, as with other inscriptions of the seventh century (found largely in
the south of the Iberian Peninsula), the text was inscribed on the base of
an altar and commemorated, alongside the date of consecration and the
name of the bishop appointed by this ritual, the relics that had been
deposited in the church.

In terms of the construction of churches in the

countryside, this evidence points to the seventh century as a period of
vigorous building activity.

Rural churches in Hispania in the seventh century: function

and context

We have already noted the famous inscription in the church of S. Juan de
Baños which refers to the construction of the building by King Recce-
suint and the identical epigraphic documentation in the church of
S. Roman de Hornija (Valladolid).

The most plausible hypothesis for

this odd duplication is that the scholar who recorded it, A. Gómez de
Castro, made a mistake when he recorded the location. But it must be
remembered that, according to the continuatio of the Historia Gothorum,
King Chindasuinth, father of Reccesuint, founded a monastery at S.
Roman de Hornija as a burial place for himself and for Reciberga, the
wife of his son. Bishop Eugenius II of Toledo also recalls in one of his
poems the construction of a private funerary church built by his brother
Evantius, according to a custom typical of the members of the aristocracy
of the time who were rich enough to finance such works.

In fact, there are burials in many of the churches cited, often monu-

mental, but unfortunately seldom described in detail, so that their

J. Vives, Iscripciones romanas de la España romana y visigoda, Monumenta Hispaniae sacra. Serie
Patrística II (Barcelona,

1969), no. 505, pp. 82–3. An analysis of the documents and its

archaeological context in C. García Merino, Historia de Uxama Argaela (Soria,

2001), pp. 82–3

who dates the inscription to the late sixth century.

Corpus Inscriptionum Latinarum II

5, 482; Inscriptiones Hispaniae Latinae, ed. altera, pars. VII.

Conventus Cordubensis, (Berlin and New York,

1995).

J. Vives, Iscripciones romanas de la España romana y visigoda, no.

308, p. 104.

On the inscriptions of these altars, see Y. Duval, ‘Projet d’enquête sur l’épigraphie martyriale en
Espagne romaine, visigothique (et byzantine)’, Antiquité Tardive

1 (1993), pp. 173–206. On the

altars themselves, see G. Ripoll and A. Chavarría, ‘El altar en Hispania. Siglos IV–X’, Hortus
Artium Medievalium

11 (2005), pp. 29–48, at pp. 30–1.

An analysis of this document is in I. Velázquez Soriano and R. Hernando, ‘Una notícia
desconcertante sobre la inscripción de San Juan de Baños ofrecida por Álvar Gómez de Castro’,
Archivo Español de Arqueología

73 (2000), pp. 295–307.

Carmen XXIX, v.

3, ed. F. Vollmer, MGH AA 14, p. 25: ‘hanc in honore dei supplex Evantius

aulam/ sacram fabricans hanc in honore dei/hic patrios cineres praeciso marmori clausi’.

Churches and aristocracies in seventh-century Spain

169

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

location and chronology cannot be easily established and it is not possible
to say if they were prepared for the founders of the buildings or if they
should be attributed to a significantly later period. At Melque (Toledo) a
monumental arcosolium tomb is located in the southern wing of the
transept;

at Mijangos (Burgos) several sarcophagi have been identified

inside the building;

at Quintanilla de las Viñas (Burgos) a tomb with

grave-goods was discovered in an annex at the lower end of the nave.

this last case, the grave-goods (possibly a liturgical jar) suggest that the
deceased may have had ecclesiastical status. In the church at Las Tapias at
Abelda de Iregua (Logroño), there is a privileged burial in the portico,

while at San Vicente del Valle (Burgos) a tomb was located near the
entrance on the axis of the altar.

The phenomenon is mirrored in other European areas,

particularly in

Lombard Italy, where sources attest to attempts by local aristocracies to
obtain a privileged burial place for the founders and their families
through the foundation of churches and monasteries.

These religious

buildings were also significant landmarks in the topography of the coun-
tryside, and played an important role in defining a new geography of
power. Archaeology in northern Italy has shown the wide diffusion
of these private funerary churches during the seventh and the eighth
centuries.

L. Caballero Zoreda and J.I. Latorre Macarrón, La iglesia y el monasterio visigodo de Santa María
de Melque (Toledo). Arqueología y Arquitectura. San Pedro de la Mata (Toledo) y Santa Comba de
Bande (Orense), Excavaciones Arqueológicas en España

109 (Madrid, 1980).

J.A. Lecanda, ‘Santa María de Mijangos: de la arquitectura paleocristiana a la altomedieval,
transformaciones arquitectónicas y litúrgicas’, in Arqueologia da antiguedade na Península Ibérica

6 (Porto, 2000), pp. 535–54.

On the tomb and its grave-goods, see F. Iñiguez Almech, ‘Algunos problemas de las viejas
iglesias españolas’, Cuadernos de trabajos de la Escuela Española de historia y Arqueología de Roma

7 (1955), p. 89 (lámina 108).

U. Espinosa, ‘La iglesia de Las Tapias (Albelda) en la arquitectura religiosa rural de época
visigoda’, III Semana de Estudios Medievales (Nájera,

1992) (Logroño, 1993), pp. 267–73.

J.A. Aparicio Bastardo and A. de la Fuente, ‘Estudio arqueológico e interpretación arqui-
tectónica de la iglesia de La Asunción en San Vicente del Valle (Burgos)’, Numantia (

1996), pp.

158–71. A later chronology for this building has been proposed in L. Caballero, F. Arce and M.

de los A. Utrero, ‘Santa María de los Arcos de Tricio (La Rioja), Santa Coloma (La Rioja) y La
Asunción en San Vicente del Valle (Burgos). Tres miembros de una familia arquitectónica’,
Arqueología de la Arquitectura

2 (2003), pp. 81–5.

See, for example, K.H. Krüger, Königsgrabkirchen des Franken, Angelsachsen und Langobarden bis
zur Mitte des

8. Jahrhunderts: ein historischer Katalog, Münstersche Mittelalter Schriften 4

(Munich,

1971); H.W. Böhme, ‘Adel und Kirche bei den Alamannen der Merowingerzeit’,

Germania

74.2 (1996), pp. 477–507; E. Hassenpflug, Das Laienbegräbnis in der Kirche.

Historisch-archäologisches Studein zu Alemannien in frühen Mittelalter, Freiburger Beiträge zur
Archäologie und Geschichte des ersten Jahrtausends (Rahden-Westf,

1999).

C. La Rocca, ‘Le aristocrazie e le loro chiese tra VIII e IX secolo in Italia settentrionale’, in R.
Salvarani, C. Andenna and G.P. Brogiolo (eds), Alle origini del Romanico, Atti delle III Giornate
di Studi medioevali, Castiglione delle Stiviere

2003 (Brescia, 2005), pp. 59–67.

G.P. Brogiolo, ‘Oratori funerari tra VII e VIII secolo nelle campagne transpadane’, Hortus
Artium Medievalium

8 (2002), pp. 8–31.

170

Alexandra Chavarría Arnau

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

We do not know anything about the wider settlement context of

Spanish churches. Common opinion is that they were annexed to mon-
asteries.

The presence of buildings in the territory around these churches

is certain and, in some cases – as at Melque or at Guarrazar – sample
excavations have revealed the presence of important structures.

However, without extensive archaeological investigations it is not possible
to identify their specific function. Considering that texts often refer to
private churches, it would not be surprising if these churches were part of
residential complexes, at least in some places.

Another interesting point is the presence of these churches not only

in the territory of the capital Toledo, and in the territories of the main
cities of the Visigothic kingdom, but also within the northern sector of
the Meseta. Elsewhere I have argued that, in late antiquity, this area was
dominated by monumental villas located in enormous landed proper-
ties belonging to the fourth- and fifth-century aristocracy, and that in
some cases (as suggested by the extraordinary villa of La Olmeda, for
example) they might even have been imperial lands.

I have also pro-

posed that these properties may have passed wholesale into Visigothic
hands early in the sixth century.

In addition to the archaeological

testimony of the so-called Visigothic necropolises (probably identifiable
with the Visigothic and local peasant population established at these
properties), the written sources contain references to the interests of the
aristocracy in these areas, such as the funerary church built by Chin-
dasuint at Valladolid, that of San Juan de Baños or the uillula of Ger-
ticos, where Reccesuinth died and where Wamba acceded to the
throne.

L. Caballero Zoreda, ‘Monasterios visigodos. Evidencias arqueológicas’, Codex Aquilariensis

(

1988), pp. 31–50.

C. Eger, ‘Guarrazar (Provinz Toledo). Bericht zu den Untersuchungen

2002 bis 2005’, Madrider

Mitteilungen

48 (2007), pp. 267–305.

A. Chavarría, El final de las villas en Hispania (siglos IV–VII) (Turnhout,

2007) and A.

Chavarría, ‘Villae tardoantiguas en el valle del Duero’, in S. Castellanos and I. Martin Viso
(eds), De Roma a los bárbaros. Poder central y horizontes locales en la cuenca del Duero (León,

2008), pp. 93–122.

A. Chavarría, ‘Dopo la fine delle ville: le campagne ispaniche in epoca visigota’, in G.P.
Brogiolo, A. Chavarría and M. Valenti (eds), Dopo la fine delle ville: Le campagne tra VI e IX
secolo,

10 Seminario sul tardo antico e l’alto medioevo (Gavi 8–10 maggio 2004), Documenti di

Archeologia

39 (Mantua, 2005), pp. 263–85 and eadem, ‘Romanos y visigodos en el valle del

Duero (siglos V–VIII)’, Lancia

6 (2004–2005), pp. 191–209.

Julián de Toledo, Historia Wambae III.

37–39: ‘Gerebantur enim ista in uillula, cui antiquitas

Gerticos nomen dedit, quae fere centum uiginti milibus ab urbe regia distans in Salamanricensi
territorio sita est’, CCSL

115 (Turnhout, 1976).

Churches and aristocracies in seventh-century Spain

171

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Visigothic churches after the Arab conquest?

The Visigothic population accepted the new political order imposed after
the Arab conquest in

711, surviving it without immediate consequences.

The written sources cite some Visigothic aristocrats making their appear-
ance at Arab courts, such as Kings Akhila and Ardo in the north-east of
the peninsula, Oppa at Toledo, Teudemir and his son Athanagildus in the
east. The latter were renowned for their estates and for their power, which
they maintained after agreements with the Arabs. An uncertain but
nonetheless extremely suggestive link could be made between this
Teudemir and the residential building of Pla de Nadal (located near
Valencia), where a monogram and a graffito with the name Tebdemir
have been found.

In this case not only architectural decoration (more

than

800 fragments of the type found in churches) but also ceramics seem

to point to a Visigothic chronology for this building,

a rare case of rural

elite residential architecture after the end of the late antique villas.

In some cases, this continuity implied the conversion to Islam, but in

other cases political change did not represent a profound break in the
lifestyle of the Christian population, which maintained its religious iden-
tity during the first centuries of Muslim dominance (Christians became
known as Mozarabes). It is very probable that the Christians continued
building and rebuilding their cult places. In this respect, the dating of the
churches of Santa Maria de Melque or Santa Lucia del Trampal (for
example) before or after

711 is of little significance, if we identify them as

private churches or monasteries which were part of aristocratic properties.

The continuity of the social order of the Visigothic period in the

decades following the Arab conquest should be underlined. The year

711

marked little more than a step in a long process. Rather, the disappear-
ance of the Visigothic aristocracy should be placed around the ninth
century, in the age of the Emirate, when Abd-Al Rahaman II established
firmer control over the country, bringing about fiscal reforms that trig-
gered discontent among the rural aristocracy and revolts in the towns as
well as in the countryside. This Ummayad centralization laid the basis of
a new Islamic social order and decreed the end of Visigothic elites.

From an archaeological perspective, this transformation process is

detectable from the mid-eighth century. Of particular interest is the

On the Arab settlement and its impact on the local society, see E. Manzano, Conquistadores,
emires y califas. Los Omeyas en la fundación de Al-Andalus (Barcelona,

2006).

For a recent synthesis of this monument with earlier bibliography, E. Juan, ‘Pla de Nadal: una
villa aulica di epoca visigota’, in G.P. Brogiolo and A. Chavarría (eds), I Longobardi. Dalla
caduta dell’Impero all’alba dell’Italia, Catalogo della mostra Torino-Novalesa

28 Settembre

2007–6 Gennaio 2008, Cinisello Balsamo (Milan, 2007), pp. 146–7.

Gutiérrez Lloret, ‘Algunas consideraciones sobre la cultura material de epoca visigoda’, pp.

102–5.

172

Alexandra Chavarría Arnau

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

situation attested in the countryside around Madrid, that is the north-
eastern sector of the territory of the capital, Toledo. From the sixth
century, this area was characterized by the presence of large nucleated
settlements (such as Gozquez, for example), with sunken huts, structures
built on stone bases, and pits. Alfonso Vigil, who has excavated and
studied these villages, writes of an economic organization probably
related to an elite, perhaps based in the towns.

These settlements date to

the first half of the sixth century and ceased functioning probably around
the mid-eighth century, revealing some kind of profound changes in the
organization of the countryside. As far as the towns are concerned, at the
site identified with Reccopolis, founded at the end of the sixth century,
the crisis of the Visigothic state may be detected in the dissolution of its
urban structures (disarticulation of the urban grid, reuse of the main
buildings), a tendency that accelerated in the course of the eighth
century. Archaeology has revealed a phase of substantial destruction in
the late eighth and ninth centuries.

In another urban settlement of the

Visigothic period, El Tolmo de Minateda, similar traces of destruction
and a radical change in the urban structure are dated to around the end
of the eighth or beginning of the ninth century.

At Mérida, a new urban

and architectural framework has been detected in several parts of the
town, with the consolidation during the ninth century of new monu-
mental building structures radically different from previous ones. These
palaces have been interpreted as evidence of the presence of a new Arab
urban elite.

Conclusions

From an archaeological point of view, the dating of the so-called ‘Visig-
othic churches’ represents a difficult task, owing to the lack of reliable
chronologies and the limits of a very stylistic approach. The aristocratic
patronage of these buildings is suggested by the location of these churches

A. Vigil Escalera, ‘Los poblados de época visigoda del Sur de Madrid: algunos aspectos
económicos y sociales’, in I Congreso del Estudios Históricos del Sur de Madrid ‘Jiménez de
Gregorio’ (Alcorcón,

2003), pp. 51–8; A. Vigil Escalera, ‘El modelo de poblamiento rural de la

Meseta y algunas cuestiones de visibilidad arqueológica’, in J. López Quiroga, A.M. Martínez
Tejera and J. Morín de Pablos (eds), Gallia e Hispania en el contexto de la presencia ‘germánica’
(ss.V–VII). Balance y perspectivas (Oxford,

2006), pp. 89–108.

Olmo, ‘Ciudad y procesos de transformación’, pp.

390–4.

S. Gutiérrez Lloret, ‘Ciudades y conquista. El fin de las ciuitates visigota y la génesis de las
mudun islámicas del sureste del al-Andalus’, in Genèse de la ville islamique en al-Andalus e tau
Magreb occidental (Madrid,

1998), pp. 137–57.

M. Alba, ‘Diacronía de la vivenda señorial de Emerita (Lusitania, Hispania): desde las domus
alto imperiales y tardoantiguas a las residencias palaciales omega (siglos I–IX)’, in G.P. Brogiolo
and A. Chavarría (ed.), Archeologia e società tra tardo antico e alto medioevo (Mantua,

2007), pp.

163–92, with references to previous works.

Churches and aristocracies in seventh-century Spain

173

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

(in the Meseta and in the territories of the main cities of the period), their
physical type (monumental in scale but not enormous in size), the quality
of their construction techniques and decoration, their function (absence
of baptisteries and, often, presence of privileged burials), as well as direct
evidence via inscriptions or monograms, and wider textual and epi-
graphic evidence. Future research must address the identification and
excavation of the residential and productive annexes attached to these
churches, in order to reconstruct the structure and organization of the
territory of which they were a part. Such investigations will provide
important data not just concerning their chronology but also their func-
tion and significance, enabling these churches to operate as historical
documents in their own right.

Università degli Studi di Padova

174

Alexandra Chavarría Arnau

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Farms and families in

ninth-century Provence

emed_295

175..201

osamond Faith

Information about the people recorded in

813–14 on the estate of St Victor de

Marseilles shows that although considered to belong to the monastery they
were an independent peasant class. Family size and structure varied: some
farms were run by the labour of the family which included unmarried sons
and ‘married-in’ sons-in-law; other farmers employed living-in servants in
husbandry. The mountain sheep farms had large groups of unmarried young
people. Inheritance systems ensured that the peasant family property remained
intact over the generations and provided support for unmarried sons who
remained to work there.

Peasants’ lives are seldom recorded, and when they are it is in the
language of their social superiors. An important source for early medi-
eval rural society are the polyptyques of Frankish monasteries. These
documents list in detail the possessions of the monastery, the status and
obligations of the people living on its estates and the size of their ten-
ancies. Since the nineteenth century a vast body of scholarship has been
devoted to editing and analysing the polyptyques with a view to estab-
lishing their dating and authorship, and the rationale of their produc-
tion and revision.

More recently they have been the basis for enquiry

into the nature of rural society with a strong emphasis on the evolution
and nature of le régime domainal, the bipartite system of the seigneurial

I am grateful to Chris Wickham and members of seminars in Oxford, London and Cambridge
for comments on an earlier draft of this paper, and to Mark Whittow for his company and
insights during an exploration of some of the St Victor properties. David Austin, Andrew
Fleming and David Siddle, colleagues on the ‘Cipières Project’, have given me an unforgettable
introduction to an upland Provençal landscape, its history, its demography and its archaeology.

The work of Y. Morimoto in this field is collected in Y. Morimoto, Etudes sur l’économie rurale
du haut Moyen Age: historiographie, régime domainal, polyptyqes carolingiens (Brussels,

2008); of

J.-P. Devroey in J.-P. Devroey, Etudes sur le grand domaine carolingien (Aldershot,

1993); of A.

Verhulst in A. Verhulst, Rural and Urban Aspects of Early Medieval Northwest Europe (Aldershot,

1992).

Early Medieval Europe

2010 18 (2) 175–201

2010 Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350

Main Street, Malden, MA

02148, USA

estate, divided between demesne land worked for the direct support of
the landowner, and tenancies. It was the contribution of A. Verhulst
that showed that this system could not be taken as typical of the Frank-
ish estate as a whole, but was a phenomenon of ‘Between Loire and
Rhine’: studies of the south and south-west have showed a very differ-
ent kind of relationship between landowner and peasant.

The infor-

mation the polyptyques give about the monastery’s tenants, their legal
status and their obligations, has informed debate about the degree to
which they were free, and in what proportions. It has been the basis of
demographic enquiries into gender balance, size of families, and the
direction of population change, and used as an indicator of economic
growth or decline.

Verhulst was exceptional in using the vocabulary of

the Flemish polyptyqes to provide valuable information about the
differences between husbandry practice on the demesne and on tenant
land, but by and large the economy of the peasant farm and the rela-
tionship between family and land, topics at the heart of rural society,
have played very little part in polyptyque studies.

This may in part be

to do with language. The compilers of the polyptyques had at their
disposal a suite of Latin terms to describe landed estates and the
peasants who lived on them. Many of these words were an inheritance
from the Roman world, both terms for land (villa, latifundium,
ager, colonica, mansus) and for people (mancipia, liber, ingenuus, servus,
colonus). These terms brought with them strong connotations of abso-
lute ownership on the part of landowners, and of servility and depen-
dence on the part of the majority of people who lived on their estates.
They have often been accepted as unproblematically descriptive, yet
when we are able to learn something about the realities of particular
farms and families these realities turn out to be at odds with the lan-
guage of the documents. We are able to confront lives and language in
exceptional detail in the records of a great estate in early ninth-century
Provence, that of St Victor de Marseilles. St Victor was already by then
a great landlord. Its records use the traditional vocabulary to describe
the people on its lands and the lands themselves. There is no doubt
that the abbey considered itself the owner of both land and people. Yet
the family strategies that the peasant farmers on its estate evolved to

J.-P. Poly, La Provence et la Société Féodale,

876–1166 (Paris, 1976); with E. Bournazel, The Feudal

Transformation

900–1200 (New York and London, 1991); P. Bonnassie, From Slavery to Feudalism

in South-Western Europe (Cambridge,

1991).

C. Hammer Jnr, ‘Family and familia in Early-Medieval Bavaria’, in R. Wall (ed.), Family Forms
in Historic Europe (Cambridge,

1983), pp. 217–48; S. Weinburger, ‘La transformation de la

société paysanne en Provence médiévale’, Annales

45.1 (1990), pp. 3–21.

M. Bloch, French Rural Society: An Essay on its Basic Characteristics, trans. J. Sondheimer
(Berkeley and Los Angeles,

1996). G. Duby, Rural Economy and Country Life in the Medieval

West, trans. C. Postan (London,

1968).

176

Rosamond Faith

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

meet the demands of the environment in which they made a living,
and the control they thereby exercised over their own lives and labour,
appear to have secured them a degree of personal autonomy and per-
manent possession of their farms which belie St Victor’s claims and the
language used to assert them.

The estate

The abbey of St Victor was the successor at Marseilles to the cenobitic
community, one of two – the other was for women – which had been
established there under the inspiration of John Cassian early in the fifth
century. By the time that we have any documented knowledge of it, St
Victor had become a landholder on a large scale, thanks to the donations
of the Provençal aristocracy and very likely from lesser people too. Its
properties in the early ninth century all fell within what had been the
Roman dioceses of Provincia narbonensis and Alpes maritimae and within
what are now the départements of Bouches du Rhône, Vaucluse, Var, and
Alpes de Haute Provence (Fig.

1). St Victor at this time had very little

property near to Marseilles itself: the spread of its possessions reflects the
landholdings and power bases of the local aristocratic families who were
its benefactors. Accessible by sea along the coast from Italy, and down the
Rhone, Provence was particularly vulnerable to invasion: by the Visigoths
in the

480s, then by the Ostrogoths and Burgundians, who settled a

considerable area of western Provence. It became an annex of the
kingdom of the Franks in

537. Saracen raids in the eighth century were

repulsed by Charles Martel who thereby established a strong power base
for the Frankish kings in the region, represented by their candidates for
the office of patricius. The abbey had a constant struggle to retain its hold
on its property in the face of depredations from local magnates, including
the families of the original donors. It was not until the tenth century that
its fortunes began to revive.

It is St Victor’s landholdings, donations and

redonations, and the associated disputes, that gave rise to the only early
documentation we have: the charters and records of lawsuits, the earliest
of the seventh century, collected in the ‘cartulary of St Victor’ in the
twelfth century and edited in the nineteenth by Benjamin Guerard
(referred to here as CSV). Among these is the document known as the

H. Labaude, ‘L’église de Marseille et l’abbaye de S. Victor à l’époque carolingienne’, in Mélanges
Offerts à F. Lot par ses Amis et ses Elèves (Paris,

1925); E. Baratier, ‘La formation et l’étendue du

temporel de l’abbaye de S. Victor’, Provence Historique

16, no. 65 (1965), pp. 395–441;

F. Ganshof, ‘Les avatars d’un domaine de l’église de Marseille à la fin du VII

au VIII

siècle’,

in Studi in Onore di Gino Luzzato, vol.

1 (Milan, 1950), pp. 50–66.

Farms and families in ninth-century Provence

177

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

ig.

roper

ties

ictor

813

–14

178

Rosamond Faith

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

polyptyque drawn up in

813 and 814 (referred to here as PSV).

PSV has

been much studied for the wealth of demographic information it con-
tains and for its value as a collection of personal names from a social level
rarely accessible to historians. More recently its potential as a resource for
investigating the history of the landscape has begun to be appreciated.

The immediate occasion of the compilation of the polyptyque was

probably the decision of the Council of Arles in

813 to order the resto-

ration of the abbey’s property.

It was drawn up on the orders of the

bishop, Wadalde, and some of the properties are under the name of the
church of Marseilles, not the abbey. The whole document took two years
to compile – quite possibly a halt was called to the work in the winter
months. Of the polyptyques of northern Francia it resembles that of St
Germain des Prés in recording the tenants and their children by name. It
does not give the extent of tenants’ holdings or list the obligations due
from them. In its format and possibly also in its intended use it bears a
closer resemblance to the lists of slaves and tenants on the Italian abbeys
of Farfa and Bobbio which have recently been analysed by Laurent Feller.

It consists of a series of entries, in most cases headed by the phrase
discriptio mancipiorum de, ‘register of the slaves of ’, followed by the name
of a villa, e.g. villa Domada, villa Bettorida, or in two cases an ager.

The loose term ‘property’ is employed here for both villa and ager, but

it is important to clear away their classical connotations. Provence had
been a rich Roman province, with all that that implies in terms of villas,
roads and towns. Most of the St Victor properties were near former
Roman centres of some type and all could be reached by a Roman road,
although we do not know anything about the state of repair of these in
the ninth century.

Powerful individuals like the benefactors of St Victor

Cartulaire de l’Abbaye de S. Victor de Marseille, ed. B. Guerard,

2 vols, Collection des documents

inédits sur l’histoire de France: Collection des cartulaires de France

8, 9 (Paris, 1857). The

polyptyque was edited by L. Delisle and A. Marion as an appendix to the cartulary at pp.

633–54

of volume

9. The cartulary is cited here as CSV and individual charters by the numbers given

by the editors. J.-P. Bregi has prepared a new edition of the polyptyque as Le Polyptyque de
l’abbaye de Marseille: essai d’une édition, but as this is unobtainable in the UK I cite the Delisle
and Marion edition as PSV and follow their practice of identifying the main properties of the
estate by a letter and the individual colonicae and bercariae within them by a number.

J.-P. Bregi, ‘Recherches sur la démographie rurale et les structures sociales au IX

siècle’, Ph.D.

thesis, University of Paris (

1975), pp. 11–15. I am extremely grateful to Professor Y. Morimoto for

kindly lending me his copy of this thesis.

Laurent Feller, ‘Liberté et servitude en Italie centrale (VII

–VIII

siècle)’, in Les Formes de la

Servitude de la Fin de l’Antiquité au Monde Moderne, Mélanges de l’Ecole Française de Rome:
Moyen Age

112 (2000), pp. 511–33. I am very grateful to Marios Costambeys for bringing this

article to my attention. It is discussed at p.

22 below.

For continuities or discontinuities in the meaning of these terms: S. Applebaum, ‘The Late
Gallo-Roman Rural Pattern in the Light of the Carolingian Capitularies’, Latomus

23 (1964),

pp.

774–87; J. Percival, ‘The Late Roman Villa-System: A Reply’, Latomus 25 (1966), pp. 134–8.

For the term villa in the early Middle Ages: A. Lognon, La Géographie de Gaule au VI

siècle

(Paris,

1878); F. Bange, ‘L’ager et la villa: structures de paysage et du peuplement dans la région

Farms and families in ninth-century Provence

179

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

still dominated the countryside but the relationship of landowners to the
rural population was no longer what it had been in the heyday of the
Roman landowning class. While the villa as a domestic building was
abandoned, the term villa remained in use to describe a territory con-
taining a range of farms and settlements.

Intensive exploitation was

limited to a few ‘headquarters’ visited by their itinerant owners. The

739

testament of Abbo, patricius of Provence, shows that he kept five of his
properties as his cortes or headquarters, worked on a bipartite manorial
system and heavily dependent on slave labour, while the freedmen-
farmers on the rest of his vast property were tied to him only by the
payment of dues owed since antiquity from their dispersed farms. Simi-
larly, on the lands of Count Leibulf in the region of Arles in the ninth
century, ‘there is . . . neither demesne nor tenures, but land which is the
master’s and houses where he stays when he wishes’.

From its nature as

the record of a particular landowner our document will never give a
complete picture of an area: it does not record farms which were either
allodial properties or belonged to another major landowner. Nevertheless,
the entire estate was scattered over a very wide area and covered a wide
range of types of territory. None of the properties was near Marseilles
itself and the furthest was

110 km away.

Properties recorded in PSV

Below is a list of properties recorded in PSV arranged according to title of
property as given in the polyptyque, and followed by the present-day
commune, arrondissement and département. Some of these identifications
are disputed, notably villa Bettorida (Fourcalquier), differently located by
Bregi, Baratier and the editors of the polyptyque.

villa Nono seu et Campania
Gardanne, Aix-en-Provence, Bouches du Rhône

villa Domada
Endoume, Marseilles, Bouches du Rhône

mâconnaise à la fin du Haut Moyen Age (IX

–XI

siècles)’, Annales

39.3 (1984), pp. 529–69;

Recueil des chartes de l’Abbaye de la Grasse, ed. E. Magnou-Nortier and A.-M. Magnou (

1996),

II, pp.

779–1119 and p. xxxix. N.D. Fustel de Coulanges, Histoire des Institutions Politiques de

l’Ancienne France,

6 vols (1889), I. L’Alleu et le domaine rurale, pp. 208–9.

Bange, ‘L’ager et la villa’; F. Favory, F. and J.L. Fiches (eds), Les Campagnes de la France
Méditerranée dans l’Antiquité au Haut Moyen Age (Paris,

1994).

P. Geary, Aristocracy in Provence: The Rhone Basin at the Dawn of the Carolingian Age (Phila-
delphia,

1985); J.-P. Poly, La Provence et la Société Féodale, 876–1166 (Paris, 1976), p. 104.

E. Sauze, ‘Le polyptyque de Wadalde: problèmes de toponomie et de topographie provençale au
IX

siècle’, Provence Historique

34 (1984), pp. 3–33.

180

Rosamond Faith

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

villa Lambiscum
Lambesc, Aix-en-Provence, Bouches du Rhône

villa Bedada
Rognes, Aix-en-Provence, Bouches du Rhône

villa Marciana
Ansouis, Apt, Vaucluse

villa Bettorida
Fourcalquier, Alpes de Haute Provence

villa vel ager Sinaca
Ciounne, Castellane, Alpes de Haute Provence

ager Caladius
La Javie, Digne, Alpes de Haute Provence

villa Virgonis
Annot, Castellane, Alpes de Haute Provence

villa Tregentia
Trigance, Draguignan, Var

villa Bergemulum
Bargemon, Draguignan, Var

villa Rovagonis
Rougon, Digne, Basses-Alpes

ager Cilianus
Seillans, Draguignan, Var

The farms

These properties were themselves collections of between three and fifty-
four farms. Under each villa or ager heading follows a list of colonice and
their names. Colonica is here translated as ‘farm’, although as will appear,
this term covers agrarian units of very different types. In contrast to the
areal measures given for peasant holdings in the northern polyptyques, in
PSV there is no indication of their size. As well as the farms described as
colonice there were a few described as vercarie (bercariae), translated here as
sheep farms. Most of these were in the high Alpine properties of ager
Caladius (Chaudol) and villa Rovagonis (Rougon). Following the name of
each farm are the names of people. In total some nine thousand individuals
are described in this way, and others are referred to but not named. Nearly
always a man comes first, with a description of his status: colonus and
mancipium are the terms most commonly used, but accola is also found. In
classical Latin these terms meant respectively a farmer, a slave and an
inhabitant or neighbour, but as we shall see they need to be considered in
the context of the people they were used to describe and the document in
which they occur. Next listed is a man’s wife, of whom it is noted if she is
an ‘outsider’, extranea. (The meaning of this term is discussed below).Then

Farms and families in ninth-century Provence

181

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

follows the name, sex and age of each of their children over the age of two,
under twos being described as ad uber, at the breast. Son, daughter, wife
and husband are the only relational terms employed. Spouses are recorded
by name or simply as extraneus/a. Girls are as well represented as boys. At
some point between the ages of twelve and fifteen the unmarried children
were no longer recorded as filius or filia, but as baccalarius or baccalaria.
Baccalarius is a medieval Latin term for an unmarried man, our modern
word bachelor derives from it. A list of dues follows the entry for each
colonica, irrespective of the number of people listed: census paid in kind,
tributum in cash and pascuum in stock (all these terms are discussed below).

Although thanks to the work of French scholars virtually all the

properties of the estate can on the evidence of place names be associated
with present-day communes or centres, it is by no means certain what
was the relationship between the collection of farms which comprised the
ninth-century property, the later nucleated settlement, and present-day
settlement. But on the assumption that each property roughly corre-
sponded to, was contained within, or contained the commune with
which it can be associated, some idea of the physical environment of the
farms, their terroir, can be obtained. Nearly all were in, or comprised,
upland or mountain valleys. They correspond very well to Sclafert’s
description of Haute Provence as a multiplicity of little regions, of small
neighbourhoods centred on more or less enclosed valleys.

Two French

landscape historians have identified the probable locations of several of
the colonicae listed for two of the properties and their work has made it
much more possible to place some of the families in PSV in their physical
setting. Marijo Chiche-Aubrun has located, with varying degrees of
certainty, the fourteen colonicae listed under Seillans (ager cilianus) and
her topographical work is summarized here, combined with information
from PSV to show in what very diverse environments the farms were set.

Chiche-Aubrun’s identification of many of the Seillans farms confirms
the opinion of many historians of early medieval Provence that it was
then an area of dispersed settlement. The village perché picturesquely
clinging to its rocky hill with a castle at the top, which was to become
characteristic of the region, was a product of the twelfth and thirteenth
centuries. Using the evidence of PSV for the property of villa rovagonis,
Rougon in Castellane, Jacques Cru comes to the same conclusion: the
farms recorded there were overwhelmingly dispersed.

The demographic

T. Sclafert, Cultures en Haut-Provence: déboisements et pâturages au Moyen Age (Paris,

1959).

Marijo Chiche-Aubrun, ‘L’Ager de Seillans du IX

siècle à nos jours’, in Bastides, Bories,

Hameaux: l’habitat dispersé en Provence, Actes des

Journées Regionales Mouans-Sartoux

15, 16

Mars

1985 (Mouans-Sartoux, [n.d.]), pp. 7–14.

Jacques Cru, ‘Le polyptyque de Wadalde (

813–814) et l’occupation du sol: l’exemple de la Villa

Rovagonis’, in Bastides, Bories, Hamaux, pp.

75–96, as well as other essays in this volume.

182

Rosamond Faith

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

detail we have from PSV shows that not only were farms dispersed, they
supported widely differing little communities.

Most of the commune of Seillans is at the edge of the most moun-

tainous region of the Var, but includes land in valleys of the rivers that
drain it. The farms recorded in PSV were distributed over the entire area
of the

15,000 acres of the present-day commune.

Three were at its

perimeter. At the extreme north colonica in cumbilis has been identified
with the area now known as ‘le plan de Comps’ at Les Combes Longues
– a very common name in the region, often used for large common
pastures. At this farm lived Aquilinus, colonus, his wife Gaudentia and
their six children. Two people belonging to the farm were absent and
enquiries were ordered to be made about them, ad requirendum. To the
north-west colonica ad manico, high and isolated, was possibly the
modern Magnan. It supported colonus Silverianus, Amada his wife and
their four young children, as well as Germana, possibly the widow of
Silverianus’ late father, and her adult son, perhaps Silverianus’ brother.
On the road from Seillans to Bargemon at colonica in vagaione, the
hamlet of Baguier, lived accola Gundofredus, his wife Germana and their
four children. Two and a half kilometres away on the west border of the
commune colonica ad pino felice took its name from the landmark hill of
La Pigne, capped by a hill-fort. This farm supported a complex group
including six working adults: the colonus Laurentius, his wife Dominica
and their two children, an elderly married day-labourer (cotidianus)
Joannis and his wife Expectada and their four children, one of whom was
married to an ‘outsider’ with whom he had four children. Following the
border south two colonice in amagnulis seem to have been somewhere
near the River Meaulx and were probably low-lying. On the first there
were four married couples and four children, on the second a couple with
six children. North-west of Seillans itself three farms show how different
family circumstances could be even within the confines of the simple
form of nuclear family. In the quartier of Saint-Pierre de Bouton, of two
farms each with a married couple, one had no children and one had
seven. East of Seillans at colonica ad liga (the name possibly derives from
a pre-Roman people centred on Borrigaille) the colonus Dominicus and
his wife Licinia had six offspring.

In some cases, as is common in Provence, there is now a village where

there is evidence of Gallo-Roman occupation, and this has made some of
Chiche-Aubrun’s identifications possible. Places like this are likely to
have been different from the outlying hill farms. One of the entries hints
at some kind of continuing, or renewed, central status. On lower-lying

In the order in which they are referred to here, these are PSV N

5, 6, 1, 14, 2, 3, 4, 8, 7, 9,

12, 13.

Farms and families in ninth-century Provence

183

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

and flatter land colonica vulpeglarias in the extreme south has been
identified with the quartier of Saint-Julien d’Oule. It was the most
important place in the Seillans property. Excavation has revealed a pres-
tigious second-century villa there. Lambert the priest and Stephanus the
deacon, recorded there in PSV, show the existence of a church. Of
the farm itself we know only the interesting fact that the first name in the
entry is that of a woman, the slave Frederada married to a free man, and
that they had two children. Colonica in avasio was at Avaye, where there
is evidence of an important Roman centre and a medieval village; and
Galine-Grasse, the suggested location of colonica in valiglas, has signs of
Gallo-Roman occupation. There are other farms, also in the part of the
commune for which there is archaeological evidence of Roman settle-
ment, although to date none is recorded for their particular location: at
Saint-Pierre de Figolas at Caillian and on the upper reaches of the Siagne,
colonica in civania.

PSV sometimes refers to two or three farms as ‘at’ the same location.

Two or three farmhouses and their outbuildings might have given the
appearance of a tiny hamlet. We can get some idea of what one such place
was like from the three farms at Esclangon, colonica in sclangone on the
huge property of Chaudol.

Esclangon is a deserted village site a thou-

sand metres up the steep slope of the south side of the valley of the River
Bés in the Alpes de Haute Provence (Fig.

2). It consists of a small plateau

backed on one side by near-vertical screed slopes. From the plateau the
vestiges of what was once a street of closely backed houses runs up to a
rocky ridge and from it a gentle slope of broad benches of open land leads
down to the valley, the site of the modern village. (People probably
relocated here for the convenience of access to the modern road, not for
any inherent deficiencies of the original site.) The land of Esclangon was
evidently capable of sustaining a substantial nucleated village and the
seventeen adults on the three colonice there in

813–14 could have made a

comfortable living, with a sheltered site, ample land for crops, and
unlimited pasture. Such groups of farms were perhaps the embryos of
later settlement patterns such as is found in the hinterland of Nice: farms
grouped together with their cultivable land around them, owned
individually, with communal grazing land further off.

An exception to

the general picture of dispersion is ager caladio, Chaudol (La Javie, Digne,
Alpes de Haute Provence). Its location at the head of a broad valley and
its possible role as in some way the administrative centre of the estate with

PSV H

44.

J.-P. Boyer, Hommes et Communautés du Haut Pays Niçois Médiévale (XIII

–XV

siècles) (Nice,

1990). I am grateful to Mark Whittow for his help and companionship in investigating

Esclangon.

184

Rosamond Faith

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

some kind of demesne farm, a colonica indominicada, had led by the early
ninth century to a significantly more socially mobile population and a
degree of nucleation, even incipient urbanization. (Chaudol is discussed
in more detail below).

Empty farms

The compilers of PSV found that they were not able to account for all the
monastery’s farms. Saracen raids from the eighth century had violently
disrupted rural society, and worse was to come. Many peasants moved to
get away from trouble. The eighth-century Life of St Porcarius reported
that ‘the Christians living between the sea and the Alps have abandoned
their towns and villages and fled to the mountains’.

One possible sign of

this is that six of St Victor’s properties in PSV had very large numbers of
farms noted as apsta, ‘empty’. In all,

56–58% of farms are recorded like

this. This fact has been interpreted as evidence of drastic depopulation,
but the term apsta, like the agri deserti of the late Roman period and the
‘waste’ of Domesday Book, rather than indicating outright abandonment
of the land, can simply record difficulty in collecting the tax or rent due

PSV H.

Bregi, ‘Recherches’, pp.

11–15.

Fig.

2 Farm A Saxo mancipium and his wife Aregia, their children Maginca,

Maurone and a baby. Magna, married to an extraneus ‘with their children’
Farm B ‘Empty’. Dignoaldus colonicus and his wife Pascasia and their five
children Ailaldus, Exuperius, Giso and Gairefredus, all unmarried and
Excisefredus aged

Farm C Dominicus colonicus and his wife, an extranea, and their four children
Maximus, unmarried, ‘who should rule the farm’, Standeus, Christidonus and a
child of

Farms and families in ninth-century Provence

185

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

from it.

While many farmsteads were indeed deserted, with no people

and no dues listed, on some of these ‘empty’ farms people were living, on
others people who should be there were ad requirendum, ‘to be enquired
after’. Some had been granted in beneficium, so must have had a value.
And land was leased in a métayage agreement in

817 at villa Marciana

(Ansouis) in spite of its high proportion of seemingly deserted farms.

‘Crisis’ is not the only possible explanation. Farmsteads and hamlets can
be abandoned while farming is carried on from a different base. Some still
owed payments for pasturage and may have been run as pasture only.
Gabrielle Démians d’Archimbaud sums up the characteristic of early
medieval settlement at Rougiers in the Var in terms which will have
resonance for English historians aware of the mobility of early medieval
settlement: ‘a certain tendency towards mobility, not far, but within the
same territory . . . a certain indifference to the built site, whose perma-
nence seems less essential than that of the cultivated area’. She has found
a ‘significant change of habitat location’ (revealed by pottery of the early
Christian period) away from the Roman centre in the valley at Rougiers
towards upland sites.

Greater security could well have been an attraction

of such inaccessible areas: a little fortified building was constructed
clinging to the side of the deserted Iron Age oppidum there. A St Victor
farm ‘at the ruins’ at villa Nono (Gardanne) might have been just such a
place, and other farms were located ‘at the plateau’, ‘at the rocks’ (several
incorporate an Indo-European root kal, stone), ‘at the rock piles’, ‘at the
round hilltop’, ‘at the caves’, ‘at the boulders’, and ‘at the citadel’.

The peasant farm economy

Of the three kinds of dues recorded, tributum was paid by a minority,
mostly by coloni, and always in cash, and censum was paid in pigs (both
mature animals and sucklers). Chickens and eggs were due from a minor-
ity of farms, especially from the property nearest to Marseilles. Most of
the farms owed pascuum, the payment – here generally of one or two
wethers (raised for their wool) – very widely found in the post-Roman
world for the right to graze sheep on uncultivated land.

The stock

A.H.M. Jones, The Later Roman Empire,

3 vols (Oxford, 1964), I, p. 213; Bregi, ‘Recherches’, p.

84; J.-P. Poly, La Provence et la Société Féodale, deals with St Victor’s ‘empty’ properties at p. 99ff.

J.-P. Devroey, ‘Mansi absi: indices de crise ou de croissance de l’économie rurale du haut moyen
age’, in idem, Etudes sur le grand domaine carolingien, section IX, discusses these at pp.

435–8.

CSV

291; PSV E 1–10.

Gabrielle Démians d’Archimbaud, Les fouilles de Rougiers (Var): contribution à l’archéologie de
l’habitat rural médiéval en pays méditerranéen (Paris,

1980), p. 233.

For similar payments, widespread in early Europe, see A. Verhulst, ‘Economic Organisation’, in
R. McKitterick (ed.), The New Cambridge Medieval History, II. c.

700–c.900 (Cambridge, 1995),

pp.

481–509, at p. 498.

186

Rosamond Faith

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

element in these dues suggests that all the farms had a mixed economy,
combining raising sheep and pigs with the cultivation of the cereals that
must have been needed for human consumption. This is equally true of
the bercariae which must, from their name, have been primarily devoted
to sheep (and goat) rearing. These do not seem to have been shielings,
temporary dwellings at summer pastures, but had cultivated land associ-
ated with them and people living there. All the farms must have included
three elements in varying proportions: cultivable land on which cereals
and winter fodder crops were grown, meadowland which provided hay,
and pasture. The terroir of the farms, if we take it to be coterminous with
that of the present-day communes, provides just such a mixture of
resources: the valley floors for hay meadow, the lower hill slopes for
arable, vine and orchard husbandry, the higher slopes of the mountain for
summer pasture.

The few farm names which can be interpreted today

give some indication – although it will be a very slight one – of the
distribution of settlement over these landscapes, and the use which was
being made of their resources.

Although no farm has a name which can be associated with arable

husbandry and renders in grain are only found very occasionally (oats and
barley at Seillans), some kind of cereals must have been widely grown.
Modern grains can be grown up to

1000 or 1500 metres. Although there

is some doubt whether this would apply to the grains of the early Middle
Ages, there is no obvious staple carbohydrate the peasants can have been
eating other than bread, although chestnuts are a possible supplement.
Most, judging by the dues they owed, reared fowl and pigs, and references
to vines, olive groves, gardens and orchards indicate places where inten-
sive cultivation was important and conditions were sheltered. Wax and
honey appear once as dues. ‘Donkey/mule pasture’ appears as a farm
name (colonica ad asinarias), and only mules or donkeys would have been
able to tackle the mountain paths: there is no sign, whether in place
names or dues, that horses or cattle were reared, either for traction or
food. Sheep were the main stock animal, valued for food (cheese, milk
and meat), for their wool, hides and horn, and also for their manure
which maintained the fertility of the thin upland soils. Studies of a high
limestone plateau in eastern Provence in the later Middle Ages show that
the smaller peasant flocks did not roam freely but were fairly strictly
confined, perhaps by moveable hurdles: at night ‘close-folded’ when they

Due to reafforestation policies modern maps often show a greater degree of tree cover than was
probably the case in the early Middle Ages and with a different distribution: G. Bertrand, ‘Pour
un histoire écologique de la France rurale’, in G. Duby and A. Wallon (eds), Histoire de la France
Rurale,

4 vols (Paris, 1975), I, pp. 37–111, at p. 94.

Farms and families in ninth-century Provence

187

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

needed protecting from a range of predators (wolves and eagles in par-
ticular), grazing and manuring small areas of land during the day.

This raises the question of transhumance. It is sometimes assumed that

the large-scale long-distance transhumance which was to be such an
important part of Provençal life in the Middle Ages, and which would
provide such an important part of St Victor’s income in the future, was
‘immemorial’.

This is not likely to have been the case here. Such a

system demanded conditions that ninth-century Provence could not
satisfy: social order, large-scale investment, the establishment of safe and
recognized transhumance routes.

But nearly all the farmers on St

Victor’s estate may well have have been practising ‘internal transhu-
mance’ or alpeggio, moving their stock around short distances at various
stages in the year: in the case of the Alpine farms wintering animals near
the farm itself, where the lambs were born, moving the young stock part
way up the mountain in spring, and further up onto the heights in the
summer. In other regions the movement may have been the other way:
down onto lower pastures for summer grazing. Such evidence as can be
gleaned from farm names, and from the general topography of the St
Victor properties, illustrates this.

Hay provided the most important element in winter fodder and so the

stock needed to be kept off it in the spring and summer growing season.
Valley-bottom and lower hill slopes provided grass which could be mown
for two, possibly three, crops of hay. (Farms ad mairolas (recte ad mariolas,
suffix -iola, field), in prato, ad pradas, and in pradinas look as if they were
set among meadows.) This meant that another essential resource for every
farm was permanent pasture, land used only for grazing, which is likely
to have been on the more sheltered hill slopes. These ‘fields’, several of
which are in mountain valleys far from today’s villages, were the valuable
hill pastures which made upland farming viable. The young stock needed
sheltered grazing which would provide an ‘early bite’ in spring and kept
them near enough to the farmstead to be supervised easily. The farm in
prima capa (recte campa) the ‘first field’ at ager caladio, may have been at

A. Fleming in D. Austin, A. Fleming, R. Faith and D. Siddle, Cipières: A Provençal Village in its
Landscape (forthcoming).

P. Coste, ‘Les origines de la transhumance en Provence: enseignements d’un enquête sur les
pâturages comtaux de

1345’, in L’Elevage en Méditerranée Occidentale (Paris, 1977). A similar

assumption underlies much of Bregi, ‘Recherches’.

P. Garnsey, ‘Mountain Economies in Southern Europe: Thoughts on the Early History, Con-
tinuity and Individuality of Mediterranean Upland Pastoralism’, in L’Uomo al Fronte al Mondo
Animale, Settimane di Studio

31 (Spoleto, 1983–5), pp. 401–51. For a region where the devel-

opment of transhumance had a very similar chronology to that of Provence, see D.R. Blanks,
‘Transhumance in the Middle Ages: The Eastern Pyrennees’, Journal of Peasant Studies

23 (1995),

pp.

64–87.

188

Rosamond Faith

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

what is today the mountain valley called Premier Champs, some five
miles from Chaudol. As animals matured they could be moved further up
the mountain.

There is little evidence of what we think of as ‘common’: an area open

to common grazing by a whole community. Rather, individuals’ rights to
pasture seem to have been envisaged as attaching to their ownership of
their farms. In the tenth century the peasants of Chaudol each claimed a
share in a named area of mountainside, but this articulation of rights and
their recording may have been brought about by local circumstances.

Chaudol is an exception to the general picture of dispersion and was
unusual among the St Victor properties in several ways. It lies in a broad
valley and was more than just a collection of farms: it had a centre, some
kind of demesne farm (a colonica indominicada) with its own pastures,
and it is possibly here that the ‘rural and urban slaves’ worked who were
referred to in an eighth-century charter. Several boys were at school or in
orders and there was a church dedicated to St Damian, served by a priest,
Beno. There was an industrial area – one farm is called colonica ad fabricas
– and by the tenth century there were (very contentious) fulling mills,
vineyards, a church, and ‘a little mountain where there was a fortress long
ago’.

Also by the tenth century a peasant community had formed,

capable of collective action based on a claim to accumulated individual
pasture rights on the mountains. It is likely that this notion was wide-
spread. The idea that ownership of a farm conveyed rights derived much
of its strength from the belief in descent from ‘founding fathers’. B.
Derouet suggests that rights in pasture developed as an extension of the
rights in the cultivable land of members of family groups who saw
themselves as descended from certain specific ‘original’ families.

As we

will see, the peasants of St Victor’s estate evolved strategies for ensuring
that their farm descended in the family. If the farm conveyed rights, and
if those rights were seen as adhering in a particular family, and if,
moreover, they derived from that family’s traditional status, then more
than securing a living was surely involved. But how they secured a living
was basic to family structure and to understand both we need to under-
stand the environment in which these people were farming.

CSV

739, 741.

PSV H; CSV

739. For a commentary on this latter text see Duby and Wallon (eds), Histoire de

la France Rurale, I, pp.

23–4.

B. Derouet, ‘Territoire et parenté: pour une mise de perspective de la communauté rurale et des
formes de reproduction familiale’, Annales

50.3 (1995), pp. 645–86. This view of common right

originating as an accumulation of individual rights in land has resonances in English rural
history: F.W. Maitland, Domesday Book and Beyond: Three Essays in the Early History of England
(Cambridge,

1987), pp. 404–10. For ‘founding fathers’ in medieval Provence, see E. Sauze and

P. Sénac, Un pays provençal, le Freinet: de l’an mille au milieu du XIII

siècle (Minerve,

1986).

Farms and families in ninth-century Provence

189

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Mixed woodland covered a large area, to judge from the many farms

named from oak, elms, yew, birch, poplar, ash and beech. English land-
scape historians have learned to appreciate the importance of wood-
pasture and the role played by trees and undergrowth as fodder, as well as
of acorns and beechnuts as mast for pigs.

Sclafert has described the

medieval mountain economy of Provence as one in which there was a
range of strategies to exploit woodland: wood-pasture was combined with
piecemeal clearances made by burning and grubbing up roots.

The

scrubland which is now so extensive appears in its typical vegetation in
the farms named from such plants as mare’s tail, rosemary, box and
broom. Above that were the dry rocky heights which can be grazed by
sheep and goats in the summer months in the year, and where farms
called ‘at the heights’ and ‘at the rocks’ and those named from bears,
wolves and deer show permanent habitation even in these tough condi-
tions. It needed a complex deployment of labour to work this demanding
landscape so as best to exploit its pasture resources. Stock had to be
moved periodically and to be in the care of knowledgeable and respon-
sible people in those periods when away from the farm itself. It has been
calculated that to support a family of just under five ‘adult equivalents’
from raising sheep, if other sources of food were available, ‘to go below
sixty heads (sic) would lead to the vicious decline of the flock, ending in
poverty’.

The shepherding work of driving, folding and guarding even a

small flock in a short-distance transhumance system would have
demanded the labour of one, possibly two, working adults. It is possible
that neighbouring farms could have pooled their flocks and employed a
shepherd, and thus reduced the amount of labour needed. Yet if farms
were in the main widely dispersed, opportunities for this would have
been limited. Ninth-century Provence was not yet the domain of the
professional shepherd like those of the Spanish Mesta, or the far-travelling
shepherds of Montaillou. And shepherding was not the only work
involved. Milking and cheese-making, possibly cropping the hay from
upland pastures, even growing small amounts of corn there, demanded a
group of adults (and their dogs), as did protecting the animals from
thieves, eagles, wolves and wild boar. There was also a further set of
labour demands. It was essential to grow on the cultivated area near the
farmstead enough food for the family, and to store enough fodder,

D. Hooke, ‘Early Medieval Woodland and the Place-Name lea¯h’, in O. Padel and D.N. Parsons
(eds), A Commodity of Good Names: Essays in Honour of Margaret Gelling (Donington, Lincoln-
shire,

2008), pp. 365–76.

Sclafert, Cultures.

G. Dahl and A. Hjort, Having Herds: Pastoral Herd Growth and Household Economy (Stock-
holm,

1976), p. 220. This study is based on an almost entirely pastoral North African economy.

In medieval Provence in the Middle Ages the standard unit for sheep was a trentain of thirty.

190

Rosamond Faith

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

whether grown as hay or collected from the waste (bracken, holly etc.), to
keep the stock alive for up to six months in the winter. Gardens, arable
fields, vineyards and orchards required considerable and skilled labour. As
will appear, the labour supply was a vital factor in the farm economy.

Households and work groups

The historical demography of medieval Europe has traditionally taken
the ‘household’ as the primary unit to be investigated. But if we want to
examine the relationship of the St Victor peasants to the land, the
household is not available for study. While it is obvious in some cases –
two parents and their young children, say – that the polyptyque records
a household, it is often impossible to know whether the group of people
registered, which might include several married couples and their chil-
dren as well as a handful of adult single people, were living as one
household or several. Nor has the archaeology of peasant housing at this
period for this region yet provided a basis for such an enquiry. However,
we can assume that all the people registered under the name of a par-
ticular farm, resident in one household or not, in some way belonged to
it or were associated with it. They were in what Peter Laslett, in his study
of pre-industrial families, calls a work group.

The St Victor evidence is in line with Laslett’s finding that the single

family farm is the most common form of peasant landholding in medi-
eval and early modern Europe. Of the

101 colonicae and bercariae whose

occupants’ family relationship can be determined, sixty-three were occu-
pied by a single family (defining a family as a group with a marriage or
blood relationship, and including as families widows and widowers and
single parents). How did a single peasant family manage the farm? There
is no doubt that some must, if they relied solely on the labour of the
family, at various stages in the life cycle have gone through periods of
desperate difficulty. Approximately one-third of all farms (thirty-one) had
either one or two adults only present, and twenty to twenty-three had no
adults beside the farming couple. Some of the latter may have been young
couples who had not yet started a family, others possibly old people
whose children had left home. Not all farms had as many as two working
adults. A few farms were run on women’s labour alone. The most likely
circumstance for a woman to have control over land in the European
middle ages was as a widow, and that is the case here: on fourteen farms
the head was a woman, six of whom were known to be widows. At villa

P. Laslett, ‘Family and Household as Work Group and Kin Group: Areas of Traditional Europe
Compared’, in Richard Wall with Jean Robin and Peter Laslett (eds), Family Forms in Historic
Europe (Cambridge,

1983), pp. 513–63.

Farms and families in ninth-century Provence

191

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Marciana (Ansouis) each of the only two farms that were occupied had
two single women as its sole occupants. A widow would be in a particu-
larly weak position, especially if her husband had died when their chil-
dren were still young, like the widow with children aged five and ten who
ran a farm at villa Sinaca (Seillans).

Shortage of labour would have been

felt most at the beginning of the farming couple’s life together and
particularly when the children were young, on the assumption that
women who were breast-feeding a baby would have been less capable of
doing their normal share of heavy work. These would have been the
difficult circumstances of the couples with young children on the four
farms at V Lambiscum (Lambesc). However, several factors meant that the
single-family farm did not remain short of labour for ever. Most children
grow up, after all. And there was a good supply of children. The popu-
lation was replacing itself from its own ranks, not from immigration, and
increasing. The mean figure of

2.9 children per couple conceals a wide

range of family size. Infant mortality was not so high that it prevented six
or seven children of some couples surviving into adult life. Women not
infrequently had a married child and a new born baby and at least one set
of twins survived infancy.

The most common source of labour on the single family farm was

adult offspring. At fourteen or (more rarely) fifteen, children were
recorded as baccalarius or baccalaria.

The assumption here is that they

were then considered adult. There is likely then to have followed a period
in which they were unmarried but capable of an adult’s work. In the case
of sons, this period may have lasted a lifetime (see below). If at least some
of the children remained at home a family could, given time, produce an
adequate labour supply from its own ranks. Hence on sixty-three single-
family farms, thirty-seven farming couples had adult offspring living at
home. Yet in spite of the need for labour, the marriage regime was not
designed to produce the maximum number of children. By southern
European standards marriage was late, was at more equal ages for men
and women, was less universal, and hence produced fewer children.

Twenty-two farms had no adult children of the farming couple present.
Nonetheless marriage practice and labour needs were to some extent
linked. Marriages which were late and at a similar age for men and
woman may well have been connected with the employment of

PSV C

1, 2, 3.

M. Zerner-Chardavoine, ‘Enfants et jeunes au IX

siècle: la démographie du polyptyque de

Marseille

813–814’, Provence Historique 126 (1981), pp. 355–84. By contrast Bregi, ‘Recherches’,

pp.

155–9, interprets the demographic evidence as showing ‘a human society in the course of

extinction’. For twins PSV H

68.

This is deduced by Zerner-Chardavoine from the highest ages for offspring recorded as infans.

Zerner-Chardavoine, ‘Enfants et jeunes’, pp.

369–70.

192

Rosamond Faith

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

unmarried people as living-in labour. Some forty-seven farms had such
single people registered, twice as many men as women (

62 : 31). These

single people are found on all types of farm – both single family and
multiple family. They were very much more numerous on the sheep
farms of the Alpine properties. Nearly all the farms at villa Virgonis
(Annot) in the Basses Alpes, for instance, had up to six single people
resident as well as the unmarried children of the farmer. It could well be
that they were ‘servants in husbandry’: living-in farm labourers. H.S.A.
Fox has illuminated the important part that such workers played in the
economy of the medieval peasant farm in the English west country.

shows that this form of labour was confined to large farms which needed
a labour supply throughout the year beyond the resources of the family.
On these farms, he calculates, even for a family which could draw on an
adequate supply of labour from within its own ranks, it was more
economical to hire a neighbour’s child and send your own out to work.
Many servants in husbandry on the St Victor farms will have been the
offspring of neighbouring farmers and some may have been relatives of
the farming couple that employed them. It is reasonable to suppose that
there was a considerable amount of intermarriage between families of
many of the farms on the St Victor estate. Personal names reveal such
links between three, possibly four of the farms at ager Caladio (Chaudol).
There may have been in fact a dense web of family connections of this
kind.

One group of workers who may be invisible simply because they

do not fall under the remit of the compilers of the polyptyque are
farmers’ slaves. There are several people specifically recorded as slaves
(mancipia) in the polyptyque, but most of these are farmers or farmers’
sons. The cotidiani occasionally recorded were likely to have been day-
labourers living on the site or in the household and so recorded there. A
long enough association with a farm might earn them expectations of
lifelong support, possibly even inheritance.

H.S.A. Fox, ‘Exploitation of the Landless by Lords and Tenants in Early Medieval England’, in
Z. Razi and R.M. Smith (eds), Medieval Society and the Manor Court (

1996), pp. 518–68 and

‘Servants, Cottagers and Tied Cottages During the Later Middle Ages: Towards a Regional
Dimension’, Rural History

6 (1995), pp. 125–54.

There were at least two naming practices at work. There were explicitly Christian names and
others which look Germanic (?Burgundian). Children’s names frequently incorporated an
element from one parent and sometimes one can deduce another forebear from whom they
were named. The claim of Weinburger, ‘Transformation de la société paysanne’ (note

3 above)

to detect racial differences between the name-pools of slaves and non-slaves is hard to sustain
in view of the fact that parents of large families used both Latin and Germanic names – as did
Stabilia and her husband at vercaria in Alesino at Chaudol, whose children were Projecta, Maria,
Nicasia, Gaugemares, Constantinus, Babo, Daniel, Leuderic and Adelaicus (PSV H

73). It is

possible that some of these were their grandchildren.

E.g. N

14 and H 67 for married quotidiani with children at Seillans and Chaudol. At Seillans

the farmer’s own children were probably too young to work and the quotidianus’ son had been
allowed to marry an outsider and raise a family, and evidently had an expectation of support

Farms and families in ninth-century Provence

193

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Farms with more than one working and resident family comprised

about a third of the total. Of these most were in villae Sinaca, Bettorida
or in ager Caladio (Seillans, Fourcalquier, Chaudol). Some complex
groups are revealed. The farm ad Cenas at Chaudol was run by Gairefre-
dus, a mancipium, and his wife Vuoldefreda. They had two small chil-
dren, a girl of five and a boy of six, and a baby. There was also a widow,
Dominique, with two adult unmarried daughters (baccalarie), and (?her)
three sons and a daughter whose ages are not given. There were in
addition, and apparently quite unrelated to these two families, two adult
unmarried men and four adult unmarried women. In all there were
eleven adults on the farm, of whom one, the widow, was probably elderly,
and one, with a child at breast, young, and eight were of the status to be
called baccalarius, adult but unmarried. This is a large working group. On
other farms the number of adults could reach over twelve.

Bregi con-

sidered these groups as examples of the phenomenon of the ‘overcrowded
manse’ observed on the great bipartite estates of northern Europe. He
resolves the contradiction of deserted farms in some areas and over-
crowded ones in others by representing them as peasant strategies in the
face of extreme poverty, to desert unproductive land and combine their
resources ‘piled up on the same holding’.

But it is not really appropriate

to compare Provençal mountain farms with access to virtually limitless
upland resources and their own heavy and complex labour demands, with
the arable-based holdings of tenants on northern estates, fixed in size and
often burdened with heavy labour rent due to the landlord.

The mountain bercarie or sheep farms are extreme forms of the mul-

tiple family unit and they have other singularities as well. They were more
likely to have supported several unmarried adults. They were also much
more likely to have people with children whose names and ages are not
given. And they were more likely to have unmarried fathers and mothers
(x cum infantes (sic) suos, baccalaria cum infantes suos). Very high illegiti-
macy rates were a characteristic of Alpine farms in the early modern
period, where ‘illegitimate children were a welcome addition to the rural
lower social strata, from which farmers could draw servants’.

Another

from – possibly the inheritance of – the farm. At Chaudol the eldest son Joannis had left home,
possibly because for a labourer’s son there was no property to inherit. Hammer, ‘Family and
familia’, p.

222 found mancipia, bond workers, ‘allotted to the tenants and sometimes . . .

related to them’ on a ninth-century Bavarian estate.

PSV H

55.

Bregi, ‘Recherches’, p.

90 and n. 84, p. 92; Y. Morimoto, ‘Sur les manses surpeuplés ou

fractionnaires dans le polyptyque de Prüm: phénomènes marginaux ou signe de décadence’, in
Campagnes médiévales: l’homme et son espace: études offertes à Robert Fossier (Paris,

1995), pp.

409–23.

P.P. Viazzo, Upland Communities: Environment, Population and Social Structure in the Alps
(Cambridge,

1989), pp. 191–2; Hammer, ‘Family and familia’, pp. 232–4 considers the

194

Rosamond Faith

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

feature of the mountain sheep farms in PSV is the frequent note that such
and such a person – and it is twice as likely to be a man as a woman – is
ad requirendum (although these are not the only places where such a note
is found). I take this to mean that not enough information was available
about them and they were to be enquired into further.

It cannot be proved that it was the large numbers of baccalarii on some

of the sheep farms who provided the essential shepherds for the flocks on
the mountain. But the facts that so many such people could not be traced
when the compilers of the polyptyque came round, that the local peas-
antry knew that these people had children but didn’t know the children’s
names, and that it was among this group that the rare cases of unmarried
parents are found, all suggest a group of young people of both sexes away
from parental control and likely to be physically some distance away from
the farmstead.

Marriage and inheritance

But the majority of parents were married and in most anthropological
studies of peasant societies marriage is seen as a crucial stage in the
fortunes of the farm and the family. What happened when an adult child
of the farm married on the lands of St Victor? Many marriages are
recorded between a named person and a spouse described as extraneus or
an extranea, ‘foreigner’ or ‘outsider’, whose name is not given. Of

144

identifiable married couples, fifty-four included a spouse who was
extraneus/a. It is likely that the polyptyque’s use of the term reflects estate
policy, and that they were not from remote regions but were people who
did not belong to the estate of St Victor. It appears that marriage within
the estate seems to have been what the abbey encouraged, possibly
prescribed. At villa Virgonis (Annot), a note tells us, ‘these men from that
county have our women’ (abent illi homines de illo commite nostras femi-
nas).Two women, both with children, are then named, and then a couple
with at least three adult children. The phrase ‘our women’ is surely
significant. There were some places where the monastery owned only
some of the land, so ‘outsiders’ may have been simply from neighbouring
families who happened to live on land which was not St Victor’s.

However, if there was a policy of encouraging marrying within the estate
it was unsuccessful, as the large numbers of marriages with outsiders
shows. It appears to have been particularly unsuccessful in the case of

unmarried unfree women with children exchanged between landowners in numerous tenth-
century cases to be women married to men from another estate, but also allows for ‘single
parents . . . (as) . . . real, not administrative, phenomena’.

PSV I

7; Hammer, ‘Family and familia’, p. 232 has a similar interpretation, regarding a lordship

as a ‘seigneurial family’.

Farms and families in ninth-century Provence

195

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

women: more women are recorded as having married extranei than men
had married extraneae. It is an important question whether marriage to an
‘outsider’ spouse resulted in a son or daughter leaving the farm, taking
their labour power and that of any subsequent children elsewhere. In
many peasant societies this was the point at which a new household was
formed, and provision was made in land or goods to support it. An
alternative practice seems to be the rule here: women who married
outsiders brought their new husbands onto the farm, where they were an
important source of additional labour in return for which they must have
been entitled to support from the farm’s resources.

For example, at a

farm at Ardonis at Chaudol the farmer Mauresindus (a slave) and his wife
Gennaria had two unmarried sons, two unmarried daughters, and two
daughters married to extranei. One of these had presumably married
quite recently, because she was still nursing a baby.

The children of

couples of whom one was extraneus/extranea – and some had been born
very recently – are named, and it is hard to imagine that such up-to-date
information would be available if the couple had moved a great distance
away. If a St Victor man married an ‘outsider’ wife he brought her to live
on the farm – but only in particular circumstances.

Heads of farms

There are important differences between the individuals, overwhelmingly
men, who head the farm entry and others, including members of their
own family. Heads of farms were much more likely to be married than
were other men. Nearly every head of a farm was either married or
possibly a widower. Men were less likely to marry outsiders than women
were, and most of the men who had done so were the heads of their
farms, listed at the beginning of the entry. But as significant as the
marriages that did take place are the marriages that did not: never is more
than one son of the farming couple – that is to say the head of the farm
and his wife – recorded as married. This is a practice which can be
interpreted differently in different contexts, but underlying it must be a
preoccupation of the ratio of labour to land. For only one son to marry
was also the practice among the slaves of the monastery of Farfa at about
this time. Laurent Feller, in his study of the regime there, considers that
the monastery encouraged this in order to provide a pool of unmarried

Viazzo, Upland Communities, found a modified version of such a system in the alpine regions
in the early modern period where an only daughter marrying would bring her husband to her
parents’ farm. E. LeRoy Ladurie, ‘Système de la coutume. Structures familiales et coutume
d’héritage en France au XVI

siècle’, Annales

27.1 (1972), pp. 825–46 describes a ‘widespread’

French–Swiss custom which allowed those who ‘married in’ to share the inheritance.

PSV H

196

Rosamond Faith

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

and landless male slaves who could be shifted around the estate to work
on opening up new land.

Adult men with no dependants would have

been equally useful on the St Victor estate. When circumstances favoured
them, landlords were certainly able to impose particular inheritance
practices on their tenants but, as is argued below, it is doubtful that the
monastery in the ninth century enjoyed such control. Inheritance
customs can equally well be seen as embodying essential peasant values
and to be family strategies aimed at securing a valued goal: the preserva-
tion of the family holding.

LeRoy Ladurie considers the extended family

group headed by the patriarch, which saw a revival in Languedoc in the
crisis of the fifteenth century, to be part of ‘the old bedrock of customary
law’.

To preserve the patrimony is a priority common to many peasant

societies. European peasants farmed in a wide diversity of conditions
and the strategies they adopted to preserve the patrimony differed
according to the rural economy of which it was a part. J.-P. Derouet has
illuminated the distinction between cultures where residence confers
rights, and those where rights were inherent in the individual.

If rights

reside in individuals they can take their share of the moveable property
and leave: the preciput system of Roman law. The patrimony is pre-
served by providing the offspring with marriage portions from moveable
goods or acquired land. In the alternative system it is residence, and by
implication sharing the work of the farm, that conveys a right to be
supported by it. This seems to have been the strategy of the St Victor
peasants to preserve their scattered, mostly upland, farms in which pas-
toralism played an important part and adult labour was valued. They
limited the number of mouths the farm had to feed by allowing only
one son to marry, while ensuring every adult male a right to support
from the farm in return for their work there as long as they remained
unmarried.

In fact the concept of ‘inheritance’ seems inappropriate to

Feller, ‘Liberté et servitude’ (see note

8).

English peasant inheritance practices varied but had much the same priorities, preserving the
holding by providing marriage portions for children who left. Where joint inheritance was the
rule, this may in practice have meant sharing use-rights as much as sharing in ownership.
Inheritance did not always favour the eldest child. The care of the retiring couple was also a
consideration – hence the practice which gave the bulk of the land to the youngest, or ‘fireside’,
child. Although sons were preferred to daughters in the line of descent, daughters inherited
jointly if there were no sons. R.J. Faith, ‘Peasant Families and Inheritance Customs in Medieval
England’, Agricultural History Review

14 (1966), pp. 77–95.

E. Le Roy Ladurie, The Peasants of Languedoc, trans. John Day (Urbana and Chicago,

1974),

31.

Derouet, ‘Territoire et parenté’ (see note

31).

This kind of strategy, with its strong roots in ideas about rights in land and their connection to
membership of a household, was not a universal peasant practice: Louis Andrieu-Assier found
that in the Cerdagne it was only the custom among the richer households – the landholdings
of the poor could not support all the members of their families. Pierre Bourdieu’s studies of

Farms and families in ninth-century Provence

197

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

the peasants on the St Victor estate. When we learn about the relation-
ship between generations, the rule is not about the transfer of land but
about the transfer of authority. There was a strong assumption that only
one person should ‘govern the farm’, colonicam regere.

This was an

important choice and it was necessary to settle in the lifetime of the
governor who should take over after his death. In five entries we are told
who is singled out as the one ‘who should govern the farm’, qui debet
colonicam regere. All were married men, three of them identifiable as the
son of the existing governor. Whether explicitly recorded like this or
not, only one person in each generation would be the governor. He
provided the managerial authority to organize the complex deployment
of labour needed. The old governor hung on until grim death, after
which his widow was probably entitled to govern the farm until hers.
During the governor’s lifetime his successor, probably his son (although
not necessarily the oldest), would stay on the farm, marrying a woman
from the estate or an ‘outsider’ and bringing up their children there. He
would be the only son to marry. His unmarried brothers, uncles and
cousins, possibly also his married sisters and their children – or those
who chose to remain – were all part of the farm population, the ‘familial
community’, well suited to meet the labour demands of a farm economy
with a transhumant element. In return for their work the farm owed
them all a living.

Women’s marriages to ‘outsiders’ brought valued

adult labour onto the farm, while the single men who remained were
useful hands but fathered no extra mouths to feed, apart from the
illegitimate. It was not so much a question of the people inheriting the
farm, but of the farm inheriting the people.

It is a characteristic of family systems which impose a high level of

celibacy, or at least a low incidence of marriage, on men that an excep-
tionally large number of boys enter the priesthood. If there are other
young men to stay at home and work on the farm, a boy can be spared.
Notable numbers of the St Victor boys and young men were in full-time
education, or in the church. Seven were recorded as being ‘at school’ or

Béarn show some families maintaining the tradition that only one son should inherit in the
teeth of the provisions of the Code Civile, which prescribed partibility. Louis Andrieu-Assier,
‘Le Play et la famille-souche des Pyrénées: politique, juridisme et science sociale’, Annales

39.3

(

1984), pp. 495–512; and Pierre Bourdieu, ‘Célibat et condition paysanne’, Etudes Rurales 5–6

(

1962), pp. 32–135. David Siddle has been very helpful in introducing me to French work on

peasant families.

For similar farm heads in central Italy, see P. Toubert, Les Structures du Latium médiéval: le
Latium méridional et la Sabine du IX

à la fin du XII

siècle,

2 vols (Rome, 1973), I, pp. 480–4.

In two cases the existing head of the farm is described as relevatus. Among upper-class families
of feudal Europe this term could be interpreted as meaning someone who has received a relief,
or payment for succession, from a prospective heir. Its meaning here is obscure, but it possibly
translates some such term as ‘his successor is decided’.

198

Rosamond Faith

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

‘at the schools’, ad scolas. Two were from one family in villa Bettorida
(Fourcalquier) where there were three other adults in the family, seven
were described as clericus, two had reached the rank of deacon, one was
a priest. The ‘schools’ could have been at Marseilles or perhaps at the
diocesan centres at Digne or Embrun.

The peasants of St Victor

While the farm provided in some way for all the family this was not a
peasant economy which ensured prosperity, or even survival, to all its
members. It was an economy of subsistence faming at best and many
families must have been vulnerable to natural and demographic crises
which could wipe them out completely. Yet underdevelopment in an
agrarian economy, by delaying the effective exploitation of their estates by
landlords, can also ensure a certain degree of independence for the
peasantry. There is little sign in Provence of the ‘manorial system’. Apart
from its colonica indominicada or demesne farm at Chaudol, there is very
little evidence that St Victor in the early ninth century kept any land in
hand as demesne or ‘inland’.

Chaudol is where we might expect to find

tied dependants and it is here that ‘rural and urban slaves’ had been
recorded. But instead we find that the people there regarded themselves
as free and were capable of defending against the abbey what they
considered to be their historic rights.

It is an important fact about the

polyptyque that it is a register of people, not of tenancies. Significantly,
the term mansus, commonly used for peasant tenancies elsewhere, is
nowhere used and was not to be in this part of the world until the tenth
or eleventh centuries. For J.-P. Poly, in Provence ‘one thing is clear – the
existence of small allods’.

PSV makes no reference to labour rent being

due from any of the farms and quite apart from the enormous difficulties
of actually collecting quantities of chickens, wethers, and so on from its
inaccessible farms, the dues St Victor was owed look more symbolic than
an effective transfer of surplus. Pascuum, paid for pasture rights, is
thought to have originated in payments used for use of the saltus, uncul-
tivated land. Tributum, owed principally but not exclusively by coloni,
was widely paid in early medieval southern France as a sign of personal

PSV F

PSV H

66 and above, p. 15. The area near the monastery itself, the ‘Paradisus’, may have served

as a small demesne farm, but none of the properties in PSV were near enough to have done so.

Poly, La Provence, p.

87ff.

R. Latouche, ‘Quelques aperçus sur la manse en Provence au XI

et au XII

siècles’, in Recueil

de Travaux Offert à M. Clovis Brunel,

2 vols (Paris, 1955), I, pp. 104–6. Poly, La Provence, p. 99.

The allodarii of Chaudol can be seen in dispute with the abbey in

1055 in CSV 738–741.

Farms and families in ninth-century Provence

199

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

dependence, as were chickens and eggs paid as eulogiae, gifts. They may
all be relics of ancient payments to landlords or to the fisc.

The legal status of the people recorded in PSV returns us to
the importance of comparing language and reality. It is a discriptio
mancipium, a register of slaves, and forty-one people in it are described as
mancipia. However their children are not necessarily described as slaves
and whether we look at family structures, effective rights to inherit land,
the capacity to be the governor of a farm and thus the head of a sizeable
agrarian enterprise, to marry, even to marry an ‘outsider’, there is no
apparent difference between those registered as slaves and those registered
as coloni.

What the dues owed by each farm principally marked was that

the people registered there belonged to St Victor. Immense care was taken
to record every man, woman and child who lived on the estate, and to
track down those on remote farms, and to record absentees. That spouses
and children, including girls, were recorded by name while people not
born on the estate – ‘outsiders’ – were not must mean that the status of
being ‘ours’ was hereditary. That children’s names contained both mat-
ronymic and patronymic elements may mean that descent was traced on
both sides. Bregi effectively sums up the purpose of PSV: it was drawn up
to ‘reaffirm the ownership by the church of land and people’. This
attachment to the estate may be the one significant inheritance from
antiquity on St Victor’s properties, a relic of the concept of origo and
adscriptio, of the individual belonging to the estate or place at which he
or she was registered for taxation. Belonging to the estate may have meant
in theory being unable to leave it, like the coloni of the ancient world,
personally free but bound to the soil. Much of the late Roman law
relating to coloni was designed to prevent them leaving their holdings and
bring them back if they did so, and the church canons continued this.
This has been taken as the most telling mark of their condition and is
primarily why they look like the forerunners of medieval serfs.

Many

aspects of the relationships between peasants and landowners in Provence
must have evolved in the Visigothic and Burgundian periods, when

These dues are substantially lower than those owed by the servi ecclesie in the Bavarian and
Allemannic laws: Fustel de Coulanges, Histoire des Institutions: L’Alleu et la Domaine Rurale, p.

382; Verhulst, ‘Economic Organisation’, p. 498.

The only difference is that only one of the boys at school was a slave, and this may be explained
by the fact that school was the pathway to ordination or the monastic life, and these demanded
free status.

For the status of the colonus in Merovingian Francia see Fustel de Coulanges, Histoire des
Institutions, IV, pp.

355–69, and under the late empire J.-M. Carrié, ‘“Le colonat du Bas-

Empire”: un mythe historique’, Opus

1 (1982), pp. 351–70 and ‘Un roman des origins: les

généalogies du Colonat du Bas-Empire’, Opus

2 (1983), pp. 205–51. Carrié disputes that the

colonate formed a legal class but his articles illustrate many aspects of their condition. I am
grateful to Peter Sarris and Jairus Banaji for their help on this question.

200

Rosamond Faith

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

powerful figures of a new type took over with the land they conquered a
mixed inheritance from the battered Roman state. Part of that inherit-
ance were the legal terms relating to the rural population. Great lords,
often themselves of barbarian stock, like Nemfidius who gave land to the
monastery in the eighth century, passed along with it ‘rural and urban
slaves, freedmen and tenants both natives and those moved there from
elsewhere’ and the right to collect dues from them.

Yet as Alice Rio

shows in a recent article, we need to look beyond the legalistic definitions
of unfreedom to what happened in practice, and when we do this we find
that ‘a single blanket condition which could be characterized either as
“slavery” or “serfdom” is nowhere to be found’.

In the ninth century St

Victor’s coloni and mancipia resembled neither the tied colonate of the
late empire nor the dependent tenants of the bipartite estate. Nor were
they in transition from one to the other. The demands that their harsh
environment had imposed on them had ensured their independence. If
they had ever been regarded as being ‘bound to the soil’ this had evolved
into a de facto entitlement to remain on the family farm, an entitlement
which a closely regulated inheritance system restricted to a sustainable
group. Thus the strategies that the St Victor peasants had evolved ensured
that the farm would provide a living for the family over generations.

Kellogg College, Oxford

‘Mancipia tam rustica quam urbana libertibus accolabus et inquilinis tam ibidem consistentibus
quam ex aliunde ibidem translatis’ (CSV

31). Ganshof, ‘Les avatars’.

Alice Rio, ‘Freedom and Unfreedom in Early Medieval Francia: The Evidence of the Legal
Formulae’, Past and Present

193 (November 2006), pp. 7–40, at p. 38.

Farms and families in ninth-century Provence

201

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

The Peace of God, the ‘weakness’ of

Robert the Pious and the struggle for

the German throne,

1023–5

emed_296

202..222

heo Riches

The author of The Deeds of the Bishops of Cambrai accused Bishops
Berold of Soissons and Warin of Beauvais of overstepping the boundaries of
episcopal authority and usurping royal rights by promoting the Peace of God
and attributed their initiative to the weakness of King Robert the Pious. This
paper argues that the author was misrepresenting the situation to hide the
vulnerability of the bishop of Cambrai during the succession of Conrad II.
Instead, Berold and Warin’s peace council was patronized by Robert the Pious
and was a symptom of French royal assertiveness in the period

1023–5. The

reasons for the Cambrai author’s distortions are to be found in the significance
of kings in the rallying of support on a local and regional level.

The year

1023 was a good one for Bishop Gerard I of Cambrai-Arras

(

1012–51).

He was as close to his patron Emperor Henry II (

1002–24) as

he had ever been since his promotion from the royal chapel, playing a key
role in the spectacular meeting of Henry and the French king Robert the
Pious (

996–1031) in August of that year.

He began to have a new

cathedral built and commissioned a new vita of its patron saint Gaugeric

There have been a number of short articles on Gerard in specialist dictionaries, including H.
Sproemberg and J.-M. Duvosquel, ‘Gérard Ier’, in Biographie Nationale de Belgique, vol.

(Brussels,

1969), pp. 286–99; E. van Mingroot, ‘Gérard Ier’, in Dictionnaire d’histoire et géog-

raphie ecclésiastiques, vol.

20 (Paris, 1984), cols 742–51 and T. Struve, ‘Gerhard I., bischof von

Cambrai’, in Lexikon des Mittelalters, vol.

4.6 (Stuttgart, 1989), cols 1311–12. For longer consid-

erations, see T. Schieffer, ‘Ein deutscher Bischof des

11. Jahrhunderts: Gerhard I. von Cambrai

(

1012–1051)’, Deutsches Archiv 1 (1937), pp. 323–60 and H. Sproemberg, ‘Gerhard I., Bischof von

Cambrai (

1012–1061 [sic])’, in M. Unger (ed.), Mittelalter und demokratische Geschichtsschrei-

bung. Ausgewählte Abhandlungen, Forschungen zur mittelalterlichen Geschichte

18 (Berlin

[East],

1971), pp. 103–18. The see was based in Cambrai (and will be referred to as such

henceforth), but in the eleventh century it was believed to have united two older dioceses in
Cambrai and Arras; whether this was in fact the case is questionable: L. Kéry, Die Errichtung des
Bistums Arras

1093/94, Beiheft der Francia 33 (Sigmaringen, 1994), pp. 211–25.

See I. Voss, ‘La rencontre entre le roi Robert II et l’empereur Henri II à Mouzon et Ivois en

1023.

Un exemple des relations franco-allemandes au Moyen Age’, Annales de l’Est

1 (1992), pp. 3–14.

Early Medieval Europe

2010 18 (2) 202–222

2010 Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350

Main Street, Malden, MA

02148, USA

from one of his canons.

Soon after, he had the same author write a

history of the diocese, the Gesta episcoporum Cameracensium, drawing
inspiration from Flodoard’s history of Reims (Cambrai’s metropolitan).

The building programme and literary production bespeak a confident
bishop, eager to inscribe his authority onto space and time.

But by

1025,

when the canon came to write about recent events, the situation had
worsened. Henry II was dead, as was Gerard’s other close political ally,
Duke Godfrey II of Lower Lotharingia (

1012–23). The duke’s successor,

along with much of the Lotharingian aristocracy, was in rebellion against
the new king, Conrad II (

1024–39; called ‘the Elder’ to distinguish him

from the rival claimant, Conrad ‘the Younger’).

Gerard’s diocese, pre-

cariously straddling the border between France and the empire and in the
gift of the German kings but canonically subject to the archbishop of
Reims, was vulnerable to attacks from a number of directions. The count
of Flanders was a constant threat and a French-backed castellan, Walter II
of Lens, even competed with the bishop for control of Cambrai and its
hinterland.

It was in this context of rapidly altered fortune that the

For the construction of the cathedral, see M. Rouche, ‘Cambrai, du comte mérovingien à
l’évêque impérial’, in L. Trenard (ed.), Histoire de Cambrai (Lille,

1982), pp. 37–8. For the author

of the third version of the Vita Gaugerici being the same as that of the Gesta, see already P. Van
den Bosch, Acta altera S. Gaugerici episcopi confessori, AA SS Aug II (Antwerp,

1735), pp. 668–70.

The dating of the Gesta is complex: see E. van Mingroot, ‘Kritisch onderzoek omtrent de
datering van de Gesta episcoporum Cameracensium’, Revue belge de philologie et d’histoire

53.2

(

1975), pp. 281–332 and T. Riches, ‘Episcopal Historiography as Archive: Some Reflections on

the Autograph of the Gesta episcoporum Cameracensium (MS Den Haag KB

75 F15)’, Jaarboek

voor middeleeuwse Geschiedenis

10 (2007), pp. 7–46. At this time Gerard may also have been

involved in the production of a Giant Bible at the monastery of St-Vaast, Arras; see D.J. Reilly,
The Art of Reform in Eleventh-Century Flanders. Gerard of Cambrai, Richard of Saint-Vanne and
the Saint-Vaast Bible (Leiden,

2006). Reilly’s interpretation of the illustrations relies on a certain

interpretation of Gerard’s political position. To cite Reilly’s work in any detail as evidence for
Gerard’s ideological position therefore threatens to make the argument circular: see, further, T.
Riches, ‘Review of Reilly, Diane J., The Art of Reform in Eleventh-Century Flanders: Gerard of
Cambrai, Richard of Saint-Vanne, and the Saint-Vaast Bible’, H-German, H-Net Reviews
(September,

2009), <http://www.h-net.org/reviews/showrev.php?id=25209>.

The Gesta was personally commissioned by Gerard and completed up to

1025 in his lifetime at

his cathedral community; see Riches, ‘Episcopal Historiography’, p.

16. In addition, Gerard can

be shown to have collaborated in the writing by providing the author with information; see
Gesta episcoporum Cameracensium, ed. L. Bethmann, MGH SS

7 (Hanover, 1846) [henceforth

Gesta epp. Cam.], III, c.

12, p. 469, c.22, p. 472. This article will therefore treat it as a safe

assumption that the Gesta reflects Gerard’s own views.

For the events around

1024/5, see, with care, H. Pabst, ‘Frankreich und Konrad der Zweite in

den Jahren

1024 und 1025’, Forschungen zur deutschen Geschichte V (1865), pp. 337–68 and below

pp.

13–17. The designations ‘Elder’ and ‘Younger’ are contemporaneous: see Wipo, Gesta

Chuonradi imperatoris, ed. H. Bresslau, MGH SRG

61 (Hanover, 1915), c. 2, p. 15.

For Flemish expansion, see F.L. Ganshof, ‘Les origines de la Flandre Impériale. Contribution à
l’histoire de l’ancien Brabant’, Annales de la Société Royale d’archéologie de Bruxelles. Mémoires,
Rapports et Documents

96 (1942/43), pp. 99–171. For the conflict with Walter and its long history,

see S. Patzold, ‘. . . inter pagensium nostrorum gladios vivimus. Zu den “Spielregeln” der Konf-
liktführung in Niederlothringen zur Zeit der Ottonen und frühen Salier’, Zeitschrift der
Savigny-Stiftung für Rechtsgeschichte, Germanistische Abteilung

118 (2001), pp. 578–99.

The struggle for the German throne, 1023–5

203

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

author of the history, the Gesta episcoporum Cameracensium, looked back
on the events of the previous, happy years.

In particular, the anonymous canon of Cambrai took the opportunity

to lambast the French king as weak and his bishops as usurping royal
rights by promoting an ‘unprecedented’ initiative, by which people were
to swear an oath to keep the peace. This was, in other words, an instance
of the phenomenon we now know as the Peace of God.

The Gesta’s

criticism has played a significant role in the historiography of the French
monarchy and the Peace, fitting, as it does, into grand narratives of weak
post-Carolingian kingship and increasingly assertive ecclesiastical reform-
ers.

It is odd, then, that so little has been written on the source itself,

either with regards to the circumstances of its composition or the reli-
ability of its claims.

This article will seek to fill this gap by examining

both issues. Firstly, it will seek to reconstruct the events around the Peace
oath which the Gesta so harshly attacks. Having established that the
Gesta’s account is unreliable, it will investigate the background against
which the account was written to determine why the author might have
wished to depict the king in the way that he did.

After describing an incident datable to the period following the death

of Archbishop Arnulf of Reims (

1021), the Gesta continues:

At that time Bishops Berold of Soissons and Warin of Beauvais saw
that on account of the weakness of the king and in payment for
[mankind’s] sins, the stability of the realm was disrupted to its foun-
dation, law was disordered, and the customs of the fathers and every
kind of justice were profaned. They thought that it would be a great
help to public order if they copied the idea of the bishops of Burgundy.
The latter, certainly lacking all authority, together made a decree that
obliged by oath everyone, including themselves, to serve peace and
justice . . . [Gerard argued] that they presumed to lay claim to what is

The name is a modern coinage; the medieval nomenclature was much less categorical. See H.
Hoffmann, Gottesfriede und Treuga Dei, Schriften der MGH

20 (Stuttgart, 1964), pp. 3–4.

This is particularly true in the work of G. Duby. See, most importantly, his ‘Gérard de Cambrai,
la paix et les trois fonctions sociales,

1024’, in Comptes rendus des séances de l’Académie des

Inscriptions et Belles-Lettres (Paris,

1976), pp. 136–46 and idem, The Three Orders. Feudal Society

Imagined, trans. Arthur Goldhammer (Chicago,

1980).

Work has begun on the Gesta in recent years. See L. Jégou, ‘L’évêque entre authorité sacrée et
exercice du pouvoir. L’exemple de Gérard de Cambrai (

1012–1051)’, Cahiers de civilisation

médiévale

47 (2004), pp. 33–55; R.M. Stein, Reality Fictions. Romance, History, and Governmental

Authority,

1025–1180 (Notre Dame, 2006), pp. 13–35; and T. Riches, ‘Bishop Gerard I of

Cambrai-Arras, the Three Orders and the Problem of Human Weakness’, in John S. Ott and
Anna Trumbore Jones (eds), The Bishop Reformed: Studies of Episcopal Power and Culture in the
Central Middle Ages (Aldershot,

2007), pp. 122–36. Stein’s work is a slightly extended version of

idem, ‘Sacred Authority and Secular Power: The Historical Argument of the Gesta episcoporum
Cameracensis [sic]’, in L. Besserman (ed.), Sacred and Secular in Medieval and Early Modern
Cultures: New Essays (New York,

2006), pp. 149–65.

204

Theo Riches

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

royal law. In this way they would throw into disorder the stability of
the holy church, which must be governed by twin persons, that is to
say the royal and the priestly. It is assigned to the latter to pray and to
the former to fight. Therefore it is for kings to curb rebellions with
their prowess, suppress wars, and spread peaceful relations, and for
bishops to urge kings to fight courageously for the preservation of the
fatherland and to pray that they win.

It has been suggested that the argument about the respective spheres of

royal and episcopal power is disingenuous, even hypocritical: the same
author who reported Gerard’s objections to bishops usurping royal pre-
rogatives also approvingly described the tenth-century Bishop Rothard of
Cambrai leading a small army to destroy the castle of Otto of Vermandois
at Vinchy, and Gerard was himself the holder of numerous royal rights,
including possession of the county of Cambrésis.

But this accusation

cannot be sustained without further evidence being taken into consider-
ation. The bishops of Cambrai had been granted their royal rights by the
German kings and emperors; Rothard’s troops belonged to the ‘fideles
imperatoris’, Counts Godfrey of Eename and Arnulf of Valenciennes.

What the Gesta seems to be describing in the case of Berold and Warin is,
in contrast, an unauthorized power grab. An accusation of hypocrisy is
only valid if the Gesta is hiding something about the circumstances of
Berold and Warin’s peace initiative. As it happens, this is precisely what
the Gesta is doing.

We know this because there is key external evidence of the French

bishops’ proposal: we have the text of the oath and it has a number of

‘Ipso in tempore videntes episcopi Beroldus Suessionensium, et Walerannus Belvacensium, prae
inbecillitate regis peccatis quidem exigentibus statum regni funditus inclinari, iura confundi,
usumque patrium et omne genus iustitiae profanari: multum rei publicae succurrere arbitrati
sunt, si Burgundiae episcoporum sententiam sequerentur. Hii nimirum totius auctoritatis
expertes, commune decretum fecerunt, ut tam sese quam omnes homines sub sacramento
constringerent, pacem videlicet et iustitiam servaturos . . . quod regalis iuris est, sibi vendicari
presumerent. Hoc etiam modo sanctae aecclesiae statum confundi, quae geminis personis,
regali videlicet ac sacerdotali, administrari precipitur. Huic enim orare, illi vero pugnare
tribuitur. Igitur regum esse, seditiones virtute compescere, bella sedare, pacis commercia dila-
tare; episcoporum vero, reges ut viriliter pro salute patriae pugnent monere, ut vincant orare.’
Gesta epp. Cam. III, c.

27, p. 474.

Gesta epp. Cam. I, c.

103, pp. 443–4. The hypocrisy of Gerard’s accusations was already spotted

by Hoffmann, Gottesfriede, pp.

58–9. See also idem, ‘Grafschaften in Bischofshand’, Deutsches

Archiv

46 (1990), pp. 391–2, 474–6.

Gesta epp. Cam. I, c.

103, p. 443. Description as ‘fideles imperatoris’: ibid., I, c. 96, p. 440.

Earlier the Gesta claimed, wrongly, that Godfrey and Arnulf had their counties at the behest of
Bruno of Cologne, Otto the Great’s brother and ‘archduke’ of Lotharingia: Gesta epp. Cam. I,
c.

95, pp. 439–40. See also H. Franz-Reinhold, ‘Die Marken Valenciennes, Eename und

Antwerpen im Rahmen der kaiserlichen Grenzsicherungspolitik an der Schelde im

10–11.

Jahrhundert’, Rheinische Vierteljahrsblätter

10 (1940), pp. 229–76.

The struggle for the German throne, 1023–5

205

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

characteristics that contradict the Cambrai gesta’s account.

Firstly, like

other oaths before it, exception is made for people participating in royal
or episcopal hosts.

In other words, it places no restriction on royal or

episcopal military action. Secondly, and most significantly, it seems to
have been formulated in the king’s and others’ presence.

The specific

occasion is generally accepted to have been the royal assembly in Com-
piègne on

1 May 1023/4.

While there is no explicit mention of the oath

in our records for the assembly, we can presume Berold was there, since
Compiègne lay in his diocese, and we know that Warin was present, since
it was there that he met and came to an agreement with Abbot Leduin of
St-Vaast, whom the Cambrai author describes as supporting Berold and
Warin’s peace proposal.

In addition, a number of other important

magnates were at the assembly, including Baldwin IV of Flanders, Bishop
Fulk I of Amiens, Richard II of Normandy, and the latter’s brother
Archbishop Robert of Rouen, so this would have been an ideal time for
any peace proposal to be discussed.

Most significantly, Gerard I was also present in Compiègne as part of

an embassy to Robert the Pious which he undertook with Abbot Richard
of St-Vanne, an embassy which was almost certainly part of the prepa-
rations either for the meeting of Robert and Henry II at Mouzon and

C. Pfister, Études sur le règne de Robert le Pieux (

996–1031), Bibliothèque de l’École des Hautes

Études

64 (Paris, 1885), pp. lx–xi.

Pace Hoffmann, Gottesfriede, p.

57, there is no reason to believe this refers to armies summoned

to deal with those who broke the peace oath, rather than royal and episcopal armies more
generally.

After the oath proper, the text continues: ‘Haec superius scripta in his verbis sequentibus jurata
sunt. Hoc audias, tu, rex Roberte, sicut in hoc brevi superius scriptum et sicut ego Warinus
episcopus hac hora novissima derationavi et sicut isti circumstantes nunc audierunt et intellex-
erunt. Sic attendam de mea parte contra illos qui hoc sacramentum hoc praesenti tempore
juraverunt et jurabunt, ab hinc usque festivitatem sancti Johannis quae est futura in mense
junio et de illa festivitate in sex annos, nisi de werra regis me sciente. Sic me Deus adjuvet et
haec sacra . . . ’ Pfister then notes, ‘encore trois ou quatre mots illisibles’. Études, p. lxi.
Hoffmann, following G. de Manteyer, suggests that the remaining words are, ‘sanctorum Dei
pignora. Amen’: Hoffmann, Gottesfriede, p.

57, n. 34.

The assembly is probably to be dated to

1 May 1023, but M. Prou has shown that it may have

taken place in

1024, cf. idem, ‘Une charte de Garin évêque de Beauvais. L’assemblée de 1023 ou

1024’, in Société nationale des antiquaires de France (ed.), Centenaire 1804–1904. Receuil de

mémoires publiés par les membres de la Société (Paris,

1904), pp. 383–98. If we are to date Warin’s

oath to this meeting, then its placing in the Gesta epp. Cam., before the description of the
dispute between Pilgrim of Cologne and Durand of Liège over the monastery of Burtscheid in
July

1023, would indicate that the 1023 date is the more likely.

Pfister, Études, p. lxxx; Gesta epp. Cam. III, c.

27, p. 474.

Hoffmann, Gottesfriede, p.

56. We know who attended from the witness list of Warin’s charter,

discussed by Prou; see idem, ‘Une charte’, p.

398. The list includes Gerard’s long-running rival

for control of Cambrai, the castellan Walter II of Lens, but not Odo II of Blois who was
therefore probably absent. For the meeting itself, see R. Bonnaud-Delamare, ‘Les institutions de
paix dans la province ecclésiastique de Reims au XIe siècle’, Bulletin philologique et historique du
Comité des travaux historiques et scientifiques,

1955–6 (Paris, 1957), p. 145. Fulk I’s successor and

relative, Fulk II of Amiens, would later (

1033–4) become involved in a peace initiative with

Corbie, see Riches, ‘Bishop Gerard I of Cambrai-Arras’, pp.

125–6.

206

Theo Riches

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Ivois in August

1023 or for the summit at Pavia with the Pope that was

supposed to result from it.

In either case, this supports the supposition

that the Compiègne assembly was the occasion for Berold and Warin’s
peace initiative, since both the Mouzon/Ivois meeting and the projected
meeting with the Pope were explicitly to do with establishing peace and
justice and ensuring religious renewal.

For all these reasons the histo-

riographical consensus that Compiègne was the occasion for Berold and
Warin’s proposal is almost certainly correct. We can therefore conclude
that the oath was done with the support of the king and that Gerard
knew this full well. In other words, the peace initiative of Berold and
Warin no more constituted an illegitimate usurpation of royal right than
Gerard’s own title of count of the Cambrésis. Still, the fact that the
Compiègne initiative certainly did have the support of the king (and
many other magnates) does not in itself falsify the Cambrai gesta’s claims
either. Gerard never explicitly levels the accusation that Robert was
opposed to the peace, only that he was weak and that the bishops
usurped his rights. It is perfectly possible that whereas the bishop
of Cambrai’s rights were granted to him by his king from a position of
strength, Berold and Warin’s initiative was foisted on Robert in a time of
weakness. Only a further examination of the context of Compiègne can
resolve the issue.

As the Gesta claims, Berold and Warin had copied their idea from the

bishops of Burgundy, specifically those at a council in Verdun-sur-le-
Doubs which Berold had attended and where an oath was sworn which
closely matches that which he and Warin proposed at Compiègne.

But

here, too, there are indications that the Gesta’s association of the Peace of
God with royal weakness is at least an exaggeration, if not a downright
fabrication. Like its northern offspring, the Verdun-sur-le-Doubs oath
explicitly reserves the right to exact military rights when participating in

As described in Gesta epp. Cam. III, c.

37, p. 480, with the former being the case if the

Compiègne assembly took place in

1023 and the latter if in 1024. But the Gesta strongly implies

that Gerard was already acting in an official capacity at Ivois, and there is little time for Gerard
to have been involved afterwards, since he was busy dealing with the castellan of Cambrai
Walter II and with rebuilding the cathedral: Gesta epp. Cam. III, cc.

38–39, 49, pp. 481, 483. This

makes a

1023 dating for Compiègne much more likely. That Henry II was at Cologne on 16 May

1023, but ill for much of April/May 1024 is additional evidence for a 1023 date: Regesta Imperii

II. Sächsisches Haus

919–1024. 4: Die Regesten des Kaiserreiches unter Heinrich II. 1002–1024, ed.

T. Graff (Vienna,

1971), nos. 1:2038–9; Annnales Quedlinburgenses, ed. Martina Giese, MGH

SRG

72 (Hanover, 2004), an. 1024, p. 574.

‘Ibi certe pacis et iusticiae summa diffinitio mutuaeque amicitiae facta reconciliatio; ibi quoque
diligentissime de pace sanctae Dei aecclesiae maxime tractatum est, et quomodo christianitati,
quae tot lapsibus patet, melius subvenire deberent’, Gesta epp. Cam. III, c.

37, p. 480.

The text of the oath was published by Leclercq in K.J. von Hefele, Histoire des conciles d’après
les documents originaux,

2nd edn, trans. H. Leclercq (Paris, 1911), vol. 4.2, pp. 1409–10. For a

discussion of Warin’s oath and its relationship to that of Verdun-sur-le-Doubs, see Bonnaud-
Delamare, ‘Les institutions de paix’, pp.

148–56.

The struggle for the German throne, 1023–5

207

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

royal or episcopal hosts.

In addition, the political context of the Verdun-

sur-le-Doubs council and its successors is described in the gesta of
the bishops of Auxerre in very different terms from the account of the
Cambrai author: instead of royal weakness, we are told about how the
rebellious aristocracy of the duchy of Burgundy submitted to Robert and
his ally Bishop Hugh of Auxerre. Indeed, according to the Auxerre gesta,
Verdun-sur-le-Doubs was the first of a number of councils held in the
Burgundian duchy, including ones in the counties of Auxerre (at Héry),
Dijon, Beaune and Lyon.

In this reading, the peace of Verdun-sur-le-

Doubs, far from being a symptom of royal impotence, represents the
imposition of royal power on a formerly recalcitrant province.

But the sources cannot be read so easily, since the Auxerre gesta is just

as unreliable as its Cambrai counterpart, if not more so. This is especially
true of its chronology. The council of Verdun-sur-le-Doubs took place
sometime between

1018 and 1022.

The Auxerre account places it ‘some

Both E. Magnou-Nortier, ‘Les mauvaises coutumes en Auvergne, Bourgogne méridionale,
Languedoc et Provence au XI

siècle: un moyen d’analyse sociale’, in Structures féodales et

féodalisme dans l’Occident méditerranéen (X

– XIII

siècles). Bilan et perspectives de Recherches,

Collection de l’École française de Rome

44 (Rome, 1980), p. 144 and H.-W. Goetz, ‘La paix de

Dieu en France autour l’an Mil: fondements et objectifs, diffusion et participants’, in Michel
Parisse and Xavier Barral i Altet (eds), Le roi de France et son royaume autour l’an Mil. Actes du
colloque Hugues Capet

987–1987, La France de l’an Mil, Paris–Senlis, 22–25 juin 1987 (Paris, 1992),

pp.

132–3 make this a centrepiece of their respective arguments.

‘Post aliquot uero annos dierum, iam dicti primores Burgundie in deditionem regis pacifice
deuenerunt. Per Hugonis tamen consilium, quicquid isdem rex facere decreuerat disposuit,
eique quod accipere placuit libentissime condonauit. Constituit ergo pro redintegranda seu
firmanda pace concilium episcoporum, ac multorum tam nobilium quam plebeiorum innu-
mere multitudinis, in comitatu quem gubernat Cabilonense, in loco qui Viridunus dicitur, ubi
etiam sanctorum reliquie diuersis regionibus delate diuersas sanitates egrotis contulerunt.
Itemque aliud concilium pro eadem re in pago Autissiodorense apud Airiacum uillam beati
Germani, in quo rex Robbertus cum episcopis et abbatibus adfuit, simul et innumerabilia
sanctorum pignera totius pene prouincie . . . Similiter et per diuersa loca uel in pagis Diuion-
ense et Belnense atque Lugdunense concilia sepius celebrari fecit.’ Gesta pontificum Autissiodo-
rensium, vol.

1, ed. M. Sot, G. Lobrichon and M. Goullet, Les classiques de l’histoire de France

au Moyen Âge

42 (Paris, 2006), pp. 251.

The council seems to have been held in late spring or early summer, since the participants
promise to abide by their oath until the impending feast of John the Baptist (

24 June) and for

seven years after that: ‘Haec omnia supradicta adtendam usque ad praesentem festivitatem
Sancti Johannes Baptistae, et ab illa usque in septem annis’, where ‘praesentem’ is understood
to mean ‘impending’ or ‘imminent’; von Hefele, Histoire des conciles, pp.

1409–10. Dates of 1023

or even

1024 are theoretically possible but unlikely; see below, n. 34. It is unclear why F.S.

Paxton thinks the traditional

1016 date is still acceptable: idem, ‘History, Historians, and the

Peace of God’, in T. Head and R. Landes (eds), The Peace of God: Social Violence and Religious
Response in France around the Year

1000 (Ithaca, 1992), p. 37, n. 79. Both Duby and Magnou-

Nortier, whom he cites for this dating, rely on Bonnaud-Delamare, who in turn does not
question Leclercq. Indeed, in an early article Magnou-Nortier noted that Leclercq’s dating
probably needed revision; see E. Nortier, ‘La foi et les convenientiae. Enquête lexicogaphique
et interpretation sociale’, in D. Buschinger (ed.), Littérature et société au Moyen Âge. Actes du
colloque des

5 et 6 mai 1978. Université de Picardie. Centre d’Études médiévales (Paris, 1978), p. 262,

20. The 1016 dating seems to rest on the assumption that Robert’s imposition of authority

on the duchy immediately followed the capture of Dijon in

1015 and the death of his enemy

208

Theo Riches

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

years’ after Robert besieged Auxerre (its bishop, Hugh, having been
expelled by Landric of Nevers in the immediate aftermath of Duke Henry
of Burgundy’s death in

1002), but that scarcely helps us since the siege of

Auxerre seems to have taken place c.

1005.

The oath makes no mention

of the king being there and neither does the gesta of Auxerre, which
attributes both the Verdun-sur-le-Doubs council and a number of later
ones to Hugh.

As Hartmut Hoffmann pointed out some time ago, this

is clearly a distortion of events which ignores the role of other bishops in
favour of the Auxerre gesta’s hero.

Both the attempt to connect the

council with a royal campaign waged about fifteen years earlier and its
attribution to the bishop are clearly designed as a justification of Hugh’s
pro-royal policies.

If we are to distrust the Cambrai gesta’s depiction of

the Peace as undermining the French monarchy, we should exercise
precisely the same scepticism towards its Auxerre counterpart’s diametri-
cally opposed version.

There is one further link between the Compiègne and Verdun-sur-le-

Doubs peace initiatives which opens the possibility that neither may have
been to Robert’s advantage: the presence at both of individuals who had
been opposed to Robert and his queen Constance at the Orléans heresy

Bishop Bruno of Langres on

31 January 1016: Pfister, Études, pp. 261–4, and M. Chaume, Les

origines du duché de Bourgogne,

3 vols (Dijon, 1925), I, pp. 476–84. There is also the thirteenth-

century evidence of Aubry of Troisfontaines, who claims Robert made his second son Henry
duke of Burgundy in

1015 (although historians have preferred to date this to 1017, to coincide

with his elder brother Hugh’s coronation): Aubry of Troisfontaines, Chronica, ed. P. Scheffer-
Boichorst, MGH SS

23 (Hanover, 1874), p. 780. Yet this entire construct relies on believing the

claim of the Auxerre gesta that the council was meant to establish Robert’s rule over the
rebellious local aristocracy.

For the background to the council of Verdun-sur-le-Doubs, see Y. Sassier, Recherches sur le
pouvoir comtal en Auxerrois du X

au début du XIII

siècle (Auxerre,

1980), pp. 30–41. Sassier still

dates the council to

1016, but it cannot possibly have taken place before 1018 since one of the

participants, Elmuin of Autun, could not have been bishop before then: Hoffmann, Gottes-
friede, p.

51. For the siege of Auxerre, see J.-L. Chassel, ‘A propos de quelques documents de

Saint-Bénigne de Dijon au

siècle’, Annales de Bourgogne

65 (1993), pp. 147–60, esp. pp. 148–9.

Although the subject of the previous clause is the king, the subject of ‘Constituit’ is the bishop,
as the phrase ‘in comitatu quem gubernat Cabilonense’ demonstrates (Hugh was also count of
Châlons-sur-Saône); see above, n.

24. Note that the Auxerre gesta mentions Robert’s presence

at Héry (for which see below, pp.

7–9), but not at Verdun-sur-le-Doubs. Hoffmann speculates

that Hugh and Berold of Soissons were there as royal representatives: Hoffmann, Gottesfriede,
p.

51.

Hoffmann, Gottesfriede, p.

51. The text of the oath itself attributes the council to Archbishop

Burchard of Lyon: ‘Burchardus Lugdunensis Ecclesiae archiepiscopus hoc pacis foedus in
concilio Verdunensi constituit’, von Hefele, Histoire des conciles, vol.

4, p. 1410.

Frustratingly, the dating of the Auxerre gesta is problematic. We are told that each section,
devoted (unlike in the Gesta episcoporum Cameracensium) to an individual bishop, was based on
notes written up as soon as each bishop died: ‘Altissiodorensis ecclesiae non ignobilis consue-
tudo est, quam cito de seculo migrat eius episcopus, ilico terminum uitae, sedis introitum, ac
precipue bene gesta ipsius conscribere’, Gesta pont. Aut., p.

279. Yet we are also told that Bishop

Geoffrey of Auxerre (

1052–76) ‘hoc uolumen de gestis pontificum renouauit’, ibid., p. 267. As

the most recent editors point out, ‘renewal’ is an ambiguous phrase, but most probably
indicates a reworking of the content: ibid., pp. xxiv–v.

The struggle for the German throne, 1023–5

209

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

trial of December

1022. Robert-Henri Bautier has revealed the political

background of this scandal.

It related directly to the presence of heretics

in Constance’s entourage, but also to the replacement of Robert’s
appointment to the see of Orléans, Thierry, with an ally of Odo II of
Blois, Odalrich, and to the rivalry between Robert and Odo more
widely.

Although again the chronology is uncertain, the inheritance of

Odo’s cousin, Count Stephen of Meaux (d.

1019x1023), had brought the

king and his subject into open aggression at around this time.

Warin had

spoken out at the council against the heretics and Odalrich was present
at Verdun-sur-le-Doubs.

Yet none of this is particularly damning. As Bernard Gowers has

pointed out, Warin, like Fulbert of Chartres, was careful not to become
irrevocably associated with either Robert or Odo.

That Warin once

spoke out on Odo’s side is not very strong evidence that his and Berold’s
Compiègne initiative was meant to undermine Robert. As for Odalrich,
he is only the last of a number of participants listed for Verdun-sur-le-
Doubs, after Archibishop Burchard of Lyon, Bishops Elmuin of Autun,
Lambert of Langres, Gauzlin of Mâcon and Godfrey of Chalon-sur-
Saône, as well as Hugh of Auxerre, Archbishop Walter of Besançon and
Berold of Soissons.

Elmuin, Lambert, Gauzlin and Godfrey were

R.H. Bautier, ‘L’hérésie d’Orléans et le mouvement intellectuel au début du XI

siècle: Docu-

ments et hypothèses’, Actes du

Congrès national des sociétés savantes, Reims

1970. Section de

philologie et d’histoire jusqu’à

1610, I: enseignement et vie intellectuelle (XI

–XVI

siècle) (Paris,

1975), pp. 63–88; see also R.I. Moore, ‘Appendix: Heresy and Politics in Early Eleventh-Century

France’, in idem, The Origins of European Dissent (Oxford,

1985), pp. 285–9.

Bautier, ‘L’hérésie d’Orléans’, pp.

77–80. Moore sees the trial as less of an attack on Robert than

on Constance, for whom Thierry had acted as chaplain: idem, ‘Appendix’, p.

287. Yet this was

also an extension of the conflict between Robert and Odo, since Robert had repudiated Odo’s
mother Bertha as his wife in order to marry Constance and thereby forge a link with the latter’s
cousin and Odo’s rival, Fulk Nerra of Anjou; see G. Duby, The Knight, the Lady and the Priest
(London,

1984), pp. 80–1.

For a maximalist reconstruction of events, see Pfister, Études, pp.

239–43; for a more cautious

approach, based on a redating of Fulbert’s letters, see The Letters and Poems of Fulbert of
Chartres, ed. F. Behrends (Oxford,

1976), pp. lxxix–lxxx.

B. Gowers, ‘Fulbert of Chartres and his Circle: Scholarship and Society in Eleventh-Century
France’, Ph.D thesis, University of Oxford (

2007), p. 186. Fulbert, too, was a supporter of

Odalrich, although he was able to work with Thierry: Fulbert, Letters, nos.

21–22, 26, 40, 42,

pp.

38–45, 48–51, 72–75.

Von Hefele, Histoire des conciles, IV, p.

1410. Odalrich is given the rather odd honorific of ‘in re

honorabilis sacerdos, in spe episcopus venerabilis’. Odalrich’s predecessor, Thierry, was last
active as bishop in January

1021 and Odalrich already witnessed a charter as bishop by June that

same year. This would imply that Verdun-sur-le-Doubs occurred in

1021, but we should not

place much weight on a single, ambiguous phrase, pace Behrends, Letters, p. lxxviii, n.

24. In any

case, Odalrich’s succession seems to have been complicated by the hostility of his archbishop,
Leotheric of Sens, thus a

1022 date remains a distinct possibility: see B. Töpfer, Volk und Kirche

zur Zeit der beginnenden Gottesfriedenbewegung in Frankreich (Berlin [East],

1957), pp. 64–5. If

we interpret the reluctance to give Odalrich an unambiguous episcopal title as reflecting
sensitivities towards the deposed Thierry, then dates of

1023 or even 1024 are possible, since pace

Töpfer, ibid., and, following him, Hoffmann, Gottesfriede, p.

51, Thierry may not have died in

210

Theo Riches

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

suffragans of the Archbishop of Lyon, who, according to the recorded
oath, was in fact the one who arranged the meeting.

For this reason the

council is best interpreted as first and foremost a provincial synod.

Indeed, Odalrich is added to the end almost as an afterthought. In
contrast, Lambert of Langres, who was much more prominent, seems to
have been an ally of Robert.

Robert’s links to the Verdun-sur-le-Doubs council therefore appear to

be ambiguous or weak and our evidence for the Compiègne meeting is
contradictory. Similar issues plague our sources regarding one of the
councils with which the Auxerre gesta links that of Verdun-sur-le-Doubs:
the council of Héry, just north of Auxerre. Like Verdun-sur-le-Doubs, the
Auxerre source describes Héry as an expression of royal strength. Unlike
Verdun-sur-le-Doubs, we have other narrative descriptions of what
occurred at Héry, notably in the Miracula s. Bercharii, Miracula s. Veroli
and Chronicon of Saint-Pierre-le-Vif in Sens.

On the one hand, they all

make a point of Robert’s participation, but on the other each is shot
through with its respective agenda. The Chronicon attributes the council
to the archbishop of Sens, relegating Robert to a supporting role,

while

the two miracula texts explicitly downplay the effectiveness of secular
power, even where this weakness is interpreted positively as a sign of
Robert’s gentleness.

1022 but could have survived to January 1024; see Behrends, Letters, p. lxxviii. But a later date

for the council seems unlikely, since Odalrich was probably fully in place by the time of the
heresy trial in December

1022.

See above, n.

28.

As Hoffmann pointed out, Gottesfriede, p.

51.

Chaume, Les origines, vol.

1, pp. 482–3.

The Chronicon Autissiodorense is sometimes quoted as a source, but this is much later (early
thirteenth century) and at least partially dependent on the Gesta pont. Aut.: see Recueil des
historiens des Gaules et de la France

10, ed. D.M. Bouquet and L. Delisle (Paris, 1760; repr. 1874),

275, c–d, and note b.

‘In tempore illo factus magnus conventus in Autissiodorensi pago, in Airiaca villa, a Leoterico
archiepiscopo . . . Ibi vero presentes fuerunt Rotbertus rex et Goslinus Bituricensis archiepis-
copus cum aliis episcopis et abbatibus et populo innumerabili.’ Chronique de Saint-Pierre-le-Vif
de Sens, dite de Clarius, ed. R.-H. Bautier and M. Giles (Paris

1979), pp. 114–16. The

archbishop’s prominence is easily explained if Bautier is right that this section of the (early
twelfth-century) Chronicon drew on a now lost mid-eleventh-century gesta of the archbishops
of Sens; see ibid., pp. ix–x, xxxvii–viii.

‘Gloriosus rex Robertus, apud villam Aireyas nomine noscitur concilium habuisse . . . ipse rex
mitissimus prae cunctis existebat, et magis mansuetudine quam feritate suos vincere
cernebat . . . ’, Miracula s. Bercharii, AASS Oct. VII.

2, V, c. 33, col. 1029, to be dated to not long

after

1085: idem, col. 989–90; ‘. . . religiosus princeps Robertus nitens pacem, sicquo modo

posset, inter dissidentes componere, iussit cunctos valentes episcopos occurrere et abbates apud
Airyacum, villam in Autissiodorensi dioecesi sitam et cum sanctorum pignoribus adesse;
quatenus si malitiae amatores minus libenter, pro terreni principatus districtione pacificari
vellent; saltem pro Dei et sanctorum eius, quos praesentes seque quodammodo exspectantes
viderent, timore, pacis concordiam et promptius firmandam exciperent; et sanctorum, in
quorum praesentia firmassent semper memores, irruptam arctius conservarent’, Miracula s.
Veroli, AASS June III, II, c.

6, col. 385. The Miracula s. Veroli are difficult to date, but seem to

The struggle for the German throne, 1023–5

211

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

We are, however, on considerably firmer ground with the chronology

of the council of Héry than we are with that of Verdun-sur-le-Doubs.
Héry is generally dated to early

1024, since 5 April that year is the terminus

ante quem for a royal diploma made out at Avallon, south-east of Auxerre,
and witnessed by ‘Bishop Godfrey of Châlons-sur-Saône and the remain-
ing bishops who were recently at the council held at Héry’.

This could

put the council towards the end of

1023,

but in any case it was held fairly

soon after the Mouzon/Ivois meeting which the Cambrai gesta is so
anxious to highlight. The chronology is important because it fits with
Robert’s other conciliar activity. Excepting the council of Verdun-sur-le-
Doubs, which, as has been shown, may be only weakly related to Robert’s
own activities, the councils of Compiègne, Mouzon/Ivois and Héry are
all clearly linked with Robert and all occurred within the twelve-month
period from May

1023 to May 1024 and therefore follow closely on the

heels of the Orléans trial of December

1022.

There is thus a strong,

circumstantial reason to believe that these events were related and that
Robert’s activity in

1023–4 sees him seizing the initiative in promoting

orthodoxy after the embarrassing loss in Orléans at the hands of Odo of
Blois.

Certainly Robert tried to use the Mouzon/Ivois meeting to win

Henry II’s support against Odo, since the Gesta episcoporum Cameracen-
sium mentions that representatives of the French king went with the

be from the later eleventh century: a reference to a severe drought (also found in the Miracla s.
Bercharii, V, cc.

36–37, col. 1030) may be to that mentioned in the Chronique de Saint-Pierre-

le-Vif de Sens and elsewhere for c.

1078; see Chronique de Saint-Pierre, p. 133 and n. 2. The most

famous example of Robert being depicted as a ‘rex mitissimus’ is, of course, Helgaud of Fleury’s
Epitoma vitae regis Rotberti Pii, ed. R.-H. Bautier and G. Labory (Paris,

1965), esp. chs 4–5, 10,

pp.

62–5, 72–5, quotation at p. 62. The Epitoma is also the earliest of the texts cited here, dating

to shortly after Robert’s death: Bautier, ibid., pp.

36–7.

‘interveniente et subscribente Gosfrido Cabilonensi episcopo cum reliquis episcopis qui inter-
fuerunt concilio nuper Ariaco habito’, PL

141, ed. J.P. Migne (Paris, 1853), col. 965. See

Catalogue des Actes de Robert II Roi de France, ed. W.M. Newman (Paris,

1937), no. 60, pp. 77–8,

esp. p.

77, n. 1 for the date. See also Pfister, Études, p. 172, n. 2 and M. Bur, La formation du

comté de Champagne, v.

950–v.1150, Mémoires des annales de l’Est 54 (Nancy, 1977), p. 163, who

mistakenly dates it to before

25 March. In fact, we have a royal diploma purportedly from the

council itself, confirming a donation to Fruttuaria by Bishop Hugh of Auxerre, but unfortu-
nately its date –

1021, with no indiction, and in the 31st year of Robert’s reign (i.e. 1018) – makes

no sense. See Newman, ibid., no.

57, pp. 74–5, esp. p. 74, n. 1 for the dating clause and the

source of Bur’s dating mistake.

As did G.D. Mansi, Sacrorum conciliorum nova amplissimia collectio

19 (Venice, 1774), cols

387–90.

The most likely chronology is that the Compiègne assembly happened on

1 May 1023, the

Mouzon/Ivois summit in August of that year, and Héry in the last months of

1023 or the first

1024. Even if Compiègne took place at the latest possible date of 1 May 1024 (see above, n.

17), this would still place all the councils within a ten-month period of August 1023 to the

beginning of May

1024; either way, the argument stands.

This is the interpetation of J.-F. Lemarignier, ‘Paix et réforme monastique en Flandre et en
Normandie autour de l’année

1023. Quelques observations’, in Droit privé et institutions région-

ales. Études historiques offertes à Jean Yver (Paris,

1976), pp. 448–54, but Lemarignier misses the

implications for the Gesta episcoporum Cameracensium account.

212

Theo Riches

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

emperor from there to Verdun to meet and hear Odo’s response to
accusations Robert had levelled against him.

The activities of

1023–4 can

therefore be seen, contrary to the view of the Cambrai gesta, as a sign of
(albeit short-lived) French royal strength – Robert was mustering support
from his bishops and even from the German king in his power struggle
with Odo. Normally, this would not necessarily be problematic for an
imperial magnate like Gerard of Cambrai: until

1023 Odo was, after all,

at war with Duke Dietrich of Upper Lotharingia

and a strong French

king who could control his greatest subjects might work to Cambrai’s
benefit, as it had done when Robert allied with Henry II against the
count of Flanders.

Unfortunately for Gerard, relations between Robert

and Odo were destined to improve at the precise time when it was
dangerous for the bishop of Cambrai.

There is no mention of Henry II settling the hostilities between Odo

and Robert either at Verdun or later,

but the two were certainly at peace

again by September

1025, since a series of letters from Fulbert of Chartres

to Robert and his queen, including one which intervenes on behalf of
their son Hugh (d.

17 September 1025), assumes friendly relations

between them and Odo.

The reconciliation seems to have come about

via Duke William V of Aquitaine, since Robert was at one of Odo’s
residences at Tours when, on William’s intervention, he confirmed a
charter issued that March at a large gathering of Aquitainian notables
under William’s leadership.

Although Odo does not seem to have been

Gesta epp. Cam. III, c.

38, p. 480. Fulbert of Chartres had been invited to the meeting but

excused himself, presumably so as not to offend Odo; see idem., Letters, no.

81, pp. 146–7.

See R. Parisot, Les origines de la Haute-Lorraine et sa première maison ducale (

959–1033) (Paris,

1909), pp. 400–7.

Gesta epp. Cam. I, c.

33, p. 414; I, c. 114, p. 452.

Henry did settle the conflict between Odo and Dietrich, in part by destroying some castles Odo
had ‘unjustly’ built: Gesta epp. Cam. III, c.

38, p. 481.

Fulbert, Letters, nos.

99–101, pp. 178–85 (no. 101, pp. 182–5 intervenes for Hugh). The Annals

of Vendôme do record peace being agreed between the parties in

1025: Annales Vindocinenses, ed.

L. Halphen, Recueil d’annales angevines et vendômoises (Paris,

1903), p. 60, but unfortunately the

relevant entry has been redated to

1026: ibid., p. 60, n. 2. The traditional dating to 1025 can still

be found in some quarters, most recently by H. Wolfram, Conrad II,

990–1036. Emperor of Three

Kingdoms, trans. D.A. Kaiser (University Park, PA,

2006), p. 74, following on from old

commentaries by scholars such as H. Pabst (‘Frankreich und Konrad’, p.

357) and H. Bresslau

(Jahrbücher des deutschen Reichs unter Konrad II, Bd.

1 (1024–1031) (Leipzig, 1879), p. 77, n. 4).

Yet it is rarely pointed out that the annals do not place this pax in the context of conflicts
between Robert and Odo, let alone between France and Germany (pace Pabst, Bresslau and
even Wolfram; but see Behrends, Letters, pp. lxxx–xxxi, n.

30), but instead in that of allowing

Odo to attack Fulk Nerra of Anjou unhindered. In other words, the author of the Vendôme
annals only notes the pax when it becomes of interest to his own narrative.

Newman, Catalogue des Actes, no.

64, pp. 80–1. The latter charter concerns a new foundation

made by Hugh of Lusignan; see Chartes de l’abbaye de Nouaillé de

678 à 1200, ed. P. Monsabert

(Poitiers,

1936), no. 104, pp. 172–4, and for Robert’s confirmation: no. 106, pp. 176–7. William

of Aquitaine was in Italy by August

1025, thereby providing a terminus ante quem for Robert’s

diploma: see below, n.

57. The occasion on which the original transaction took place must have

The struggle for the German throne, 1023–5

213

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

present at the March assembly, it was probably linked in some way to
negotiations between William and an anti-Conrad faction of Lombards,
and it was these negotiations that were apparently the catalyst for Odo to
seek peaceful relations with Robert.

Thus their relationship changed in

the period encompassing the death of Henry II and the drawn-out
accession of Conrad II (who was only recognized by the Lotharingians in
December

1025), and during which the Gesta episcoporum Cameracensium

was written.

An old school of nationalist historiography argued that after the

death of Henry II on

13 July 1024 there was a French conspiracy to seize

Italy, Lotharingia and the kingdom of Burgundy from the newly
crowned Conrad II.

While this certainly forces dynastic politics into a

nineteenth-century, nation-state mould, Conrad did face threats to his
power in these areas, involving magnates from the west. The evidence
is unfortunately patchy: little is known for sure about the transition
from Henry II to Conrad II, but this does not mean there was no
danger. As Timothy Reuter pointed out, the apparent smoothness of
the succession may be largely owing to the absence of any commentator
such as Thietmar of Merseburg to describe any unrest.

Later sources,

in particular Wipo’s Gesta Chuonradi II imperatoris, of course empha-
sized the naturalness of Conrad’s acquisition of the throne.

The evi-

dential gaps notwithstanding, we still know of three challenges to the

been an impressive event, since Hugh’s charter is witnessed by William, his two sons and his
wife, William of Angoulême and his son, Bishops Isembert of Poitiers, Islo of Saintes, Roho of
Angoulême, Arnaud of Périgueux and Jordan of Limoges, as well as the abbots of the monas-
teries of Poitou and the canons of St-Hilaire, Poitiers, whose land was being used for the new
foundation. On the basis of this turnout, Thomas Head has argued that the March assembly
was the occasion of a regular Aquitainian Peace council held every five years under ducal
authority; see T. Head, ‘The Development of the Peace of God in Aquitaine (

970–1005)’,

Speculum

74 (1999), p. 684.

Alfred Richard suggested that the March assembly was probably used by William to organize
the governance of his territory for when he and his son left for Italy to seek the Italian throne;
see A. Richard, Historie des comtes de Poitou,

778–1203 (Paris, 1903) vol. 1, pp. 182–3 and below,

14. He based his argument on the comments of Isembert of Poitiers cited below, n. 57. We

have evidence that the Italian issue spurred the reconciliation between Odo and Robert in a
letter written by Fulbert of Chartres, through whom Odo asks Robert not to harm him and
instead to hear what the Italian ambassadors had to say: Fulbert, Letters, no.

97, pp. 176–7. If

Richard is right, then Robert’s presence at Tours confirms the success of Fulbert’s letter.

Henry II’s death is mentioned in Gesta epp. Cam. III, cc.

36 and 38, and thus brackets the

description of the Mouzon/Ivois summit in c.

37. He had been ill for much of 1024: Annales

Quedlinburgenses, an.

1024, p. 574. For the dating of the Gesta, see above, n. 4. For the final

capitulation of the Lower Lotharingians, see Wolfram, Conrad II, p.

74.

Pabst, ‘Frankreich und Konrad’, pp.

337–68, where Pabst places the events in the context of

France and Germany’s struggle for a dominant position against the other and in central Europe.
This struggle was not to end until one was fully broken by the other, or a solution ‘in wahrhaft
nationalen Sinne’ appeared: idem, p.

339.

T. Reuter, Germany in the Early Middle Ages

800–1050 (Harlow, 1991), p. 187.

Wipo, Gesta Chuonradi, ch.

2, pp. 13–20.

214

Theo Riches

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

new ruler: from Italy, from the kingdom of Burgundy and from
Lotharingia.

Within a year of Henry II’s death, before Conrad’s accession was

secure, a group of Italian magnates attempted to prevent Conrad from
becoming king of the Lombards by offering the crown elsewhere.

It was

apparently first offered to Robert the Pious, who rejected it, and then to
Duke William V of Aquitaine for his son.

William even went to Italy

to investigate his support and he had Fulk Nerra of Anjou ask Robert to
help Dietrich of Upper Lotharingia’s son and co-duke, Frederick (who
was also the stepfather of Conrad the Younger), in his rebellion against
the new king.

While William eventually turned down the offer, for our

purposes it is noteworthy that Odo was involved in these negotiations in
some unknown capacity, including negotiations with Robert.

Although

the death in September

1025 of Hugh, Robert’s son and successor, put an

end to any involvement on the French king’s part,

we know that Gerard

of Cambrai feared his intervention.

Both recent biographies of Conrad address these events only intermittently, perhaps because
each threat dissipated fairly quickly for its own reasons, without having a significant impact on
Conrad’s own activities; see F.-R. Erkens, Konrad II. (um

990–1039). Herrschaft und Reich des

ersten Salierkaisers (Regensburg,

1998), pp. 57–8, 64–72; and Wolfram, Conrad II, pp. 57, 64–7,

73–4, 95–8, 183–4.

For a good analysis of these events, see C. Brühl, Deutschland – Frankreich. Die Geburt zweier
Völker (Cologne,

1990), pp. 674–7. The often-stated belief that they were secular magnates

emerges from William’s letter to Leo of Vercelli after the affair was over, where he refers to
‘quidam primorum Italiae’ wanting him to depose ‘ex voluntate eorum episcopos qui essent
Italiae’ and replace them with others of their choice; see Fulbert, Letters, no.

113, pp. 202–3. But

note that this may refer to only one element of the conspiracy, whose help William could
nevertheless not do without. Indeed the fact that William explicitly excludes Bishop Alricus of
Asti (along with Margrave Manfred of Turin) from these corrupt ‘primores’ implies that the
group could include ecclesiastical as well as secular princes. We can date the events to late
summer and early autumn of

1025 because Bishop Isembert of Poitiers excused himself from

attending the consecration of Hubert of Angers’s cathedral on

16 August 1025 because William,

‘having met with the Italians [habito consilio cum Italis]’, had instructed Isembert and his
colleagues Islo of Saintes and Roho of Angoulême to take care of some affairs: Fulbert, Letters,
no.

102, pp. 186–7. A later letter from Isembert to Archbishop Arnulf of Tours mentions that

William ‘has gone on an expedition from which he is not expected to return until

16 October

[abest in expeditionem profectus, ut aiunt non rediturus usque xvii kalendas Nouembris]’, Fulbert,
Letters, no.

110, pp. 196–7.

Fulbert, Letters, nos.

103–4, pp. 186–9. Most of our information on these events comes from

Fulbert’s letters, but they are also briefly outlined by Adhémar of Chabannes, Chronicon III, c.

62,

ed. P. Bourgain, Corpus christianorum continuatio mediaeualis

129 (Turnhout, 1999), p. 183.

Fulbert, Letters, no.

104, p. 188.

See above, n.

51.

As noted already by Bresslau, Jahrbücher Konrad II, vol.

1, p. 111 and Pfister, Études, p. 266.

Gesta epp. Cam. III, c.

50, p. 485. There is no other evidence that an invasion attempt was

planned, but the concern is understandable given that King Lothar of West Francia had
launched a dangerous attack on Aachen less than fifty years before; the attack and Otto II’s
retaliation are described in Gesta epp. Cam. I, c.

97–98, pp. 440–1. For how these events were

remembered, see T. Riches, ‘The Carolingian Capture of Aachen in

978 and its Historiographi-

cal Footprint’, in P. Fouracre and D. Ganz (eds), Frankland. The Franks and the World of the
Early Middle Ages. Essays in Honour of Dame Jinty Nelson (Manchester,

2008), pp. 191–208.

The struggle for the German throne, 1023–5

215

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

The Burgundian kingdom seems only tangentially related to these

events. King Rudolf III had agreed to make Henry II his heir, but when
Henry predeceased him, Rudolf refused to extend the agreement to
Conrad. Although Rudolf seems to have been acting independently of
any plans of Odo, William or Robert,

his rejection of Conrad provides

a clue as to Odo’s motivation for supporting William’s Italian bid. Odo
was Rudolf’s nephew and might have hoped that the pressure on Conrad
in Italy and Lotharingia would enable Rudolf to break away leaving Odo
as his heir. It was while claiming the kingdom after Rudolf’s

1032 death

that Odo met his own end.

More relevant to our interests is the Lotharingian constellation, as

examined by Robert Parisot.

Both Thietmar of Merseburg and the Gesta

episcoporum Cameracensium relate how Count Gerard of Alsace, the
brother-in-law of Henry II’s empress Cunigunde, had, in

1017, attacked

Duke Godfrey of Lower Lotharingia, but been defeated.

Count Ger-

ard’s young nephew, the future Conrad II, was injured while fighting on
his uncle’s side.

The Cambrai gesta records that the city’s bishop was an

ally of Duke Godfrey in this struggle and that, ‘together with other
supporters of the emperor [they] put down with much difficulty those
rebellions which [relatives] of the empress’s sister often launched against
the stability of the realm’.

According to Thietmar, Gerard of Alsace and

Godfrey only made peace in

1018 because the emperor forced them to.

Wipo, Gesta Chuonradi, ch.

8, p. 31. Rudolf’s unwillingness to renew the agreement may reflect

a decline in the threat posed by Count Otto-William of the Franche-Comté either after Robert
the Pious’s intervention in the neighbouring Burgundian duchy and imposition of his son Henry
as duke or simply because Otto-William was old (he died in

1026, probably in his sixties). The

description of Otto-William as count of the Franche-Comté is anachronistic, but is used to avoid
confusion with the duchy and kingdom of Burgundy. See, further, Chaume, Les origines, vol.

pp.

473–91; R. Poupardin, Le Royaume de Bourgogne (888–1038). Étude sur les origines du royaume

d’Arles (Paris,

1907), pp. 121–36 and H.-D. Kahl, ‘Die Angliederung Burgunds an das mittela-

lterliche Imperium. Zum geschichtlichen Hintergrund des Schatzfundes von Corcelles-près-
Payerne’, Schweizerische Numismatische Rundschau

98 (1969), pp. 27–9.

1037: Wipo, Gesta Chuonradi, ch. 34, p. 56. See also Wolfram, Conrad II, pp. 239–46; Erkens,

Konrad II., pp.

158–64.

Parisot, Les origines de la Haute-Lorraine, pp.

409–18.

Thietmar of Merseburg, Chronicon, ed. R. Holtzmann, MGH SRG, ns

9 (Berlin, 1935), VII, c.

62, p. 474; Gesta epp. Cam. III, c. 11, p. 469. For these events, see also Wolfram, Conrad II, p.

39 and Erkens, Konrad II., pp. 35–6.

Thietmar, Chronicon VII, c.

62, pp. 476–7. In a poem on the occasion of Conrad’s imperial

coronation, Wipo praised his willingness to support his family and friends: Wipo, Cantilena in
Chuonradum II. Factum imperatorem, ed. H. Bresslau, MGH SRG

61 (Hanover, 1915), p. 104,

verse

6a.

‘Post bellum vero Gerardi quod iam diximus, illas seditiones, quas sororii imperatoris contra
statum regni saepius incitabant, domnus episcopus duxque Godefridus una cum aliis fidelibus
imperatoris multo labore sopierant . . . ’ Gesta epp. Cam. III, c.

13, p. 470. Indeed, the Gesta

claims that Bishop Gerard’s intervention had been partly responsible for Henry II making
Godfrey a duke: ibid. III, c.

7, p. 468.

Thietmar, Chronicon IX, c.

17, p. 249. Conrad’s problems with Henry II would continue into

1019, although he was probably in imperial favour again by 1020; see Erkens, Konrad II.,

pp.

36–7.

216

Theo Riches

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Godfrey died in

1023, being succeeded as duke of Lower Lotharingia by

his brother Gozelo, and dynastic rivalry was compounded by conflict
over Godfrey’s inheritance.

For the author of the Gesta episcoporum

Cameracensium, it was Gozelo who led the resistance against Conrad,
taking oaths from the bishops of Lotharingia and beyond that they would
not support Conrad without his consent.

It is clear why the Gesta should

have this emphasis: the previous alliance of bishop and duke had been
suddenly ended by the new duke’s rebellion.

William of Aquitaine saw

the political landscape otherwise: he viewed Duke Frederick of Upper
Lotharingia as the leader of the Lotharingian rebels.

Although we have no evidence that Robert actually was considering an

invasion of Lotharingia, the Cambrai author clearly thought so.

French king had dramatically invaded the area within living memory,
almost capturing the emperor and empress in the process.

But the

Cambrai author also had more recent reasons to have an ambivalent
attitude toward the French king, since his power was an unavoidable
factor in the politics of the region and could be used either to the
advantage or disadvantage of the bishopric. Thus Walter II of Lens, the
castellan of Cambrai and the bishop’s greatest enemy, acquired Robert,
alongside Odo of Blois, as intercessors,

and is listed as having been

present at Compiègne; yet Robert’s influence was welcome when he
helped rein in Baldwin IV of Flanders.

The fear that Robert’s hostility

M. Werner, ‘Der Herzog von Lothringen in salischer Zeit’, in S. Weinfurter (ed.), Die Salier und
das Reich. Band

1. Salier, Adel und Reichsverfassung (Sigmaringen, 1991), pp. 417–20. As Werner

demonstrates, the problems over inheritance pre-date Henry II’s death. Henry had already tried
to prevent the concentration of titles and lands in the hands of Godfrey’s brothers by granting
the county of Drenthe, which he had probably given Godfrey in

1018, to the bishop of Utrecht;

see ibid., p.

418, n. 287. The accession of a new emperor with a history of hostility against the

family only exacerbated tensions that were already present.

‘Quorum [i.e. the electors of Conrad II] ordinationi dux Gothilo, princeps videlicet Lotharien-
sium, contraire voluit; episcoposque Coloniae, Noviomagi, Virduni, Trajecti, Leodii allocutus,
sacramentum a singulis accepit, nonnisi eius consensu manus se ei daturos neque ad eum ituros.
Hoc idem dux Theodericus comesque Haynocensium Raginerius cum sibi conplicibus sacra-
mento firmaverunt.’ Gesta epp. Cam. III, c.

50, p. 485. The reference to ‘Noviomagus’ must be

incorrect – Nijmegen was not a diocese and its closest episcopal seat, Utrecht, is listed separately.
Pabst’s correction to Noyon (‘Noviomensis’) must be right, although we need not share his shock
at the participation of a ‘französischer Unterthan und nichts weiter als das’ in imperial politics:
‘Frankreich und Konrad’, p.

354, esp. n. 2. See also Werner, ‘Der Herzog’, p. 437 and n. 382.

Werner, ‘Der Herzog’, p.

442.

See above, p.

14. Wipo confirms William’s impression: he specifically names Frederick as having

stormed out of Conrad’s election in September

1024 and also as one of the leaders of the

German rebels in

1025 alongside Ernst II of Swabia and Conrad the Younger; see Wipo, Gesta

Chuonradi, ch.

2, p. 19, and ch. 10, p. 32.

Gesta epp. Cam. III, c.

50, p. 485.

See above, n.

62.

Gesta epp. Cam. III, c.

3, p. 467. In fact, neither appeared personally to support Walter, sending

Harduin of Noyon and others instead. See also Gesta epp. Cam. III, c.

42, p. 481. For Walter at

Compiègne, see above, n.

19.

See above, n.

47.

The struggle for the German throne, 1023–5

217

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

would enable and combine with Walter’s activities and the extension of
Flemish power in a ‘perfect storm’ is clearly visible in the Gesta’s account
of the succession crisis.

At the same time, the conflict between Dietrich

of Upper Lotharingia and Odo of Blois had been set aside in

1023, so

there remained no obstacle to cooperation between the French and
Lotharingian magnates.

When the bishop’s previously most reliable ally,

the duke of Lower Lotharingia, also came out against Conrad, and so
potentially alongside Robert, Odo and by extension Baldwin and Walter,
the bishop was left politically alone.

Bishop Gerard had no political leverage, but he may well have hoped

for ideological influence.

The Gesta depicts the ideal relationship

between the two kingdoms in his account of the Mouzon/Ivois meetings.
The portrayal comes immediately after a description of an omen por-
tending the deaths of Duke Godfrey and Emperor Henry, and is thus
explicitly elegaic in tone.

Henry is said to have gone to Ivois intending

to speak with Robert about the state of the kingdom, but not so much
about worldly as about spiritual matters.

Henry’s humility is praised,

and when the two kings meet, their alliance (amicitia) is depicted as part
of a project to establish peace and justice and a renewed Christendom.

They then arranged to meet again in Italy for a council of prelates from
both sides of the Alps. When the Gesta describes the giving of gifts, it

See the quotation below, n.

80.

See above, n.

48.

The Cambrai gesta’s account of how Gerard manoeuvres in this situation is telling. It shows him
living up to the ideals of the bishop as counsellor and mediator, but it was quite clearly making
the best of a bad situation: ‘The Lord Bishop had not agreed to any of these pacts, but once he
understood that they did not have the best of intentions, he kept trying to bring them back into
the fold of peace. Meanwhile he withdrew from the [German] king’s presence, lest he insulted
[the Lotharingian magnates], but cleared himself of the stain of suspicion by sending embassies,
as was his duty. He was just as careful to placate the French king with gifts, so that [Robert]
would not seize [the diocese] before his intended invasion of the whole kingdom. Moreover he
discreetly restrained Count Baldwin, in case he was fooled by Walter’s lies into building up his
fortifications in Cambrai’ (‘His omnibus pactionibus non accesserat domnus episcopus, sed
conabatur eos ad pacis redigere gratiam, postquam cognovit eorum minus bene sanam senten-
tiam. Interim suspendit suum a regis praesentia gradum, ne ipsis fieret scandalum offensionis;
directis tamen officiose legatis, macula se exuit suspicionis. Nihilominus regem Francorum
placare muneribus studuit, ne sibi primitus usurpationem inferret, quam toto regno facere ad
consilium habuit. Balduinum preterea comitem repressit modeste, ne sibi munitiones con-
strueret Cameraci, Walteri corruptus fraudulentiis’), Gesta epp. Cam. III, c.

50, p. 485. The idea

that Robert was in a position to invade ‘the whole kingdom’ is a fantasy. Even if this were only
to refer to Lotharingia, it makes no sense given the position of the Lotharingian dukes vis-à-vis
Conrad.

For similar argumentation about other aspects of Gerard of Cambrai’s activities, see Jégou,
‘L’évêque entre authorité sacrée’, and Stein, Reality Fictions, pp.

13–63.

Gesta epp. Cam. III, c.

36, p. 480.

‘ibi scilicet cum Rotberto rege colloquium habiturus, sed et de statu imperii, ac non tantum de
mundanis verum de spiritualibus locuturus’, Gesta epp. Cam. III, c.

37, p. 480.

See above, n.

21. For the political meaning of ‘amicitia’, see Gerd Althoff, Verwandte, Freunde

und Getreue: Zum politischen Stellenwert der Gruppenbindungen im früheren Mittelalter (Darm-
stadt,

1990), esp. pp. 85–119.

218

Theo Riches

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

explicitly mentions the archbishop of Cologne, Gerard himself and Duke
Godfrey of Lower Lotharingia.

Henry’s subsequent brokerage of peace

at Verdun between Odo and Dietrich of Upper Lotharingia then serves as
a perfect example of a king who ‘suppress[es] wars and spread[s] peaceful
relations’.

Thus in one account, the Gesta demonstrates the role of ideal

kingship in ecclesiastical renewal, both between kingdoms and in Chris-
tendom generally, while placing the role of Lotharingia and its magnates
in the foreground and demonstrating how a good king dealt with unau-
thorized warfare. And this was all written with the foreknowledge that
Henry’s successor Conrad would be opposed by Godfrey’s successor
Gozelo and, potentially, by Robert the Pious as well.

If Robert is seen here positively in his role as a protector of the church,

there remains the question of why Gerard objected to the peace oath
suggested by Warin of Beauvais and Berold of Soissons. He does seem to
have objected to peace oaths in general on the principle that they set such
a high bar that they damned everybody, and thereby robbed the bishops
of one of their most powerful discretionary tools – the power to bind and
loose through excommunication and reconciliation.

Certainly, caught

as he was in a web of overlapping ‘hierarchies of rights’, both ecclesiastical
and secular, the bishop of Cambrai was a stickler for correct legal form.

But the reason advanced here – that enforcing peace is a matter for kings
not bishops – is clearly spurious,

and the accusation that it is suggested

because of the weakness of the king is belied by Robert’s activities at

By the time of writing, Pilgrim would have been known to have capitulated especially quickly
to Conrad II: within three weeks of refusing to elect Conrad he was crowning Conrad’s wife,
Gisela, queen. See Erkens, Konrad II., pp.

54–5. The Gesta describes how the foresworn bishops

became the butts of popular derision: Gesta epp. Cam. III, c.

50, p. 485.

See above, n.

48.

‘So this decree was dangerous to everyone, namely that everyone had to swear [an oath] or be
subject to anathema . . . Scarcely anyone escaped the crime of perjury.’ (‘Hoc ergo decretum
periculosum esse omnibus: omnes videlicet aut iurare aut anathemati subiacere . . . Vix enim
paucissimi crimen periurii evaserunt’), Gesta epp. Cam. III, c.

27, p. 474. See also Riches,

‘Bishop Gerard I’.

A distinctive example from Gerard’s career is when he turned down Henry II’s offer to be
consecrated at the new foundation at Bamberg, despite the honour, and insisted on going to
Reims. Henry gave him a pontifical so that he would not ‘be irregularly consecrated according
to the undisciplined customs of the French’ (‘indisciplinatis moribus Karlensium inregulariter
ordinaretur’), Gesta epp. Cam. III.

2, p. 466. For hierarchies of rights, see H. Vollrath, ‘Konf-

liktwahrnehmung und Konfliktdarstellung in erzählenden Quellen des

11. Jahrhunderts’, in Die

Salier und das Reich. Band

3. Gesellschaftlicher und ideengeschichtlicher Wandel im Reich der Salier,

292–3. Vollrath contrasts a ‘hierarchisch gegliederten Rechtsordnung’ where legal contradic-

tions could be resolved through rational justification, with the logic of the feud, where each
person had his own legal status (Rechtsstand), made up of the sum of all that person’s rights,
possessions and honours, and whose injury demanded immediate reparation through revenge.
This paper differs from Vollrath’s in seeing eleventh-century society as containing multiple,
parallel hierarchies of rights, whose interaction was not clearly regulated, rather than two
qualitatively different legal conceptions, one ‘rational’ and one characteristic of the feud.

See above, p.

The struggle for the German throne, 1023–5

219

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Compiègne, Mouzon/Ivois and Héry. Instead, these both betray Gerard’s
concerns at the time of writing, during the succession crisis. Gerard’s
ideal, as set out in the Mouzon/Ivois account, was for the peaceful
cooperation of kings, episcopacy and dukes in the ordering of Christen-
dom. In this political order, each bishop or magnate took his place as
subject of a king, in Gerard’s case, the German king or emperor. It was on
this order that Gerard relied to maintain his authority as bishop against
his numerous enemies. At the time the Gesta was being written, this order
was broken, since the duke, and many of Gerard’s close episcopal col-
leagues, refused to accept the new king, and the foreign king threatened
to break his God-given bounds. In this context, Robert’s attempt to
overcome the stigma of the Orléans heresy trial by putting himself at the
head of church councils was a threat to Gerard’s authority.

It was

therefore rewritten as the overturning of the natural order. Only in
cooperation with the German emperor did the Cambrai gesta allow
Robert a legitimate role in the renewal of Christendom. The implication
was that in a time without such a figure, the French king needed to leave
well alone. The peace initiative of Berold and Warin, therefore, posed not
just a threat to Gerard’s spiritual power by perjuring everyone without
regard to episcopal oversight, but also his political influence by strength-
ening Robert’s position at a time when Cambrai was exposed and Lothar-
ingia in turmoil. As a result, Robert’s efforts at ecclesiastical renewal and
the propagation of peace that were so prized at Mouzon/Ivois became, in
the absence of Henry II, targets of authorial scorn.

If this was the content of Gerard’s ideological critique, it remains to

be investigated precisely how this was supposed to influence the situ-
ation in the bishop’s favour. In particular, the question of who read or
heard the stories contained in the Cambrai gesta becomes key. Insofar as
its subsequent manuscript tradition can be held to tell us anything
about its intended or expected audience, then the targets of its persua-
sion were the members of the diocese’s various religious communities.

This presumption finds a certain amount of support in two facts: that
the Cambrai author devoted the second book of the Gesta to brief
summaries of each community of the diocese, and that he omitted the
life of Cambrai’s bishop-saint Gaugeric from his history on the
assumption that anyone reading the Gesta would also have access to a

In this light, Gerard’s ‘discovery’ of very similar heretics in January

1025, at the height of the

crisis, looks like ideological competition with the French king, see Moore, Origins of European
Dissent, pp.

288–9.

The method and audience of the author of the Gesta episcoporum Cameracensium are considered
further by Riches, ‘Episcopal Historiography’, pp.

37–42. See, for gesta more generally, M. Sot,

Gesta episcoporum, gesta abbatum (Turnhout,

1981), pp. 45–7.

220

Theo Riches

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

library containing the relevant vita.

Certainly, we know from the use

of Flodoard of Reims’s Historia of Reims by both the Cambrai author
and Folcuin of Lobbes (for the Gesta abbatum Lobbiensium) that gesta
would be consulted by members of religious communities writing a
history of their own institutions.

While we cannot rule out a secular,

even a royal or imperial audience, the most sensible and cautious con-
clusion would be to assume an intended audience of local clergy,
monks and canons, stretching into a presumed future. The question
then remains as to why these relatively lowly figures would be the target
of such abstract ideological argumentation as to the legitimate activities
of the French king in God’s order.

The answer to this must lie in the role such clergy played in the

politicking of regional aristocrats and in the assemblies at which mag-
nates at all levels profiled themselves and sought to push through deci-
sions favourable to their interests.

The opinions of a humble provincial

canon may have been unimportant against the backdrop of such a sig-
nificant assembly as Mouzon/Ivois, but those of Abbot Leduin of St-Vaast
most certainly did matter at the assembly of Compiègne, and in later
years when Gerard competed with the castellan Walter II over the control
of a peace initiative in his own diocese, the question of the local clergy’s
loyalties may well have been significant.

The story of the nun of Nivelles

who gave a relic to Lambert of Louvain, her lover and an inveterate rebel
against imperial authority (d.

12 September 1015), in order to protect him

from being killed at the battle of Florennes (

12 September 1015) may seem

quaint to a modern reader, but it can also be read as illustrative of an
everyday competition to mobilize localized religious resources on one side

Gesta epp. Cam. I, c.

12, pp. 407–8.

Folcuin describes his ‘research trip’ to Reims in Gesta abbatum Lobbiensium, ed. G. Waitz,
MGH SS

4 (Hanover, 1841), c. 7, pp. 58–9.

See, for early medieval assemblies in general, T. Reuter, ‘Assembly Politics in Western Europe
from the Eighth Century to the Twelfth’, in P. Linehan and J. L. Nelson (eds), The Medieval
World (London,

2001), pp. 432–50. In this period, the public element of such assemblies

generally took place after the work of advice-giving, discussion and decision-making had been
done behind closed doors in order to preserve the outward illusion of consensus; see G. Althoff,
‘Colloquium familiare – colloquium secretum – colloquium publicum. Beratung im politis-
chen Leben des früheren Mittelalters’, in idem, Spielregeln der Politik im Mittelalter: Kommu-
nikation in Frieden und Fehde (Darmstadt,

1997), pp. 157–84. The issue of who was included in

the smaller, decision-making circle was, however, a subject of fierce competition; see Steffen
Patzold, ‘Konsens und Konkurrenz. Überlegungen zu einer aktuellen Forschungskonzept der
Mediävistik’, Frühmittelalterliche Studien

41 (2007), pp. 75–104, who refers to the ‘tonange-

benden Kreis’.

Gesta epp. Cam. III, c.

53, pp. 486–7. According to the Gesta account, the two competed to win

over a crowd gathered to approve this peace initiative, although whether this crowd included
local clergy is not specified. Gerard gave a speech that insisted on a bishop’s right to decide who
did and who did not belong to the Christian community, perhaps the most important
‘tonangebenden Kreis’ of all! See Riches, ‘Bishop Gerard I’.

The struggle for the German throne, 1023–5

221

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

or another of ongoing, wider political battles.

Bishops, like counts and

even castellans, were the middle management of the medieval political
hierarchy.

It was important for them to have both the support of local

actors like the crowds who flocked to religious festivals (including Peace
councils) or even incontinent, relic-distributing nuns, as well as that of
high-level magnates like Robert the Pious and Godfrey II of Lotharingia.
Winning the former away from opponents was partly a question of
demonstrating that those opponents’ support from the latter was illegiti-
mate. That Walter was backed by Robert the Pious would count for
nothing in the eyes of the amassed crowd if they had previously been
taught to see Robert as overstepping his authority.

If we accept this model, we also acquire a more sophisticated idea of

the relation between short-term political power and more widely appli-
cable principles of religious and secular hierarchy and order. There need
be no choice between explanations of Peace activity that emphasize
locality and limited political aims on the one hand, and on the other
those that appeal to fundamental theological tensions and seek to place
such action into a longer history of reform culminating in the dramatic
events of the later eleventh century.

Instead, we are looking at local

actors manoeuvring for position and deploying high-level ideological
arguments, which then create the ideological battle lines around which
later reform would circle. The history of the events around the

1023 Peace

and its connection with the struggle for the German throne shows us how
it was precisely in dealing with the petty conflicts characteristic of every-
day history that contemporaries were forced to create the ideological
constructions that would live on in the reform battles and gesta manu-
scripts alike.

The story is told in Gesta epp. Cam. III, c.

12, pp. 469, according to which the relic was

successful until God knocked it free from Lambert’s clothing during the fighting. Lambert’s
opponent was Bishop Gerard’s ally Duke Godfrey II of Lower Lotharingia.

For the business metaphor, albeit differently applied, see T. Reuter, ‘Ein Europa der Bischöfe.
Das Zeitalter Burchards von Worms’, in W. Hartmann (ed.), Bischof Burchard von Worms

1000–1025 (Mainz, 2000), pp. 27–8.

For two subtle examples of the former approach, which the present article also seeks to adopt,
see J. Bowman, ‘Councils, Memories and Mills: The Early Development of the Peace of God
in Catalonia,’ Early Medieval Europe

8 (1999), pp. 99–129, especially his comment at p. 116

about the importance to contemporaries of ‘micropolitical and microsocial considerations’, and
T. Head, ‘Peace and Power in Europe Around the Year

1000’, Essays in Medieval Studies 23

(

2006), pp. 1–17, who nonetheless is careful to emphasize the intellectual sophistication of the

initiators of the Peace.

222

Theo Riches

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Book reviews

emed_297

223..242

Vivarium in Context. By Samuel Barnish, Lellia Cracco Ruggini,
Luciana Cuppo, Ronald Marchese and Marlene Breu. Centre for Medi-
aeval Studies Leonard Boyle, Vicenza. Vicenza: Pozzo Publishers.

2008.

122 pp. + 32 colour plates. $ 25. ISBN 978 88 902035 1 X.

This book gathers four essays in the fields of late Roman history, medieval
Latin and art history. The collection has a peculiar format. It does not
specify an editor, nor does it give an introduction to the essays. In the
absence of these, the reader has to turn to the blurb to get a sense of its
intellectual framework. Here we learn that the aim of the book is to
explore the cultural roots, contemporary relations and impact of
Vivarium, the monastery founded by Magnus Aurelius Cassiodorus
(c.

485–c.585) in his native Bruttium.

Returning to the essays themselves, it is easy to see that only the first

two fulfil this aim. Barnish’s and Cracco Ruggini’s contributions link
Cassiodorus’ monastic project firmly back to his world-view as a secular
office holder. Barnish considers the significance of Cassiodorus’ Variae
as twofold. In the first instance they can be read as a product of the
Roman aristocracy coming to terms with the role of the senate under
Ostrogothic government, by tapping into contemporary eastern political
thought on senatorial libertas reminiscent of the Roman republic. Yet,
Cassiodorus deviates from eastern models, by developing the biblical
prototype of Joseph as provider and instructor for the office of Praeto-
rian Prefect. This definition of office-holding as being held together not
by any secular framework, but by a Christian code of conduct, Barnish
claims, is what makes Cassiodorus’ work enduring for the early Middle
Ages.

Cracco Ruggini analyses Cassiodorus’ interpretation of scientific

knowledge in the Variae in the light of his educational projects. Cas-
siodorus took perhaps a more egalitarian view of scientific knowledge
than his contemporary Boethius, who advocated a traditional Aristotelian
pre-eminence of pure science over technical arts. It is Cassiodorus’ inter-
est in the public and social utilitas of science as expressed in the Variae

Early Medieval Europe

2010 18 (2) 223–242

2010 Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350

Main Street, Malden, MA

02148, USA

that led him to encourage his monastic community to copy Greek
medical encyclopedic texts translated in Ravenna, the former focal point
of Cassiodorus’ life. Vivarium therefore came to preserve a tradition of
technical arts in a world increasingly defined by a Christian concept of
nature as untouchable.

The connection of the two last essays of the collection with the theme

of Vivarium becomes increasingly hazy. Luciana Cuppo investigates a
twelfth-century catalogue of Roman bishops, the Chronica Pontificum,
deriving perhaps from an exemplar from seventh-century Bobbio (pp.

69–75 give a transcription of the text from Vat. lat. 1348). She concludes
that this catalogue is a composition independent from the Roman Liber
pontificalis that can be tied closely to northern Italian support for the
Three Chapters. Cuppo rightly claims that it is perhaps time to reassess
the Liber pontificalis’ monopoly on our view of the early medieval papacy.
However, the links with Vivarium remain unexplored. Only the blurb,
not the essay, tells us about the transmission of the Chronica Pontificum
with the chronicle of Mellitus, a Vivarian edition of Isidore’s chronicle,
which may shed light on the still poorly understood transmission of
Vivarian manuscripts to monasteries such as Bobbio.

Finally, Ronald Marchese and Marlene Breu explore the revival of

Armenian religious embroidery, known as the ‘Constantinople Style’
(

1700–1900), under the Armenian Patriarchate in Istanbul. Marchese and

Breu’s essay is hampered by sloppy editing (an entire section is repro-
duced twice on pp.

83–92), yet it is a careful investigation of the signifi-

cance of religious textiles for the ways in which the Armenian
congregation participated in liturgy. It is also detailed about the artistic
tradition of the Constantinople Style in Byzantine pictorial motifs.
However, by no means does it live up to the claim, stated in the blurb,
to link the motifs found on Armenian liturgical vestments to ‘those at
Saint Clement’s, the Roman basilica renovated by Mercurius, one of
Cassiodorus’ friends, better known as pope John II’. Perhaps to do so
would have been impossible anyway, as John’s connection to the extant
frescoes in the lower church of St Clement’s would seem to be extremely
debatable.

The website of the Centre for Medieval Studies Leonard Boyle, under

whose auspices this collection of essays is published, promotes the book
as ‘[p]articularly suited for upper undergraduate and graduate students’.
However, to establish the relationship between some of the essays and the
overarching theme of Vivarium as a sixth-century cultural centre takes a
certain amount of knowledge and inspiration, something a student audi-
ence may not be alone in finding challenging.

JULIA HILLNER

University of Sheffield

224

Book reviews

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

The Earls of Mercia: Lordship and Power in Late Anglo-Saxon
England. By Stephen Baxter. Oxford and New York: Oxford University
Press.

2007. xviii + 300 pp. £66. ISBN 978 0 19 923098 3.

This monograph offers a penetrating examination of lordship in late
Anglo-Saxon England. The analysis is organized around the Leofwine-
sons, who put forth the earls in Mercia between

994 and 1071. The

continuity of the family’s lordship during this turbulent and decisive
period of internal intrigue and external interventions presents a unique
opportunity for exploring the networks of lordship which bound earls,
kings and local lords to one another in the Anglo-Saxon realm.

After a brisk introduction, which frames the argument and sketches

out the royalist pre-occupations of the literature, Chapter

2 examines the

‘Identity of the House of Leofwine’. ‘Identifying the Leofwinesons’ might
have been a more appropriate title, since the chapter presents a meticu-
lous prosopographical reconstruction of the earls and other major figures
among the family, rather than an investigation of identity per se. The next
four chapters elucidate the structures of lordship: the earls’ administrative
roles within the realm, the family’s landholdings and patronage of mon-
asteries, and the bonds which tied lords and retainers to one another.

These chapters reveal the earls to have been powerful presences in local

politics, forces for unity because of the national distribution of their
holdings, and yet checked within their earldoms by a robust layer of royal
officials. The crowning argument is a revision of Maitland’s conclusions
about the weakness of commendation. Baxter presents a careful exposi-
tion of commendatory bonds, their function relative to other lordly
rights and obligations, and their importance for understanding both the
resilience, as well as the instability, of late Anglo-Saxon lordship. The
study concludes with an insightful investigation of the collapse of the
Leofwinesons’ lordship after the Conquest. Damaged by defeat and
further weakened by William’s shrewd manoeuvring over the next five
years to undermine the earls, the structures that had supported the
Leofwinesons’ power crumbled.

The analysis rests mainly on a sober yet productive treatment of

charter evidence, narrative accounts, and the rich Domesday survey of
landholdings at the time of King Edward. It makes creative use of the
evocative Vision of Earl Leofric to put some flesh on the bones of the terse
charter and narrative records. It also draws upon a number of sources
preserved, and often redacted, during the post-Conquest period, though
their deployment is cautious and assiduously qualified.

The study comes down on the side of the maximalist interpretation of

the late Anglo-Saxon kingdom. The earls were quite powerful, dominat-
ing localities under their sway and commanding, to the surprise of this

Book reviews

225

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

continental scholar anyway, extensive and impressively wealthy lordships.
Despite their power, the earls were susceptible to deposition by Anglo-
Saxon kings in ways which would have been nearly impossible on the
continent.

The Earls of Mercia fits within a larger trend among studies of medieval

lordship which has shifted the focus to the middle layers of power and
emphasized the ties that bound social and political actors into an inter-
dependent network. That being the case, the study might have profitably
brought its examination of lordship into a tighter comparison with
continental scholarship. The moments where this does occur are brief
and tentative, and often based on now aging studies.

This criticism is a quibble, representing what one might wish were

there for the purposes of broadening the reach of this perceptive study,
yet not central to its sharply argued tenets. Specialists should admire the
trenchant yet respectful consideration and recalibration of the scholarship
on Anglo-Saxon political culture, and continentalists will find much
fruitful, comparative material to ponder about the exercise and configu-
ration of lordship in early medieval Europe.

HANS HUMMER

Wayne State University

Beatus Vir: Studies in Early English and Norse Manuscripts in
Memory of Phillip Pulsiano. Edited by A.N. Doane and Kirsten Wolf.
Medieval and Renaissance Texts and Studies

319. Tempe: Arizona Center

for Medieval and Renaissance Studies.

2006. xxix + 545 pp. + 47 b/w

figures. $

65 / £49. ISBN 978 0 86698 364 8.

Phillip Pulsiano was a brilliant young Anglo-Saxonist, energetic and
influential, who died of cancer in

2000, at the age of forty-four. He had

a wide range of scholarly interests, in Old Norse studies as well as
Anglo-Saxon, and among his most significant work was his research
into Old English psalter glosses. He also edited a number of important
books, including the invaluable Medieval Scandinavia: An Encyclopedia
and (with Elaine Treharne) the Blackwell Companion to Anglo-Saxon
Literature. The essays in this volume have all been specially written to
honour his memory, and all examine topics close to his heart, above all
the compilation, interpretation, and preservation of manuscripts. It is
a tightly focused collection, with approximately two-thirds of the
essays devoted to Anglo-Saxon manuscripts, one-third to Norse. The
volume also includes a moving biographical sketch of Phillip Pulsiano
by the editors (one of whom is his widow), and a helpful list of his
publications.

226

Book reviews

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Two outstanding essays look at Ælfric’s Lives of Saints, using individual

manuscripts as a departure point for more wide-ranging discussions.
Joyce Hill reviews the contents list in British Library, Cotton Julius E.vii,
and argues that its compiler proceeded not on the modern notion of
discrete texts (as W.W. Skeat did in his standard edition of the Lives of
Saints), but rather on the liturgical notion of readings belonging to
particular days. Meanwhile, in a fascinating discussion, Jonathan Wilcox
works outwards from British Library, Cotton Caligula A.xiv to consider
the question of the original readership of the Lives. These two essays are
complemented by Kevin Kiernan’s reconstructive digital work on the
much-damaged manuscript British Library, Cotton Otho B.x, which
contains a fragment of Ælfric’s Julian and Basilissa.

There are five other essays on Anglo-Saxon topics. Gernot Wieland

examines the Arator manuscript British Library, Royal

15.A.v, and argues

that its three parts belonged together right from the start. A.N. Doane
supplies a detailed reconstruction of the Werden Glossary, now scattered
in various fragments. Joseph McGowan looks at the practice of ‘elliptical
glossing’, a form of glossarial abbreviation used when an element is
common to two or more compounds. Elaine Treharne provides a further
instalment from her ongoing work on the uses of Old English in the
post-Conquest period, here considering homiletic manuscripts. And J.R.
Hall scrutinizes three disputed readings in the manuscript of Beowulf,
paying close attention to the evidence of the Thorkelin transcripts.

On the Old Norse side, the volume contains two editions (both with

parallel translations) and two studies. Ólafur Halldórsson, in an essay
translated by Peter Foote, edits diplomatically the text of Danakonungatal
in Copenhagen, Royal Library, Barth. D.III fol, and offers an account of
how its information parallels, or differs from, that found in other sources.
Marianne Kalinke supplies a normalized edition of Jóhannes saga gull-
mans, a text full of interest for the influence of Low German on Icelandic
literature and language. Kirsten Wolf considers Copenhagen, Arnamag-
naean Institute

429 12mo, a manuscript from the Icelandic nunnery of

Kirkjubær, but concludes that it furnishes no evidence for female scribal
activity. And Stephanie Würth provides a very helpful review of the
interconnected manuscript transmission of Trójumanna saga and Breta
sögur.

There are also two essays that examine the post-medieval reception of

manuscripts. Peter Lucas explores the world of seventeenth-century anti-
quarianism, by trying to identify the specific manuscripts from which
Abraham Wheelock devised his distinctive typeface for the printing of
Anglo-Saxon. Finally, in the last essay of the collection, Andrew Prescott
offers an engrossing account of the history of the organization and storage
of manuscripts in the old British Library in Bloomsbury – an appropriate

Book reviews

227

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

place to conclude, since the Manuscripts Students’ Room there was
where Phillip Pulsiano conducted much of his research, and was one of
his favourite places.

MATTHEW TOWNEND

University of York

Medieval Dublin VIII. Proceedings of the Friends of Medieval
Dublin Symposium

2006. Edited by Seán Duffy. Dublin: Four Courts

Press.

2008. 328 pp. + 76 b/w figures and 30 b/w plates. £40 (hardback);

19.95 (paperback). ISBN 978 1 84682 042 7 (hardback); ISBN 978 1

84682 043 4 (paperback).

This is the eighth volume in the excellent Medieval Dublin series, all
edited by Seán Duffy. These have appeared annually since

2000 and are

both a written record of presentations given at the Friends’ symposia plus
other contributions. The format allows swift preliminary results from
archaeological work and historical studies in Dublin and its surrounding
area, as well as broader interpretative and analytical papers. Thus the
whole series provides easily accessible, modestly priced reading for
anyone seeking the latest results of research on medieval Dublin.

Volume VIII has three excavation reports (on the Golden Lane, city

wall and Merrion Castle sites); two papers on the longphort and on
Man, Ireland and England; three analyses of annals and chronicles; a
review of pre-Romanesque churches in the Dublin area; environmental
history extracted from a late medieval register; and finally, Howard
Clarke’s memoir of the first thirty years of the Friends of Medieval
Dublin.

Myles and Michael Gibbons revisit the problem of longphuirt and

urbanization in Ireland. They show that the term longphort was not an
ethnic indicator, since there are hints of its use before the Viking Age and
its meaning changed through time. It meant a stronghold in the four-
teenth century, and was used for a camp, later a hunting-booth, in
medieval Scotland. Looking at possible parallels outside Ireland, they
point out that the Repton enclosure was too small to protect the Viking
army camped there in

873–4 reported in the Anglo-Saxon Chronicle, nor

could it effectively have protected a fleet on the River Trent. Longphuirt
were, they speculate, citadels within much larger dispersed riverine or
coastal settlements where ships might have been protected by timber
barriers constructed in the water. Examples are known from Scandinavia
including the one at Hominde, Lolland, Denmark, although the majority
post-date AD

1000. In their opinion, evidence for similar structures

should be sought by fieldworkers in Ireland.

228

Book reviews

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

O’Donovan’s report of an excavation close to the pre-Norman church

of St Michael le Pole by the Dublin Archaeological Research Team,
extends our knowledge of this area close to the River Poddle and the black
pool and adjacent to Linzi Simpson’s excavations reported in Volume VI.
All the recent fieldwork has uncovered small numbers of furnished burials
of the ninth century, suggested by O’Donovan to be part of a widely
dispersed but definable grave field. These contrast with the densely filled
cemetery just south of the church, with its

268 unaccompanied burials

dating c.

700–1200. Simpson has speculated that the earliest Norse settle-

ment in Dublin, of the first quarter of the ninth century, might have been
in this area. This report emphasizes the apparent continuity between the
early Christian and Hiberno-Norse phases, and contrasts that with the
disruption caused by the later Anglo-Norman occupation of the area.

Siobhán Gerahty’s key study of botanical evidence from Fishamble

Street (

1996) showed the variety of plants identifiable from the environ-

mental evidence recovered from Hiberno-Norse Dublin. However, as
Teresa Bolger points out in her paper in the volume under review,
excavated evidence is unable to show the extent to which particular crops
were grown. Using Archbishop Alen’s early sixteenth-century register, a
compilation of documents from c.

1172, she is able to show significant

changes in land use in Dublin’s medieval county over some

300 years.

This is a new approach with some potential for understanding agricul-
tural change.

Linzi Simpson’s own report of an excavation of a short section of

Dublin’s city wall on the south side, shows conclusively that what was
previously thought to be Anglo-Norman construction on top of
Hiberno-Norse work is in fact a single phase of construction and part of
an ambitious building programme by the Anglo-Normans.

Anyone wanting the latest results of work on medieval Dublin need go

no further than this excellent series ably edited by Seán Duffy. Volume
VIII maintains the high standard of content and production set in
Volume I.

DEREK GORE

University of Exeter

La reine Brunehaut. By Bruno Dumézil. Paris: Fayard.

2008. 559 pp. 6

illustrations.

€29. ISBN 978 2 213 63170 7.

If called upon to vote for the leading créateurs/créatrices de la civilisation
médiévale, how many would expect that amongst the potential candi-
dates would be Brunechildis, the Visigothic princess who married the
Merovingian Sigebert I (

561–75), and became the mother, grandmother

Book reviews

229

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

and great-grandmother of four more Frankish kings, only to perish in a
savage purge that extinguished this entire branch of the dynasty in

613?

Yet this is the status claimed for her by Professor Dumézil, who insists
that ‘grace à elle et à son entourage, l’autorité de l’État, le principe d’un
impôt équitable, le fonctionnariat antique, l’universalité du droit écrit et
la littérature classique vécurent un été indien’ (p.

421). If he is correct,

then she has been sorely misjudged and is deserving of a book of over five
hundred and fifty pages extolling her merits. Admittedly, the literary
evidence that relates to her directly could probably be printed on two
sides of a sheet of A

4 paper, and perhaps in consequence the period of

her active life has here to be prefaced by a long lead-in of over one
hundred pages that extends chronologically all the way back to the third
century. Even once arrived in the mid-

560s, when Brunechildis makes

her first personal appearance in the historical record, there is a measured
stateliness about the narrative that extends, like ivy, in all directions,
entwining itself around every possible interpretation and many totally
improbable ones. The frequent appearance of question marks and of
phrases such as sans doute, as well as much use of the subjunctive,
underlines the hypothetical nature of much that is being suggested. We
are also given surprisingly privileged access to the queen’s mind, being
told with confidence of her very thoughts and often the precise motives
for her actions. Those who have previously encountered one or more of
the legion of voluminous books devoted to French monarchs and their
spouses, however short-reigned or obscure, and mostly published by
either Fayard or Tallandier, will recognize that this is another example of
that peculiarly Gallic tradition of haute-vulgarisation applied to the genre
of royal biography. Such books are always, as in this case, written by
distinguished scholars, and usually include both foot- or endnotes and
selected passages from primary sources, sometimes – as here – given in
the original as well as in French translation. They are thus more serious
in character and appearance than many equivalent works of popular
history in other languages. They are also nicely produced in a stiff cover,
are provided with illustrations (now usually in colour) and maps, and
collectively might be thought to look good in a row in a bookcase, with
their titles prominently displayed on their spines. Perhaps it is this role
as cultural status symbol that requires them to be so long, and thus in
many cases also so florid and verbose in style, and so padded and
fantastical in their intellectual content, defying normal rules of historical
argument. In the present case an article on ‘Queen Brunechildis and the
Indian Summer of Equitable Taxation’ might have served its author
much better.

ROGER COLLINS

University of Edinburgh

230

Book reviews

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Frankland. The Franks and the World of the Early Middle Ages.
Essays in Honour of Dame Jinty Nelson. Edited by Paul Fouracre and
David Ganz. Manchester: Manchester University Press.

2008. xvi + 340

pp. £

55. ISBN 9780719076885.

So many historians have been inspired by the work, advice and encour-
agement of Professor Dame Jinty Nelson, and are mindful of it, that the
challenge for the editors of this splendid essay collection in her honour
was to ensure the volume would be of manageable proportions. The
solution adopted was to anchor the book to Jinty’s home territory, the
weekly ‘Earlier Middle Ages’ IHR seminar, so most of the contributors
here have a London connection, whether as former students or as former
colleagues; but the two hundred or so names in the tabula gratulatoria of
historians from over a dozen countries remind the reader of the wide
respect the honoranda enjoys.

Though all connected by some link with ‘Frankland’, the essays form

an impressively diverse and cross-cutting collection, testimony to the
glinting diversity of Jinty’s own thought, which does not make
summary easy. This review will take its cue from the book’s unusual
index. Eschewing traditional alphabetical lists of people and places, it
instead organizes its references according to five classifications, which
will guide the discussion of the essays in this review. Admittedly, cat-
egorizing the essays in this way not only does them all an injustice, as
they are all broad-ranging, it also runs counter to the index’s immediate
point, which is to encourage intellectual cross-referencing, not segrega-
tion. Nevertheless, as its wider purpose is clearly to link the essays to
Jinty’s own interests, in other words to give us a sense of how the
honoranda herself might read them, it seems an appropriate perspective
to adopt.

The theme of the first of these classifications, ‘Living in the Present’, is

most obviously articulated in contributions by Alice Rio on Merovingian
and Carolingian formularies’ value as sources for early medieval history;
by Matthew Innes on the important and importantly repressed Carolin-
gian strand of thought on heresy initiated by Boniface; and by John
Gillingham’s evaluation of codes of battlefield conduct, which it would
seem left increasing room for living in the present, so long as one was
deemed part of the chivalric society.

Several papers speak to the second, closely connected category,

‘Power in the Present’. Two, by Simon MacLean and David Bates, are
on queenship: on the continent in the tenth century and in the Anglo-
Norman realms in the eleventh and early twelfth, respectively. Rachel
Stone and Sarah Hamilton pay attention to particular paths through
which power was exercised in England and the continent – in terms of

Book reviews

231

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

affective lordship for the former, and episcopal excommunication, or
rather post-excommunication reconciliation, for the latter – both
revealing perhaps unexpectedly strong evidence for difference. Susan
Reynolds brings the collective exercise of power in the earlier Middle
Ages into focus through the issue of compulsory purchase, observing
that more research is required, while Stephen Baxter puts the Anglo-
Saxon prosopographical database PASE, in whose establishment Jinty
played a leading role, through its paces, to elicit insight into the
exercise of power by late Anglo-Saxon aristocratic networks and
individuals.

Essays by Alan Thacker and Paul Fouracre resonate well with the third

classification, ‘Power of the Past’. Thacker’s study confirms the influence
of Gregory the Great’s plans for organizing his new Anglo-Saxon con-
verts, but highlights too the relevance of later developments in church
hierarchy. Fouracre concentrates on a rupture in historical memory which
swiftly rendered King Dagobert II so obscure that he was essentially
unknowable already by the ninth century, and would remain so till the
modern period.

The fourth classification, ‘Medieval Teaching’, is particularly clearly

represented through the contributions of David Ganz (an edition and
translation of a challenging but fascinating early medieval teaching
dialogue), Theo Riches (the historiographical afterlife and impact of
King Lothar’s assault on Aachen and Emperor Otto II’s counter-
attack (

978) in west and east Frankland), and Paul Kershaw (the ninth-

century scholarly reception of Bede, emphasizing its international
character). Finally, a fifth classification, ‘Thinking about Frankland’,
is best illustrated by Wendy Davies’s detailed study of the early
modern and modern historiography of ninth-century Breton–Frankish
relations.

If these classifications do represent to some extent how Jinty might

read this volume, there is still surely one aspect missing, namely the
intellectual pleasure one imagines these high-quality pieces will gener-
ate. The collection does not strictly reflect the substance of her work
(there is nothing substantial on coronation, on Hincmar of Rheims, or
on the ‘Alfredian moment’, for example), it reflects instead the perspec-
tives, and the angles, and some of the people, she has inspired and
continues to inspire: a more flexible, and fruitful, approach. Of the
fifteen essays included, there is, to borrow her own phrase justly applied
in the introduction to her own prolific publication record, ‘not a
knocker amongst them’. This volume is a fitting tribute to a renowned
scholar.

C.M.A. WEST

University of Sheffield

232

Book reviews

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Codex Carolinus: Päpstliche Epistolographie im

8. Jahrhundert. By

Achim Thomas Hack. Päpste und Papsttum

35. Stuttgart: Anton Hierse-

mann.

2006. xx + 1290 pp. (in two volumes). €340. ISBN 978 3 7772

0609 7.

This massive and magisterial work is a stupendous achievement. Strictly
speaking, Hack’s book is a minute and comprehensive study of the
ninety-nine letters that make up the Codex Epistolaris Carolinus, that is
the collection of letters dispatched by Popes Gregory III, Zachary,
Stephen II, Paul I, Stephen III and Hadrian I to Charles Martel, Pippin
III, Charlemagne and Carloman, and then Charlemagne. The title is not
original; it was supplied by Peter Lambech in

1673 and then in 1760

Gaetano Cennis assigned the now familiar Codex Carolinus. Between
December of

790 and August of 791, probably at Regensburg, Charle-

magne ordered this collection to be made because many letters were
already in poor condition. The collection should have had the imperial
letters too, but these have vanished. The collection survives today in a
single manuscript, Vienna, ÖNB,

449. This manuscript was apparently

prepared for Archbishop Willibert of Cologne between

871 and 899.

Following in the footsteps of other scholars, Hack tells the interesting
story of how the manuscript slowly migrated to Vienna. The letters were
first printed by Jacob Gretser in

1613 and have been printed a further ten

times. The standard edition is the one in MGH Epistolae Carolini Aevi

by Wilhelm Gundlach, an edition whose various deficiencies have long
been noted. In

1962 Franz Unterkircher published a fine facsimile.

Hack seems confident that the Vienna manuscript is a faithful copy of

the (lost) original. It was written by four hands in fine Caroline minus-
cule. There are corrections in a contemporary hand and in a later hand,
along with many additional ones by Sebastian Tengnagl in the seven-
teenth century who took it upon himself to clean up early medieval
grammar and orthography. The manuscript’s ‘Foreword’ provides some
useful details about the collection. The letters are preceded by lemmata in
majuscules and in red ink ranging from a few lines to half a page. A few
were displaced – perhaps an inattentive copyist? In three places, space was
left but no lemma was added. CC

15 has only a lemma; perhaps that letter

was already illegible. The lemmata seem to function both as a register and
as head-notes. Judging from dorsal notes on some St Gall and Fulda
documents, it is possible that we have here evidence of Carolingian
archival practice. The work may have been carried out in the scriptorium
of St Emmeram; the Carolingian court was in residence at Regensburg
for almost two years. It is possible that Angilramn of Metz, chancellor
after Fulrad of St Denis, supervised the project. Hack reflects on the fact
that this collection is roughly contemporary with the Annales regni

Book reviews

233

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Francorum and the Opus Caroli Regis in a period of fairly intense Franko-
papal relations. A manuscript now in Wolfenbüttel (HAB, Helmstadten-
sis

254) has the famous Capitulare de villis, the Brevium exempla, and ten

letters of Leo III to Charlemagne. This small collection may signal a
continuing effort to collect the papal correspondence with Francia.

Hack’s basic methodology is quite straightforward. He takes up one

after another the formal aspects of each letter in the collection and then
offers generalizations about the collection as a whole. Throughout he is
attentive to formulaic language. This provides him with ample opportu-
nity for very close readings of the letters but it also serves another project.
Nineteenth-century Wissenschaft called both Briefen and Urkunden acta
but always lavished more attention upon Urkunden (documents) in the
(mistaken?) belief that Briefen (letters) are too full of stylistic and literary
peculiarities to be studied ‘scientifically’. Although Hack does not lay
down the ground rules for a sort of Briefdiplomatik, his work certainly
tends usefully in that direction. Hack also studies closely topoi in pre-
cisely Curtius’ sense. The formal epistolary aspects that Hack studies are
protocols, eschatocols, embola, concluding prayers, and good wishes.
Ninety-four of the ninety-nine letters have prayers, usually at the end but
sometimes elsewhere in the body. Hack’s discussion of these prayers is
particularly long, interesting and informative. He notes, for some
examples, the distinctive vocabulary of the prayers, the gestures that
accompanied them, the places where prayers were offered, the people
who did the praying, the addressee of the prayers (usually God), the
specific events that occasioned prayers, and the structure of the prayers
for well-being and success in the here and in the hereafter. The prayers
might signal urgency, uncertainty, satisfaction or thanksgiving. By
reading, characterizing, and comparing them so carefully, Hack can draw
comparisons with liturgical and biblical sources while also adding to
older studies of political thought. Another long section studies forms of
address. Hack looks at the use of the singular and the plural, concrete
intitulations, and a huge array of abstract terms (excellentia, eximietas,
clementia, serenitas, etc.). As an example of the kinds of things Hack
discovers in this exploration, papal abstracts tended to stress the pope’s
dangerous position while other people’s abstracts emphasized positive
personal or moral qualities.

After his detailed treatment of the formal, and often formulaic, dimen-

sions of the letters, Hack turns to two wider historical matters: Commu-
nicative contexts and gifts. Under the former heading, Hack looks at the
actual purpose of the letters, references to other letters (very helpful for
identifying deperdita which he lists in an appendix), irregularities (e.g.
letters that were altered or intercepted), the envoys who bore the letters
(an appendix has an exceptionally valuable prosopography of papal and

234

Book reviews

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Frankish envoys), underlying principles of diplomacy, hints about oral
messages transmitted by the bearers of letters, evidence for instructions to
envoys, and reception ceremonies. Under the latter heading, Hack treats
in great detail (

162 pp.) all the evidence for gifts mentioned in the letters.

This discussion ranges from economic issues reaching back to Alfons
Dopsch and Henri Pirenne and coming down to Michael McCormick, to
anthropological theories such as those of the well-known Marcel Mauss
and those of a number of less well-known figures. Hack expertly charac-
terizes both the gifts themselves and the language used to describe them.

Another important feature of Hack’s methodology is comparative and

contextual within the immense genre of epistolography. The book opens
with a wonderful treatment of the epistolary art from ancient Mesopota-
mia down to its ‘explosion’ in late antiquity. The source material and the
scholarship on it provide Hack with important opportunities to consider
matters of style as well as topoi. In his analysis of the Codes Carolinus,
however, Hack develops repeatedly and in detail comparisons with the
letters of Gregory I and with those of Boniface and Lull.

Less strictly speaking, Hack proffers not food but a banquet for

thought for intellectual, ecclesiastical and political historians of the early
Middle Ages. I already mentioned the ways in which he situates the Codex
Carolinus within the history of epistolography and within the historiog-
raphy of diplomatics. He offers many helpful comments on the quality of
available information – ‘intelligence’ we might say – in the early Middle
Ages. Early medieval moral ideas and values are constantly on display in
this book. There is a great deal here about Latinity and learning.

The book concludes with a series of appendices, tables, and a spec-

tacular bibliography. I have mentioned a couple of these appendices
already but let me now draw direct attention to them. They include:

Protocols and Eschatocols from all papal letters from Gregory I to Leo
III;

2) all the extant or known letters of the Carolingians from Charles

Martel to Louis the Pious;

3) a register of all papal, Carolingian, or

third-party deperdita;

4) a prosopography of all papal and Carolingian

envoys;

5) a careful analysis of all gifts mentioned in the letters of Gregory

I, Boniface and Lull, and the Codex Carolinus. The tables include con-
cordances of the Vienna manuscript and all previous editions, all efforts
to date the letters (none is transmitted with a date), and the letters as they
are referenced in the Jahrbücher of Heinrich Hahn (

714–41), Theodor

Breysig (

741–51), Ludwig Oelsner (Pippin III), and Sigurd Abel and

Bernhard Simson (Charlemagne). These tables are not pedantry run
amok but essential aids to research.

Hack’s is a long, dense and richly detailed study. It would be a real

shame if it wound up being consulted only by a narrow range of special-
ists. Hack bears his learning gently and humbly. He repeatedly indicates

Book reviews

235

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

where further research is needed. Graduate students can find dozens of
dissertation topics in Hack’s suggestions. For the faint of heart, Hack
concludes with an elegant ‘Resümee und Ausblick’. Pages

869 to 889 do

indeed provide a fine resumé of the whole work. Readers might wish to
start here and then turn to the carefully delineated heads and sub-heads
to follow up any particular issue. In the Ausblick, Hack returns to new
material, namely to a careful discussion of papal relations with the
Lombards and then with the Franks.

Achim Thomas Hack’s Codex Carolinus is one of the most important

and inspiring works I have ever read.

THOMAS F.X. NOBLE

University of Notre Dame

Bekehrungsmotive. Untersuchungen zum Christianisierungsprozess
im römischen Westreich und seinen romanisch-germanischen Nach-
folgern (

4.–8. Jahrhundert). By Daniel König. Husum: Matthiesen

Verlag.

2008. 640 pp. ISBN 978 3 7868 1493 1.

The conversion of Europe in late antiquity and the early Middle Ages has
hardly lacked for monographs and new studies in the past fifteen years, so
it is a brave decision to try and bring something new to the field. König
here sets out to study the professed motivations for conversion as a way
of shedding light on what made religious change attractive or necessary.
He professes an attachment, not unreasonably, to a contextualized view of
the processes of conversion in which the experience of one generation or
place will not be the same as that of the next. This is, as the author
admits, difficult territory because most of the source materials are
detached from the processes they describe, or else are prone to moralizing
rhetoric. König’s solution is simply to examine the logic of triggers and
consequences as a means of generating ‘a spectrum of potential conver-
sion motivations, which can be measured against criteria of plausibility’.
What follows is a systematic breakdown of the material, flagged by a
four-page list of sections and subsections in small print to give it some
pretence at brevity. The first main section provides brief summaries of the
conversion of the Roman empire, the Goths, Suevi, Burgundians, Franks
and Lombards – all largely historiographical in flavour but without much
attempt to judge the relative merits of different arguments, except to
announce a preference for the work of Ramsay MacMullen above all. The
next section is the main act: a survey of the different motivations for
conversion, divided into ‘Persuasion’, ‘Compensation’ and ‘The Pursuit
of Advantage’. Under ‘Persuasion’ König surveys a variety of conversion
processes, from the rhetorical outpourings of Lactantius to the ways in

236

Book reviews

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

which martyr cults generated a sense of moral authority; these illustrate
both the importance of the pre-existing dynamics of religious and social
cultures in the late antique world, and the importance of changing role
models as the targets of conversion changed. ‘Compensation’ focuses on
what Christianity offered weaker sections of society: charity for the poor,
hope for the disenfranchised, and so forth. Most of the arguments are
familiar but some, notably the alleged ‘identity crisis’ which the Ger-
manic gentes resolved by adopting Arianism, are redeployed to good effect
within the theme. The material benefits of conversion are also evident in
the final section, which concentrates on conversion motivated by political
expediency and self-interest. Again the examples and conclusions are
familiar: Constantine recognized the unifying power of monotheism,
Clovis needed the cooperation of Catholic bishops for administration,
and so on.

One of the strengths of this book is that it takes difference seriously

within this framework. König illustrates neatly and wisely how, for
example, the conversion of a third-century Roman aristocrat was not like
that of a sixth-century Frankish warrior, either in the triggers for reli-
gious change (whether the individual converted as a result of persuasion
or group action following battle), or in the context and meaning of such
changes. This of course leads to the setting out of a number of possible
motivations and results in the inevitable conclusion that there were
many different processes dependent on context, not all of which lend
themselves to simple categorization. While this is probably a good con-
clusion, it is at the same time not necessarily a particularly satisfying one.
The hows and whys are not neglected, but they do lack depth and do
rely on easy ‘Roman/Germanic’ ‘pre-/post-Constantine’ divisions, par-
ticularly when compared to the nominally less ‘scientific’ but more
revealing books on the conversion of Europe by Richard (here ‘Robert’)
Fletcher and Peter Brown. The use of sources, moreover, does not help
to generate much that is new. Despite noting repeatedly that the sources
are prone to rhetoric, moralizing etc., König usually presents them at
face value in order to illustrate the relevant themes; there is never any
elaborated discussion on, for example, how Gregory of Tours’s many
dispositions and strategies may have tailored his accounts of Clovis or
Gundobad on the subject of Catholics versus Arians. We could debate to
what extent such an approach is appropriate within a typology of con-
version if common themes are found across source types and contexts;
but without a nuanced sense of the sources, the book strays close to
becoming a repackaging of MacMullen and E.A. Thompson within a
heightened ‘scientific’ framework. And therein perhaps lies the problem:
by the end I had read very little that I had not read elsewhere as an
undergraduate. In the end Bekehrungsmotive is a rigorous attempt to

Book reviews

237

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

provide a new model for understanding religious change, but with not
that much to say.

JAMES PALMER

University of St Andrews

Unpredictability and Presence. Norwegian Kingship in the High
Middle Ages. By Hans Jacob Orning. Translated by Alan Crozier. The
Northern World. North Europe and the Baltic c.

400–1700 AD. Peoples,

Economies and Cultures vol.

38. Leiden and Boston: Brill. 2007. 377 pp.

€ 110. ISBN 978 90 04 16661 5.

A key theme in medieval Norwegian historiography has been the unifi-
cation of the country under a single royal authority in the twelfth and
thirteenth centuries. While Unpredictability and Presence focuses on royal
authority, the study adopts a rather unconventional approach to the
topic. Instead of mapping the institutional development of the Norwe-
gian crown, Orning is primarily interested in the personal relations of the
Norwegian kings with their subjects. From the introduction it is apparent
that the author has been inspired by (what can be termed) the ‘legal /
anthropological school’ of medieval studies, in particular the studies of
Fredric L. Cheyette and Stephen D. White on dispute settlements, and
those of Geoffrey Koziol and Gerd Althoff on political rituals. The
author’s approach to the Old Norse sagas has also been influenced by
scholars such as Jesse Byock and William Ian Miller, who from the

1980s

applied these texts to the study of social history. In his exploration of the
‘rules of the political game’ Unpredictability and Presence has a kindred
spirit in Lars Hermanson’s Släkt, vänner och makt. En studie av elitens
politiska kultur i

1100-talets Danmark (Kindred, Friends, and Power. A

Study of the Elite’s Political Culture in Twelfth-Century Denmark) (Göte-
borg,

2000) which also eschewed the traditional approach to Danish

medieval history in favour of an exploration of kinship and power struc-
tures. However, unlike Lars Hermanson, Orning does not deal with the
‘political game’ amongst the elite at a sub-regal level, but focuses on
relations between the Norwegian king and his magnates.

The Old Norse sagas about the Norwegian kings are a valuable source

for the conduct of medieval politics and have been used extensively by
scholars such as Sverre Bagge in his Society and Politics in Snorri Sturlu-
son’s Heimskringla,

1991. However, whereas Bagge relied exclusively on

Snorri Sturluson’s compilation of Kings’ sagas from c.

1230, which deal

with the history of Norway from prehistory to the late twelfth century,
Orning focuses on the contemporary or near-contemporary sagas which
were commissioned or composed under the tutelage of the Norwegian

238

Book reviews

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

kings Sverrir Sigurðarson (

1177–1202) and Hákon Hákonarson

(

1217–63). Moreover, Bagge did not assume that the portrayal of the

kings’ conduct and that of other characters in Heimskringla could illu-
minate the ‘political game’ of eleventh- and twelfth-century Norway; in
fact he argued that Heimskringla likely reflects the values and procedures
in Snorri Sturluson’s time. Conversely, Orning maintains that while the
factual accuracy of the sagas is at times questionable, their portrayal of
how kings interacted with their subjects must reflect what contemporar-
ies considered possible or likely: ‘The sagas’ descriptions are culturally
meaningful, even if they are not necessarily correct in every detail’
(p.

314).

As noted, the main sources for Orning’s analysis are the Icelandic

Kings’ sagas, most importantly The Saga of Sverrir, composed c.

1200, and

the Saga of Hákon Hákonarson the old, completed

1264–5. Whereas the

former saga tells of Sverrir’s struggle for the throne and his subsequent
efforts to keep his competitors at bay, The Saga of Hákon depicts a
relatively secure king who rules by the grace of God. Orning’s analysis
clearly shows the differences in how Hákon and Sverrir interacted with
their retainers, magnates and peasants, both within Norway proper and
in its dependent dominions. The title Unpredictability and Presence cap-
tures the very essence of Orning’s principal thesis. By the thirteenth
century the Norwegian kings claimed to rule by the grace of God yet this
ideological projection was inevitably constrained by political reality. Orn-
ing’s close reading of the sagas shows how the kings had two weapons in
their arsenal that they could employ to attract and retain the loyalty of the
magnates: namely, the king’s physical presence in a society where author-
ity was essentially invested in the person rather than the office, and the
magnate’s uncertainty of how their king would react to insubordination
or outright rebellion.

This is simply a brief summary of Orning’s nuanced and often quite

sophisticated analysis which essentially draws general observations from a
rather limited corpus of texts. It is unfortunate that the author does not
quote more extensively from the principal sagas, since textual nuances are
often pivotal to an analysis of this kind and it would have been useful to
have had the most important passages cited in toto, perhaps in an appen-
dix. The dearth of direct quotations cannot be blamed on a lack of space,
for we are treated to an introductory chapter on methodology and prior
research which runs to some fifty pages. This chapter and the somewhat
repetitious style of the book may reflect the origins of the study in a
Norwegian doctoral dissertation published in

2004.

Orning’s focus on the ‘rules of the game’ provides a rather one-

dimensional view of society where Christianity and particularly the
church are almost invisible. Further dimensions can, and should, be

Book reviews

239

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

added, a point that Orning himself acknowledges in the epilogue, which
addresses some of the criticism aimed at the Norwegian version. These
are minor criticisms. Unpredictability and Presence is an innovative and
thought-provoking study which successfully demonstrates how the analy-
sis of well-known sources from a fresh perspective can open up new fields
of investigation. It will be interesting to observe what legacy it will leave
in the historiography of medieval Scandinavia.

HAKI ANTONSSON

University College London

Sutton Hoo and its Landscape: The Context of Monuments. By Tom
Williamson. Oxford: Windgather.

2008. vii + 154 pp. £20. ISBN 978 1

90511 925 7.

After seventy years of archaeological research at the famous seventh-
century barrow cemetery at Sutton Hoo, Suffolk, landscape history
heavyweight Tom Williamson presents the first systematic and in-depth
study of its landscape context.

Williamson begins by adeptly reviewing the archaeology of Sutton

Hoo and advocating an approach to the cemetery’s location that melds
landscape history and a phenomenology of landscape. He reconstructs
the landscape of the Sandlings from prehistory to the present, using
historic maps, place names, Domesday Book and archaeological survey
data. He regards Sutton Hoo as located on the edge of a sparsely popu-
lated upland heath overlooking the settled tidal estuary of the River
Deben (Chapters

2 and 3). His picture seems restricted because of the

choice to employ a study area using the Deben as its western boundary;
land on the other side of the Deben is omitted regardless of the fact it is
close to, and within sight of, the cemetery. Despite this, Williamson
paints a vivid sketch of the early Anglo-Saxon landscape for the area.

Williamson next explores the socio-political and symbolic context of

the barrow’s location by considering the experience of the barrows in
their immediate setting. He questions whether prehistoric monuments
influenced the cemetery’s location and uses a GIS (Geographic Informa-
tion Systems) viewshed analysis to suggest that the barrows were not
positioned to prioritize a ‘commanding’ location. Instead, he notes the
relative height of the hill and the view of the river from the barrows,
framed by the sides of a shallow embayment. As the cemetery evolved,
Williamson argues that the visual relationship with the river was adapted
and emphasized, culminating in the two ship-burials on the site.

Williamson’s approach is laudable but seems incomplete. His GIS

viewshed analysis methodology is not adequately explained to the reader.

240

Book reviews

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Also, his approach lacks a control sample of alternative locations for the
cemetery, while a systematic comparison between Sutton Hoo and other
early Anglo-Saxon cemetery locations in the vicinity could have strength-
ened his argument. Likewise, a consideration of the locations of other
princely graves across the country must await future work. Williamson
could have taken his interpretations further by reference to recent impor-
tant work on the topic of burial location, particularly the studies by Sarah
Semple and Stuart Brookes.

If these are limitations, it does not prevent Williamson from convinc-

ingly showing that it is pitifully inadequate to explain the location of the
Sutton Hoo mounds in terms of their prominence from the river valley
alone. Yet Williamson has replaced this simple explanation by prioritizing
the static ‘gaze’ of the river from the mounds. Previous work has already
gone further down the road that Williamson’s conclusions clearly point
to. For instance, I have already suggested in print that there is a strong
relationship between seventh-century barrow burials and wide and ‘com-
manding’ views from their locations connected to principal routes of
movement and territorial boundaries. In different ways, Martin Carver
and I have both argued that Sutton Hoo’s location can be understood in
terms of visual interaction with routes of movement along the Deben
valley and across it. Williamson makes a strong case that the river was the
‘centre of the world’ for those living on or around it, but his necessary
landscape focus demands a marriage with a maritime perspective. For
instance, Martin Carver has described Sutton Hoo as not peripheral
when water-borne travel is taken into account. He has seen Sutton Hoo
as located at the ‘front door of the [East Anglian] kingdom’ connected by
‘fast sea routes’ to settlements along the coast and up other estuaries, as
well as with territories across the North Sea.

Adopting a wider regional approach (Chapter

5), Williamson next

robustly and convincingly refutes the argument that Sutton Hoo was an
East Saxon burial ground. Instead the evidence suggests that Sutton Hoo
was situated just within an ‘Anglian’ ‘North Sea Province’ that was largely
determined by environment: soil and water catchments in the longue
durée of prehistory. A detailed comparison of early Anglo-Saxon artefacts
and burial traditions in Essex and Suffolk might reveal whether this
landscape division is so clear-cut in the sixth and seventh centuries AD.

In conclusion, Williamson’s study is innovative and fascinating, man-

aging different scales and approaches in an original case study of historic
landscape analysis. This book is essential reading for students of Sutton
Hoo, early medieval archaeology and landscape archaeology. I believe
Williamson when he asserts that early medieval landscape studies should
embrace both ‘traditional’ and elements of ‘phenomenological’
approaches to landscape. For explicit and thorough debates on early

Book reviews

241

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

medieval burials and their landscape settings beyond Sutton Hoo,
however, readers need to look elsewhere in the recent literature and
towards future studies.

HOWARD WILLIAMS

University of Chester

242

Book reviews

Early Medieval Europe

2010 18 (2)

2010 Blackwell Publishing Ltd

Document Outline

18.2.1
18.2.2
18.2.3
18.2.4
18.2.5

Wyszukiwarka

Podobne podstrony:
Early Medieval Europe Vol XVIII Is 1
Journal of Medieval History Vol XXV Is 1
Journal of Medieval History Vol XXV Is 2
Selected Secondary Sources Rituals and ceremonies from late medieval Europe to early America
North European trading centres and the Early Medieval craftsman Craftsman at Aarhus north eastern Sc
4 Nowe prądy w pedagogice europejskiej XVII i XVIII wieku
The causes of early medieval climate change
The Other Witches The Male Witch of Early Modern Europe a thesis by Arnon Ram (2006)
Exercise of royal power in early medieval
John Marenbon Early Medieval Philosophy 480 1150, An Introduction (2002)
Madness, Religion and the State in Early Modern Europe review
Wood, Hair and Beards in the Early Medieval West
Davis; Medieval Europe
digging ditches in early med europe
Crossfit vol 19 Mar 2004 WHAT IS CROSSFIT
The European Union is the world
IS Historia integracji europejskiej 2011
2014 vol 09 POLITYKA BEZPIECZEŃSTWA ENERGETYCZNEGO UNII EUROPEJSKIEJ W REGIONIE MORZA KASPIJSKIEGO

więcej podobnych podstron

Early Medieval Europe Vol XVIII Is 2

Document Outline