Writing the Barbarian Past
For use by the Author only | © 2016 Koninklijke Brill NV
Brill’s Series on the Early
Middle Ages
Continuation of The Transformation
of the Roman World
Managing Editor
Bonnie Effros (University of Florida)
Editorial Board
Deborah Deliyannis (Indiana University)
Edward James (University College Dublin)
Walter Pohl (Austrian Academy of Sciences)
VOLumE 24
The titles published in this series are listed at brill.com/bsem
For use by the Author only | © 2016 Koninklijke Brill NV
Writing the Barbarian Past
Studies in Early Medieval Historical Narrative
By
Shami Ghosh
LEIDEN | BOSTON
For use by the Author only | © 2016 Koninklijke Brill NV
Cover Illustration: Hildebrandslied. Origenes, lat. Liber Sapientiae. Liber Iesu filii Sirach. Shelf mark: 2° Ms.
theol. 54, fol. 1r. With kind permission of the University Library of Kassel. Landesbibliothek und
Murhardsche Bibliothek der Stadt Kassel.
This publication has been typeset in the multilingual “Brill” typeface. With over 5,100 characters covering
Latin, ipa, Greek, and Cyrillic, this typeface is especially suitable for use in the humanities. For more
information, please see www.brill.com/brill-typeface.
issn 1878-4879
isbn 978-90-04-30522-9 (hardback)
isbn 978-90-04-30581-6 (e-book)
Copyright 2016 by Koninklijke Brill nv, Leiden, The Netherlands.
Koninklijke Brill nv incorporates the imprints Brill, Brill Nijhoff and Hotei Publishing.
All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system,
or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
without prior written permission from the publisher.
Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill nv provided
that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive,
Suite 910, Danvers, ma 01923, usa.
Fees are subject to change.
This book is printed on acid-free paper.
For use by the Author only | © 2016 Koninklijke Brill NV
Carissimae uxori meae
∵
For use by the Author only | © 2016 Koninklijke Brill NV
For use by the Author only | © 2016 Koninklijke Brill NV
Contents
Acknowledgements ix
Abbreviations xi
1 Introduction 1
The Barbarian Past and Early Medieval Historical Narrative
Barbarians and Romans, Christians and Pagans: Cultural
Contact in Late Antiquity 11
Historical Writing in Late Antiquity and the Early Middle Ages
3
27
2 The Gothic Histories of Jordanes and Isidore 39
The Goths and Rome: Historical Background 39
The De origine actibusque Getarum of Jordanes:
Background and Summary 42
Jordanes’s Sources: Oral or Written? 46
Paganism and Arianism in the Getica 60
Goths and Romans: The Purpose of the Gothic Past in the Getica 63
Isidore of Seville’s Historia Gothorum: Background,
Summary, and Sources 69
Religious Identities in Isidore’s Historia Gothorum 74
Goths, Romans and Barbarians in Isidore’s Historia Gothorum 81
The Function of Gothic History: Isidore and Jordanes Compared 87
3 The Origins of the Franks 93
The Historical Background to Frankish Historiography 93
Gregory, Fredegar, and the lhf: Background and Summaries
The Trojan Origin of the Franks 99
The Sources for the Trojan Myth 104
The Function of the Frankish Distant Past 110
95
4 Paul the Deacon and the Ancient History of the Lombards 115
The Early History of the Lombards: Background and Sources 115
Paul the Deacon and his Historia Langobardorum 117
Lombard Oral Tradition in the Historia Langobardorum 121
Catholics, Romans, and Lombards in the Historia Langobardorum 141
For use by the Author only | © 2016 Koninklijke Brill NV
viii
contents
5 A ‘Germanic’ Hero in Latin and the Vernacular: Waltharius
and Waldere 153
Waltharius and Waldere: Authorship, Content, and Historical
Background 155
Christianity in Waltharius 163
Waltharius and Germanic Oral Tradition 170
The Distant Past and its Function: Heroic Narrative as Light
Entertainment
178
6 Looking Back to a Troubled Past: Beowulf and Anglo-Saxon Historical
Consciousness 184
Anglo-Saxon England: Origins, Narratives, and Literary Culture 184
Christianity in Beowulf: The Pagan Past as a Problem 197
Beowulf, Germanic Tradition, and the Anglo-Saxon Past 212
7 Vernacular Oral Tradition and The ‘Germanic’ Past 222
Oral Vernacular Historical Material 225
“Fashionable Gothicism”? The Value of the ‘Germanic’ Past
236
8 Conclusions 257
Bibliography 267
Index 305
For use by the Author only | © 2016 Koninklijke Brill NV
Acknowledgements
This book originated in a doctoral dissertation completed at the Centre for
Medieval Studies in the University of Toronto (defended in August 2009),
which it now supersedes; it is a pleasant duty to acknowledge here my debts to
the various people and institutions without whose manifold aid I would not
have been able to complete this work.
I am grateful to the Centre for Medieval Studies for the award of a graduate
scholarship for the duration of my studies there (2004–9), and I am deeply
indebted to the staff and officers of the Centre, in particular Lawrin Armstrong,
John Magee, Andy Orchard, and above all Grace Desa, for their assistance during these years. In addition, I was the recipient of an Ontario Graduate
Scholarship (2005–6) and a Canada Graduate Scholarship (2006–9) for the
four years of my doctoral work; I thank the Ontario Ministry of Training,
Colleges and Universities, and the Social Sciences and Humanities Research
Council (Government of Canada) for their support of my studies. The dissertation was written while I was one of the inaugural fellows at the Jackman
Humanities Institute in the University of Toronto, and I should like to thank
the faculty fellows for electing me, and the director, Bob Gibbs, the associate
director, Kim Yates, and the administrator, Monica Toffoli, for providing me
with such a pleasant and efficient space to complete my graduate studies. The
last stages of turning the dissertation into a book were completed during my
tenure of a Mellon Postdoctoral Fellowship at the Pontifical Institute of
Mediaeval Studies (2014–15), and I thank the Institute for appointing me to this
post. The library of the Institute has been an invaluable resource for all the
research I have done in Toronto over the years; and without the efforts of the
staff at the Interlibrary Loan office at the John P. Robarts Library I would have
found it impossible to gain access to many essential works consulted in preparing this book.
The dissertation out of which this monograph emerged was supervised by
Nick Everett, Sandy Murray and Andy Orchard, and I am deeply grateful to
them for their patience, criticisms, advice, and support. I am particularly grateful to Andy for his continuing encouragement and the example of his teaching
and scholarship before, during, and after the years I spent producing this work,
without which I would certainly have achieved much less, and with a much
heavier spirit. I should also like to thank my examiners, Tom Hill and Ian
McDougall, for their many useful comments. In addition, I am grateful to Rob
Getz and Markus Stock for their comments and criticisms on earlier drafts of
portions of this book. The constructive criticism from the editorial board and
For use by the Author only | © 2016 Koninklijke Brill NV
x
acknowledgements
reader for this series has been very helpful in forcing me to clarify my arguments and eliminate some errors. At Brill, Marcella Mulder has once again
been an exemplary (and patient) editor, and I am grateful to her and Paige
Sammartino for producing this monograph with maximum efficiency and
minimum fuss. Lawrin Armstrong’s manifold support and encouragement on
all issues connected with my scholarly vocation has been invaluable.
Scholarship has occupied the working hours of most of my adult life; but
those and the rest of my hours would have little meaning were they not also
filled with my wife’s presence. I am grateful that I can dedicate my second
book, like my first, to her—with all my love.
For use by the Author only | © 2016 Koninklijke Brill NV
Abbreviations
AbäG
aqdgm
arca
asc
ase
aspr 3
béfar
bsem
cbt
cccm
ccsl
céa mâ
céa sa
celama
csmlt
da
dlh
ehd
ehr
eme
erga
Amsterdamer Beiträge zur älteren Germanistik.
Ausgewählte Quellen zur deutschen Geschichte des Mittelalters (=
Freiherr vom Stein Gedächtnisausgabe).
arca Classical and Medieval Texts, Papers and Monographs.
Anglo-Saxon Chronicle:
A = Janet M. Bately, ed. 1986. The Anglo-Saxon Chronicle: A Collaborative
Edition, vol. 3: ms A. Cambridge: D.S. Brewer.
B = Simon Taylor, ed. 1983. The Anglo-Saxon Chronicle: A Collaborative
Edition, vol. 4: ms B. Cambridge: D.S. Brewer.
C = Katherine O’Brien O’Keefe, ed. 2000. The Anglo-Saxon Chronicle: A
Collaborative Edition, vol. 5: ms C. Cambridge: D.S. Brewer.
D = G.P. Cubbin, ed. 1996. The Anglo-Saxon Chronicle: A Collaborative
Edition, vol. 6: ms D. Cambridge: D.S. Brewer.
Anglo-Saxon England.
George Philip Krapp and Elliott Van Kirk Dobbie, eds 1936. Anglo-Saxon
Poetic Records, vol. 3: The Exeter Book. New York: Columbia University
Press.
Bibliothèque des écoles françaises d’Athènes et de Rome.
Brill’s Series on the Early Middle Ages.
Cultures, Beliefs and Traditions: Medieval and Early Modern Peoples.
Corpus Christianorum, Continuatio Mediaevalis.
Corpus Christianorum, Series Latina.
Collection des Études Augustiniennes, Série Moyen Âge et Temps
Modernes.
Collection des Études Augustiniennes, Série Antiquité.
Cultural Encounters in Late Antiquity and the Middle Ages.
Cambridge Studies in Medieval Life and Thought, Fourth Series.
Deutsches Archiv zur Erforschung des Mittelalters.
Rudolf Buchner, ed. and trans. 1967. Gregorii episcopi Turonensis
Historiarum libri decem. 2 vols. aqdgm 2. Darmstadt: Wissenschaftliche
Buchgesellschaft.
Dorothy Whitelock, ed. 1979. English Historical Documents, c. 500–1042,
2nd edn. London: Eyre Methuen.
English Historical Review
Early Medieval Europe
Ergänzungsbände zum Reallexikon der germanischen Altertumskunde.
For use by the Author only | © 2016 Koninklijke Brill NV
xii
Etym.
FMSt
gl
hd
he
hg
Hist. Brit.
hl
íf
jla
kut
lcl
lhf
mgh
mgm
Mlat Jb
mst
abbreviations
W.M. Lindsay, ed. 1911. Isidori Hispalensis episcopi Etymologiarum sive
originum libri xx. Oxford: Clarendon Press.
Frühmittelalterliche Studien.
Germania Latina.
Habelts Dissertationsdrucke, Reihe alte Geschichte.
Bède le Vénérable: Histoire ecclésiastique du peuple anglais, ed. Michael
Lapidge, trans. Pierre Monat and Philippe Robin, with introduction and
notes by André Crépin. 3 vols. Sources Chrétiennes 489–91. Paris:
Éditions du Cerf.
Cristóbal Rodríguez Alonso, ed. and trans. 1975. Las historias de los godos,
vandalos y suevos de Isidoro de Sevilla. Collección fuentes y estudios de
historia leonesa 13. León: Centro de Estudios e Investigación “San
Isidoro”, Archivo Histórico Dicesano, Caja de Ahorros y Monte de Piedad
de León.
“Historia Britonnum”. In Edmond Faral, ed. and trans. 1969 [orig. 1929].
La légende arthurienne: Études et documents, vol. 3: 2–62. Bibliothèque de
l’École des Haute Études 257. Paris: Champion.
Ludwig Bethmann and Georg Waitz, eds. 1878. “Historia Langobardorum”.
In mgh SRL: 12–187. Hanover: Hahn.
Íslenzk fornrit.
Journal of Late Antiquity.
Kohlhammer Urban-Taschenbücher.
Loeb Classical Library.
Bruno Krusch, ed. 1888. “Liber Historiae Francorum”. In mgh srm 2:
215–328. Hanover: Hahn.
Monumenta Germaniae Historica.
aa
Auctores Antiquissimi.
Epp
Epistolae.
ll
Leges.
Poet.
Poetae Latini medii aevi.
srg
Scriptores rerum Germanicarum in usum scholarum
separatim editi.
srg ns
Scriptores rerum Germanicarum nova series.
srl
Scriptores rerum Langobardicarum.
srm
Scriptores rerum Merovingicarum.
ss
Scriptores.
Monographien zur Geschichte des Mittelalters.
Mittellateinisches Jahrbuch.
Mittellateinische Studien und Texte.
For use by the Author only | © 2016 Koninklijke Brill NV
abbreviations
ogl
p &p
Pan. lat.
pbb
pg
pl
plre
rwaw
s &t
sem
Sett.
sha
sm
toes
trhs
trw
tth
tumh
Var.
viög
vkf
ZfdA
xiii
Ludwig Bethmann and Georg Waitz, eds. 1878. “Origo gentis
Langobardorum”. In mgh SRL: 2–6. Hanover: Hahn.
Past and Present.
R.A.B. Mynors, ed. 1964. xii panegyrici Latini. Oxford: Clarendon Press.
Beiträge zur Geschichte der deutschen Sprache und Literatur.
J.-P. Migne, ed. 1857–1912. Patrologiae cursus completus, Series graeca.
J.-P. Migne, ed. 1839–64. Patrologiae cursus completus, Series latina.
A.H.M. Jones, J.R. Martindale, and J. Morris. 1971–92. The Prosopography
of the Later Roman Empire. 3 vols. Cambridge: Cambridge University
Press.
Rheinisch-westfälische Akademie der Wissenschaften, Vorträge.
Studies and Texts.
Studies in the Early Middle Ages.
Settimane di studio del Centro italiano di studi sull’alto medioevo.
Studies in Historical Archaeoethnology.
Sammlung Metzler.
Toronto Old English Series.
Transactions of the Royal Historical Society.
Transformation of the Roman World.
Translated Texts for Historians.
Texte und Untersuchungen zur mittelhochdeutschen Heldenepik.
Theodor Mommsen, ed. 1894. Cassiodori Senatoris Variae. mgh aa 12.
Berlin: Weidmann.
Veröffentlichungen des Instituts für Österreichische
Geschichtsforschung.
Veröffentlichungen der Kommission für Frühmittelalterforschung.
Zeitschrift für deutsches Altertum.
For use by the Author only | © 2016 Koninklijke Brill NV
For use by the Author only | © 2016 Koninklijke Brill NV
chapter 1
Introduction
On Christmas Day in the year 800, Charles, king of the Franks and known to
posterity as Karolus magnus or Charlemagne, was crowned Roman emperor
by Pope Leo iii in a ceremony at Rome. For the first time since 476, there was
again a Roman emperor in the west, albeit one who ruled over only a limited
portion of what had once been the western empire, along with some regions
east of the Rhine that had always been beyond the empire’s boundaries. One
of the principal sources for the life of Charlemagne is the Vita Karoli magni by
his courtier Einhard, a text composed in fluent classicising Latin, and modelled in many respects on Suetonius’s imperial biographies.1 Although, according to Einhard, Charlemagne had trouble writing, he nevertheless appears to
have learnt Latin, and was clearly Christian;2 like his empire, Charlemagne was
manifestly at least in some respects—including his religion—‘Roman’. But the
fluently Latinate Einhard calls himself a “homo barbarus”; and immediately
after telling us about his patron’s coronation, he also informs us that Charlemagne
ordered “barbara et antiquissima carmina” to be written, a sign of the value the
new emperor placed, it could be argued, on an aspect of his ‘barbarian’
heritage.3
Like Charlemagne, Einhard was a Frank, and a native speaker of a Germanic
language; although Charlemagne ruled over many regions that had been within
the Roman empire, including its very heartlands in Italy, he was himself from
northern Europe, and the core of Carolingian power lay in what are now northern France, western Germany, and Belgium, rather than Rome and central
Italy. Charlemagne himself seems to have had some reverence for Theoderic
the Great, the Ostrogothic king of Italy in the sixth century: he had a statue of
Theoderic removed from Ravenna and transported to Aachen.4 This has been
seen as an indication that Charlemagne valued Theoderic as a great ‘Germanic’
ruler of antiquity.5 The ‘barbarian’ inheritance, on this reading, was thus of
1 Holder-Egger (ed. 1911; the coronation is reported at c. 28); on Einhard and Suetonius, see further Innes (1997). On the sources for and implications of the imperial coronation, see Becher
(2003): 81–119; Collins (1998): 141–59; McKitterick (2008): 114–18; Schieffer (2004).
2 Holder-Egger (ed. 1911): c. 25.
3 Holder-Egger (ed. 1911): Prologus; c. 29.
4 Deliyannis (ed. 2006): c. 94.
5 Löwe (1952): 394–8.
© koninklijke brill nv, leiden, ���6 | doi �0.��63/97890043058�6_00�
For use by the Author only | © 2016 Koninklijke Brill NV
2
chapter 1
some importance for the new Roman emperor, most likely because he was
himself still, in some respects at least, not fully Roman.
Yet there can be no disputing the centrality of the Roman inheritance for
Carolingian Europe, and indeed for early medieval western Europe altogether.6
After 476, there was no longer a Roman emperor in the west, and the regions
that had once been imperial provinces were all, by the early sixth century,
factually independent post-imperial kingdoms, ruled in most cases by the
descendants of peoples who had once been a military class within late-imperial
society, but who were generally, even into the sixth century and certainly for
most of the fifth, not native speakers of Latin, largely illiterate, and either not
Christian, or Arian rather than Catholic. However, all of the various peoples
who established these post-imperial polities in western Europe eventually
adopted Catholic Christianity (and therefore the Roman religion), as well as
the Latin language, the (primarily Latin) written word both for use in government and for cultural production, and many aspects of late Roman administration. These were all parts of the Roman inheritance; and as a promoter of
precisely many of these very Roman aspects of cultural and administrative life
in the process that has been termed the ‘Carolingian Renaissance’, Charlemagne
was perhaps quite appropriately a Roman emperor.7
Charlemagne’s imperial status notwithstanding, even the Franks had originally been, in Roman perception, ‘barbarians’, and one of the most fundamental aspects of early medieval cultural history is—to put it crudely—precisely
the process of ‘barbarians’ becoming (more) Roman. This process did not, however, by any means imply the elimination of non-Roman aspects of the ‘barbarian’ heritage: a simple reading of Einhard’s words cited above would suggest
that at least one part of that heritage—ancient barbarian songs—were still
valued at the very highest levels of this Romanising society. We could thus argue
that the value placed both on “barbara carmina” (for example) and Roman,
Latin texts—the production of which reach unprecedented heights during the
Carolingian Renaissance—embodies the synthesis of early medieval culture,
which can be said to have emerged out of a conjunction of different elements,
6 There is a vast body of work on the significance of Rome in early medieval culture; see for
example McKitterick (2014); Riché (1995): 153–201; Julia Smith (2005): 28–31; 255–92; Wickham
(2009): 200–1; 561–2. On the importance of the Rome as a political model and source of legitimacy, not least in the Carolingian era, see also most recently Heather (2013).
7 Of the innumerable studies of various aspects of the ‘Carolingian Renaissance’, see e.g. the
essays in McKitterick (ed. 1994) for useful surveys; on the Latin/Roman inheritance, see in
particular Brown (1994); Garrison (1994); Garrison (2000); Innes (1997); and Wood (2014), in
addition to the works cited in the previous note.
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
3
some more Roman than others. Needless to say, the evolution of early medieval
culture from its diverse inheritances has been the subject of much previous
scholarship; the present monograph provides an examination of one very specific aspect of this process, namely the ways in which texts dealing explicitly
with a past that was in some manner ‘barbarian’ treated that past in relation to
the Roman aspect of the cultural inheritance.
The Barbarian Past and Early Medieval Historical Narrative8
Narratives concerning the origins and ancient histories of the peoples who
established post-Roman kingdoms in western Europe were not, in the form in
which they have come down to us in the texts examined below, also contained
within the traditions of classical Roman or Greek secular historiography or
ethnography, nor within the tradition of ecclesiastical history. The narratives
studied in the following chapters are all, in one way or another, evidence of
methods of utilising both the Roman and the non-Roman elements of early
medieval cultural heritage to provide an expression of some sort of coherent
sense of historical consciousness in the present within which these texts were
composed.
How did early medieval writers present aspects of the distant past that were
not Christian (or at least not orthodox), not Roman, and not a part of the written Latin historical memory inherited from Rome and the Roman Church?
How were these two facets of the cultural memory of the early middle ages
related to each other in the narratives that explicitly have a barbarian, nonChristian past as their theme? These are the questions that the present work
seeks to address, based on analyses of a number of discrete texts from quite
8 Henceforth I use the term ‘barbarian’ without inverted commas to refer to non-Roman peoples who would have been encompassed by the Roman usage of the term, and thus in general
interchangeably with ‘non-Roman’. No value judgement is implied by my use of the term,
and hopefully the peoples concerned are sufficiently far removed from any living peoples
and polities that my choice of vocabulary will be inoffensive. For a defence of the usage of
‘barbarian’ vs (for example) ‘Germanic’ or ‘non-Roman’, see James (2009): 5–8; cf. however
Haubrichs (2011): 28, n. 60. The term ‘barbarian’ and its connotations are discussed in more
detail, with further references, in the next section of this chapter. The ‘barbarian past’,
although non-Roman, could also be Christian, and I qualify the phrase further with ‘preChristian’ or ‘pre-Catholic’ as appropriate. The Jewish past was also, of course, both nonRoman and pre-Christian, but none of the peoples who are the subjects of the texts discussed
below had any Jewish or genuinely biblical heritage, so my usage will hopefully be unambiguous in this respect.
For use by the Author only | © 2016 Koninklijke Brill NV
4
chapter 1
different contexts. While there is rarely any explicit relationship between the
contexts of the evidence used here, the studies presented below nevertheless
do reveal some trends in terms of the place of the barbarian past in early medieval culture. What is equally important to note, however, is that different
aspects of the past served different functions within different elements of society: the values and historical consciousness of the militarised lay aristocracy
were not necessarily identical to those of the clergy, and this difference is evident from comparing the perspectives adopted by, for example, the authors of
the ‘national’ histories examined below, and those who wrote ecclesiastical
histories during the same period.9 Nevertheless, another result of this study is
the demonstration that even with regard just to historical consciousness, there
was no fundamental barrier between the lay and ecclesiastical spheres.
There exists a small corpus of texts that are devoted to the histories of the
Goths, Franks, and Lombards from their origins to the author’s present, and
these are, in the present work, interpreted as efforts to conjoin a sense of an
independent identity as Goths, Franks, and Lombards, with the inherited
Roman and Christian historical traditions. These texts are the Gothic histories
of Jordanes (De origine actibusque Getarum, hereafter Getica; c.551–2) and
Isidore of Seville (Historia Gothorum, hereafter hg; c.630); the Fredegar chronicle (c.660) and the Liber historiae Francorum (hereafter lhf; c.727), both of
which provide narratives about the origins of the Franks; and Paul the Deacon’s
Historia Langobardorum (hl; c.790). Chapters 2–4 of this book are devoted to
these narratives, which are, with the exception of Fredegar, all explicitly
‘national’ histories in that they are concerned exclusively or primarily with the
history of one people that was also the ruling class of a post-Roman kingdom,
and with which the author clearly identified. (The Fredegar chronicle is an
exception in that it starts as a universal chronicle, but it too increasingly focuses
on the Franks, and it is clear that Frankish history is the primary concern.)
In the latter part of this book, I examine two epics, the Latin Waltharius
(composed in the ninth or tenth century) and the Old English Beowulf (composed at some point between c.700 and c.1000), and finally some shorter
vernacular texts, along with other evidence regarding the place of the barbar9 Here and in the following pages, the term ‘national’ history is used as a shorthand for a history that is devoted to the past of a single people. In other words, it excludes universal histories or world chronicles, as well as the vernacular legendary matter that deals with many
different peoples in a single narrative. The use of the word ‘national’ is simply a circumlocution for ‘relating to a single people’, and should not be taken to have any relation to modern
concepts of nationality and nationhood. For a succinct survey of ‘national’ histories, see
Martínez Pizarro (2003); the relevant specialist scholarship is cited and discussed more fully
at appropriate points below.
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
5
ian past in contemporary historical consciousness. All the works studied in
second part of the present monograph tell of a distant past that was not contained, in the way these narratives report it, in any prior written historical tradition; all of these texts appear to derive primarily from oral vernacular sources.
None of them were origin narratives, and there is evidence (presented and
discussed in Chapter 7) that such tales of distant heroes that existed in an oral
milieu were condemned in clerical circles. Since they could not have had any
kind of immediate function beyond the provision of entertainment and the
conveying of secular values, clerical censure of these narratives was not easy to
overcome, and since writing was largely controlled by the Church, few such
texts survive from our period. They are, nevertheless, important evidence for
the existence of a secular historical culture that found expression chiefly in the
vernacular and in an oral milieu.
All of these texts are representative of a secular historical consciousness that
was nevertheless deeply imbued with Christianity, and indeed survives within
a written, and for the most part Latin form—the medium and language of the
Church, though not exclusively of the Church. They are also all narratives that
were not just about the past, but about a distant past that had fundamental
differences, in all of these texts, from the present within which they were produced: this distant past was not Roman; it was not (explicitly) Christian; and it
had lived in a language that was not Latin. Nevertheless, all the works examined below betray evidence of a form of reconciliation between a perception of
a barbarian, non-Christian heritage, and a Christian and Roman/Latin inheritance that had a very tangible presence.
The texts studied here are not the only works that could have been examined for the purposes of the present monograph. Other ‘national’ histories are
extant from the early middle ages, notably the Historia Brittonum (c.830),
Widukind of Corvey’s Rerum gestarum Saxonicarum libri tres (967–73), and
Dudo of St Quentin’s De moribus et actis primorum Normanniiæ ducum (996–
1015). The first of these is something of an anomaly, since it is aware of the real
Roman past of its subjects, as well as their post-Roman history; all three works,
however, perform a process of integrating Roman/Christian/Latin, and barbarian or otherwise non-Roman elements, within a more or less coherent historical consciousness.10
10
On Widukind, see fundamentally Beumann (1950); Beumann (1970); Beumann (1982); see
also Eggert and Pätzold (1984): 206–22; Mortensen (1995); Plassmann (2006): 243–64; and
for the text, see Hirsch and Lohmann (eds 1935). On Dudo, see Gatti and Degl’Innocenti
(eds 199); and Plassmann (2006): 265–89; for the text, see Lair (ed. 1865). The relevant
scholarship on the Historia Brittonum is cited in Chapter 6.
For use by the Author only | © 2016 Koninklijke Brill NV
6
chapter 1
As such, it might appear to have been logical to extend the present book to
encompass studies of these histories, not least because they present useful comparanda and examples of how approaches to this process did (or did not)
change over time and across a wider range of historical contexts. However, any
work of scholarship must stop somewhere: a number of even later histories also,
albeit in a variety of different ways, undertake the same task of providing a synthesis between a barbarian past and a Roman inheritance, including, for example, the highly erudite and voluminous Gesta Danorum of Saxo Grammaticus
(c.1208), and it would not be possible to keep expanding the scope of this monograph without sacrificing the gains of detailed close readings of the individual
texts.11 I shall argue below that Paul the Deacon’s Historia Langobardorum is the
first extant ‘national’ history genuinely to provide such a synthesis with a minimum of embarrassment regarding the cleavages (potential or real) within the
story it tells; this makes it an appropriate point to curtail the examination of
‘national’ histories and move on to other material.12
For the most part, the Latin ‘national’ histories discussed below have not been
studied in conjunction with the epics and brief vernacular narratives also examined in the present monograph: the one group is thought to be ‘history’, while
the other is ‘literature’, and they have accordingly largely been examined by
scholars of different disciplines.13 Here I seek to bridge this divide, since both
sets of texts were evidence of a historical consciousness seeking to express a
view about a particular kind of past. While even in the middle ages history and
literature might not have been perceived to be identical, we must recall that
human consciousness does not respect modern disciplinary boundaries: the
11
12
13
On Saxo, see the essays in Friis-Jensen (ed. 1981); Friis-Jensen (1987); Friis-Jensen (1992);
and the recent introduction of Riis (2006); for the text, see Friis-Jensen and Fisher (ed.
and trans. 2015).
By this I do not mean, of course, that there would be no value to comparing the texts
examined here with others that are not, including but not restricted to those named
above; see for example Plassmann (2006), and from a more theoretical perspective with a
wider range of comparanda, Graus (1975). For a broad-ranging survey of ‘national’ histories in the middle ages, see Kersken (1995).
An exception that briefly compares a vernacular heroic epic (Beowulf) with some Latin
histories (including those of Paul the Deacon and Jordanes) with regard to narrative style is
presented by Scheil (2008). There is, of course, a large body of Germanist literary scholarship that has mined some of the Latin narratives examined below for reflexes of Germanic
heroic legend for use as comparanda to the extant vernacular texts; the relevant works are
cited at appropriate points below, but it should be noted that the purpose of this scholarship has rarely been a detailed examination of the Latin narratives as what they are—
Latin narratives—and there has been to my knowledge no detailed prior examination in
comparative perspective of historical consciousness as manifest in these two types of text.
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
7
historical consciousness of a people, past or present, will comprise, and be influenced by, more than just texts that might be perceived formally as ‘history’ of
some sort. After all, our own conception of our past is shaped at least as much
(arguably more) by historical fiction, films, and television dramas as it is by the
works of academic historians; there is no reason to believe that early medieval
historical consciousness was more respectful of the segregation between genres
that modern academic disciplinary boundaries might seek to impose on it.
All of these texts show that however much we might feel that there might
have been sharp divides between a pre-Christian and Christian period, between
a Latinate and vernacular culture, between barbarian and Roman pasts, or
between the milieux of the clergy and the laity, these pairs of opposites are not
necessarily the best way of understanding early medieval culture. The texts
examined below do not, for the most part, present the past in terms of such
oppositions, but rather stress continuities over cleavages. While the present work
is not an effort to provide a holistic interpretation of early medieval culture and
how it reconciled its varied inheritances, it does intend to add to the many recent
reassessments of the nature of this culture from one specific angle, the attitude
towards the past, and specifically, towards precisely that distant past that we
might perceive today as having been discontinuous with the present of early
medieval societies. For this purpose, an interdisciplinary approach seemed to me
to be useful, and I hope my readers, whether students of literature or history, will
also gain some stimulus from my transgression of the disciplinary boundaries.
A further rationale behind this selection of texts also has a historiographical
justification. A ‘Germanic’ spectre has long loomed over the study of the early
middle ages: historians and philologists alike have sought evidence of the survival of some sort of ‘Germanic’ antiquity in the extant literary and material
survivals from this period, and have often felt it appropriate, for this reason, to
interpret the extant evidence with reference to other sources that are also—
for better or worse reasons—thought to be ‘Germanic’, though not necessarily
contemporary or belonging to the same historical context. Furthermore, the
cultural heritage of this period has often been understood in oppositional
terms specifically as Roman and ‘Germanic’, not just Roman and barbarian.
The early period of encounters between ‘Germanic’ peoples and Rome has
been seen as a ‘heroic age’ (c.300–c.600, though in some versions it stretches
even further back), commemorated in later narrative verse traditions in the
vernacular, reflections of which are thought to be extant even in early medieval Latin texts, including some of the works to be discussed below.14
14
On the concept of a ‘heroic age’ in a broader context, see fundamentally Chadwick (1912),
and Bowra (1957). Specifically on a ‘Germanic’ heroic age and its reflection in (later)
extant vernacular works, see in particular the classic statements of Heusler (1905), and
For use by the Author only | © 2016 Koninklijke Brill NV
8
chapter 1
It is certainly the case that the Goths, Lombards, Franks, Angles, Saxons, and
Burgundians—the peoples who figure as the principal protagonists and/or primary or initial audiences of the narratives to be examined below—were all
Germanic peoples, in that their vernacular tongue belonged to the Germanic
sub-group of the Indo-European family of languages. It is also the case that the
corpus of what literary scholars define as Germanic heroic poetry does contain
narratives that have as a historical core events that took place largely in the
period c.300–c.600—insofar as any of these narratives can in fact be related to
any sort of historical realities at all. But there is little evidence from before the
eighth century, at least, for any sense even of an awareness of an inter-relatedness
among these peoples, and certainly not of any perception among them of any
significance of such inter-relatedness—any sort of knowledge of and meaning
granted to a common ‘Germanentum’, or ‘Germanic-ness’, that has any relation
to the burden of significance such a concept has borne in modern scholarship.
Furthermore, the historical links between the extant heroic texts and any verifiable historical fact are both invariably slender and often quite tenuous, and
therefore should not be overvalued. And finally, as we shall see in Chapter 7,
even when we find writers displaying a knowledge of the linguistic kinship of
the Germanic languages, this did not imply either a strict division between
‘Germanic’ and Roman cultural heritages, nor a perception of a sense of a common identity of any kind derived from any other form of kinship, cultural or
otherwise, among the Germanic-speaking peoples.15
All the texts discussed below concern themselves primarily with a Germanic
people—in the linguistic sense. The epics and poetic texts examined in the
latter chapters also contain material extant either only in Germanic vernaculars, or also, and primarily, in those vernaculars. Jordanes’s Getica, Paul’s hl, as
15
Heusler (1943): 155–60, as well as Heusler (1909); see further, for critiques, restatements,
and refinements of the concept of a ‘Germanic’ heroic age and its commemoration in
vernacular poetry, Andersson (1988b); Ebenbauer (1988); Haubrichs (2002); Haug (1975);
and Haug (1994). For a broader, comparative perspective, see Reichl (2000). A recent,
thorough, and insightful synthesis is provided by Harris (2012). My own views have been
presented in Ghosh (2007). For critiques that are in essence rejections of the concept
altogether (or at least its validity as an analytical category for the understanding of the
origins of vernacular heroic poetry in the Germanic languages), see Frank (1991); and
Goffart (2002).
For critiques of the concept of an early ‘Germanic culture’, see Amory (1997): 326–31;
Goffart (1980): 12–29; Goffart (1995); Goffart (2006): 187–229; and Halsall (2007): 22–4;
118–31. For a very recent effort to reinstate a (weaker) concept of “pan-Germanic identity”,
see Neidorf (2013a). The points made above are discussed in more detail in Chapter 7
below. See also, on classical perceptions of ‘Germanic’ peoples, Isaac (2004): 427–39.
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
9
well as Waltharius, Beowulf, and the shorter vernacular texts have all been seen
as expressions, in some form or another, of ‘Germanic’ culture. (Indeed, this is
even true of the Histories of Gregory of Tours.) They are thus a useful sample
with which not only to make the point that there was, by the end of the eighth
century at the latest, really no perception of a great divide between barbarian
and Roman inheritances at least with regard to historical consciousness, but
also to reiterate the point that—once again specifically in the context of historical consciousness—there appears to have been no special significance
attached to any ‘Germanic’ identity or heritage.16
The studies below present analyses of literary texts that have a particular
historical relevance, and are read with due attention to their historical context (insofar as this can be determined); but this book is not intended to be a
study of early medieval historical consciousness at a more general level. It
does not seek to match the results of analyses of specific literary texts to
other, non-literary forms of evidence to arrive at a more comprehensive
understanding of how early medieval people thought about their past(s).17
Nor does it intend to use analyses of this particular aspect of historical consciousness to come to conclusions regarding early medieval notions of group
identity. Other kinds of sources present their own problems of interpretation,
and are not easily amenable to the same kind of interrogation as literary narratives, thus rendering a comparison of the evidence of differing genres an
16
17
A note on usage is thus apposite here: ‘Germanic’, when used without inverted commas,
refers solely to language. A Germanic people, in my lexicon, is simply a people that speaks
a Germanic language, defined as Germanic on the basis of purely linguistic criteria that
have nothing whatsoever to do with ethnicity or other markers of identity. This need not
mean that any of these Germanic peoples knew that their native tongue was a Germanic
language; that they felt any kind of identity with other peoples who spoke a Germanic language; or even that they particularly cared that the language they spoke was Germanic
and not something else. A phrase such as ‘Germanic tradition’ does not, in my usage,
imply any sort of bond between peoples beyond that of language; it refers only to the fact
that narrative elements contained in this tradition were known in more than one
Germanic language, and that across Germanic languages there might have been similarities of poetic form caused not least by the similarities of the languages. There are no further implications to my use of the phrase ‘Germanic tradition’. The term ‘Germanic’ used
without inverted commas therefore does not, in this study, carry any implications of any
kind of ancient, more or less unified, ‘Germanic’ culture which would allow one to interpret the practices or textual remains of one Germanic-speaking group by reference to the
practices or textual remains of another.
In this regard, the present study follows the examples of Coumert (2007), Plassmann
(2006), and Reydellet (1981).
For use by the Author only | © 2016 Koninklijke Brill NV
10
chapter 1
exercise fraught with difficulty.18 Concentrating solely on narrative texts,
which are explicitly intended to convey to readers stories and views about the
distant past, allows me to address the question of the role of this past in a
more well-defined manner, focusing on the issues of religious difference and
the distinctions between Romans and barbarians. I shall show, over the
course of the following pages, that a pattern can be identified regarding the
way these issues were treated across a lengthy period of time. My arguments
are valid in the first instance for the texts examined here, but I believe that a
demonstration, on the basis of these examples, of the existence of this pattern can also contribute something to our understanding of secular historical
culture in the early middle ages, and the place in it of the distant, barbarian,
non-biblical past.
However, given how little we know about the authors, their audiences, and
the reception of their texts, it is difficult to say much about the extent to which
the views presented in these works reflect widespread ideas about the past. The
reception history of the texts can tell us a little: Jordanes, Fredegar, the lhf and
the hl were widely read in the ninth and tenth centuries at least,19 and the
references to figures of the past in vernacular poetry examined in Chapter 7
below suggest that narratives about them were widely known. Some of the
ideas contained in the ‘national’ histories seem to have resonated with later
audiences; it is difficult to know how much they did so among contemporaries,
or even the extent to which they reflected a common historical consciousness
in the periods in which they were written. Once again, the common elements
in these works—deriving from an effort to stress continuity between distant
origins and the present, regardless of the religious difference and the gap
between a barbarian past and a present imbued with a very Roman heritage—
suggests that since several authors had similar concerns, these reflected a view
more broadly prevalent at least among the secular aristocracy. It is also likely
that the authors of these texts were responding to audience expectations,
which were thus in turn reflected in their narratives.20 To say more, however,
would be purely speculative.
18
19
20
Good examples of studies using a multitude of different sources to attempt to understand
ethnic identities in this period are Amory (1997), and Koch (2012). For a study exposing
the complexities involved in using law-codes to understand ethnicity, see Amory (1993).
For knowledge of Jordanes’s Getica, Fredegar, the lhf and Paul’s hl in the ninth century
and later, see Innes (2000): 243–4; Lapidge (2006): 160; 239; 272; McKitterick (1989): 238–41;
McKitterick (2004b): 13–15; 37–8; 49–51; 55–8; 75–83; 201; 212. Isidore’s hg does not seem to
have enjoyed a wide reception.
Plassmann (2006): 22–3.
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
11
These then, are the questions posed in this book and the constraints within
which they are addressed. The individual texts will be given a more detailed
introduction in each of the following chapters, along with a brief (and very
basic) overview of the historical background regarding the people(s) concerned, their earlier contact with Rome and Christianity in the historical
record, and the relevant narrative sources available in addition to the texts to
be examined here. In the remainder of the present chapter I provide, first, a
more general introduction to the historical context of cultural contact between
Romans and barbarians in late antiquity and the scholarship on barbarian
identity; and second, a (necessarily very brief) survey of the other kinds of historical narrative extant from this period, including both texts that were demonstrably known to the authors of the works to be examined here, as well as
others that must be understood as part of the intellectual and cultural furniture of the audience of these texts.
Barbarians and Romans, Christians and Pagans: Cultural Contact
in Late Antiquity21
The term ‘barbarian’ already had a long history by our period; the Romans had
inherited it from the Greeks and generally used the term to indicate foreign,
exotic, and (in their view) uncivilised peoples around their borders.22 There
was a fairly prominent discourse in late antiquity about barbarians needing to
be ‘tamed’ by Roman emperors, who are presented as having to keep the peace
and defend Rome against the savage barbarians; and while the sources show
21
22
The following paragraphs present what is a necessarily superficial synthesis of a vast and
often fractious body of scholarship; for a recent useful, clear, and balanced survey, see
James (2009). Other important recent works include Halsall (2007), and Heather (2009);
see further the essays collected in Mitchell and Greatrex (eds 2000), and Mathisen and
Shanzer (eds 2011). For reasons of space—and relevance to the primary topic of the present monograph—I do not engage in detailed discussion of these or other works, nor cite
profusely the earlier scholarship.
For a brief history of the use of the term, see Gillett (2009): 397–402; see further the surveys of barbarians and the term ‘barbarus’ and its cognates in late antiquity (and modern
scholarship) in James (2009): 1–20; Maas (2012); and Mathisen (2011). On classical perceptions of barbarians and other ‘others’, cf. the contrasting views in Isaac (2004); Gruen
(2011); and Woolf (2011).
For use by the Author only | © 2016 Koninklijke Brill NV
12
chapter 1
barbarians being settled within the empire, the literary scheme of rebellious
settler-barbarians is also quite common.23
It is easy to follow blindly the Roman rhetoric that differentiates starkly
between Romans and barbarians, but it would be unwise to do so. Just as the
narratives examined below do not operate with such a strict dichotomy
between Romans and barbarians, it is also the case that in actual fact, the
boundaries were less sharp than they might appear; and even the late antique
sources do not present such a singular, undifferentiated image of barbarians.24
As Ralph Mathisen and Danuta Shanzer recently stated, we should not operate
with a model of “cultural and social segregation of Roman and barbarian populations […] Romans and barbarians interacted in every way imaginable, social,
cultural, political, and religious”.25
The Roman empire had, for most of its history, engaged in the process of
absorbing barbarians (and other outsiders of one sort or another) within its
economic, legal, political, and cultural sphere; after the grant of citizenship to
all free men within the empire in 212, distinctions between Romans and barbarians seem to have become, in law and practice, even less clear than earlier.
It appears to be the case that barbarians now became citizens of the empire as
long as they resided within its boundaries, and it has been suggested that many
people who might still have been classified as barbarians were nevertheless
now able to “enjoy the benefits of Roman civil law without having to give up
23
24
25
Ladner (1976). On Roman perceptions of barbarians in late antiquity, see further Halsall
(2007): 45–56; Maas (2012); Wickham (2009): 44–9. Specifically on how barbarians are
presented by fourth- and fifth-century historians, see Rohrbacher (2002): 207–36. For
detailed studies on the very influential Christian historian Orosius’s views of barbarians
in the early fifth century, see further Goetz (1980b); Inglebert (1996): 566–8; and van
Nuffelen (2012): 170–85; for a recent comparison of Orosius and Augustine’s treatment of
barbarians, see Clark (2011). Orosius and his significance are briefly discussed in the next
section of the present chapter.
As Edward James states, it is likely that we “have a more black-and-white view of ethnic
difference than the Romans or barbarians”: James (2009): 126. In a similar vein, HansWerner Goetz has suggested, with regard to Merovingian sources, that “ethnische
Abgrenzungen [sind] (allein) ein Problem der modernen Forschung”: Goetz (2004): 555.
On the difficulty of determining, from the sources, what criteria were used to distinguish
between Roman and barbarian in the sixth century, see further Greatrex (2000); on the use
of the term ‘barbarus’ and its potentially mixed connotations in Gaul/Francia c.400–c.700,
see Wood (2011). With regard to classical antiquity as well, Erich Gruen has recently
argued that Romans “had far more mixed, nuanced, and complex opinions about other
peoples”: Gruen (2011): 3; but cf. Isaac (2004).
Mathisen and Shanzer (2011): 4.
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
13
their own cultural identity”.26 Whatever aspects of non-Roman culture they
might have retained, these Roman barbarians, by functioning as Roman citizens, were also clearly Roman; and despite a legal prohibition from the 370s,
there appear to have been no real de facto restrictions of marriage between
barbarians and Romans (at least among the elites), another suggestion that in
fact, once within Roman boundaries, the distinctions could become increasingly blurred.27
The circumstances in which barbarians became Roman appear to have
been principally military. Many were recruited by one means or another as
soldiers in the Roman army, and from the third century onwards at the latest,
it became common practice for some aspects of the defence of the empire to
be delegated to ‘warlords’, leaders of armed groups settled at the borders of the
empire, either just within or immediately without those borders.28 Some of
these groups were incorporated within the Roman army, while others, the socalled ‘foederati’, functioned as allies.29 Over the course of the fourth and fifth
centuries, many of these warlords became generals leading not just their ‘own’
troops (that is to say, some sort of barbarian grouping), but whole Roman
armies of mixed origin. Indeed, already in the fourth century, a number of
them held the title of ‘magister militum’ or ‘magister militum utriusque militiae’: commander-in-chief of the Roman armed forces in a province or in the
whole of the western or eastern empire. Some even advanced to becoming
consul.30 Many of these barbarian-Roman military commanders were among
the groups labelled (by the Romans) ‘Franks’ or ‘Goths’ (as well as other groups
not among those to be examined in the present work).31 Indeed, there are a
number of attested marriages between such barbarian military leaders and
26
27
28
29
30
31
Maas (2012): 63.
The issues relating to law and citizenship have been examined in most detail by Ralph
Mathisen: Mathisen (2006); Mathisen (2012). Specifically on the marriage laws, see Sivan
(1996). See further the useful survey of barbarians in contact with Rome from the third to
the late fifth century in James (2009): 33–75; 194–201.
Another mode of contact that could lead to forms of intercultural communication was
captivity, also obviously in the context of military action: Grey (2011); Lenski (2011).
On the ‘foederati’ and the recruitment of barbarians in the Roman army, see in brief Jones
(1964): 611–13; 619–23; for a more recent survey discussing the ‘foederati’ and their significance, see Stickler (2007); and see further James (2009): 161–73. For a more general survey
of the late Roman army, see Jones (1964): 607–86.
On these developments, see Liebeschuetz (2007); MacGeorge (2002); Whittaker (1994):
243–78.
Some examples of high-ranking barbarians in the Roman army in the fourth century who
are thought to be Franks or Goths, and attested in contemporary or near-contemporary
For use by the Author only | © 2016 Koninklijke Brill NV
14
chapter 1
Roman aristocrats; the extent to which such cross-cultural kin-relationships
were common lower down the social scale cannot really be determined, but
there seems to be no particular reason to believe that there was any significant
level of formal segregation.32
We have to conclude, therefore, that whatever the cause of the ‘fall’ of Rome,
it was not the result of a sudden influx of completely exotic, uncivilised barbarians: the military leaders who succeeded imperial rule and established kingdoms on the continent in the post-imperial west belonged to peoples that, in
every case, had had at least a century of contact with the Roman empire and
often lived within its bounds, perhaps even as Roman citizens; had collaborated more or less closely with the Roman army; and had in most cases in fact
been in some manner a part of the Roman army. While we can know little
about the level of ‘Romanisation’ of the lower ranks, there is every reason to
believe of the military commanders that, whatever other cultural affiliations
they might have retained, they were also quite Roman.33
In other words: the barbarians who established post-imperial kingdoms had
most certainly partaken of a great deal of Roman culture, whatever other
32
33
sources (listed in chronological order of their first appointment to high military office):
Gaiso (Goth): magister militum(?) (350); consul (351) (plre i: 380).
Silvanus (Frank): tribune (351); magister peditum (352/353–5); Augustus (355) (plre i:
840–1).
Mallobaudes (Frank): tribune (354–5); comes domesticorum and rex Francorum
(according to Ammianus) (378) (plre i: 539; Ammianus: xxxi.x,6).
Flavius Merobaudes (Frank): magister peditum (375–88?); consul (377; 383; 388) (plre i:
598–9).
Flavius Bauto (Frank): magister militum (c.380–5); consul (385) (plre i: 159–60).
Arbogastes (Frank): comes rei militaris (380); magister militum (c.388–94) (plre i: 95–7).
Modares (Goth): magister militum (380–2) (plre i: 605).
Flavius Richomeres (Frank): magister militum (383; 388–93); consul (384) (plre i:
765–6).
Gainas (Goth): comes rei militaris (395–9); magister utriusque militiae (399–400)
(plre i: 379–80).
Flavius Fravitta (Goth): magister militum (395/400); consul (401) (plre i: 372–3).
On the generals’ marriages, see Liebeschuetz (2007): 482; 489; see further, on documented
kin-relationships between barbarians and Romans (principally with regard to the elites),
Blockley (1982); Claude (1989); Demandt (1989); James (2009): 194–6; Krautschick (1989);
Soraci (1974). Even if there was little legal enforcement of any prohibition on intermarriage, one should not automatically assume that it was common; the example of modern
multicultural societies in, for example, Canada or the uk, demonstrates that even after
two generations, cross-cultural marriages can remain relatively rare.
On the potential for conflicted loyalties or identities among these military commanders,
see Chauviot (1984).
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
15
identities they might have had in addition; and while the forms and meaning
of Roman-ness were also arguably altered as a result of this contact in some
ways, to assume that the culture of the barbarian groups remained untouched
by Roman contact is implausible, and not supported by the extant evidence.34
Moreover, we must recall that these kingdoms were, initially at least, established as clients of the empire, rather than as truly independent polities; the
process of becoming properly ‘post-Roman’ was a long one, and involved a
good deal of Roman-ness as well, not least with regard to the religion of the
people within these kingdoms, all of which were, by c.500 at the latest, predominantly, if not completely, Christian (though not necessarily Catholic).35
That being said, one should not push the argument for continuity too far.
There was certainly also some amount of social and economic disruption in
the fifth and sixth centuries;36 and however much the western kingdoms
inherited from Rome in terms of law, language, religion, and administrative
practice, the barbarian settlers in these regions also brought with them their
34
35
36
Guy Halsall makes the important point that even the supposed ‘barbarisation’ of the late
Roman army in terms of its practices, clothing, and customs, is overstated: Halsall (2007):
102–9. See further, with regard to dress, hairstyles, and customs in Ostrogothic Italy,
Amory (1997): 338–47; Arnold (2013); and James (2009): 168–70. A detailed study of one
example, the Vandals in Africa, is provided by von Rummel (2007), who concludes that
what we can know of their material culture provides no evidence that it was in fact particularly distinct as non-Roman. For a recent argument that the Gothic kingdom in Italy
was in fact very ‘Roman’, while some Italo-Romans in it also began to adopt some ‘Gothic’
traits, see Arnold (2014). None of this detracts from the fact that the “cultural sharing was
a two-way street” in this period that resulted in a “polyethnic cultural world, with cultural
frontiers between Romans and barbarians that were increasingly permeable in both
directions” (Mathisen and Shanzer 2011: 4).
For a useful recent (and succinct) survey on the establishment and early history of the
western kingdoms, see Kulikowski (2012); cf. the varying perspectives given in Halsall
(2007): 220–319; Heather (2009): 266–385; James (2009): 76–94; and Wickham (2009):
76–149. A stimulating discussion of Roman/Italian views of fifth-century Gaul and the
process of that region’s becoming ‘post-Roman’ while also remaining in many respects
‘Roman’ is provided in Arnold (2014): 235–61. England is a rather different case from the
continental successors to the empire with regard to the points made in this and the following paragraphs; see James (2009): 202–4; and Ward-Perkins (2000); further discussion
and references are given below in Chapter 6.
Cf. e.g. Goffart (1980); and Ward-Perkins (2005). Some sort of middle ground between
these two positions is probably the most plausible way of interpreting the period. Since
the heated debate regarding the modalities of settlement and the extent of disruption
involved is not really pertinent to my subject, I avoid further discussion of this topic,
though I cite without comment the more fundamental works of scholarship on it where
appropriate. For the most significant recent salvoes, see (in chronological order) Goffart
(2006); Halsall (2007); Heather (2009); Goffart (2010); Halsall (2010).
For use by the Author only | © 2016 Koninklijke Brill NV
16
chapter 1
own, non-Roman heritages. These settlers were always a minority of the population, but this need not necessarily mean that the barbarian aspects of their
culture were swiftly or comprehensively swamped by the culture of the Roman
majority.37 While it is important to bear in mind the fact that, with the exception of England and some border regions, in almost all the lands that had been
within the western Roman empire, the vernacular tongue was eventually a
Latin-derived (that is to say, Romance) rather than Germanic language, we
should also remember that linguistic and cultural identity are not necessarily
identical. It is clear from the example of Paul the Deacon’s hl (as we shall see
in Chapter 4) that although the Lombard language was most likely no longer in
common use and the Lombards where thoroughly ‘Italian’ by his time, some
sense of cultural identity still imbued the narratives of pre-Christian, nonRoman, non-Italian origins he transmitted, which concerned a period in which
the Lombards had spoken a different language.
While religious identity has sometimes been seen as an important marker of
either Roman or barbarian allegiance, the continental barbarians even of the
fifth century, and certainly of the sixth, appear to have been overwhelmingly
Christian. Many (though not all) were Arians and thus did not accept the
authority over them of the bishop of Rome. This might suggest that Arianism
was a distinctively barbarian trait in the fifth century; but it was also selfevidently equally a part of the Roman inheritance rather than a specifically
barbarian attribute: Christianity, in whatever form, had only come to barbarians through contact with Rome. Thus to view the Arian religion of many barbarian groups as being necessarily something that made them non-Roman is
not really an accurate perspective on the matter, even though the distinction in
Christian denomination could clearly be a significant cultural divide for at least
some witnesses, not least the clerical authors on whom we depend for most of
our evidence. The distinction between Arian and Nicene Christians is not sufficient, however, for us to postulate a clear distinction between ‘barbaritas’ and
‘Romanitas’ separating people who lived in the same regions—particularly
because precisely the fact of their being clergy arguably meant that the authors
of our sources might have made religious differences appear to be more divisive than they might have been in actual fact in the daily lives (or even in the
political actions) of most people (or even just the secular ruling classes).
Leaving aside the hotly debated legal and economic issues (which are less
relevant for the subject of this monograph), with regard to culture alone, it is
clear that barbarian identity in the period between c.300 and c.600 was not
static, and was also neither impervious to, nor wholly submerged by, the contact
37
On this point see also Wickham (2009): 97–102; 105–7.
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
17
with and lasting influence of Rome. Much ink has been spilt on the ways in
which the barbarian and Roman cultural inheritances interacted to shape ethnic identities in the fifth, sixth, and seventh centuries in particular, and some of
the relevant scholarship will be discussed at appropriate points in the following
chapters. For the purposes of the present work, the most important aspects of
the debate on early medieval ethnic identity have to do with the age, authenticity, value, and function of the narratives concerning the past, and views on this
issue diverge considerably. Given that much of the discussion has been based
precisely on narratives about the distant past, a survey of the contours of the
debate and a presentation of my own position is apposite here—particularly
since, as Chris Wickham recently stated, “no one in the rest of Late Antique
studies gets as upset about anything” as do those who debate (‘Germanic’) ethnicity in this period.38
The participants in the debate are often divided into two camps, the socalled ‘Vienna School’, and the so-called ‘Toronto/Goffart School’. The position
of the latter can sometimes seem to be that there is in fact no authentic material (that is to say, genuinely ancient, or at least genuinely barbarian, and not
derived from written Roman sources) in any of the narratives, and nothing in
them derives from oral tradition of any kind: these narratives are entirely constructed on the basis of earlier Latin and Greek texts, and even the vernacular
poetry extant from the ninth century and later was stimulated by a reading of
Jordanes rather than originating in any vernacular oral tradition.39 At the
other end of the scale, the Vienna School can read information provided in an
eighth-century source in the light of later, thirteenth-century material from
Scandinavia, primarily on the basis of both sources being ‘Germanic’; ‘oral tradition’ is seen as the source for the origin narratives, which are thought to contain something that is, because it “does not fit” with the rest of the narrative,
indeed an authentic ethnic memory that served as a marker of identity.40
This summary is, of course, something of a caricature; both sides of the
debate are (normally, but not invariably) rather more sophisticated, and both
sides have produced work that (normally, but not invariably) contains much
38
39
40
Wickham (2012): 552.
For the extreme position, see Frank (1991); Goffart (1988): passim; and Goffart (2002); cf.
Chapter 7 below; and Ghosh (2007): 229; 242–7.
For the extreme position, see Pohl (2000b): 16; and Pohl (2002): 228–9 (quote); 233; see
also e.g. Wolfram (1994); and cf. Ghosh (2007): 245; 247–51. Note that Walter Pohl appears
to be willing to change his views somewhat—rather more so, it seems to me, than both
Walter Goffart and Herwig Wolfram; thus the Pohl position in 2000 or 2002 is not necessarily identical with that of 2013.
For use by the Author only | © 2016 Koninklijke Brill NV
18
chapter 1
that has greatly enriched our understanding of the period. It is also certainly
the case that there is really no monolithic ‘school’, as the individual scholars
who have been grouped into one or another camp often have quite different
approaches to the issues at hand. Nevertheless, it is equally certainly the case
that the debate is often presented and carried out in precisely such starkly
polarised terms as my caricature suggests. It seems to me, however, that—
perhaps because this subject seems to excite passions in a manner unusual in
medieval studies—these kinds of positions tend to make the discussion both
unnecessarily polemical, and ultimately rather sterile, partly because the argument and evidence can both get drowned in the polemic.
The fact is that the extant narratives are indeed substantially derived from
borrowings from or adaptations of a clearly Roman (or Greco-Roman) inheritance (whether or not the producers of these texts thought of it in that manner), but equally clearly often contain some material for which no extant
Roman source can be found. This does not, of course, mean that this material
is necessarily ancient, or derived from some sort of long-enduring ‘oral tradition’, nor even that it is free of written, Latin, and Roman influence. But it does
exist, and we cannot dismiss it, nor later (ninth- and tenth-century) evidence
for something that can only be called an oral tradition of historical narrative in
Germanic languages, as unimportant, or simply created from the reading of
Latin texts. These points are elaborated on in some detail at appropriate points
in the chapters below.
The principal contributions of the Vienna School and those scholars associated with it have had to do with the theory of ethnogenesis.41 In the context of
41
For the major expositions of the theory itself, and supportive contributions to the debate
from a variety of perspectives, see, in chronological order, Wenskus (1961); Wolfram (1979);
Wolfram (1981); Geary (1983); Geary (1988); Wolfram (1990); Wood (1990); Wolfram (1994);
Pohl (1991); Pohl (1994b); Pohl (1998a); Pohl (1998b); Wolfram (1998); Geary (1999); Pohl
(1999); Geary (2002); Pohl (2002); and most recently Pohl (2013). The most succinct and
recent summary of the theory and its major critics is presented in Maas (2012): 74–7; the
most detailed (if excessively polemical) historiographical analysis of ethnogenesis theory
and its development is in Murray (2002); see further also the surveys of the debate in
Gillett (2006); and James (2009): 102–28. In addition, for useful recent surveys of the modern historiography on barbarian identity in relation to modern national identity, see
Wood (2008), and in much greater detail, Ian N. Wood (2013), which should be read along
with Fouracre (2014). Ian Wood’s recent works perform the salutary service of placing the
present debate within the larger perspective of modern scholarship on these issues from
the eighteenth century onwards; while this history of historical writing is obviously not a
subject one could expect most scholars of the early middle ages to be as deeply immersed
in as Wood, the longer perspective is nevertheless something we should all at least be
more aware of than is normally the case. Such awareness might even help to reduce just
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
19
late antique and early medieval studies, ethnogenesis theory has been almost
exclusively concerned with groups that are thought to have had a ‘Germanic’
identity.42 The most prominent proponents of this theory have been Herwig
Wolfram and Walter Pohl, although it originated, in this context, in the work
of Reinhard Wenskus. Other prominent scholars outside Vienna whose work
has been sympathetic to ethnogenesis theory include Patrick Geary and Ian
Wood, who—like Pohl—use the concept in a more flexible manner than
Wenskus and Wolfram, with less emphasis on the concept of a more or less
stable ‘Traditionskern’. Patrick Amory’s monograph on Ostrogothic Italy has
also been claimed as an example of the use of ethnogenesis theory.43
The core components of ethnogenesis theory are that ethnic identity was
flexible, and was constructed in a manner that included Roman and barbarian
influence; much of the recent scholarship sympathetic to this theory would
agree that ethnic identity was a ‘situational construct’, created and utilised
because of and within particular historical situations. This is in itself not only
plausible, but almost certainly true of the nature of identity in the period in
question (and, mutatis mutandis, in other periods as well); the problems arise
when one has to define more closely how and in what ways ethnicity or ethnic
(or other forms of) identity was flexible and constructed; which aspects were
malleable and how much and to what; what situations could and did give rise
to what sorts of constructs; and precisely whose identity we are talking about.
In other words: while it is not difficult to agree that identity was ‘constructed’
and not something that was static through history, it is far more difficult to gain
any clarity on the manner of its construction and the factors that conditioned
how it was constructed. It is also extremely hard even to be sure what precisely
the identity was that was being constructed.
A further key component of ethnogenesis theory is (or at least used to be)
that the material presented in the extant narratives, while constructed for
contemporary purposes, also contained a ‘Traditionskern’ (‘kernel of tradition’), which was indeed ancient, and transmitted over the generations some
42
43
how much grief is caused to modern scholars by the contours of the debate. In this context, see already Graus (1975): 240–393.
For discussions of the concept of ethnogenesis in the context of theories regarding ethnic
identities among non-Germanic peoples, see e.g. Berend, Urbańczyk, and Wiszewski
(2014): 61–81; Curta (2001): 18–35; and Curta (2005).
Amory’s work does not present itself as explicitly in support of ethnogenesis theory, and
is in fact quite critical; see Amory (1997): 34–9; 306–7; and for the claim that this book uses
the ethnogenesis “approach”, see Gillett (2009): 408. Nevertheless, it is fair to state that
Amory certainly views identity as a ‘situational construct’, though in fact he sees it as far
more flexible and fluid than most exponents of ethnogenesis theory.
For use by the Author only | © 2016 Koninklijke Brill NV
20
chapter 1
authentic ethnic memory comprising legendary matter about ancestors and
their heroic deeds, and was borne by members of a core group. The ‘Traditionskern’ was not, however, tied to race, and the identity contained in it transferred outwards from members of the core group to others who assimilated
to it, often from completely different ethnic backgrounds. What bound this
group together in a common ethnic identity was thus a common adherence
to a particular narrative of origins.
This theory has been extensively debated; this is not the place to enter into
the discussion in any detail.44 To my mind the principal flaw in ethnogenesis
theory (at least in its more detailed and ambitious iterations) is precisely the
attempt to arrive at broadly valid conclusions about the nature of ethnic identity when the extant source base provides too little evidence for any such conclusions: the existence of narratives of origins does not really provide us with a
guide to how to interpret them or understand their significance among any
contemporary audience; in most cases, we do not even know who the audience
was. This problem in the use of ethnogenesis theory is sometimes compounded
by resorting to rather dubious sorts of evidence from much later to shore up
the case for interpreting the early sources in a particular manner.
Beyond the fact that there is far too little evidence to form the basis of a
theory that aims (at least in some iterations) to be as comprehensive and
cogent as this one, there are two points that are most pertinent to the arguments of this monograph. The first concerns the existence and significance of
the ‘Traditionskern’; the second, the extent to which there was any ‘Germanic’
identity in this period, and the legitimacy of using material from one ‘Germanic’
source to interpret another. It will be argued in the following chapters that
many of the texts that presented ‘national’ histories did indeed draw on oral
sources of some sort, but also that we know too little about these sources and
their provenance to dignify them with the term ‘Traditionskern’ in the sense
meant by the proponents of ethnogenesis theory. It is impossible to know how
ancient or unchanging any of this material was, and it is equally impossible to
discern the extent to which it served as a motor of identity-formation—even
44
For extensive and severe critiques, see Gillett (ed. 2002), along with the response from
Pohl (2002). For less polemical but nevertheless critical appraisals of ethnogenesis theory
and what can be known of early medieval identity, see Halsall (2007): 455–82; and James
(2009): 102–11. Goetz has made some useful, rather non-committal contributions to the
discussion of many of these issues, with regard specifically to the Frankish kingdom:
Goetz (2003a); Goetz (2003b); Goetz (2004). In addition, the essays in Mitchell and
Greatrex (eds 2000), and Mathisen and Shanzer (eds 2011), also provide counterpoints
that are not (or at least not explicitly) of the ‘Toronto/Goffart School’.
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
21
within the contexts of the texts within which these oral-derived narratives are
transmitted, let alone more generally among the supposed ethnic group whose
tradition these narratives are supposed to reproduce.
With regard to ‘Germanic’ identity, even the current proponents of the ethnogenesis theory would agree (I hope) that there was really no such thing in
this period. Although it is true that narratives concerning one Germanicspeaking people are often (much later) preserved by another, I shall argue
below that in fact we only begin to see some evidence of an awareness of specifically linguistic kinship rather late in this period, and even that has little to
do with any sense of a shared cultural or ethnic identity of any sort; furthermore, it is only because of the language that the extant narratives are preserved, not because of any sense that they belonged to a specifically ‘Germanic’,
common inheritance.45
Nevertheless, there is a case to be made—or at least to be answered—for a
‘Germanic’ heroic age and its influence on later material: the extant vernacular
(and much later) texts do, after all, almost invariably concern events that, insofar
as they have a genuinely historical antecedent, took place in precisely the period
known to earlier scholarship as the ‘migration age’ or ‘Völkerwanderungszeit’;
and even where, as in a poem such as Widsith, all we get are names, when it is
Germanic-speakers who are mentioned (and not all the names are of Germanicspeakers!), these are generally persons who, insofar as they have a genuine historical antecedent, also lived in the ‘Völkerwanderungszeit’. These were the years
between c.300 and c.600, when ‘Germanic’ peoples came into extensive, increasing, and often violent contact with Rome, and eventually established their own
post-imperial kingdoms.46 But the later existence in Germanic vernaculars of
narratives about events that took place during these centuries does not necessarily tell us anything at all about this period and the cultural identities of the
45
46
This point was already made with the example of the Burgundians in Ghosh (2007):
247–52.
The concept of the ‘Völkerwanderungszeit’ has itself been severely called into question in
the past decades, and for many scholars completely discredited. It is certainly a fact that
some amount of migration did take place, though the significance of this fact, and whether
it should be allowed to characterise our view of a whole period, is a problem that is still
debated. For a recent spirited reassertion of the importance of migration in the shaping of
late antiquity (and indeed the middle ages that followed altogether) that is rather more
sophisticated than the earlier ‘Völkerwanderungszeit’ theories, and is much nourished by
more recent theories of migration in modern contexts, see Heather (2009); cf. however
Goffart (2006); Halsall (2007); Halsall (2014); and James (2009): 174–92, for other views on
the extent, nature, importance, chronology, causes, and consequences of migration.
For use by the Author only | © 2016 Koninklijke Brill NV
22
chapter 1
people these narratives are about; and thus a great deal of caution is required in
disentangling the threads of what ought to be a healthy collaboration between
philology, literary scholarship, and history. Although one should not make too
much of the similarities of some aspects of ethnogenesis theory (at least in some
versions) to the discipline of Germanic philology, it is certainly the case that its
exponents have sometimes relied on the presence of Germanic names and legendary or heroic material contained in later vernacular sources, just as literary
scholars have plundered the earlier (Latin) narrative material to find the historical background for those later vernacular tales; and like the philologists, some
historians have not been averse to taking the existence of narratives about one
people (the Burgundians, for example) in the vernacular of another (Icelanders,
for example) as indicative of the existence or memory of some sort of ‘Germanic
antiquity’, or at the very least some sort of basic linkage across early Germanic
cultures—sufficient, at any rate, to make it legitimate to use later sources in
a Germanic language to interpret the culture of earlier peoples who spoke a
Germanic language.47
I see no reason to reject the view that the contact with Rome was of epochal
significance for various extra-Roman peoples of the period c.300–c.600, including many who spoke Germanic languages, though it was not equally significant
for all Germanic-speakers, nor exclusively for Germanic-speakers; and I must
stress that many Germanic-speakers who preserved narratives of this period—
thirteenth-century Icelanders, for example—were far less affected by this
encounter with Rome than the subjects of those narratives—fifth-century
Burgundians, for example. Nor do I reject the likelihood that some historical
events of this period were indeed commemorated in oral material, probably
even a formal ‘oral tradition’;48 nor even the probability that the later extant
47
48
There is a danger in this procedure in that philologists and literary historians quite justifiably tend to contextualise their work on the basis of historical scholarship, so that when
some aspects of that historical scholarship are themselves—equally justifiably—based
on the judgements of philologists and literary scholars, a kind of circularity is produced
that can be hard to break out of.
‘Oral tradition’ is a term I shall use rarely in the following pages; when I do use this phrase,
I refer to a formal historical tradition, normally in verse, cultivated by persons in some
manner skilled in reciting and/or composing memorial narratives. By ‘formal’ in this context, I mean that the narratives, while not necessarily stable from a modern point of view,
were—unless about the immediate past—thought to be authentically ancient, with
some value deriving from their antiquity, and therefore were thought to remain stable in
some way, and arguably, even by modern standards, might have retained a stable core.
Generally, I prefer phrases like ‘oral sources’ or ‘oral material’, by which I mean matter that
was not written, but was not necessarily considered ancient, nor necessarily belonged to
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
23
vernacular poetry derives in some manner from this earlier oral tradition.
However, I must stress that we have no knowledge—and no way of increasing
our knowledge—of what exactly this oral tradition might have contained;
how, at any given point, it related to its past, functioned within its present, and
was transmitted further into the future; how and whether the content of this
oral tradition changed over time; and why certain stories survived and others
did not. We also know little of the relationship between any formal oral tradition on the one hand, and both written texts and more informal oral material
of various kinds on the other. Furthermore, by the time we get to the written
vernacular texts, most of which survive only from manuscripts of the twelfth
and thirteenth century (or indeed later), all the places where this material was
preserved had been thoroughly imbued with both Latinity and Christianity;
this is true equally with regard to the few earlier survivals, as we shall see below.
This should not, I hasten to add, lead us to believe that the later vernacular
stories were derived from a reading of, for example, Jordanes; there is no evidence for that whatsoever.49 Nevertheless, this fact is, I believe, more than
enough reason to disallow the use of later sources to understand material from
the earlier period; and therefore ambitious statements regarding the ‘oral culture of the barbarians’ and how it might relate to the nature of the ‘heroic age’,
and the meaning of this ‘heroic age’ in constituting any kind of ‘Germanic’
identity in our period (or later), are best avoided, since they can only be based
on the thinnest and most tenuous of evidence.50
While acknowledging the tremendous industry displayed in the vast body
of work on early medieval ethnic identity, the great erudition of those who
49
50
any sort of prized memorial tradition; such matter could be derived from written sources,
but garbled in the course of person-to-person oral transmission; it could be court gossip,
travellers’ anecdotes, the oral explications of a text by a teacher, and so on. For a useful
study of the interaction between a Latin education and these kinds of informal orality, see
Innes (1998) on Notker I of St Gall. The standard handbook on various kinds of oral tradition is Vansina (1985), who, however, includes “historical gossip” in his definition of oral
tradition; for a broader survey of medieval orality, see most recently Reichl (2012). Kuhn
(1961) argues that historical and heroic narratives could and did exist outside formal traditions, but still operates with excessively schematised notions of lay, legend, and heroic
narrative.
This point is discussed further in Chapter 7 below.
Apart from Chapter 7 below, see also the detailed discussion of these points in Ghosh
(2007); see further the useful and more thorough and theoretical critical musings of
František Graus regarding what can and cannot be known of historical traditions, and the
limits of speculation: Graus (1975): 1–28, et passim. For a more positive view of how much
can be known of the ‘oral culture of the barbarians’, cf. Richter (1994).
For use by the Author only | © 2016 Koninklijke Brill NV
24
chapter 1
have produced it, and indeed even the manifold usefulness of that scholarship
in many respects, my own views tend more towards those of, for example,
Timothy Reuter and Susan Reynolds, namely that we cannot know very much;
and that what we can know suggests that we need to operate with far more
flexible (and cautious) means of analysis than has normally been the case—on
any side of the great divides in the scholarship.51 Particularly given the number
of studies that have been devoted to ethnic and other forms of identity in this
period, it seems to me worth repeating (as Reuter cogently and elegantly suggested) that in fact we know (and can know) very little about what people
actually felt about their own ethnicity and/or identity—astoundingly little,
given the number of quite detailed and erudite claims made about ethnic
identity in this period. It is worth recalling also that the written material until
the sixth century was invariably produced by Romans presenting a perspective
that was also largely Roman; and even the later works are written in Latin by
(Catholic) Christian writers, which means at the very least that whatever is
contained in these works, it is not undiluted barbarian matter.
The extent to which any of this material is representative of a broader sense
of ethnic identity cannot be determined, and it seems to me prudent not to
make over-confident claims in this respect. The main extant kinds of written
sources that might provide some information about identity are law-codes and
narratives; both are highly problematic genres in many respects, and all the
texts available to us are thoroughly infused with Roman-ness in one way or
another. At a broader level, we know almost nothing about most people—even
most elite people—in the post-imperial kingdoms: at most, we have a name, a
profession, and sometimes a religious affiliation and the name of some other
family member. To hang on this minimal information interpretations about
ethnic identity and what it meant takes us into the realm of pure speculation,
backed up by almost no fact.52
51
52
See Reuter (2006b); and Reynolds (1998); both provide judicious and unpartisan discussions of the issues; see in addition James (2009): 102–28, et passim; and Wickham (2012).
The complexities involved in trying to ascertain people’s identities are demonstrated in
the case studies of Amory (1997), and Koch (2012), on Ostrogothic Italy and Visigothic
Spain respectively; see further Amory (1993), and Amory (1994) on the Burgundians. For a
recent discussion of the limits of archaeological evidence with regard to ethnic identity,
see Halsall (2011). The extent and limits of what can be known about a broader population
base beyond members of a royal family are presented in Amory’s prosopography of Goths
and potential Goths in Ostrogothic Italy: Amory (1997): 348–486; note that rather more
information survives for this post-imperial kingdom than most others, and as Amory
demonstrates, it is still woefully insufficient for any sort of conclusiveness in most cases.
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
25
In light of the foregoing remarks, it will not come as a surprise to readers of
the present work that I offer no grand theory regarding ethnic or national identity and historical consciousness as manifest in origin narratives or other narratives about the barbarian past. My effort is rather to interpret the texts on their
own terms, without making broader claims regarding how representative they
might be of any larger, widely-held notions of identity. That there are certain
common elements or approaches across a wide range of texts is suggestive, and
the implications of this point will be discussed further in the concluding chapter of this monograph. Let it be said already, however, that the commonalities
point, in my view, not to a common sense of ‘Germanic’ identity, nor to the
establishment of distinct and discrete ethnic identities based on origin narratives, but rather to similar strategies for dealing with cultural inheritances that
were heterogenous and complex, and the use, perhaps, of origin narratives to
bolster a sense of political cohesion that need not have had very much to do
with most people’s sense of their own identity.
For the moment, the main point to be established is that each text needs to
be read as an artefact that makes certain statements and takes certain positions, but we can have little or no knowledge regarding the broader validity
and acceptance of what we can discern from these texts within the societies in
which they were produced. It seems to me undeniable that these texts were
indeed making an effort at some sort of representation and preservation of a
group identity; but, equally, it is impossible to discern precisely whose group
identity that was, and how representative these texts were of contemporary
self-perceptions of ethnic or other forms of identity.
It has been plausibly suggested that origin narratives and stories about the
distant past can contribute to a notion of identity, in that they can provide a
particular group with a narrative of where it came from, and how it reached its
present geographical, temporal, political, and social location; by sharing in this
narrative as a collective a sense of group cohesion can be fostered.53 Equally, it
has been argued that narratives of the distant past, because they provide the
stories of the origins of present, can serve as a means of conferring legitimacy
to claims to power in that present.54 Certainly it is likely that it is precisely
because of this value of the past that the ‘national’ histories examined below
were important enough to be written down; the fact that we do not find efforts
simply to assimilate all aspects of the Catholic, Latinate, and in many respects
53
54
Coumert (2007): 9; see further Neville (2001), who suggests, drawing on Andersen (2006),
that medieval narratives of origins were efforts to create ‘imagined communities’ by
means of a narrative of shared origins.
Assmann (1992): 70–2; Goetz (2000): 187, et passim; Plassmann (2006): 22–3.
For use by the Author only | © 2016 Koninklijke Brill NV
26
chapter 1
Romanised present to a Roman and Catholic past shows that for the ruling lay
class and at least some of the clergy—those who wrote secular histories—the
barbarian aspects of the past were valuable enough to be preserved, and in a
manner in which the discontinuity between a pagan or heretic distant past
and a Catholic present was glossed over in favour of a narrative that provided
a more or less seamless continuity between past and present.55
However, an inevitable problem with modern theories about the uses of the
past in past societies is that—at least for the period that the present monograph is concerned with—we generally lack the evidence to demonstrate that
the past really did serve whatever function we attribute to it. It is almost certainly the case that in the middle ages as in later periods, the past had a ‘social
function’ as a kind of ‘cultural memory’ with a political and social role within
the society with which that past was identified, and traditions were ‘invented’
in order to shore up precisely this sense of identification. But it is also the case
that from this period we normally have just the one text, with little evidence of
how it related to other texts that might have had to do with identity and the
past, and no evidence whatsoever with regard to how what we read in these
individual texts related to actual practices of negotiating identity in the political and social spheres, even just of the elites, let alone society more broadly.56
We will see below that in fact, it appears that what is presented in many of
these texts might really be ‘invented traditions’; but the function of these
inventions, their antiquity and sources, generally remain opaque to us. Thus, in
the first instance, these texts are here examined solely as individual texts, and
statements about them should not be taken to have any broader significance
regarding the function of the past and the nature of national or ethnic identity.
I return briefly to these issues, and the possibility of making broader claims
regarding early medieval historical consciousness, in the concluding chapter
of the present work.
55
56
As Jan Assmann pointed out, insofar as the narrative of the past is used to constitute a communal identity, this is done by means of stressing both uniqueness—difference from other
groups with other communal identities—and continuity, which is achieved by blanking
out, wherever possible, transformative breaks in the past: Assmann (1992): 39–40.
In addition to Assmann (1992), and Andersen (2006), important theoretical works on
these issues include Fentress and Wickham (1992); Graus (1975); Hobsbawm (1972); and
Hobsbawm (1983). While these problems do not really apply to Hobsbawm’s work since
he deals with periods for which there is more evidence, they are not addressed in a manner in the other works that is quite satisfactory enough, to my mind, to allow the theories
to be applied easily to the texts examined below.
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
27
Historical Writing in Late Antiquity and the Early Middle Ages
Each of the texts examined below belonged within a quite specific historical
and historiographical context, beyond the general context presented in the
previous pages; this will be briefly introduced at the beginning of each of the
following chapters. These texts also need to be set against the background of
the traditions of secular and Christian historical writing in late antiquity, as
this was the historiographical inheritance that influenced, more broadly, the
traditions of historical writing and the understanding of the past in the early
middle ages, within which context our texts must be understood.
In this section of the present chapter, I provide a brief, general overview of
Latin historiography in this period.57 Although the division is somewhat schematic, the main forms of historical writing can be grouped into the broad categories of universal histories and world chronicles, which encompassed secular
and religious matter on a large scale; ecclesiastical histories, which focused on
religious communities; and secular histories.
Universal Histories and World Chronicles
Universal histories and world chronicles aimed to portray the history of all peoples from creation to the present. The most influential in the Latin west were the
chronicle of Eusebius, in its modified form in Jerome’s Latin translation (c.380),
and the Historiae adversus paganos of Orosius (c.417). The Eusebius–Jerome
chronicle comprises short entries under each year, giving brief notices of events
from all over the world. These are often drawn from other sources, and are
selected with the particular theological significance of the structure of world
57
Useful surveys approaching the early medieval traditions of historiography and their late
antique heritage from different perspectives are provided by Allen (2003), Bonamente
(2003), Croke (2012), Ray (1974), Winkelmann (2003), and Woods (2009); a broader context
for the fourth, fifth, and sixth centuries is given by Rohrbacher (2002), and the essays collected in Marasco (ed. 2003). On the novelty of a ‘Christian’ historiography in late antiquity, see further Momigliano (1963). For very useful analyses of late antique Latin Christian
reactions to Roman history and historiographical traditions, see Inglebert (1996), and
Inglebert (2001): 289–554. For a comparative study of pagan and Christian approaches to
historical writing and the ancient history of Rome, see Sehlmeyer (2009). In the following,
I cover the Latin traditions only, since apart from Jordanes, none of the texts discussed
below betrays much awareness of Greek historiography. See, however, the relevant chapters in Marasco (ed. 2003); in addition, for the principal relevant historians, see Blockley
(1981); Cameron (1985); Kaldellis (2004); Paschoud (1975); Paschoud (2006); Rohrbacher
(2002): 64–92; 108–34; and the studies of Inglebert already cited.
For use by the Author only | © 2016 Koninklijke Brill NV
28
chapter 1
history in mind.58 There were a number of continuations in the fifth and sixth
centuries, and the chronicles were well-known, and widely read and copied.59
Isidore of Seville, and later Bede, produced both world chronicles as well as
more specific histories concerned with individual regions. The Fredegar chronicle also starts off as a continuation of Eusebius–Jerome, and only in the latter
parts begins to be independent; even so, it retains something of the character of
a chronicle, though now distinctly less universal. The chronicle tradition, while
not otherwise particularly influential in any direct manner on the texts examined below, was certainly one of the most prominent forms of historical writing
throughout the early middle ages, and was thus one of the fundamental pillars
of the historical consciousness of literate people in the period.60
The universal history as exemplified by the work of Orosius did not present
a synoptic view of the past. Rather, it followed particular great empires of
antiquity from their rise to their fall, and Orosius formulated a theological conception of historical change according to which power is transferred from one
empire to another until reaching its final, highest point in Rome. For Orosius,
Rome was the natural culmination of the progress of political and cultural
development, and the truth of this notion is supported by the fact that Christ
was born in the Roman empire and a Roman citizen.61 Orosius, far more effectively than Eusebius–Jerome, formulated an idea of history in which political
power and religious salvation were inextricably linked, and also developed the
concept of a Christian ‘translatio imperii’ (the transfer of legitimate imperial
rule from one empire to another). This allowed later authors to compose histories of post-imperial kingdoms and empires that nevertheless fulfilled a key
58
59
60
61
Zecchini (2003): 318–19. On Eusebius and his context, see further Burgess and Kulikowski
(2013): 96–126; and Winkelmann (2003): 3–10; 18–31.
On the Eusebius–Jerome chronicle and reactions to it in Latin Christianity up to the
chronicle of Sulpicius (written c.405), see the overview in Burgess and Kulikowski (2013):
119–31; for a more detailed analysis, see Inglebert (1996): 153–295; 357–93. On the chronicle
tradition in the fifth century, see Muhlberger (1990). On sixth-century chronicles, see
Croke (2001); Favrod (ed. 1993): 11–60; Markus (1986).
For the chronicle tradition up to the twelfth century, von den Brincken’s survey remains
indispensable, along with the more recent work of Burgess and Kulikowski, which provides greater detail regarding the antique and Mediterranean origins of the genre: von
den Brincken (1957); Burgess and Kulikowski (2013).
He is called “ciuis Romanus” twice: Orosius: vi.xxii,8; viii.iii,4. Fundamental studies of
Orosius are Goetz (1980a), and Inglebert (1996): 507–89; and in addition the more recent
work of van Nuffelen (2012), who sees Orosius’s historical thought in a less purely theological light than Goetz and Inglebert (but cf. Goetz 2014). Brief surveys are provided in
Rohrbacher (2002): 135–49, and Zecchini (2003): 319–29.
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
29
role in salvation history, following the model of Rome in Orosius’s presentation. Orosius’s history was one of the most widely read historical works
throughout the middle ages, with many imitators, continuators, and writers
who drew profusely on him;62 the authors of all the Latin texts analysed below
seem to have been familiar with at least some part of his work. (This might not
be true of the Waltharius-poet, but given both how widely Orosius was read
and copied, and this poet’s erudition, the above statement probably applies in
this case too.) Yet unlike the narratives examined here, Orosius’s ultimate aim
was to formulate a concept of salvation history in which even the political
aspect of the past was subordinate to its religious significance. In this respect
not least, his universal history is fundamentally different from most of the texts
examined in the present study (Isidore’s Gothic history is something of an
exception to this statement), in which salvation history appears hardly to play
any part at all.
Ecclesiastical Histories
Ecclesiastical histories aimed to depict the history of the Christian Church in
the broadest sense, namely as a community of believers; although, in varying
degrees, political history was also included, it was subordinate to the larger plan
of salvation history, and was often incorporated only insofar as it had directly to
do with matters of religion and the hierarchy of the Church. Although there
were a number of later Greek ecclesiastical histories, it is Eusebius, once again,
who provided the impetus for later writers with his Historia ecclesiastica, which
was widely known in the Latin west in Rufinus’s Latin adaptation.63 Perhaps
the most outstanding example of this genre from the early middle ages is Bede’s
Historia ecclesiastica gentis Anglorum (he; c.731), which, although it has a
regional focus on England and does include political events as well, focuses on
conversion narratives, the institutional history of the Church in England, and
stories of holy men of various stripes.64 Gregory of Tours’s Decem libri historiarum (dlh; c.590–4) are not an ecclesiastical history in the Bedan sense, and
Gregory includes a wide variety of manifestly profane narratives. But his work
too is, at its core, concerned with the community of believers and the fate of
62
63
64
von den Brincken (1957): 80–6; Goetz (1980a): 148–65; Hillgarth (1992); Werner (1987).
Winkelmann (2003): 3–10; 18–31 (on Eusebius); Inglebert (1996): 325–55; Rohrbacher
(2002): 93–107 (Rufinus).
Fundamental studies of Bede: Goffart (1988): 235–328; and Higham (2006); for a historical
(but not theological) commentary, see Wallace-Hadrill (1988); for the broader context, see
the essays in DeGregorio (ed. 2010). See also the further references cited in the first section of Chapter 6.
For use by the Author only | © 2016 Koninklijke Brill NV
30
chapter 1
their faith, rather than with politics; secular history is included more as a means
of illustrating the ills of a worldly life than for its own intrinsic interest.65 Both
Gregory and Bede were widely read during the whole of our period.66
Ecclesiastical histories intended to establish the historical inevitability of
Christianity and prove a particular theological point, which normally hinged
on the eventual conversion to Christianity of all humanity, and pointed the
way forward to the Last Judgement; in this sense, they had much in common
with the theological conception of history in Orosius. They differed in the relative narrowness of the subject matter they covered, not just in terms of its
ecclesiastical focus, but also because these works tended to focus on the history of one nation or region; they had no pretensions to being ‘universal’ histories, and thus could—as in the case of both Bede and Gregory—appear to have
some characteristics of ‘national’ histories.
Secular Histories in Late Antiquity
While ecclesiastical histories and chronicles generally provided a specifically
Christian view of the past (though the extent to and manner in which this
shaped the narrative could vary widely), this period also saw the production of
a number of secular histories. These were perhaps less influential in that they
were less widely read, but some, such as the Breviarum of Eutropius, evidently
did also provide important models for later authors within the period. These
texts were, naturally enough, concerned with the history of Rome, and ranged
from those that covered the whole period from its founding to the historians’
present (the minor Latin historians of the fourth century), through the more
expansive imperial biographies of the Historia Augusta, to the voluminous
Tacitean ‘Zeitgeschichte’ of Ammianus Marcellinus. Also produced in this
period was a short narrative about Roman origins, the Origo gentis Romanae,
which, in its mingling of varied kinds of sources ranging from what seem to be
mythical to more factual accounts, appears to be in some ways comparable to
some of the ‘national’ histories to be discussed below.
There are four brief histories of Rome from the second half of the fourth
century that appear to draw on common sources and on each other, of which
65
66
Fundamental for Gregory: Heinzelmann (1994b); for an important alternative reading, see
Goffart (1988): 112–234; further references are given below in Chapter 3.
On the dissemination of he: Colgrave and Mynors (ed. and trans. 1969): xxxix–lxx; Lapidge
(2006): 119; 138; 142; 145; 149; 154; 166; 234; 238; 240; 243; 248; 255; 268; Crépin (2005): 50–60;
67–8. On the dissemination of dlh: Goffart (1987); Heinzelmann (1994b): 167–75; Lapidge
(2006): 212; 259; 305; McKitterick (2004b): 152; Reimitz (2003); on his use by Fredegar, the
lhf and Paul, see the apparatus in the respective editions, and Chapters 3–4 below.
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
31
Eutropius’s Breviarum ab urbe condita seems to have been the best known for
the period under consideration in this monograph.67 This text is, as the title
indicates, a brief history of the Roman state from Romulus down to 364, and
was written c.369.68 Not long before, Sextus Aurelius Victor—like Eutropius,
a career civil servant—had composed his Liber de Caesaribus, a history of
emperors from 31bc to AD361 (when his work was published); and in c.370,
Festus, again a civil servant, composed his Breviarum, a shorter work covering
the same period as Eutropius’s Breviarum.69 Finally, there is the anonymous
Epitome de Caesaribus from the end of the fourth century.70
A common source for these short histories is believed to be a now lost text
known to modern scholarship as the Kaisergeschichte, which is also thought to
be a source for Jerome’s continuation of Eusebius, and is supposed to have
been composed at some point in the fourth century. None of these histories
have a noticeably prominent religious angle to their narratives; none of them
was written by a Christian. All of these works continued to be read; Eutropius’s
history in particular was used by Jerome and many later Christian historians,
and also formed the basis for Paul the Deacon’s Historia Romana (though Paul
added considerably to it). Aurelius Victor and the Epitome de Caesaribus seem
to have been less well known until the ninth century, but from that point
onwards probably did exercise some influence on later Latin writers. None of
these histories, however, were as extensively and widely received in our period
as either Orosius or the Eusebius–Jerome chronicle. They nevertheless need to
be borne in mind as alternative, resolutely non-religious models of historical
writing available to later historians: while in terms of the brevity which they
cover events they might have something in common with the chronicle tradition, in terms of the approach to history and the choice of material, they are
quite different.
67
68
69
70
On the minor Latin historians of the fourth century, in addition to the survey in Bonamente
(2003), see den Boer (1972); and Rohrbacher (2002): 42–63.
For introductory material on Eutropius, apart from the edition of Santini (ed. 1979); see
Bird (trans. 1993): vii–lvii, along with Bird’s extensive commentary: 71–164; den Boer
(1972): 114–72; Rohrbacher (2002): 49–58.
On Aurelius Victor, in addition to the edition of Pichlmayr and Gruendel (eds 1966), see
Bird (1984); Bird (trans. 1994): vii–xxx, along with Bird’s extensive commentary: 55–207;
den Boer (1972): 19–113; Rohrbacher (2002): 42–8. On Festus, see den Boer (1972): 173–223;
Eadie (ed. 1967): 1–41, along with Eadie’s extensive commentary: 70–153; Rohrbacher
(2002): 59–63.
On this text, in addition to the edition of Pichlmayr and Gruendel (eds 1966), see
Schlumberger (1974).
For use by the Author only | © 2016 Koninklijke Brill NV
32
chapter 1
The other principal secular Latin histories are the Historia Augusta, a compilation of imperial biographies from the later fourth century covering the
emperors from Hadrian to the sons of Carus, and largely anonymous; and the
contemporary history of Ammianus Marcellinus. The Historia Augusta has
long been a thorn in the side of scholars: it is inconsistent in style and content,
as well as in its treatment of different emperors, and it is generally no longer
thought to be a particularly reliable historical source; nor is it normally considered to have any particular literary merit.71 The text’s date cannot be firmly
determined, with estimates ranging from the 360 s to the decade after 395; and
even the fourth-century dating is not necessarily secure. It is a work of pagan
historiography, a “reaction to Christianity triumphant and intolerant” that
showcases the greatness of the pagan and tolerant Roman past as a counterpoint to the intolerant Christian present.72 Although there is little evidence
that it was used by any of the historians studied below, its existence is an example of yet another way of looking at the past that focused on secular achievement and the biographies of rulers, enlivened by what some modern scholars
have suggested is a good dose of fiction.
Ammianus Marcellinus, in contrast, has been seen as an exceptionally factual, sober historian of his own time. He also has his bias, of course, which is
not a Christian one—though precisely what the attitude of his Res gestae to
the new religion is remains a matter of some dispute.73 A Greek-speaker himself, Ammianus chose to write his Res gestae in Latin. The extant text begins in
353 and extends to 390; the first thirteen books have been lost, but Ammianus
says that he presented a history from the principate of Nerva onwards.74
Ammianus was an administrator and staff officer posted in various parts of the
Roman empire; despite his own personal experience, he often appears to have
drawn on written sources for his descriptions rather than presenting an eyewitness account.
Because of the period in which he wrote, he is naturally an especially informative source for contact and conflict with various barbarian groups, and he is particularly useful as providing the principal contemporary narratives concerning
71
72
73
74
On the Historia Augusta, in addition to the edition of Hohl, Samberger, and Seyfarth, see
the introductory survey in Birley (2003); see further especially the fundamental studies of
Barnes (1978); Syme (1968); and Syme (1971).
Birley (2003): 144.
There is a vast body of work on Ammianus, and there appears often to be very little consensus. For introductions, see Sabbah (2003); and Rohrbacher (2002): 14–41; see further
the important studies of Barnes (1998); Blockley (1975); Demandt (1965); Kelly (2008);
Matthews (1989); and Syme (1968).
Ammianus: xxxi.xvi, 9.
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
33
the Goths at a time when they were having an ever-greater impact on the
empire, as well as being the earliest real source for Gothic history that is likely
to be in any manner reliable; Ammianus can indeed be used as a check against
some aspects of Jordanes’s narrative of the same period.75 Like the histories of
Tacitus, Ammianus’s history is formulated in a rather dense Latin, and like
Tacitus, he has much to say about Roman contacts and conflicts with barbarians. Ammianus has therefore, like Tacitus, been much utilised by modern
scholars interested in barbarians. Like Tacitus once again, it is hard to know
both just how accurate he is about the barbarians, and just how much influence his work had on historians in the following four centuries: there is little
evidence of direct use. As his Res gestae are ‘Zeitgeschichte’ rather than an origin narrative, they are of little immediate relevance to our understanding of
how the distant past was treated by the later ‘national’ histories.
In contrast, the Origo gentis Romanae was explicitly about the distant past.
It is a rather unusual text that exists only from fifteenth-century manuscripts,
but was apparently composed in the late fourth century and provides a narrative of Roman origins based both on Virgil and on other sources, not all of
which have as yet been properly identified.76 Unlike some of the other works
just discussed, this text has attracted relatively little attention in the scholarship, and seems not to have excited the imagination of any of the many scholars dealing with later narratives of barbarian (as opposed to Roman)
origins.77
Presenting as it does a rather heterogenous use of sources and a combination of what seem to be both more factual and more fictional elements, it
might be a useful comparandum as a specifically Roman prose narrative of
origins that, even if no direct influence can be shown, might nevertheless be
well worth studying alongside the many stories of barbarian origins that are
extant. In particular, it is a demonstration that not just early medieval barbarians, but also non-Christian Romans in late antiquity were interested in
a mythologising prose narrative of origins that synthesised a number of different and not always harmonious versions of the distant past, and thereby
75
76
77
References to some of the relevant passages are provided in the next chapter; for Ammianus’s
value regarding the Alamanni, another contemporary barbarian group frequently in conflict
with Rome, see Zotz (1998).
On this text, in addition to the edition of Pichlmayr and Gruendel (eds 1966), see most
recently the extensive material in Sehlmeyer (ed. and trans. 2004): 7–27; 65–160; and the
discussion in the context of its place among other contemporary histories in Sehlmeyer
(2009): passim; see further Momigliano (1958); and Christopher Smith (2005).
See, however, the brief comments of Pohl (2014): 410–11; 413; this paper appeared after the
paragraphs above had already been written.
For use by the Author only | © 2016 Koninklijke Brill NV
34
chapter 1
encompassed many elements that do not seem to ‘fit’ with the rest of the narrative, but seem nevertheless clearly to derive from written rather than oral
sources. It is also a demonstration that the sort of eclectic manner of collecting
material relating to ‘ethnic’ or ‘national’ origins that we find in, inter alia,
Jordanes’s Gothic history, was by no means invented by or unique to the
authors of the later ‘national’ histories of what had earlier been barbarian
peoples.78
By the fifth century, therefore, there was a wide range of more or less wellknown narratives, providing a number of different models for how history
could and should be written. With the exception of the Breviarium of Eutropius
(and to a lesser extent that of Festus), most of the historical narratives extant
from late antiquity that were widely received and served as models in the early
middle ages followed the paradigms either of ecclesiastical history, or universal history or world chronicle; the most well-known models for historical writing were thus dominated by the perspective of salvation history, and this
perspective is accordingly shared by the majority of early medieval histories.
Although they do not exclude religion, the authors even of the Latin histories to be considered below—Jordanes, Isidore, Fredegar, the lhf-author and
Paul—seem to be more concerned simply with providing a secular narrative
from the distant past to the present or very recent past, without really considering the way any of this fits into the perspective of salvation history.79 They
are thus perhaps more akin to the Historia Augusta in its form as serial biography and the Origo gentis Romanae in its quest for (what appear to us to be
mythical) origins cobbled together apparently indiscriminately from a variety
of sources. Somewhat frustratingly, however, there seems to be little evidence
that either of these texts were indeed read by the authors of the works to be
examined below; Eutropius’s Breviarum was more widely known, but in terms
of style, form, and historical ideology appears not to have had as much influence as the text’s dissemination might suggest, except insofar as it tends to
avoid religious issues.
The differences between the histories to be examined below and those just
discussed notwithstanding, there are a few points to be made regarding the
inheritance that the latter group of texts represent. To begin with, let me reiterate that we should always be aware that there was a broad palette of approaches
78
79
I am grateful to Sandy Murray for alerting me to the existence and possible significance of
this text.
Although Isidore does appear to present the Goths within the framework of salvation
history, this aspect is not prominent in his historical works, unlike in his other writings.
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
35
to the past among later Roman historians writing in Latin, and it should not
surprise us therefore that early medieval approaches to the past could be
equally varied. While it is true that the challenges faced by early medieval historians were different at least in one respect—late Roman authors did not have
to try and integrate a barbarian past (with little written material about it at
least in its furthest reaches) with a Roman inheritance, and could rely on
extensive written sources for the early history of Rome—there was also one
basic challenge that they shared with the Christian Roman historians of the
fourth and fifth centuries (and indeed earlier), namely the need to integrate a
pre-Christian past, and indeed a pre-Christian cultural heritage inherent in the
very language in which the texts were written, with a Christian present.80
This was a task approached in different ways, but one aspect common to
most Christian histories was an effort to stress the providential place within
Christian history of the people or polity that was the focus of the history being
written. Such a focus on salvation history is not to be found in most of our
‘national’ histories; nevertheless, one lesson that the authors of the latter texts
might have learnt was the importance of papering over the cracks that could
appear from too close a scrutiny of the differences between the pagan past and
the Christian present. One of the methods of harmonising past and present was
euhemerism: although originating in pre-Christian thought as a means of interpreting pagan myths, euhemerism—an interpretation of gods as exceptional
humans who were worshipped as deities—was adopted by some Christians in
late antiquity as a method of explaining the worship of pagan gods, and continued to be employed by some Christian writers throughout the middle ages.
From the beginning of its use by Christian writers, euhemerism could be
applied with a simply historicising and thereby often apologetic function, without moral critique; it could, however, equally well be used as a means of explicitly denigrating pagan practices and the pagan past.81
The salvific aspect of history, a focus on which was one of the principal
means by which the authors of ecclesiastical histories were able to pass over
potential breaks between past and present, was not prominent in the later
80
81
Admittedly, for Christian Roman historians, the problem was arguably less a matter of
legitimising Roman rule, and more a question of not wanting to abandon Roman culture.
For basic introductions, see Cooke (1927); Orchard (1997): 101–4; von See (1989); and
Thraede (1966). See more broadly Inglebert (1996), for studies of how individual Christian
Roman historians approached the problems posed by the pre-Christian Roman past; the
earliest Christian writers to present a euhemeristic reading of that past were Minucius
Felix (late second or early third century) and Lactantius (c.250–c.325) (Inglebert 1996:
105–44). On the relevance of this point for Jordanes, for example, see Chapter 2 at n. 90.
For use by the Author only | © 2016 Koninklijke Brill NV
36
chapter 1
‘national’ histories. Yet it is possible, indeed likely, that the authors of the latter
learnt just as much from their more religiously-oriented predecessors as from
the secular or pagan historians of late antiquity—and in fact probably more.
Arguably, their task was in some respects to produce a synthesis of both
approaches: to provide a secular history of a people that had formed a polity,
and to do so from the origins of that people to the present in a manner that
portrayed the present in a flattering light, and therefore could not present too
negative an image of the past; but also, precisely because of this last requirement, to provide a history that would not offend the current religious sensibilities because of the difference in religion in the past. Unlike late antique
Christian historians, who had to find a way of coming to terms only with the
non-Christian aspect of their Roman cultural inheritance, but not with the language or medium per se in which the rest of that inheritance was transmitted,
the historians of the post-imperial, barbarian kingdoms had to harmonise
their current, Romanised, Christian, Latin present with a barbarian past that
had not already been through a process of assimilation in the Latin written
word to the Roman inheritance.
In the following chapters I shall argue that the ‘national’ histories I examine
were written as a means of recording and codifying a narrative of the distant
past that established a continuity between that distant past and the present.
This sense of continuity was arguably important to the secular aristocracy,
since, despite the fact that the distant past was pagan or heretic, continuity
with ancestral greatness was one of the principal means of legitimising rule in
the present. Because of this lay interest, therefore, narratives about this distant
past were sufficiently important that they needed to be recorded in the now
authoritative medium of Latin writing.
But the impulse for writing these texts did not simply emanate from the
secular aristocracy: the existence of these Latin ‘national’ histories also shows
how closely related were the spheres of Church and court, and furthermore, it
demonstrates that it was acceptable to present aspects of the distant past not
related to religion in a positive light, as long as ways could be found to diminish
the presence of elements of that past that could have been interpreted as
opposed Christianity. In the person of Isidore especially, but also to differing
extents in the other cases presented below, it could be argued that the interests
of Church and state were so closely related that a stress on continuities rather
than religious difference was probably as important to at least some members
of the ecclesiastical hierarchy as to the secular aristocracy; and in any case the
former were invariably linked to the latter by ties of blood.
Because writing, with the partial exception of administrative literacy, was
dominated by the Church in our period, the bias of most extant written sources
For use by the Author only | © 2016 Koninklijke Brill NV
Introduction
37
leans heavily towards clerical attitudes, which were conditioned to a much
greater degree by concerns of religion and religious morality. With the exception
of the texts examined in this book, it is only from around the turn of the millennium, and in some parts of western Europe from considerably later, that we
begin to find in any significant quantity narrative literature that gives expression
primarily to a lay culture and historical consciousness: for most of the early middle ages, there was scarcely any written cultural production that was essentially
of a secular nature. (I exclude here administrative and documentary material.)
Nevertheless, two final points must be noted here, which will recur throughout the following pages. The first is that there was no great gulf between the
ecclesiastical and lay aristocracy: from the very beginnings of the medieval
period, senior churchmen were closely linked with secular affairs, from urban
administration to military service; the clergy were often expected to commemorate members of the secular aristocracy; and it was common for at least one if
not more members of lay aristocratic families to join the Church at an early
age: thus the ecclesiastical and lay hierarchies were literally linked by blood.82
The second important point is that although the bulk of the written evidence that survives betrays a marked ecclesiastical bias, we cannot assume
that no secular education or culture existed: there was a large lay aristocracy,
which was certainly closely connected to and influenced by the Church, but
would have had some kind of lay education (in, among other things, the arts of
war and politics), and secular values that we cannot assume would have been
identical to those of the Church. This is a further significant difference in context between the secular Latin histories reviewed in the previous pages, and
the works to be examined below: late antique secular Latin histories were written for and embedded within a lay culture just as much as early medieval secular histories were; but it was (at least with regard to the elites) a highly literate
secular culture, within which the written and Latin cultural heritage was
82
The literature on secular aspects of the early medieval Church is vast; see, in general,
Wickham (2009): 184–90. The following is a very small sample of important studies on
more specific issues: on episcopal administration and relationship with secular political
hierarchies, see Heinzelmann (1988); and Kaiser (1988); on monastic commemoration of
secular aristocrats, see Freise (1985); McKitterick (2004b): 162–73; Oexle (1976); Schmid
(1965); and Schmid (1979); on military service and the Church, see Prinz (1971); on family
relationships between the secular and clerical elites, see Schmid (1965); and Wormald
(1978): 49–58; on royal patronage of churches and written, Latin cultural production, see
Hen (2007); on these topics and the royal promotion of ecclesiastical reform and particular forms of piety and religious representation, see McKitterick (2008): 292–380.
For use by the Author only | © 2016 Koninklijke Brill NV
38
chapter 1
fundamental.83 In contrast, early medieval secular culture (even of the elites)
was, to a large extent, not set down in writing; we can know little about it, relative to what we know about religious culture in the same period.
The subjects of the present study are all texts: written material, produced (at
least in their its form) in all cases by persons with some form of religious education in Latin. These works are among the relatively few forms of literary evidence both of an independent lay aristocratic culture, and of the extent to
which this was influenced by and in turn influenced the culture of the ecclesiastical elite. They constitute, therefore, the principal extant narrative evidence
of a secular historical consciousness; they are in some cases a window into a
secular culture that lived primarily in an oral context and in the vernacular
rather than in Latin writing; and they are also, in all cases, witness to the interaction between the secular and ecclesiastical milieux. They are not, however,
anything approaching the totality of expressions of that secular culture, even
just with regard to historical consciousness and the distant past; much more
must have existed that we can never know about, and our conclusions must
therefore be appropriately cautious.
83
On this point, see e.g. Jones (1964): 988–91; 997–1006; 1007–12; 1021–4; and Wickham
(2009): 29–31.
For use by the Author only | © 2016 Koninklijke Brill NV